On Wed, 28 Mar 2012 10:55:13 -0600
Tom Tromey <tromey@redhat.com> wrote:
> >>>>> "Siddhesh" == Siddhesh Poyarekar <siddhesh@redhat.com> writes:
> 
> Siddhesh> I took a separate approach from the one I took in:
> Siddhesh> http://sourceware.org/ml/gdb-patches/2012-02/msg00403.html
> 
> Siddhesh> and removed the overloaded meaning of the bitpos location
> Siddhesh> variable to fix PR symtab/7259. In the following patch, I
> Siddhesh> introduce a separate field_location union member 'enumval'
> Siddhesh> which can accept LONGEST and hence expand enum values to
> Siddhesh> 64-bit signed values. With this change, bitpos now only is
> Siddhesh> used for (non-negative) offsets into structures, since the
> Siddhesh> other overload of bitpos (range bounds) were already
> Siddhesh> separated into struct range_bound.
> 
> I think this also needs an update in py-type.c:convert_field.

Thanks, I missed that. I have attached a modified patch with this
change (and other changes resulting from it) included. Updated
ChangeLog below.

Regards,
Siddhesh


ChangeLog:

2012-03-21  Siddhesh Poyarekar  <siddhesh@redhat.com>

	PR symtab/7259:
	* gdbtypes.h (enum field_loc_kind): New
	FIELD_LOC_KIND_ENUMVAL.
	(struct main_type.fields): Adjust loc_kind and bitsize to
	accommodate FIELD_LOC_KIND_ENUMVAL.
	(struct main_type.fields.field_location): New enumval.
	New macros: FIELD_ENUMVAL, SET_FIELD_ENUMVAL and
	TYPE_FIELD_ENUMVAL.
	* ada-exp.y (convert_char_literal): use TYPE_FIELD_ENUMVAL if
	type is TYPE_CODE_ENUM.
	* ada-lang.c (ada_discrete_type_high_bound): Ditto.
	(ada_discrete_type_low_bound, value_val_atr): Ditto.
	(ada_identical_enum_types_p, pos_atr): Ditto.
	* ada-typeprint.c (print_enum_type): Ditto.
	* ada-valprint.c (print_optional_low_bound): Ditto.
	(ada_print_scalar, ada_val_print_1): Ditto.
	* c-typeprint.c (c_type_print_base): Ditto.
	* coffread.c (coff_read_enum_type): Ditto.
	* dwarf2read.c (process_enumeration_scope): Ditto.
	* gdbtypes.c (get_discrete_bounds): Ditto
	(recursive_dump_type, copy_type_recursive): Ditto.
	* m2-typeprint.c (m2_enum): Ditto.
	* mdebugread.c (parse_symbol): Ditto.
	* p-typeprint.c (pascal_type_print_base): Ditto.
	* python/py-type.c (check_types_equal, convert_field): Ditto.
	* stabsread.c (read_enum_type): Ditto.
	* typepint.c (print_type_scalar): Ditto.
	* valprint.c (generic_val_print): Ditto.
	* gdb-gdb.py (class StructMainTypePrettyPrinter): Ditto.
	(TypeFlagsPrinter): Use field.enumval instead of field.bitpos.
	* python/lib/gdb/printing.py (FlagEnumerationPrinter): Ditto.
	* python/lib/gdb/types.py (make_enum_dict): Ditto.

testsuite/ChangeLog:

2012-03-21  Siddhesh Poyarekar  <siddhesh@redhat.com>

	PR symtab/7259:
	* gdb.base/enumval.exp: New test case.
	* gdb.base/enumval.c: New test case.
	* gdb.python/py-type.exp (test_enums): Modify test to look for
	field.enumval instead of field.bitpos.