Jerome Guitton [Thu, 29 Nov 2012 16:28:10 +0000 (16:28 +0000)]
Full view of interface-wide types
For displaying the full view of a class-wide object, GDB relies on
the assumption that this view will have the same address as the
address of the object. In the case of simple inheritance, this
assumption is correct; the proper type is deduced by decoding
the tag of the object and converting the result to this full-view
type.
Consider for example an abstract class Shape, a child Circle
which implements an interface Drawable, and the corresponding
following objects:
Then the type specific data and the expanded name of the tag is read
from there:
(gdb) p my_shape'tag
$3 = (access ada.tags.dispatch_table) 0x8059ec4 (classes.circle)
To get the full view, the debugger converts to the corresponding type:
(gdb) p {classes.circle}0x8063e28
$4 = (center => (x => 1, y => 2), radius => 3)
Now, in the case of multiple inheritance, the assumption does not hold
anymore. The address that we have usually points to some
place lower. The offset to the original address is saved in the field
Offset_To_Top of the metadata that are above the tag, at address
obj'tag - 8. In the case of my_shape, this offset is 0:
The following change handles this relocation in the most common cases.
Remaining cases that are still to be investigated are signaled by
comments.
gdb/ChangeLog:
* ada-lang.h (ada_tag_value_at_base_address): New function
declaration.
* ada-lang.c (is_ada95_tag, ada_tag_value_at_base_address):
New functions.
(ada_to_fixed_type_1, ada_evaluate_subexp): Let ada_tag_base_address
relocate the class-wide value if need be.
(ada_value_struct_elt, ada_value_ind, ada_coerce_ref):
Let ada_tag_value_at_base_address relocate the class-wide access/ref
before dereferencing it.
* ada-valprint.c (ada_val_print_1): Relocate to base address
before displaying the content of an interface-wide ref.
gdb/testsuite/ChangeLog:
* gdb.ada/ptype_tagged_param.exp: Adjust expected output in
ptype test.
Jerome Guitton [Thu, 29 Nov 2012 16:26:12 +0000 (16:26 +0000)]
Handle other cases than EVAL_NORMAL in the default case
In the evaluation of an expression in Ada mode, the default case
unwraps the argument unconditionally. For an object of a variant
record type, this unwrapping builds a fixed type from the
specification of the variant type and the actual values of the
object's discriminants. It means that unwrapping needs the "proper"
value for the object, not just a zero value with the proper type.
When not in EVAL_NORMAL, we cannot assume that the evaluation returns
such a proper value; it may well return a zero value of the
appropriate type e.g in EVAL_AVOID_SIDE_EFFECTS. It is wrong to try to
unwrap in that case.
In particular, a problem shows up when using expression of the form
{VARIANT_TYPE}OBJ. GDB first evaluates this expression in
EVAL_AVOID_SIDE_EFFECTS to compute the type, the evaluation of OBJ
in most cases returns a zero value of its type, and as UNOP_MEMVAL
is mapped to the default case its evaluation ends up trying to
read memory around address 0.
gdb/ChangeLog:
* ada-lang.c (ada_evaluate_subexp): Unwrap only in EVAL_NORMAL.
2009-10-19 Edgar E. Iglesias <edgar@axis.com>
Hans-Peter Nilsson <hp@axis.com>
* elf32-cris.c (cris_elf_relocate_section) <case R_CRIS_32_DTPREL>:
Don't subtract the size of the TLS block for non-shared objects
from the relocation.
Tom Tromey [Wed, 28 Nov 2012 18:48:38 +0000 (18:48 +0000)]
PR gdb/14290:
* solib-darwin.c (gdb_bfd_mach_o_fat_extract): New function.
(darwin_solib_get_all_image_info_addr_at_init, darwin_bfd_open):
Use it.
* gdb_bfd.h (gdb_bfd_mark_parent): Declare.
* gdb_bfd.c (gdb_bfd_mark_parent): New function.
(gdb_bfd_openr_next_archived_file): Use it.
Roland McGrath [Tue, 27 Nov 2012 17:26:11 +0000 (17:26 +0000)]
binutils/testsuite/
* lib/binutils-common.exp (is_zlib_supported): New function.
* lib/utils-lib.exp (run_dump_test): If as options include
--compress-debug-sections and zlib is not available, report
the test as unsupported.
* binutils-all/compress.exp: Bail out if zlib is not available.
* binutils-all/objdump.exp (objdump compressed debug):
Mark unsupported if zlib is not available.
* binutils-all/readelf.exp (readelf_compressed_wa_test): Likewise.
gas/testsuite/
* lib/gas-defs.exp (run_dump_test): If as options include
--compress-debug-sections and zlib is not available, report
the test as unsupported.
ld/testsuite/
* ld-elf/compress.exp: Bail out if zlib is not supported.
* lib/ld-lib.exp (run_dump_test): If as options include
--compress-debug-sections and zlib is not available, report
the test as unsupported.
The recent patch for %fsr handling had a couple of tiny mistakes
that cause the build to fail on sparc-solaris. Fixed thusly.
gdb/ChangeLog:
* sparc-sol2-nat.c (supply_gregset): Fix first parameter in
call to sparc_supply_fpregset.
(fill_fpregset): Fix first parameter in call to
sparc_collect_fpregset.
Tested on sparc-solaris by rebuilding GDB. Will check it in.
Yao Qi [Tue, 27 Nov 2012 07:59:12 +0000 (07:59 +0000)]
gdb/
2012-11-27 Daniel Jacobowitz <dan@codesourcery.com>
Yao Qi <yao@codesourcery.com>
* eval.c (evaluate_subexp_standard): Add handling of
TYPE_CODE_MEMBERPTR when calling functions. Correct the
result of ptype for calling a TYPE_CODE_METHODPTR.
gdb/testsuite/
2012-11-27 Daniel Jacobowitz <dan@codesourcery.com>
* gdb.cp/member-ptr.cc (class Diamond): Add func_ptr.
(func): New function.
(main): Initialize diamond.func_ptr and add diamond_pfunc_ptr.
* gdb.cp/member-ptr.exp: Add new tests for ptype and for
pointers to members with pointer-to-function type.
Joel Brobecker [Mon, 26 Nov 2012 17:41:44 +0000 (17:41 +0000)]
Do not crash when calling GDB with empty executable name.
% gdb -q ''
... or ...
% gdb -q
(gdb) file ''
... both cause GDB to crash with an invalid free. This is because
exec_file_attach is attempting to free a string that has not been
allocated. The string is only allocated if openp is successful.
But in the case of this obviously invalid filename, openp fails,
and leaves scratch_pathname uninitialized, thus causing the xfree
to fail.
The fix is to enable the associated cleanup after we have verified
that openp was successful.
gdb/ChangeLog (By Keith Seitz <keiths@redhat.com>):
* exec.c (exec_file_attach): Move cleanup after verifying that
memory has in fact been allocated.
H.J. Lu [Wed, 21 Nov 2012 21:31:18 +0000 (21:31 +0000)]
Set output maxpagesize when rewriting program header
bfd/
PR binutils/14493
* elf.c (copy_elf_program_header): When rewriting program
header, set the output maxpagesize to the maximum alignment
of input PT_LOAD segments.
Roland McGrath [Wed, 21 Nov 2012 20:13:04 +0000 (20:13 +0000)]
bfd/
* elf-nacl.c (segment_nonexecutable_and_has_contents): Renamed to ...
(segment_eligible_for_headers): ... this. Take new arguments
MAXPAGESIZE and SIZEOF_HEADERS. Return false if the first section's
start address doesn't leave space for the headers.
(nacl_modify_segment_map): Update caller.
Roland McGrath [Wed, 21 Nov 2012 19:56:38 +0000 (19:56 +0000)]
ld/
* ld.h (ld_config_type): New flag member separate_code.
* emultempl/elf32.em
(gld${EMULATION_NAME}_before_parse): Set it based on $SEPARATE_CODE.
* ldlang.c (ldlang_override_segment_assignment): If it's set, then
always return TRUE when SEC_CODE differs between the sections.
Mike Frysinger [Tue, 20 Nov 2012 21:02:36 +0000 (21:02 +0000)]
gdb: include help aliases in help command completion
There are a bunch of aliases that get used with help, but the current
command completion logic does not include those when doing completions.
Since the framework is already mostly in place, extend complete_on_cmdlist
slightly to pass down the ignore_help_classes flag like is done with the
existing lookup command logic.
Now you can do:
(gdb) help use<tab>
and get back:
(gdb) help user-defined
Yao Qi [Tue, 20 Nov 2012 02:55:10 +0000 (02:55 +0000)]
gdb/testsuite/
2012-11-20 Yao Qi <yao@codesourcery.com>
* gdb.mi/mi-cmd-param-changed.exp (test_command_param_changed):
Don't test 'maint set show-debug-regs'.
Use command 'set remotecache' instead of
'set circular-trace-buffer'.
Keith Seitz [Fri, 16 Nov 2012 20:54:30 +0000 (20:54 +0000)]
PR c++/13615
* cp-namespace.c (cp_lookup_symbol_in_namespace): Add SEARCH
parameter and pass it to lookup_symbol_file.
(cp_lookup_symbol_imports): Tell cp_lookup_symbol_in_namespace
to search base classes.
(cp_lookup_symbol_namespace): Likewise.
(lookup_namespace_scope): Likewise.
(lookup_symbol_file): Add SEARCH parameter.
If SEARCH is non-zero and no symbol is found, lookup the class
and call cp_lookup_nested_symbol.
(find_symbol_in_baseclass): New function.
(cp_lookup_nested_symbol): Do not let
cp_lookup_symbol_in_namespace search through base classes.
Do that later when there is no global symbol match.
PR c++/13615
* gdb.cp/baseenum.cc: New file.
* gdb.cp/baseenum.exp: New file.
* gdb.cp/derivation.cc (A): Add copyright.
Add a typedef.
(B): Use A::value_type instead of int. Change all references.
(D): Use value_type instead of int. Change all references.
(E): Likewise.
(F); Likewise.
(Z): New class.
(ZZ): New class.
(N, Base, Derived): New namespace and classes.
(main): Add instances of Z and ZZ.
Make sure all symbols from N are kept.
* gdb.cp/derivation.exp: Update typedef changes in tests.
Add tests for class typedefs both before and after starting
the inferior.
Add tests for searching for a typedef while stopped in a
method.
Doug Evans [Fri, 16 Nov 2012 19:43:39 +0000 (19:43 +0000)]
* main.c (gdb_datadir_provided): New static global.
(get_init_files): If --data-directory is provided,
and SYSTEM_GDBINIT lives in data-directory, look for it there.
* NEWS: Mention it.
doc/
* gdb.texinfo (System-wide configuration): If the system-wide init
file lives in the data-directory, and --data-directory is provided,
look for it there.
Yufeng Zhang [Fri, 16 Nov 2012 18:33:00 +0000 (18:33 +0000)]
bfd/ChangeLog
2012-11-16 Joey Ye <joey.ye@arm.com>
* elf32-arm.c (elf32_arm_final_link_relocate,
case R_ARM_THM_ALU_PREL_11_0, case R_ARM_THM_PC12): Align address of
the place being relocated.
(elf32_arm_final_link_relocate, case R_ARM_THM_PC8): Align address
of the place being relocated and truncate addend.
(Pa): New macro.
projects / thirdparty / binutils-gdb.git / log
