With this patch, we e.g. get:
[Switching to task 2 "task_list(1)"]
[Current task is 2 "task_list(1)"]
instead of
[Switching to task 2]
[Current task is 2]
The logic to produce the taskno optionally followed
by the task name has been factorized in the task_to_str function.
Task names are output between double quotes in the new messages,
similarly to what GDB does for thread names.
However, no quotes are put around task names in 'info tasks' Name column.
This was discussed with Tom, that preferred no quotes there, while I
was more in favour of visual consistency.
I discussed with a few more users, which led to (exactly) 50% preferring
quotes and 50% preferring no quotes :).
To arrive to the decision to remove the quotes, the following "killing args"
were used:
* To have quotes or to not have quotes, that is the question; yes
but not *THE* question :).
* If there is not a clear majority that prefers quotes, better to
not disturb the existing user basis for a (somewhat) irrelevant
aspect.
* The opinion of the reviewer has more weight.
So, compared to the previous version, this version remotes the quotes
in 'info tasks'.
It improves the alignement of 'info tasks' output.
With this patch, we get:
(gdb) info task
ID TID P-ID Pri State Name
* 1 555555759030 48 Runnable main_task
2 555555759e30 1 48 Selective Wait mit
(gdb)
instead of
(gdb) info task
ID TID P-ID Pri State Name
* 1 555555759030 48 Runnable main_task
2 555555759e30 1 48 Selective Wait mit
(gdb)
(e.g. the first one properly shows parent and priority under the
correct header).
This is version 4 of the 'task name' patch.
Compared to version 3, the changes are:
output task names between quotes but not in 'info tasks'
gdb/ChangeLog
2019-09-12 Philippe Waroquiers <philippe.waroquiers@skynet.be>
* NEWS: Announce that Ada task names are now shown at more places,
and between quotes (except in info task output).
* gdb/ada-tasks.c (task_to_str): New function.
(display_current_task_id): Call task_to_str.
(task_command_1): Likewise.
(print_ada_task_info): In non-mi mode, Properly align headers and data
when task-id length is > 9 (9 is the default for a 32 bits CORE_ADDR).
gdb/doc/ChangeLog
2019-09-12 Philippe Waroquiers <philippe.waroquiers@skynet.be>
* gdb.texinfo (Ada Tasks): Tell the task name is printed, update
examples.
gdb/testsuite/ChangeLog
2019-09-12 Philippe Waroquiers <philippe.waroquiers@skynet.be>
* gdb.ada/rdv_wait.exp: Update to new task names.
* gdb.base/task_switch_in_core.exp: Likewise.
* gdb.base/info_sources_base.c: Likewise.
For some reason, commit 906e58cab5 2008-05-15 excluded input
statements with BFD_LINKER_CREATED BFDs from being printed to map
files. This isn't ideal since it loses claimed plugin BFDs.
* ldlang.c (print_input_statement): Do not exclude linker created
BFDs.
This fixes a problem with commit 128bf1fe60, a patch I made
2019-08-06. Apparently it is possible to trigger the assertion I
added during an LTO bootstrap, something I haven't reproduced.
However, I did find a case triggered by an odd linker script feature
that allows a file to be loaded from the script without specifying
that file on the command line. Regarding input sections:
"When you use a file name which is not an archive:file specifier
and does not contain any wild card characters, the linker will
first see if you also specified the file name on the linker command
line or in an INPUT command. If you did not, the linker will
attempt to open the file as an input file, as though it appeared on
the command line."
So putting
.foo : { foo.a(*) }
into a script supposedly extracts foo.a into .foo. Except it doesn't,
since this feature is meant for object files only. Well anyway,
assuming --whole-archive was given on the command line, foo.a contains
a -flto object and no other objects involved were -flto then we'll hit
the assert due to files added like foo.a here *not* having their input
statement put on the general statement list. Why these are not put on
the statement list isn't obvious but it has been that way since commit
193c5f93a1 in 1994.
PR 24981
* ldlang.c (lang_process): Remove assertion. Comment.
It's been a long-standing nuisance that gdb reported unaligned accesses
on Solaris/SPARC as SIGSEGV, contrary to the shells and truss which
correctly report SIGBUS instead.
I could trace this down to the fault handling code in procfs.c
(procfs_target::wait): when pr_why is set to PR_FAULTED, the current
code sets the signal based on the fault number. For one, the code gets
this wrong for FLTACCESS (the unaligned access case) where it uses
SIGSEGV. What's worse, it's completely unnecessary to make up the
signal number inside gdb. Instead, it should just take what procfs
reports to avoid mismatches, which is what this patch does. I've
completely removed the explicit handling of the various fault codes: for
one, the list has already been incomplete, lacking FLTCPCOVF which
existed since at least Solaris 8. Besides, there's no reason to error
out on unknown fault codes: either the fault causes a signal which can
then be reported from procfs, or it doesn't (as for FLTPAGE) and no
reporting is necessary.
Tested on sparcv9-sun-solaris2.11 and x86_64-pc-solaris2.11. Also
spot-checked manually for a couple of cases (unaligned access, division
by 0, NULL pointer dereference).
* procfs.c (procfs_target::wait) <PR_FAULTED>: Get signal from
prstatus.pr_lwp.pr_info instead of making it up.
This simplifies memory management. I've also changed some global variables
to std::string accordingly (which store the result of these functions),
but not all because some are used with add_setshow_optional_filename_cmd
which requires a char*.
gdb/ChangeLog:
2019-09-11 Christian Biesinger <cbiesinger@google.com>
* auto-load.c (auto_load_expand_dir_vars): Update.
* defs.h (gdb_datadir): Change to std::string.
(python_libdir): Likewise.
(relocate_gdb_directory): Change return type to std::string.
* guile/guile.c (gdbscm_data_directory): Update.
(initialize_scheme_side): Update.
* jit.c (jit_reader_dir): Change to std::string.
(jit_reader_load_command): Update.
* main.c (gdb_datadir): Change to std::string.
(python_libdir): Likewise.
(set_gdb_data_directory): Update.
(relocate_path): Change to return std::string.
(relocate_gdb_directory): Change to return std::string.
(relocate_gdbinit_path_maybe_in_datadir): Update.
(captured_main_1): Update.
* python/python.c (do_start_initialization): Update.
* top.c (show_gdb_datadir): Update.
* xml-syscall.c (xml_init_syscalls_info): Update.
(init_syscalls_info): Update.
This simplifies get_init_files and makes it possible to reuse
this code in an upcoming patch for SYSTEM_GDBINIT_DIR.
gdb/ChangeLog:
2019-09-11 Christian Biesinger <cbiesinger@google.com>
* main.c (relocate_gdbinit_path_maybe_in_datadir): Factor this code
out of get_init_files.
(get_init_files): Update.
To avoid manual memory management.
Tested on buildbot.
gdb/ChangeLog:
2019-09-11 Christian Biesinger <cbiesinger@google.com>
* main.c (get_init_files): Change to use std::string.
(captured_main_1): Update.
(print_gdb_help): Update.
The code used to have undefined behaviour when template parameter is float and
host_float is NaN, because it attempted to convert NaN value to LONGEST at the
last statement. This frequently caused crashes on tests that checked "info
all-registers" (at least when the code is compiled with clang; I didn't test
with gdb).
gdb/ChangeLog:
*target-float.c (host_float_ops<T>::to_longest): Update
implementation.
This parameter is really a boolean, so change the type accordingly
and update the callers.
This is for symbol_set_names, add_psymbol_to_bcache, and
add_psymbol_to_list.
minimal_symbol_reader::record_full was already passing a bool
to symbol_set_names.
gdb/ChangeLog:
2019-09-11 Christian Biesinger <cbiesinger@google.com>
* dbxread.c (read_dbx_symtab): Update.
* dwarf2read.c (load_partial_dies): Update.
* mdebugread.c (parse_partial_symbols): Update.
(handle_psymbol_enumerators): Update.
* psympriv.h (add_psymbol_to_list): Change type of copy_names to bool.
* psymtab.c (add_psymbol_to_bcache): Likewise.
(add_psymbol_to_list): Likewise.
* symtab.c (symbol_set_names): Likewise.
* symtab.h (symbol_set_names): Likewise.
* xcoffread.c (scan_xcoff_symtab): Update.
This introduces a new bfd_set_filename function, which is then used in
various spots in gdb. This allows for the removal of some casts.
bfd/ChangeLog
2019-09-11 Tom Tromey <tom@tromey.com>
* opncls.c (bfd_set_filename): New function.
* bfd-in2.h: Regenerate.
gdb/ChangeLog
2019-09-11 Tom Tromey <tom@tromey.com>
* symfile-mem.c (symbol_file_add_from_memory): Use
bfd_set_filename.
* solib-darwin.c (darwin_bfd_open): Use bfd_set_filename.
* solib-aix.c (solib_aix_bfd_open): Use bfd_set_filename.
This constifies the argument to make_tempname and make_tempdir,
removing some casts.
I initially thought that this obsoleted the allocation in
write_archive, but write_archive closes the BFD before using the name,
so this appears not to be the case.
binutils/ChangeLog
2019-09-11 Tom Tromey <tom@tromey.com>
* objcopy.c (copy_archive): Update.
* bucomm.h (make_tempname, make_tempdir): Make argument const.
* bucomm.c (make_tempname, make_tempdir): Make argument const.
This updates the "save gdb-index" documentation to reflect the new
state of Ada support. It also corrects an existing grammatical error.
gdb/doc/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* gdb.texinfo (Index Files): Update Ada text.
This adds Ada support to the cc-with-tweaks.exp board file, so that we
can test Ada this way. The cc-with-tweaks.sh script already works
reasonably well as a wrapper for gnatmake.
gdb/testsuite/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* boards/cc-with-tweaks.exp: Set GNATMAKE_FOR_TARGET.
This patch adds support for Ada to .debug_names. I opted to leave
.gdb_index alone, because in my view it is a defunct format.
gdb/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* dwarf-index-write.c (write_psymbols): Extend error message.
(debug_names::insert): Add Ada code.
(debug_names::write_psymbols): Remove Ada check.
(debug_names) <m_string_obstack>: New member.
* dwarf2read.c (gdb_index_symbol_name_matcher): Remove.
(gdb_index_symbol_name_matcher::matches): Remove.
(mapped_index_base::find_name_components_bounds): Add "lang"
parameter.
(mapped_index_base::build_name_components): Also split names
according to Ada syntax.
(dw2_expand_symtabs_matching_symbol): Loop over languages. Change
type of "match_callback".
(check_match, check_find_bounds_finds)
(dw2_expand_symtabs_matching): Update.
(dw2_debug_names_iterator): Add new constructor.
(dw2_debug_names_map_matching_symbols): New function.
(dw2_debug_names_expand_symtabs_matching): Update.
(dwarf2_debug_names_functions): Use
dw2_debug_names_map_matching_symbols.
An internal Ada test case showed that the .debug_names code does not
compute the same list of file names as the partial symbol reader. In
particular, the partial symbol reader uses the DW_AT_name of the CU:
/* Allocate a new partial symbol table structure. */
filename = dwarf2_string_attr (comp_unit_die, DW_AT_name, cu);
if (filename == NULL)
filename = "";
pst = create_partial_symtab (per_cu, filename);
This patch changes the .debug_names reader to follow.
gdb/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* dwarf2read.c (dw2_get_file_names_reader): Add the
CU's file name to the results.
gdb/testsuite/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* gdb.ada/dgopt.exp: New file.
* gdb.ada/dgopt/x.adb: New file.
This patch further simplifies the map_matching_symbols callback, by
having it take a lookup_name_info rather than a plain string.
gdb/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* ada-lang.c (add_nonlocal_symbols): Combine calls to
map_matching_symbols. Update.
* dwarf2read.c (dw2_map_matching_symbols): Update.
* psymtab.c (match_partial_symbol): Change type; update.
(psym_map_matching_symbols): Likewise.
* symfile-debug.c (debug_qf_map_matching_symbols): Change
type; update.
* symfile.h (struct quick_symbol_functions)
<map_matching_symbols>: Change "name" to be a lookup_name_info.
Remove "match".
This introduces a new helper function,
iterate_over_symbols_terminated, and changes psym_map_matching_symbols
to use it. A subsequent patch will introduce a new user of this
function in the DWARF reader.
gdb/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* psymtab.c (map_block): Remove.
(psym_map_matching_symbols): Use iterate_over_symbols_terminated.
* symtab.c (iterate_over_symbols_terminated): New function.
* symtab.c (iterate_over_symbols_terminated): Declare.
This changes iterate_over_symbols to return a bool. This allows it to
be reused in another context in a subsequent patch.
gdb/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* ada-lang.c (ada_iterate_over_symbols): Return bool.
* language.h (struct language_defn) <la_iterate_over_symbols>:
Return bool.
* symtab.c (iterate_over_symbols): Return bool.
* symtab.h (iterate_over_symbols): Return bool.
This changes map_matching_symbols to take a
symbol_found_callback_ftype, rather than separate callback and data
parameters. This enables a future patch to clean up some existing
code so that it can more readily be shared.
gdb/ChangeLog
2019-09-10 Tom Tromey <tromey@adacore.com>
* ada-lang.c (aux_add_nonlocal_symbols): Change type.
(add_nonlocal_symbols): Update.
* dwarf2read.c (dw2_map_matching_symbols): Change type.
* psymtab.c (map_block, psym_map_matching_symbols): Change type.
* symfile-debug.c (debug_qf_map_matching_symbols): Change type.
* symfile.h (struct quick_symbol_functions) <map_matching_symbols>:
Change type of "callback". Remove "data".
This patch adds support for MVE VCTP instruction in assembler.
gas ChangeLog:
2019-09-10 Srinath Parvathaneni <srinath.parvathaneni@arm.com>
* config/tc-arm.c (M_MNEM_vctp): Add new Mnemonic.
(do_mve_vctp): Add function to encode VCTP instruction.
* testsuite/gas/arm/mve-vctp-bad.d: New test.
* testsuite/gas/arm/mve-vctp-bad.l: Likewise.
* testsuite/gas/arm/mve-vctp-bad.s: Likewise.
* testsuite/gas/arm/mve-vctp.d: Likewise.
* testsuite/gas/arm/mve-vctp.s: Likewise.
* DW_UT_skeleton and DW_UT_split_compile compilation units have dwo ids
to match the compilation unit in the skeleton and .dwo files. The dwo_id is
in the header.
Tested with CC=/usr/bin/gcc (version 8.3.0) against master branch (also with
-gsplit-dwarf and -gdwarf-4 flags) and there was no increase in the set of
tests that fails.
This is part of an effort to support DWARF 5 in gdb.
gdb/ChangeLog:
* dwarf2read.c (comp_unit_head): Update comment.
(dwarf2_dwo_name): New function declaration.
(dwarf_unit_type_name): New function declaration.
(read_comp_unit_head): Add support for new compilation units,
DW_UT_partial, DW_UT_skeleton, DW_UT_split_compile, DW_UT_split_type.
Particularly, DW_UT_skeleton and DW_UT_split_compile have dwo_id
(currently named as "signature") in their header. Also clarify error
messages.
(lookup_dwo_id): New function. Returns the dwo id of the given
compile unit.
(lookup_dwo_unit): Use the new lookup_dwo_id function.
(init_cutu_and_read_dies): Use the new dwarf2_dwo_name and lookup_dwo_id
functions.
(create_dwo_cu_reader): Use the added lookup_dwo_id function.
(dwarf2_dwo_name): Get the dwo name if present.
(dwarf_unit_type_name): Convert DW_UT_* types to string for diagnostic
purposes.
This changes the "info win" command to use ui-out. This yields
somewhat nicer table output.
gdb/ChangeLog
2019-09-09 Tom Tromey <tom@tromey.com>
* tui/tui-win.c (tui_all_windows_info): Use ui_out.
Valgrind reports the below leak.
Make the variable progname_copy static, so that Valgrind continues
to find a pointer to the memory given to Python.
Note that the comment in do_start_initialization and the Python documentation
indicates that the progname given to Py_SetProgramName cannot be freed.
However, in Python 3.7.4, Py_SetProgramName does:
void
Py_SetProgramName(const wchar_t *program_name)
{
...
PyMem_RawFree(_Py_path_config.program_name);
_Py_path_config.program_name = _PyMem_RawWcsdup(program_name);
So, it looks like 3.7.4 Python duplicates its argument, which explains
the leak found by Valgrind.
It looks better to respect the doc and not have GDB freeing the string
given to Py_SetProgramName, and avoid the leak error by declaring
the progname_copy static.
This will work with Python versions that really use this string without
duplicating it, and avoids a leak report for Python version that duplicates
it.
==4023== 200 bytes in 1 blocks are definitely lost in loss record 4,545 of 7,116^M
==4023== at 0x4C29F33: malloc (vg_replace_malloc.c:307)^M
==4023== by 0x446D27: xmalloc (alloc.c:60)^M
==4023== by 0x657C77: do_start_initialization (python.c:1610)^M
==4023== by 0x657C77: _initialize_python() (python.c:1823)^M
==4023== by 0x75FE24: initialize_all_files() (init.c:231)^M
==4023== by 0x708A94: gdb_init(char*) (top.c:2242)^M
==4023== by 0x5E7460: captured_main_1 (main.c:857)^M
==4023== by 0x5E7460: captured_main (main.c:1161)^M
==4023== by 0x5E7460: gdb_main(captured_main_args*) (main.c:1186)^M
==4023== by 0x4122D4: main (gdb.c:32)^M
gdb/ChangeLog
2019-09-09 Philippe Waroquiers <philippe.waroquiers@skynet.be>
* python/python.c (do_start_initialization): Make progname_copy static,
to avoid a leak report.
If a TUI window has a long title, it can overflow the title line.
This changes the TUI to use just the tail part of the title in this
case.
gdb/ChangeLog
2019-09-08 Tom Tromey <tom@tromey.com>
* tui/tui-wingeneral.c (box_win): Truncate long window titles.
gdb/testsuite/ChangeLog
2019-09-08 Tom Tromey <tom@tromey.com>
* gdb.tui/resize.exp: Remove setup_xfail.
* gdb.tui/regs.exp: Remove setup_xfail.
* gdb.tui/basic.exp: Remove setup_xfail.
The only two values valid to pass to the block_index parameter of
quick_symbol_functions::lookup_symbol are GLOBAL_BLOCK and STATIC_BLOCK,
part of enum block_enum. Change the type of that parameter to
block_enum.
Change also the block_index field of dw2_symtab_iterator in the same
way.. This makes it consistent with dw2_debug_names_iterator, which
already uses block_enum for its block_index field.
This is a follow-up to this thread:
https://sourceware.org/ml/gdb-patches/2019-08/msg00097.html
gdb/ChangeLog:
* dwarf2read.c (struct dw2_symtab_iterator) <block_index>:
Change type to gdb::optional<block_enum>.
(dw2_symtab_iter_init): Change block_index parameter type
to gdb::optional<block_enum>.
(dw2_lookup_symbol): Change block_index parameter
type to block_enum.c
(dw2_debug_names_lookup_symbol): Likewise.
* psymtab.c (psym_lookup_symbol): Likewise.
* symfile-debug.c (debug_qf_lookup_symbol): Likewise.
* symfile.h (struct quick_symbol_functions) <lookup_symbol>:
Likewise.
These parameters are only used as bools. This also
renames "flag" to "relocatable" to make it clearer
what it does.
gdb/ChangeLog:
2019-09-06 Christian Biesinger <cbiesinger@google.com>
* defs.h (relocate_gdb_directory): Change int to bool in
signature and rename flag to relocatable.
* main.c (relocate_path): Likewise.
(relocate_gdb_directory): Likewise.
lang_input_section_type** is better than lang_statement_union_type**
here since that is in fact the type passed to the qsort predicate,
and I think it's worth factoring out the common indirections.
* emultempl/beos.em (sort_by_file_name): Use appropriate
intermediary variables.
(sort_by_section_name): Likewise.
This restores a line that has been dropped when the auto-import feature
of the PE-COFF linker was overhauled about one year. It is necessary
for GDB to properly resolve extern symbol in DLLs.
ld/ChangeLog
* pe-dll.c (pe_find_data_imports): Replace again the original name
of the undefined symbol with the __imp_ prefixed one after it is
resolved.
In gdb.base/ui-redirect.exp, the "save breakpoint" command is used to write
the current breakpoints to a file, but the actual output is not verified.
Consequently, the test has regressed in that the "print 1" command associated
with a breakpoint on main is removed by a subsequent runto_main, which first
deletes all breakpoints:
...
(gdb) break main
Breakpoint 1 at 0x4004d7: file start.c, line 34.
(gdb) commands
Type commands for breakpoint(s) 1, one per line.
End with a line saying just "end".
> PASS: gdb.base/ui-redirect.exp: commands
print 1
> PASS: gdb.base/ui-redirect.exp: print 1
end
(gdb) PASS: gdb.base/ui-redirect.exp: end
delete breakpoints
Delete all breakpoints? (y or n) y
...
and consequently the "save breakpoint" output is missing the breakpoint
command for main:
...
break main
- commands
- print 1
- end
break foo
break bar
...
Fix this by replacing "gdb_breakpoint main" with runto_main, and verifying the
"save breakpoints" output.
Tested on x86_64-linux.
gdb/testsuite/ChangeLog:
2019-09-05 Tom de Vries <tdevries@suse.de>
* lib/gdb.exp (cmp_file_string): New proc.
* gdb.base/ui-redirect.exp: Replace "gdb_breakpoint main" with
runto_main. Verify save breakpoints output.
PC-relative relocs typically use the addend in adjusting what they are
relative to. For example:
bcl 20,31,1f
1: mflr 12
addi 12,12,xxx-1b
generates "R_PPC64_REL16 xxx+0x4" for the addi (when little-endian).
The addend reflects the fact that you want the offset relative to the
previous insn not the current one in this case.
So the question is, will we ever want to do something like that for an
instruction using R_PPC64_GOT_PCREL34? I thought so at the time I
first implemented support in ld but at the time I think the hardware
was possibly going to support pcrel+offset+reg addressing. In which
case you might want something like:
load_big_offset_into_r2
pld 3,sym-big_offset@got@pcrel(2)
which would be a way of supporting more than 8G offsets from code to
the GOT. We could do the same with
load_big_offset_into_r2
pla 9,sym-big_offset@got@pcrel
ldx 3,9,2
However, this is really a poor version of TOC-pointer relative code.
So let's go with an addend on R_PPC64_GOT_PCREL34 meaning that
sym+addend should be put in a GOT entry, and the relocation calculate
the pc-relative offset to that GOT entry.
Note that this is an extension to the ABI, which says (by the
expression given for GOT relocs) that non-zero addends on GOT and PLT
relocs are ignored. This is true for all GOT/PLT relocs, not just the
pcrel ones.
* elf64-ppc.c (ppc64_elf_check_relocs): Interpret an addend in
GOT_PCREL and PLT_PCREL relocs as affecting the value stored
in the GOT/PLT entry rather than affecting the offset to that
GOI/PLT entry.
(ppc64_elf_edit_toc, ppc64_elf_relocate_section): Likewise.
The loads and stores handled in the second instruction of a sequence
marked by R_PPC64_PCREL_OPT may be a prefix instruction. For example:
pld ra,symbol@got@pcrel
0:
pld rt,off(ra)
.reloc 0b-8,R_PPC64_PCREL_OPT,(.-8)-(0b-8)
can be optimised to
pld rt,symbol+off@pcrel
pnop
* elf64-ppc.c (xlate_pcrel_opt): Handle prefix loads and stores
in second instruction.
(ppc64_elf_relocate_section): Likewise.
