===================================================================
*************** gfc_trans_structure_assign (tree dest, g
gfc_se se;
gfc_start_block (&block);
- cm = expr->ts.u.derived->components;
if (expr->ts.u.derived->from_intmod == INTMOD_ISO_C_BINDING
&& (expr->ts.u.derived->intmod_sym_id == ISOCBINDING_PTR
*************** gfc_trans_structure_assign (tree dest, g
return gfc_finish_block (&block);
}
+ /* Make sure that the derived type has been completely built. */
+ if (!expr->ts.u.derived->backend_decl
+ || !TYPE_FIELDS (expr->ts.u.derived->backend_decl))
+ {
+ tmp = gfc_typenode_for_spec (&expr->ts);
+ gcc_assert (tmp);
+ }
+
+ cm = expr->ts.u.derived->components;
+
+
if (coarray)
gfc_init_se (&se, NULL);
*************** gfc_trans_structure_assign (tree dest, g
gfc_add_expr_to_block (&block, tmp);
}
field = cm->backend_decl;
+ gcc_assert(field);
tmp = fold_build3_loc (input_location, COMPONENT_REF, TREE_TYPE (field),
dest, field, NULL_TREE);
if (!c->expr)
===================================================================
***************
+ ! { dg-do compile }
+ !
+ ! Test the fix for PR69654 in which the derived type 'ty_foo2' was
+ ! not completely built in time for initialization thereby causing an ICE.
+ !
+ ! Contributed by Hossein Talebi <talebi.hossein@gmail.com>
+ !
+ Module foo_pointers_class
+ implicit none
+ type :: ty_foo_pointers
+ integer :: scale=0
+ integer,pointer :: universe_ulogfile => NULL()
+ class(*),pointer :: foo => NULL()
+ end type ty_foo_pointers
+
+ type :: ty_part_ptrs
+ character(len=80),pointer :: part_name => NULL()
+ class(*),pointer :: part_fem => NULL()
+ end type
+
+ type :: ty_class_basis
+ integer :: id=0
+ end type ty_class_basis
+
+ type :: ty_store_sclass
+ class(ty_class_basis),allocatable :: OBJ
+ end type ty_store_sclass
+ End Module foo_pointers_class
+
+ Module foo_class
+ use foo_pointers_class
+ implicit none
+ type,extends(ty_class_basis) :: ty_foo2
+ character(200) :: title
+ logical :: isInit=.false.
+ type(ty_foo_pointers) :: foo
+ end type ty_foo2
+ ENd Module foo_class
+
+
+ Module foo_scripts_mod
+ implicit none
+ contains
+
+ subroutine foo_script1
+ use foo_class, only: ty_foo2
+ implicit none
+ type(ty_foo2) :: foo2
+ integer i
+
+ Call foo_init2(foo2)
+ end subroutine foo_script1
+
+ subroutine foo_init2(self)
+ use foo_class, only: ty_foo2
+ type(ty_foo2),target :: self
+ self%isInit=.true.
+ end subroutine foo_init2
+
+ End Module foo_scripts_mod
As I remarked in PR, this fix probably comes 1,379 days too late. I am not at all sure that I understand why I couldn't see the problem because it is rather trivial. I am open to not adding the second gcc_assert - it does seem like overkill. Regtested on FC30/x86_64 - OK for trunk and ....? Paul 2019-11-14 Paul Thomas <pault@gcc.gnu.org> PR fortran/69654 * trans-expr.c (gfc_trans_structure_assign): Move assignment to 'cm' after treatment of C pointer types and test that the type has been completely built before it. Add an assert that the backend_decl for each component exists. 2019-11-14 Paul Thomas <pault@gcc.gnu.org> PR fortran/69654 * gfortran.dg/derived_init_6.f90: New test.