[RFA] dwarf2out.c - subrange_type cleanup
Joel Brobecker
brobecker@gnat.com
Thu Jan 15 08:43:00 GMT 2004
Hello,
this is the second cleanup patch to dwarf2out.c wrt subrange_type DIEs
emission. This is a followup on:
http://gcc.gnu.org/ml/gcc-patches/2004-01/msg00693.html
We decided to explore the possibility of removing the checks for
- TYPE_NAME
- TREE_UNSIGNED
- is_ada
In a previous patch, I first removed the check for TYPE_NAME. This
patch now removes the checks for is_ada() and TYPE_UNSIGNED.
I have sent a patch to gdb-patches adding support for subrange_type
DIEs: http://sources.redhat.com/ml/gdb-patches/2004-01/msg00380.html.
It might be wiser to wait for the GDB patch to be applied before this
GCC patch goes in (but the change can still be reviewed anytime, though :)
2004-01-15 J. Brobecker <brobecker at gnat dot com>
* dwarf2out.c (is_subrange_type): Renamed from is_ada_subrange_type().
Remove checks for is_ada() and TREE_UNSIGNED.
(subrange_type_die): Emit a byte_size attribute if the subrange
type size is different from the base type size.
(modified_type_die): Replace call to is_ada_subrange_type() by
call to is_subrange_type().
Bootstrapped on x86-linux, and tested against the testsuite and the
GDB testsuite (GDB was built with the patch I mentionned above).
OK to apply?
Thanks,
--
Joel
-------------- next part --------------
Index: dwarf2out.c
===================================================================
RCS file: /cvs/gcc/gcc/gcc/dwarf2out.c,v
retrieving revision 1.475
diff -u -p -r1.475 dwarf2out.c
--- dwarf2out.c 14 Jan 2004 21:12:02 -0000 1.475
+++ dwarf2out.c 15 Jan 2004 06:58:13 -0000
@@ -3703,7 +3703,7 @@ static void output_file_names (void);
static dw_die_ref base_type_die (tree);
static tree root_type (tree);
static int is_base_type (tree);
-static bool is_ada_subrange_type (tree);
+static bool is_subrange_type (tree);
static dw_die_ref subrange_type_die (tree, dw_die_ref);
static dw_die_ref modified_type_die (tree, int, int, dw_die_ref);
static int type_is_enum (tree);
@@ -7812,24 +7812,14 @@ simple_type_size_in_bits (tree type)
emitted as a subrange type. */
static inline bool
-is_ada_subrange_type (tree type)
+is_subrange_type (tree type)
{
- /* We should use a subrange type in the following situations:
- - For Ada modular types: These types are stored as integer subtypes
- of an unsigned integer type;
- - For subtypes of an Ada enumeration type: These types are stored
- as integer subtypes of enumeral types.
-
- This subrange type is mostly for the benefit of debugger users.
- A nameless type would therefore not be very useful, so no need
- to generate a subrange type in these cases. */
tree subtype = TREE_TYPE (type);
- if (is_ada ()
- && TREE_CODE (type) == INTEGER_TYPE
+ if (TREE_CODE (type) == INTEGER_TYPE
&& subtype != NULL_TREE)
{
- if (TREE_CODE (subtype) == INTEGER_TYPE && TREE_UNSIGNED (subtype))
+ if (TREE_CODE (subtype) == INTEGER_TYPE)
return true;
if (TREE_CODE (subtype) == ENUMERAL_TYPE)
return true;
@@ -7846,6 +7836,7 @@ subrange_type_die (tree type, dw_die_ref
dw_die_ref subtype_die;
dw_die_ref subrange_die;
tree name = TYPE_NAME (type);
+ const HOST_WIDE_INT size_in_bytes = int_size_in_bytes (type);
if (context_die == NULL)
context_die = comp_unit_die;
@@ -7864,6 +7855,13 @@ subrange_type_die (tree type, dw_die_ref
add_name_attribute (subrange_die, IDENTIFIER_POINTER (name));
}
+ if (int_size_in_bytes (TREE_TYPE (type)) != size_in_bytes)
+ {
+ /* The size of the subrange type and its base type do not match,
+ so we need to generate a size attribute for the subrange type. */
+ add_AT_unsigned (subrange_die, DW_AT_byte_size, size_in_bytes);
+ }
+
if (TYPE_MIN_VALUE (type) != NULL)
add_bound_info (subrange_die, DW_AT_lower_bound,
TYPE_MIN_VALUE (type));
@@ -7966,7 +7964,7 @@ modified_type_die (tree type, int is_con
#endif
item_type = TREE_TYPE (type);
}
- else if (is_ada_subrange_type (type))
+ else if (is_subrange_type (type))
mod_type_die = subrange_type_die (type, context_die);
else if (is_base_type (type))
mod_type_die = base_type_die (type);
More information about the Gcc-patches
mailing list