Pedro Alves [Thu, 23 May 2013 17:17:50 +0000 (17:17 +0000)]
range stepping: gdbserver (x86 GNU/Linux)
This patch adds support for range stepping to GDBserver, teaching it
about vCont;r.
It'd be easy to enable this for all hardware single-step targets
without needing the linux_target_ops hook, however, at least PPC needs
special care, due to the fact that PPC atomic sequences can't be
hardware single-stepped through, a thing which GDBserver doesn't know
about. So this leaves the support limited to x86/x86_64.
gdb/
2013-05-23 Pedro Alves <palves@redhat.com>
* NEWS: Mention GDBserver range stepping support.
gdb/gdbserver/
2013-05-23 Yao Qi <yao@codesourcery.com>
Pedro Alves <palves@redhat.com>
* linux-low.c (lwp_in_step_range): New function.
(linux_wait_1): If the thread was range stepping and stopped
outside the stepping range, report the stop to GDB. Otherwise,
continue stepping. Add range stepping debug output.
(linux_set_resume_request): Copy the step range from the resume
request to the lwp.
(linux_supports_range_stepping): New.
(linux_target_ops) <supports_range_stepping>: Set to
linux_supports_range_stepping.
* linux-low.h (struct linux_target_ops)
<supports_range_stepping>: New field.
(struct lwp_info) <step_range_start, step_range_end>: New fields.
* linux-x86-low.c (x86_supports_range_stepping): New.
(the_low_target) <supports_range_stepping>: Set to
x86_supports_range_stepping.
* server.c (handle_v_cont): Handle 'r' action.
(handle_v_requests): Append ";r" if the target supports range
stepping.
* target.h (struct thread_resume) <step_range_start,
step_range_end>: New fields.
(struct target_ops) <supports_range_stepping>:
New field.
(target_supports_range_stepping): New macro.
Pedro Alves [Thu, 23 May 2013 17:15:35 +0000 (17:15 +0000)]
range stepping: gdb
This patch teaches GDB to take advantage of target-assisted range
stepping. It adds a new 'r ADDR1,ADDR2' action to vCont (vCont;r),
meaning, "step once, and keep stepping as long as the thread is in the
[ADDR1,ADDR2) range".
Rationale:
When user issues the "step" command on the following line of source,
a = b + c + d * e - a;
GDB single-steps every single instruction until the program reaches a
new different line. E.g., on x86_64, that line compiles to:
IOW, a lot of roundtrips between GDB and GDBserver.
If we add a new command to the RSP, meaning "keep stepping and don't
report a stop until the program goes out of the [0x08048434,
0x08048451) address range", then the RSP traffic can be reduced down
to:
As number of packets is reduced dramatically, the performance of
stepping source lines is much improved.
In case something is wrong with range stepping on the stub side, the
debug info or even gdb, this adds a "set/show range-stepping" command
to be able to turn range stepping off.
gdb/
2013-05-23 Yao Qi <yao@codesourcery.com>
Pedro Alves <palves@redhat.com>
* gdbthread.h (struct thread_control_state) <may_range_step>: New
field.
* infcmd.c (step_once, until_next_command): Enable range stepping.
* infrun.c (displaced_step_prepare): Disable range stepping.
(resume): Disable range stepping if stepping over a breakpoint or
we have software watchpoints. If range stepping is enabled,
assert the thread is in the stepping range.
(clear_proceed_status_thread): Clear may_range_step.
(handle_inferior_event): Disable range stepping as soon as we know
the thread that hit the event. Re-enable it whenever we're going
to step with a step range.
* remote.c (struct vCont_action_support) <r>: New field.
(use_range_stepping): New global.
(remote_vcont_probe): Handle 'r' action.
(append_resumption): Append an 'r' action if the thread may range
step.
(show_range_stepping): New function.
(set_range_stepping): New function.
(_initialize_remote): Call add_setshow_boolean_cmd to register the
'set range-stepping' and 'show range-stepping' commands.
* NEWS: Mention range stepping, the new vCont;r action, and the
new "set/show range-stepping" commands.
gdb/doc/
2013-05-23 Yao Qi <yao@codesourcery.com>
Pedro Alves <palves@redhat.com>
* gdb.texinfo (Packets): Document 'vCont;r'.
(Continuing and Stepping): Document target-assisted range
stepping, and the 'set range-stepping' and 'show range-stepping'
commands.
Pedro Alves [Thu, 23 May 2013 17:12:51 +0000 (17:12 +0000)]
Factor out in-stepping-range checks.
This adds a function for doing within-thread's-stepping-range checks,
and converts a couple spots to use it. Following patches will add
more uses.
gdb/
2013-05-23 Yao Qi <yao@codesourcery.com>
Pedro Alves <palves@redhat.com>
* gdbthread.h (pc_in_thread_step_range): New declaration.
* thread.c (pc_in_thread_step_range): New function.
* infrun.c (handle_inferior_event): Use it.
Joel Brobecker [Thu, 23 May 2013 06:00:53 +0000 (06:00 +0000)]
Document new <data-dir>/system-gdbinit area
gdb/doc/ChangeLog:
* gdb.texinfo (Installed System-wide Configuration Scripts):
Add subsection describing the scripts now available under
the data-dir's system-gdbbinit subdirectory.
* NEWS: Add entry announcing the availability of system-wide
configuration scripts for ElinOS and Wind River Linux.
Alan Modra [Thu, 23 May 2013 03:35:59 +0000 (03:35 +0000)]
* format.c (bfd_check_format_matches): Don't match a target in
targ_selvecs if some other target is a better match. If
targets implement match priority, fall back to the first of
the best matches.
Keith Seitz [Wed, 22 May 2013 21:16:18 +0000 (21:16 +0000)]
* ada-lang.c (is_known_support_routine): Add explicit free of
'func_name' from find_frame_funname.
(ada_unhandled_exception_name_addr_from_raise): Add cleanups
for func_name from find_frame_funname.
* python/py-frame.c (frapy_name): Add explicit free of
'name' from find_frame_funname.
* stack.c (find_frame_funname): Add comment explaining that
funcp must be freed by the caller.
Return copy of symbol names instead of pointers.
(print_frame): Add a cleanup for 'funname' from
find_frame_funname.
* stack.h (find_frame_funname): Remove "const" from
'funname' parameter.
Tom Tromey [Wed, 22 May 2013 20:51:49 +0000 (20:51 +0000)]
PR c++/15401:
* c-valprint.c (c_value_print): Use value_addr for
references. Convert back to reference type with value_ref.
gdb/testsuite
* gdb.cp/class2.cc (main): New local 'aref'.
* gdb.cp/class2.exp: Check printing of 'aref'.
gas/
2013-05-22 Jürgen Urban <JuergenUrban@gmx.de>
* config/tc-mips.c (macro): Handle M_LQC2_AB and M_SQC2_AB.
gas/testsuite/
2013-05-22 Jürgen Urban <JuergenUrban@gmx.de>
* gas/mips/r5900-full.s, gas/mips/r5900-full.d: Add tests for LQ
and SQ macros.
* gas/mips/r5900-vu0.s, gas/mips/r5900-vu0.d: New test.
* gas/mips/mips.exp: Run it.
Eli Zaretskii [Wed, 22 May 2013 16:18:12 +0000 (16:18 +0000)]
Fix reporting of DLL unload events on MS-Windows.
gdb/windows-nat.c (handle_unload_dll): Don't call solib_add for the
unloaded DLL, it will be done by handle_solib_event. See
http://sourceware.org/ml/gdb-patches/2013-05/msg00713.html for the
details.
Phil Muldoon [Wed, 22 May 2013 12:27:13 +0000 (12:27 +0000)]
2013-05-22 Phil Muldoon <pmuldoon@redhat.com>
* ui-out.c: Create typedef ui_out_level_p and define vector
operations for that type.
(struct ui_out): Use a vector instead of an array.
(current_level): Return level from a vector.
(push_level): Create a level in a vector.
(pop_level): Delete a level in a vector.
(ui_out_new): Create initial level zero level, and store in a
vector.
(ui_out_destroy): Add vector cleanup.
Pedro Alves [Wed, 22 May 2013 09:31:44 +0000 (09:31 +0000)]
Let the ARI know gdb_Py_DECREF is OK.
The ARI complains with:
> gdb/python/python-internal.h:177: code: editCase function: Function name starts lower case but has uppercased letters.
gdb/python/python-internal.h:177:gdb_Py_DECREF (void *op)
gdb_Py_DECREF is just wrapping a python macro that happens to be mixed case.
gdb/
2013-05-22 Pedro Alves <palves@redhat.com>
* python/python-internal.h (gdb_Py_DECREF): Tag with
"ARI: editCase function".
Cary Coutant [Tue, 21 May 2013 21:14:40 +0000 (21:14 +0000)]
gold/
* symtab.h (Symbol::is_cxx_vtable): New function.
* target-reloc.h (relocate_section): Check for vtable symbol.
* testsuite/Makefile.am (missing_key_func.sh): New test case.
* testsuite/Makefile.in: Regenerate.
* testsuite/missing_key_func.cc: New test source.
* testsuite/missing_key_func.sh: New test script.
Cary Coutant [Tue, 21 May 2013 20:56:13 +0000 (20:56 +0000)]
gold/
* object.cc (Sized_relobj_file::get_symbol_location_info): Set
type of enclosing symbol.
(Relocate_info::location): Check symbol type when describing symbol.
* object.h (Symbol_location_info): Remove unused line_number;
add enclosing_symbol_type.
* testsuite/debug_msg.sh: Adjust expected output.
Pedro Alves [Tue, 21 May 2013 20:53:21 +0000 (20:53 +0000)]
py_decref: Don't check for NULL before calling Py_DECREF.
The only difference between Py_DECREF and Py_XDECREF is that the latter allows passing
in a NULL object, while the former prohibits it. Given that, it's natural to expect
the same from py_decref vs py_xdecref.
gdb/
2013-05-21 Pedro Alves <palves@redhat.com>
* python/py-prettyprint.c (apply_val_pretty_printer): Check
whether PRINTER is NULL before installing a Py_DECREF cleanup.
* python/py-utils.c (py_decref): Don't check for NULL before
calling Py_DECREF.
Pedro Alves [Tue, 21 May 2013 20:52:30 +0000 (20:52 +0000)]
Centralize workaround for Python 2.6's Py_DECREF.
Wrap/redefine Py_DECREF ourselves, avoiding the need for uses to care
about extra braces due to the fact that Python only started wrapping Py_DECREF
in 'do {} while (0)' after 2.6.
gdb/
2013-05-21 Pedro Alves <palves@redhat.com>
* python/py-utils.c (py_decref): Remove extra braces.
(gdb_pymodule_addobject): Remove extra braces.
* python-internal.h (gdb_Py_DECREF): New static inline function.
(Py_DECREF): Redefine as calling gdb_Py_DECREF.
Keith Seitz [Tue, 21 May 2013 19:11:50 +0000 (19:11 +0000)]
* gdb.base/filesym.exp: Use gdb_test_multiple instead of
gdb_expect.
Add test to flush the remaining input buffer so that this
file passes testsuite/12649.
Jan Kratochvil [Tue, 21 May 2013 08:16:10 +0000 (08:16 +0000)]
gdb/
Code cleanup: constification.
* solib.c (solib_ops): Make return type and ops variable type const.
(set_solib_ops): Make the new_ops parameter and ops variable const.
(solib_find, solib_map_sections, clear_so, free_so, update_solib_list)
(solib_add, solib_keep_data_in_core, clear_solib)
(solib_create_inferior_hook, in_solib_dynsym_resolve_code)
(reload_shared_libraries, solib_global_lookup): Make the ops variable
const.
* solib.h (set_solib_ops): Make the new_ops parameter const.
Joel Brobecker [Tue, 21 May 2013 06:50:12 +0000 (06:50 +0000)]
Add new system-gdbinit infrastructure
gdb/ChangeLog:
* data-directory/Makefile.in (SYSTEM_GDBINIT_SRCDIR): New
variable.
(VPATH): Add SYSTEM_GDBINIT_SRCDIR.
(SYSTEM_GDBINIT_DIR, SYSTEM_GDBINIT_INSTALL_DIR)
(SYSTEM_GDBINIT_FILES): New variables.
(all): Add stamp-system-gdbinit.
(stamp-system-gdbinit): New rule.
(clean-system-gdbinit, install-system-gdbinit)
(uninstall-system-gdbinit): New rules. Make them .PHONY.
(install-only): Add dependency on install-system-gdbinit.
(uninstall): Add dependency on uninstall-system-gdbinit.
(clean): Add dependency on clean-system-gdbinit.
* system-gdbinit/elinos.py: New file.
* system-gdbinit/wrs-linux.py: New file.
Tom Tromey [Mon, 20 May 2013 20:24:49 +0000 (20:24 +0000)]
* python/py-inferior.c (gdbpy_inferiors): Update. Hoist
get_addr_from_python calls out of TRY_CATCH.
(infpy_write_memory, infpy_search_memory): Likewise.
* python/py-utils.c (get_addr_from_python): Return negative
value on error. Use TRY_CATCH.
* python/python-internal.h (get_addr_from_python): Use
CPYCHECKER_NEGATIVE_RESULT_SETS_EXCEPTION.
Tom Tromey [Mon, 20 May 2013 20:19:54 +0000 (20:19 +0000)]
* python/py-event.h (gdbpy_initialize_event_generic): Use
CPYCHECKER_NEGATIVE_RESULT_SETS_EXCEPTION.
* python/py-evts.c (add_new_registry): Use
CPYCHECKER_NEGATIVE_RESULT_SETS_EXCEPTION.
* python/python-internal.h
(CPYCHECKER_NEGATIVE_RESULT_SETS_EXCEPTION): New macro.
Tom Tromey [Mon, 20 May 2013 20:16:57 +0000 (20:16 +0000)]
* python/py-event.h (evpy_emit_event): Use
CPYCHECKER_STEALS_REFERENCE_TO_ARG.
* python/python-internal.h (CPYCHECKER_STEALS_REFERENCE_TO_ARG):
New macro.
Tom Tromey [Mon, 20 May 2013 20:10:03 +0000 (20:10 +0000)]
* python/py-threadevent.c (get_event_thread): Use
CPYCHECKER_RETURNS_BORROWED_REF.
* python/python-internal.h (CPYCHECKER_RETURNS_BORROWED_REF):
New define.
(pspace_to_pspace_object, objfile_to_objfile_object)
(find_thread_object): Use it.
Doug Evans [Mon, 20 May 2013 17:24:21 +0000 (17:24 +0000)]
When reading CU, stay in DWO. Be more tolerent of bad debug info.
For Fission.
* dwarf2read.c (struct dwarf2_per_cu_data): New member
reading_dwo_directly.
(struct signatured_type): New member dwo_unit.
(struct die_reader_specs): New member comp_dir.
(create_signatured_type_table_from_index): Use malloc for
all_type_units instead of objfile's obstack.
(create_all_type_units): Ditto.
(fill_in_sig_entry_from_dwo_entry): New function.
(add_type_unit): New function.
(lookup_dwo_signatured_type): New function.
(lookup_dwp_signatured_type): New function.
(lookup_signatured_type): New arg cu. All callers updated.
(init_cu_die_reader): Initialize comp_dir.
(read_cutu_die_from_dwo): New arg stub_comp_dir. All callers updated.
Change assert of matching type signatures to call error on mismatch.
(lookup_dwo_unit): Add assert.
(init_tu_and_read_dwo_dies): New function.
(init_cutu_and_read_dies): Call it.
(build_type_unit_groups): Handle case of no type unit groups created.
(hash_dwo_file, eq_dwo_file): Handle missing comp_dir.
(lookup_dwo_cutu): Tweak complaint.
(dwarf2_free_abbrev_table): Check for NULL abbrev_table.
(dwarf2_per_objfile_free): Free all_type_units.
Alan Modra [Mon, 20 May 2013 10:16:31 +0000 (10:16 +0000)]
* config/tc-ppc.c (md_apply_fix): Hoist code common to insn
and data fixups performing shift/high adjust/sign extension on
fieldval. Sink fx_pcrel handling and checks. Use fixP->fx_size
when writing data fixups rather than recalculating size.
Joel Brobecker [Mon, 20 May 2013 09:45:13 +0000 (09:45 +0000)]
[dwarf] Mark all functions as prototyped except C functions.
This makes sure that the types of the arguments are taken into account
when performing an inferior function call to a non-C (or C-like)
function. In particular, this makes sure that the arguments are
appropriatly converted to the correct type.
For instance, on x86_64-linux, with the following Ada code:
procedure Set_Float (F : Float) is
begin
Global_Float := F;
end Set_Float;
The following sequence shows that Float arguments are incorrectly
passed (Ada's Float type is the equivalent of type "float" in C):
It shows that the argument was passed as a double.
The code responsible for doing appropriate type conversions
for the arguments (value_arg_coerce) found that our function
was not prototyped, and thus could not use typing information
for the arguments. Instead, it defaulted to the value of "set
coerce-float-to-double", which by default is true, to determine
the argument type.
This patch fixes the problem by setting the PROTOTYPE flag
for all functions of any language except C and Objective C.
gdb/ChangeLog:
* dwarf2read.c (prototyped_function_p): New function.
(read_subroutine_type): Use it.
Joel Brobecker [Mon, 20 May 2013 09:14:24 +0000 (09:14 +0000)]
De-indent example code in rs6000-aix-tdep.c (ARI fix)
This patch de-indents the code provided as a comment explaining
how the code declaring the ld_info32_desc and ld_info64_desc globals
was generated. The intent is to avoid an ARI warning about a macro
not starting at column zero of the line.
gdb/ChangeLog:
* rs6000-aix-tdep.c: De-indent some example code provided
as a comment.
* ppc-linux-nat.c (ppc_linux_region_ok_for_hw_watchpoint): Check if the
region is ok for a hardware watchpoint using the new ptrace interface
on Power servers.
Doug Evans [Fri, 17 May 2013 18:09:06 +0000 (18:09 +0000)]
* NEWS: Mention new maintenance commands check-symtabs, and
expand-symtabs, and renamed check-psymtabs.
* psymtab.c (maintenance_check_psymtabs): Renamed from
maintenance_check_symtabs. Only process already-expanded symbol
tables.
(_initialize_psymtab): Update.
* symmisc.c (maintenance_check_symtabs): New function.
(maintenance_expand_name_matcher): New function
(maintenance_expand_file_matcher): New function
(maintenance_expand_symtabs): New function.
(_initialize_symmisc): Add "mt check-symtabs" and "mt expand-symtabs"
commands.
doc/
* gdb.texinfo (Maintenance Commands): Update doc for
"maint check-psymtabs". Add doc for "maint check-symtabs",
"maint expand-symtabs".
testsuite/
* gdb.base/maint.exp: Update test for "maint check-psymtabs".
Add tests for "maint check-symtabs", "maint expand-symtabs".
Phil Muldoon [Fri, 17 May 2013 08:34:18 +0000 (08:34 +0000)]
2013-05-17 Phil Muldoon <pmuldoon@redhat.com>
* frame.c (frame_stash): Convert to htab.
(frame_addr_hash): New function.
(frame_addr_hash_eq): New function.
(frame_stash_create): Convert function to create
a hash table.
(frame_stash_add): Convert function to add an entry to a hash
table.
(frame_stash_find): Convert function to search the hash table.
(frame_stash_invalidate): Convert function to empty the hash
table.
(get_frame_id): Only add to stash if a frame_id is created.
(_initialize_frame): Call frame_stash_create.