[Bug fortran/66366] ICE on invalid with non-allocatable CLASS variable [OOP]

[abensonca at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66366

--- Comment #3 from Andrew Benson abensonca at gmail dot com ---
Note the name conflict between:

  type :: spsf

and

type(h5) :: spsf

in the previous comment. Changing the name of either to make them distinct
removes the ICE.

[Bug fortran/66366] ICE on invalid with non-allocatable CLASS variable [OOP]

[abensonca at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66366

--- Comment #2 from Andrew Benson abensonca at gmail dot com ---
This reduced test case produces an ICE with similar backtrace, although it
doesn't share the problem of having a non-allocatable/pointer CLASS variable
(as far as I can tell):

module sps
  type :: spsf
  end type spsf
  type :: h5
   contains
 procedure :: c = hC
  end type h5
contains
  subroutine frf()
type(h5) :: spsf
call spsf%c()
  end subroutine frf
  subroutine hC(s)
class(h5), intent(inout) :: s
  end subroutine hC
end module sps

[Bug fortran/66366] ICE on invalid with non-allocatable CLASS variable [OOP]

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66366

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

   What|Removed |Added

   Keywords||error-recovery,
   ||ice-on-invalid-code
 Status|UNCONFIRMED |NEW
   Last reconfirmed||2015-06-03
 Ever confirmed|0   |1

--- Comment #1 from Dominique d'Humieres dominiq at lps dot ens.fr ---
With gfortran 4.5.4 I get

pr66366.f90:9.27:

 class(t1d) :: transfer
   1
Error: Component 'transfer' with CLASS at (1) must be allocatable or pointer

twice, from 4.6 to 4.8, I get a segmentation fault, and from 4.9 to trunk 6.0,
I get the ICE after the error.