more (catastrophic) internal compiler errors when initializing derived types

[joseph_battelle]

Couple more ICE's related to structure constructors. This time the base type is not empty but the empty structure constructor relies on default init of both base and derived type. Also on a typo I came up with a separate "catastrophic" error that should be a syntax error instead. See vd2.

Compiling with Intel Visual Fortran 11.1.065 [IA-32]...

[fortran]module M6
    type :: b
        integer :: i=0
    end type b
  
    type, extends(b) :: d
        integer :: n=0
    end type d

contains
    
    subroutine foo ()
    type(d) :: vd1 = d() !fortcom: Fatal: There has been an internal compiler error (C0000005)
    type(d) :: vd2 = d(b=5) !catastrophic error: **Internal compiler error: internal abort**
!---------------------^
    type(d) :: vd3 = d(5) !OK
    type(d) :: vd4 = d(b=b(5), 5) !OK
    type(d) :: vd5 = d(b()) !OK
    end subroutine foo    
        
end module M6[/fortran]

[Steven_L_Intel1 · Accepted solution]

Joseph,

I'm sorry to see you encountering so many issues - we do appreciate your bringing them to our attention so that they can be fixed. I have escalated this one as issue DPD200157328.

[joseph_battelle]

Thanks. I don't mind at all. I'm excited to see you get closer to full F2003 support in your next major release. As long as you guys keep inching towards eventual support of the full standard I'll continue to report issues I come across.

-joe

[Steven_L_Intel1]

I expect this issue to be fixed in our November release.