https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84487
--- Comment #19 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
Unfortunately, this patch
Index: class.c
===================================================================
--- class.c (Revision 269895)
+++ class.c (Arbeitskopie)
@@ -911,6 +911,9 @@ finalize_component (gfc_expr *expr, gfc_symbol *de
if (!comp_is_finalizable (comp))
return;
+ if (comp->finalized)
+ return;
+
e = gfc_copy_expr (expr);
if (!e->ref)
e->ref = ref = gfc_get_ref ();
@@ -1038,6 +1041,7 @@ finalize_component (gfc_expr *expr, gfc_symbol *de
sub_ns);
gfc_free_expr (e);
}
+ comp->finalized = true;
}
@@ -2290,9 +2294,10 @@ gfc_find_derived_vtab (gfc_symbol *derived)
gfc_find_symbol (name, ns, 0, &vtype);
if (vtype == NULL)
{
- gfc_component *c;
+ gfc_component *c, *csave;
gfc_symbol *parent = NULL, *parent_vtab = NULL;
bool rdt = false;
+ bool has_init = false;
/* Is this a derived type with recursive allocatable
components? */
@@ -2299,6 +2304,8 @@ gfc_find_derived_vtab (gfc_symbol *derived)
c = (derived->attr.unlimited_polymorphic
|| derived->attr.abstract) ?
NULL : derived->components;
+ csave = c;
+
for (; c; c= c->next)
if (c->ts.type == BT_DERIVED
&& c->ts.u.derived == derived)
@@ -2307,6 +2314,15 @@ gfc_find_derived_vtab (gfc_symbol *derived)
break;
}
+ for (c = csave; c; c = c->next)
+ {
+ if (c->initializer)
+ {
+ has_init = true;
+ break;
+ }
+ }
+
gfc_get_symbol (name, ns, &vtype);
if (!gfc_add_flavor (&vtype->attr, FL_DERIVED, NULL,
&gfc_current_locus))
@@ -2383,7 +2399,7 @@ gfc_find_derived_vtab (gfc_symbol *derived)
c->ts.type = BT_DERIVED;
c->ts.u.derived = derived;
if (derived->attr.unlimited_polymorphic
- || derived->attr.abstract)
+ || derived->attr.abstract || !has_init)
c->initializer = gfc_get_null_expr (NULL);
else
{
Index: gfortran.h
===================================================================
--- gfortran.h (Revision 269895)
+++ gfortran.h (Arbeitskopie)
@@ -1094,6 +1094,7 @@ typedef struct gfc_component
struct gfc_typebound_proc *tb;
/* When allocatable/pointer and in a coarray the associated token. */
tree caf_token;
+ bool finalized;
}
gfc_component;
leads to testsuite failures (allocate_with_source_5.f90,
allocate_with_source_6.f90 and possibly others).
So, it seems as if the BSS idea might have some merit. Ideas on how
to implement this, anybody?
