Re: [patch] Re: Regression: field type preservation: 7.0 -> 7.0.1+HEAD

[Jan Kratochvil <jan.kratochvil@redhat.com>]

On Sun, 03 Jan 2010 05:57:17 +0100, Joel Brobecker wrote:
> So, my proposal, if the other maintainers agree, is to document
> the side-effect of check_typedef (sets the typedef TYPE_LENGTH)
> as this appears to be a fully-intended behavior, and then do:

OK, attached.  As I find the type system [censored] I am never sure how it can
be used in a sane way.

OK to check-in with this wording?

No regressions on {x86_64,x86_64-m32,i686}-fedora12-linux-gnu.


Thanks,
Jan


gdb/
2010-01-03  Jan Kratochvil  <jan.kratochvil@redhat.com>
	    Joel Brobecker  <brobecker@adacore.com>

	* gdbtypes.c (check_typedef): New comment on type length.
	* value.c (allocate_value_lazy): Remove the unused atype variable.  New
	comment on type length.
	(value_primitive_field): Keep the original TYPE value, new comment.

gdb/testsuite/
2010-01-03  Jan Kratochvil  <jan.kratochvil@redhat.com>

	* gdb.mi/var-cmd.c (do_bitfield_tests): Change "V.sharable" type to
	"uint_for_mi_testing".

--- a/gdb/gdbtypes.c
+++ b/gdb/gdbtypes.c
@@ -1342,12 +1342,17 @@ stub_noname_complaint (void)
    symbols which contain a full definition for the type.
 
    This used to be coded as a macro, but I don't think it is called 
-   often enough to merit such treatment.  */
+   often enough to merit such treatment.
 
-/* Find the real type of TYPE.  This function returns the real type,
+   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.
+
+   If TYPE is a TYPE_CODE_TYPEDEF, its length is (also) set to the length of
+   the target type instead of zero.  However, in the case of TYPE_CODE_TYPEDEF
+   check_typedef can still return different type than the original TYPE
+   pointer.  */
 
 struct type *
 check_typedef (struct type *type)
--- a/gdb/testsuite/gdb.mi/var-cmd.c
+++ b/gdb/testsuite/gdb.mi/var-cmd.c
@@ -494,7 +494,7 @@ void do_bitfield_tests ()
     mi_create_varobj V d "create varobj for Data"
     mi_list_varobj_children "V" {
         {"V.alloc" "alloc" "0" "int"}
-        {"V.sharable" "sharable" "0" "unsigned int"}
+        {"V.sharable" "sharable" "0" "uint_for_mi_testing"}
     } "list children of Data"
     mi_check_varobj_value V.sharable 3 "access bitfield"
     :*/
--- a/gdb/value.c
+++ b/gdb/value.c
@@ -254,7 +254,14 @@ struct value *
 allocate_value_lazy (struct type *type)
 {
   struct value *val;
-  struct type *atype = check_typedef (type);
+
+  /* Call check_typedef on our type to make sure that, if TYPE
+     is a TYPE_CODE_TYPEDEF, its length is set to the length
+     of the target type instead of zero.  However, we do not
+     replace the typedef type by the target type, because we want
+     to keep the typedef in order to be able to set the VAL's type
+     description correctly.  */
+  check_typedef (type);
 
   val = (struct value *) xzalloc (sizeof (struct value));
   val->contents = NULL;
@@ -1873,7 +1880,14 @@ value_primitive_field (struct value *arg1, int offset,
 
   CHECK_TYPEDEF (arg_type);
   type = TYPE_FIELD_TYPE (arg_type, fieldno);
-  type = check_typedef (type);
+
+  /* Call check_typedef on our type to make sure that, if TYPE
+     is a TYPE_CODE_TYPEDEF, its length is set to the length
+     of the target type instead of zero.  However, we do not
+     replace the typedef type by the target type, because we want
+     to keep the typedef in order to be able to print the type
+     description correctly.  */
+  check_typedef (type);
 
   /* Handle packed fields */