gdb/gdb-6.8-bz377541-vla.patch

diff --git a/gdb/ada-lang.c b/gdb/ada-lang.c
index a6d5757..5c44b52 100644
--- a/gdb/ada-lang.c
+++ b/gdb/ada-lang.c
@@ -1666,8 +1666,8 @@ ada_type_of_array (struct value *arr, int bounds)
         return NULL;
       while (arity > 0)
         {
-          struct type *range_type = alloc_type (objf);
-          struct type *array_type = alloc_type (objf);
+          struct type *range_type = alloc_type (objf, NULL);
+          struct type *array_type = alloc_type (objf, NULL);
           struct value *low = desc_one_bound (descriptor, arity, 0);
           struct value *high = desc_one_bound (descriptor, arity, 1);
           arity -= 1;
@@ -1774,9 +1774,9 @@ packed_array_type (struct type *type, long *elt_bits)
   if (TYPE_CODE (type) != TYPE_CODE_ARRAY)
     return type;
 
-  new_type = alloc_type (TYPE_OBJFILE (type));
   new_elt_type = packed_array_type (ada_check_typedef (TYPE_TARGET_TYPE (type)),
                                     elt_bits);
+  new_type = alloc_type (TYPE_OBJFILE (type), new_elt_type);
   create_array_type (new_type, new_elt_type, TYPE_INDEX_TYPE (type));
   TYPE_FIELD_BITSIZE (new_type, 0) = *elt_bits;
   TYPE_NAME (new_type) = ada_type_name (type);
@@ -6790,7 +6790,7 @@ variant_field_index (struct type *type)
 static struct type *
 empty_record (struct objfile *objfile)
 {
-  struct type *type = alloc_type (objfile);
+  struct type *type = alloc_type (objfile, NULL);
   TYPE_CODE (type) = TYPE_CODE_STRUCT;
   TYPE_NFIELDS (type) = 0;
   TYPE_FIELDS (type) = NULL;
@@ -6847,7 +6847,7 @@ ada_template_to_fixed_record_type_1 (struct type *type,
         nfields++;
     }
 
-  rtype = alloc_type (TYPE_OBJFILE (type));
+  rtype = alloc_type (TYPE_OBJFILE (type), NULL);
   TYPE_CODE (rtype) = TYPE_CODE_STRUCT;
   INIT_CPLUS_SPECIFIC (rtype);
   TYPE_NFIELDS (rtype) = nfields;
@@ -7034,7 +7034,8 @@ template_to_static_fixed_type (struct type *type0)
         new_type = static_unwrap_type (field_type);
       if (type == type0 && new_type != field_type)
         {
-          TYPE_TARGET_TYPE (type0) = type = alloc_type (TYPE_OBJFILE (type0));
+          TYPE_TARGET_TYPE (type0) = type = alloc_type (TYPE_OBJFILE (type0),
+							NULL);
           TYPE_CODE (type) = TYPE_CODE (type0);
           INIT_CPLUS_SPECIFIC (type);
           TYPE_NFIELDS (type) = nfields;
@@ -7079,7 +7080,7 @@ to_record_with_fixed_variant_part (struct type *type, const gdb_byte *valaddr,
   else
     dval = dval0;
 
-  rtype = alloc_type (TYPE_OBJFILE (type));
+  rtype = alloc_type (TYPE_OBJFILE (type), NULL);
   TYPE_CODE (rtype) = TYPE_CODE_STRUCT;
   INIT_CPLUS_SPECIFIC (rtype);
   TYPE_NFIELDS (rtype) = nfields;
@@ -7251,7 +7252,7 @@ to_fixed_array_type (struct type *type0, struct value *dval,
       if (elt_type0 == elt_type)
         result = type0;
       else
-        result = create_array_type (alloc_type (TYPE_OBJFILE (type0)),
+        result = create_array_type (alloc_type (TYPE_OBJFILE (type0), NULL),
                                     elt_type, TYPE_INDEX_TYPE (type0));
     }
   else
@@ -7281,7 +7282,7 @@ to_fixed_array_type (struct type *type0, struct value *dval,
           struct type *range_type =
             to_fixed_range_type (TYPE_FIELD_NAME (index_type_desc, i),
                                  dval, TYPE_OBJFILE (type0));
-          result = create_array_type (alloc_type (TYPE_OBJFILE (type0)),
+          result = create_array_type (alloc_type (TYPE_OBJFILE (type0), NULL),
                                       result, range_type);
         }
       if (!ignore_too_big && TYPE_LENGTH (result) > varsize_limit)
@@ -9546,7 +9547,7 @@ to_fixed_range_type (char *name, struct value *dval, struct objfile *objfile)
       if (L < INT_MIN || U > INT_MAX)
 	return raw_type;
       else
-	return create_range_type (alloc_type (objfile), raw_type, 
+	return create_range_type (alloc_type (objfile, NULL), raw_type, 
 				  discrete_type_low_bound (raw_type),
 				  discrete_type_high_bound (raw_type));
     }
@@ -9611,7 +9612,7 @@ to_fixed_range_type (char *name, struct value *dval, struct objfile *objfile)
 
       if (objfile == NULL)
         objfile = TYPE_OBJFILE (base_type);
-      type = create_range_type (alloc_type (objfile), base_type, L, U);
+      type = create_range_type (alloc_type (objfile, NULL), base_type, L, U);
       TYPE_NAME (type) = name;
       return type;
     }
diff --git a/gdb/c-typeprint.c b/gdb/c-typeprint.c
index 0929516..2aaf9ad 100644
--- a/gdb/c-typeprint.c
+++ b/gdb/c-typeprint.c
@@ -559,7 +559,12 @@ c_type_print_varspec_suffix (struct type *type, struct ui_file *stream,
 	fprintf_filtered (stream, ")");
 
       fprintf_filtered (stream, "[");
-      if (TYPE_LENGTH (TYPE_TARGET_TYPE (type)) > 0
+      if (TYPE_ARRAY_BOUND_IS_DWARF_BLOCK (type, 1))
+	{
+	  /* No _() - printed sources should not be locale dependent.  */
+	  fprintf_filtered (stream, "variable");
+	}
+      else if (TYPE_LENGTH (TYPE_TARGET_TYPE (type)) > 0
 	&& !TYPE_ARRAY_UPPER_BOUND_IS_UNDEFINED (type))
 	fprintf_filtered (stream, "%d",
 			  (TYPE_LENGTH (type)
diff --git a/gdb/coffread.c b/gdb/coffread.c
index 6059d68..6a24c2c 100644
--- a/gdb/coffread.c
+++ b/gdb/coffread.c
@@ -346,7 +346,7 @@ coff_alloc_type (int index)
      We will fill it in later if we find out how.  */
   if (type == NULL)
     {
-      type = alloc_type (current_objfile);
+      type = alloc_type (current_objfile, NULL);
       *type_addr = type;
     }
   return type;
diff --git a/gdb/dwarf2expr.c b/gdb/dwarf2expr.c
index 75a4ec7..aa8ab33 100644
--- a/gdb/dwarf2expr.c
+++ b/gdb/dwarf2expr.c
@@ -752,6 +752,13 @@ execute_stack_op (struct dwarf_expr_context *ctx,
 	  ctx->initialized = 0;
 	  goto no_push;
 
+	case DW_OP_push_object_address:
+	  if (ctx->get_object_address == NULL)
+	    error (_("DWARF-2 expression error: DW_OP_push_object_address must "
+	           "have a value to push."));
+	  result = (ctx->get_object_address) (ctx->baton);
+	  break;
+
 	default:
 	  error (_("Unhandled dwarf expression opcode 0x%x"), op);
 	}
diff --git a/gdb/dwarf2expr.h b/gdb/dwarf2expr.h
index 7047922..a287b6f 100644
--- a/gdb/dwarf2expr.h
+++ b/gdb/dwarf2expr.h
@@ -67,10 +67,10 @@ struct dwarf_expr_context
      The result must be live until the current expression evaluation
      is complete.  */
   unsigned char *(*get_subr) (void *baton, off_t offset, size_t *length);
+#endif
 
   /* Return the `object address' for DW_OP_push_object_address.  */
   CORE_ADDR (*get_object_address) (void *baton);
-#endif
 
   /* The current depth of dwarf expression recursion, via DW_OP_call*,
      DW_OP_fbreg, DW_OP_push_object_address, etc., and the maximum
diff --git a/gdb/dwarf2loc.c b/gdb/dwarf2loc.c
index cad3db8..65751a4 100644
--- a/gdb/dwarf2loc.c
+++ b/gdb/dwarf2loc.c
@@ -107,6 +107,9 @@ struct dwarf_expr_baton
 {
   struct frame_info *frame;
   struct objfile *objfile;
+  /* From DW_TAG_variable's DW_AT_location (not DW_TAG_type's
+     DW_AT_data_location) for DW_OP_push_object_address.  */
+  CORE_ADDR object_address;
 };
 
 /* Helper functions for dwarf2_evaluate_loc_desc.  */
@@ -163,22 +166,32 @@ dwarf_expr_frame_base (void *baton, gdb_byte **start, size_t * length)
       *start = find_location_expression (symbaton, length,
 					 get_frame_address_in_block (frame));
     }
-  else
+  else if (SYMBOL_OPS (framefunc) == &dwarf2_locexpr_funcs)
     {
       struct dwarf2_locexpr_baton *symbaton;
+
       symbaton = SYMBOL_LOCATION_BATON (framefunc);
-      if (symbaton != NULL)
-	{
-	  *length = symbaton->size;
-	  *start = symbaton->data;
-	}
-      else
-	*start = NULL;
+      gdb_assert (symbaton != NULL);
+      *start = symbaton->data;
+      *length = symbaton->size;
+    }
+  else if (SYMBOL_OPS (framefunc) == &dwarf2_missing_funcs)
+    {
+      struct dwarf2_locexpr_baton *symbaton;
+
+      symbaton = SYMBOL_LOCATION_BATON (framefunc);
+      gdb_assert (symbaton == NULL);
+      *start = NULL;
+      *length = 0;	/* unused */
     }
+  else
+    internal_error (__FILE__, __LINE__,
+		    _("Unsupported SYMBOL_OPS %p for \"%s\""),
+		    SYMBOL_OPS (framefunc), SYMBOL_PRINT_NAME (framefunc));
 
   if (*start == NULL)
     error (_("Could not find the frame base for \"%s\"."),
-	   SYMBOL_NATURAL_NAME (framefunc));
+	   SYMBOL_PRINT_NAME (framefunc));
 }
 
 /* Using the objfile specified in BATON, find the address for the
@@ -191,6 +204,119 @@ dwarf_expr_tls_address (void *baton, CORE_ADDR offset)
   return target_translate_tls_address (debaton->objfile, offset);
 }
 
+static CORE_ADDR
+dwarf_expr_object_address (void *baton)
+{
+  struct dwarf_expr_baton *debaton = baton;
+
+  /* The message is suppressed in DWARF_BLOCK_EXEC.  */
+  if (debaton->object_address == 0)
+    error (_("Cannot resolve DW_OP_push_object_address for a missing object"));
+
+  return debaton->object_address;
+}
+
+/* Address of the variable we are currently referring to.  It is set from
+   DW_TAG_variable's DW_AT_location (not DW_TAG_type's DW_AT_data_location) for
+   DW_OP_push_object_address.  */
+
+static CORE_ADDR object_address;
+
+/* Callers use object_address_set while their callers use the result set so we
+   cannot run the cleanup at the local block of our direct caller.  Still we
+   should reset OBJECT_ADDRESS at least for the next GDB command.  */
+
+static void
+object_address_cleanup (void *prev_save_voidp)
+{
+  CORE_ADDR *prev_save = prev_save_voidp;
+
+  object_address = *prev_save;
+  xfree (prev_save);
+}
+
+/* Set the base address - DW_AT_location - of a variable.  It is being later
+   used to derive other object addresses by DW_OP_push_object_address.
+
+   It would be useful to sanity check ADDRESS - such as for some objects with
+   unset VALUE_ADDRESS - but some valid addresses may be zero (such as first
+   objects in relocatable .o files).  */
+
+void
+object_address_set (CORE_ADDR address)
+{
+  CORE_ADDR *prev_save;
+
+  prev_save = xmalloc (sizeof *prev_save);
+  *prev_save = object_address;
+  make_cleanup (object_address_cleanup, prev_save);
+
+  object_address = address;
+}
+
+/* Evaluate DWARF expression at DATA ... DATA + SIZE with its result readable
+   by dwarf_expr_fetch (RETVAL, 0).  FRAME parameter can be NULL to call
+   get_selected_frame to find it.  Returned dwarf_expr_context freeing is
+   pushed on the cleanup chain.  */
+
+static struct dwarf_expr_context *
+dwarf_expr_prep_ctx (struct frame_info *frame, gdb_byte *data,
+		     unsigned short size, struct dwarf2_per_cu_data *per_cu)
+{
+  struct dwarf_expr_context *ctx;
+  struct dwarf_expr_baton baton;
+
+  if (!frame)
+    frame = get_selected_frame (NULL);
+
+  baton.frame = frame;
+  baton.objfile = dwarf2_per_cu_objfile (per_cu);
+  baton.object_address = object_address;
+
+  ctx = new_dwarf_expr_context ();
+  ctx->gdbarch = get_objfile_arch (baton.objfile);
+  ctx->addr_size = dwarf2_per_cu_addr_size (per_cu);
+  ctx->baton = &baton;
+  ctx->read_reg = dwarf_expr_read_reg;
+  ctx->read_mem = dwarf_expr_read_mem;
+  ctx->get_frame_base = dwarf_expr_frame_base;
+  ctx->get_tls_address = dwarf_expr_tls_address;
+  ctx->get_object_address = dwarf_expr_object_address;
+
+  make_cleanup ((make_cleanup_ftype *) free_dwarf_expr_context, ctx);
+
+  dwarf_expr_eval (ctx, data, size);
+
+  /* It was used only during dwarf_expr_eval.  */
+  ctx->baton = NULL;
+
+  return ctx;
+}
+
+/* Evaluate DWARF expression at DLBATON expecting it produces exactly one
+   CORE_ADDR result on the DWARF stack stack.  */
+
+CORE_ADDR
+dwarf_locexpr_baton_eval (struct dwarf2_locexpr_baton *dlbaton)
+{
+  struct dwarf_expr_context *ctx;
+  CORE_ADDR retval;
+  struct cleanup *back_to = make_cleanup (null_cleanup, 0);
+
+  ctx = dwarf_expr_prep_ctx (NULL, dlbaton->data, dlbaton->size,
+			     dlbaton->per_cu);
+  if (ctx->num_pieces > 0)
+    error (_("DW_OP_*piece is unsupported for DW_FORM_block"));
+  else if (ctx->in_reg)
+    error (_("Register result is unsupported for DW_FORM_block"));
+
+  retval = dwarf_expr_fetch (ctx, 0);
+
+  do_cleanups (back_to);
+
+  return retval;
+}
+
 /* Evaluate a location description, starting at DATA and with length
    SIZE, to find the current location of variable VAR in the context
    of FRAME.  */
@@ -200,8 +326,8 @@ dwarf2_evaluate_loc_desc (struct symbol *var, struct frame_info *frame,
 			  struct dwarf2_per_cu_data *per_cu)
 {
   struct value *retval;
-  struct dwarf_expr_baton baton;
   struct dwarf_expr_context *ctx;
+  struct cleanup *back_to = make_cleanup (null_cleanup, 0);
 
   if (size == 0)
     {
@@ -211,19 +337,8 @@ dwarf2_evaluate_loc_desc (struct symbol *var, struct frame_info *frame,
       return retval;
     }
 
-  baton.frame = frame;
-  baton.objfile = dwarf2_per_cu_objfile (per_cu);
+  ctx = dwarf_expr_prep_ctx (frame, data, size, per_cu);
 
-  ctx = new_dwarf_expr_context ();
-  ctx->gdbarch = get_objfile_arch (baton.objfile);
-  ctx->addr_size = dwarf2_per_cu_addr_size (per_cu);
-  ctx->baton = &baton;
-  ctx->read_reg = dwarf_expr_read_reg;
-  ctx->read_mem = dwarf_expr_read_mem;
-  ctx->get_frame_base = dwarf_expr_frame_base;
-  ctx->get_tls_address = dwarf_expr_tls_address;
-
-  dwarf_expr_eval (ctx, data, size);
   if (ctx->num_pieces > 0)
     {
       int i;
@@ -261,6 +376,10 @@ dwarf2_evaluate_loc_desc (struct symbol *var, struct frame_info *frame,
     {
       CORE_ADDR address = dwarf_expr_fetch (ctx, 0);
 
+      /* object_address_set called here is required in ALLOCATE_VALUE's
+	 CHECK_TYPEDEF for the object's possible DW_OP_push_object_address.  */
+      object_address_set (address);
+
       retval = allocate_value (SYMBOL_TYPE (var));
       VALUE_LVAL (retval) = lval_memory;
       set_value_lazy (retval, 1);
@@ -269,7 +388,7 @@ dwarf2_evaluate_loc_desc (struct symbol *var, struct frame_info *frame,
 
   set_value_initialized (retval, ctx->initialized);
 
-  free_dwarf_expr_context (ctx);
+  do_cleanups (back_to);
 
   return retval;
 }
@@ -587,7 +706,7 @@ static int
 loclist_describe_location (struct symbol *symbol, struct ui_file *stream)
 {
   /* FIXME: Could print the entire list of locations.  */
-  fprintf_filtered (stream, "a variable with multiple locations");
+  fprintf_filtered (stream, _("a variable with multiple locations"));
   return 1;
 }
 
@@ -603,16 +722,56 @@ loclist_tracepoint_var_ref (struct symbol * symbol, struct agent_expr * ax,
 
   data = find_location_expression (dlbaton, &size, ax->scope);
   if (data == NULL)
-    error (_("Variable \"%s\" is not available."), SYMBOL_NATURAL_NAME (symbol));
+    error (_("Variable \"%s\" is not available."), SYMBOL_PRINT_NAME (symbol));
 
   dwarf2_tracepoint_var_ref (symbol, ax, value, data, size);
 }
 
-/* The set of location functions used with the DWARF-2 expression
-   evaluator and location lists.  */
+/* The set of location functions used with the DWARF-2 location lists.  */
 const struct symbol_ops dwarf2_loclist_funcs = {
   loclist_read_variable,
   loclist_read_needs_frame,
   loclist_describe_location,
   loclist_tracepoint_var_ref
 };
+
+static struct value *
+missing_read_variable (struct symbol *symbol, struct frame_info *frame)
+{
+  struct dwarf2_loclist_baton *dlbaton = SYMBOL_LOCATION_BATON (symbol);
+
+  gdb_assert (dlbaton == NULL);
+  error (_("Unable to resolve variable \"%s\""), SYMBOL_PRINT_NAME (symbol));
+}
+
+static int
+missing_read_needs_frame (struct symbol *symbol)
+{
+  return 0;
+}
+
+static int
+missing_describe_location (struct symbol *symbol, struct ui_file *stream)
+{
+  fprintf_filtered (stream, _("a variable we are unable to resolve"));
+  return 1;
+}
+
+static void
+missing_tracepoint_var_ref (struct symbol *symbol, struct agent_expr *ax,
+			    struct axs_value *value)
+{
+  struct dwarf2_loclist_baton *dlbaton = SYMBOL_LOCATION_BATON (symbol);
+
+  gdb_assert (dlbaton == NULL);
+  error (_("Unable to resolve variable \"%s\""), SYMBOL_PRINT_NAME (symbol));
+}
+
+/* The set of location functions used with the DWARF-2 evaluator when we are
+   unable to resolve the symbols.  */
+const struct symbol_ops dwarf2_missing_funcs = {
+  missing_read_variable,
+  missing_read_needs_frame,
+  missing_describe_location,
+  missing_tracepoint_var_ref
+};
diff --git a/gdb/dwarf2loc.h b/gdb/dwarf2loc.h
index 76577f1..bf46761 100644
--- a/gdb/dwarf2loc.h
+++ b/gdb/dwarf2loc.h
@@ -71,5 +71,11 @@ struct dwarf2_loclist_baton
 
 extern const struct symbol_ops dwarf2_locexpr_funcs;
 extern const struct symbol_ops dwarf2_loclist_funcs;
+extern const struct symbol_ops dwarf2_missing_funcs;
+
+extern void object_address_set (CORE_ADDR address);
+
+extern CORE_ADDR dwarf_locexpr_baton_eval
+  (struct dwarf2_locexpr_baton *dlbaton);
 
 #endif /* dwarf2loc.h */
diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
index 925cfd0..481cdfc 100644
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -1024,7 +1024,14 @@ static void store_in_ref_table (struct die_info *,
 
 static unsigned int dwarf2_get_ref_die_offset (struct attribute *);
 
-static int dwarf2_get_attr_constant_value (struct attribute *, int);
+enum dwarf2_get_attr_constant_value
+  {
+    dwarf2_attr_unknown,
+    dwarf2_attr_const,
+    dwarf2_attr_block
+  };
+static enum dwarf2_get_attr_constant_value dwarf2_get_attr_constant_value
+  (struct attribute *attr, int *val_return);
 
 static struct die_info *follow_die_ref (struct die_info *,
 					struct attribute *,
@@ -1079,6 +1086,9 @@ static void age_cached_comp_units (void);
 
 static void free_one_cached_comp_unit (void *);
 
+static void fetch_die_type_attrs (struct die_info *die, struct type *type,
+				  struct dwarf2_cu *cu);
+
 static struct type *set_die_type (struct die_info *, struct type *,
 				  struct dwarf2_cu *);
 
@@ -1098,6 +1108,9 @@ static void dwarf2_clear_marks (struct dwarf2_per_cu_data *);
 
 static struct type *get_die_type (struct die_info *die, struct dwarf2_cu *cu);
 
+static struct dwarf2_locexpr_baton *dwarf2_attr_to_locexpr_baton
+  (struct attribute *attr, struct dwarf2_cu *cu);
+
 /* Try to locate the sections we need for DWARF 2 debugging
    information and return true if we have enough to do something.  */
 
@@ -3900,7 +3913,7 @@ dwarf2_add_member_fn (struct field_info *fip, struct die_info *die,
   /* The name is already allocated along with this objfile, so we don't
      need to duplicate it for the type.  */
   fnp->physname = physname ? physname : "";
-  fnp->type = alloc_type (objfile);
+  fnp->type = alloc_type (objfile, NULL);
   this_type = read_type_die (die, cu);
   if (this_type && TYPE_CODE (this_type) == TYPE_CODE_FUNC)
     {
@@ -4084,7 +4097,7 @@ quirk_gcc_member_function_pointer (struct die_info *die, struct dwarf2_cu *cu)
     return NULL;
 
   domain_type = TYPE_TARGET_TYPE (TYPE_FIELD_TYPE (pfn_type, 0));
-  type = alloc_type (objfile);
+  type = alloc_type (objfile, NULL);
   smash_to_method_type (type, domain_type, TYPE_TARGET_TYPE (pfn_type),
 			TYPE_FIELDS (pfn_type), TYPE_NFIELDS (pfn_type),
 			TYPE_VARARGS (pfn_type));
@@ -4121,7 +4134,7 @@ read_structure_type (struct die_info *die, struct dwarf2_cu *cu)
   if (type)
     return type;
 
-  type = alloc_type (objfile);
+  type = alloc_type (objfile, NULL);
   INIT_CPLUS_SPECIFIC (type);
   name = dwarf2_name (die, cu);
   if (name != NULL)
@@ -4334,7 +4347,7 @@ read_enumeration_type (struct die_info *die, struct dwarf2_cu *cu)
   struct attribute *attr;
   const char *name;
 
-  type = alloc_type (objfile);
+  type = alloc_type (objfile, NULL);
 
   TYPE_CODE (type) = TYPE_CODE_ENUM;
   name = dwarf2_full_name (die, cu);
@@ -4484,6 +4497,29 @@ process_enumeration_scope (struct die_info *die, struct dwarf2_cu *cu)
   new_symbol (die, this_type, cu);
 }
 
+/* Create a new array dimension referencing its target type TYPE.
+
+   Multidimensional arrays are internally represented as a stack of
+   singledimensional arrays being referenced by their TYPE_TARGET_TYPE.  */
+
+static struct type *
+create_single_array_dimension (struct type *type, struct type *range_type,
+			       struct die_info *die, struct dwarf2_cu *cu)
+{
+  type = create_array_type (NULL, type, range_type);
+
+  /* These generic type attributes need to be fetched by
+     evaluate_subexp_standard <multi_f77_subscript>'s call of
+     value_subscripted_rvalue only for the innermost array type.  */
+  fetch_die_type_attrs (die, type, cu);
+
+  /* These generic type attributes are checked for allocated/associated
+     validity while accessing FIELD_LOC_KIND_DWARF_BLOCK.  */
+  fetch_die_type_attrs (die, range_type, cu);
+
+  return type;
+}
+
 /* Extract all information from a DW_TAG_array_type DIE and put it in
    the DIE's type field.  For now, this only handles one dimensional
    arrays.  */
@@ -4497,7 +4533,7 @@ read_array_type (struct die_info *die, struct dwarf2_cu *cu)
   struct type *element_type, *range_type, *index_type;
   struct type **range_types = NULL;
   struct attribute *attr;
-  int ndim = 0;
+  int ndim = 0, i;
   struct cleanup *back_to;
   char *name;
 
@@ -4544,16 +4580,11 @@ read_array_type (struct die_info *die, struct dwarf2_cu *cu)
   type = element_type;
 
   if (read_array_order (die, cu) == DW_ORD_col_major)
-    {
-      int i = 0;
-      while (i < ndim)
-	type = create_array_type (NULL, type, range_types[i++]);
-    }
-  else
-    {
-      while (ndim-- > 0)
-	type = create_array_type (NULL, type, range_types[ndim]);
-    }
+    for (i = 0; i < ndim; i++)
+      type = create_single_array_dimension (type, range_types[i], die, cu);
+  else /* (read_array_order (die, cu) == DW_ORD_row_major) */
+    for (i = ndim - 1; i >= 0; i--)
+      type = create_single_array_dimension (type, range_types[i], die, cu);
 
   /* Understand Dwarf2 support for vector types (like they occur on
      the PowerPC w/ AltiVec).  Gcc just adds another attribute to the
@@ -4909,29 +4940,95 @@ read_tag_string_type (struct die_info *die, struct dwarf2_cu *cu)
   struct objfile *objfile = cu->objfile;
   struct type *type, *range_type, *index_type, *char_type;
   struct attribute *attr;
-  unsigned int length;
+  int length;
+
+  index_type = builtin_type_int32;
+  /* RANGE_TYPE is allocated from OBJFILE, not OBJFILE_INTERNAL.  */
+  range_type = alloc_type (objfile, index_type);
+  /* LOW_BOUND and HIGH_BOUND are set for real below.  */
+  range_type = create_range_type (range_type, index_type, 0, -1);
+
+  /* C/C++ should probably have the low bound 0 but C/C++ does not use
+     DW_TAG_string_type.  */
+  TYPE_LOW_BOUND (range_type) = 1;
 
   attr = dwarf2_attr (die, DW_AT_string_length, cu);
-  if (attr)
-    {
-      length = DW_UNSND (attr);
-    }
-  else
-    {
-      /* check for the DW_AT_byte_size attribute */
+  switch (dwarf2_get_attr_constant_value (attr, &length))
+    {
+    case dwarf2_attr_const:
+      /* We currently do not support a constant address where the location
+	 should be read from - DWARF2_ATTR_BLOCK is expected instead.  See
+	 DWARF for the DW_AT_STRING_LENGTH vs. DW_AT_BYTE_SIZE difference.  */
+      /* PASSTHRU */
+    case dwarf2_attr_unknown:
       attr = dwarf2_attr (die, DW_AT_byte_size, cu);
-      if (attr)
-        {
-          length = DW_UNSND (attr);
-        }
-      else
-        {
-          length = 1;
-        }
+      switch (dwarf2_get_attr_constant_value (attr, &length))
+	{
+	case dwarf2_attr_unknown:
+	  length = 1;
+	  /* PASSTHRU */
+	case dwarf2_attr_const:
+	  TYPE_HIGH_BOUND (range_type) = length;
+	  break;
+	case dwarf2_attr_block:
+	  TYPE_RANGE_BOUND_SET_DWARF_BLOCK (range_type, 1);
+	  TYPE_FIELD_DWARF_BLOCK (range_type, 1) =
+					dwarf2_attr_to_locexpr_baton (attr, cu);
+	  TYPE_DYNAMIC (range_type) = 1;
+	  break;
+	}
+      break;
+    case dwarf2_attr_block:
+      /* Security check for a size overflow.  */
+      if (DW_BLOCK (attr)->size + 2 < DW_BLOCK (attr)->size)
+	{
+	  TYPE_HIGH_BOUND (range_type) = 1;
+	  break;
+	}
+      /* Extend the DWARF block by a new DW_OP_deref/DW_OP_deref_size
+	 instruction as DW_AT_string_length specifies the length location, not
+	 its value.  */
+      {
+	struct dwarf2_locexpr_baton *length_baton;
+	struct attribute *size_attr;
+
+	length_baton = obstack_alloc (&cu->comp_unit_obstack,
+				      sizeof (*length_baton));
+	length_baton->per_cu = cu->per_cu;
+	length_baton->data = obstack_alloc (&cu->comp_unit_obstack,
+					    DW_BLOCK (attr)->size + 2);
+	memcpy (length_baton->data, DW_BLOCK (attr)->data,
+		DW_BLOCK (attr)->size);
+
+	/* DW_AT_BYTE_SIZE existing together with DW_AT_STRING_LENGTH specifies
+	   the size of an integer to fetch.  */
+
+	size_attr = dwarf2_attr (die, DW_AT_byte_size, cu);
+	if (size_attr)
+	  {
+	    length_baton->size = DW_BLOCK (attr)->size + 2;
+	    length_baton->data[DW_BLOCK (attr)->size] = DW_OP_deref_size;
+	    length_baton->data[DW_BLOCK (attr)->size + 1]
+							 = DW_UNSND (size_attr);
+	    if (length_baton->data[DW_BLOCK (attr)->size + 1]
+	        != DW_UNSND (size_attr))
+	      complaint (&symfile_complaints,
+	                 _("DW_AT_string_length's DW_AT_byte_size integer "
+			   "exceeds the byte size storage"));
+	  }
+	else
+	  {
+	    length_baton->size = DW_BLOCK (attr)->size + 1;
+	    length_baton->data[DW_BLOCK (attr)->size] = DW_OP_deref;
+	  }
+
+	TYPE_RANGE_BOUND_SET_DWARF_BLOCK (range_type, 1);
+	TYPE_FIELD_DWARF_BLOCK (range_type, 1) = length_baton;
+	TYPE_DYNAMIC (range_type) = 1;
+      }
+      break;
     }
 
-  index_type = builtin_type_int32;
-  range_type = create_range_type (NULL, index_type, 1, length);
   type = create_string_type (NULL, range_type);
 
   return set_die_type (die, type, cu);
@@ -5025,7 +5122,6 @@ static struct type *
 read_typedef (struct die_info *die, struct dwarf2_cu *cu)
 {
   struct objfile *objfile = cu->objfile;
-  struct attribute *attr;
   const char *name = NULL;
   struct type *this_type;
 
@@ -5131,8 +5227,8 @@ read_subrange_type (struct die_info *die, struct dwarf2_cu *cu)
   struct type *base_type;
   struct type *range_type;
   struct attribute *attr;
-  int low = 0;
-  int high = -1;
+  int low, high, byte_stride_int;
+  enum dwarf2_get_attr_constant_value high_type;
   char *name;
   
   base_type = die_type (die, cu);
@@ -5145,42 +5241,90 @@ read_subrange_type (struct die_info *die, struct dwarf2_cu *cu)
 		     0, NULL, cu->objfile);
     }
 
-  if (cu->language == language_fortran)
-    { 
-      /* FORTRAN implies a lower bound of 1, if not given.  */
-      low = 1;
-    }
+  /* LOW_BOUND and HIGH_BOUND are set for real below.  */
+  range_type = create_range_type (NULL, base_type, 0, -1);
 
-  /* FIXME: For variable sized arrays either of these could be
-     a variable rather than a constant value.  We'll allow it,
-     but we don't know how to handle it.  */
   attr = dwarf2_attr (die, DW_AT_lower_bound, cu);
-  if (attr)
-    low = dwarf2_get_attr_constant_value (attr, 0);
+  switch (dwarf2_get_attr_constant_value (attr, &low))
+    {
+    case dwarf2_attr_unknown:
+      if (cu->language == language_fortran)
+	{
+	  /* FORTRAN implies a lower bound of 1, if not given.  */
+	  low = 1;
+	}
+      else
+        {
+	  /* According to DWARF we should assume the value 0 only for
+	     LANGUAGE_C and LANGUAGE_CPLUS.  */
+	  low = 0;
+	}
+      /* PASSTHRU */
+    case dwarf2_attr_const:
+      TYPE_LOW_BOUND (range_type) = low;
+      if (low >= 0)
+	TYPE_UNSIGNED (range_type) = 1;
+      break;
+    case dwarf2_attr_block:
+      TYPE_RANGE_BOUND_SET_DWARF_BLOCK (range_type, 0);
+      TYPE_FIELD_DWARF_BLOCK (range_type, 0) = dwarf2_attr_to_locexpr_baton
+								     (attr, cu);
+      TYPE_DYNAMIC (range_type) = 1;
+      /* For setting a default if DW_AT_UPPER_BOUND would be missing.  */
+      low = 0;
+      break;
+    }
 
   attr = dwarf2_attr (die, DW_AT_upper_bound, cu);
-  if (attr)
-    {       
-      if (attr->form == DW_FORM_block1)
-        {
-          /* GCC encodes arrays with unspecified or dynamic length
-             with a DW_FORM_block1 attribute.
-             FIXME: GDB does not yet know how to handle dynamic
-             arrays properly, treat them as arrays with unspecified
-             length for now.
-
-             FIXME: jimb/2003-09-22: GDB does not really know
-             how to handle arrays of unspecified length
-             either; we just represent them as zero-length
-             arrays.  Choose an appropriate upper bound given
-             the lower bound we've computed above.  */
-          high = low - 1;
-        }
-      else
-        high = dwarf2_get_attr_constant_value (attr, 1);
+  high_type = dwarf2_get_attr_constant_value (attr, &high);
+  if (high_type == dwarf2_attr_unknown)
+    {
+      attr = dwarf2_attr (die, DW_AT_count, cu);
+      high_type = dwarf2_get_attr_constant_value (attr, &high);
+      /* It does not hurt but it is needlessly ineffective in check_typedef.  */
+      if (high_type != dwarf2_attr_unknown)
+      	{
+	  TYPE_RANGE_HIGH_BOUND_IS_COUNT (range_type) = 1;
+	  TYPE_DYNAMIC (range_type) = 1;
+	}
+      /* Pass it now as the regular DW_AT_upper_bound.  */
+    }
+  switch (high_type)
+    {
+    case dwarf2_attr_unknown:
+      TYPE_RANGE_UPPER_BOUND_IS_UNDEFINED (range_type) = 1;
+      high = low - 1;
+      /* PASSTHRU */
+    case dwarf2_attr_const:
+      TYPE_HIGH_BOUND (range_type) = high;
+      break;
+    case dwarf2_attr_block:
+      TYPE_RANGE_BOUND_SET_DWARF_BLOCK (range_type, 1);
+      TYPE_FIELD_DWARF_BLOCK (range_type, 1) = dwarf2_attr_to_locexpr_baton
+								     (attr, cu);
+      TYPE_DYNAMIC (range_type) = 1;
+      break;
     }
 
-  range_type = create_range_type (NULL, base_type, low, high);
+  /* DW_AT_bit_stride is currently unsupported as we count in bytes.  */
+  attr = dwarf2_attr (die, DW_AT_byte_stride, cu);
+  switch (dwarf2_get_attr_constant_value (attr, &byte_stride_int))
+    {
+    case dwarf2_attr_unknown:
+      break;
+    case dwarf2_attr_const:
+      if (byte_stride_int == 0)
+	complaint (&symfile_complaints,
+		   _("Found DW_AT_byte_stride with unsupported value 0"));
+      TYPE_BYTE_STRIDE (range_type) = byte_stride_int;
+      break;
+    case dwarf2_attr_block:
+      TYPE_RANGE_BOUND_SET_DWARF_BLOCK (range_type, 2);
+      TYPE_FIELD_DWARF_BLOCK (range_type, 2) = dwarf2_attr_to_locexpr_baton
+								     (attr, cu);
+      TYPE_DYNAMIC (range_type) = 1;
+      break;
+    }
 
   name = dwarf2_name (die, cu);
   if (name)
@@ -7468,10 +7612,12 @@ var_decode_location (struct attribute *attr, struct symbol *sym,
      (i.e. when the value of a register or memory location is
      referenced, or a thread-local block, etc.).  Then again, it might
      not be worthwhile.  I'm assuming that it isn't unless performance
-     or memory numbers show me otherwise.  */
+     or memory numbers show me otherwise.
+     
+     SYMBOL_CLASS may get overriden by dwarf2_symbol_mark_computed.  */
 
-  dwarf2_symbol_mark_computed (attr, sym, cu);
   SYMBOL_CLASS (sym) = LOC_COMPUTED;
+  dwarf2_symbol_mark_computed (attr, sym, cu);
 }
 
 /* Given a pointer to a DWARF information entry, figure out if we need
@@ -7504,6 +7650,8 @@ new_symbol (struct die_info *die, struct type *type, struct dwarf2_cu *cu)
 					     sizeof (struct symbol));
       OBJSTAT (objfile, n_syms++);
       memset (sym, 0, sizeof (struct symbol));
+      /* Some methods are called without checking SYMBOL_OPS validity.  */
+      SYMBOL_OPS (sym) = &dwarf2_missing_funcs;
 
       /* Cache this symbol's name and the name's demangled form (if any).  */
       SYMBOL_LANGUAGE (sym) = cu->language;
@@ -9309,26 +9457,35 @@ dwarf2_get_ref_die_offset (struct attribute *attr)
   return result;
 }
 
-/* Return the constant value held by the given attribute.  Return -1
-   if the value held by the attribute is not constant.  */
+/* (*val_return) is filled only if returning dwarf2_attr_const.  */
 
-static int
-dwarf2_get_attr_constant_value (struct attribute *attr, int default_value)
+static enum dwarf2_get_attr_constant_value
+dwarf2_get_attr_constant_value (struct attribute *attr, int *val_return)
 {
+  if (attr == NULL)
+    return dwarf2_attr_unknown;
   if (attr->form == DW_FORM_sdata)
-    return DW_SND (attr);
-  else if (attr->form == DW_FORM_udata
-           || attr->form == DW_FORM_data1
-           || attr->form == DW_FORM_data2
-           || attr->form == DW_FORM_data4
-           || attr->form == DW_FORM_data8)
-    return DW_UNSND (attr);
-  else
     {
-      complaint (&symfile_complaints, _("Attribute value is not a constant (%s)"),
-                 dwarf_form_name (attr->form));
-      return default_value;
+      *val_return = DW_SND (attr);
+      return dwarf2_attr_const;
     }
+  if (attr->form == DW_FORM_udata
+      || attr->form == DW_FORM_data1
+      || attr->form == DW_FORM_data2
+      || attr->form == DW_FORM_data4
+      || attr->form == DW_FORM_data8)
+    {
+      *val_return = DW_UNSND (attr);
+      return dwarf2_attr_const;
+    }
+  if (attr->form == DW_FORM_block
+      || attr->form == DW_FORM_block1
+      || attr->form == DW_FORM_block2
+      || attr->form == DW_FORM_block4)
+    return dwarf2_attr_block;
+  complaint (&symfile_complaints, _("Attribute value is not a constant (%s)"),
+             dwarf_form_name (attr->form));
+  return dwarf2_attr_unknown;
 }
 
 /* THIS_CU has a reference to PER_CU.  If necessary, load the new compilation
@@ -10106,6 +10263,34 @@ attr_form_is_constant (struct attribute *attr)
     }
 }
 
+/* Convert DW_BLOCK into struct dwarf2_locexpr_baton.  ATTR must be a DW_BLOCK
+   attribute type.  */
+
+static struct dwarf2_locexpr_baton *
+dwarf2_attr_to_locexpr_baton (struct attribute *attr, struct dwarf2_cu *cu)
+{
+  struct dwarf2_locexpr_baton *baton;
+
+  gdb_assert (attr_form_is_block (attr));
+
+  baton = obstack_alloc (&cu->objfile->objfile_obstack, sizeof (*baton));
+  baton->per_cu = cu->per_cu;
+  gdb_assert (baton->per_cu);
+
+  /* Note that we're just copying the block's data pointer
+     here, not the actual data.  We're still pointing into the
+     info_buffer for SYM's objfile; right now we never release
+     that buffer, but when we do clean up properly this may
+     need to change.  */
+  baton->size = DW_BLOCK (attr)->size;
+  baton->data = DW_BLOCK (attr)->data;
+  gdb_assert (baton->size == 0 || baton->data != NULL);
+
+  return baton;
+}
+
+/* SYM may get its SYMBOL_CLASS overriden on invalid ATTR content.  */
+
 static void
 dwarf2_symbol_mark_computed (struct attribute *attr, struct symbol *sym,
 			     struct dwarf2_cu *cu)
@@ -10135,35 +10320,24 @@ dwarf2_symbol_mark_computed (struct attribute *attr, struct symbol *sym,
       SYMBOL_OPS (sym) = &dwarf2_loclist_funcs;
       SYMBOL_LOCATION_BATON (sym) = baton;
     }
+  else if (attr_form_is_block (attr))
+    {
+      SYMBOL_OPS (sym) = &dwarf2_locexpr_funcs;
+      SYMBOL_LOCATION_BATON (sym) = dwarf2_attr_to_locexpr_baton (attr, cu);
+    }
   else
     {
-      struct dwarf2_locexpr_baton *baton;
+      dwarf2_invalid_attrib_class_complaint ("location description",
+					     SYMBOL_NATURAL_NAME (sym));
 
-      baton = obstack_alloc (&cu->objfile->objfile_obstack,
-			     sizeof (struct dwarf2_locexpr_baton));
-      baton->per_cu = cu->per_cu;
-      gdb_assert (baton->per_cu);
+      /* Some methods are called without checking SYMBOL_OPS validity.  */
+      SYMBOL_OPS (sym) = &dwarf2_missing_funcs;
+      SYMBOL_LOCATION_BATON (sym) = NULL;
 
-      if (attr_form_is_block (attr))
-	{
-	  /* Note that we're just copying the block's data pointer
-	     here, not the actual data.  We're still pointing into the
-	     info_buffer for SYM's objfile; right now we never release
-	     that buffer, but when we do clean up properly this may
-	     need to change.  */
-	  baton->size = DW_BLOCK (attr)->size;
-	  baton->data = DW_BLOCK (attr)->data;
-	}
-      else
-	{
-	  dwarf2_invalid_attrib_class_complaint ("location description",
-						 SYMBOL_NATURAL_NAME (sym));
-	  baton->size = 0;
-	  baton->data = NULL;
-	}
-      
-      SYMBOL_OPS (sym) = &dwarf2_locexpr_funcs;
-      SYMBOL_LOCATION_BATON (sym) = baton;
+      /* For functions a missing DW_AT_frame_base does not optimize out the
+	 whole function definition, only its frame base resolving.  */
+      if (attr->name == DW_AT_location)
+	SYMBOL_CLASS (sym) = LOC_OPTIMIZED_OUT;
     }
 }
 
@@ -10438,6 +10612,31 @@ offset_and_type_eq (const void *item_lhs, const void *item_rhs)
   return ofs_lhs->offset == ofs_rhs->offset;
 }
 
+/* Fill in generic attributes applicable for type DIEs.  */
+
+static void
+fetch_die_type_attrs (struct die_info *die, struct type *type,
+		      struct dwarf2_cu *cu)
+{
+  struct attribute *attr;
+
+  attr = dwarf2_attr (die, DW_AT_data_location, cu);
+  if (attr_form_is_block (attr))
+    TYPE_DATA_LOCATION_DWARF_BLOCK (type) = dwarf2_attr_to_locexpr_baton (attr,
+									  cu);
+  gdb_assert (!TYPE_DATA_LOCATION_IS_ADDR (type));
+
+  attr = dwarf2_attr (die, DW_AT_allocated, cu);
+  if (attr_form_is_block (attr))
+    TYPE_ALLOCATED (type) = dwarf2_attr_to_locexpr_baton (attr, cu);
+  gdb_assert (!TYPE_NOT_ALLOCATED (type));
+
+  attr = dwarf2_attr (die, DW_AT_associated, cu);
+  if (attr_form_is_block (attr))
+    TYPE_ASSOCIATED (type) = dwarf2_attr_to_locexpr_baton (attr, cu);
+  gdb_assert (!TYPE_NOT_ASSOCIATED (type));
+}
+
 /* Set the type associated with DIE to TYPE.  Save it in CU's hash
    table if necessary.  For convenience, return TYPE.  */
 
@@ -10446,6 +10645,8 @@ set_die_type (struct die_info *die, struct type *type, struct dwarf2_cu *cu)
 {
   struct dwarf2_offset_and_type **slot, ofs;
 
+  fetch_die_type_attrs (die, type, cu);
+
   if (cu->type_hash == NULL)
     {
       gdb_assert (cu->per_cu != NULL);
diff --git a/gdb/eval.c b/gdb/eval.c
index 038334b..3e2c320 100644
--- a/gdb/eval.c
+++ b/gdb/eval.c
@@ -40,6 +40,7 @@
 #include "regcache.h"
 #include "user-regs.h"
 #include "valprint.h"
+#include "dwarf2loc.h"
 
 #include "gdb_assert.h"
 
@@ -671,6 +672,7 @@ evaluate_subexp_standard (struct type *expect_type,
   long mem_offset;
   struct type **arg_types;
   int save_pos1;
+  struct cleanup *old_chain;
 
   pc = (*pos)++;
   op = exp->elts[pc].opcode;
@@ -1529,7 +1531,10 @@ evaluate_subexp_standard (struct type *expect_type,
 
       /* First determine the type code we are dealing with.  */
       arg1 = evaluate_subexp (NULL_TYPE, exp, pos, noside);
+      old_chain = make_cleanup (null_cleanup, 0);
+      object_address_set (VALUE_ADDRESS (arg1));
       type = check_typedef (value_type (arg1));
+      do_cleanups (old_chain);
       code = TYPE_CODE (type);
 
       if (code == TYPE_CODE_PTR)
@@ -1963,13 +1968,19 @@ evaluate_subexp_standard (struct type *expect_type,
       {
 	int subscript_array[MAX_FORTRAN_DIMS];
 	int array_size_array[MAX_FORTRAN_DIMS];
+	int byte_stride_array[MAX_FORTRAN_DIMS];
 	int ndimensions = 1, i;
 	struct type *tmp_type;
 	int offset_item;	/* The array offset where the item lives */
+	CORE_ADDR offset_byte;	/* byte_stride based offset  */
+	unsigned element_size;
 
 	if (nargs > MAX_FORTRAN_DIMS)
 	  error (_("Too many subscripts for F77 (%d Max)"), MAX_FORTRAN_DIMS);
 
+	old_chain = make_cleanup (null_cleanup, 0);
+	object_address_set (VALUE_ADDRESS (arg1));
+
 	tmp_type = check_typedef (value_type (arg1));
 	ndimensions = calc_f77_array_dims (type);
 
@@ -1999,6 +2010,9 @@ evaluate_subexp_standard (struct type *expect_type,
 	    upper = f77_get_upperbound (tmp_type);
 	    lower = f77_get_lowerbound (tmp_type);
 
+	    byte_stride_array[nargs - i - 1] =
+					TYPE_ARRAY_BYTE_STRIDE_VALUE (tmp_type);
+
 	    array_size_array[nargs - i - 1] = upper - lower + 1;
 
 	    /* Zero-normalize subscripts so that offsetting will work. */
@@ -2017,17 +2031,25 @@ evaluate_subexp_standard (struct type *expect_type,
 	      tmp_type = check_typedef (TYPE_TARGET_TYPE (tmp_type));
 	  }
 
-	/* Now let us calculate the offset for this item */
+	/* Kept for the f77_get_upperbound / f77_get_lowerbound calls above.  */
+	do_cleanups (old_chain);
 
-	offset_item = subscript_array[ndimensions - 1];
+	/* Now let us calculate the offset for this item */
 
-	for (i = ndimensions - 1; i > 0; --i)
-	  offset_item =
-	    array_size_array[i - 1] * offset_item + subscript_array[i - 1];
+	offset_item = 0;
+	offset_byte = 0;
 
-	/* Construct a value node with the value of the offset */
+	for (i = ndimensions - 1; i >= 0; --i)
+	  {
+	    offset_item *= array_size_array[i];
+	    if (byte_stride_array[i] == 0)
+	      offset_item += subscript_array[i];
+	    else
+	      offset_byte += subscript_array[i] * byte_stride_array[i];
+	  }
 
-	arg2 = value_from_longest (builtin_type_int32, offset_item);
+	element_size = TYPE_LENGTH (TYPE_TARGET_TYPE (tmp_type));
+	offset_byte += offset_item * element_size;
 
 	/* Let us now play a dirty trick: we will take arg1 
 	   which is a value node pointing to the topmost level
@@ -2037,7 +2059,7 @@ evaluate_subexp_standard (struct type *expect_type,
 	   returns the correct type value */
 
 	deprecated_set_value_type (arg1, tmp_type);
-	return value_subscripted_rvalue (arg1, arg2, 0);
+	return value_subscripted_rvalue (arg1, offset_byte);
       }
 
     case BINOP_LOGICAL_AND:
@@ -2624,12 +2646,13 @@ evaluate_subexp_with_coercion (struct expression *exp,
     {
     case OP_VAR_VALUE:
       var = exp->elts[pc + 2].symbol;
+      /* address_of_variable will call object_address_set for check_typedef.  */
+      val = address_of_variable (var, exp->elts[pc + 1].block);
       type = check_typedef (SYMBOL_TYPE (var));
       if (TYPE_CODE (type) == TYPE_CODE_ARRAY
 	  && CAST_IS_CONVERSION)
 	{
 	  (*pos) += 4;
-	  val = address_of_variable (var, exp->elts[pc + 1].block);
 	  return value_cast (lookup_pointer_type (TYPE_TARGET_TYPE (type)),
 			     val);
 	}
@@ -2681,9 +2704,13 @@ evaluate_subexp_for_sizeof (struct expression *exp, int *pos)
 
     case OP_VAR_VALUE:
       (*pos) += 4;
-      type = check_typedef (SYMBOL_TYPE (exp->elts[pc + 2].symbol));
-      return
-	value_from_longest (size_type, (LONGEST) TYPE_LENGTH (type));
+      /* We do not need to call read_var_value but the object evaluation may
+	 need to have executed object_address_set which needs valid
+	 SYMBOL_VALUE_ADDRESS of the symbol.  Still VALUE returned by
+	 read_var_value we left as lazy.  */
+      type = value_type (read_var_value (exp->elts[pc + 2].symbol,
+					deprecated_safe_get_selected_frame ()));
+      return value_from_longest (size_type, (LONGEST) TYPE_LENGTH (type));
 
     default:
       val = evaluate_subexp (NULL_TYPE, exp, pos, EVAL_AVOID_SIDE_EFFECTS);
diff --git a/gdb/f-lang.h b/gdb/f-lang.h
index 711bdba..123a783 100644
--- a/gdb/f-lang.h
+++ b/gdb/f-lang.h
@@ -28,6 +28,10 @@ extern void f_error (char *);	/* Defined in f-exp.y */
 extern void f_print_type (struct type *, char *, struct ui_file *, int,
 			  int);
 
+extern const char *f_object_address_data_valid_print_to_stream
+  (struct type *type, struct ui_file *stream);
+extern void f_object_address_data_valid_or_error (struct type *type);
+
 extern int f_val_print (struct type *, const gdb_byte *, int, CORE_ADDR,
 			struct ui_file *, int,
 			const struct value_print_options *);
diff --git a/gdb/f-typeprint.c b/gdb/f-typeprint.c
index 6c9668f..852b9a8 100644
--- a/gdb/f-typeprint.c
+++ b/gdb/f-typeprint.c
@@ -31,7 +31,7 @@
 #include "gdbcore.h"
 #include "target.h"
 #include "f-lang.h"
-
+#include "dwarf2loc.h"
 #include "gdb_string.h"
 #include <errno.h>
 
@@ -48,6 +48,34 @@ void f_type_print_varspec_prefix (struct type *, struct ui_file *,
 void f_type_print_base (struct type *, struct ui_file *, int, int);
 
 
+const char *
+f_object_address_data_valid_print_to_stream (struct type *type,
+					     struct ui_file *stream)
+{
+  const char *msg;
+
+  msg = object_address_data_not_valid (type);
+  if (msg != NULL)
+    {
+      /* Assuming the content printed to STREAM should not be localized.  */
+      fprintf_filtered (stream, "<%s>", msg);
+    }
+
+  return msg;
+}
+
+void
+f_object_address_data_valid_or_error (struct type *type)
+{
+  const char *msg;
+
+  msg = object_address_data_not_valid (type);
+  if (msg != NULL)
+    {
+      error (_("Cannot access it because the %s."), _(msg));
+    }
+}
+
 /* LEVEL is the depth to indent lines by.  */
 
 void
@@ -57,6 +85,9 @@ f_print_type (struct type *type, char *varstring, struct ui_file *stream,
   enum type_code code;
   int demangled_args;
 
+  if (f_object_address_data_valid_print_to_stream (type, stream) != NULL)
+    return;
+
   f_type_print_base (type, stream, show, level);
   code = TYPE_CODE (type);
   if ((varstring != NULL && *varstring != '\0')
@@ -166,6 +197,9 @@ f_type_print_varspec_suffix (struct type *type, struct ui_file *stream,
 
   QUIT;
 
+  if (TYPE_CODE (type) != TYPE_CODE_TYPEDEF)
+    CHECK_TYPEDEF (type);
+
   switch (TYPE_CODE (type))
     {
     case TYPE_CODE_ARRAY:
diff --git a/gdb/f-valprint.c b/gdb/f-valprint.c
index 5721041..a4d69fb 100644
--- a/gdb/f-valprint.c
+++ b/gdb/f-valprint.c
@@ -54,15 +54,17 @@ int f77_array_offset_tbl[MAX_FORTRAN_DIMS + 1][2];
 /* The following macro gives us the size of the nth dimension, Where 
    n is 1 based. */
 
-#define F77_DIM_SIZE(n) (f77_array_offset_tbl[n][1])
+#define F77_DIM_COUNT(n) (f77_array_offset_tbl[n][1])
 
-/* The following gives us the offset for row n where n is 1-based. */
+/* The following gives us the element size for row n where n is 1-based. */
 
-#define F77_DIM_OFFSET(n) (f77_array_offset_tbl[n][0])
+#define F77_DIM_BYTE_STRIDE(n) (f77_array_offset_tbl[n][0])
 
 int
 f77_get_lowerbound (struct type *type)
 {
+  f_object_address_data_valid_or_error (type);
+
   if (TYPE_ARRAY_LOWER_BOUND_IS_UNDEFINED (type))
     error (_("Lower bound may not be '*' in F77"));
 
@@ -72,14 +74,17 @@ f77_get_lowerbound (struct type *type)
 int
 f77_get_upperbound (struct type *type)
 {
+  f_object_address_data_valid_or_error (type);
+
   if (TYPE_ARRAY_UPPER_BOUND_IS_UNDEFINED (type))
     {
-      /* We have an assumed size array on our hands.  Assume that
-	 upper_bound == lower_bound so that we show at least 1 element.
-	 If the user wants to see more elements, let him manually ask for 'em
-	 and we'll subscript the array and show him.  */
+      /* We have an assumed size array on our hands.  As type_length_get
+	 already assumes a length zero of arrays with underfined bounds VALADDR
+	 passed to the Fortran functions does not contained the real inferior
+	 memory content.  User should request printing of specific array
+	 elements instead.  */
 
-      return f77_get_lowerbound (type);
+      return f77_get_lowerbound (type) - 1;
     }
 
   return TYPE_ARRAY_UPPER_BOUND_VALUE (type);
@@ -135,24 +140,29 @@ f77_create_arrayprint_offset_tbl (struct type *type, struct ui_file *stream)
       upper = f77_get_upperbound (tmp_type);
       lower = f77_get_lowerbound (tmp_type);
 
-      F77_DIM_SIZE (ndimen) = upper - lower + 1;
+      F77_DIM_COUNT (ndimen) = upper - lower + 1;
+
+      F77_DIM_BYTE_STRIDE (ndimen) =
+        TYPE_ARRAY_BYTE_STRIDE_VALUE (tmp_type);
 
       tmp_type = TYPE_TARGET_TYPE (tmp_type);
       ndimen++;
     }
 
-  /* Now we multiply eltlen by all the offsets, so that later we 
+  /* Now we multiply eltlen by all the BYTE_STRIDEs, so that later we
      can print out array elements correctly.  Up till now we 
-     know an offset to apply to get the item but we also 
+     know an eltlen to apply to get the item but we also
      have to know how much to add to get to the next item */
 
   ndimen--;
   eltlen = TYPE_LENGTH (tmp_type);
-  F77_DIM_OFFSET (ndimen) = eltlen;
+  if (F77_DIM_BYTE_STRIDE (ndimen) == 0)
+    F77_DIM_BYTE_STRIDE (ndimen) = eltlen;
   while (--ndimen > 0)
     {
-      eltlen *= F77_DIM_SIZE (ndimen + 1);
-      F77_DIM_OFFSET (ndimen) = eltlen;
+      eltlen *= F77_DIM_COUNT (ndimen + 1);
+      if (F77_DIM_BYTE_STRIDE (ndimen) == 0)
+	F77_DIM_BYTE_STRIDE (ndimen) = eltlen;
     }
 }
 
@@ -172,34 +182,34 @@ f77_print_array_1 (int nss, int ndimensions, struct type *type,
 
   if (nss != ndimensions)
     {
-      for (i = 0; (i < F77_DIM_SIZE (nss) && (*elts) < options->print_max); i++)
+      for (i = 0; (i < F77_DIM_COUNT (nss) && (*elts) < options->print_max); i++)
 	{
 	  fprintf_filtered (stream, "( ");
 	  f77_print_array_1 (nss + 1, ndimensions, TYPE_TARGET_TYPE (type),
-			     valaddr + i * F77_DIM_OFFSET (nss),
-			     address + i * F77_DIM_OFFSET (nss),
+			     valaddr + i * F77_DIM_BYTE_STRIDE (nss),
+			     address + i * F77_DIM_BYTE_STRIDE (nss),
 			     stream, recurse, options, elts);
 	  fprintf_filtered (stream, ") ");
 	}
-      if (*elts >= options->print_max && i < F77_DIM_SIZE (nss)) 
+      if (*elts >= options->print_max && i < F77_DIM_COUNT (nss))
 	fprintf_filtered (stream, "...");
     }
   else
     {
-      for (i = 0; i < F77_DIM_SIZE (nss) && (*elts) < options->print_max;
+      for (i = 0; i < F77_DIM_COUNT (nss) && (*elts) < options->print_max;
 	   i++, (*elts)++)
 	{
 	  val_print (TYPE_TARGET_TYPE (type),
-		     valaddr + i * F77_DIM_OFFSET (ndimensions),
+		     valaddr + i * F77_DIM_BYTE_STRIDE (ndimensions),
 		     0,
-		     address + i * F77_DIM_OFFSET (ndimensions),
+		     address + i * F77_DIM_BYTE_STRIDE (ndimensions),
 		     stream, recurse, options, current_language);
 
-	  if (i != (F77_DIM_SIZE (nss) - 1))
+	  if (i != (F77_DIM_COUNT (nss) - 1))
 	    fprintf_filtered (stream, ", ");
 
 	  if ((*elts == options->print_max - 1)
-	      && (i != (F77_DIM_SIZE (nss) - 1)))
+	      && (i != (F77_DIM_COUNT (nss) - 1)))
 	    fprintf_filtered (stream, "...");
 	}
     }
@@ -251,6 +261,9 @@ f_val_print (struct type *type, const gdb_byte *valaddr, int embedded_offset,
   CORE_ADDR addr;
   int index;
 
+  if (f_object_address_data_valid_print_to_stream (type, stream) != NULL)
+    return 0;
+
   CHECK_TYPEDEF (type);
   switch (TYPE_CODE (type))
     {
diff --git a/gdb/findvar.c b/gdb/findvar.c
index 1048887..b9e7711 100644
--- a/gdb/findvar.c
+++ b/gdb/findvar.c
@@ -35,6 +35,7 @@
 #include "user-regs.h"
 #include "block.h"
 #include "objfiles.h"
+#include "dwarf2loc.h"
 
 /* Basic byte-swapping routines.  GDB has needed these for a long time...
    All extract a target-format integer at ADDR which is LEN bytes long.  */
@@ -382,27 +383,16 @@ symbol_read_needs_frame (struct symbol *sym)
 /* Given a struct symbol for a variable,
    and a stack frame id, read the value of the variable
    and return a (pointer to a) struct value containing the value. 
-   If the variable cannot be found, return a zero pointer.  */
+   If the variable cannot be found, return a zero pointer.
+   We have to first find the address of the variable before allocating struct
+   value to return as its size may depend on DW_OP_PUSH_OBJECT_ADDRESS possibly
+   used by its type.  */
 
 struct value *
 read_var_value (struct symbol *var, struct frame_info *frame)
 {
-  struct value *v;
   struct type *type = SYMBOL_TYPE (var);
   CORE_ADDR addr;
-  int len;
-
-  if (SYMBOL_CLASS (var) == LOC_COMPUTED
-      || SYMBOL_CLASS (var) == LOC_REGISTER)
-    /* These cases do not use V.  */
-    v = NULL;
-  else
-    {
-      v = allocate_value (type);
-      VALUE_LVAL (v) = lval_memory;	/* The most likely possibility.  */
-    }
-
-  len = TYPE_LENGTH (type);
 
   if (symbol_read_needs_frame (var))
     gdb_assert (frame);
@@ -410,31 +400,39 @@ read_var_value (struct symbol *var, struct frame_info *frame)
   switch (SYMBOL_CLASS (var))
     {
     case LOC_CONST:
-      /* Put the constant back in target format.  */
-      store_signed_integer (value_contents_raw (v), len,
-			    (LONGEST) SYMBOL_VALUE (var));
-      VALUE_LVAL (v) = not_lval;
-      return v;
+      {
+	/* Put the constant back in target format.  */
+	struct value *v = allocate_value (type);
+	VALUE_LVAL (v) = not_lval;
+	store_signed_integer (value_contents_raw (v), TYPE_LENGTH (type),
+			      (LONGEST) SYMBOL_VALUE (var));
+	return v;
+      }
 
     case LOC_LABEL:
-      /* Put the constant back in target format.  */
-      if (overlay_debugging)
-	{
-	  CORE_ADDR addr
-	    = symbol_overlayed_address (SYMBOL_VALUE_ADDRESS (var),
-					SYMBOL_OBJ_SECTION (var));
-	  store_typed_address (value_contents_raw (v), type, addr);
-	}
-      else
-	store_typed_address (value_contents_raw (v), type,
-			      SYMBOL_VALUE_ADDRESS (var));
-      VALUE_LVAL (v) = not_lval;
-      return v;
+      {
+	/* Put the constant back in target format.  */
+	struct value *v = allocate_value (type);
+	VALUE_LVAL (v) = not_lval;
+	if (overlay_debugging)
+	  {
+	    CORE_ADDR addr
+	      = symbol_overlayed_address (SYMBOL_VALUE_ADDRESS (var),
+					  SYMBOL_OBJ_SECTION (var));
+	    store_typed_address (value_contents_raw (v), type, addr);
+	  }
+	else
+	  store_typed_address (value_contents_raw (v), type,
+				SYMBOL_VALUE_ADDRESS (var));
+	return v;
+      }
 
     case LOC_CONST_BYTES:
       {
-	memcpy (value_contents_raw (v), SYMBOL_VALUE_BYTES (var), len);
+	struct value *v = allocate_value (type);
 	VALUE_LVAL (v) = not_lval;
+	memcpy (value_contents_raw (v), SYMBOL_VALUE_BYTES (var),
+		TYPE_LENGTH (type));
 	return v;
       }
 
@@ -476,12 +474,23 @@ read_var_value (struct symbol *var, struct frame_info *frame)
       break;
 
     case LOC_BLOCK:
-      if (overlay_debugging)
-	VALUE_ADDRESS (v) = symbol_overlayed_address
-	  (BLOCK_START (SYMBOL_BLOCK_VALUE (var)), SYMBOL_OBJ_SECTION (var));
-      else
-	VALUE_ADDRESS (v) = BLOCK_START (SYMBOL_BLOCK_VALUE (var));
-      return v;
+      {
+        CORE_ADDR addr;
+	struct value *v;
+
+	if (overlay_debugging)
+	  addr = symbol_overlayed_address
+	    (BLOCK_START (SYMBOL_BLOCK_VALUE (var)), SYMBOL_OBJ_SECTION (var));
+	else
+	  addr = BLOCK_START (SYMBOL_BLOCK_VALUE (var));
+	/* ADDR is set here for ALLOCATE_VALUE's CHECK_TYPEDEF for
+	   DW_OP_push_object_address.  */
+	object_address_set (addr);
+	v = allocate_value (type);
+	VALUE_ADDRESS (v) = addr;
+	VALUE_LVAL (v) = lval_memory;
+	return v;
+      }
 
     case LOC_REGISTER:
     case LOC_REGPARM_ADDR:
@@ -499,7 +508,6 @@ read_var_value (struct symbol *var, struct frame_info *frame)
 	      error (_("Value of register variable not available."));
 
 	    addr = value_as_address (regval);
-	    VALUE_LVAL (v) = lval_memory;
 	  }
 	else
 	  {
@@ -542,18 +550,33 @@ read_var_value (struct symbol *var, struct frame_info *frame)
       break;
 
     case LOC_OPTIMIZED_OUT:
-      VALUE_LVAL (v) = not_lval;
-      set_value_optimized_out (v, 1);
-      return v;
+      {
+        struct value *v = allocate_value (type);
+
+	VALUE_LVAL (v) = not_lval;
+	set_value_optimized_out (v, 1);
+	return v;
+      }
 
     default:
       error (_("Cannot look up value of a botched symbol."));
       break;
     }
 
-  VALUE_ADDRESS (v) = addr;
-  set_value_lazy (v, 1);
-  return v;
+  {
+    struct value *v;
+
+    /* ADDR is set here for ALLOCATE_VALUE's CHECK_TYPEDEF for
+       DW_OP_PUSH_OBJECT_ADDRESS.  */
+    object_address_set (addr);
+    v = allocate_value (type);
+    VALUE_ADDRESS (v) = addr;
+    VALUE_LVAL (v) = lval_memory;
+
+    set_value_lazy (v, 1);
+
+    return v;
+  }
 }
 
 /* Install default attributes for register values.  */
@@ -590,10 +613,11 @@ struct value *
 value_from_register (struct type *type, int regnum, struct frame_info *frame)
 {
   struct gdbarch *gdbarch = get_frame_arch (frame);
-  struct type *type1 = check_typedef (type);
   struct value *v;
 
-  if (gdbarch_convert_register_p (gdbarch, regnum, type1))
+  type = check_typedef (type);
+
+  if (gdbarch_convert_register_p (gdbarch, regnum, type))
     {
       /* The ISA/ABI need to something weird when obtaining the
          specified value from this register.  It might need to
@@ -607,7 +631,7 @@ value_from_register (struct type *type, int regnum, struct frame_info *frame)
       VALUE_FRAME_ID (v) = get_frame_id (frame);
       VALUE_REGNUM (v) = regnum;
       gdbarch_register_to_value (gdbarch,
-				 frame, regnum, type1, value_contents_raw (v));
+				 frame, regnum, type, value_contents_raw (v));
     }
   else
     {
diff --git a/gdb/gdbtypes.c b/gdb/gdbtypes.c
index 8102321..e6190cf 100644
--- a/gdb/gdbtypes.c
+++ b/gdb/gdbtypes.c
@@ -38,6 +38,8 @@
 #include "cp-abi.h"
 #include "gdb_assert.h"
 #include "hashtab.h"
+#include "dwarf2expr.h"
+#include "dwarf2loc.h"
 
 /* These variables point to the objects
    representing the predefined C data types.  */
@@ -146,6 +148,23 @@ static void print_bit_vector (B_TYPE *, int);
 static void print_arg_types (struct field *, int, int);
 static void dump_fn_fieldlists (struct type *, int);
 static void print_cplus_stuff (struct type *, int);
+static void type_init_refc (struct type *new_type, struct type *parent_type);
+
+/* A reference count structure for the type reference count map.  Each
+   type in a hierarchy of types is mapped to the same reference
+   count.  */
+struct type_refc_entry
+{
+  /* One type in the hierarchy.  Each type in the hierarchy gets its
+     own slot.  */
+  struct type *type;
+
+  /* A pointer to the shared reference count.  */
+  int *refc;
+};
+
+/* The hash table holding all reference counts.  */
+static htab_t type_refc_table;
 
 
 /* Alloc a new type structure and fill it with some defaults.  If
@@ -154,23 +173,25 @@ static void print_cplus_stuff (struct type *, int);
    structure by xmalloc () (for permanent types).  */
 
 struct type *
-alloc_type (struct objfile *objfile)
+alloc_type (struct objfile *objfile, struct type *parent)
 {
   struct type *type;
 
   /* Alloc the structure and start off with all fields zeroed.  */
 
-  if (objfile == NULL)
+  switch ((long) objfile)
     {
+    case (long) OBJFILE_INTERNAL:
+    case (long) OBJFILE_MALLOC:
       type = XZALLOC (struct type);
       TYPE_MAIN_TYPE (type) = XZALLOC (struct main_type);
-    }
-  else
-    {
+      break;
+    default:
       type = OBSTACK_ZALLOC (&objfile->objfile_obstack, struct type);
       TYPE_MAIN_TYPE (type) = OBSTACK_ZALLOC (&objfile->objfile_obstack,
 					      struct main_type);
       OBJSTAT (objfile, n_types++);
+      break;
     }
 
   /* Initialize the fields that might not be zero.  */
@@ -180,6 +201,9 @@ alloc_type (struct objfile *objfile)
   TYPE_VPTR_FIELDNO (type) = -1;
   TYPE_CHAIN (type) = type;	/* Chain back to itself.  */
 
+  if (objfile == NULL)
+    type_init_refc (type, parent);
+
   return (type);
 }
 
@@ -194,16 +218,24 @@ alloc_type_instance (struct type *oldtype)
 
   /* Allocate the structure.  */
 
-  if (TYPE_OBJFILE (oldtype) == NULL)
-    type = XZALLOC (struct type);
-  else
-    type = OBSTACK_ZALLOC (&TYPE_OBJFILE (oldtype)->objfile_obstack,
-			   struct type);
-
+  switch ((long) TYPE_OBJFILE (oldtype))
+    {
+    case (long) OBJFILE_INTERNAL:
+    case (long) OBJFILE_MALLOC:
+      type = XZALLOC (struct type);
+      break;
+    default:
+      type = OBSTACK_ZALLOC (&TYPE_OBJFILE (oldtype)->objfile_obstack,
+			     struct type);
+      break;
+    }
   TYPE_MAIN_TYPE (type) = TYPE_MAIN_TYPE (oldtype);
 
   TYPE_CHAIN (type) = type;	/* Chain back to itself for now.  */
 
+  if (TYPE_OBJFILE (oldtype) == NULL)
+    type_init_refc (type, oldtype);
+
   return (type);
 }
 
@@ -248,7 +280,7 @@ make_pointer_type (struct type *type, struct type **typeptr)
 
   if (typeptr == 0 || *typeptr == 0)	/* We'll need to allocate one.  */
     {
-      ntype = alloc_type (TYPE_OBJFILE (type));
+      ntype = alloc_type (TYPE_OBJFILE (type), type);
       if (typeptr)
 	*typeptr = ntype;
     }
@@ -260,6 +292,9 @@ make_pointer_type (struct type *type, struct type **typeptr)
       smash_type (ntype);
       TYPE_CHAIN (ntype) = chain;
       TYPE_OBJFILE (ntype) = objfile;
+
+      /* Callers may only supply storage if there is an objfile.  */
+      gdb_assert (objfile);
     }
 
   TYPE_TARGET_TYPE (ntype) = type;
@@ -328,7 +363,7 @@ make_reference_type (struct type *type, struct type **typeptr)
 
   if (typeptr == 0 || *typeptr == 0)	/* We'll need to allocate one.  */
     {
-      ntype = alloc_type (TYPE_OBJFILE (type));
+      ntype = alloc_type (TYPE_OBJFILE (type), type);
       if (typeptr)
 	*typeptr = ntype;
     }
@@ -340,6 +375,9 @@ make_reference_type (struct type *type, struct type **typeptr)
       smash_type (ntype);
       TYPE_CHAIN (ntype) = chain;
       TYPE_OBJFILE (ntype) = objfile;
+
+      /* Callers may only supply storage if there is an objfile.  */
+      gdb_assert (objfile);
     }
 
   TYPE_TARGET_TYPE (ntype) = type;
@@ -388,7 +426,7 @@ make_function_type (struct type *type, struct type **typeptr)
 
   if (typeptr == 0 || *typeptr == 0)	/* We'll need to allocate one.  */
     {
-      ntype = alloc_type (TYPE_OBJFILE (type));
+      ntype = alloc_type (TYPE_OBJFILE (type), type);
       if (typeptr)
 	*typeptr = ntype;
     }
@@ -398,6 +436,9 @@ make_function_type (struct type *type, struct type **typeptr)
       objfile = TYPE_OBJFILE (ntype);
       smash_type (ntype);
       TYPE_OBJFILE (ntype) = objfile;
+
+      /* Callers may only supply storage if there is an objfile.  */
+      gdb_assert (objfile);
     }
 
   TYPE_TARGET_TYPE (ntype) = type;
@@ -643,7 +684,7 @@ lookup_memberptr_type (struct type *type, struct type *domain)
 {
   struct type *mtype;
 
-  mtype = alloc_type (TYPE_OBJFILE (type));
+  mtype = alloc_type (TYPE_OBJFILE (type), NULL);
   smash_to_memberptr_type (mtype, domain, type);
   return (mtype);
 }
@@ -655,7 +696,7 @@ lookup_methodptr_type (struct type *to_type)
 {
   struct type *mtype;
 
-  mtype = alloc_type (TYPE_OBJFILE (to_type));
+  mtype = alloc_type (TYPE_OBJFILE (to_type), NULL);
   TYPE_TARGET_TYPE (mtype) = to_type;
   TYPE_DOMAIN_TYPE (mtype) = TYPE_DOMAIN_TYPE (to_type);
   TYPE_LENGTH (mtype) = cplus_method_ptr_size (to_type);
@@ -696,19 +737,20 @@ create_range_type (struct type *result_type, struct type *index_type,
 		   int low_bound, int high_bound)
 {
   if (result_type == NULL)
-    result_type = alloc_type (TYPE_OBJFILE (index_type));
+    result_type = alloc_type (TYPE_OBJFILE (index_type), index_type);
   TYPE_CODE (result_type) = TYPE_CODE_RANGE;
   TYPE_TARGET_TYPE (result_type) = index_type;
   if (TYPE_STUB (index_type))
     TYPE_TARGET_STUB (result_type) = 1;
   else
     TYPE_LENGTH (result_type) = TYPE_LENGTH (check_typedef (index_type));
-  TYPE_NFIELDS (result_type) = 2;
+  TYPE_NFIELDS (result_type) = 3;
   TYPE_FIELDS (result_type) = TYPE_ZALLOC (result_type,
 					   TYPE_NFIELDS (result_type)
 					   * sizeof (struct field));
   TYPE_LOW_BOUND (result_type) = low_bound;
   TYPE_HIGH_BOUND (result_type) = high_bound;
+  TYPE_BYTE_STRIDE (result_type) = 0;
 
   if (low_bound >= 0)
     TYPE_UNSIGNED (result_type) = 1;
@@ -727,6 +769,9 @@ get_discrete_bounds (struct type *type, LONGEST *lowp, LONGEST *highp)
   switch (TYPE_CODE (type))
     {
     case TYPE_CODE_RANGE:
+      if (TYPE_RANGE_UPPER_BOUND_IS_UNDEFINED (type) 
+	  || TYPE_RANGE_LOWER_BOUND_IS_UNDEFINED (type))
+	return -1;
       *lowp = TYPE_LOW_BOUND (type);
       *highp = TYPE_HIGH_BOUND (type);
       return 1;
@@ -805,30 +850,65 @@ create_array_type (struct type *result_type,
 
   if (result_type == NULL)
     {
-      result_type = alloc_type (TYPE_OBJFILE (range_type));
+      result_type = alloc_type (TYPE_OBJFILE (range_type), range_type);
     }
+  else
+    {
+      /* Callers may only supply storage if there is an objfile.  */
+      gdb_assert (TYPE_OBJFILE (result_type));
+    }
+
   TYPE_CODE (result_type) = TYPE_CODE_ARRAY;
   TYPE_TARGET_TYPE (result_type) = element_type;
-  if (get_discrete_bounds (range_type, &low_bound, &high_bound) < 0)
-    low_bound = high_bound = 0;
-  CHECK_TYPEDEF (element_type);
-  /* Be careful when setting the array length.  Ada arrays can be
-     empty arrays with the high_bound being smaller than the low_bound.
-     In such cases, the array length should be zero.  */
-  if (high_bound < low_bound)
-    TYPE_LENGTH (result_type) = 0;
-  else
-    TYPE_LENGTH (result_type) =
-      TYPE_LENGTH (element_type) * (high_bound - low_bound + 1);
   TYPE_NFIELDS (result_type) = 1;
   TYPE_FIELDS (result_type) =
     (struct field *) TYPE_ZALLOC (result_type, sizeof (struct field));
+  /* FIXME: type alloc.  */
   TYPE_INDEX_TYPE (result_type) = range_type;
   TYPE_VPTR_FIELDNO (result_type) = -1;
 
-  /* TYPE_FLAG_TARGET_STUB will take care of zero length arrays */
+  /* DWARF blocks may depend on runtime information like
+     DW_OP_PUSH_OBJECT_ADDRESS not being available during the
+     CREATE_ARRAY_TYPE time.  */
+  if (TYPE_RANGE_BOUND_IS_DWARF_BLOCK (range_type, 0)
+      || TYPE_RANGE_BOUND_IS_DWARF_BLOCK (range_type, 1)
+      || TYPE_RANGE_UPPER_BOUND_IS_UNDEFINED (range_type) 
+      || TYPE_RANGE_LOWER_BOUND_IS_UNDEFINED (range_type) 
+      || get_discrete_bounds (range_type, &low_bound, &high_bound) < 0)
+    {
+      low_bound = 0;
+      high_bound = -1;
+    }
+
+  /* Be careful when setting the array length.  Ada arrays can be
+     empty arrays with the high_bound being smaller than the low_bound.
+     In such cases, the array length should be zero.  TYPE_TARGET_STUB needs to
+     be checked as it may have dependencies on DWARF blocks depending on
+     runtime information not available during the CREATE_ARRAY_TYPE time.  */
+  if (high_bound < low_bound || TYPE_TARGET_STUB (element_type))
+    TYPE_LENGTH (result_type) = 0;
+  else
+    {
+      CHECK_TYPEDEF (element_type);
+      TYPE_LENGTH (result_type) =
+	TYPE_LENGTH (element_type) * (high_bound - low_bound + 1);
+    }
+
+  if (TYPE_DYNAMIC (range_type))
+    TYPE_DYNAMIC (result_type) = 1;
+
+  /* Multidimensional dynamic arrays need to have all the outer dimensions
+     dynamic to update the outer TYPE_TARGET_TYPE pointer with the new type
+     with statically evaluated dimensions.  */
+  if (TYPE_DYNAMIC (element_type))
+    TYPE_DYNAMIC (result_type) = 1;
+
   if (TYPE_LENGTH (result_type) == 0)
-    TYPE_TARGET_STUB (result_type) = 1;
+    {
+      /* The real size will be computed for specific instances by
+	 CHECK_TYPEDEF.  */
+      TYPE_TARGET_STUB (result_type) = 1;
+    }
 
   return (result_type);
 }
@@ -865,7 +945,12 @@ create_set_type (struct type *result_type, struct type *domain_type)
 {
   if (result_type == NULL)
     {
-      result_type = alloc_type (TYPE_OBJFILE (domain_type));
+      result_type = alloc_type (TYPE_OBJFILE (domain_type), domain_type);
+    }
+  else
+    {
+      /* Callers may only supply storage if there is an objfile.  */
+      gdb_assert (TYPE_OBJFILE (result_type));
     }
   TYPE_CODE (result_type) = TYPE_CODE_SET;
   TYPE_NFIELDS (result_type) = 1;
@@ -1368,6 +1453,65 @@ stub_noname_complaint (void)
   complaint (&symfile_complaints, _("stub type has NULL name"));
 }
 
+/* Calculate the memory length of array TYPE.
+
+   TARGET_TYPE should be set to `check_typedef (TYPE_TARGET_TYPE (type))' as
+   a performance hint.  Feel free to pass NULL.  Set FULL_SPAN to return the
+   size incl. the possible padding of the last element - it may differ from the
+   cleared FULL_SPAN return value (the expected SIZEOF) for non-zero
+   TYPE_BYTE_STRIDE values.  */
+
+static CORE_ADDR
+type_length_get (struct type *type, struct type *target_type, int full_span)
+{
+  struct type *range_type;
+  int count;
+  CORE_ADDR byte_stride = 0;	/* `= 0' for a false GCC warning.  */
+  CORE_ADDR element_size;
+
+  if (TYPE_CODE (type) != TYPE_CODE_ARRAY
+      && TYPE_CODE (type) != TYPE_CODE_STRING)
+    return TYPE_LENGTH (type);
+
+  /* Avoid executing TYPE_HIGH_BOUND for invalid (unallocated/unassociated)
+     Fortran arrays.  The allocated data will never be used so they can be
+     zero-length.  */
+  if (object_address_data_not_valid (type))
+    return 0;
+
+  range_type = TYPE_INDEX_TYPE (type);
+  if (TYPE_RANGE_LOWER_BOUND_IS_UNDEFINED (range_type)
+      || TYPE_RANGE_UPPER_BOUND_IS_UNDEFINED (range_type))
+    return 0;
+  count = TYPE_HIGH_BOUND (range_type) - TYPE_LOW_BOUND (range_type) + 1;
+  /* It may happen for wrong DWARF annotations returning garbage data.  */
+  if (count < 0)
+    warning (_("Range for type %s has invalid bounds %d..%d"),
+	     TYPE_NAME (type), TYPE_LOW_BOUND (range_type),
+	     TYPE_HIGH_BOUND (range_type));
+  /* The code below does not handle count == 0 right.  */
+  if (count <= 0)
+    return 0;
+  if (full_span || count > 1)
+    {
+      /* We do not use TYPE_ARRAY_BYTE_STRIDE_VALUE (type) here as we want to
+         force FULL_SPAN to 1.  */
+      byte_stride = TYPE_BYTE_STRIDE (range_type);
+      if (byte_stride == 0)
+        {
+	  if (target_type == NULL)
+	    target_type = check_typedef (TYPE_TARGET_TYPE (type));
+	  byte_stride = type_length_get (target_type, NULL, 1);
+	}
+    }
+  if (full_span)
+    return count * byte_stride;
+  if (target_type == NULL)
+    target_type = check_typedef (TYPE_TARGET_TYPE (type));
+  element_size = type_length_get (target_type, NULL, 1);
+  return (count - 1) * byte_stride + element_size;
+}
+
 /* Added by Bryan Boreham, Kewill, Sun Sep 17 18:07:17 1989.
 
    If this is a stubbed struct (i.e. declared as struct foo *), see if
@@ -1384,7 +1528,8 @@ stub_noname_complaint (void)
 /* Find the real type of TYPE.  This function returns the real type,
    after removing all layers of typedefs and completing opaque or stub
    types.  Completion changes the TYPE argument, but stripping of
-   typedefs does not.  */
+   typedefs does not.  Still original passed TYPE will have TYPE_LENGTH
+   updated.  FIXME: Remove this dependency (only ada_to_fixed_type?).  */
 
 struct type *
 check_typedef (struct type *type)
@@ -1420,7 +1565,7 @@ check_typedef (struct type *type)
 	  if (sym)
 	    TYPE_TARGET_TYPE (type) = SYMBOL_TYPE (sym);
 	  else					/* TYPE_CODE_UNDEF */
-	    TYPE_TARGET_TYPE (type) = alloc_type (NULL);
+	    TYPE_TARGET_TYPE (type) = alloc_type (NULL, NULL);
 	}
       type = TYPE_TARGET_TYPE (type);
     }
@@ -1494,34 +1639,37 @@ check_typedef (struct type *type)
         }
     }
 
-  if (TYPE_TARGET_STUB (type))
+  /* copy_type_recursive automatically makes the resulting type containing only
+     constant values expected by the callers of this function.  */
+  if (TYPE_DYNAMIC (type))
+    {
+      htab_t copied_types;
+      struct type *type_old = type;
+
+      copied_types = create_copied_types_hash (NULL);
+      type = copy_type_recursive (type, copied_types);
+      htab_delete (copied_types);
+
+      gdb_assert (TYPE_DYNAMIC (type) == 0);
+    }
+
+  if (!currently_reading_symtab
+      && (TYPE_TARGET_STUB (type) || TYPE_DYNAMIC (type)))
     {
-      struct type *range_type;
       struct type *target_type = check_typedef (TYPE_TARGET_TYPE (type));
 
+      if (TYPE_DYNAMIC (type))
+	TYPE_TARGET_TYPE (type) = target_type;
       if (TYPE_STUB (target_type) || TYPE_TARGET_STUB (target_type))
 	{
 	  /* Empty.  */
 	}
       else if (TYPE_CODE (type) == TYPE_CODE_ARRAY
-	       && TYPE_NFIELDS (type) == 1
-	       && (TYPE_CODE (range_type = TYPE_INDEX_TYPE (type))
-		   == TYPE_CODE_RANGE))
+	       || TYPE_CODE (type) == TYPE_CODE_STRING)
 	{
 	  /* Now recompute the length of the array type, based on its
-	     number of elements and the target type's length.
-	     Watch out for Ada null Ada arrays where the high bound
-	     is smaller than the low bound.  */
-	  const int low_bound = TYPE_LOW_BOUND (range_type);
-	  const int high_bound = TYPE_HIGH_BOUND (range_type);
-	  int nb_elements;
-	
-	  if (high_bound < low_bound)
-	    nb_elements = 0;
-	  else
-	    nb_elements = high_bound - low_bound + 1;
-	
-	  TYPE_LENGTH (type) = nb_elements * TYPE_LENGTH (target_type);
+	     number of elements and the target type's length.  */
+	  TYPE_LENGTH (type) = type_length_get (type, target_type, 0);
 	  TYPE_TARGET_STUB (type) = 0;
 	}
       else if (TYPE_CODE (type) == TYPE_CODE_RANGE)
@@ -1529,9 +1677,12 @@ check_typedef (struct type *type)
 	  TYPE_LENGTH (type) = TYPE_LENGTH (target_type);
 	  TYPE_TARGET_STUB (type) = 0;
 	}
+      TYPE_DYNAMIC (type) = 0;
     }
+
   /* Cache TYPE_LENGTH for future use.  */
   TYPE_LENGTH (orig_type) = TYPE_LENGTH (type);
+
   return type;
 }
 
@@ -1753,7 +1904,7 @@ init_type (enum type_code code, int length, int flags,
 {
   struct type *type;
 
-  type = alloc_type (objfile);
+  type = alloc_type (objfile, NULL);
   TYPE_CODE (type) = code;
   TYPE_LENGTH (type) = length;
 
@@ -1783,15 +1934,24 @@ init_type (enum type_code code, int length, int flags,
   if (flags & TYPE_FLAG_FIXED_INSTANCE)
     TYPE_FIXED_INSTANCE (type) = 1;
 
-  if ((name != NULL) && (objfile != NULL))
-    {
-      TYPE_NAME (type) = obsavestring (name, strlen (name), 
-				       &objfile->objfile_obstack);
-    }
-  else
-    {
-      TYPE_NAME (type) = name;
-    }
+  if (name)
+    switch ((long) objfile)
+      {
+      case (long) OBJFILE_INTERNAL:
+	TYPE_NAME (type) = name;
+	break;
+      case (long) OBJFILE_MALLOC:
+	TYPE_NAME (type) = xstrdup (name);
+	break;
+#if 0 /* OBJFILE_MALLOC duplication now.  */
+      case (long) NULL:
+	internal_error (__FILE__, __LINE__,
+			_("OBJFILE pointer NULL should be OBJFILE_* instead"));
+#endif
+      default:
+	TYPE_NAME (type) = obsavestring (name, strlen (name), 
+					 &objfile->objfile_obstack);
+      }
 
   /* C++ fancies.  */
 
@@ -1803,6 +1963,10 @@ init_type (enum type_code code, int length, int flags,
     {
       INIT_CPLUS_SPECIFIC (type);
     }
+
+  if (!objfile)
+    type_incref (type);
+
   return (type);
 }
 
@@ -2916,33 +3080,47 @@ type_pair_eq (const void *item_lhs, const void *item_rhs)
 }
 
 /* Allocate the hash table used by copy_type_recursive to walk
-   types without duplicates.  We use OBJFILE's obstack, because
-   OBJFILE is about to be deleted.  */
+   types without duplicates.   */
 
 htab_t
 create_copied_types_hash (struct objfile *objfile)
 {
-  return htab_create_alloc_ex (1, type_pair_hash, type_pair_eq,
-			       NULL, &objfile->objfile_obstack,
-			       hashtab_obstack_allocate,
-			       dummy_obstack_deallocate);
+  if (objfile == NULL)
+    {
+      /* NULL OBJFILE is for TYPE_DYNAMIC types already contained in
+	 OBJFILE_MALLOC memory, such as those from VALUE_HISTORY_CHAIN.  Table
+	 element entries get allocated by xmalloc - so use xfree.  */
+      return htab_create (1, type_pair_hash, type_pair_eq, xfree);
+    }
+  else
+    {
+      /* Use OBJFILE's obstack, because OBJFILE is about to be deleted.  Table
+	 element entries get allocated by xmalloc - so use xfree.  */
+      return htab_create_alloc_ex (1, type_pair_hash, type_pair_eq,
+				   xfree, &objfile->objfile_obstack,
+				   hashtab_obstack_allocate,
+				   dummy_obstack_deallocate);
+    }
 }
 
-/* Recursively copy (deep copy) TYPE, if it is associated with
-   OBJFILE.  Return a new type allocated using malloc, a saved type if
-   we have already visited TYPE (using COPIED_TYPES), or TYPE if it is
-   not associated with OBJFILE.  */
+/* A helper for copy_type_recursive.  This does all the work.
+   REPRESENTATIVE is a pointer to a type.  This is used to register
+   newly-created types in the type_refc_table.  Initially it pointer
+   to a NULL pointer, but it is filled in the first time a type is
+   copied.  OBJFILE is used only for an assertion checking.  */
 
-struct type *
-copy_type_recursive (struct objfile *objfile, 
-		     struct type *type,
-		     htab_t copied_types)
+static struct type *
+copy_type_recursive_1 (struct objfile *objfile, 
+		       struct type *type,
+		       htab_t copied_types,
+		       struct type **representative)
 {
   struct type_pair *stored, pair;
   void **slot;
   struct type *new_type;
 
-  if (TYPE_OBJFILE (type) == NULL)
+  if (TYPE_OBJFILE (type) == OBJFILE_INTERNAL
+      || (objfile == OBJFILE_MALLOC && !TYPE_DYNAMIC (type)))
     return type;
 
   /* This type shouldn't be pointing to any types in other objfiles;
@@ -2954,11 +3132,15 @@ copy_type_recursive (struct objfile *objfile,
   if (*slot != NULL)
     return ((struct type_pair *) *slot)->new;
 
-  new_type = alloc_type (NULL);
+  new_type = alloc_type (OBJFILE_MALLOC, *representative);
+  if (!*representative)
+    *representative = new_type;
 
   /* We must add the new type to the hash table immediately, in case
-     we encounter this type again during a recursive call below.  */
-  stored = obstack_alloc (&objfile->objfile_obstack, sizeof (struct type_pair));
+     we encounter this type again during a recursive call below.  Memory could
+     be allocated from OBJFILE in the case we will be removing OBJFILE, this
+     optimization is missed and xfree is called for it from COPIED_TYPES.  */
+  stored = xmalloc (sizeof (*stored));
   stored->old = type;
   stored->new = new_type;
   *slot = stored;
@@ -2968,6 +3150,13 @@ copy_type_recursive (struct objfile *objfile,
   *TYPE_MAIN_TYPE (new_type) = *TYPE_MAIN_TYPE (type);
   TYPE_OBJFILE (new_type) = NULL;
 
+  /* Pre-clear the fields processed by delete_main_type.  If DWARF block
+     evaluations below call error we would leave an unfreeable TYPE.  */
+  TYPE_TARGET_TYPE (new_type) = NULL;
+  TYPE_VPTR_BASETYPE (new_type) = NULL;
+  TYPE_NFIELDS (new_type) = 0;
+  TYPE_FIELDS (new_type) = NULL;
+
   if (TYPE_NAME (type))
     TYPE_NAME (new_type) = xstrdup (TYPE_NAME (type));
   if (TYPE_TAG_NAME (type))
@@ -2976,12 +3165,45 @@ copy_type_recursive (struct objfile *objfile,
   TYPE_INSTANCE_FLAGS (new_type) = TYPE_INSTANCE_FLAGS (type);
   TYPE_LENGTH (new_type) = TYPE_LENGTH (type);
 
+  if (TYPE_ALLOCATED (new_type))
+    {
+      gdb_assert (!TYPE_NOT_ALLOCATED (new_type));
+
+      if (!dwarf_locexpr_baton_eval (TYPE_ALLOCATED (new_type)))
+        TYPE_NOT_ALLOCATED (new_type) = 1;
+      TYPE_ALLOCATED (new_type) = NULL;
+    }
+
+  if (TYPE_ASSOCIATED (new_type))
+    {
+      gdb_assert (!TYPE_NOT_ASSOCIATED (new_type));
+
+      if (!dwarf_locexpr_baton_eval (TYPE_ASSOCIATED (new_type)))
+        TYPE_NOT_ASSOCIATED (new_type) = 1;
+      TYPE_ASSOCIATED (new_type) = NULL;
+    }
+
+  if (!TYPE_DATA_LOCATION_IS_ADDR (new_type)
+      && TYPE_DATA_LOCATION_DWARF_BLOCK (new_type))
+    {
+      if (TYPE_NOT_ALLOCATED (new_type)
+          || TYPE_NOT_ASSOCIATED (new_type))
+	TYPE_DATA_LOCATION_DWARF_BLOCK (new_type) = NULL;
+      else
+	{
+	  TYPE_DATA_LOCATION_IS_ADDR (new_type) = 1;
+	  TYPE_DATA_LOCATION_ADDR (new_type) = dwarf_locexpr_baton_eval
+				    (TYPE_DATA_LOCATION_DWARF_BLOCK (new_type));
+	}
+    }
+
   /* Copy the fields.  */
   if (TYPE_NFIELDS (type))
     {
       int i, nfields;
 
       nfields = TYPE_NFIELDS (type);
+      TYPE_NFIELDS (new_type) = nfields;
       TYPE_FIELDS (new_type) = XCALLOC (nfields, struct field);
       for (i = 0; i < nfields; i++)
 	{
@@ -2990,8 +3212,8 @@ copy_type_recursive (struct objfile *objfile,
 	  TYPE_FIELD_BITSIZE (new_type, i) = TYPE_FIELD_BITSIZE (type, i);
 	  if (TYPE_FIELD_TYPE (type, i))
 	    TYPE_FIELD_TYPE (new_type, i)
-	      = copy_type_recursive (objfile, TYPE_FIELD_TYPE (type, i),
-				     copied_types);
+	      = copy_type_recursive_1 (objfile, TYPE_FIELD_TYPE (type, i),
+				       copied_types, representative);
 	  if (TYPE_FIELD_NAME (type, i))
 	    TYPE_FIELD_NAME (new_type, i) = 
 	      xstrdup (TYPE_FIELD_NAME (type, i));
@@ -3010,6 +3232,16 @@ copy_type_recursive (struct objfile *objfile,
 				  xstrdup (TYPE_FIELD_STATIC_PHYSNAME (type,
 								       i)));
 	      break;
+	    case FIELD_LOC_KIND_DWARF_BLOCK:
+	      /* `struct dwarf2_locexpr_baton' is too bound to its objfile so
+		 it is expected to be made constant by CHECK_TYPEDEF.  */
+	      if (TYPE_NOT_ALLOCATED (new_type)
+		  || TYPE_NOT_ASSOCIATED (new_type))
+		SET_FIELD_DWARF_BLOCK (TYPE_FIELD (new_type, i), NULL);
+	      else
+		SET_FIELD_BITPOS (TYPE_FIELD (new_type, i),
+		   dwarf_locexpr_baton_eval (TYPE_FIELD_DWARF_BLOCK (type, i)));
+	      break;
 	    default:
 	      internal_error (__FILE__, __LINE__,
 			      _("Unexpected type field location kind: %d"),
@@ -3018,17 +3250,32 @@ copy_type_recursive (struct objfile *objfile,
 	}
     }
 
+  /* Convert TYPE_RANGE_HIGH_BOUND_IS_COUNT into a regular bound.  */
+  if (TYPE_CODE (type) == TYPE_CODE_RANGE
+      && TYPE_RANGE_HIGH_BOUND_IS_COUNT (type))
+    {
+      TYPE_RANGE_HIGH_BOUND_IS_COUNT (new_type) = 0;
+      TYPE_HIGH_BOUND (new_type) = TYPE_LOW_BOUND (type)
+				   + TYPE_HIGH_BOUND (type) - 1;
+    }
+
+  /* Both FIELD_LOC_KIND_DWARF_BLOCK and TYPE_RANGE_HIGH_BOUND_IS_COUNT were
+     possibly converted.  */
+  TYPE_DYNAMIC (new_type) = 0;
+
   /* Copy pointers to other types.  */
   if (TYPE_TARGET_TYPE (type))
     TYPE_TARGET_TYPE (new_type) = 
-      copy_type_recursive (objfile, 
-			   TYPE_TARGET_TYPE (type),
-			   copied_types);
+      copy_type_recursive_1 (objfile, 
+			     TYPE_TARGET_TYPE (type),
+			     copied_types,
+			     representative);
   if (TYPE_VPTR_BASETYPE (type))
     TYPE_VPTR_BASETYPE (new_type) = 
-      copy_type_recursive (objfile,
-			   TYPE_VPTR_BASETYPE (type),
-			   copied_types);
+      copy_type_recursive_1 (objfile,
+			     TYPE_VPTR_BASETYPE (type),
+			     copied_types,
+			     representative);
   /* Maybe copy the type_specific bits.
 
      NOTE drow/2005-12-09: We do not copy the C++-specific bits like
@@ -3046,6 +3293,20 @@ copy_type_recursive (struct objfile *objfile,
   return new_type;
 }
 
+/* Recursively copy (deep copy) TYPE.  Return a new type allocated using
+   malloc, a saved type if we have already visited TYPE (using COPIED_TYPES),
+   or TYPE if it is not associated with OBJFILE.  */
+
+struct type *
+copy_type_recursive (struct type *type,
+		     htab_t copied_types)
+{
+  struct type *representative = NULL;
+
+  return copy_type_recursive_1 (TYPE_OBJFILE (type), type, copied_types,
+				&representative);
+}
+
 /* Make a copy of the given TYPE, except that the pointer & reference
    types are not preserved.
    
@@ -3059,7 +3320,7 @@ copy_type (const struct type *type)
 
   gdb_assert (TYPE_OBJFILE (type) != NULL);
 
-  new_type = alloc_type (TYPE_OBJFILE (type));
+  new_type = alloc_type (TYPE_OBJFILE (type), NULL);
   TYPE_INSTANCE_FLAGS (new_type) = TYPE_INSTANCE_FLAGS (type);
   TYPE_LENGTH (new_type) = TYPE_LENGTH (type);
   memcpy (TYPE_MAIN_TYPE (new_type), TYPE_MAIN_TYPE (type),
@@ -3068,6 +3329,242 @@ copy_type (const struct type *type)
   return new_type;
 }
 
+static void delete_type (struct type *type);
+
+/* A helper for delete_type which deletes a main_type and the things to which
+   it refers.  TYPE is a type whose main_type we wish to destroy.  */
+
+static void
+delete_main_type (struct main_type *main_type)
+{
+  int i;
+  void **slot;
+  struct
+    {
+      struct main_type *main_type;
+    } type_local = { main_type }, *type = &type_local;
+
+  gdb_assert (TYPE_OBJFILE (type) == OBJFILE_MALLOC);
+
+  xfree (TYPE_NAME (type));
+  xfree (TYPE_TAG_NAME (type));
+
+  for (i = 0; i < TYPE_NFIELDS (type); ++i)
+    {
+      xfree (TYPE_FIELD_NAME (type, i));
+
+      if (TYPE_FIELD_LOC_KIND (type, i) == FIELD_LOC_KIND_PHYSNAME)
+	xfree (TYPE_FIELD_STATIC_PHYSNAME (type, i));
+    }
+  xfree (TYPE_FIELDS (type));
+
+  /* Strangely, HAVE_CPLUS_STRUCT will return true when there isn't
+     one at all.  */
+  gdb_assert (!HAVE_CPLUS_STRUCT (type) || !TYPE_CPLUS_SPECIFIC (type));
+
+  xfree (TYPE_MAIN_TYPE (type));
+}
+
+/* Store `struct main_type *' entries which got `struct type *' deleted.  */
+
+static htab_t deleted_main_types_hash;
+
+/* To be called before any call of delete_type.  */
+
+static void
+delete_type_begin (void)
+{
+  gdb_assert (deleted_main_types_hash == NULL);
+
+  deleted_main_types_hash = htab_create_alloc (10, htab_hash_pointer,
+			    htab_eq_pointer, NULL, xcalloc, xfree);
+}
+
+/* Helper for delete_type_finish.  */
+
+static int
+delete_type_finish_traverse (void **slot, void *unused)
+{
+  struct main_type *main_type = *slot;
+
+  delete_main_type (main_type);
+
+  return 1;
+}
+
+/* To be called after all the calls of delete_type.  Each MAIN_TYPE must have
+   either none or all of its TYPE entries deleted.  */
+
+static void
+delete_type_finish (void)
+{
+  htab_traverse (deleted_main_types_hash, delete_type_finish_traverse, NULL);
+
+  htab_delete (deleted_main_types_hash);
+  deleted_main_types_hash = NULL;
+}
+
+/* Delete TYPE and remember MAIN_TYPE it references.  TYPE must have been
+   allocated using xmalloc -- not using an objfile.  You must wrap calls of
+   this function by delete_type_begin and delete_type_finish.  */
+
+static void
+delete_type (struct type *type)
+{
+  void **slot;
+
+  if (!type)
+    return;
+
+  if (TYPE_OBJFILE (type) == OBJFILE_INTERNAL)
+    return;
+  gdb_assert (TYPE_OBJFILE (type) == OBJFILE_MALLOC);
+
+  slot = htab_find_slot (deleted_main_types_hash, TYPE_MAIN_TYPE (type),
+			 INSERT);
+  gdb_assert (!*slot);
+  *slot = TYPE_MAIN_TYPE (type);
+
+  xfree (type);
+}
+
+/* Hash function for type_refc_table.  */
+
+static hashval_t
+type_refc_hash (const void *p)
+{
+  const struct type_refc_entry *entry = p;
+  return htab_hash_pointer (entry->type);
+}
+
+/* Equality function for type_refc_table.  */
+
+static int
+type_refc_equal (const void *a, const void *b)
+{
+  const struct type_refc_entry *left = a;
+  const struct type_refc_entry *right = b;
+  return left->type == right->type;
+}
+
+/* Insert the new type NEW_TYPE into the table.  Does nothing if
+   NEW_TYPE has an objfile.  If PARENT_TYPE is not NULL, then NEW_TYPE
+   will be inserted into the same hierarchy as PARENT_TYPE.  In this
+   case, PARENT_TYPE must already exist in the reference count map.
+   If PARENT_TYPE is NULL, a new reference count is allocated and set
+   to one.  */
+
+static void
+type_init_refc (struct type *new_type, struct type *parent_type)
+{
+  int *refc;
+  void **slot;
+  struct type_refc_entry *new_entry;
+
+  if (TYPE_OBJFILE (new_type))
+    return;
+
+  if (parent_type)
+    {
+      struct type_refc_entry entry, *found;
+      entry.type = parent_type;
+      found = htab_find (type_refc_table, &entry);
+      gdb_assert (found);
+      refc = found->refc;
+    }
+  else
+    {
+      refc = xmalloc (sizeof (int));
+      *refc = 0;
+    }
+
+  new_entry = XNEW (struct type_refc_entry);
+  new_entry->type = new_type;
+  new_entry->refc = refc;
+
+  slot = htab_find_slot (type_refc_table, new_entry, INSERT);
+  gdb_assert (!*slot);
+  *slot = new_entry;
+}
+
+/* Increment the reference count for TYPE.  */
+
+void
+type_incref (struct type *type)
+{
+  struct type_refc_entry entry, *found;
+
+  if (TYPE_OBJFILE (type))
+    return;
+
+  entry.type = type;
+  found = htab_find (type_refc_table, &entry);
+  gdb_assert (found);
+  ++*(found->refc);
+}
+
+/* A traverse callback for type_refc_table which removes any entry
+   whose reference count pointer is REFC.  REFC may be NULL to delete all the
+   unused entries - use such cleanup only in the GDB idle state as GDB code
+   does not necessarily reference county TYPEs during its processing.  */
+
+static int
+type_refc_remove (void **slot, void *refc)
+{
+  struct type_refc_entry *entry = *slot;
+
+  if (entry->refc == refc || (refc == NULL && *entry->refc == 0))
+    {
+      delete_type (entry->type);
+
+      xfree (entry);
+      htab_clear_slot (type_refc_table, slot);
+    }
+
+  return 1;
+}
+
+/* Decrement the reference count for TYPE.  If TYPE has no more
+   references, delete it.  */
+
+void
+type_decref (struct type *type)
+{
+  struct type_refc_entry entry, *found;
+
+  if (TYPE_OBJFILE (type))
+    return;
+
+  entry.type = type;
+  found = htab_find (type_refc_table, &entry);
+  gdb_assert (found);
+  --*(found->refc);
+  if (*(found->refc) == 0)
+    {
+      void *refc = found->refc;
+
+      /* Clear all table entries referring to this count.  CHECK: Should not be
+         the deletion delayed till free_all_types?  */
+      delete_type_begin ();
+      htab_traverse (type_refc_table, type_refc_remove, refc);
+      delete_type_finish ();
+
+      /* Delete the reference count itself.  */
+      xfree (refc);
+    }
+}
+
+/* Free all the types that have been allocated (except for those released).
+   Called after each command, successful or not.  */
+
+void
+free_all_types (void)
+{
+  delete_type_begin ();
+  htab_traverse (type_refc_table, type_refc_remove, NULL);
+  delete_type_finish ();
+}
+
 static struct type *
 build_flt (int bit, char *name, const struct floatformat **floatformats)
 {
@@ -3105,7 +3602,7 @@ build_complex (int bit, char *name, struct type *target_type)
       return builtin_type_error;
     }
   t = init_type (TYPE_CODE_COMPLEX, 2 * bit / TARGET_CHAR_BIT,
-		 0, name, (struct objfile *) NULL);
+		 0, name, OBJFILE_INTERNAL);
   TYPE_TARGET_TYPE (t) = target_type;
   return t;
 }
@@ -3119,56 +3616,56 @@ gdbtypes_post_init (struct gdbarch *gdbarch)
   builtin_type->builtin_void =
     init_type (TYPE_CODE_VOID, 1,
 	       0,
-	       "void", (struct objfile *) NULL);
+	       "void", OBJFILE_INTERNAL);
   builtin_type->builtin_char =
     init_type (TYPE_CODE_INT, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
 	       (TYPE_FLAG_NOSIGN
                 | (gdbarch_char_signed (gdbarch) ? 0 : TYPE_FLAG_UNSIGNED)),
-	       "char", (struct objfile *) NULL);
+	       "char", OBJFILE_INTERNAL);
   builtin_type->builtin_signed_char =
     init_type (TYPE_CODE_INT, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
 	       0,
-	       "signed char", (struct objfile *) NULL);
+	       "signed char", OBJFILE_INTERNAL);
   builtin_type->builtin_unsigned_char =
     init_type (TYPE_CODE_INT, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
 	       TYPE_FLAG_UNSIGNED,
-	       "unsigned char", (struct objfile *) NULL);
+	       "unsigned char", OBJFILE_INTERNAL);
   builtin_type->builtin_short =
     init_type (TYPE_CODE_INT, 
 	       gdbarch_short_bit (gdbarch) / TARGET_CHAR_BIT,
-	       0, "short", (struct objfile *) NULL);
+	       0, "short", OBJFILE_INTERNAL);
   builtin_type->builtin_unsigned_short =
     init_type (TYPE_CODE_INT, 
 	       gdbarch_short_bit (gdbarch) / TARGET_CHAR_BIT,
 	       TYPE_FLAG_UNSIGNED, "unsigned short", 
-	       (struct objfile *) NULL);
+	       OBJFILE_INTERNAL);
   builtin_type->builtin_int =
     init_type (TYPE_CODE_INT, 
 	       gdbarch_int_bit (gdbarch) / TARGET_CHAR_BIT,
-	       0, "int", (struct objfile *) NULL);
+	       0, "int", OBJFILE_INTERNAL);
   builtin_type->builtin_unsigned_int =
     init_type (TYPE_CODE_INT, 
 	       gdbarch_int_bit (gdbarch) / TARGET_CHAR_BIT,
 	       TYPE_FLAG_UNSIGNED, "unsigned int", 
-	       (struct objfile *) NULL);
+	       OBJFILE_INTERNAL);
   builtin_type->builtin_long =
     init_type (TYPE_CODE_INT, 
 	       gdbarch_long_bit (gdbarch) / TARGET_CHAR_BIT,
-	       0, "long", (struct objfile *) NULL);
+	       0, "long", OBJFILE_INTERNAL);
   builtin_type->builtin_unsigned_long =
     init_type (TYPE_CODE_INT, 
 	       gdbarch_long_bit (gdbarch) / TARGET_CHAR_BIT,
 	       TYPE_FLAG_UNSIGNED, "unsigned long", 
-	       (struct objfile *) NULL);
+	       OBJFILE_INTERNAL);
   builtin_type->builtin_long_long =
     init_type (TYPE_CODE_INT,
 	       gdbarch_long_long_bit (gdbarch) / TARGET_CHAR_BIT,
-	       0, "long long", (struct objfile *) NULL);
+	       0, "long long", OBJFILE_INTERNAL);
   builtin_type->builtin_unsigned_long_long =
     init_type (TYPE_CODE_INT,
 	       gdbarch_long_long_bit (gdbarch) / TARGET_CHAR_BIT,
 	       TYPE_FLAG_UNSIGNED, "unsigned long long", 
-	       (struct objfile *) NULL);
+	       OBJFILE_INTERNAL);
   builtin_type->builtin_float
     = build_flt (gdbarch_float_bit (gdbarch), "float",
 		 gdbarch_float_format (gdbarch));
@@ -3187,26 +3684,26 @@ gdbtypes_post_init (struct gdbarch *gdbarch)
   builtin_type->builtin_string =
     init_type (TYPE_CODE_STRING, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
 	       0,
-	       "string", (struct objfile *) NULL);
+	       "string", OBJFILE_INTERNAL);
   builtin_type->builtin_bool =
     init_type (TYPE_CODE_BOOL, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
 	       0,
-	       "bool", (struct objfile *) NULL);
+	       "bool", OBJFILE_INTERNAL);
 
   /* The following three are about decimal floating point types, which
      are 32-bits, 64-bits and 128-bits respectively.  */
   builtin_type->builtin_decfloat
     = init_type (TYPE_CODE_DECFLOAT, 32 / 8,
 	        0,
-	       "_Decimal32", (struct objfile *) NULL);
+	       "_Decimal32", OBJFILE_INTERNAL);
   builtin_type->builtin_decdouble
     = init_type (TYPE_CODE_DECFLOAT, 64 / 8,
 	       0,
-	       "_Decimal64", (struct objfile *) NULL);
+	       "_Decimal64", OBJFILE_INTERNAL);
   builtin_type->builtin_declong
     = init_type (TYPE_CODE_DECFLOAT, 128 / 8,
 	       0,
-	       "_Decimal128", (struct objfile *) NULL);
+	       "_Decimal128", OBJFILE_INTERNAL);
 
   /* Pointer/Address types.  */
 
@@ -3245,27 +3742,28 @@ gdbtypes_post_init (struct gdbarch *gdbarch)
     init_type (TYPE_CODE_INT, 
 	       gdbarch_addr_bit (gdbarch) / 8,
 	       TYPE_FLAG_UNSIGNED,
-	       "__CORE_ADDR", (struct objfile *) NULL);
+	       "__CORE_ADDR", OBJFILE_INTERNAL);
 
 
   /* The following set of types is used for symbols with no
      debug information.  */
   builtin_type->nodebug_text_symbol =
     init_type (TYPE_CODE_FUNC, 1, 0, 
-	       "<text variable, no debug info>", NULL);
+	       "<text variable, no debug info>", OBJFILE_INTERNAL);
   TYPE_TARGET_TYPE (builtin_type->nodebug_text_symbol) =
     builtin_type->builtin_int;
   builtin_type->nodebug_data_symbol =
     init_type (TYPE_CODE_INT, 
 	       gdbarch_int_bit (gdbarch) / HOST_CHAR_BIT, 0,
-	       "<data variable, no debug info>", NULL);
+	       "<data variable, no debug info>", OBJFILE_INTERNAL);
   builtin_type->nodebug_unknown_symbol =
     init_type (TYPE_CODE_INT, 1, 0,
-	       "<variable (not text or data), no debug info>", NULL);
+	       "<variable (not text or data), no debug info>",
+	       OBJFILE_INTERNAL);
   builtin_type->nodebug_tls_symbol =
     init_type (TYPE_CODE_INT, 
 	       gdbarch_int_bit (gdbarch) / HOST_CHAR_BIT, 0,
-	       "<thread local variable, no debug info>", NULL);
+	       "<thread local variable, no debug info>", OBJFILE_INTERNAL);
 
   return builtin_type;
 }
@@ -3276,6 +3774,9 @@ _initialize_gdbtypes (void)
 {
   gdbtypes_data = gdbarch_data_register_post_init (gdbtypes_post_init);
 
+  type_refc_table = htab_create_alloc (20, type_refc_hash, type_refc_equal,
+				       NULL, xcalloc, xfree);
+
   /* FIXME: The following types are architecture-neutral.  However,
      they contain pointer_type and reference_type fields potentially
      caching pointer or reference types that *are* architecture
@@ -3284,47 +3785,47 @@ _initialize_gdbtypes (void)
   builtin_type_int0 =
     init_type (TYPE_CODE_INT, 0 / 8,
 	       0,
-	       "int0_t", (struct objfile *) NULL);
+	       "int0_t", OBJFILE_INTERNAL);
   builtin_type_int8 =
     init_type (TYPE_CODE_INT, 8 / 8,
 	       TYPE_FLAG_NOTTEXT,
-	       "int8_t", (struct objfile *) NULL);
+	       "int8_t", OBJFILE_INTERNAL);
   builtin_type_uint8 =
     init_type (TYPE_CODE_INT, 8 / 8,
 	       TYPE_FLAG_UNSIGNED | TYPE_FLAG_NOTTEXT,
-	       "uint8_t", (struct objfile *) NULL);
+	       "uint8_t", OBJFILE_INTERNAL);
   builtin_type_int16 =
     init_type (TYPE_CODE_INT, 16 / 8,
 	       0,
-	       "int16_t", (struct objfile *) NULL);
+	       "int16_t", OBJFILE_INTERNAL);
   builtin_type_uint16 =
     init_type (TYPE_CODE_INT, 16 / 8,
 	       TYPE_FLAG_UNSIGNED,
-	       "uint16_t", (struct objfile *) NULL);
+	       "uint16_t", OBJFILE_INTERNAL);
   builtin_type_int32 =
     init_type (TYPE_CODE_INT, 32 / 8,
 	       0,
-	       "int32_t", (struct objfile *) NULL);
+	       "int32_t", OBJFILE_INTERNAL);
   builtin_type_uint32 =
     init_type (TYPE_CODE_INT, 32 / 8,
 	       TYPE_FLAG_UNSIGNED,
-	       "uint32_t", (struct objfile *) NULL);
+	       "uint32_t", OBJFILE_INTERNAL);
   builtin_type_int64 =
     init_type (TYPE_CODE_INT, 64 / 8,
 	       0,
-	       "int64_t", (struct objfile *) NULL);
+	       "int64_t", OBJFILE_INTERNAL);
   builtin_type_uint64 =
     init_type (TYPE_CODE_INT, 64 / 8,
 	       TYPE_FLAG_UNSIGNED,
-	       "uint64_t", (struct objfile *) NULL);
+	       "uint64_t", OBJFILE_INTERNAL);
   builtin_type_int128 =
     init_type (TYPE_CODE_INT, 128 / 8,
 	       0,
-	       "int128_t", (struct objfile *) NULL);
+	       "int128_t", OBJFILE_INTERNAL);
   builtin_type_uint128 =
     init_type (TYPE_CODE_INT, 128 / 8,
 	       TYPE_FLAG_UNSIGNED,
-	       "uint128_t", (struct objfile *) NULL);
+	       "uint128_t", OBJFILE_INTERNAL);
 
   builtin_type_ieee_single =
     build_flt (-1, "builtin_type_ieee_single", floatformats_ieee_single);
@@ -3344,15 +3845,15 @@ _initialize_gdbtypes (void)
   builtin_type_void =
     init_type (TYPE_CODE_VOID, 1,
 	       0,
-	       "void", (struct objfile *) NULL);
+	       "void", OBJFILE_INTERNAL);
   builtin_type_true_char =
     init_type (TYPE_CODE_CHAR, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
 	       0,
-	       "true character", (struct objfile *) NULL);
+	       "true character", OBJFILE_INTERNAL);
   builtin_type_true_unsigned_char =
     init_type (TYPE_CODE_CHAR, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
 	       TYPE_FLAG_UNSIGNED,
-	       "true character", (struct objfile *) NULL);
+	       "true character", OBJFILE_INTERNAL);
 
   add_setshow_zinteger_cmd ("overload", no_class, &overload_debug, _("\
 Set debugging of C++ overloading."), _("\
diff --git a/gdb/gdbtypes.h b/gdb/gdbtypes.h
index c90b6d7..86df022 100644
--- a/gdb/gdbtypes.h
+++ b/gdb/gdbtypes.h
@@ -209,6 +209,11 @@ enum type_instance_flag_value
 
 #define TYPE_TARGET_STUB(t)	(TYPE_MAIN_TYPE (t)->flag_target_stub)
 
+/* Type needs to be evaluated on each CHECK_TYPEDEF and its results must not be
+   sticky.  */
+
+#define TYPE_DYNAMIC(t)		(TYPE_MAIN_TYPE (t)->flag_dynamic)
+
 /* Static type.  If this is set, the corresponding type had 
  * a static modifier.
  * Note: This may be unnecessary, since static data members
@@ -266,6 +271,36 @@ enum type_instance_flag_value
 
 #define TYPE_NOTTEXT(t)		(TYPE_MAIN_TYPE (t)->flag_nottext)
 
+/* Is HIGH_BOUND a low-bound relative count (1) or the high bound itself (0)?  */
+
+#define TYPE_RANGE_HIGH_BOUND_IS_COUNT(range_type) \
+  (TYPE_MAIN_TYPE (range_type)->flag_range_high_bound_is_count)
+
+/* Not allocated.  TYPE_ALLOCATED(t) must be NULL in such case.  If this flag
+   is unset and TYPE_ALLOCATED(t) is NULL then the type is allocated.  If this
+   flag is unset and TYPE_ALLOCATED(t) is not NULL then its DWARF block
+   determines the actual allocation state.  */
+
+#define TYPE_NOT_ALLOCATED(t)	(TYPE_MAIN_TYPE (t)->flag_not_allocated)
+
+/* Not associated.  TYPE_ASSOCIATED(t) must be NULL in such case.  If this flag
+   is unset and TYPE_ASSOCIATED(t) is NULL then the type is associated.  If
+   this flag is unset and TYPE_ASSOCIATED(t) is not NULL then its DWARF block
+   determines the actual association state.  */
+
+#define TYPE_NOT_ASSOCIATED(t)	(TYPE_MAIN_TYPE (t)->flag_not_associated)
+
+/* Address of the actual data as for DW_AT_data_location.  Its dwarf block must
+   not be evaluated unless both TYPE_NOT_ALLOCATED and TYPE_NOT_ASSOCIATED are
+   false.  If TYPE_DATA_LOCATION_IS_ADDR set then TYPE_DATA_LOCATION_ADDR value
+   is the actual data address value.  If unset and
+   TYPE_DATA_LOCATION_DWARF_BLOCK is NULL then the value is the normal
+   VALUE_ADDRESS copy.  If unset and TYPE_DATA_LOCATION_DWARF_BLOCK is not NULL
+   then its DWARF block determines the actual data address.  */
+
+#define TYPE_DATA_LOCATION_IS_ADDR(t) \
+  (TYPE_MAIN_TYPE (t)->flag_data_location_is_addr)
+
 /* Constant type.  If this is set, the corresponding type has a
  * const modifier.
  */
@@ -352,6 +387,11 @@ struct main_type
   unsigned int flag_stub_supported : 1;
   unsigned int flag_nottext : 1;
   unsigned int flag_fixed_instance : 1;
+  unsigned int flag_dynamic : 1;
+  unsigned int flag_range_high_bound_is_count : 1;
+  unsigned int flag_not_allocated : 1;
+  unsigned int flag_not_associated : 1;
+  unsigned int flag_data_location_is_addr : 1;
 
   /* Number of fields described for this type.  This field appears at
      this location because it packs nicely here.  */
@@ -414,6 +454,20 @@ struct main_type
 
   struct type *target_type;
 
+  /* For DW_AT_data_location.  */
+  union
+    {
+      struct dwarf2_locexpr_baton *dwarf_block;
+      CORE_ADDR addr;
+    }
+  data_location;
+
+  /* For DW_AT_allocated.  */
+  struct dwarf2_locexpr_baton *allocated;
+
+  /* For DW_AT_associated.  */
+  struct dwarf2_locexpr_baton *associated;
+
   /* For structure and union types, a description of each field.
      For set and pascal array types, there is one "field",
      whose type is the domain type of the set or array.
@@ -795,9 +849,9 @@ extern void allocate_cplus_struct_type (struct type *);
 #define TYPE_POINTER_TYPE(thistype) (thistype)->pointer_type
 #define TYPE_REFERENCE_TYPE(thistype) (thistype)->reference_type
 #define TYPE_CHAIN(thistype) (thistype)->chain
-/* Note that if thistype is a TYPEDEF type, you have to call check_typedef.
-   But check_typedef does set the TYPE_LENGTH of the TYPEDEF type,
-   so you only have to call check_typedef once.  Since allocate_value
+/* Note that if thistype is a TYPEDEF, ARRAY or STRING type, you have to call
+   check_typedef.  But check_typedef does set the TYPE_LENGTH of the TYPEDEF
+   type, so you only have to call check_typedef once.  Since allocate_value
    calls check_typedef, TYPE_LENGTH (VALUE_TYPE (X)) is safe.  */
 #define TYPE_LENGTH(thistype) (thistype)->length
 #define TYPE_OBJFILE(thistype) TYPE_MAIN_TYPE(thistype)->objfile
@@ -807,23 +861,44 @@ extern void allocate_cplus_struct_type (struct type *);
 #define TYPE_NFIELDS(thistype) TYPE_MAIN_TYPE(thistype)->nfields
 #define TYPE_FIELDS(thistype) TYPE_MAIN_TYPE(thistype)->fields
 #define TYPE_TEMPLATE_ARGS(thistype) TYPE_CPLUS_SPECIFIC(thistype)->template_args
+#define TYPE_DATA_LOCATION_DWARF_BLOCK(thistype) TYPE_MAIN_TYPE (thistype)->data_location.dwarf_block
+#define TYPE_DATA_LOCATION_ADDR(thistype) TYPE_MAIN_TYPE (thistype)->data_location.addr
+#define TYPE_ALLOCATED(thistype) TYPE_MAIN_TYPE (thistype)->allocated
+#define TYPE_ASSOCIATED(thistype) TYPE_MAIN_TYPE (thistype)->associated
 
 #define TYPE_INDEX_TYPE(type) TYPE_FIELD_TYPE (type, 0)
 #define TYPE_LOW_BOUND(range_type) TYPE_FIELD_BITPOS (range_type, 0)
 #define TYPE_HIGH_BOUND(range_type) TYPE_FIELD_BITPOS (range_type, 1)
-
-/* Moto-specific stuff for FORTRAN arrays */
-
-#define TYPE_ARRAY_UPPER_BOUND_IS_UNDEFINED(arraytype) \
-   (TYPE_FIELD_ARTIFICIAL(TYPE_INDEX_TYPE((arraytype)),1))
+#define TYPE_BYTE_STRIDE(range_type) TYPE_FIELD_BITPOS (range_type, 2)
+
+/* Whether we should use TYPE_FIELD_DWARF_BLOCK (and not TYPE_FIELD_BITPOS).  */
+#define TYPE_RANGE_BOUND_IS_DWARF_BLOCK(range_type, fieldno) \
+  (TYPE_FIELD_LOC_KIND (range_type, fieldno) == FIELD_LOC_KIND_DWARF_BLOCK)
+#define TYPE_RANGE_BOUND_SET_DWARF_BLOCK(range_type, fieldno) \
+  (TYPE_FIELD_LOC_KIND (range_type, fieldno) = FIELD_LOC_KIND_DWARF_BLOCK)
+#define TYPE_ARRAY_BOUND_IS_DWARF_BLOCK(array_type, fieldno) \
+  TYPE_RANGE_BOUND_IS_DWARF_BLOCK (TYPE_INDEX_TYPE (array_type), fieldno)
+
+/* Unbound arrays, such as GCC array[]; at end of struct.  */
+#define TYPE_RANGE_LOWER_BOUND_IS_UNDEFINED(rangetype) \
+   TYPE_FIELD_ARTIFICIAL((rangetype),0)
+#define TYPE_RANGE_UPPER_BOUND_IS_UNDEFINED(rangetype) \
+   TYPE_FIELD_ARTIFICIAL((rangetype),1)
 #define TYPE_ARRAY_LOWER_BOUND_IS_UNDEFINED(arraytype) \
-   (TYPE_FIELD_ARTIFICIAL(TYPE_INDEX_TYPE((arraytype)),0))
-
-#define TYPE_ARRAY_UPPER_BOUND_VALUE(arraytype) \
-   (TYPE_HIGH_BOUND(TYPE_INDEX_TYPE((arraytype))))
+   TYPE_RANGE_LOWER_BOUND_IS_UNDEFINED (TYPE_INDEX_TYPE (arraytype))
+#define TYPE_ARRAY_UPPER_BOUND_IS_UNDEFINED(arraytype) \
+   TYPE_RANGE_UPPER_BOUND_IS_UNDEFINED (TYPE_INDEX_TYPE (arraytype))
 
 #define TYPE_ARRAY_LOWER_BOUND_VALUE(arraytype) \
-   (TYPE_LOW_BOUND(TYPE_INDEX_TYPE((arraytype))))
+  TYPE_LOW_BOUND (TYPE_INDEX_TYPE (arraytype))
+#define TYPE_ARRAY_UPPER_BOUND_VALUE(arraytype) \
+  TYPE_HIGH_BOUND (TYPE_INDEX_TYPE (arraytype))
+/* TYPE_BYTE_STRIDE (TYPE_INDEX_TYPE (arraytype)) with a fallback to the
+   element size if no specific stride value is known.  */
+#define TYPE_ARRAY_BYTE_STRIDE_VALUE(arraytype)		\
+  (TYPE_BYTE_STRIDE (TYPE_INDEX_TYPE (arraytype)) == 0	\
+   ? TYPE_LENGTH (TYPE_TARGET_TYPE (arraytype))		\
+   : TYPE_BYTE_STRIDE (TYPE_INDEX_TYPE (arraytype)))
 
 /* C++ */
 
@@ -1078,6 +1153,16 @@ extern struct type *builtin_type_error;
    (TYPE_UNSIGNED(t) ? UMIN_OF_SIZE(TYPE_LENGTH(t)) \
     : MIN_OF_SIZE(TYPE_LENGTH(t)))
 
+/* Virtual OBJFILE used for builtin types.  */
+#define OBJFILE_INTERNAL ((struct objfile *) 1L)
+
+/* Virtual OBJFILE used for types allocated by malloc.  FIXME: Currently
+   backward compatible with the old NULL value; fix then also init_type.  */
+#define OBJFILE_MALLOC ((struct objfile *) 0L)
+
+#define OBJFILE_IS_VIRTUAL(objfile) ((objfile) == OBJFILE_INTERNAL	\
+				     || (objfile) == OBJFILE_MALLOC)
+
 /* Allocate space for storing data associated with a particular type.
    We ensure that the space is allocated using the same mechanism that
    was used to allocate the space for the type structure itself.  I.E.
@@ -1087,18 +1172,18 @@ extern struct type *builtin_type_error;
    builtin types), then the data space will be allocated with xmalloc,
    the same as for the type structure. */
 
-#define TYPE_ALLOC(t,size)  \
-   (TYPE_OBJFILE (t) != NULL  \
-    ? obstack_alloc (&TYPE_OBJFILE (t) -> objfile_obstack, size) \
-    : xmalloc (size))
+#define TYPE_ALLOC(t,size)						\
+   (OBJFILE_IS_VIRTUAL (TYPE_OBJFILE (t))				\
+    ? xmalloc (size)							\
+    : obstack_alloc (&TYPE_OBJFILE (t) -> objfile_obstack, size))
 
-#define TYPE_ZALLOC(t,size)  \
-   (TYPE_OBJFILE (t) != NULL  \
-    ? memset (obstack_alloc (&TYPE_OBJFILE (t)->objfile_obstack, size),  \
-	      0, size)  \
-    : xzalloc (size))
+#define TYPE_ZALLOC(t,size)						\
+   (OBJFILE_IS_VIRTUAL (TYPE_OBJFILE (t))				\
+    ? xzalloc (size)							\
+    : memset (obstack_alloc (&TYPE_OBJFILE (t)->objfile_obstack, size),	\
+	      0, size))
 
-extern struct type *alloc_type (struct objfile *);
+extern struct type *alloc_type (struct objfile *, struct type *);
 
 extern struct type *init_type (enum type_code, int, int, char *,
 			       struct objfile *);
@@ -1172,6 +1257,16 @@ extern struct type *create_range_type (struct type *, struct type *, int,
 extern struct type *create_array_type (struct type *, struct type *,
 				       struct type *);
 
+extern CORE_ADDR type_range_any_field_internal (struct type *range_type,
+						int fieldno);
+
+extern int type_range_high_bound_internal (struct type *range_type);
+
+extern int type_range_count_bound_internal (struct type *range_type);
+
+extern CORE_ADDR type_range_byte_stride_internal (struct type *range_type,
+						  struct type *element_type);
+
 extern struct type *create_string_type (struct type *, struct type *);
 
 extern struct type *create_set_type (struct type *, struct type *);
@@ -1263,10 +1358,15 @@ extern void maintenance_print_type (char *, int);
 
 extern htab_t create_copied_types_hash (struct objfile *objfile);
 
-extern struct type *copy_type_recursive (struct objfile *objfile,
-					 struct type *type,
+extern struct type *copy_type_recursive (struct type *type,
 					 htab_t copied_types);
 
 extern struct type *copy_type (const struct type *type);
 
+extern void type_incref (struct type *type);
+
+extern void type_decref (struct type *type);
+
+extern void free_all_types (void);
+
 #endif /* GDBTYPES_H */
diff --git a/gdb/jv-lang.c b/gdb/jv-lang.c
index 0728866..cc82cb4 100644
--- a/gdb/jv-lang.c
+++ b/gdb/jv-lang.c
@@ -302,7 +302,7 @@ type_from_class (struct value *clas)
   if (type != NULL)
     return type;
 
-  type = alloc_type (objfile);
+  type = alloc_type (objfile, NULL);
   TYPE_CODE (type) = TYPE_CODE_STRUCT;
   INIT_CPLUS_SPECIFIC (type);
 
diff --git a/gdb/mdebugread.c b/gdb/mdebugread.c
index 7cbcc59..e507c3b 100644
--- a/gdb/mdebugread.c
+++ b/gdb/mdebugread.c
@@ -4696,7 +4696,7 @@ new_type (char *name)
 {
   struct type *t;
 
-  t = alloc_type (current_objfile);
+  t = alloc_type (current_objfile, NULL);
   TYPE_NAME (t) = name;
   TYPE_CPLUS_SPECIFIC (t) = (struct cplus_struct_type *) &cplus_struct_default;
   return t;
diff --git a/gdb/mi/mi-main.c b/gdb/mi/mi-main.c
index b905a9e..83c6b48 100644
--- a/gdb/mi/mi-main.c
+++ b/gdb/mi/mi-main.c
@@ -1317,6 +1317,7 @@ mi_cmd_execute (struct mi_parse *parse)
   struct cleanup *cleanup;
   int i;
   free_all_values ();
+  free_all_types ();
 
   current_token = xstrdup (parse->token);
   cleanup = make_cleanup (free_current_contents, &current_token);
diff --git a/gdb/printcmd.c b/gdb/printcmd.c
index a51ba68..55a39f5 100644
--- a/gdb/printcmd.c
+++ b/gdb/printcmd.c
@@ -878,6 +878,11 @@ print_command_1 (char *exp, int inspect, int voidprint)
   else
     val = access_value_history (0);
 
+  /* Do not try to OBJECT_ADDRESS_SET here anything.  We are interested in the
+     source variable base addresses as found by READ_VAR_VALUE.  The value here
+     can be already a calculated expression address inappropriate for
+     DW_OP_push_object_address.  */
+
   if (voidprint || (val && value_type (val) &&
 		    TYPE_CODE (value_type (val)) != TYPE_CODE_VOID))
     {
diff --git a/gdb/stabsread.c b/gdb/stabsread.c
index 3457784..976a5a7 100644
--- a/gdb/stabsread.c
+++ b/gdb/stabsread.c
@@ -322,7 +322,7 @@ dbx_alloc_type (int typenums[2], struct objfile *objfile)
 
   if (typenums[0] == -1)
     {
-      return (alloc_type (objfile));
+      return (alloc_type (objfile, NULL));
     }
 
   type_addr = dbx_lookup_type (typenums);
@@ -332,7 +332,7 @@ dbx_alloc_type (int typenums[2], struct objfile *objfile)
      We will fill it in later if we find out how.  */
   if (*type_addr == 0)
     {
-      *type_addr = alloc_type (objfile);
+      *type_addr = alloc_type (objfile, NULL);
     }
 
   return (*type_addr);
diff --git a/gdb/testsuite/gdb.arch/x86_64-vla-typedef-foo.S b/gdb/testsuite/gdb.arch/x86_64-vla-typedef-foo.S
new file mode 100644
index 0000000..66f7a39
--- /dev/null
+++ b/gdb/testsuite/gdb.arch/x86_64-vla-typedef-foo.S
@@ -0,0 +1,455 @@
+	.file	"x86_64-vla-typedef.c"
+	.section	.debug_abbrev,"",@progbits
+.Ldebug_abbrev0:
+	.section	.debug_info,"",@progbits
+.Ldebug_info0:
+	.section	.debug_line,"",@progbits
+.Ldebug_line0:
+	.text
+.Ltext0:
+.globl foo
+	.type	foo, @function
+foo:
+.LFB2:
+	.file 1 "x86_64-vla-typedef.c"
+	.loc 1 22 0
+	pushq	%rbp
+.LCFI0:
+	movq	%rsp, %rbp
+.LCFI1:
+	subq	$64, %rsp
+.LCFI2:
+	movl	%edi, -36(%rbp)
+	.loc 1 22 0
+	movq	%rsp, %rax
+	movq	%rax, -48(%rbp)
+	.loc 1 23 0
+	movl	-36(%rbp), %edx
+	movslq	%edx,%rax
+	subq	$1, %rax
+	movq	%rax, -24(%rbp)
+	.loc 1 24 0
+	movslq	%edx,%rax
+	addq	$15, %rax
+	addq	$15, %rax
+	shrq	$4, %rax
+	salq	$4, %rax
+	subq	%rax, %rsp
+	movq	%rsp, -56(%rbp)
+	movq	-56(%rbp), %rax
+	addq	$15, %rax
+	shrq	$4, %rax
+	salq	$4, %rax
+	movq	%rax, -56(%rbp)
+	movq	-56(%rbp), %rax
+	movq	%rax, -16(%rbp)
+	.loc 1 27 0
+	movl	$0, -4(%rbp)
+	jmp	.L2
+.L3:
+	.loc 1 28 0
+	movl	-4(%rbp), %esi
+	movl	-4(%rbp), %eax
+	movl	%eax, %ecx
+	movq	-16(%rbp), %rdx
+	movslq	%esi,%rax
+	movb	%cl, (%rdx,%rax)
+	.loc 1 27 0
+	addl	$1, -4(%rbp)
+.L2:
+	movl	-4(%rbp), %eax
+	cmpl	-36(%rbp), %eax
+	jl	.L3
+	.loc 1 30 0
+	.globl	break_here
+break_here:
+	movq	-16(%rbp), %rax
+	movb	$0, (%rax)
+	movq	-48(%rbp), %rsp
+	.loc 1 31 0
+	leave
+	ret
+.LFE2:
+	.size	foo, .-foo
+	.section	.debug_frame,"",@progbits
+.Lframe0:
+	.long	.LECIE0-.LSCIE0
+.LSCIE0:
+	.long	0xffffffff
+	.byte	0x1
+	.string	""
+	.uleb128 0x1
+	.sleb128 -8
+	.byte	0x10
+	.byte	0xc
+	.uleb128 0x7
+	.uleb128 0x8
+	.byte	0x90
+	.uleb128 0x1
+	.align 8
+.LECIE0:
+.LSFDE0:
+	.long	.LEFDE0-.LASFDE0
+.LASFDE0:
+	.long	.Lframe0
+	.quad	.LFB2
+	.quad	.LFE2-.LFB2
+	.byte	0x4
+	.long	.LCFI0-.LFB2
+	.byte	0xe
+	.uleb128 0x10
+	.byte	0x86
+	.uleb128 0x2
+	.byte	0x4
+	.long	.LCFI1-.LCFI0
+	.byte	0xd
+	.uleb128 0x6
+	.align 8
+.LEFDE0:
+	.section	.eh_frame,"a",@progbits
+.Lframe1:
+	.long	.LECIE1-.LSCIE1
+.LSCIE1:
+	.long	0x0
+	.byte	0x1
+	.string	"zR"
+	.uleb128 0x1
+	.sleb128 -8
+	.byte	0x10
+	.uleb128 0x1
+	.byte	0x3
+	.byte	0xc
+	.uleb128 0x7
+	.uleb128 0x8
+	.byte	0x90
+	.uleb128 0x1
+	.align 8
+.LECIE1:
+.LSFDE1:
+	.long	.LEFDE1-.LASFDE1
+.LASFDE1:
+	.long	.LASFDE1-.Lframe1
+	.long	.LFB2
+	.long	.LFE2-.LFB2
+	.uleb128 0x0
+	.byte	0x4
+	.long	.LCFI0-.LFB2
+	.byte	0xe
+	.uleb128 0x10
+	.byte	0x86
+	.uleb128 0x2
+	.byte	0x4
+	.long	.LCFI1-.LCFI0
+	.byte	0xd
+	.uleb128 0x6
+	.align 8
+.LEFDE1:
+	.text
+.Letext0:
+	.section	.debug_loc,"",@progbits
+.Ldebug_loc0:
+.LLST0:
+	.quad	.LFB2-.Ltext0
+	.quad	.LCFI0-.Ltext0
+	.value	0x2
+	.byte	0x77
+	.sleb128 8
+	.quad	.LCFI0-.Ltext0
+	.quad	.LCFI1-.Ltext0
+	.value	0x2
+	.byte	0x77
+	.sleb128 16
+	.quad	.LCFI1-.Ltext0
+	.quad	.LFE2-.Ltext0
+	.value	0x2
+	.byte	0x76
+	.sleb128 16
+	.quad	0x0
+	.quad	0x0
+	.section	.debug_info
+	.long	.Ldebug_end - .Ldebug_start
+.Ldebug_start:
+	.value	0x2
+	.long	.Ldebug_abbrev0
+	.byte	0x8
+	.uleb128 0x1
+	.long	.LASF2
+	.byte	0x1
+	.long	.LASF3
+	.long	.LASF4
+	.quad	.Ltext0
+	.quad	.Letext0
+	.long	.Ldebug_line0
+	.uleb128 0x2
+	.byte	0x1
+	.string	"foo"
+	.byte	0x1
+	.byte	0x16
+	.byte	0x1
+	.quad	.LFB2
+	.quad	.LFE2
+	.long	.LLST0
+	.long	0x83
+	.uleb128 0x3
+	.long	.LASF5
+	.byte	0x1
+	.byte	0x15
+	.long	0x83
+	.byte	0x2
+	.byte	0x91
+	.sleb128 -52
+.Ltag_typedef:
+	.uleb128 0x4
+	.long	.LASF6
+	.byte	0x1
+	.byte	0x17
+	.long	.Ltag_array_type - .debug_info
+	.uleb128 0x5	/* Abbrev Number: 5 (DW_TAG_variable) */
+	.long	.LASF0
+	.byte	0x1
+	.byte	0x18
+#if 1
+	.long	.Ltag_typedef - .debug_info
+#else
+	/* Debugging only: Skip the typedef indirection.  */
+	.long	.Ltag_array_type - .debug_info
+#endif
+	/* DW_AT_location: DW_FORM_block1: start */
+	.byte	0x3
+	.byte	0x91
+	.sleb128 -32
+#if 0
+	.byte	0x6	/* DW_OP_deref */
+#else
+	.byte	0x96	/* DW_OP_nop */
+#endif
+	/* DW_AT_location: DW_FORM_block1: end */
+	.uleb128 0x6
+	.string	"i"
+	.byte	0x1
+	.byte	0x19
+	.long	0x83
+	.byte	0x2
+	.byte	0x91
+	.sleb128 -20
+	.byte	0x0
+	.uleb128 0x7
+	.byte	0x4
+	.byte	0x5
+	.string	"int"
+.Ltag_array_type:
+	.uleb128 0x8	/* Abbrev Number: 8 (DW_TAG_array_type) */
+	.long	0xa0 + (2f - 1f)	/* DW_AT_type: DW_FORM_ref4 */
+	.long	0x9d + (2f - 1f)	/* DW_AT_sibling: DW_FORM_ref4 */
+1:	/* DW_AT_data_location: DW_FORM_block1: start */
+	.byte	2f - 3f	/* length */
+3:
+	.byte	0x97	/* DW_OP_push_object_address */
+	.byte	0x6	/* DW_OP_deref */
+2:	/* DW_AT_data_location: DW_FORM_block1: end */
+	.uleb128 0x9
+	.long	0x9d + (2b - 1b)	/* DW_AT_type: DW_FORM_ref4 */
+	.byte	0x3
+	.byte	0x91
+	.sleb128 -40
+	.byte	0x6
+	.byte	0x0
+	.uleb128 0xa
+	.byte	0x8
+	.byte	0x7
+	.uleb128 0xb
+	.byte	0x1
+	.byte	0x6
+	.long	.LASF1
+	.byte	0x0
+.Ldebug_end:
+	.section	.debug_abbrev
+	.uleb128 0x1
+	.uleb128 0x11
+	.byte	0x1
+	.uleb128 0x25
+	.uleb128 0xe
+	.uleb128 0x13
+	.uleb128 0xb
+	.uleb128 0x3
+	.uleb128 0xe
+	.uleb128 0x1b
+	.uleb128 0xe
+	.uleb128 0x11
+	.uleb128 0x1
+	.uleb128 0x12
+	.uleb128 0x1
+	.uleb128 0x10
+	.uleb128 0x6
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0x2
+	.uleb128 0x2e
+	.byte	0x1
+	.uleb128 0x3f
+	.uleb128 0xc
+	.uleb128 0x3
+	.uleb128 0x8
+	.uleb128 0x3a
+	.uleb128 0xb
+	.uleb128 0x3b
+	.uleb128 0xb
+	.uleb128 0x27
+	.uleb128 0xc
+	.uleb128 0x11
+	.uleb128 0x1
+	.uleb128 0x12
+	.uleb128 0x1
+	.uleb128 0x40
+	.uleb128 0x6
+	.uleb128 0x1
+	.uleb128 0x13
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0x3
+	.uleb128 0x5
+	.byte	0x0
+	.uleb128 0x3
+	.uleb128 0xe
+	.uleb128 0x3a
+	.uleb128 0xb
+	.uleb128 0x3b
+	.uleb128 0xb
+	.uleb128 0x49
+	.uleb128 0x13
+	.uleb128 0x2
+	.uleb128 0xa
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0x4
+	.uleb128 0x16
+	.byte	0x0
+	.uleb128 0x3
+	.uleb128 0xe
+	.uleb128 0x3a
+	.uleb128 0xb
+	.uleb128 0x3b
+	.uleb128 0xb
+	.uleb128 0x49
+	.uleb128 0x13
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0x5
+	.uleb128 0x34
+	.byte	0x0
+	.uleb128 0x3
+	.uleb128 0xe
+	.uleb128 0x3a
+	.uleb128 0xb
+	.uleb128 0x3b
+	.uleb128 0xb
+	.uleb128 0x49
+	.uleb128 0x13
+	.uleb128 0x2
+	.uleb128 0xa
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0x6
+	.uleb128 0x34
+	.byte	0x0
+	.uleb128 0x3
+	.uleb128 0x8
+	.uleb128 0x3a
+	.uleb128 0xb
+	.uleb128 0x3b
+	.uleb128 0xb
+	.uleb128 0x49
+	.uleb128 0x13
+	.uleb128 0x2
+	.uleb128 0xa
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0x7
+	.uleb128 0x24
+	.byte	0x0
+	.uleb128 0xb
+	.uleb128 0xb
+	.uleb128 0x3e
+	.uleb128 0xb
+	.uleb128 0x3
+	.uleb128 0x8
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0x8	/* Abbrev Number: 8 (DW_TAG_array_type) */
+	.uleb128 0x1
+	.byte	0x1
+	.uleb128 0x49	/* DW_AT_type */
+	.uleb128 0x13	/* DW_FORM_ref4 */
+	.uleb128 0x1	/* DW_AT_sibling */
+	.uleb128 0x13	/* DW_FORM_ref4 */
+	.uleb128 0x50	/* DW_AT_data_location */
+	.uleb128 0xa	/* DW_FORM_block1 */
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0x9
+	.uleb128 0x21
+	.byte	0x0
+	.uleb128 0x49	/* DW_AT_type */
+	.uleb128 0x13	/* DW_FORM_ref4 */
+	.uleb128 0x2f
+	.uleb128 0xa
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0xa
+	.uleb128 0x24
+	.byte	0x0
+	.uleb128 0xb
+	.uleb128 0xb
+	.uleb128 0x3e
+	.uleb128 0xb
+	.byte	0x0
+	.byte	0x0
+	.uleb128 0xb
+	.uleb128 0x24
+	.byte	0x0
+	.uleb128 0xb
+	.uleb128 0xb
+	.uleb128 0x3e
+	.uleb128 0xb
+	.uleb128 0x3
+	.uleb128 0xe
+	.byte	0x0
+	.byte	0x0
+	.byte	0x0
+	.section	.debug_pubnames,"",@progbits
+	.long	0x16
+	.value	0x2
+	.long	.Ldebug_info0
+	.long	0xa8
+	.long	0x2d
+	.string	"foo"
+	.long	0x0
+	.section	.debug_aranges,"",@progbits
+	.long	0x2c
+	.value	0x2
+	.long	.Ldebug_info0
+	.byte	0x8
+	.byte	0x0
+	.value	0x0
+	.value	0x0
+	.quad	.Ltext0
+	.quad	.Letext0-.Ltext0
+	.quad	0x0
+	.quad	0x0
+	.section	.debug_str,"MS",@progbits,1
+.LASF0:
+	.string	"array"
+.LASF5:
+	.string	"size"
+.LASF3:
+	.string	"x86_64-vla-typedef.c"
+.LASF6:
+	.string	"array_t"
+.LASF1:
+	.string	"char"
+.LASF4:
+	.string	"gdb.arch"
+.LASF2:
+	.string	"GNU C 4.3.2 20081105 (Red Hat 4.3.2-7)"
+	.ident	"GCC: (GNU) 4.3.2 20081105 (Red Hat 4.3.2-7)"
+	.section	.note.GNU-stack,"",@progbits
diff --git a/gdb/testsuite/gdb.arch/x86_64-vla-typedef.c b/gdb/testsuite/gdb.arch/x86_64-vla-typedef.c
new file mode 100644
index 0000000..b809c4e
--- /dev/null
+++ b/gdb/testsuite/gdb.arch/x86_64-vla-typedef.c
@@ -0,0 +1,43 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2008 Free Software Foundation, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 3 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
+
+#if 0
+
+void
+foo (int size)
+{
+  typedef char array_t[size];
+  array_t array;
+  int i;
+
+  for (i = 0; i < size; i++)
+    array[i] = i;
+
+  array[0] = 0;	/* break-here */
+}
+
+#else
+
+int
+main (void)
+{
+  foo (26);
+  foo (78);
+  return 0;
+}
+
+#endif
diff --git a/gdb/testsuite/gdb.arch/x86_64-vla-typedef.exp b/gdb/testsuite/gdb.arch/x86_64-vla-typedef.exp
new file mode 100644
index 0000000..b05411e
--- /dev/null
+++ b/gdb/testsuite/gdb.arch/x86_64-vla-typedef.exp
@@ -0,0 +1,64 @@
+# Copyright 2009 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+# Test DW_AT_data_location accessed through DW_TAG_typedef intermediate.
+
+if ![istarget "x86_64-*-*"] then {
+    verbose "Skipping over gdb.arch/x86_64-vla-typedef.exp test made only for x86_64."
+    return
+}
+
+set testfile x86_64-vla-typedef
+set srcasmfile ${testfile}-foo.S
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}
+set binobjfile ${objdir}/${subdir}/${testfile}-foo.o
+if  { [gdb_compile "${srcdir}/${subdir}/${srcasmfile}" "${binobjfile}" object {}] != "" } {
+    untested "Couldn't compile test program"
+    return -1
+}
+if  { [gdb_compile "${srcdir}/${subdir}/${srcfile} ${binobjfile}" "${binfile}" executable {debug}] != "" } {
+    untested "Couldn't compile test program"
+    return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+if ![runto_main] {
+    untested x86_64-vla-typedef
+    return -1
+}
+
+gdb_breakpoint "break_here"
+
+gdb_continue_to_breakpoint "break_here"
+
+gdb_test "whatis array" "type = array_t" "first: whatis array"
+
+gdb_test "ptype array" "type = char \\\[26\\\]" "first: ptype array"
+
+gdb_test "p array\[1\]" "\\$\[0-9\] = 1 '\\\\001'"
+gdb_test "p array\[2\]" "\\$\[0-9\] = 2 '\\\\002'"
+gdb_test "p array\[3\]" "\\$\[0-9\] = 3 '\\\\003'"
+gdb_test "p array\[4\]" "\\$\[0-9\] = 4 '\\\\004'"
+
+gdb_continue_to_breakpoint "break_here"
+
+gdb_test "whatis array" "type = array_t" "second: whatis array"
+
+gdb_test "ptype array" "type = char \\\[78\\\]" "second: ptype array"
diff --git a/gdb/testsuite/gdb.base/vla-overflow.c b/gdb/testsuite/gdb.base/vla-overflow.c
new file mode 100644
index 0000000..c5d5ee0
--- /dev/null
+++ b/gdb/testsuite/gdb.base/vla-overflow.c
@@ -0,0 +1,30 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2008 Free Software Foundation, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 3 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
+
+#include <stdlib.h>
+
+int
+main (int argc, char **argv)
+{
+  int array[argc];
+
+  array[0] = array[0];
+
+  abort ();
+
+  return 0;
+}
diff --git a/gdb/testsuite/gdb.base/vla-overflow.exp b/gdb/testsuite/gdb.base/vla-overflow.exp
new file mode 100644
index 0000000..7203a48
--- /dev/null
+++ b/gdb/testsuite/gdb.base/vla-overflow.exp
@@ -0,0 +1,108 @@
+# Copyright 2008 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+# We could crash in:
+# #0  block_linkage_function (bl=0x0) at ../../gdb/block.c:69
+# #1  in dwarf_block_get_frame_base (...) at ../../gdb/dwarf2block.c:97
+#   97	  framefunc = block_linkage_function (get_frame_block (frame, NULL));
+# #2  in execute_stack_op (...) at ../../gdb/dwarf2expr.c:496
+# #3  in dwarf_block_exec_core () at ../../gdb/dwarf2block.c:156
+# #4  dwarf_block_exec (...) at ../../gdb/dwarf2block.c:206
+# #5  in range_type_count_bound_internal (...) at ../../gdb/gdbtypes.c:1430
+# #6  in create_array_type (...) at ../../gdb/gdbtypes.c:840
+# ...
+# #21 in psymtab_to_symtab (...) at ../../gdb/symfile.c:292
+# ...
+# #29 in backtrace_command_1 () at ../../gdb/stack.c:1273
+
+set testfile vla-overflow
+set shfile ${objdir}/${subdir}/${testfile}-gdb.sh
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}
+if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
+    untested "Couldn't compile test program"
+    return -1
+}
+
+set f [open "|getconf PAGESIZE" "r"]
+gets $f pagesize
+close $f
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+set pid_of_gdb [exp_pid -i [board_info host fileid]]
+
+if { [runto_main] < 0 } {
+    untested vla-overflow
+    return -1
+}
+
+# Get the GDB memory size when we stay at main.
+
+proc memory_v_pages_get {} {
+    global pid_of_gdb pagesize
+    set fd [open "/proc/$pid_of_gdb/statm"]
+    gets $fd line
+    close $fd
+    # number of pages of virtual memory
+    scan $line "%d" drs
+    return $drs
+}
+
+set pages_found [memory_v_pages_get]
+
+set mb_reserve 10
+verbose -log "pages_found = $pages_found, mb_reserve = $mb_reserve"
+set kb_found [expr $pages_found * $pagesize / 1024]
+set kb_permit [expr $kb_found + 1 * 1024 + $mb_reserve * 1024]
+verbose -log "kb_found = $kb_found, kb_permit = $kb_permit"
+
+# Create the ulimit wrapper.
+set f [open $shfile "w"]
+puts $f "#! /bin/sh"
+puts $f "ulimit -v $kb_permit"
+puts $f "exec $GDB \"\$@\""
+close $f
+remote_exec host "chmod +x $shfile"
+
+gdb_exit
+set GDBold $GDB
+set GDB "$shfile"
+gdb_start
+set GDB $GDBold
+
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+set pid_of_gdb [exp_pid -i [board_info host fileid]]
+
+# Check the size again after the second run.
+# We must not stop in main as it would cache `array' and never crash later.
+
+gdb_run_cmd
+
+verbose -log "kb_found before abort() = [expr [memory_v_pages_get] * $pagesize / 1024]"
+
+gdb_test "" "Program received signal SIGABRT, Aborted..*" "Enter abort()"
+
+verbose -log "kb_found in abort() = [expr [memory_v_pages_get] * $pagesize / 1024]"
+
+# `abort' can get expressed as `*__GI_abort'.
+gdb_test "bt" "in \[^ \]*abort \\(.* in main \\(.*" "Backtrace after abort()"
+
+verbose -log "kb_found in bt after abort() = [expr [memory_v_pages_get] * $pagesize / 1024]"
diff --git a/gdb/testsuite/gdb.base/vla.c b/gdb/testsuite/gdb.base/vla.c
new file mode 100644
index 0000000..e1f3ed1
--- /dev/null
+++ b/gdb/testsuite/gdb.base/vla.c
@@ -0,0 +1,55 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2008 Free Software Foundation, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 3 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
+
+#include <string.h>
+
+void
+marker (void)
+{
+}
+
+void
+bar (char *a, char *b, char *c, int size)
+{
+  memset (a, '1', size);
+  memset (b, '2', size);
+  memset (c, '3', 48);
+}
+
+void
+foo (int size)
+{
+  char temp1[size];
+  char temp3[48];
+
+  temp1[size - 1] = '\0';
+  {
+    char temp2[size];
+
+    bar (temp1, temp2, temp3, size);
+
+    marker ();	/* break-here */
+  }
+}
+
+int
+main (void)
+{
+  foo (26);
+  foo (78);
+  return 0;
+}
diff --git a/gdb/testsuite/gdb.base/vla.exp b/gdb/testsuite/gdb.base/vla.exp
new file mode 100644
index 0000000..5da7378
--- /dev/null
+++ b/gdb/testsuite/gdb.base/vla.exp
@@ -0,0 +1,62 @@
+# Copyright 2008 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+set testfile vla
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}
+if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
+    untested "Couldn't compile test program"
+    return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+if ![runto_main] {
+    untested vla
+    return -1
+}
+
+gdb_breakpoint [gdb_get_line_number "break-here"]
+
+gdb_continue_to_breakpoint "break-here"
+
+gdb_test "whatis temp1" "type = char \\\[variable\\\]" "first: whatis temp1"
+gdb_test "whatis temp2" "type = char \\\[variable\\\]" "first: whatis temp2"
+gdb_test "whatis temp3" "type = char \\\[48\\\]" "first: whatis temp3"
+
+gdb_test "ptype temp1" "type = char \\\[26\\\]" "first: ptype temp1"
+gdb_test "ptype temp2" "type = char \\\[26\\\]" "first: ptype temp2"
+gdb_test "ptype temp3" "type = char \\\[48\\\]" "first: ptype temp3"
+
+gdb_test "p temp1" " = '1' <repeats 26 times>" "first: print temp1"
+gdb_test "p temp2" " = '2' <repeats 26 times>" "first: print temp2"
+gdb_test "p temp3" " = '3' <repeats 48 times>" "first: print temp3"
+
+gdb_continue_to_breakpoint "break-here"
+
+gdb_test "whatis temp1" "type = char \\\[variable\\\]" "second: whatis temp1"
+gdb_test "whatis temp2" "type = char \\\[variable\\\]" "second: whatis temp2"
+gdb_test "whatis temp3" "type = char \\\[48\\\]" "second: whatis temp3"
+
+gdb_test "ptype temp1" "type = char \\\[78\\\]" "second: ptype temp1"
+gdb_test "ptype temp2" "type = char \\\[78\\\]" "second: ptype temp2"
+gdb_test "ptype temp3" "type = char \\\[48\\\]" "second: ptype temp3"
+
+gdb_test "p temp1" " = '1' <repeats 78 times>" "second: print temp1"
+gdb_test "p temp2" " = '2' <repeats 78 times>" "second: print temp2"
+gdb_test "p temp3" " = '3' <repeats 48 times>" "second: print temp3"
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-stripped.c b/gdb/testsuite/gdb.dwarf2/dw2-stripped.c
new file mode 100644
index 0000000..1f02d90
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-stripped.c
@@ -0,0 +1,42 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2004 Free Software Foundation, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+ 
+   You should have received a copy of the GNU General Public License
+   along with this program; if not, write to the Free Software
+   Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
+   USA.  */
+
+
+/* The function `func1' traced into must have debug info on offset > 0;
+   (DW_UNSND (attr)).  This is the reason of `func0' existence.  */
+
+void
+func0(int a, int b)
+{
+}
+
+/* `func1' being traced into must have some arguments to dump.  */
+
+void
+func1(int a, int b)
+{
+  func0 (a,b);
+}
+
+int
+main(void)
+{
+  func1 (1, 2);
+  return 0;
+}
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-stripped.exp b/gdb/testsuite/gdb.dwarf2/dw2-stripped.exp
new file mode 100644
index 0000000..1c6e84a
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-stripped.exp
@@ -0,0 +1,79 @@
+# Copyright 2006 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+# 
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+# 
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+
+# Minimal DWARF-2 unit test
+
+# This test can only be run on targets which support DWARF-2.
+# For now pick a sampling of likely targets.
+if {![istarget *-*-linux*]
+    && ![istarget *-*-gnu*]
+    && ![istarget *-*-elf*]
+    && ![istarget *-*-openbsd*]
+    && ![istarget arm-*-eabi*]
+    && ![istarget powerpc-*-eabi*]} {
+    return 0  
+}
+
+set testfile "dw2-stripped"
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}.x
+
+remote_exec build "rm -f ${binfile}"
+
+# get the value of gcc_compiled
+if [get_compiler_info ${binfile}] {
+    return -1
+}
+
+# This test can only be run on gcc as we use additional_flags=FIXME
+if {$gcc_compiled == 0} {
+    return 0
+}
+
+if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug additional_flags=-ggdb3}] != "" } {
+    return -1
+}
+
+remote_exec build "objcopy -R .debug_loc ${binfile}"
+set strip_output [remote_exec build "objdump -h ${binfile}"]
+
+set test "stripping test file preservation"
+if [ regexp ".debug_info " $strip_output]  {
+    pass "$test (.debug_info preserved)"
+} else {
+    fail "$test (.debug_info got also stripped)"
+}
+
+set test "stripping test file functionality"
+if [ regexp ".debug_loc " $strip_output]  {
+    fail "$test (.debug_loc still present)"
+} else {
+    pass "$test (.debug_loc stripped)"
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+# For C programs, "start" should stop in main().
+
+gdb_test "start" \
+         ".*main \\(\\) at .*" \
+         "start"
+gdb_test "step" \
+         "func.* \\(.*\\) at .*" \
+         "step"
diff --git a/gdb/testsuite/gdb.fortran/dynamic.exp b/gdb/testsuite/gdb.fortran/dynamic.exp
new file mode 100644
index 0000000..77a1203
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/dynamic.exp
@@ -0,0 +1,156 @@
+# Copyright 2007 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+# 
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+# 
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  
+
+# This file was written by Jan Kratochvil <jan.kratochvil@redhat.com>.
+
+# This file is part of the gdb testsuite.  It contains tests for dynamically
+# allocated Fortran arrays.
+# It depends on the GCC dynamic Fortran arrays DWARF support:
+# 	http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22244
+
+set testfile "dynamic"
+set srcfile ${testfile}.f90
+set binfile ${objdir}/${subdir}/${testfile}
+
+if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug f77 quiet}] != "" } {
+    untested "Couldn't compile ${srcfile}"
+    return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+if ![runto MAIN__] then {
+    perror "couldn't run to breakpoint MAIN__"
+    continue
+}
+
+gdb_breakpoint [gdb_get_line_number "varx-init"]
+gdb_continue_to_breakpoint "varx-init"
+gdb_test "p varx" "\\$\[0-9\]* = <(object|the array) is not allocated>" "p varx unallocated"
+gdb_test "ptype varx" "type = <(object|the array) is not allocated>" "ptype varx unallocated"
+gdb_test "p varx(1,5,17)" "(Cannot access it|Unable to access the object) because the (object|array) is not allocated\\." "p varx(1,5,17) unallocated"
+gdb_test "p varx(1,5,17)=1" "(Cannot access it|Unable to access the object) because the (object|array) is not allocated\\." "p varx(1,5,17)=1 unallocated"
+gdb_test "ptype varx(1,5,17)" "(Cannot access it|Unable to access the object) because the (object|array) is not allocated\\." "ptype varx(1,5,17) unallocated"
+
+gdb_breakpoint [gdb_get_line_number "varx-allocated"]
+gdb_continue_to_breakpoint "varx-allocated"
+# $1 = (( ( 0, 0, 0, 0, 0, 0) ( 0, 0, 0, 0, 0, 0) --- , 0) ) ( ( 0, 0, ...) ...) ...)
+gdb_test "ptype varx" "type = real(\\(kind=4\\)|\\*4) \\(6,5:15,17:28\\)" "ptype varx allocated"
+# Intel Fortran Compiler 10.1.008 uses -1 there, GCC uses 1.
+gdb_test "p l" "\\$\[0-9\]* = (\\.TRUE\\.|4294967295)" "p l if varx allocated"
+
+gdb_breakpoint [gdb_get_line_number "varx-filled"]
+gdb_continue_to_breakpoint "varx-filled"
+gdb_test "p varx(2, 5, 17)" "\\$\[0-9\]* = 6"
+gdb_test "p varx(1, 5, 17)" "\\$\[0-9\]* = 7"
+gdb_test "p varx(2, 6, 18)" "\\$\[0-9\]* = 8"
+gdb_test "p varx(6, 15, 28)" "\\$\[0-9\]* = 9"
+# The latter one is for the Intel Fortran Compiler 10.1.008 pointer type.
+gdb_test "p varv" "\\$\[0-9\]* = (<(object|the array) is not associated>|.*(Cannot access it|Unable to access the object) because the object is not associated.)" "p varv unassociated"
+gdb_test "ptype varv" "type = (<(object|the array) is not associated>|.*(Cannot access it|Unable to access the object) because the object is not associated.)" "ptype varv unassociated"
+
+gdb_breakpoint [gdb_get_line_number "varv-associated"]
+gdb_continue_to_breakpoint "varv-associated"
+gdb_test "p varx(3, 7, 19)" "\\$\[0-9\]* = 6" "p varx(3, 7, 19) with varv associated"
+gdb_test "p varv(3, 7, 19)" "\\$\[0-9\]* = 6" "p varv(3, 7, 19) associated"
+# Intel Fortran Compiler 10.1.008 uses -1 there, GCC uses 1.
+gdb_test "p l" "\\$\[0-9\]* = (\\.TRUE\\.|4294967295)" "p l if varv associated"
+gdb_test "ptype varx" "type = real(\\(kind=4\\)|\\*4) \\(6,5:15,17:28\\)" "ptype varx with varv associated"
+# Intel Fortran Compiler 10.1.008 uses the pointer type.
+gdb_test "ptype varv" "type = (PTR TO -> \\( )?real(\\(kind=4\\)|\\*4) \\(6,5:15,17:28\\)\\)?" "ptype varv associated"
+
+gdb_breakpoint [gdb_get_line_number "varv-filled"]
+gdb_continue_to_breakpoint "varv-filled"
+gdb_test "p varx(3, 7, 19)" "\\$\[0-9\]* = 10" "p varx(3, 7, 19) with varv filled"
+gdb_test "p varv(3, 7, 19)" "\\$\[0-9\]* = 10" "p varv(3, 7, 19) filled"
+
+gdb_breakpoint [gdb_get_line_number "varv-deassociated"]
+gdb_continue_to_breakpoint "varv-deassociated"
+# The latter one is for the Intel Fortran Compiler 10.1.008 pointer type.
+gdb_test "p varv" "\\$\[0-9\]* = (<(object|the array) is not associated>|.*(Cannot access it|Unable to access the object) because the object is not associated.)" "p varv deassociated"
+gdb_test "ptype varv" "type = (<(object|the array) is not associated>|.*(Cannot access it|Unable to access the object) because the object is not associated.)" "ptype varv deassociated"
+gdb_test "p l" "\\$\[0-9\]* = \\.FALSE\\." "p l if varv deassociated"
+gdb_test "p varv(1,5,17)" "(Cannot access it|Unable to access the object) because the (object|array) is not associated\\."
+gdb_test "ptype varv(1,5,17)" "(Cannot access it|Unable to access the object) because the (object|array) is not associated\\."
+
+gdb_breakpoint [gdb_get_line_number "varx-deallocated"]
+gdb_continue_to_breakpoint "varx-deallocated"
+gdb_test "p varx" "\\$\[0-9\]* = <(object|the array) is not allocated>" "p varx deallocated"
+gdb_test "ptype varx" "type = <(object|the array) is not allocated>" "ptype varx deallocated"
+gdb_test "p l" "\\$\[0-9\]* = \\.FALSE\\." "p l if varx deallocated"
+gdb_test "p varx(1,5,17)" "(Cannot access it|Unable to access the object) because the (object|array) is not allocated\\." "p varx(1,5,17) deallocated"
+gdb_test "ptype varx(1,5,17)" "(Cannot access it|Unable to access the object) because the (object|array) is not allocated\\." "ptype varx(1,5,17) deallocated"
+
+gdb_breakpoint [gdb_get_line_number "vary-passed"]
+gdb_continue_to_breakpoint "vary-passed"
+# $1 = (( ( 1, 1, 1, 1, 1, 1) ( 1, 1, 1, 1, 1, 1) --- , 1) ) ( ( 1, 1, ...) ...) ...)
+gdb_test "p vary" "\\$\[0-9\]* = \\(\[()1, .\]*\\)"
+
+gdb_breakpoint [gdb_get_line_number "vary-filled"]
+gdb_continue_to_breakpoint "vary-filled"
+gdb_test "ptype vary" "type = real(\\(kind=4\\)|\\*4) \\(10,10\\)"
+gdb_test "p vary(1, 1)" "\\$\[0-9\]* = 8"
+gdb_test "p vary(2, 2)" "\\$\[0-9\]* = 9"
+gdb_test "p vary(1, 3)" "\\$\[0-9\]* = 10"
+# $1 = (( ( 3, 3, 3, 3, 3, 3) ( 3, 3, 3, 3, 3, 3) --- , 3) ) ( ( 3, 3, ...) ...) ...)
+gdb_test "p varw" "\\$\[0-9\]* = \\(\[()3, .\]*\\)"
+
+gdb_breakpoint [gdb_get_line_number "varw-almostfilled"]
+gdb_continue_to_breakpoint "varw-almostfilled"
+gdb_test "ptype varw" "type = real(\\(kind=4\\)|\\*4) \\(5,4,3\\)"
+gdb_test "p varw(3,1,1)=1" "\\$\[0-9\]* = 1"
+# $1 = (( ( 6, 5, 1, 5, 5, 5) ( 5, 5, 5, 5, 5, 5) --- , 5) ) ( ( 5, 5, ...) ...) ...)
+gdb_test "p varw" "\\$\[0-9\]* = \\( *\\( *\\( *6, *5, *1,\[()5, .\]*\\)" "p varw filled"
+# "up" works with GCC but other Fortran compilers may copy the values into the
+# outer function only on the exit of the inner function.
+gdb_test "finish" ".*call bar \\(y, x\\)"
+gdb_test "p z(2,4,5)" "\\$\[0-9\]* = 3"
+gdb_test "p z(2,4,6)" "\\$\[0-9\]* = 6"
+gdb_test "p z(2,4,7)" "\\$\[0-9\]* = 5"
+gdb_test "p z(4,4,6)" "\\$\[0-9\]* = 1"
+
+gdb_breakpoint [gdb_get_line_number "varz-almostfilled"]
+gdb_continue_to_breakpoint "varz-almostfilled"
+# GCC uses the pointer type here, Intel Fortran Compiler 10.1.008 does not.
+gdb_test "ptype varz" "type = (PTR TO -> \\( )?real(\\(kind=4\\)|\\*4) \\(\\*\\)\\)?"
+# Intel Fortran Compiler 10.1.008 has a bug here - (2:11,7:7)
+# as it produces DW_AT_lower_bound == DW_AT_upper_bound == 7.
+gdb_test "ptype vart" "type = (PTR TO -> \\( )?real(\\(kind=4\\)|\\*4) \\(2:11,7:\\*\\)\\)?"
+gdb_test "p varz" "\\$\[0-9\]* = \\(\\)"
+gdb_test "p vart" "\\$\[0-9\]* = \\(\\)"
+gdb_test "p varz(3)" "\\$\[0-9\]* = 4"
+# maps to foo::vary(1,1)
+gdb_test "p vart(2,7)" "\\$\[0-9\]* = 8"
+# maps to foo::vary(2,2)
+gdb_test "p vart(3,8)" "\\$\[0-9\]* = 9"
+# maps to foo::vary(1,3)
+gdb_test "p vart(2,9)" "\\$\[0-9\]* = 10"
+
+set test "quit #1"
+gdb_test_multiple "quit" $test {
+    -re "The program is running.  Quit anyway \\(and kill it\\)\\? \\(y or n\\) " {
+	pass $test
+    }
+}
+set test "quit #2"
+gdb_test_multiple "y" $test {
+    eof {
+	pass $test
+    }
+}
diff --git a/gdb/testsuite/gdb.fortran/dynamic.f90 b/gdb/testsuite/gdb.fortran/dynamic.f90
new file mode 100644
index 0000000..0f43564
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/dynamic.f90
@@ -0,0 +1,98 @@
+! Copyright 2007 Free Software Foundation, Inc.
+!
+! This program is free software; you can redistribute it and/or modify
+! it under the terms of the GNU General Public License as published by
+! the Free Software Foundation; either version 2 of the License, or
+! (at your option) any later version.
+!
+! This program is distributed in the hope that it will be useful,
+! but WITHOUT ANY WARRANTY; without even the implied warranty of
+! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+! GNU General Public License for more details.
+!
+! You should have received a copy of the GNU General Public License
+! along with this program; if not, write to the Free Software
+! Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+!
+! Ihis file is the Fortran source file for dynamic.exp.
+! Original file written by Jakub Jelinek <jakub@redhat.com>.
+! Modified for the GDB testcase by Jan Kratochvil <jan.kratochvil@redhat.com>.
+
+subroutine baz
+  real, target, allocatable :: varx (:, :, :)
+  real, pointer :: varv (:, :, :)
+  real, target :: varu (1, 2, 3)
+  logical :: l
+  allocate (varx (1:6, 5:15, 17:28))            ! varx-init
+  l = allocated (varx)
+  varx(:, :, :) = 6                             ! varx-allocated
+  varx(1, 5, 17) = 7
+  varx(2, 6, 18) = 8
+  varx(6, 15, 28) = 9
+  varv => varx                                  ! varx-filled
+  l = associated (varv)
+  varv(3, 7, 19) = 10                           ! varv-associated
+  varv => null ()                               ! varv-filled
+  l = associated (varv)
+  deallocate (varx)                             ! varv-deassociated
+  l = allocated (varx)
+  varu(:, :, :) = 10                            ! varx-deallocated
+  allocate (varv (1:6, 5:15, 17:28))
+  l = associated (varv)
+  varv(:, :, :) = 6
+  varv(1, 5, 17) = 7
+  varv(2, 6, 18) = 8
+  varv(6, 15, 28) = 9
+  deallocate (varv)
+  l = associated (varv)
+  varv => varu
+  varv(1, 1, 1) = 6
+  varv(1, 2, 3) = 7
+  l = associated (varv)
+end subroutine baz
+subroutine foo (vary, varw)
+  real :: vary (:, :)
+  real :: varw (:, :, :)
+  vary(:, :) = 4                                ! vary-passed
+  vary(1, 1) = 8
+  vary(2, 2) = 9
+  vary(1, 3) = 10
+  varw(:, :, :) = 5                             ! vary-filled
+  varw(1, 1, 1) = 6
+  varw(2, 2, 2) = 7                             ! varw-almostfilled
+end subroutine foo
+subroutine bar (varz, vart)
+  real :: varz (*)
+  real :: vart (2:11, 7:*)
+  varz(1:3) = 4
+  varz(2) = 5                                   ! varz-almostfilled
+  vart(2,7) = vart(2,7)
+end subroutine bar
+program test
+  interface
+    subroutine foo (vary, varw)
+    real :: vary (:, :)
+    real :: varw (:, :, :)
+    end subroutine
+  end interface
+  interface
+    subroutine bar (varz, vart)
+    real :: varz (*)
+    real :: vart (2:11, 7:*)
+    end subroutine
+  end interface
+  real :: x (10, 10), y (5), z(8, 8, 8)
+  x(:,:) = 1
+  y(:) = 2
+  z(:,:,:) = 3
+  call baz
+  call foo (x, z(2:6, 4:7, 6:8))
+  call bar (y, x)
+  if (x (1, 1) .ne. 8 .or. x (2, 2) .ne. 9 .or. x (1, 2) .ne. 4) call abort
+  if (x (1, 3) .ne. 10) call abort
+  if (z (2, 4, 6) .ne. 6 .or. z (3, 5, 7) .ne. 7 .or. z (2, 4, 7) .ne. 5) call abort
+  if (any (y .ne. (/4, 5, 4, 2, 2/))) call abort
+  call foo (transpose (x), z)
+  if (x (1, 1) .ne. 8 .or. x (2, 2) .ne. 9 .or. x (1, 2) .ne. 4) call abort
+  if (x (3, 1) .ne. 10) call abort
+end
diff --git a/gdb/testsuite/gdb.fortran/string.exp b/gdb/testsuite/gdb.fortran/string.exp
new file mode 100644
index 0000000..ab72206
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/string.exp
@@ -0,0 +1,72 @@
+# Copyright 2008 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+# 
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+# 
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  
+
+# This file was written by Jan Kratochvil <jan.kratochvil@redhat.com>.
+
+# This file is part of the gdb testsuite.  It contains tests for Fortran
+# strings with dynamic length.
+
+set testfile "string"
+set srcfile ${testfile}.f90
+set binfile ${objdir}/${subdir}/${testfile}
+
+if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug f77 quiet}] != "" } {
+    untested "Couldn't compile ${srcfile}"
+    return -1
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+if ![runto MAIN__] then {
+    perror "couldn't run to breakpoint MAIN__"
+    continue
+}
+
+gdb_breakpoint [gdb_get_line_number "var-init"]
+gdb_continue_to_breakpoint "var-init"
+gdb_test "ptype c" "type = character(\\(kind=1\\)|\\*1)"
+gdb_test "ptype d" "type = character(\\(kind=8\\)|\\*8)"
+gdb_test "ptype e" "type = character(\\(kind=4\\)|\\*4)"
+gdb_test "ptype f" "type = character(\\(kind=4\\)|\\*4) \\(7,8:10\\)"
+gdb_test "ptype *e" "Attempt to take contents of a non-pointer value."
+gdb_test "ptype *f" "type = character(\\(kind=4\\)|\\*4) \\(7\\)"
+gdb_test "p c" "\\$\[0-9\]* = 'c'"
+gdb_test "p d" "\\$\[0-9\]* = 'd       '"
+gdb_test "p e" "\\$\[0-9\]* = 'g   '"
+gdb_test "p f" "\\$\[0-9\]* = \\(\\( 'h   ', 'h   ', 'h   ', 'h   ', 'h   ', 'h   ', 'h   '\\) \\( 'h   ', 'h   ', 'h   ', 'h   ', 'h   ', 'h   ', 'h   '\\) \\( 'h   ', 'h   ', 'h   ', 'h   ', 'h   ', 'h   ', 'h   '\\) \\)"
+gdb_test "p *e" "Attempt to take contents of a non-pointer value."
+gdb_test "p *f" "Attempt to take contents of a non-pointer value."
+
+gdb_breakpoint [gdb_get_line_number "var-finish"]
+gdb_continue_to_breakpoint "var-finish"
+gdb_test "p e" "\\$\[0-9\]* = 'e   '" "p e re-set"
+gdb_test "p f" "\\$\[0-9\]* = \\(\\( 'f   ', 'f   ', 'f   ', 'f   ', 'f   ', 'f   ', 'f   '\\) \\( 'f2  ', 'f   ', 'f   ', 'f   ', 'f   ', 'f   ', 'f   '\\) \\( 'f   ', 'f   ', 'f   ', 'f   ', 'f   ', 'f   ', 'f   '\\) \\)" "p *f re-set"
+
+set test "quit #1"
+gdb_test_multiple "quit" $test {
+    -re "The program is running.  Quit anyway \\(and kill it\\)\\? \\(y or n\\) " {
+	pass $test
+    }
+}
+set test "quit #2"
+gdb_test_multiple "y" $test {
+    eof {
+	pass $test
+    }
+}
diff --git a/gdb/testsuite/gdb.fortran/string.f90 b/gdb/testsuite/gdb.fortran/string.f90
new file mode 100644
index 0000000..226dc5d
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/string.f90
@@ -0,0 +1,37 @@
+! Copyright 2008 Free Software Foundation, Inc.
+!
+! This program is free software; you can redistribute it and/or modify
+! it under the terms of the GNU General Public License as published by
+! the Free Software Foundation; either version 2 of the License, or
+! (at your option) any later version.
+!
+! This program is distributed in the hope that it will be useful,
+! but WITHOUT ANY WARRANTY; without even the implied warranty of
+! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+! GNU General Public License for more details.
+!
+! You should have received a copy of the GNU General Public License
+! along with this program; if not, write to the Free Software
+! Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+!
+! Ihis file is the Fortran source file for dynamic.exp.
+! Original file written by Jakub Jelinek <jakub@redhat.com>.
+! Modified for the GDB testcase by Jan Kratochvil <jan.kratochvil@redhat.com>.
+
+subroutine foo (e, f)
+  character (len=1) :: c
+  character (len=8) :: d
+  character (len=*) :: e
+  character (len=*) :: f (1:7, 8:10)
+  c = 'c'
+  d = 'd'
+  e = 'e'                                       ! var-init
+  f = 'f'
+  f(1,9) = 'f2'
+  c = 'c'                                       ! var-finish
+end subroutine foo
+  character (len=4) :: g, h (1:7, 8:10)
+  g = 'g'
+  h = 'h'
+  call foo (g, h)
+end
diff --git a/gdb/top.c b/gdb/top.c
index d5ef706..6045e21 100644
--- a/gdb/top.c
+++ b/gdb/top.c
@@ -377,6 +377,7 @@ execute_command (char *p, int from_tty)
     }
   
   free_all_values ();
+  free_all_types ();
 
   /* Force cleanup of any alloca areas if using C alloca instead of
      a builtin alloca.  */
diff --git a/gdb/typeprint.c b/gdb/typeprint.c
index 1f824fa..60a4c5b 100644
--- a/gdb/typeprint.c
+++ b/gdb/typeprint.c
@@ -35,6 +35,7 @@
 #include "gdb_string.h"
 #include "exceptions.h"
 #include "valprint.h"
+#include "dwarf2loc.h"
 #include <errno.h>
 
 extern void _initialize_typeprint (void);
@@ -76,6 +77,9 @@ void
 type_print (struct type *type, char *varstring, struct ui_file *stream,
 	    int show)
 {
+  if (show >= 0)
+    type = check_typedef (type);
+
   LA_PRINT_TYPE (type, varstring, stream, show, 0);
 }
 
@@ -115,7 +119,8 @@ whatis_exp (char *exp, int show)
 {
   struct expression *expr;
   struct value *val;
-  struct cleanup *old_chain = NULL;
+  /* Required at least for the object_address_set call.  */
+  struct cleanup *old_chain = make_cleanup (null_cleanup, NULL);
   struct type *real_type = NULL;
   struct type *type;
   int full = 0;
@@ -126,12 +131,13 @@ whatis_exp (char *exp, int show)
   if (exp)
     {
       expr = parse_expression (exp);
-      old_chain = make_cleanup (free_current_contents, &expr);
+      make_cleanup (free_current_contents, &expr);
       val = evaluate_type (expr);
     }
   else
     val = access_value_history (0);
 
+  object_address_set (VALUE_ADDRESS (val));
   type = value_type (val);
 
   get_user_print_options (&opts);
@@ -168,8 +174,7 @@ whatis_exp (char *exp, int show)
   type_print (type, "", gdb_stdout, show);
   printf_filtered ("\n");
 
-  if (exp)
-    do_cleanups (old_chain);
+  do_cleanups (old_chain);
 }
 
 static void
diff --git a/gdb/valarith.c b/gdb/valarith.c
index f38cdb8..8e103cf 100644
--- a/gdb/valarith.c
+++ b/gdb/valarith.c
@@ -164,9 +164,9 @@ an integer nor a pointer of the same type."));
 struct value *
 value_subscript (struct value *array, struct value *idx)
 {
-  struct value *bound;
   int c_style = current_language->c_style_arrays;
   struct type *tarray;
+  LONGEST index = value_as_long (idx);
 
   array = coerce_ref (array);
   tarray = check_typedef (value_type (array));
@@ -179,13 +179,26 @@ value_subscript (struct value *array, struct value *idx)
       get_discrete_bounds (range_type, &lowerbound, &upperbound);
 
       if (VALUE_LVAL (array) != lval_memory)
-	return value_subscripted_rvalue (array, idx, lowerbound);
+	{
+	  if (index >= lowerbound && index <= upperbound)
+	    {
+	      CORE_ADDR element_size = TYPE_LENGTH (TYPE_TARGET_TYPE (tarray));
+	      CORE_ADDR offset = (index - lowerbound) * element_size;
+
+	      return value_subscripted_rvalue (array, offset);
+	    }
+	  error (_("array or string index out of range"));
+	}
 
       if (c_style == 0)
 	{
-	  LONGEST index = value_as_long (idx);
 	  if (index >= lowerbound && index <= upperbound)
-	    return value_subscripted_rvalue (array, idx, lowerbound);
+	    {
+	      CORE_ADDR element_size = TYPE_LENGTH (TYPE_TARGET_TYPE (tarray));
+	      CORE_ADDR offset = (index - lowerbound) * element_size;
+
+	      return value_subscripted_rvalue (array, offset);
+	    }
 	  /* Emit warning unless we have an array of unknown size.
 	     An array of unknown size has lowerbound 0 and upperbound -1.  */
 	  if (upperbound > -1)
@@ -194,49 +207,52 @@ value_subscript (struct value *array, struct value *idx)
 	  c_style = 1;
 	}
 
-      if (lowerbound != 0)
-	{
-	  bound = value_from_longest (value_type (idx), (LONGEST) lowerbound);
-	  idx = value_binop (idx, bound, BINOP_SUB);
-	}
-
+      index -= lowerbound;
       array = value_coerce_array (array);
     }
 
   if (c_style)
-    return value_ind (value_ptradd (array, idx));
+    {
+      struct value *idx;
+
+      idx = value_from_longest (builtin_type_int32, index);
+      return value_ind (value_ptradd (array, idx));
+    }
   else
     error (_("not an array or string"));
 }
 
-/* Return the value of EXPR[IDX], expr an aggregate rvalue
-   (eg, a vector register).  This routine used to promote floats
-   to doubles, but no longer does.  */
+/* Return the value of *((void *) ARRAY + ELEMENT), ARRAY an aggregate rvalue
+   (eg, a vector register).  This routine used to promote floats to doubles,
+   but no longer does.  OFFSET is zero-based with 0 for the lowermost existing
+   element, it must be expressed in bytes (therefore multiplied by
+   check_typedef (TYPE_TARGET_TYPE (array_type)).  */
 
 struct value *
-value_subscripted_rvalue (struct value *array, struct value *idx, int lowerbound)
+value_subscripted_rvalue (struct value *array, CORE_ADDR offset)
 {
   struct type *array_type = check_typedef (value_type (array));
   struct type *elt_type = check_typedef (TYPE_TARGET_TYPE (array_type));
-  unsigned int elt_size = TYPE_LENGTH (elt_type);
-  LONGEST index = value_as_long (idx);
-  unsigned int elt_offs = elt_size * longest_to_int (index - lowerbound);
   struct value *v;
 
-  if (index < lowerbound || elt_offs >= TYPE_LENGTH (array_type))
-    error (_("no such vector element"));
+  /* Do not check TYPE_LENGTH (array_type) as we may have been given the
+     innermost dimension of a multi-dimensional Fortran array where its length
+     is shorter than the possibly accessed element offset.  */
 
   v = allocate_value (elt_type);
   if (VALUE_LVAL (array) == lval_memory && value_lazy (array))
     set_value_lazy (v, 1);
   else
-    memcpy (value_contents_writeable (v),
-	    value_contents (array) + elt_offs, elt_size);
+    {
+      unsigned int elt_size = TYPE_LENGTH (elt_type);
+      memcpy (value_contents_writeable (v),
+	      value_contents (array) + offset, elt_size);
+    }
 
   set_value_component_location (v, array);
   VALUE_REGNUM (v) = VALUE_REGNUM (array);
   VALUE_FRAME_ID (v) = VALUE_FRAME_ID (array);
-  set_value_offset (v, value_offset (array) + elt_offs);
+  set_value_offset (v, value_offset (array) + offset);
   return v;
 }
 
diff --git a/gdb/valops.c b/gdb/valops.c
index 9810f2b..c3a48b4 100644
--- a/gdb/valops.c
+++ b/gdb/valops.c
@@ -38,6 +38,7 @@
 #include "cp-support.h"
 #include "dfp.h"
 #include "user-regs.h"
+#include "dwarf2loc.h"
 
 #include <errno.h>
 #include "gdb_string.h"
@@ -371,8 +372,6 @@ value_cast (struct type *type, struct value *arg2)
 	  new_length = val_length / element_length;
 	  if (val_length % element_length != 0)
 	    warning (_("array element type size does not divide object size in cast"));
-	  /* FIXME-type-allocation: need a way to free this type when
-	     we are done with it.  */
 	  range_type = create_range_type ((struct type *) NULL,
 					  TYPE_TARGET_TYPE (range_type),
 					  low_bound,
@@ -568,6 +567,64 @@ value_one (struct type *type, enum lval_type lv)
   return val;
 }
 
+/* object_address_set must be already called before this function.  */
+
+const char *
+object_address_data_not_valid (struct type *type)
+{
+  /* Attributes are present only at the target type of a typedef.  Make the
+     call conditional as it would otherwise loop through type_length_get.  */
+  if (TYPE_CODE (type) == TYPE_CODE_TYPEDEF)
+    CHECK_TYPEDEF (type);
+
+  /* DW_AT_associated has a preference over DW_AT_allocated.  */
+  if (TYPE_NOT_ASSOCIATED (type)
+      || (TYPE_ASSOCIATED (type) != NULL
+	  && 0 == dwarf_locexpr_baton_eval (TYPE_ASSOCIATED (type))))
+    return N_("object is not associated");
+
+  if (TYPE_NOT_ALLOCATED (type)
+      || (TYPE_ALLOCATED (type) != NULL
+	  && 0 == dwarf_locexpr_baton_eval (TYPE_ALLOCATED (type))))
+    return N_("object is not allocated");
+
+  return NULL;
+}
+
+/* Return non-zero if the variable is valid.  If it is valid the function
+   may store the data address (DW_AT_DATA_LOCATION) of TYPE at *ADDRESS_RETURN.
+   You must set *ADDRESS_RETURN as VALUE_ADDRESS (VAL) before calling this
+   function.  If no DW_AT_DATA_LOCATION is present for TYPE the address at
+   *ADDRESS_RETURN is left unchanged.  ADDRESS_RETURN must not be NULL, use
+   object_address_data_not_valid () for just the data validity check.  */
+
+int
+object_address_get_data (struct type *type, CORE_ADDR *address_return)
+{
+  gdb_assert (address_return != NULL);
+
+  object_address_set (*address_return);
+
+  /* TYPE_DATA_LOCATION_DWARF_BLOCK / TYPE_DATA_LOCATION_ADDR are present only
+     at the target type of a typedef.  */
+  CHECK_TYPEDEF (type);
+
+  if (object_address_data_not_valid (type) != NULL)
+    {
+      /* Do not try to evaluate DW_AT_data_location as it may even crash
+	 (it would just return the value zero in the gfortran case).  */
+      return 0;
+    }
+
+  if (TYPE_DATA_LOCATION_IS_ADDR (type))
+    *address_return = TYPE_DATA_LOCATION_ADDR (type);
+  else if (TYPE_DATA_LOCATION_DWARF_BLOCK (type) != NULL)
+    *address_return
+      = dwarf_locexpr_baton_eval (TYPE_DATA_LOCATION_DWARF_BLOCK (type));
+
+  return 1;
+}
+
 /* Return a value with type TYPE located at ADDR.
 
    Call value_at only if the data needs to be fetched immediately;
@@ -637,11 +694,19 @@ value_fetch_lazy (struct value *val)
   allocate_value_contents (val);
   if (VALUE_LVAL (val) == lval_memory)
     {
-      CORE_ADDR addr = VALUE_ADDRESS (val) + value_offset (val);
-      int length = TYPE_LENGTH (check_typedef (value_enclosing_type (val)));
+      CORE_ADDR addr = VALUE_ADDRESS (val);
 
-      if (length)
-	read_memory (addr, value_contents_all_raw (val), length);
+      if (object_address_get_data (value_type (val), &addr))
+	{
+	  struct type *type = value_enclosing_type (val);
+	  int length = TYPE_LENGTH (check_typedef (type));
+
+	  if (length)
+	    {
+	      addr += value_offset (val);
+	      read_memory (addr, value_contents_all_raw (val), length);
+	    }
+	}
     }
   else if (VALUE_LVAL (val) == lval_register)
     {
@@ -1036,7 +1101,8 @@ address_of_variable (struct symbol *var, struct block *b)
 
   val = value_of_variable (var, b);
 
-  if ((VALUE_LVAL (val) == lval_memory && value_lazy (val))
+  if ((VALUE_LVAL (val) == lval_memory && value_lazy (val)
+       && object_address_get_data (type, &VALUE_ADDRESS (val)))
       || TYPE_CODE (type) == TYPE_CODE_FUNC)
     {
       CORE_ADDR addr = VALUE_ADDRESS (val);
@@ -1145,6 +1211,7 @@ struct value *
 value_coerce_array (struct value *arg1)
 {
   struct type *type = check_typedef (value_type (arg1));
+  CORE_ADDR address;
 
   /* If the user tries to do something requiring a pointer with an
      array that has not yet been pushed to the target, then this would
@@ -1154,8 +1221,12 @@ value_coerce_array (struct value *arg1)
   if (VALUE_LVAL (arg1) != lval_memory)
     error (_("Attempt to take address of value not located in memory."));
 
+  address = VALUE_ADDRESS (arg1);
+  if (!object_address_get_data (type, &address))
+    error (_("Attempt to take address of non-valid value."));
+
   return value_from_pointer (lookup_pointer_type (TYPE_TARGET_TYPE (type)),
-			     (VALUE_ADDRESS (arg1) + value_offset (arg1)));
+			     address + value_offset (arg1));
 }
 
 /* Given a value which is a function, return a value which is a pointer
@@ -2989,8 +3060,6 @@ value_slice (struct value *array, int lowbound, int length)
       || lowbound + length - 1 > upperbound)
     error (_("slice out of range"));
 
-  /* FIXME-type-allocation: need a way to free this type when we are
-     done with it.  */
   slice_range_type = create_range_type ((struct type *) NULL,
 					TYPE_TARGET_TYPE (range_type),
 					lowbound, 
diff --git a/gdb/value.c b/gdb/value.c
index 4d4329e..47739c9 100644
--- a/gdb/value.c
+++ b/gdb/value.c
@@ -225,7 +225,9 @@ allocate_value_lazy (struct type *type)
   val->next = all_values;
   all_values = val;
   val->type = type;
+  type_incref (type);
   val->enclosing_type = type;
+  type_incref (type);
   VALUE_LVAL (val) = not_lval;
   VALUE_ADDRESS (val) = 0;
   VALUE_FRAME_ID (val) = null_frame_id;
@@ -269,13 +271,9 @@ struct value *
 allocate_repeat_value (struct type *type, int count)
 {
   int low_bound = current_language->string_lower_bound;		/* ??? */
-  /* FIXME-type-allocation: need a way to free this type when we are
-     done with it.  */
   struct type *range_type
   = create_range_type ((struct type *) NULL, builtin_type_int32,
 		       low_bound, count + low_bound - 1);
-  /* FIXME-type-allocation: need a way to free this type when we are
-     done with it.  */
   return allocate_value (create_array_type ((struct type *) NULL,
 					    type, range_type));
 }
@@ -335,6 +333,8 @@ value_type (struct value *value)
 void
 deprecated_set_value_type (struct value *value, struct type *type)
 {
+  type_incref (type);
+  type_decref (value->type);
   value->type = type;
 }
 
@@ -552,6 +552,9 @@ value_free (struct value *val)
 {
   if (val)
     {
+      type_decref (val->type);
+      type_decref (val->enclosing_type);
+
       if (VALUE_LVAL (val) == lval_computed)
 	{
 	  struct lval_funcs *funcs = val->location.computed.funcs;
@@ -655,6 +658,8 @@ value_copy (struct value *arg)
     val = allocate_value_lazy (encl_type);
   else
     val = allocate_value (encl_type);
+  type_incref (arg->type);
+  type_decref (val->type);
   val->type = arg->type;
   VALUE_LVAL (val) = VALUE_LVAL (arg);
   val->location = arg->location;
@@ -693,6 +698,7 @@ set_value_component_location (struct value *component, struct value *whole)
     VALUE_LVAL (component) = VALUE_LVAL (whole);
 
   component->location = whole->location;
+
   if (VALUE_LVAL (whole) == lval_computed)
     {
       struct lval_funcs *funcs = whole->location.computed.funcs;
@@ -700,6 +706,8 @@ set_value_component_location (struct value *component, struct value *whole)
       if (funcs->copy_closure)
         component->location.computed.closure = funcs->copy_closure (whole);
     }
+
+  object_address_get_data (value_type (whole), &VALUE_ADDRESS (component));
 }
 
 
@@ -830,6 +838,25 @@ show_values (char *num_exp, int from_tty)
       num_exp[1] = '\0';
     }
 }
+
+/* Sanity check for memory leaks and proper types reference counting.  */
+
+static void
+value_history_cleanup (void *unused)
+{
+  while (value_history_chain)
+    {
+      struct value_history_chunk *chunk = value_history_chain;
+      int i;
+
+      for (i = 0; i < ARRAY_SIZE (chunk->values); i++)
+      	value_free (chunk->values[i]);
+
+      value_history_chain = chunk->next;
+      xfree (chunk);
+    }
+  value_history_count = 0;
+}
 
 /* Internal variables.  These are variables within the debugger
    that hold values assigned by debugger commands.
@@ -1067,12 +1094,21 @@ preserve_one_value (struct value *value, struct objfile *objfile,
 		    htab_t copied_types)
 {
   if (TYPE_OBJFILE (value->type) == objfile)
-    value->type = copy_type_recursive (objfile, value->type, copied_types);
+    {
+      /* No need to decref the old type here, since we know it has no
+	 reference count.  */
+      value->type = copy_type_recursive (value->type, copied_types);
+      type_incref (value->type);
+    }
 
   if (TYPE_OBJFILE (value->enclosing_type) == objfile)
-    value->enclosing_type = copy_type_recursive (objfile,
-						 value->enclosing_type,
-						 copied_types);
+    {
+      /* No need to decref the old type here, since we know it has no
+	 reference count.  */
+      value->enclosing_type = copy_type_recursive (value->enclosing_type,
+						   copied_types);
+      type_incref (value->enclosing_type);
+    }
 }
 
 /* Update the internal variables and value history when OBJFILE is
@@ -1461,6 +1497,8 @@ value_static_field (struct type *type, int fieldno)
 struct value *
 value_change_enclosing_type (struct value *val, struct type *new_encl_type)
 {
+  type_incref (new_encl_type);
+  type_decref (val->enclosing_type);
   if (TYPE_LENGTH (new_encl_type) > TYPE_LENGTH (value_enclosing_type (val))) 
     val->contents =
       (gdb_byte *) xrealloc (val->contents, TYPE_LENGTH (new_encl_type));
@@ -1516,6 +1554,8 @@ value_primitive_field (struct value *arg1, int offset,
 	  memcpy (value_contents_all_raw (v), value_contents_all_raw (arg1),
 		  TYPE_LENGTH (value_enclosing_type (arg1)));
 	}
+      type_incref (type);
+      type_decref (v->type);
       v->type = type;
       v->offset = value_offset (arg1);
       v->embedded_offset = (offset + value_embedded_offset (arg1)
@@ -1944,4 +1984,6 @@ init-if-undefined VARIABLE = EXPRESSION\n\
 Set an internal VARIABLE to the result of the EXPRESSION if it does not\n\
 exist or does not contain a value.  The EXPRESSION is not evaluated if the\n\
 VARIABLE is already initialized."));
+
+  make_final_cleanup (value_history_cleanup, NULL);
 }
diff --git a/gdb/value.h b/gdb/value.h
index aa43365..ad5306a 100644
--- a/gdb/value.h
+++ b/gdb/value.h
@@ -348,6 +348,10 @@ extern struct value *value_from_decfloat (struct type *type,
 					  const gdb_byte *decbytes);
 extern struct value *value_from_string (char *string);
 
+extern const char *object_address_data_not_valid (struct type *type);
+extern int object_address_get_data (struct type *type,
+				    CORE_ADDR *address_return);
+
 extern struct value *value_at (struct type *type, CORE_ADDR addr);
 extern struct value *value_at_lazy (struct type *type, CORE_ADDR addr);
 
@@ -658,5 +662,7 @@ extern struct value *value_allocate_space_in_inferior (int);
 
 extern struct value *value_of_local (const char *name, int complain);
 
-extern struct value * value_subscripted_rvalue (struct value *array, struct value *idx, int lowerbound);
+extern struct value *value_subscripted_rvalue (struct value *array,
+					       CORE_ADDR offset);
+
 #endif /* !defined (VALUE_H) */