When building gdb on ubuntu 16.04 with gcc 5.4.0, and running the gdb
testsuite we run into:
...
FAIL: gdb.linespec/explicit.exp: complete after -line: \
cmd complete "b -line argument " (timeout)
...
The failure is reproducible outside the testsuite like this:
...
$ gdb -q build/gdb/testsuite/outputs/gdb.linespec/explicit/explicit \
-ex "complete b -line argument"
Reading symbols from \
build/gdb/testsuite/outputs/gdb.linespec/explicit/explicit...
terminate called after throwing an instance of 'std::length_error'
what(): basic_string::_M_create
Aborted (core dumped)
...
The problem is here in complete_command:
...
completion_result result = complete (arg, &word, "e_char);
std::string arg_prefix (arg, word - arg);
if (result.number_matches != 0)
...
The problem is that the word variable is not initialized when
result.number_matches == 0, but the variable is still used in the arg_prefix
initialization.
Fix this by guarding the arg_prefix initialization with the
'result.number_matches != 0' test.
Build and tested on x86_64-linux.
gdb/ChangeLog:
2019-05-21 Tom de Vries <tdevries@suse.de>
PR cli/24587
* cli/cli-cmds.c (complete_command): Fix use of unitialized variable.
When building gdb on ubuntu 16.04 with gcc 5.4.0, and running the gdb
testsuite we run into failures due test-cases requiring at least c++1.
Fix this by adding -std=c++11 to those test-cases.
Tested on x86_64-linux.
gdb/testsuite/ChangeLog:
2019-05-21 Tom de Vries <tdevries@suse.de>
* gdb.arch/amd64-eval.exp: Require c++11.
* gdb.base/max-depth.exp: Same.
* gdb.compile/compile-cplus-array-decay.exp: Same.
* gdb.cp/meth-typedefs.exp: Same.
* gdb.cp/subtypes.exp: Same.
* gdb.cp/temargs.exp: Same.
This matches the kill in gdb_file_cmd, and ensures that the command is not
sent to the gdb.in file.
When gdb.in is used as a batch file, any kill commands run before the target
is started will cause gdb to stop processing commands.
gdb/testsuite/ChangeLog:
* lib/gdbserver-support.exp (gdbserver_run): Mark kill as optional.
PR gdb/18644 is caused by GDB using the wrong floating point format
for gfortran's 16-byte floating point type, including when the 16-byte
float is used as the component of a 32-byte complex type.
This commit addresses the issue in two places, first in i386-tdep.c,
there is already some code to force the use of floatformats_ia64_quad
for specific named types, this is extended to include the type names
that gfortran uses for its 16-byte floats.
Second, the builtin 16-byte float type (in f-lang.c) is changed so it
no longer uses gdbarch_long_double_format. On i386 this type is not
16-bytes, but is smaller, this is not what gfortran is expecting.
Instead we now use gdbarch_floatformat_for_type and ask for a
16-byte (128 bit) type using the common gfortran type name. This is
then spotted in i386-tdep.c (thanks to the first change above) and we
again get floatformats_ia64_quad returned.
This patch was tested on X86-64/GNU-Linux using '--target_board=unix'
and '--target_board=unix/-m32', and resolves all of the known failures
associated with PR gdb/18644. I've also added the test case from the
original bug report.
gdb/ChangeLog:
PR gdb/18644:
* f-lang.c (build_fortran_types): Use floatformats_ia64_quad for
16-byte floats.
* i386-tdep.c (i386_floatformat_for_type): Use
floatformats_ia64_quad for the 16-byte floating point component
within a fortran 32-byte complex number.
gdb/testsuite/ChangeLog:
PR gdb/18644
* gdb.fortran/complex.exp: Remove setup_kfail calls.
* gdb.fortran/printing-types.exp: Add new test.
* gdb.fortran/printing-types.f90: Add 16-byte real variable for
testing.
* gdb.fortran/type-kinds.exp (test_cast_1_to_type_kind): Remove
setup_kfail call.
Adds a constructor to 'struct cu_partial_die_info' and disables the
default constructor, preventing partially initialised instances from
being created.
Update 'find_partial_die' to return a const struct.
Users of 'find_partial_die' are updated to take account of the above
two changes.
There should be no user visible changes after this commit.
gdb/ChangeLog:
* dwarf2read.c (struct cu_partial_die_info): Add constructor,
delete default constructor.
(find_partial_die): Update to return const struct.
(partial_die_parent_scope): Move variable declaration into scope
of its use and change its type to auto.
(guess_partial_die_structure_name): Likewise.
(partial_die_info::fixup): Likewise.
find_and_open_source casts away const, but hasn't needed to in a
while. This removes the cast and a strangely hostile comment.
gdb/ChangeLog
2019-05-17 Tom Tromey <tromey@adacore.com>
* source.c (find_and_open_source): Remove cast.
I noticed that annotate_source takes a "char *", but really should
take a "const char *". This patch fixes this.
gdb/ChangeLog
2019-05-17 Tom Tromey <tromey@adacore.com>
* annotate.c (annotate_source): Make "filename" const.
* annotate.h (annotate_source): Use const.
TRANSCRIPT is superseeded by the .in, .cmd and .debug files, and
can be removed.
gdb/testsuite/ChangeLog
* README (Running the Testsuite): Change example.
(Testsuite Parameters): Remove TRANSCRIPT.
* lib/gdb.exp: Remove TRANSCRIPT check.
Add "replay" to the list of GDBSERVER_DEBUG options. This will
cause a gdbserver.replay file to be written to the test output
directory.
At the same time switch this to a comma separated list in order
to easily handle all possible options.
The replay log is created by GDB, but has been added to
GDBSERVER_DEBUG as it is only required for gdbserver tests. To
enable it, the gdb_debug_init is overridden to allow the additional
checking, before calling the original function.
gdb/testsuite/ChangeLog:
* README (Testsuite Parameters): Add replay logging to
GDBSERVER_DEBUG.
(gdbserver,debug): Refer to GDBSERVER_DEBUG.
* lib/gdbserver-support.exp (gdbserver_start): Treat gdbserverdebug
as a comma separated list.
(gdb_debug_init): Override procedure.
When spawning gdb or gdbserver create a .cmd file in the test output
directory containing the full command line, ensuring the current gdb
instance is appended to the files so that they can be quickly matched
to the corresponding gdb.in file.
gdb/testsuite/ChangeLog:
* lib/gdb.exp (default_gdb_spawn): Call gdb_write_cmd_file.
(gdb_write_cmd_file): New procedure.
* lib/gdbserver-support.exp (gdbserver_start): Call
gdbserver_write_cmd_file.
(gdbserver_write_cmd_file): New proedure.
When debugging testsuite failures, it can be awkward parsing gdb.log to
obtain all the commands run in order to manually re-run the test.
This patch adds the functionality to save all gdb commands to the file gdb.in
when the testsuite is run. The file is saved in the directory for the test and
if gdb is restarted then .1, .2, .3 etc is added to the filename.
Once a test has been run, the .in file can be used to re-run the test in the
following way:
gdb -x outputs/gdb.store/gdb.in outputs/gdb.store/store
The code works by intercepting send_gdb. I've added a TYPE to ensure that any
commands that would destroy the playback are kept from the log (for example the
Y from an answer to a y/n question).
Adds library function standard_output_file_with_gdb_instance to open a file
postfixed with count of the gdb instance. Ensure this count is reset when a new
.exp script is run.
I've re-run a random selection of .in files to check they do not error. Logs with
commands such as "attach <pid>" will not directly work when re-run.
gdb/testsuite/ChangeLog:
* lib/gdb.exp (gdb_unload): Mark Y as an answer.
(delete_breakpoints): Likewise.
(gdb_run_cmd): Likewise.
(gdb_start_cmd): Likewise.
(gdb_starti_cmd): Likewise.
(gdb_internal_error_resync): Likewise.
(gdb_test_multiple): Likewise.
(gdb_reinitialize_dir): Likewise.
(default_gdb_exit): Likewise.
(gdb_file_cmd): Mark kill as optional.
(default_gdb_start): Call gdb_stdin_log_init.
(send_gdb): Call gdb_stdin_log_write.
(rerun_to_main): Mark Y as an answer.
(gdb_stdin_log_init): New function.
(gdb_stdin_log_write): Likewise.
Fix up all failures encountered when running the testsuite with
GDB_DEBUG="infrun".
Some tests rely on enabling debugging for various components. With
debugging on, this will be lost to the debug file.
Disable separate tty for mi tests when debugging. This currently
does not work.
disasm.c should send errors to the stderr instead of the logfile.
Note that enabling debug for other components might still cause
additional errors above what has been fixed here.
gdb/ChangeLog:
* disasm.c (set_disassembler_options): Send errors to stderr.
gdb/testsuite/ChangeLog:
* gdb.base/breakpoint-in-ro-region.exp: Disable when debugging.
* gdb.base/debug-expr.exp: Likewise.
* gdb.base/foll-fork.exp: Likewise.
* gdb.base/foll-vfork.exp: Likewise.
* gdb.base/fork-print-inferior-events.exp: Likewise.
* gdb.base/gdb-sigterm.exp: Likewise.
* gdb.base/gdbinit-history.exp: Likewise.
* gdb.base/osabi.exp: Likewise.
* gdb.base/sss-bp-on-user-bp-2.exp: Likewise.
* gdb.base/ui-redirect.exp: Likewise.
* gdb.gdb/unittest.exp: Likewise.
* gdb.mi/mi-break.exp: Disable separate-mi-tty when debugging.
* gdb.mi/mi-watch.exp: Likewise.
* gdb.mi/new-ui-mi-sync.exp: Likewise.
* gdb.mi/user-selected-context-sync.exp: Likewise.
* gdb.python/python.exp: Disable debug test when debugging.
* gdb.threads/check-libthread-db.exp: Disable when debugging.
* gdb.threads/signal-while-stepping-over-bp-other-thread.exp:
Likewise.
* gdb.threads/stepi-random-signal.exp: Likewise.
Add both board option and environment variable which enables gdb
debug via a comma separated list and sends it to the file gdb.debug,
located in the output directory for the current test. Document this.
Add support for the environment variable in the Makefile.
The testsuite can be run with gdb debug enabled in the following way:
make check GDB_DEBUG="infrun,target,remote"
A Test with multiple invocations of GDB will all append debug to the
same log file.
gdb/testsuite/ChangeLog:
* Makefile.in: Pass through GDB_DEBUG.
* README (Testsuite Parameters): Add GDB_DEBUG.
(gdb,debug): Add board setting.
* lib/gdb.exp (default_gdb_start): Start debugging.
(gdb_debug_enabled): New procedure.
(gdb_debug_init): Likewise.
Currently, when logging is enabled, output will be sent to both a
logfile and standard terminal output. The redirect option sends output
only to the logfile. This includes all debug output.
Add the option to redirect debug output seperately to normal
output, using the cli command:
set logging debugredirect on
By setting this and enabling logging, all output and debug will
be sent to the logfile. The user will still see all output but
no debug output.
This causes a change in behaviour for anyone currently using
logging redirect, as now only output will be redirected. Users
will have to issue the additional command above to also redirect
debug.
Expand ui-redirect.exp cover the changes.
gdb/ChangeLog:
* cli/cli-interp.c (struct saved_output_files): Add saved entry.
(cli_interp_base::set_logging): Check debug_redirect.
* cli/cli-interp.h (set_logging): Add debug_redirect parameter.
* cli/cli-logging.c (debug_redirect): Add static variable.
(pop_output_files): Add default param.
(handle_redirections): Print debug setting.
(show_logging_command): Likewise.
(_initialize_cli_logging): Add debugredirect command.
* interps.c (current_interp_set_logging): Add debug_redirect
parameter.
* interps.h (set_logging): Add debug_redirect parameter.
(current_interp_set_logging): Likewise.
* mi/mi-common.h: Likewise.
* mi/mi-interp.c (mi_interp::set_logging): Likewise.
gdb/testsuite/ChangeLog:
* gdb.base/ui-redirect.exp: Add debug redirect tests.
Instead of using two bools to decide if the files should close when tee_file
is closed, make file one stay open and file two close. This simplifies the
use cases for it.
Inline the make_logging_output into the calling functions (the logic here
looks ugly in order to simplify a later change).
Expand ui-redirect.exp to cover the changes, similar to mi-logging.exp.
gdb/ChangeLog:
* cli/cli-interp.c (cli_interp_base::set_logging): Create tee_file
directly.
* cli/cli-interp.h (make_logging_output): Remove declaration.
* cli/cli-logging.c (make_logging_output): Remove function.
* mi/mi-interp.c (mi_interp::set_logging): Create tee_file
directly.
* ui-file.c (tee_file::tee_file): Remove bools.
(tee_file::~tee_file): Remove deletes.
* ui-file.h (tee_file): Remove bools.
gdb/testsuite/ChangeLog:
* gdb.base/ui-redirect.exp: Test redirection.
There is a CLI command 'complete' intended to use with emacs. Such a command
would also be useful for MI frontends, when separate CLI and MI channels cannot
be used. For example, on Windows (because of lack of PTYs) or when GDB is used
through SSH session.
This commit adds a new '-complete' MI command.
gdb/Changelog:
2019-01-28 Jan Vrany <jan.vrany@fit.cvut.cz>
* mi/mi-cmds.h (mi_cmd_complete): New function.
* mi/mi-main.c (mi_cmd_complete): Likewise.
* mi/mi-cmds.c: Define new MI command -complete.
* NEWS: Mention new -complete command.
gdb/doc/ChangeLog:
2019-01-28 Jan Vrany <jan.vrany@fit.cvut.cz>
* gdb.texinfo (Miscellaneous GDB/MI Commands): Document new
MI command -complete.
gdb/testsuite/ChangeLog:
2019-01-28 Jan Vrany <jan.vrany@fit.cvut.cz>
* gdb.mi/mi-complete.exp: New file.
* gdb.mi/mi-complete.cc: Likewise.
Extract completion logic from CLI complete_command() into a new
helper function complete().
gdb/Changelog:
* completer.h (complete): New function.
* completer.c (complete): Likewise.
* cli/cli-cmds.c: (complete_command): Update to use new complete()
function defined in completer.h.
When running gdb using AddressSanitizer, and loading a cc1plus binary built
with profiledbootstrap and -flto, we run into a heap-use-after-free error:
...
$ LD_PRELOAD=/usr/lib64/libasan.so.3 ./gdb -batch cc1plus
==26855==ERROR: AddressSanitizer: heap-use-after-free on address \
0x62100ad8a8b0 at pc 0x7f13803cc9e3 bp 0x7ffe55b0d090 sp 0x7ffe55b0c840
READ of size 47 at 0x62100ad8a8b0 thread T0
#0 0x7f13803cc9e2 (/usr/lib64/libasan.so.3+0x3e9e2)
#1 0x5e7a0d in typename_concat gdb/dwarf2read.c:22661
#2 0x5c6437 in partial_die_full_name gdb/dwarf2read.c:8876
#3 0x5c6555 in add_partial_symbol gdb/dwarf2read.c:8893
#4 0x5c6ecf in add_partial_subprogram gdb/dwarf2read.c:9156
#5 0x5c5e90 in scan_partial_symbols gdb/dwarf2read.c:8668
#6 0x5c6c0a in add_partial_namespace gdb/dwarf2read.c:9081
#7 0x5c5f99 in scan_partial_symbols gdb/dwarf2read.c:8702
#8 0x5c48b6 in process_psymtab_comp_unit_reader gdb/dwarf2read.c:8056
#9 0x5c3c1f in init_cutu_and_read_dies gdb/dwarf2read.c:7689
#10 0x5c4c03 in process_psymtab_comp_unit gdb/dwarf2read.c:8140
#11 0x5c58a2 in dwarf2_build_psymtabs_hard gdb/dwarf2read.c:8500
#12 0x5c0d03 in dwarf2_build_psymtabs(objfile*) gdb/dwarf2read.c:6337
#13 0x612359 in read_psyms gdb/elfread.c:1311
#14 0x798a64 in require_partial_symbols(objfile*, int) gdb/psymtab.c:115
#15 0x867d7b in read_symbols gdb/symfile.c:821
#16 0x8683d9 in syms_from_objfile_1 gdb/symfile.c:1000
#17 0x8684a1 in syms_from_objfile gdb/symfile.c:1017
#18 0x868873 in symbol_file_add_with_addrs gdb/symfile.c:1124
#19 0x868b0a in symbol_file_add_from_bfd(bfd*, char const*, \
enum_flags<symfile_add_flag>, std::vector<other_sections, \
std::allocator<other_sections> >*, \
enum_flags<objfile_flag>, objfile*) gdb/symfile.c:1204
#20 0x868b64 in symbol_file_add(char const*, \
enum_flags<symfile_add_flag>, \
std::vector<other_sections, \
std::allocator<other_sections> >*, \
enum_flags<objfile_flag>) gdb/symfile.c:1217
#21 0x868c39 in symbol_file_add_main_1 gdb/symfile.c:1240
#22 0x868bd0 in symbol_file_add_main(char const*, \
enum_flags<symfile_add_flag>) gdb/symfile.c:1231
#23 0x71f1b2 in symbol_file_add_main_adapter gdb/main.c:395
#24 0x71f10e in catch_command_errors gdb/main.c:372
#25 0x71ff5f in captured_main_1 gdb/main.c:1043
#26 0x72045d in captured_main gdb/main.c:1163
#27 0x7204c8 in gdb_main(captured_main_args*) gdb/main.c:1188
#28 0x40fd7d in main gdb/gdb.c:32
#29 0x7f137e300f49 in __libc_start_main (/lib64/libc.so.6+0x20f49)
#30 0x40fc89 in _start (/data/gdb_versions/devel/build/gdb/gdb+0x40fc89)
0x62100ad8a8b0 is located 944 bytes inside of 4064-byte region \
[0x62100ad8a500,0x62100ad8b4e0)
freed by thread T0 here:
#0 0x7f13804523a0 in __interceptor_free (/usr/lib64/libasan.so.3+0xc43a0)
#1 0x435e44 in xfree<void> gdb/common/common-utils.h:60
#2 0xa82c25 in call_freefun libiberty/obstack.c:103
#3 0xa83098 in _obstack_free libiberty/obstack.c:280
#4 0x4367da in auto_obstack::~auto_obstack() gdb/gdb_obstack.h:101
#5 0x5ed72c in dwarf2_cu::~dwarf2_cu() gdb/dwarf2read.c:25341
#6 0x5fb5bb in std::default_delete<dwarf2_cu>::operator()(dwarf2_cu*) const \
/usr/include/c++/7/bits/unique_ptr.h:78
#7 0x5f7334 in std::unique_ptr<dwarf2_cu, \
std::default_delete<dwarf2_cu> >::~unique_ptr() \
/usr/include/c++/7/bits/unique_ptr.h:268
#8 0x5c3ce5 in init_cutu_and_read_dies gdb/dwarf2read.c:7624
#9 0x5c4c03 in process_psymtab_comp_unit gdb/dwarf2read.c:8140
#10 0x5c58a2 in dwarf2_build_psymtabs_hard gdb/dwarf2read.c:8500
#11 0x5c0d03 in dwarf2_build_psymtabs(objfile*) gdb/dwarf2read.c:6337
#12 0x612359 in read_psyms gdb/elfread.c:1311
#13 0x798a64 in require_partial_symbols(objfile*, int) gdb/psymtab.c:115
#14 0x867d7b in read_symbols gdb/symfile.c:821
#15 0x8683d9 in syms_from_objfile_1 gdb/symfile.c:1000
#16 0x8684a1 in syms_from_objfile gdb/symfile.c:1017
#17 0x868873 in symbol_file_add_with_addrs gdb/symfile.c:1124
#18 0x868b0a in symbol_file_add_from_bfd(bfd*, char const*, \
enum_flags<symfile_add_flag>, std::vector<other_sections, \
std::allocator<other_sections> >*, \
enum_flags<objfile_flag>, objfile*) gdb/symfile.c:1204
#19 0x868b64 in symbol_file_add(char const*, \
enum_flags<symfile_add_flag>, std::vector<other_sections, \
std::allocator<other_sections> >*, \
enum_flags<objfile_flag>) gdb/symfile.c:1217
#20 0x868c39 in symbol_file_add_main_1 gdb/symfile.c:1240
#21 0x868bd0 in symbol_file_add_main(char const*, \
enum_flags<symfile_add_flag>) gdb/symfile.c:1231
#22 0x71f1b2 in symbol_file_add_main_adapter gdb/main.c:395
#23 0x71f10e in catch_command_errors gdb/main.c:372
#24 0x71ff5f in captured_main_1 gdb/main.c:1043
#25 0x72045d in captured_main gdb/main.c:1163
#26 0x7204c8 in gdb_main(captured_main_args*) gdb/main.c:1188
#27 0x40fd7d in main gdb/gdb.c:32
#28 0x7f137e300f49 in __libc_start_main (/lib64/libc.so.6+0x20f49)
previously allocated by thread T0 here:
#0 0x7f13804526b8 in __interceptor_malloc (/usr/lib64/libasan.so.3+0xc46b8)
#1 0x5114b5 in xmalloc gdb/common/common-utils.c:44
#2 0xa82bd5 in call_chunkfun libiberty/obstack.c:94
#3 0xa82eda in _obstack_newchunk libiberty/obstack.c:206
#4 0x477310 in allocate_on_obstack::operator new(unsigned long, obstack*) \
gdb/gdb_obstack.h:117
#5 0x5dea8c in load_partial_dies gdb/dwarf2read.c:18571
#6 0x5c487f in process_psymtab_comp_unit_reader gdb/dwarf2read.c:8054
#7 0x5c3c1f in init_cutu_and_read_dies gdb/dwarf2read.c:7689
#8 0x5c4c03 in process_psymtab_comp_unit gdb/dwarf2read.c:8140
#9 0x5c58a2 in dwarf2_build_psymtabs_hard gdb/dwarf2read.c:8500
#10 0x5c0d03 in dwarf2_build_psymtabs(objfile*) gdb/dwarf2read.c:6337
#11 0x612359 in read_psyms gdb/elfread.c:1311
#12 0x798a64 in require_partial_symbols(objfile*, int) gdb/psymtab.c:115
#13 0x867d7b in read_symbols gdb/symfile.c:821
#14 0x8683d9 in syms_from_objfile_1 gdb/symfile.c:1000
#15 0x8684a1 in syms_from_objfile gdb/symfile.c:1017
#16 0x868873 in symbol_file_add_with_addrs gdb/symfile.c:1124
#17 0x868b0a in symbol_file_add_from_bfd(bfd*, char const*, \
enum_flags<symfile_add_flag>, \
std::vector<other_sections, \
std::allocator<other_sections> >*, \
enum_flags<objfile_flag>, objfile*) gdb/symfile.c:1204
#18 0x868b64 in symbol_file_add(char const*, enum_flags<symfile_add_flag>, \
std::vector<other_sections, \
std::allocator<other_sections> >*, \
enum_flags<objfile_flag>) gdb/symfile.c:1217
#19 0x868c39 in symbol_file_add_main_1 gdb/symfile.c:1240
#20 0x868bd0 in symbol_file_add_main(char const*, \
enum_flags<symfile_add_flag>) gdb/symfile.c:1231
#21 0x71f1b2 in symbol_file_add_main_adapter gdb/main.c:395
#22 0x71f10e in catch_command_errors gdb/main.c:372
#23 0x71ff5f in captured_main_1 gdb/main.c:1043
#24 0x72045d in captured_main gdb/main.c:1163
#25 0x7204c8 in gdb_main(captured_main_args*) gdb/main.c:1188
#26 0x40fd7d in main gdb/gdb.c:32
#27 0x7f137e300f49 in __libc_start_main (/lib64/libc.so.6+0x20f49)
...
This error happens as follows.
The function find_partial_die has a cu argument, but returns a pdi which may
or may not be from that cu:
...
/* Find a partial DIE at OFFSET, which may or may not be in CU,
except in the case of .debug_types DIEs which do not reference
outside their CU (they do however referencing other types via
DW_FORM_ref_sig8). */
static struct partial_die_info *
find_partial_die (sect_offset sect_off, int offset_in_dwz, struct dwarf2_cu *cu)
...
So the pdi returned by find_partial_die here in partial_die_parent_scope may
be from another cu:
...
partial_die_parent_scope (struct partial_die_info *pdi,
struct dwarf2_cu *cu)
{
const char *grandparent_scope;
struct partial_die_info *parent, *real_pdi;
/* We need to look at our parent DIE; if we have a DW_AT_specification,
then this means the parent of the specification DIE. */
real_pdi = pdi;
while (real_pdi->has_specification)
real_pdi = find_partial_die (real_pdi->spec_offset,
real_pdi->spec_is_dwz, cu);
parent = real_pdi->die_parent;
...
in which case both real_pdi and parent will be not from cu, but from another
one, say cu2.
Subsequently, cu's comp_unit_obstack is used to set parent->scope:
...
parent->scope = typename_concat (&cu->comp_unit_obstack,
grandparent_scope,
parent->name, 0, cu);
...
So, we use cu->comp_unit_obstack to assign a value to the scope field of
a pdi belonging to cu2, and when cu is deleted, the scope field points to a
freed value.
Fix this by making find_partial_die return the cu corresponding to the
returned pdi, and handling this at the call sites.
Tested on x86_64-linux.
gdb/ChangeLog:
2019-05-17 Tom de Vries <tdevries@suse.de>
PR gdb/24094
* dwarf2read.c (struct cu_partial_die_info): New struct.
(find_partial_die): Return cu_partial_die_info.
(partial_die_parent_scope, guess_partial_die_structure_name)
(partial_die_info::fixup): Handle new return type of find_partial_die.
This patch simplifies the code of
stap-probe.c:stap_parse_register_operand by making "regname" an
std::string. No functionality change.
I'm this code's maintainer, so I'm pushing this as it's a fairly
trivial patch.
gdb/ChangeLog:
2019-05-16 Sergio Durigan Junior <sergiodj@redhat.com>
* stap-probe.c (stap_parse_register_operand): Make "regname" an
"std::string", simplifying the algorithm.
I think the string formatting for complaints was messed up by Tom's
patch to simplify the complaint mechanism. This small and trivial
patch fixes them.
Pushed as obvious.
gdb/ChangeLog:
2019-05-16 Sergio Durigan Junior <sergiodj@redhat.com>
* stap-probe.c (handle_stap_probe): Fix complaint formatting.
(stap_static_probe_ops::get_probes): Likewise.
This patch contains three very small improvement on the logic of some
operations we do on stap-probe.c. They don't change what the code
does.
Pushed as obvious.
gdb/ChangeLog:
2019-05-16 Sergio Durigan Junior <sergiodj@redhat.com>
* stap-probe.c (stap_parse_register_operand): Make "if (*p->arg ==
'-')" and "else if".
(stap_parse_single_operand): Join checks for
"gdbarch_stap_parse_special_token_p" and
"gdbarch_stap_parse_special_token" in the same "if" statement.
Invert check when verifying for operation on register
displacement.
Some functions's comments were not entirely correct on stap-probe.c,
so this patch updates them.
Pushed as obvious.
gdb/ChangeLog:
2019-05-16 Sergio Durigan Junior <sergiodj@redhat.com>
* stap-probe.c (stap_get_opcode): Update comment.
(stap_get_expected_argument_type): Likewise.
(handle_stap_probe): Likewise.
This simple patch converts a bunch of "int"s to "bool" on stap-probe.c
and on the stap-related code present on i386-tdep.c.
Pushed as obvious (+ I'm the maintainer of this code).
gdb/ChangeLog:
2019-05-16 Sergio Durigan Junior <sergiodj@redhat.com>
* i386-tdep.c (i386_stap_parse_special_token_triplet): Change
return type to 'bool'. Adjust comment. Use 'bool' when
appropriate.
(i386_stap_parse_special_token_three_arg_disp): Likewise.
* stap-probe.c (stap_parse_argument_1): Likewise.
(stap_is_operator): Likewise.
(stap_is_generic_prefix): Likewise.
(stap_is_register_prefix): Likewise.
(stap_is_register_indirection_prefix): Likewise.
(stap_is_integer_prefix): Likewise.
(stap_generic_check_suffix): Likewise.
(stap_check_integer_suffix): Likewise.
(stap_check_register_suffix): Likewise.
(stap_check_register_indirection_suffix): Likewise.
(stap_parse_register_operand): Likewise.
(stap_parse_single_operand): Likewise.
(stap_parse_argument_1): Likewise.
(stap_probe::get_argument_count): Likewise.
(stap_is_operator): Likewise.
John Marshall pointed out that darwin-nat.c fails to build:
CXX darwin-nat.o
../../../binutils-gdb/gdb/darwin-nat.c:1709:8: error: must use 'class' tag to refer to type 'thread_info' in this scope
for (thread_info *it : all_threads ())
/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk/usr/include/mach/thread_act.h:240:15: note:
class 'thread_info' is hidden by a non-type declaration of 'thread_info' here
kern_return_t thread_info
Mach has a thread_info() function declared in that header, which darwin-nat.c #includes.
This patch fixes the problem by reintroducing the struct keyword.
gdb/ChangeLog
2019-05-16 Tom Tromey <tromey@adacore.com>
* darwin-nat.c (thread_info_from_private_thread_info): Add struct
keyword to foreach.
Add tests for calling sizeof on indexed and sliced arrays, and on
pointers to arrays. These are all things that currently work, but
were previously untested.
gdb/testsuite/ChangeLog:
* gdb.fortran/vla-sizeof.exp: Add tests of sizeof applied to
indexed and sliced arrays, and pointers to arrays.
I happened to be looking at this code and noticed we could replace ints
by bools at a few places.
gdb/ChangeLog:
* linux-thread-db.c (try_thread_db_load_1): Change return type
to bool.
(try_thread_db_load): Likewise.
(try_thread_db_load_from_pdir_1): Likewise.
(try_thread_db_load_from_pdir): Likewise.
(try_thread_db_load_from_sdir): Likewise.
(try_thread_db_load_from_dir): Likewise.
(thread_db_load_search): Likewise.
(has_libpthread): Likewise.
(thread_db_load): Likewise.
Hi,
Ref.: https://bugzilla.redhat.com/show_bug.cgi?id=1708192https://bugzilla.redhat.com/show_bug.cgi?id=1708786
During the Fedora RPM build process, gdb-add-index is invoked to
extract the DWARF index from the binary, and GDB will segfault because
dwarf2read.c:parse_definition_macro's 'body' variable is NULL.
The underlying problem is that Fedora's rpm-build's "debugedit"
program will silently corrupt .debug_macro strings when a binary is
compiled with -g3. This is being taken care of by Mark Wielaard,
here:
https://bugzilla.redhat.com/show_bug.cgi?id=1708786
However, I still feel it's important to make GDB more resilient
against invalid DWARF input, so I'm proposing this rather simple patch
to catch the situation when "body == NULL" (i.e., it's probably been
corrupted) and issue a complaint. This is not a real fix to the
problem, of course, but at least GDB is able to finish without
segfaulting.
OK for master?
gdb/ChangeLog:
2019-05-15 Sergio Durigan Junior <sergiodj@redhat.com>
Ref.: https://bugzilla.redhat.com/show_bug.cgi?id=1708192
* dwarf2read.c (dwarf_decode_macro_bytes): Check whether 'body' is
NULL, and complain if that's the case.
Previously, the stack unwinder searched through consecutive bytes for values
which it thought might be the start of a stack mutating operation.
This was error prone, because such bytes could also be the operands of other
instructions. This change uses the opcodes api to interpret the code in each
frame.
gdb/ChangeLog:
* s12z-tdep.c (push_pull_get_stack_adjustment): New function.
(advance, posn, abstract_read_memory): New functions.
[struct mem_read_abstraction]: New struct.
(s12z_frame_cache): Use opcodes API to interpret stack frame code.
coerce_unspec_val_to_type does:
set_value_address (result, value_address (val));
However, this is only valid for lval_memory. This patch changes this
code to only set the address for lval_memory values.
This seems like an ordinary oversight in coerce_unspec_val_to_type,
and a test case would be difficult to write, so I'm submitting it
without a test case.
Tested on x86-64 Fedora 29; plus using an Ada program that exhibits
the bug (but which cannot be shared).
gdb/ChangeLog
2019-05-14 Tom Tromey <tromey@adacore.com>
* ada-lang.c (coerce_unspec_val_to_type): Only set address when
value is not lval_memory.
The "automatic symbol index cache" entry in the docs is missing entries in the
concept and command indices. Add them.
gdb/doc/ChangeLog:
2019-05-14 Tom de Vries <tdevries@suse.de>
* gdb.texinfo (Automatic symbol index cache): Add concept and command
index entries.
A base address selection entry in a location list consist of two (constant or
relocated) address offsets. The two offsets are the same size as an address
on the target machine.
The test-case gdb.dwarf2/dw2-skip-prologue.S encodes a base address selection
entry using .4byte, which is incorrect for 8-byte pointer size. [ Which
triggers an assert in dwz, see PR dwz/24172. ]
Fix this by using PTRBYTE instead.
Tested on x86_64-linux.
gdb/testsuite/ChangeLog:
2019-05-14 Tom de Vries <tdevries@suse.de>
* gdb.dwarf2/dw2-skip-prologue.S (.debug_loc): Fix base address
selection entry encoding.
Currently, the entry for the command "maint info selftests" in the Command,
Variable, and Function Index is listed at '"', rather than next to the other
"maint info" commands.
Fix this by removing the superfluous quoting in the @kindex entry.
gdb/doc/ChangeLog:
2019-05-14 Tom de Vries <tdevries@suse.de>
* gdb.texinfo (Maintenance Commands): Remove superfluous
quoting on command index entry for "maint info selftests".
Since commit
b4be1b064860 ("Fix MI output for multi-location breakpoints")
we get this error when building with --disable-gdbmi:
CXXLD gdb
/home/smarchi/src/binutils-gdb/gdb/breakpoint.c:6358: error: undefined reference to 'mi_multi_location_breakpoint_output_fixed(ui_out*)'
This is due to breakpoint.c using a function defined in mi/mi-main.c, even
though mi/mi-main.c isn't included in the build.
To fix it, use the flags feature of ui_out. mi_ui_out has the new
fix_multi_location_breakpoint_output flag set for versions >= 3. Also,
move the global variable fix_multi_location_breakpoint_output to
breakpoint.c, so it can be read there even when we build without MI. I
renamed it to fix_multi_location_breakpoint_output_globally so it
doesn't clash with the new enumerator.
gdb/ChangeLog:
* breakpoint.h (fix_multi_location_breakpoint_output_globally):
New variable declaration.
* breakpoint.c (fix_multi_location_breakpoint_output_globally):
New variable.
(print_one_breakpoint): Use ui_out::test_flags and new global
variable to compute use_fixed_output.
* mi/mi-main.h (mi_multi_location_breakpoint_output_fixed):
Remove.
* mi/mi-main.c (fix_multi_location_breakpoint_output): Remove.
(mi_multi_location_breakpoint_output_fixed): Remove.
(mi_cmd_fix_multi_location_breakpoint_output): Adjust to set the
new variable.
* mi/mi-out.c (mi_ui_out::mi_ui_out): Set
fix_multi_location_breakpoint_output flag if version >= 3.
* ui-out.h (enum ui_out_flag)
<fix_multi_location_breakpoint_output>: New enumerator.
When running the gdb.base/index-cache.exp test case with the
cc-with-dwz-m board, I noticed that the final executable didn't actually
contain a .gnu_debugaltlink section with the name of the external dwz
file:
$ readelf --debug-dump=links testsuite/outputs/gdb.base/index-cache/index-cache
* empty *
Running dwz by hand, I realized it's because dwz complains that the
output .debug_info section is empty and fails:
$ gcc ~/src/binutils-gdb/gdb/testsuite/gdb.base/index-cache.c -g3 -O0 -o a && cp a b
$ dwz -m foo a b
dwz: foo: .debug_info section not present
$ echo $?
1
This is because index-cache.c is trivial (just an empty main) and dwz
doesn't find anything to factor out to the dwz file. [1]
I think that cc-with-tweaks should fail in this scenario: if the user
asks for an external dwz file to be generated (the -m flag), then it
should be an error if cc-with-tweaks doesn't manage to produce an
executable with the proper link to this external dwz file. Otherwise,
the test runs with a regular non-dwzified executable, which gives a
false sense of security about whether the feature under test works with
dwzified executables.
So this patch adds checks for that after invoking dwz. It also removes
the 2>&1 to allow the error message to be printed like so:
Running /home/smarchi/src/binutils-gdb/gdb/testsuite/gdb.base/index-cache.exp ...
gdb compile failed, dwz: /home/smarchi/build/binutils-gdb/gdb/testsuite/outputs/gdb.base/index-cache/index-cache.dwz: .debug_info section not present
- In the -m case (multi-file compression), we check if the expected output file
exists.
- In the -z case (single-file compression), we check if the file
contents has changed. This should catch cases where dwz doesn't modify the
file because it's not worth it.
It was chosen not to check for dwz's exit code, as it is not very
reliable up to dwz 0.12.
With this patch, fewer tests will pass than before with the
cc-with-dwz and cc-with-dwz-m boards, but those were false positives
anyway, as the test ran with regular executables.
[1] Note that dwz has been patched by Tom de Vries to work correctly in
this case, so we can use dwz master to run the test:
https://sourceware.org/git/?p=dwz.git;a=commit;h=08becc8b33453b6d013a65e7eeae57fc1881e801
gdb/ChangeLog:
* contrib/cc-with-tweaks.sh: Validate dwz's work.
I found out recently that some users didn't know that the Python
pretty-printers "children" method should compute its result lazily.
This has been a good idea since the earliest days, but wasn't
mentioned in the docs. This patch adds some text to this effect.
gdb/doc/ChangeLog
2019-05-10 Tom Tromey <tromey@adacore.com>
* python.texi (Pretty Printing API): Mention lazy computation for
"children".
This patch adds a completion function to the "catch exception"
and "catch handlers" commands.
Tested on x86-64 Fedora 29; reviewed off-list by Joel.
gdb/ChangeLog
2019-05-10 Tom Tromey <tromey@adacore.com>
* ada-lang.c (catch_ada_completer): New function.
(_initialize_ada_language): Use it.
gdb/testsuite/ChangeLog
2019-05-10 Tom Tromey <tromey@adacore.com>
* gdb.ada/info_exc.exp: Add "complete" test.
This patch makes a few minor improvements to the catchpoint
documentation:
* "catch exception" and "catch handlers" now mention the argument in
the @item.
* "catch exception unhandled" is moved to be closer to "catch
exception", rather than after "catch handlers".
* "catch load" and "catch unload" now wrap the argument in @var.
gdb/doc/ChangeLog
2019-05-10 Tom Tromey <tromey@adacore.com>
* gdb.texinfo (Set Catchpoints): Add text for parameter to "catch
exception" and "catch handlers". Move "catch exception unhandled"
text. Use @var for "catch load" and "catch unload"
I noticed a couple of spots where a "char *" was used where a
"const char *" made more sense. This patch fixes both of them.
Tested by rebuilding.
gdb/ChangeLog
2019-05-10 Tom Tromey <tromey@adacore.com>
* thread.c (print_thread_info): Make "requested_threads" const.
* gdbthread.h (print_thread_info): Make "requested_threads"
const.
* ada-tasks.c (print_ada_task_info): Make "taskno_str" const.
* ada-lang.h (print_ada_task_info): Make "taskno_str" const.
The test-case gdb.arch/amd64-tailcall-self.exp fails here:
...
if ![runto b] {
return -1
}
...
like:
...
(gdb) file build/gdb/testsuite/outputs/gdb.arch/amd64-tailcall-self/\
amd64-tailcall-self
Reading symbols from build/gdb/testsuite/outputs/gdb.arch/\
amd64-tailcall-self/amd64-tailcall-self...
Dwarf Error: Cannot find DIE at 0x1f5 referenced from DIE at 0x107 [in \
module build/gdb/testsuite/outputs/gdb.arch/amd64-tailcall-self/\
amd64-tailcall-self]
...
The problem is that in amd64-tailcall-self.S, CU-relative references are
assigned .debug_info section relative values. [ This is similar to the
problem fixed by "Fix gdb.arch/amd64-entry-value-paramref.S". ]
Fix this by assigning CU-relative references instead.
Tested on x86_64-linux.
gdb/testsuite/ChangeLog:
2019-05-09 Tom de Vries <tdevries@suse.de>
* gdb.arch/amd64-tailcall-self.S: Make DW_FORM_ref4 references
CU-relative.
The file gdb.arch/amd64-entry-value-paramref.S contains a DIE for function
bar:
...
DIE29: .uleb128 0x2 # (DIE (0x29) DW_TAG_subprogram)
.ascii "bar\0" # DW_AT_name
.byte 0x1 # DW_AT_decl_file (gdb.arch/amd64-entry-value-paramref.cc)
.byte 0x15 # DW_AT_decl_line
.long DIE45 # DW_AT_type
.byte 0x1 # DW_AT_inline
...
which refers to DIE45:
...
DIE45: .uleb128 0x4 # (DIE (0x45) DW_TAG_base_type)
.byte 0x4 # DW_AT_byte_size
.byte 0x5 # DW_AT_encoding
.ascii "int\0" # DW_AT_name
...
using a form DW_FORM_ref4:
...
.uleb128 0x2 # (abbrev code)
.uleb128 0x2e # (TAG: DW_TAG_subprogram)
.byte 0x1 # DW_children_yes
...
.uleb128 0x49 # (DW_AT_type)
.uleb128 0x13 # (DW_FORM_ref4)
...
However, the DW_FORM_ref4 is a CU-relative reference, while using a label for
the value will result in a section-relative value.
So, if linked in object files contain dwarf info and are placed in the
.debug_info section before the compilation units generated from
amd64-entry-value-paramref.S, then the referenced type is at 0x108:
...
<1><108>: Abbrev Number: 4 (DW_TAG_base_type)
<109> DW_AT_byte_size : 4
<10a> DW_AT_encoding : 5 (signed)
<10b> DW_AT_name : int
...
but the reference will point to a non-existing DIE at 0x1cf:
...
<1><f0>: Abbrev Number: 2 (DW_TAG_subprogram)
<f1> DW_AT_name : bar
<f5> DW_AT_decl_file : 1
<f6> DW_AT_decl_line : 21
<f7> DW_AT_type : <0x1cf>
<fb> DW_AT_inline : 1 (inlined)
...
which happens to cause a GDB internal error described in PR23270 - "GDB
internal error: dwarf2read.c:18656: internal-error: could not find partial
DIE 0x1b7 in cache".
Fix the invalid DWARF by making the reference value CU-relative:
...
- .long DIE45 # DW_AT_type
+ .long DIE45 - .Ldebug_info0 # DW_AT_type
...
Tested on x86_64-linux.
gdb/testsuite/ChangeLog:
2019-05-09 Tom de Vries <tdevries@suse.de>
* gdb.arch/amd64-entry-value-paramref.S: Make DW_FORM_ref4 references
CU-relative.
This changes gdbtypes.c to use the type-safe registry API.
gdb/ChangeLog
2019-05-08 Tom Tromey <tom@tromey.com>
* gdbtypes.c (objfile_type_data): Change type.
(objfile_type, _initialize_gdbtypes): Update.
This changes dwarf2-frame.c to use the type-safe registry API.
gdb/ChangeLog
2019-05-08 Tom Tromey <tom@tromey.com>
* dwarf2-frame.c (dwarf2_frame_objfile_data): Change type.
(dwarf2_frame_find_fde, dwarf2_build_frame_info)
(_initialize_dwarf2_frame): Update.
This changes objc-lang.c to use the type-safe registry API.
gdb/ChangeLog
2019-05-08 Tom Tromey <tom@tromey.com>
* objc-lang.c (objc_objfile_data): Change type.
(find_methods): Update.
(_initialize_objc_lang): Remove.
This changes stabsread.c to use the type-safe registry API.
gdb/ChangeLog
2019-05-08 Tom Tromey <tom@tromey.com>
* stabsread.c (rs6000_builtin_type_data): Change type.
(rs6000_builtin_type, _initialize_stabsread): Update.
mips_pdr_data is unused, so this patch removes it.
gdb/ChangeLog
2019-05-08 Tom Tromey <tom@tromey.com>
* mips-tdep.c (mips_pdr_data): Remove.
(_initialize_mips_tdep): Update.
