[Bug fortran/65397] [Coarrays][OOP] ICE in resolve_ordinary_assign

[Bug fortran/65397] [Coarrays][OOP] ICE in resolve_ordinary_assign

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65397

--- Comment #1 from Tobias Burnus <burnus at gcc dot gnu.org> ---
See also gfortran.dg/coarray_38.f90 of
https://gcc.gnu.org/ml/fortran/2015-03/msg00057.html (assuming it will
eventually appear on the server)

[Bug fortran/65397] New: [Coarrays][OOP] ICE in resolve_ordinary_assign

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65397

            Bug ID: 65397
           Summary: [Coarrays][OOP] ICE in resolve_ordinary_assign
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: burnus at gcc dot gnu.org

The following valid program ICEs with -fcoarray=lib in resolve_ordinary_assign.

subroutine five
implicit none
type t
  integer, pointer :: a
  integer :: b
end type t
type t2
  class(t), allocatable :: caf2[:]
end type t2
class(t), save, allocatable :: caf[:]
type(t) :: x
type(t2) :: y

x = caf[4]     ! <<<< ICE
x = y%caf2[5]  ! <<<< ICE
end subroutine five

[Bug fortran/65397] [Coarrays][OOP] ICE in resolve_ordinary_assign

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65397

--- Comment #2 from Tobias Burnus <burnus at gcc dot gnu.org> ---
The problem is in resolve_ordinary_assign (resolve.c):

9428      /* Assign the 'data' of a class object to a derived type.  */
9429      if (lhs->ts.type == BT_DERIVED
9430          && rhs->ts.type == BT_CLASS)
9431        gfc_add_data_component (rhs);

But the RHS is:

(gdb) p rhs->expr_type
$2 = EXPR_FUNCTION
(gdb) p rhs->symtree->name
$3 = 0x7ffff1f45f90 "_F.caf_get"

And adding "._data" fails in class.c as:

211       gfc_symbol *derived = e->symtree->n.sym->ts.u.derived;
(gdb) p e->symtree->n.sym->ts.u.derived
$4 = (gfc_symbol *) 0x0

[Bug fortran/65397] [Coarrays][OOP] ICE in resolve_ordinary_assign

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65397

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2015-03-13
     Ever confirmed|0                           |1

--- Comment #3 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
Confirmed on trunk (5.0 r221188) with -fcoarray=lib. The test compiles with
-fcoarray=single and with 4.9.2 with -fcoarray=lib: is it a regression?