This introduces a new "iconv_wrapper" class, to be used in
convert_between_encodings. This allows the removal of cleanup_iconv.
gdb/ChangeLog
2017-04-12 Tom Tromey <tom@tromey.com>
* charset.c (iconv_wrapper): New class.
(cleanup_iconv): Remove.
(convert_between_encodings): Use it.
This changes increment_reading_symtab to return a scoped_restore, then
fixes up the users.
gdb/ChangeLog
2017-04-12 Tom Tromey <tom@tromey.com>
* symfile.h (increment_reading_symtab): Update type.
* symfile.c (decrement_reading_symtab): Remove.
(increment_reading_symtab): Return a scoped_restore_tmpl<int>.
* psymtab.c (psymtab_to_symtab): Update.
* dwarf2read.c (dw2_instantiate_symtab): Update.
This introduces gdb_dlhandle_up, a unique_ptr that can close a
dlopen'd library. All the functions working with dlopen handles are
updated to use this new type.
I did not try to build this on Windows.
gdb/ChangeLog
2017-04-12 Tom Tromey <tom@tromey.com>
* jit.c (struct jit_reader): Declare separately. Add constructor
and destructor. Change type of "handle".
(loaded_jit_reader): Define separately.
(jit_reader_load): Update. New "new".
(jit_reader_unload_command): Use "delete".
* gdb-dlfcn.h (struct dlclose_deleter): New.
(gdb_dlhandle_up): New typedef.
(gdb_dlopen, gdb_dlsym): Update types.
(gdb_dlclose): Remove.
* gdb-dlfcn.c (gdb_dlopen): Return a gdb_dlhandle_up.
(gdb_dlsym): Change type of "handle".
(make_cleanup_dlclose): Remove.
(dlclose_deleter::operator()): Rename from gdb_dlclose.
* compile/compile-c-support.c (load_libcc): Update.
This changes find_pcs_for_symtab_line to return a std::vector. This
allows the removal of some cleanups.
gdb/ChangeLog
2017-04-12 Tom Tromey <tom@tromey.com>
* symtab.h (find_pcs_for_symtab_line): Change return type.
* symtab.c (find_pcs_for_symtab_line): Change return type.
* python/py-linetable.c (build_line_table_tuple_from_pcs): Change
type of "vec". Update.
(ltpy_get_pcs_for_line): Update.
* linespec.c (decode_digits_ordinary): Update.
This introduces command_line_up, a unique_ptr for command_line
objects, and changes many places to use it. This removes a number of
cleanups.
Command lines are funny in that sometimes they are reference counted.
Once there is more C++-ification of some of the users, perhaps all of
these can be changed to use shared_ptr instead.
gdb/ChangeLog
2017-04-12 Tom Tromey <tom@tromey.com>
* tracepoint.c (actions_command): Update.
* python/python.c (python_command, python_interactive_command):
Update.
* mi/mi-cmd-break.c (mi_cmd_break_commands): Update.
* guile/guile.c (guile_command): Update.
* defs.h (read_command_lines, read_command_lines_1): Return
command_line_up.
(command_lines_deleter): New struct.
(command_line_up): New typedef.
* compile/compile.c (compile_code_command)
(compile_print_command): Update.
* cli/cli-script.h (get_command_line, copy_command_lines): Return
command_line_up.
(make_cleanup_free_command_lines): Remove.
* cli/cli-script.c (get_command_line, read_command_lines_1)
(copy_command_lines): Return command_line_up.
(while_command, if_command, read_command_lines, define_command)
(document_command): Update.
(do_free_command_lines_cleanup, make_cleanup_free_command_lines):
Remove.
* breakpoint.h (breakpoint_set_commands): Change type of
"commands".
* breakpoint.c (breakpoint_set_commands): Change type of
"commands". Update.
(do_map_commands_command, update_dprintf_command_list)
(create_tracepoint_from_upload): Update.
I posit that this makes them easier to find.
The other day while working on the wchar_t patch, I had a bit of
trouble finding the DJGPP/go32 tdep bits. My initial reaction was
looking for a go32-specific tdep file, but there's none.
Confirmed that a --host=i586-pc-msdosdjgpp GDB still builds
successfully and includes the i386-go32-tdep.o object.
Confirmed that an --enable-targets=all build of GDB on x86-64
GNU/Linux includes the DJGPP/go32 bits too.
gdb/ChangeLog:
2017-04-12 Pedro Alves <palves@redhat.com>
* Makefile.in (ALL_TARGET_OBS): Add i386-go32-tdep.o.
* configure.tgt: Handle i[34567]86-*-go32* and
i[34567]86-*-msdosdjgpp*.
* i386-tdep.c (i386_svr4_reg_to_regnum):
Make extern.
(i386_go32_init_abi, i386_coff_osabi_sniffer): Moved to
i386-go32-tdep.c.
(_initialize_i386_tdep): DJGPP bits moved to i386-go32-tdep.c.
* i386-go32-tdep.c: New file.
* i386-tdep.h (tdesc_i386_mmx, i386_svr4_reg_to_regnum): New
declarations.
Obvious fix for:
aix-thread.c: In function 'char* pd_status2str(int)':
aix-thread.c:163:33: error: deprecated conversion from string constant to 'char*' [-Werror=write-strings]
case PTHDB_SUCCESS: return "SUCCESS";
^
gdb/ChangeLog:
* aix-thread.c (pd_status2str): Change return type to const char *.
i386_gdbarch_init already does this unconditionally for all x86 ports.
Tested on x86-64 Fedora 23.
gdb/ChangeLog:
2017-04-12 Pedro Alves <palves@redhat.com>
* i386-tdep.c (i386_elf_init_abi, i386_go32_init_abi): Remove
calls to set_gdbarch_gnu_triplet_regexp.
GDB is currently not aware that wchar_t is a built-in type in C++
mode. This is usually not a problem because the debug info describes
the type, so when you have a program loaded, you don't notice this.
However, if you try expressions involving wchar_t before a program is
loaded, gdb errors out:
(gdb) p (wchar_t)-1
No symbol table is loaded. Use the "file" command.
(gdb) p L"hello"
No type named wchar_t.
(gdb) ptype L"hello"
No type named wchar_t.
This commit teaches gdb about the type. After:
(gdb) p (wchar_t)-1
$1 = -1 L'\xffffffff'
(gdb) p L"hello"
$2 = L"hello"
(gdb) ptype L"hello"
type = wchar_t [6]
Unlike char16_t/char32_t, unfortunately, the underlying type of
wchar_t is implementation dependent, both size and signness. So this
requires adding a couple new gdbarch hooks.
I grepped the GCC code base for WCHAR_TYPE and WCHAR_TYPE_SIZE, and it
seems to me that the majority of the ABIs have a 4-byte signed
wchar_t, so that's what I made the default for GDB too. And then I
looked for which ports have a 16-bit and/or unsigned wchar_t, and made
GDB follow suit.
gdb/ChangeLog:
2017-04-12 Pedro Alves <palves@redhat.com>
PR gdb/21323
* c-lang.c (cplus_primitive_types) <cplus_primitive_type_wchar_t>:
New enum value.
(cplus_language_arch_info): Register cplus_primitive_type_wchar_t.
* gdbtypes.h (struct builtin_type) <builtin_wchar>: New field.
* gdbtypes.c (gdbtypes_post_init): Create the "wchar_t" type.
* gdbarch.sh (wchar_bit, wchar_signed): New per-arch values.
* gdbarch.h, gdbarch.c: Regenerate.
* aarch64-tdep.c (aarch64_gdbarch_init): Override
gdbarch_wchar_bit and gdbarch_wchar_signed.
* alpha-tdep.c (alpha_gdbarch_init): Likewise.
* arm-tdep.c (arm_gdbarch_init): Likewise.
* avr-tdep.c (avr_gdbarch_init): Likewise.
* h8300-tdep.c (h8300_gdbarch_init): Likewise.
* i386-nto-tdep.c (i386nto_init_abi): Likewise.
* i386-tdep.c (i386_go32_init_abi): Likewise.
* m32r-tdep.c (m32r_gdbarch_init): Likewise.
* moxie-tdep.c (moxie_gdbarch_init): Likewise.
* nds32-tdep.c (nds32_gdbarch_init): Likewise.
* rs6000-aix-tdep.c (rs6000_aix_init_osabi): Likewise.
* sh-tdep.c (sh_gdbarch_init): Likewise.
* sparc-tdep.c (sparc32_gdbarch_init): Likewise.
* sparc64-tdep.c (sparc64_init_abi): Likewise.
* windows-tdep.c (windows_init_abi): Likewise.
* xstormy16-tdep.c (xstormy16_gdbarch_init): Likewise.
gdb/testsuite/ChangeLog:
2017-04-12 Pedro Alves <palves@redhat.com>
PR gdb/21323
* gdb.cp/wide_char_types.c: Include <wchar.h>.
(wchar): New global.
* gdb.cp/wide_char_types.exp (wide_char_types_program)
(do_test_wide_char, wide_char_types_no_program, top level): Add
wchar_t testing.
While the C++ standard says that char16_t and char32_t are unsigned types:
Types char16_t and char32_t denote distinct types with the same size,
signedness, and alignment as uint_least16_t and uint_least32_t,
respectively, in <cstdint>, called the underlying types.
... gdb treats them as signed currently:
(gdb) p (char16_t)-1
$1 = -1 u'\xffff'
There are actually two places in gdb that hardcode these types:
- gdbtypes.c:gdbtypes_post_init, when creating the built-in types,
seemingly used by the "x /s" command (judging from commit 9a22f0d0).
- dwarf2read.c, when reading base types with DW_ATE_UTF encoding
(which is what is used for these types, when compiling for C++11 and
up). Despite the comment, the type created does end up used.
Both places need fixing. But since I couldn't tell why dwarf2read.c
needs to create a new type, I've made it use the per-arch built-in
types instead, so that the types are only created once per arch
instead of once per objfile. That seems to work fine.
While writting the test, I noticed that the C++ language parser isn't
actually aware of these built-in types, so if you try to use them
without a program that uses them, you get:
(gdb) set language c++
(gdb) ptype char16_t
No symbol table is loaded. Use the "file" command.
(gdb) ptype u"hello"
No type named char16_t.
(gdb) p u"hello"
No type named char16_t.
That's fixed by simply adding a couple entries to C++'s built-in types
array in c-lang.c. With that, we get the expected:
(gdb) ptype char16_t
type = char16_t
(gdb) ptype u"hello"
type = char16_t [6]
(gdb) p u"hello"
$1 = u"hello"
gdb/ChangeLog:
2017-04-12 Pedro Alves <palves@redhat.com>
PR c++/21323
* c-lang.c (cplus_primitive_types) <cplus_primitive_type_char16_t,
cplus_primitive_type_char32_t>: New enum values.
(cplus_language_arch_info): Register cplus_primitive_type_char16_t
and cplus_primitive_type_char32_t.
* dwarf2read.c (read_base_type) <DW_ATE_UTF>: If bit size is 16 or
32, use the archtecture's built-in type for char16_t and char32_t,
respectively. Otherwise, fallback to init_integer_type as before,
but make the type unsigned, and issue a complaint.
* gdbtypes.c (gdbtypes_post_init): Make char16_t and char32_t unsigned.
gdb/testsuite/ChangeLog:
2017-04-12 Pedro Alves <palves@redhat.com>
PR c++/21323
* gdb.cp/wide_char_types.c: New file.
* gdb.cp/wide_char_types.exp: New file.
As a preparation for the next patch, which will move fork_inferior
from GDB to common/ (and therefore share it with gdbserver), it is
interesting to convert a few functions to C++.
This patch touches functions related to parsing command-line arguments
to the inferior (see gdb/fork-child.c:breakup_args), the way the
arguments are stored on fork_inferior (using std::vector instead of
char **), and the code responsible for dealing with argv also on
gdbserver.
I've taken this opportunity and decided to constify a few arguments to
fork_inferior/create_inferior as well, in order to make the code
cleaner. And now, on gdbserver, we're using xstrdup everywhere and
aren't checking for memory allocation failures anymore, as requested
by Pedro:
<https://sourceware.org/ml/gdb-patches/2017-03/msg00191.html>
Message-Id: <025ebdb9-90d9-d54a-c055-57ed2406b812@redhat.com>
Pedro Alves wrote:
> On the "== NULL" check: IIUC, the old NULL check was there to
> handle strdup returning NULL due to out-of-memory.
> See NULL checks and comments further above in this function.
> Now that you're using a std::vector, that doesn't work or make
> sense any longer, since if push_back fails to allocate space for
> its internal buffer (with operator new), our operator new replacement
> (common/new-op.c) calls malloc_failure, which aborts gdbserver.
>
> Not sure it makes sense to handle out-of-memory specially in
> the gdb/rsp-facing functions nowadays (maybe git blame/log/patch
> submission for that code shows some guidelines). Maybe (or, probably)
> it's OK to stop caring about it, but then we should consistently remove
> left over code, by using xstrdup instead and remove the NULL checks.
IMO this refactoring was very good to increase the readability of the
code as well, because some parts of the argument handling were
unnecessarily confusing before.
gdb/ChangeLog:
2017-04-12 Sergio Durigan Junior <sergiodj@redhat.com>
* common/common-utils.c (free_vector_argv): New function.
* common/common-utils.h: Include <vector>.
(free_vector_argv): New prototype.
* darwin-nat.c (darwin_create_inferior): Rewrite function
prototype in order to constify "exec_file" and accept a
"std::string" for "allargs".
* fork-child.c: Include <vector>.
(breakup_args): Rewrite function, using C++.
(fork_inferior): Rewrite function header, constify "exec_file_arg"
and accept "std::string" for "allargs". Update the code to
calculate "argv" based on "allargs". Update calls to "exec_fun"
and "execvp".
* gnu-nat.c (gnu_create_inferior): Rewrite function prototype in
order to constify "exec_file" and accept a "std::string" for
"allargs".
* go32-nat.c (go32_create_inferior): Likewise.
* inf-ptrace.c (inf_ptrace_create_inferior): Likewise.
* infcmd.c (run_command_1): Constify "exec_file". Use
"std::string" for inferior arguments.
* inferior.h (fork_inferior): Update prototype.
* linux-nat.c (linux_nat_create_inferior): Rewrite function
prototype in order to constify "exec_file" and accept a
"std::string" for "allargs".
* nto-procfs.c (procfs_create_inferior): Likewise.
* procfs.c (procfs_create_inferior): Likewise.
* remote-sim.c (gdbsim_create_inferior): Likewise.
* remote.c (extended_remote_run): Update code to accept
"std::string" as argument.
(extended_remote_create_inferior): Rewrite function prototype in
order to constify "exec_file" and accept a "std::string" for
"allargs".
* rs6000-nat.c (super_create_inferior): Likewise.
(rs6000_create_inferior): Likewise.
* target.h (struct target_ops) <to_create_inferior>: Likewise.
* windows-nat.c (windows_create_inferior): Likewise.
gdb/gdbserver/ChangeLog:
2017-04-12 Sergio Durigan Junior <sergiodj@redhat.com>
* server.c: Include <vector>.
<program_argv, wrapper_argv>: Convert to std::vector.
(start_inferior): Rewrite function to use C++.
(handle_v_run): Likewise. Update code that calculates the argv
based on the vRun packet; use C++.
(captured_main): Likewise.
At the end of linux_nat_detach the main_lwp is deleted (delete_lwp).
This is problematic as during detach (detach_one_lwp and
linux_fork_detach) main_lwp already gets freed. Thus calling
delete_lwp causes a read after free. Fix it by removing the
unnecessary delete_lwp.
gdb/ChangeLog:
2017-04-11 Philipp Rudo <prudo@linux.vnet.ibm.com>
* linux-nat.c (linux_nat_detach): Remove delete_lwp call.
Pedro's recent commits enabling -Wwrite-strings has changed a bit the
logic of info_osdata. Now, 'type' is always non-NULL, so we have to
check if it's an empty string instead of NULL. One of the checks was
fixed, but there is another that was left behind. This commit fixes
it.
gdb/ChangeLog:
2017-04-10 Sergio Durigan Junior <sergiodj@redhat.com>
PR gdb/21364
* osdata.c (info_osdata): Check if 'type' is an empty string
instead of NULL.
I build GDB with asan, and run test case hook-stop.exp, and threadapply.exp,
I got the following asan error,
=================================================================^M
^[[1m^[[31m==2291==ERROR: AddressSanitizer: heap-use-after-free on address 0x6160000999c4 at pc 0x000000826022 bp 0x7ffd28a8ff70 sp 0x7ffd28a8ff60^M
^[[1m^[[0m^[[1m^[[34mREAD of size 4 at 0x6160000999c4 thread T0^[[1m^[[0m^M
#0 0x826021 in release_stop_context_cleanup ../../binutils-gdb/gdb/infrun.c:8203^M
#1 0x72798a in do_my_cleanups ../../binutils-gdb/gdb/common/cleanups.c:154^M
#2 0x727a32 in do_cleanups(cleanup*) ../../binutils-gdb/gdb/common/cleanups.c:176^M
#3 0x826895 in normal_stop() ../../binutils-gdb/gdb/infrun.c:8381^M
#4 0x815208 in fetch_inferior_event(void*) ../../binutils-gdb/gdb/infrun.c:4011^M
#5 0x868aca in inferior_event_handler(inferior_event_type, void*) ../../binutils-gdb/gdb/inf-loop.c:44^M
....
^[[1m^[[32m0x6160000999c4 is located 68 bytes inside of 568-byte region [0x616000099980,0x616000099bb8)^M
^[[1m^[[0m^[[1m^[[35mfreed by thread T0 here:^[[1m^[[0m^M
#0 0x7fb0bc1312ca in __interceptor_free (/usr/lib/x86_64-linux-gnu/libasan.so.2+0x982ca)^M
#1 0xb8c62f in xfree(void*) ../../binutils-gdb/gdb/common/common-utils.c:100^M
#2 0x83df67 in free_thread ../../binutils-gdb/gdb/thread.c:207^M
#3 0x83dfd2 in init_thread_list() ../../binutils-gdb/gdb/thread.c:223^M
#4 0x805494 in kill_command ../../binutils-gdb/gdb/infcmd.c:2595^M
....
Detaching from program: /home/yao.qi/SourceCode/gnu/build-with-asan/gdb/testsuite/outputs/gdb.threads/threadapply/threadapply, process 2399^M
=================================================================^M
^[[1m^[[31m==2387==ERROR: AddressSanitizer: heap-use-after-free on address 0x6160000a98c0 at pc 0x00000083fd28 bp 0x7ffd401c3110 sp 0x7ffd401c3100^M
^[[1m^[[0m^[[1m^[[34mREAD of size 4 at 0x6160000a98c0 thread T0^[[1m^[[0m^M
#0 0x83fd27 in thread_alive ../../binutils-gdb/gdb/thread.c:741^M
#1 0x844277 in thread_apply_all_command ../../binutils-gdb/gdb/thread.c:1804^M
....
^M
^[[1m^[[32m0x6160000a98c0 is located 64 bytes inside of 568-byte region [0x6160000a9880,0x6160000a9ab8)^M
^[[1m^[[0m^[[1m^[[35mfreed by thread T0 here:^[[1m^[[0m^M
#0 0x7f59a7e322ca in __interceptor_free (/usr/lib/x86_64-linux-gnu/libasan.so.2+0x982ca)^M
#1 0xb8c62f in xfree(void*) ../../binutils-gdb/gdb/common/common-utils.c:100^M
#2 0x83df67 in free_thread ../../binutils-gdb/gdb/thread.c:207^M
#3 0x83dfd2 in init_thread_list() ../../binutils-gdb/gdb/thread.c:223^M
This patch fixes the issue by deleting thread_info object if it is
deletable, otherwise, mark it as exited (by set_thread_exited).
Function set_thread_exited is shared from delete_thread_1. This patch
also moves field "refcount" to private and methods incref and
decref. Additionally, we stop using "ptid_t" in
"struct current_thread_cleanup" to reference threads, instead we use
"thread_info" directly. Due to this change, we don't need
restore_current_thread_ptid_changed anymore.
gdb:
2017-04-10 Yao Qi <yao.qi@linaro.org>
PR gdb/19942
* gdbthread.h (thread_info::deletable): New method.
(thread_info::incref): New method.
(thread_info::decref): New method.
(thread_info::refcount): Move it to private.
* infrun.c (save_stop_context): Call inc_refcount.
(release_stop_context_cleanup): Likewise.
* thread.c (set_thread_exited): New function.
(init_thread_list): Delete "tp" only it is deletable, otherwise
call set_thread_exited.
(delete_thread_1): Call set_thread_exited.
(current_thread_cleanup) <inferior_pid>: Remove.
<thread>: New field.
(restore_current_thread_ptid_changed): Removed.
(do_restore_current_thread_cleanup): Adjust.
(restore_current_thread_cleanup_dtor): Don't call
find_thread_ptid.
(set_thread_refcount): Use dec_refcount.
(make_cleanup_restore_current_thread): Adjust.
(thread_apply_all_command): Call inc_refcount.
(_initialize_thread): Don't call
observer_attach_thread_ptid_changed.
This patch hoists code on marking thread as exited, so more code is shared
for two different paths (thread_info is deleted or is not deleted).
gdb:
2017-04-10 Yao Qi <yao.qi@linaro.org>
* thread.c (delete_thread_1): Hoist code on marking thread as
exited.
When trying to build for x86_64-w64-mingw32:
/home/simark/src/binutils-gdb/gdb/windows-nat.c: In function ‘void windows_detach(target_ops*, const char*, int)’:
/home/simark/src/binutils-gdb/gdb/windows-nat.c:1915:20: error: converting to ‘ptid_t’ from initializer list would use explicit constructor ‘constexpr ptid_t::ptid_t(int, long int, long int)’
ptid_t ptid = {-1};
^
Fixed by initializing ptid with the minus_one_ptid variable.
gdb/ChangeLog:
* windows-nat.c (windows_detach): Initialize ptid with
minus_one_ptid.
I grew a bit tired of using ptid_get_{lwp,pid,tid} and friends, so I decided to
make it a bit easier to use by making it a proper class. The fields are now
private, so it's not possible to change a ptid_t field by mistake.
The new methods of ptid_t map to existing functions/practice like this:
ptid_t (pid, lwp, tid) -> ptid_build (pid, lwp, tid)
ptid_t (pid) -> pid_to_ptid (pid)
ptid.is_pid () -> ptid_is_pid (ptid)
ptid == other -> ptid_equal (ptid, other)
ptid != other -> !ptid_equal (ptid, other)
ptid.pid () -> ptid_get_pid (ptid)
ptid.lwp_p () -> ptid_lwp_p (ptid)
ptid.lwp () -> ptid_get_lwp (ptid)
ptid.tid_p () -> ptid_tid_p (ptid)
ptid.tid () -> ptid_get_tid (ptid)
ptid.matches (filter) -> ptid_match (ptid, filter)
I've replaced the implementation of the existing functions with calls to
the new methods. People are encouraged to gradually switch to using the
ptid_t methods instead of the functions (or we can change them all in
one pass eventually).
Also, I'm not sure if it's worth it (because of ptid_t's relatively
small size), but I have made the functions and methods take ptid_t
arguments by const reference instead of by value.
gdb/ChangeLog:
* common/ptid.h (struct ptid): Change to...
(class ptid_t): ... this.
<ptid_t>: New constructors.
<pid, lwp_p, lwp, tid_p, tid, is_pid, operator==, operator!=,
matches>: New methods.
<make_null, make_minus_one>: New static methods.
<pid>: Rename to...
<m_pid>: ...this.
<lwp>: Rename to...
<m_lwp>: ...this.
<tid>: Rename to...
<m_tid>: ...this.
(ptid_build, ptid_get_pid, ptid_get_lwp, ptid_get_tid, ptid_equal,
ptid_is_pid, ptid_lwp_p, ptid_tid_p, ptid_match): Take ptid arguments
as references, move comment to class ptid_t.
* common/ptid.c (null_ptid, minus_one_ptid): Initialize with
ptid_t static methods.
(ptid_build, pid_to_ptid, ptid_get_pid, ptid_get_tid,
ptid_equal, ptid_is_pid, ptid_lwp_p, ptid_tid_p, ptid_match):
Take ptid arguments as references, implement using ptid_t methods.
* unittests/ptid-selftests.c: New file.
* Makefile.in (SUBDIR_UNITTESTS_SRCS): Add
unittests/ptid-selftests.c.
(SUBDIR_UNITTESTS_OBS): Add unittests/ptid-selftests.o.
gdb/gdbserver/ChangeLog:
* server.c (handle_v_cont): Initialize thread_resume::thread
with null_ptid.
GDB fails to build for Windows host with Python 2 support enabled due
to PyFile_FromString's second argument being of type char * and being
passed a (const) string literal. This parameter is input only so this
commit fixes the issue by casting to char *.
2017-04-06 Thomas Preud'homme <thomas.preudhomme@arm.com>
gdb/
* python/python.c (python_run_simple_file): Cast mode literal to
non-const char pointer as expected by PyFile_FromString.
The calls to ptid_equal in ptid_lwp_p and ptid_tid_p that compare the
argument to minus_one_ptid and null_ptid are not necessary. The calls
in question are:
if (ptid_equal (minus_one_ptid, ptid)
|| ptid_equal (null_ptid, ptid))
return 0;
minus_one_ptid is { .pid = -1, .lwp = 0, .tid = 0 }
null_ptid is { .pid = 0, .lwp = 0, .tid = 0 }
If the ptid argument is either of them, the statements
return (ptid_get_lwp (ptid) != 0);
and
return (ptid_get_tid (ptid) != 0);
will yield the same result (0/false).
gdb/ChangeLog:
* common/ptid.c (ptid_lwp_p, ptid_tid_p): Remove comparison with
minus_one_ptid and null_ptid.
AFAIK GDB is now free from -Wwrite-strings warnings. A few warnings may
be left behind in some host-specific code, but those should be few and
easy to fix.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* warning.m4 (build_warnings): Remove -Wno-write-strings.
* configure: Regenerate.
gdb/gdbserver/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* configure: Regenerate.
Compiling GDB with -Wwrite-strings flags this code in gdb/proc-api.c:
static char *procfs_filename = "procfs_trace";
as needing a cast. However, this variable is a command variable, and
as such it's incorrect to initialize it to a literal, since when you
use the corresponding set command, gdb frees the old string...
I didn't manage to fully build Solaris gdb (fails for other reasons),
but I confirmed that the system GDB on Solaris 11 crashes when running
this command:
(gdb) set procfs-file foo
Segmentation Fault (core dumped)
So I don't think this commit can make it worse than the status quo.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* proc-api.c (procfs_filename): Don't initialize
procfs_filename.
(prepare_to_trace): Assume procfs_filename is non-NULL.
(_initialize_proc_api): Give procfs_filename a default value here.
The main motivation here is avoiding having to write a couple casts
like these:
if (!arg)
- arg = "";
+ arg = (char *) "";
in catch_exception_command_1 and catch_exec_command_1.
That requires making ep_parse_optional_if_clause and
check_for_argument take pointers to const strings. I then tried
propagating the resulting constification all the way, but that was
spiraling out of control, so instead I settled for keeping const and
non-const overloads.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* break-catch-throw.c (handle_gnu_v3_exceptions): Constify
'cond_string' parameter.
(extract_exception_regexp): Constify 'string' parameter.
(catch_exception_command_1): Constify.
* breakpoint.c (init_catchpoint)
(create_fork_vfork_event_catchpoint): Constify 'cond_string'
parameter.
(ep_parse_optional_if_clause, catch_fork_command_1)
(catch_exec_command_1): Constify.
* breakpoint.h (init_catchpoint): Constify 'cond_string'
parameter.
(ep_parse_optional_if_clause): Constify.
* cli/cli-utils.c (remove_trailing_whitespace)
(check_for_argument): Constify.
* cli/cli-utils.h (remove_trailing_whitespace): Constify and add
non-const overload.
(check_for_argument): Likewise.
This is ugly, but it's just making the uglyness explicit.
All these places would better be calling some API function directly
instead of going through execute_command & friends...
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* event-top.c (command_line_handler): Add cast to execute_command
call.
* record-btrace.c (cmd_record_btrace_bts_start)
(cmd_record_btrace_pt_start, cmd_record_btrace_start)
(cmd_record_btrace_start): Add cast to execute_command call.
* record-full.c (record_full_goto_insn):
* record.c (record_start, record_stop): Add cast to
execute_command_to_string calls.
(cmd_record_start): Add cast to execute_command calls.
-Wwrite-strings flags code like:
static char *keywords[] = {"command", "from_tty", "to_string", NULL };
as needing "(char *)" casts, because string literals are "const char []".
We can get rid of the casts by changing the array type like this:
- static char *keywords[] = {"command", "from_tty", "to_string", NULL };
+ static const char *keywords[] = {"command", "from_tty", "to_string", NULL };
However, passing the such array to PyArg_ParseTupleAndKeywords no longer
works OOTB, because PyArg_ParseTupleAndKeywords expects a "char **":
PyArg_ParseTupleAndKeywords(PyObject *args, PyObject *kw,
const char *format,
char *keywords[], ...);
and "const char **" is not implicitly convertible to "char **". C++
is more tolerant that C here WRT aliasing, and a const_cast<char **>
is fine. However, to avoid having all callers do the cast themselves,
this commit defines a gdb_PyArg_ParseTupleAndKeywords function here
with a corresponding 'keywords' parameter type that does the cast in a
single place.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* python/python-internal.h (gdb_PyArg_ParseTupleAndKeywords): New
static inline function.
* python/py-arch.c (archpy_disassemble): Constify 'keywords'
array and use gdb_PyArg_ParseTupleAndKeywords.
* python/py-cmd.c (cmdpy_init): Likewise.
* python/py-finishbreakpoint.c (bpfinishpy_init): Likewise.
* python/py-inferior.c (infpy_read_memory, infpy_write_memory)
(infpy_search_memory): Likewise.
* python/py-objfile.c (objfpy_add_separate_debug_file)
(gdbpy_lookup_objfile): Likewise.
* python/py-symbol.c (gdbpy_lookup_symbol)
(gdbpy_lookup_global_symbol): Likewise.
* python/py-type.c (gdbpy_lookup_type): Likewise.
* python/py-value.c (valpy_lazy_string, valpy_string): Likewise.
* python/python.c (execute_gdb_command, gdbpy_write, gdbpy_flush):
Likewise.
Unfortunately, PyGetSetDef's 'name' and 'doc' members are 'char *'
instead of 'const char *', meaning that in order to list-initialize
PyGetSetDef arrays using string literals requires writing explicit
'char *' casts. For example:
static PyGetSetDef value_object_getset[] = {
- { "address", valpy_get_address, NULL, "The address of the value.",
+ { (char *) "address", valpy_get_address, NULL,
+ (char *) "The address of the value.",
NULL },
- { "is_optimized_out", valpy_get_is_optimized_out, NULL,
- "Boolean telling whether the value is optimized "
+ { (char *) "is_optimized_out", valpy_get_is_optimized_out, NULL,
+ (char *) "Boolean telling whether the value is optimized "
"out (i.e., not available).",
NULL },
- { "type", valpy_get_type, NULL, "Type of the value.", NULL },
- { "dynamic_type", valpy_get_dynamic_type, NULL,
- "Dynamic type of the value.", NULL },
- { "is_lazy", valpy_get_is_lazy, NULL,
- "Boolean telling whether the value is lazy (not fetched yet\n\
+ { (char *) "type", valpy_get_type, NULL,
+ (char *) "Type of the value.", NULL },
+ { (char *) "dynamic_type", valpy_get_dynamic_type, NULL,
+ (char *) "Dynamic type of the value.", NULL },
+ { (char *) "is_lazy", valpy_get_is_lazy, NULL,
+ (char *) "Boolean telling whether the value is lazy (not fetched yet\n\
from the inferior). A lazy value is fetched when needed, or when\n\
the \"fetch_lazy()\" method is called.", NULL },
{NULL} /* Sentinel */
We have ~20 such arrays, and I first wrote a patch that fixed all of
them like that... It's not pretty...
One way to make these a bit less ugly would be add a new macro that
hides the casts, like:
#define GDBPY_GSDEF(NAME, GET, SET, DOC, CLOSURE) \
{ (char *) NAME, GET, SET, (char *) DOC, CLOSURE }
and then use it like:
static PyGetSetDef value_object_getset[] = {
GDBPY_GSDEF ("address", valpy_get_address, NULL,
"The address of the value.", NULL),
GDBPY_GSDEF ("is_optimized_out", valpy_get_is_optimized_out, NULL,
"Boolean telling whether the value is optimized ", NULL),
{NULL} /* Sentinel */
};
But since we have C++11, which gives us constexpr and list
initialization, I thought of a way that requires no changes where the
arrays are initialized:
We add a new type that extends PyGetSetDef (called gdb_PyGetSetDef),
and add constexpr constructors that accept const 'name' and 'doc', and
then list/aggregate initialization simply "calls" these matching
constructors instead.
I put "calls" in quotes, because given "constexpr", it's all done at
compile time, and there's no overhead either in binary size or at run
time. In fact, we get identical binaries, before/after this change.
Unlike the fixes that fix some old Python API to match the API of more
recent Python, this switches to using explicit "gdb_PyGetSetDef"
everywhere, just to be clear that we are using our own version of it.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* python/python-internal.h (gdb_PyGetSetDef): New type.
* python/py-block.c (block_object_getset)
(breakpoint_object_getset): Now a gdb_PyGetSetDef array.
* python/py-event.c (event_object_getset)
(finish_breakpoint_object_getset): Likewise.
* python/py-inferior.c (inferior_object_getset): Likewise.
* python/py-infthread.c (thread_object_getset): Likewise.
* python/py-lazy-string.c (lazy_string_object_getset): Likewise.
* python/py-linetable.c (linetable_entry_object_getset): Likewise.
* python/py-objfile.c (objfile_getset): Likewise.
* python/py-progspace.c (pspace_getset): Likewise.
* python/py-record-btrace.c (btpy_insn_getset, btpy_call_getset):
Likewise.
* python/py-record.c (recpy_record_getset): Likewise.
* python/py-symbol.c (symbol_object_getset): Likewise.
* python/py-symtab.c (symtab_object_getset, sal_object_getset):
Likewise.
* python/py-type.c (type_object_getset, field_object_getset):
Likewise.
* python/py-value.c (value_object_getset): Likewise.
When building against Python 2.7, -Wwrite-strings flags several cases
of passing a string literal to Python functions that expect a "char
*". This commit addresses the issue like we already handle several
other similar cases -- wrap the Python API with our own fixed
version that adds the necessary constification.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* python/python-internal.h (gdb_PyObject_CallMethod)
(gdb_PyErr_NewException, gdb_PySys_GetObject, gdb_PySys_SetPath):
New functions.
(GDB_PYSYS_SETPATH_CHAR, PyObject_CallMethod, PyErr_NewException)
(PySys_GetObject, PySys_SetPath): New macros.
src/gdb/gdbserver/win32-low.c:1499:39: error: ISO C++ forbids converting a string constant to 'char*' [-Werror=write-strings]
ourstatus->value.execd_pathname = "Main executable";
^
This reporting via TARGET_WAITKIND_EXECD it's totally unnecessary.
get_child_debug_event returns a TARGET_WAITKIND_SPURIOUS by default,
which works just as well here, and is what the equivalent code in
gdb/windows-nat.c does too.
gdb/gdbserver/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* win32-low.c (get_child_debug_event)
<CREATE_PROCESS_DEBUG_EVENT>: Don't report TARGET_WAITKIND_EXECD.
Report TARGET_WAITKIND_SPURIOUS instead.
-Wwrite-strings flags this assignment as requiring a cast:
port = STDIO_CONNECTION_NAME;
because 'port' is a "char *", and STDIO_CONNECTION_NAME is a string
literal.
gdb/gdbserver/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* remote-utils.c (remote_prepare, remote_open): Constify.
* remote-utils.h (remote_prepare, remote_open): Constify.
* server.c (captured_main): Constify 'port' handling.
-Wwrite-strings flags this attempt to convert a string literal to
"char *":
info_osdata_command ("", 0);
info_osdata_command is a command function. We could address this by
simply passing NULL instead of "". However, I went a little bit
further and added a new function that is called by both the CLI and
MI.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* mi/mi-cmd-info.c (mi_cmd_info_os): Call info_osdata instead of
info_osdata_command.
* osdata.c (info_osdata_command): Rename to ...
(info_osdata): ... this. Constify 'type' parameter, and remove
the 'from_tty' parameter. Accept NULL TYPE.
(info_osdata_command): New function.
* osdata.h (info_osdata_command): Remove declaration.
(info_osdata): New declaration.
-Wwrite-strings flags several cases of missing casts around
initializations like:
static char *gdb_completer_command_word_break_characters =
" \t\n!@#$%^&*()+=|~`}{[]\"';:?/>.<,";
Obviously these could/should be const. However, while at it, there's
no need for these variables to be pointers instead of arrays. They
are never changed to point to anything else.
Unfortunately, readline's rl_completer_word_break_characters is
"char *", not "const char *". So we always need a cast somewhere. The
approach taken here is to add a new
set_rl_completer_word_break_characters function that becomes the only
place that writes to rl_completer_word_break_characters, and thus the
single place that needs the cast.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* ada-lang.c (ada_completer_word_break_characters): Now a const
array.
(ada_get_gdb_completer_word_break_characters): Constify.
* completer.c (gdb_completer_command_word_break_characters)
(gdb_completer_file_name_break_characters)
(gdb_completer_quote_characters): Now const arrays.
(get_gdb_completer_quote_characters): Constify.
(set_rl_completer_word_break_characters): New function.
(set_gdb_completion_word_break_characters)
(complete_line_internal): Use it.
* completer.h (get_gdb_completer_quote_characters): Constify.
(set_rl_completer_word_break_characters): Declare.
* f-lang.c (f_word_break_characters): Constify.
* language.c (default_word_break_characters): Constify.
* language.h (language_defn::la_word_break_characters): Constify.
(default_word_break_characters): Constify.
* top.c (init_main): Use set_rl_completer_word_break_characters.
-Wwrite-strings flags these initializations as requiring a cast.
However, these variables are command variables, and as such point to
heap-allocated memory. The initial allocation is always done when the
corresponding command is registered. E.g.,:
dprintf_function = xstrdup ("printf");
add_setshow_string_cmd ("dprintf-function", class_support,
&dprintf_function, _("\
Set the function to use for dynamic printf"), _("\
Show the function to use for dynamic printf"), NULL,
update_dprintf_commands, NULL,
&setlist, &showlist);
That's why we never reach a bogus attempt to free these string
literals.
So, just drop the incorrect initializations.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* breakpoint.c (dprintf_function, dprintf_channel): Don't initialize.
* tracepoint.c (default_collect): Don't initialize.
There's one call in the file that passes a string literal, like:
init_shared_buffer (&va_arg_name, "__VA_ARGS__",
strlen ("__VA_ARGS__"));
Instead of adding a cast here, make init_shared_buffer take a 'const
char *', and remove the several casts in the file that are made
obsolete.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* macroexp.c (macro_buffer::shared): Now a bool.
(init_buffer): Update.
(init_shared_buffer): Constify 'addr' parameter.
(substitute_args, expand, macro_expand, macro_expand_next): Remove
casts.
The memory disassemble_info::disassembler_options points to is always
owned by the client. I.e., that field is an non-owning, observing
pointer. Thus const makes sense.
Are the include/ and opcodes/ bits OK?
Tested on x86_64 Fedora 23, built with --enable-targets=all.
include/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* dis-asm.h (disassemble_info) <disassembler_options>: Now a
"const char *".
(next_disassembler_option): Constify.
opcodes/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* arc-dis.c (parse_option, parse_disassembler_options): Constify.
* arm-dis.c (parse_arm_disassembler_options): Constify.
* ppc-dis.c (powerpc_init_dialect): Constify local.
* vax-dis.c (parse_disassembler_options): Constify.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* arm-tdep.c (show_disassembly_style_sfunc): Constify local.
* disasm.c (set_disassembler_options): Constify local.
* i386-tdep.c (i386_print_insn): Remove cast and FIXME comment.
This commit adds a test for the fix of PR 21352.
gdb/testsuite/ChangeLog:
2017-04-05 Sergio Durigan Junior <sergiodj@redhat.com>
PR gdb/21352
* gdb.trace/tsv.exp: Add test for "tsave -r".
This is an obvious fix for PR 21352. The problem is that the argument
parsing loop is not using an "else if" where it should, and therefore
the '-r' option ends up unrecognized by GDB.
gdb/ChangeLog:
2017-04-05 Sergio Durigan Junior <sergiodj@redhat.com>
PR gdb/21352
* tracefile.c (tsave_command): Fix argument parsing for '-r'
option.
I see the following test fail from time to time, due to the racy test
in gdb.threads/thread-specific-bp.exp.
continue -a^M
Continuing.^M
^M
Thread 1 "thread-specific" hit Breakpoint 4, end () at binutils-gdb/gdb/testsuite/gdb.threads/thread-specific-bp.c:29^M
29 }^M
(gdb) [Thread 0x40322460 (LWP 12950) exited]^M
Thread-specific breakpoint 3 deleted - thread 2 no longer in the thread list.^M
FAIL: gdb.threads/thread-specific-bp.exp: non-stop: continue to end (timeout)
This patch changes gdb_test to gdb_test_multiple to match prompt only
instead of both prompt and anchor.
gdb/testsuite:
2017-04-05 Yao Qi <yao.qi@linaro.org>
* gdb.threads/thread-specific-bp.exp (check_thread_specific_breakpoint):
Use gdb_test_multiple, and don't match anchor.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* MAINTAINERS (Global Maintainers): Add Simon Marchi.
(Write After Approval): Remove Simon Marchi.
Commit ecfb656c37 ("dwarf2read.c: Make dir_index and
file_name_index strong typedefs") added a use of gdb::optional that
triggers bogus -Wmaybe-uninitialized warnings:
GCC trunk is complaining like this:
../../binutils-gdb/gdb/dwarf2read.c: In function void read_formatted_entries(bfd*, const gdb_byte**, line_header*, const comp_unit_head*, void (*)(line_header*, const char*, dir_index, unsigned int, unsigned int)):
../../binutils-gdb/gdb/dwarf2read.c:17779:65: error: fe.file_entry::length may be used uninitialized in this function [-Werror=maybe-uninitialized]
callback (lh, fe.name, fe.d_index, fe.mod_time, fe.length);
^
../../binutils-gdb/gdb/dwarf2read.c:17779:65: error: *((void*)& fe +8) may be used uninitialized in this function [-Werror=maybe-uninitialized]
../../binutils-gdb/gdb/dwarf2read.c:17779:65: error: fe.file_entry::mod_time may be used uninitialized in this function [-Werror=maybe-uninitialized]
../../binutils-gdb/gdb/dwarf2read.c:17779:65: error: fe.file_entry::name may be used uninitialized in this function [-Werror=maybe-uninitialized]
While some older GCCs are complaining like this:
../../binutils-gdb/gdb/dwarf2read.c: In function void read_formatted_entries(bfd*, const gdb_byte**, line_header*, const comp_unit_head*, void (*)(line_header*, const char*, dir_index, unsigned int, unsigned int)):
../../binutils-gdb/gdb/dwarf2read.c:17779:65: error: uint may be used uninitialized in this function [-Werror=maybe-uninitialized]
callback (lh, fe.name, fe.d_index, fe.mod_time, fe.length);
Looking around the web, I see that boost's optional implementation
triggers this kind of issue often too. See:
http://www.boost.org/doc/libs/1_63_0/libs/optional/doc/html/boost_optional/tutorial/gotchas/false_positive_with__wmaybe_uninitialized.html
I noticed that replacing the gdb::optional uses with real C++17
std::optional uses against GCC 7/trunk makes the warnings go away.
Looking at the implementation, AFAICS, libstdc++ always initializes
its "empty" union payload member (_M_empty, which is defined as an
empty class, like ours). I.e., all payload types have this ctor:
struct _Optional_payload.....
{
constexpr _Optional_payload()
: _M_empty() {}
The constexpr makes a diference too. Without it, GCC7 still warns.
So I'm applying the same treatment to our gdb::optional.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* common/gdb_optional.h (optional::optional): Make constexpr and
initialize m_dummy.
The bsd-uthread target supports an old thread library ("libc_r") that
was last included in FreeBSD release 6.4 released in 2008. However,
this library has not been used as the default library since FreeBSD
5.0 (2003) and 4.11 (2005). Thread support for modern FreeBSD binaries
is provided via "normal" LWP support in core files and the native
FreeBSD target.
gdb/ChangeLog:
* amd64-fbsd-tdep.c: Remove "bsd-uthread.h" include.
(amd64fbsd_jmp_buf_reg_offset): Remove.
(amd64fbsd_supply_uthread): Remove function.
(amd64fbsd_collect_uthread): Remove function.
(amd64fbsd_init_abi): Don't set bsd-uthread callbacks.
* configure.tgt (i[34567]86-*-freebsd*): Remove bsd-uthread.o.
(x86_64-*-freebsd*): Remove bsd-uthread.o.
(fbsd-nat.c): Update comment.
* i386-fbsd-tdep.c: Remove "bsd-uthread.h" include.
(i386fbsd_jmp_buf_reg_offset): Remove.
(i386fbsd_supply_uthread): Remove function.
(i386fbsd_collect_uthread): Remove function.
(i386fbsd_init_abi): Don't set bsd-uthread callbacks.
FreeBSD last shipped a release for Alpha (6.3) in 2008.
This also removes support for GNU/kFreeBSD on Alpha.
gdb/ChangeLog:
* Makefile.in (ALL_64_TARGET_OBS): Remove alpha-fbsd-tdep.o.
(ALLDEPFILES): Remove alpha-fbsd-tdep.c
* NEWS: Mention that support for FreeBSD/alpha was removed.
* alpha-fbsd-tdep.c: Delete file.
* config/alpha/fbsd.mh: Delete file.
* configure.host: Delete alpha*-*-freebsd* and
alpha*-*-kfreebsd*-gnu.
* configure.tgt: Delete alpha*-*-freebsd* and
alpha*-*-kfreebsd*-gnu.
While I was looking at the file, I noticed that this struct could be
nicely converted to a class. As I was progressing, I ended up moving
all state machine actual internal state manipulation to methods of
lnp_state_machine, essentially decoupling DWARF parsing from state
tracking. I also noticed that the lnp_reader_state doesn't really
serve any good use, so that's eliminated in the process.
gdb/ChangeLog:
2017-04-04 Pedro Alves <palves@redhat.com>
* dwarf2read.c (lnp_state_machine): Now a class. Initialize all
data fields, make them private and add "m_" prefixes.
(lnp_state_machine::lnp_state_machine): New ctor.
(record_line, check_line_address, handle_set_discriminator)
(handle_set_address, handle_advance_pc, handle_special_opcode)
(handle_advance_line, handle_set_file, handle_negate_stmt)
(handle_const_add_pc, handle_fixed_advance_pc, handle_copy)
(end_sequence, advance_line): New methods.
(m_gdbarch, m_record_lines_p): New fields.
(lnp_reader_state): Delete.
(dwarf_record_line): Rename to ...
(lnp_state_machine::record_line): ... adjust.
(init_lnp_state_machine): Delete.
(lnp_state_machine::lnp_state_machine): New.
(check_line_address): Rename to ...
(lnp_state_machine::check_line_address): This.
(dwarf_decode_lines_1): Remove reference to "reader_state".
Adjust lnp_state_machine having a non-default ctor. Use bool.
State machine internal state manipulation moved to
lnp_state_machine methods.
A while ago, back when GDB was a C program, the sect_offset and
cu_offset types were made structs in order to prevent incorrect mixing
of those offsets. Now that we require C++11, we can make them
integers again, while keeping the safety, by exploiting "enum class".
We can add a bit more safety, even, by defining operators that the
types _should_ support, helping making the suspicious uses stand out
more.
Getting at the underlying type is done with the new to_underlying
function added by the previous patch, which also helps better spot
where do we need to step out of the safety net. Mostly, that's around
parsing the DWARF, and when we print the offset for complaint/debug
purposes. But there are other occasional uses.
Since we have to define the sect_offset/cu_offset types in a header
anyway, I went ahead and generalized/library-fied the idea of "offset"
types, making it trivial to add more such types if we find a use. See
common/offset-type.h and the DEFINE_OFFSET_TYPE macro.
I needed a couple generaly-useful preprocessor bits (e.g., yet another
CONCAT implementation), so I started a new common/preprocessor.h file.
I included units tests covering the "offset" types API. These are
mostly compile-time tests, using SFINAE to check that expressions that
shouldn't compile (e.g., comparing unrelated offset types) really are
invalid and would fail to compile. This same idea appeared in my
pending enum-flags revamp from a few months ago (though this version
is a bit further modernized compared to what I had posted), and I plan
on reusing the "check valid expression" bits added here in that
series, so I went ahead and defined the CHECK_VALID_EXPR macro in its
own header -- common/valid-expr.h. I think that's nicer regardless.
I was borderline between calling the new types "offset" types, or
"index" types, BTW. I stuck with "offset" simply because that's what
we're already calling them, mostly.
gdb/ChangeLog:
2017-04-04 Pedro Alves <palves@redhat.com>
* Makefile.in (SUBDIR_UNITTESTS_SRCS): Add
unittests/offset-type-selftests.c.
(SUBDIR_UNITTESTS_OBS): Add offset-type-selftests.o.
* common/offset-type.h: New file.
* common/preprocessor.h: New file.
* common/traits.h: New file.
* common/valid-expr.h: New file.
* dwarf2expr.c: Include "common/underlying.h". Adjust to use
sect_offset and cu_offset strong typedefs throughout.
* dwarf2expr.h: Adjust to use sect_offset and cu_offset strong
typedefs throughout.
* dwarf2loc.c: Include "common/underlying.h". Adjust to use
sect_offset and cu_offset strong typedefs throughout.
* dwarf2read.c: Adjust to use sect_offset and cu_offset strong
typedefs throughout.
* gdbtypes.h: Include "common/offset-type.h".
(cu_offset): Now an offset type (strong typedef) instead of a
struct.
(sect_offset): Likewise.
(union call_site_parameter_u): Rename "param_offset" field to
"param_cu_off".
* unittests/offset-type-selftests.c: New file.
This should help catch mistakes related to mixing the 1-based DWARF
indexes with 0-based std::vector indexes, since the new types do not
implicitly convert to anything.
The change in read_formatted_entries relates to the fact that doing
the seemingly simpler:
- uintp = &fe.dir_index;
+ uintp = (unsigned int *) &fe.dir_index;
would be undefined C/C++. So to address that, I made the function
extract the form before assigning to the file_entry. It felt natural
to use gdb::optional for "do I have this value", and this is what
motivated the previous patch that added the missing observer methods
to gdb::optional.
gdb/ChangeLog:
2017-04-04 Pedro Alves <palves@redhat.com>
* common/underlying.h: New file.
* dwarf2read.c: Include "common/gdb_optional.h" and
"common/underlying.h".
(dir_index, file_name_index): New types.
(file_entry): Use them.
(file_entry::include): Use to_underlying.
(line_header::add_file_name): Use dir_index.
(read_formatted_entries): Use gdb::optional. Read form before
writting to file_entry.
(dwarf_decode_line_header): Use dir_index.
(lnp_state_machine::current_file): Use to_underlying.
(lnp_state_machine::file): Change type to file_name_index.
(dwarf_record_line): Use to_underlying.
(init_lnp_state_machine): Use file_name_index.
(dwarf_decode_lines_1): Use dir_index and file_name_index.
Currently, gdb::optional is really minimal and can only be used for
lazy initialization. There's no way to get at the value contained
inside the optinal. This commit corrects that, by adding observer
methods, mostly copied from libstdc++'s implementation of C++17
std::optional.
This will be used in the following patch.
gdb/ChangeLog:
2017-04-04 Pedro Alves <palves@redhat.com>
* common/gdb_optional.h (gdb::optiona): Add operator->, operator*,
operator bool, has_value and get methods.
This starts off as replacing a couple custom open coded vector
implementations in the file with std::vector, and then the rest falls
off of that. I.e., use new/delete instead of XCNEW/xfree, add
ctors/dtors/initializers where appropriate. And then use
std::unique_ptr instead of cleanups. Some functions became methods,
and in a couple spots, some single-use callback functions that would
have to be tweaked anyway are converted to lambdas instead.
gdb/ChangeLog:
2017-04-04 Pedro Alves <palves@redhat.com>
* dwarf2read.c (struct file_entry): Add ctors, and initialize all
fields.
(line_header): Initialize all data fields. Change type of
standard_opcode_lengths to std::unique_ptr<unsigned char[]>.
Change type of include_dirs to std::vector<const char *>. Remove
num_include_dirs, include_dirs_size. Change type of file_names to
std::vector<file_entry>. Remove num_file_names, file_names_size.
(line_header::line_header): New.
(line_header::add_include_dir, line_header::add_file_name): New
methods.
(line_header::include_dir_at): Remove NULL check.
(line_header::file_name_at): Add const overload.
(line_header_up): New unique_ptr typedef.
(dw2_get_file_names_reader): Use line_header_up. Adjust to use
std::vector. Remove free_line_header call.
(dwarf2_build_include_psymtabs): Use line_header_up. Remove
free_line_header call.
(free_cu_line_header): Delete.
(handle_DW_AT_stmt_list, handle_DW_AT_stmt_list)
(setup_type_unit_groups): Use line_header_up instead of cleanups.
Adjust to use std::vector.
(free_line_header): Delete.
(free_line_header_voidp): Use delete.
(add_include_dir): Replace with ...
(line_header::add_include_dir): ... this method. Use std::vector.
(add_file_name): Replace with ...
(line_header::add_file_name): ... this method. Use std::vector.
(add_include_dir_stub): Delete.
(read_formatted_entries): Remove memset.
(dwarf_decode_line_header): Return a line_header_up instead of a
raw pointer. Remove cleanup handling. Pass lambdas to
read_formatted_entries. Adjust to use line_header methods.
(dwarf_decode_lines_1): Adjust to use line_header methods.
(dwarf_decode_lines, file_file_name, file_full_name): Adjust to
use std::vector.
In some situations, the dependency tracking files in .deps can refer to
source files that were removed or renamed, leading to errors like:
make: *** No rule to make target `version.c', needed by `version.o'. Stop.
This patch makes the clean target clear the .deps directory, which gives
the user a chance to recover from the error wihtout knowing about the
internals of the build system.
It is already done for GDB. See here for more details:
https://sourceware.org/ml/gdb-patches/2009-03/msg00000.html
gdb/gdbserver/ChangeLog:
* Makefile.in (clean): Clear .deps.
It's the only place in the codebase that uses "struct ptid", so change
it to ptid_t for consistency.
gdb/ChangeLog:
* remote.c (set_general_thread, set_continue_thread): Use ptid_t
instead of struct ptid.
I noticed that there were some missing files in gdbserver's gitignore
(some generated register format .c files). Of course the easy fix would
be to add those files to .gitignore, but I think we can do a better job,
so that we don't have to worry about adding generated files to
.gitignore or the clean Makefile target.
I suggest naming all generated source files -generated.c. This way, we
can use a single rule in .gitignore and do a "rm -f *-generated.c" to
clean them up.
New in v2:
- Don't rename version.o and xml-builtin.o
gdb/gdbserver/ChangeLog:
* .gitignore: Remove generated files, replace with wildcard.
* (clean): Replace removal of generated files with wildcard.
(version.c): Replace with...
(version-generated.c): ...this.
(xml-builtin.c): Replace with...
(xml-builtin-generated.c): ...this.
(%-ipa.o: %-generated.c, %.o: %-generated.c): New rules.
(%.c: *regformats*): Replace with...
(%-generated.c: *regformats*): ...this.
The two-tier lexer in gdb/d-exp.y, which resolves fully qualified names
missed a case where `module.type' was not being classified as one token.
And so when the grammar takes over, it matched the remaining tokens
against the rule `TypeExp . IdentifierExp', where we were expecting to
instead match cast expression `( TypeExp ) UnaryExpression'.
Adding a case for TYPE_CODE_MODULE in type_aggregate_p means that
classify_inner_name will get a chance to lookup the symbol.
This was noticed when using `watch -l', and got the rather confusing
response:
A syntax error in expression, near `) 0x0add4e55'.
So it's been included in the testsuite, along with another test that
does effectively the same expression, but explicitly.
gdb/ChangeLog:
* d-exp.y (type_aggregate_p): Treat TYPE_CODE_MODULE as being
aggregate-like.
gdb/testsuite/ChangeLog:
* gdb.dlang/watch-loc.c: New file.
* gdb.dlang/watch-loc.exp: New file.
$ gdb rustc
Reading symbols from rustc...Reading symbols from /usr/lib/debug/usr/bin/rustc.debug...done.
done.
warning: Invalid entry in .debug_gdb_scripts section
/usr/bin/rustc
Section Headers:
[Nr] Name Type Address Off Size ES Flg Lk Inf Al
[15] .debug_gdb_scripts PROGBITS 00000000000008ed 0008ed 000022 00 AMS 0 0 1
/usr/lib/debug/usr/bin/rustc.debug
Section Headers:
[Nr] Name Type Address Off Size ES Flg Lk Inf Al
[15] .debug_gdb_scripts NOBITS 00000000000008ed 000280 000022 00 AMS 0 0 1
There remains questionable whether bfd_get_section_by_name() should not return
an error for !SEC_LOAD but I haven't investigated that.
gdb/ChangeLog
2017-03-29 Jan Kratochvil <jan.kratochvil@redhat.com>
* auto-load.c (auto_load_section_scripts): Check SEC_HAS_CONTENTS.
gdb/testsuite/ChangeLog
2017-03-29 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdb.python/py-section-script.exp (sepdebug): New testcases.
This patch adds constructor and destructor to thread_info.
gdb:
2017-03-29 Yao Qi <yao.qi@linaro.org>
* gdbthread.h (struct thread_info): Declare constructor and
destructor. Add some in-class member initializers.
* thread.c (free_thread): Remove.
(init_thread_list): Call delete instead of free_thread.
(new_thread): Call thread_info constructor.
(thread_info::thread_info): New function.
(thread_info::~thread_info): New function.
(delete_thread_1): Call delete instead of free_thread.
(make_cleanup_restore_current_thread): Move tp and frame to
inner block.
Add a prologue analysis that recognizes all instructions that may happen in
compiler-generated prologue, including various stores, core register moves,
subtraction and ENTER_S instruction that does a lot of prologue actions through
microcode.
Testcases cover various prologue scenarios, including instructions that are
spread across multiple 16-bit encodings (for example there are 7 encodings of
store instruction).
gdb/ChangeLog:
yyyy-mm-dd Anton Kolesov <anton.kolesov@synopsys.com>
* arc-tdep.c (arc_frame_cache): Add support for prologue analysis.
(arc_skip_prologue): Likewise.
(arc_make_frame_cache): Likewise.
(arc_pv_get_operand): New function.
(arc_is_in_prologue): Likewise.
(arc_analyze_prologue): Likewise.
(arc_print_frame_cache): Likewise.
(MAX_PROLOGUE_LENGTH): New constant.
gdb/doc/ChangeLog:
yyyy-mm-dd Anton Kolesov <anton.kolesov@synopsys.com>
* gdb.texinfo (Synopsys ARC): Document "set debug arc 2".
gdb/testsuite/ChangeLog:
yyyy-mm-dd Anton Kolesov <anton.kolesov@synopsys.com>
* gdb.arch/arc-analyze-prologue.S: New file.
* gdb.arch/arc-analyze-prologue.exp: Likewise.
Add disassembler helper for GDB, that uses opcodes structure arc_instruction
and adds convenience functions to handle instruction operands. This interface
solves at least those problems with arc_instruction:
* Some instructions, like "push_s", have implicit operands which are not
directly present in arc_instruction.
* Operands of particular meaning, like branch/jump targets, have various
locations and meaning depending on type of branch/target.
* Access to operand value is abstracted into a separate function, so callee
code shouldn't bother if operand value is an immediate value or in a
register.
Testcases included in this commit are fairly limited - they test exclusively
branch instructions, something that will be used in software single stepping.
Most of the other parts of this disassembler helper are tested during prologue
analysis testing.
gdb/ChangeLog:
yyyy-mm-dd Anton Kolesov <anton.kolesov@synopsys.com>
* configure.tgt: Add arc-insn.o.
* arc-tdep.c (arc_delayed_print_insn): Make non-static.
(dump_arc_instruction_command): New function.
(arc_fprintf_disasm): Likewise.
(arc_disassemble_info): Likewise.
(arc_insn_get_operand_value): Likewise.
(arc_insn_get_operand_value_signed): Likewise.
(arc_insn_get_memory_base_reg): Likewise.
(arc_insn_get_memory_offset): Likewise.
(arc_insn_get_branch_target): Likewise.
(arc_insn_dump): Likewise.
(arc_insn_get_linear_next_pc): Likewise.
* arc-tdep.h (arc_delayed_print_insn): Add function declaration.
(arc_disassemble_info): Likewise.
(arc_insn_get_branch_target): Likewise.
(arc_insn_get_linear_next_pc): Likewise.
* NEWS: Mention new "maint print arc arc-instruction".
gdb/doc/ChangeLog:
yyyy-mm-dd Anton Kolesov <anton.kolesov@synopsys.com>
* gdb.texinfo (Synopsys ARC): Add "maint print arc arc-instruction".
gdb/testsuite/ChangeLog:
yyyy-mm-dd Anton Kolesov <anton.kolesov@synopsys.com>
* gdb.arch/arc-decode-insn.S: New file.
* gdb.arch/arc-decode-insn.exp: Likewise.
Add an "arc" sublist to "maintenance print" command list. The list is empty
for now, its purpose is to contain commands that print internal state of some
ARC-specific structures.
gdb/ChangeLog:
yyyy-mm-dd Anton Kolesov <anton.kolesov@synopsys.com>
* arc-tdep (maintenance_print_arc_list): New variable.
(maintenance_print_arc_command): New function.
Add ARC_LIMM_REGNUM to arc_regnum enumeration and assign a number 62 to it.
This ensures that for core registers internal register numbers in this enum are
the same as architectural numbers. This allows to use internal register
numbers in the contexts where architectural number is implied, for example when
disassembling instruction during prologue analysis.
gdb/ChangeLog:
yyyy-mm-dd Anton Kolesov <anton.kolesov@synopsys.com>
* arc-tdep.c (core_v2_register_names, core_arcompact_register_names)
Add "limm" and "reserved".
(arc_cannot_fetch_register, arc_cannot_store_register): Add
ARC_RESERVED_REGNUM and ARC_LIMM_REGNUM.
* arc-tdep.h (arc_regnum): Likewise.
Provide aceess to the THREADPTR register to remote gdb.
gdb/gdbserver/
2017-03-27 Max Filippov <jcmvbkbc@gmail.com>
* linux-xtensa-low.c (regnum::R_THREADPTR): New enum member.
(xtensa_fill_gregset): Call collect_register_by_name for
threadptr register.
(xtensa_store_gregset): Call supply_register_by_name for
threadptr register.
There are no a0-a15 pseudoregisters at the top of the register set in
call0 registers layout. All registers above gdbarch_num_regs (gdbarch)
are privileged. Treat them as such. This fixes the following gdb
assertion seen when 'finish' command is invoked:
regcache.c:649: internal-error: register_status
regcache_raw_read(regcache*, int, gdb_byte*):
Assertion `regnum >= 0 && regnum < regcache->descr->nr_raw_registers'
failed.
gdb/
2017-03-27 Max Filippov <jcmvbkbc@gmail.com>
* xtensa-tdep.c (xtensa_pseudo_register_read): Treat all
registers above gdbarch_num_regs (gdbarch) as privileged in
call0 ABI.
Correctly handle a0- registers on requests from remote gdb. This fixes
'Register 1 is not available'
and subsequent assertion in the remote gdb connecting to the gdbserver:
'findvar.c:291: internal-error: value_of_register_lazy:
Assertion `frame_id_p(get_frame_id (frame))' failed.'
The register structure is the same for windowed and call0 ABIs because
currently linux kernel internally requires windowed registers, so they
are always present.
gdb/gdbserver/
2017-03-27 Max Filippov <jcmvbkbc@gmail.com>
* linux-xtensa-low.c (xtensa_fill_gregset): Call collect_register
for all registers in a0_regnum..a0_regnum + C0_NREGS range.
(xtensa_store_gregset): Call supply_register for all registers in
a0_regnum..a0_regnum + C0_NREGS range.
Correctly handle a0- registers. This allows debugging call0 code in
linux natively.
The register structure is the same for windowed and call0 ABIs because
currently linux kernel internally requires windowed registers, so they are
always present.
gdb/
2017-03-27 Max Filippov <jcmvbkbc@gmail.com>
* xtensa-linux-nat.c (fill_gregset): Call regcache_raw_collect
for a single specified register or for all registers in
a0_base..a0_base + C0_NREGS range.
(supply_gregset_reg): Call regcache_raw_supply for a single
specified register or for all registers in a0_base..a0_base +
C0_NREGS range.
Both gdb and gdbserver need this definition to properly work with call0
ABI. Make it available to both.
gdb/
2017-03-27 Max Filippov <jcmvbkbc@gmail.com>
* arch/xtensa.h (C0_NREGS): Add definition.
* xtensa-tdep.c (C0_NREGS): Remove definition.
This fixes segfault in native gdb because isa is not initialized at the
point of call to xtensa_isa_maxlength.
gdb/
2017-03-27 Max Filippov <jcmvbkbc@gmail.com>
* xtensa-tdep.c (xtensa_scan_prologue, call0_analyze_prologue):
Drop xtensa_default_isa initialization.
(xtensa_gdbarch_init): Initialize xtensa_default_isa.
Multiple places in dwarf2read.c open code 1-based to 0-based index
conversion and check for out of bounds accesses to lh->include_dirs
and lh->file_names. This commit factors those out to a couple methods
and uses them throughout.
gdb/ChangeLog:
2017-03-27 Pedro Alves <palves@redhat.com>
* dwarf2read.c (file_entry) <dir_index>: Add comment.
(file_entry::include_dir): New method.
(line_header::include_dir_at, line_header::file_name_at): New
methods.
(setup_type_unit_groups, setup_type_unit_groups)
(psymtab_include_file_name): Simplify using the new methods.
(lnp_state_machine) <the_line_header>: New field.
<file>: Add comment.
(lnp_state_machine::current_file): New method.
(dwarf_record_line): Simplify using the new methods.
(init_lnp_state_machine): Initialize the "the_line_header" field.
(dwarf_decode_lines_1, dwarf_decode_lines, file_file_name):
Simplify using the new methods.
The demangler exports the cplus_demangle_fill_component function that
clients should use to initialize demangle_component components that
use the "s_binary" union member. cp-name-parser.y uses it in some
places, via the fill_comp wrapper, but not all. Several places
instead use a GDB-specific "make_empty" function. Because this
function does not call any of the demangler "fill" functions, we had
to patch it recently to clear the allocated demangle_component's
"d_printing" field, which is supposedly a "private" demangler field.
To avoid such problems in the future, this commit switches those
places to use "fill_comp" instead, and eliminates the "make_empty"
function.
gdb/ChangeLog:
2017-03-27 Pedro Alves <palves@redhat.com>
* cp-name-parser.y (make_empty): Delete.
(demangler_special, nested_name, ptr_operator, array_indicator)
(direct_declarator, declarator_1): Use fill_comp instead of
make_empty.
The declarations of gdb_xml_debug and gdb_xml_error are passing "0" as
"first-to-check" argument to ATTRIBUTE_PRINTF, as if they were va_args
functions. Consequently, the arguments to gdb_xml_debug /
gdb_xml_error aren't being checked against the format strings.
With that fixed, a couple obvious bugs are exposed, both fixed by this
commit.
gdb/ChangeLog:
2017-03-27 Pedro Alves <palves@redhat.com>
* xml-support.h (gdb_xml_debug): Pass a "first-to-check" argument
to ATTRIBUTE_PRINTF.
* solib-target.c (library_list_start_list): Print "string" not
"version".
* xml-tdesc.c (tdesc_start_field): Pass "field_name" to
gdb_xml_error call.
Eliminates several uses of cleanups.
Tested on x86_64 Fedora 23 with Python 2 and 3.
gdb/ChangeLog
2017-03-27 Pedro Alves <palves@redhat.com>
* dwarf2read.c (struct file_and_directory): New.
(dwarf2_get_dwz_file): Adjust to use std::string.
(dw2_get_file_names_reader): Adjust to use file_and_directory.
(find_file_and_directory): Adjust to return a file_and_directory
object.
(read_file_scope): Adjust to use file_and_directory. Remove
make_cleanup/do_cleanups calls.
(open_and_init_dwp_file): Adjust to use std::string. Remove
make_cleanup/do_cleanups calls.
* python/python.c (do_start_initialization): Adjust to ldirname
returning a std::string.
* utils.c (ldirname): Now returns a std::string.
* utils.h (ldirname): Change return type to std::string.
* xml-syscall.c (xml_init_syscalls_info): Adjust to ldirname
returning a std::string.
* xml-tdesc.c (file_read_description_xml): Likewise.
gdb was segfaulting during backtrace on a binary here, where
fe->dir_index parsed from the DWARF info was seen to access beyond the
provided include_dirs array.
This commit bounds the access to entries actually written to the
array, and was verified to output the backtrace correctly.
gdb/ChangeLog:
* dwarf2read.c (setup_type_unit_groups): Ensure dir_index doesn't
reference beyond the 'lh->include_dirs' array before accessing to
it.
(psymtab_include_file_name): Likewise.
(dwarf_decode_lines_1): Likewise.
(dwarf_decode_lines): Likewise.
(file_file_name): Likewise.
Now that the to_fetch_registers, to_store_registers and
to_prepare_to_store target methods don't rely on the value of
inferior_ptid anymore, we can remove a bunch of now unnecessary setting
and restoring of inferior_ptid.
The asserts added recently in target_fetch_registers and
target_store_registers, which validate that inferior_ptid matches the
regcache's ptid, must go away. It's the whole point of this effort, to
not require inferior_ptid to have a particular value when calling these
functions.
One thing that I noticed is how sol-thread.c's ps_lgetregs and friends
use the current value of inferior_ptid instead of what's passed as
argument (ph->ptid), unlike proc-service.c's versions of the same
functions. Is it expected? I left it like this in the current patch,
but unless there's a good reason for it to be that way, I guess we
should make it use the parameter.
gdb/ChangeLog:
* fbsd-tdep.c (fbsd_corefile_thread): Don't set/restore
inferior_ptid.
* proc-service.c (ps_lgetregs, ps_lsetregs, ps_lgetfpregs,
ps_lsetfpregs): Likewise.
* regcache.c (regcache_raw_update, regcache_raw_write): Likewise.
* sol-thread.c (ps_lgetregs, ps_lsetregs, ps_lgetfpregs,
ps_lsetfpregs): Likewise.
* target.c (target_fetch_registers, target_store_registers):
Remove asserts.
This patch fixes the bug of handling PRFM instruction. PRFM is documented
in a table with other load and store instructions, but it doesn't do any
load or store. This patch also adds a unit test to PRFM instruction.
gdb:
2017-03-23 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_process_record_test): Declare.
(_initialize_aarch64_tdep): Register it.
(aarch64_record_load_store): Handle PRFM instruction.
(aarch64_process_record_test): New function.
The patch "Fix memory leak in python.c:do_start_initialization"
(https://sourceware.org/ml/gdb-patches/2017-03/msg00407.html) introduced a
compilation error on some platforms:
../../binutils-gdb/gdb/python/python.c: In function bool do_start_initialization():
../../binutils-gdb/gdb/python/python.c:1556:16: error: invalid conversion from const void* to void* [-fpermissive]
xfree (libdir);
^
This is fixed by removing the constness of libdir's data type.
gdb/ChangeLog:
* python/python.c (do_start_initialization): Remove 'const' from
data type of libdir.
When intializing Python the path to the python binary is build the
following way
progname = concat (ldirname (python_libdir), SLASH_STRING, "bin",
SLASH_STRING, "python", (char *) NULL);
This is problematic as both concat and ldirname allocate memory for the
string they return. Thus the memory allocated by ldirname cannot be
accessed afterwards causing a memory leak. Fix it by temporarily storing
libdir in a variable and xfree it after concat.
gdb/ChangeLog:
python/python.c (do_start_initialization): Fix memory leak.
The linux_nat_xfer_partial does a conversion of inferior_ptid: if it's
an LWP (ptid::lwp != 0), it builds a new ptid with the lwp as
the pid and assigns that temporarily to inferior_ptid. For example, if
inferior_ptid is:
{ .pid = 1234, .lwp = 1235 }
it will assign this to inferior_ptid for the duration of the call:
{ .pid = 1235, .lwp = 0 }
Instead of doing this, this patch teaches the inf-ptrace implementation
of xfer_partial to deal with ptids representing lwps by using
get_ptrace_pid.
Also, in linux_proc_xfer_spu and linux_proc_xfer_partial, we use ptid_get_lwp
instead of ptid_get_pid. While not strictly necessary, since the content of
/proc/<pid> and /proc/<lwp> should be the same, it's a bit safer, because:
- some files under /proc/<pid>/ may not work if the <pid> thread is
running, just like ptrace requires a stopped thread. The current
thread's lwp id is more likely to be in the necessary state (stopped).
- if the leader (<pid>) had exited and is thus now zombie, then several
files under "/proc/<pid>" won't work, while they will if you use
"/proc/<lwp>".
The testsuite found no regression on native amd64 linux.
gdb/ChangeLog:
* inf-ptrace.c (inf_ptrace_xfer_partial): Get pid from ptid
using get_ptrace_pid.
* linux-nat.c (linux_nat_xfer_partial): Don't set/restore
inferior_ptid.
(linux_proc_xfer_partial, linux_proc_xfer_spu): Use lwp of
inferior_ptid instead of pid.
There's no reason to use @code for Python the name of a programming
language.
gdb/doc:
2017-03-22 Yao Qi <yao.qi@linaro.org>
* python.texi (Inferiors In Python): Remove @code from Python.
Both aarch64-tdep.c and arm-tdep.c defines a class instruction_reader, which
violates ODR, but linker doesn't an emit error. I fix this issue by wrapping
them by anonymous namespace, but I think it is better to apply this for all
locally used classes.
If it is a good idea to put locally used class into anonymous namespace, we
should document this rule into GDB coding convention, or even GCC coding
convention. Note that anonymous namespace has been used in GCC but GCC
coding convention doesn't mention the it.
gdb:
2017-03-22 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c: Wrap locally used classes in anonymous
namespace.
* arm-tdep.c: Likewise.
* linespec.c: Likewise.
* ui-out.c: Likewise.
Pick up missing bits from the patch merged in.
2017-03-22 Jonah Graham <jonah@kichwacoders.com>
PR gdb/19637
* python/lib/gdb/printer/bound_registers.py: Import sys.
Use the ptid from the regcache so we don't depend on the current value
of the inferior_ptid global.
Also, change how the current thread is passed to sub-functions. The
windows_fetch_inferior_registers function sets current_thread then calls
do_windows_fetch_inferior_registers, which reads current_thread. This
very much looks like passing a parameter through a global variable. I
think it would be more straightforward to pass the thread as a
parameter.
gdb/ChangeLog:
* windows-nat.c (do_windows_fetch_inferior_registers): Add
windows_thread_info parameter and use it instead of
current_thread.
(windows_fetch_inferior_registers): Don't set current_thread,
pass the thread to do_windows_fetch_inferior_registers. Use
ptid from regcache instead of inferior_ptid.
(do_windows_store_inferior_registers): Add windows_thread_info
parameter and use it instead of current_thread.
(windows_store_inferior_registers): Don't set current_thread,
pass the thread to do_windows_store_inferior_registers. Use
ptid from regcache instead of inferior_ptid.
I get this when trying to build for --host=x68_64-w64-mingw32:
/home/emaisin/src/binutils-gdb/gdb/ser-mingw.c: In function 'void ser_windows_raw(serial*)':
/home/emaisin/src/binutils-gdb/gdb/ser-mingw.c:166:8: error: 'struct serial' has no member named 'current_timeout'
scb->current_timeout = 0;
^~~~~~~~~~~~~~~
It is just a leftover from
9bcbdca808
PR remote/21188: Fix remote serial timeout
gdb/ChangeLog:
* ser-mingw.c (ser_windows_raw): Remove reference to
struct serial::current_timeout.
While at it, decode also properly one-bit flags for %fsr (accrued and
current exception flags were mixed up).
ChangeLog entry:
2017-03-21 Ivo Raisr <ivo.raisr@oracle.com>
PR tdep/20928
* gdb/sparc-tdep.h (gdbarch_tdep) <sparc64_ccr_type>: New field.
* gdb/sparc64-tdep.c (sparc64_ccr_type): New function.
(sparc64_fsr_type): Fix %fsr decoding.
ChangeLog entry for testsuite:
2017-03-21 Ivo Raisr <ivo.raisr@oracle.com>
PR tdep/20928
* gdb.arch/sparc64-regs.exp: New file.
* gdb.arch/sparc64-regs.S: Likewise.
This changes the return type of "gdb.BtraceInstruction.data ()" from
"memoryview" to "buffer" on Python 2.7 and below, similar to what
"gdb.Inferior.read_memory ()" does.
The implementations of to_fetch_registers/to_store_registers in the spu
code use some functions that rely on inferior_ptid. It's simpler for
now to set/restore inferior_ptid.
gdb/ChangeLog:
* spu-linux-nat.c (spu_fetch_inferior_registers,
spu_store_inferior_registers): Use ptid from regcache, set and
restore inferior_ptid.
* spu-multiarch.c (spu_fetch_registers, spu_store_registers):
Likewise.
This patch adds tests for the initial rvalue reference support patchset. All
of the new tests are practically mirrored regular references tests and, except
for the demangler ones, are introduced in new files, which are set to be
compiled with -std=gnu++11. Tested are printing of rvalue reference types and
values, rvalue reference parameters in function overloading, demangling of
function names containing rvalue reference parameters, casts to rvalue
reference types, application of the sizeof operator to rvalue reference types
and values, and support for rvalue references within the gdb python module.
gdb/ChnageLog
PR gdb/14441
* NEWS: Mention support for rvalue references in GDB and python.
* doc/gdb.texinfo (C Plus Plus Expressions): Mention that GDB
supports both lvalue and rvalue references.
gdb/testsuite/ChangeLog
PR gdb/14441
* gdb.cp/demangle.exp: Add rvalue reference tests.
* gdb.cp/rvalue-ref-casts.cc: New file.
* gdb.cp/rvalue-ref-casts.exp: New file.
* gdb.cp/rvalue-ref-overload.cc: New file.
* gdb.cp/rvalue-ref-overload.exp: New file.
* gdb.cp/rvalue-ref-params.cc: New file.
* gdb.cp/rvalue-ref-params.exp: New file.
* gdb.cp/rvalue-ref-sizeof.cc: New file.
* gdb.cp/rvalue-ref-sizeof.exp: New file.
* gdb.cp/rvalue-ref-types.cc: New file.
* gdb.cp/rvalue-ref-types.exp: New file.
* gdb.python/py-rvalue-ref-value-cc.cc: New file.
* gdb.python/py-rvalue-ref-value-cc.exp: New file.
This patch introduces changes to rank_one_type() dealing with ranking an rvalue
reference type when selecting a best viable function from a set of candidate
functions. The 4 new added rules for rvalue references are:
1) An rvalue argument cannot be bound to a non-const lvalue reference parameter
and an lvalue argument cannot be bound to an rvalue reference parameter.
[C++11 13.3.3.1.4p3]
2) If a conversion to one type of reference is an identity conversion, and a
conversion to the second type of reference is a non-identity conversion, choose
the first type. [C++11 13.3.3.2p3]
3) An rvalue should be first tried to bind to an rvalue reference, and then to
an lvalue reference. [C++11 13.3.3.2p3]
4) An lvalue reference to a function gets higher priority than an rvalue
reference to a function. [C++11 13.3.3.2p3]
This patch is not exactly correct. See c++/15372 for tracking overload
resolution bugs.
gdb/ChangeLog
PR gdb/14441
* gdbtypes.c (rank_one_type): Implement overloading
resolution rules regarding rvalue references.
This patch adds the ability to inspect rvalue reference types and values using
the gdb python module. This is achieved by creating two wrappers for
valpy_reference_value(), using the ReferenceExplorer class to handle the
objects of rvalue reference types and placing necessary checks for a
TYPE_CODE_RVALUE_REF type code next to the checks for a TYPE_CODE_REF type
code.
gdb/ChangeLog
PR gdb/14441
* doc/python.texi (Types in Python): Add TYPE_CODE_RVALUE_REF to
table of constants.
* python/lib/gdb/command/explore.py: Support exploring values
of rvalue reference types.
* python/lib/gdb/types.py: Implement get_basic_type() for
rvalue reference types.
* python/py-type.c (pyty_codes) <TYPE_CODE_RVALUE_REF>: New
constant.
* python/py-value.c (valpy_getitem): Add an rvalue reference
check.
(valpy_reference_value): Add new parameter "refcode".
(valpy_lvalue_reference_value, valpy_rvalue_reference_value):
New wrappers for valpy_reference_value().
* python/py-xmethods.c (gdbpy_get_xmethod_result_type)
(gdbpy_invoke_xmethod): Likewise.
Make gdb DWARF reader understand the DW_TAG_rvalue_reference type tag. Handling
of this tag is done in the existing read_tag_reference_type() function, to
which we add a new parameter representing the kind of reference type
(lvalue vs rvalue).
gdb/ChangeLog
PR gdb/14441
* dwarf2read.c (process_die, read_type_die_1): Handle the
DW_TAG_rvalue_reference_type DIE.
(read_tag_reference_type): Add new parameter "refcode".
This patch provides the ability to print out names of rvalue reference types
and values of those types. This is done in full similarity to regular
references, and as with them, we don't print out "const" suffix because all
rvalue references are const.
gdb/ChangeLog
PR gdb/14441
* c-typeprint.c (c_print_type, c_type_print_varspec_prefix)
(c_type_print_modifier, c_type_print_varspec_suffix)
(c_type_print_base): Support printing rvalue reference types.
* c-valprint.c (c_val_print, c_value_print): Support printing
rvalue reference values.
This patch implements correct parsing of C++11 rvalue reference typenames.
This is done in full similarity to the handling of regular references by adding
a '&&' token handling in c-exp.y, defining an rvalue reference type piece, and
implementing a follow type derivation in follow_types().
gdb/ChangeLog
PR gdb/14441
* c-exp.y (ptr_operator): Handle the '&&' token in the typename.
* parse.c (insert_type): Change assert statement.
(follow_types): Handle rvalue reference types.
* parser-defs.h (enum type_pieces) <tp_rvalue_reference>: New
constant.
Parameterize value_ref() by the kind of reference type the value of which
is requested. Change all callers to use the new API.
gdb/ChangeLog
PR gdb/14441
* ada-lang.c (ada_evaluate_subexp): Adhere to the new
value_ref() interface.
* c-valprint.c (c_value_print): Likewise.
* infcall.c (value_arg_coerce): Likewise.
* python/py-value.c (valpy_reference_value): Likewise.
* valops.c (value_cast, value_reinterpret_cast)
(value_dynamic_cast, typecmp): Likewise.
(value_ref): Parameterize by kind of return value reference type.
* value.h (value_ref): Add new parameter "refcode".
Parameterize lookup_reference_type() and make_reference_type() by the kind of
reference type we want to look up. Create two wrapper functions
lookup_{lvalue,rvalue}_reference_type() for lookup_reference_type() to simplify
the API. Change all callers to use the new API.
gdb/Changelog
PR gdb/14441
* dwarf2read.c (read_tag_reference_type): Use
lookup_lvalue_reference_type() instead of lookup_reference_type().
* eval.c (evaluate_subexp_standard): Likewise.
* f-exp.y: Likewise.
* gdbtypes.c (make_reference_type, lookup_reference_type):
Generalize with rvalue reference types.
(lookup_lvalue_reference_type, lookup_rvalue_reference_type): New
convenience wrappers for lookup_reference_type().
* gdbtypes.h (make_reference_type, lookup_reference_type): Add a
reference kind parameter.
(lookup_lvalue_reference_type, lookup_rvalue_reference_type): Add
wrappers for lookup_reference_type().
* guile/scm-type.c (gdbscm_type_reference): Use
lookup_lvalue_reference_type() instead of lookup_reference_type().
* guile/scm-value.c (gdbscm_value_dynamic_type): Likewise.
* parse.c (follow_types): Likewise.
* python/py-type.c (typy_reference, typy_lookup_type): Likewise.
* python/py-value.c (valpy_get_dynamic_type, valpy_getitem):
Likewise.
* python/py-xmethods.c (gdbpy_get_xmethod_result_type)
(gdbpy_invoke_xmethod): Likewise.
* stabsread.c: Provide extra argument to make_reference_type()
call.
* valops.c (value_ref, value_rtti_indirect_type): Use
lookup_lvalue_reference_type() instead of lookup_reference_type().
This patch introduces preliminal definitions regarding C++11 rvalue references
to the gdb type system. In addition to an enum type_code entry, a field in
struct type and an accessor macro for that which are created similarly to the
lvalue references counterparts, we also introduce a TYPE_REFERENCE convenience
macro used to check for both kinds of references simultaneously as they are
equivalent in many contexts.
gdb/Changelog
PR gdb/14441
* gdbtypes.h (enum type_code) <TYPE_CODE_RVALUE_REF>: New constant.
(TYPE_IS_REFERENCE): New macro.
(struct type): Add rvalue_reference_type field.
(TYPE_RVALUE_REFERENCE_TYPE): New macro.
This change adds the MI equivalent for the "info sharedlibrary"
command. The command was already partially documented but ignored as
it was not implemented. The new MI command works similarly to the CLI
command, taking an optional regular expression as an argument and
outputting the library information.
I included a test for the new command in mi-solib.exp.
gdb/doc/ChangeLog:
* gdb.texinfo (gdb/mi Symbol Query Commands): Document new MI
command file-list-shared-libraries
(GDB/MI Async Records): Update documentation of library-loaded with new
field.
gdb/ChangeLog:
* NEWS: Add an entry about new '-file-list-shared-libraries' command.
* mi/mi-cmd-file.c (mi_cmd_file_list_shared_libraries):
New function definition.
* mi/mi-cmds.c (mi_cmds): Add -file-list-shared-libraries command.
* mi/mi-cmds.h (mi_cmd_file_list_shared_libraries):
New function declaration.
* mi/mi-interp.c (mi_output_solib_attribs): New Function.
* mi/mi-interp.h: New file.
* solib.c (info_sharedlibrary_command): Replace for loop with
ALL_SO_LIBS macro
* solib.h (update_solib_list): New function declaration.
(so_list_head): Move macro.
* solist.h (ALL_SO_LIBS): New macro.
gdb/testsuite/ChangeLog:
* gdb.mi/mi-solib.exp (test_file_list_shared_libraries):
New procedure.
Signed-off-by: Marc-Andre Laperle <marc-andre.laperle@ericsson.com>
When using mi_gdb_test, if it fails because of the presence of
unexpected output, the error message is only the message passed as
the argument with no indication that there was an unexpected output.
This change adds an additional text to the failure message to
indicate that there was an unexpected output.
gdb/testsuite/ChangeLog:
* lib/mi-support.exp (mi_gdb_test): Add additional message
for unexpected output.
Signed-off-by: Marc-Andre Laperle <marc-andre.laperle@ericsson.com>
The target parameter in both solib_add and update_solib_list
functions is not used anymore and as not been used for a while. This
change removes the parameter to clean up the code a little bit.
gdb/ChangeLog:
* infcmd.c (post_create_inferior): Remove unused argument in
call to solib_add.
* remote.c (remote_start_remote): Likewise.
* solib-frv.c (frv_fetch_objfile_link_map): Likewise.
* solib-svr4.c: (svr4_fetch_objfile_link_map): Likewise.
(enable_break): Likewise.
* solib.c (update_solib_list): Remove unused target argument
and its documentation.
(solib_add): Remove unused target argument. Remove unused
argument in call to update_solib_list.
(info_sharedlibrary_command): Remove unused argument in call
to update_solib_list.
(sharedlibrary_command): Remove unused argument in call to
solib_add.
(handle_solib_event): Likewise.
(reload_shared_libraries): Likewise.
* solib.h (solib_add): Remove unused target argument.
Signed-off-by: Marc-Andre Laperle <marc-andre.laperle@ericsson.com>
The test case examine-backward.exp issues the command "x/-s" after the end
of the first string in TestStrings, but without making sure that this
string is preceded by a string terminator. Thus GDB may spuriously print
some random characters from before that string, and then the test fails.
This patch assures that TestStrings is preceded by a string terminator.
gdb/testsuite/ChangeLog:
* gdb.base/examine-backward.c (Barrier): New character array
constant, to ensure that TestStrings is preceded by a string
terminator.
On s390x targets GDB can not handle displaced stepping correctly for some
relative branch instructions, such as cij (compare immediate and branch
relative). When setting a breakpoint on such an instruction and
single-stepping over it, the branch is never taken. This is because the
check in s390_displaced_step_fixup for relative branch instructions is
incomplete.
Instead of completing the list of relative branch instructions to check
against, this patch just treats relative branches and non-branching
instructions in the same way and adjusts the PC with the negated
displacement in both cases.
gdb/ChangeLog:
* s390-linux-tdep.c (is_rsi, is_rie): Remove functions.
(s390_displaced_step_fixup): Cover relative branches with the
default fixup handling. This fixes lack of support for some
relative branch instructions.
This is one of the rare to_fetch/store_registers implementations that will
still rely (for now) on inferior_ptid (because of the memory read/write
operations). We therefore have to add a save/restore of inferior_ptid. We'll
be able to remove it when we make the memory operations accept the ptid as a
parameter.
gdb/ChangeLog:
* bsd-uthread.c (bsd_uthread_fetch_registers,
bsd_uthread_store_registers): Use ptid from regcache, set and
restore inferior_ptid.
As Gareth McMullin <gareth@blacksphere.co.nz> reports at
<https://sourceware.org/ml/gdb-patches/2017-02/msg00560.html>, the
timeout mechanism in ser-unix.c was broken by commit 048094acc
("target remote: Don't rely on immediate_quit (introduce quit
handlers)").
Instead of applying a local fix, and since we now finally always use
interrupt_select [1], let's get rid of hardwire_readchar entirely, and
use ser_base_readchar instead, which has similar timeout handling,
except for the bug.
Smoke tested with:
$ socat -d -d pty,raw,echo=0 pty,raw,echo=0
2017/03/14 14:08:13 socat[4994] N PTY is /dev/pts/14
2017/03/14 14:08:13 socat[4994] N PTY is /dev/pts/15
2017/03/14 14:08:13 socat[4994] N starting data transfer loop with FDs [3,3] and [5,5]
$ gdbserver /dev/pts/14 PROG
$ gdb PROG -ex "tar rem /dev/pts/15"
and then a few continues/ctrl-c's, plus killing gdbserver and socat.
[1] - See FIXME comments being removed.
gdb/ChangeLog:
2017-03-17 Pedro Alves <palves@redhat.com>
PR remote/21188
* ser-base.c (ser_base_wait_for): Add comment.
(do_ser_base_readchar): Improve comment based on the ser-unix.c's
version.
* ser-unix.c (hardwire_raw): Remove reference to
scb->current_timeout.
(wait_for, do_hardwire_readchar, hardwire_readchar): Delete.
(hardwire_ops): Install ser_base_readchar instead of
hardwire_readchar.
* serial.h (struct serial) <current_timeout, timeout_remaining>:
Remove fields.
Fix this the same way gdb/python/lib/gdb/printing.py handles it.
gdb/Changelog:
2017-03-17 Jonah Graham <jonah@kichwacoders.com>
PR gdb/19637
* python/lib/gdb/printer/bound_registers.py: Add support for
Python 3.
Recently I fixed a bug that caused a DW_OP_implicit_pointer with non-zero
offset into a DW_OP_implicit_value to be handled incorrectly on big-endian
targets. GDB ignored the offset and copied the wrong bytes:
https://sourceware.org/ml/gdb-patches/2017-01/msg00251.html
But there is still a similar issue when a DW_OP_implicit_pointer points
into a DW_OP_stack_value instead; and again, the offset is ignored. There
is an important difference, though: While implicit values are treated like
blocks of data and anchored at the lowest-addressed byte, stack values
traditionally contain integer numbers and are anchored at the *least
significant* byte. Also, stack values do not come in varying sizes, but
are cut down appropriately when used. Thus, on big-endian targets the
scenario looks like this (higher addresses shown right):
|<- - - - - Stack value - - - - - - ->|
| |
|<- original object ->|
|
| offset ->|####|
^^^^
de-referenced
implicit pointer
(Note how the original object's size influences the position of the
de-referenced implicit pointer within the stack value. This is not the
case for little-endian targets, where the original object starts at offset
zero within the stack value.)
This patch implements the logic indicated in the above diagram and adds an
appropriate test case. A new function dwarf2_fetch_die_type_sect_off is
added; it is used for retrieving the original object's type, so its size
can be determined. That type is passed to dwarf2_evaluate_loc_desc_full
via a new parameter.
gdb/ChangeLog:
* dwarf2loc.c (indirect_synthetic_pointer): Get data type of
pointed-to DIE and pass it to dwarf2_evaluate_loc_desc_full.
(dwarf2_evaluate_loc_desc_full): New parameter subobj_type; rename
byte_offset to subobj_byte_offset. Fix the handling of
DWARF_VALUE_STACK on big-endian targets when coming via an
implicit pointer.
(dwarf2_evaluate_loc_desc): Adjust call to
dwarf2_evaluate_loc_desc_full.
* dwarf2loc.h (dwarf2_fetch_die_type_sect_off): New declaration.
* dwarf2read.c (dwarf2_fetch_die_type_sect_off): New function.
gdb/testsuite/ChangeLog:
* lib/dwarf.exp: Add support for DW_OP_implicit_pointer.
* gdb.dwarf2/nonvar-access.exp: Add test for stack value location
and implicit pointer into such a location.
This patch adds the support for these instructions in arm process
record.
gdb:
2017-03-16 Yao Qi <yao.qi@linaro.org>
* arm-tdep.c (thumb_record_misc): Decode CBNZ, CBZ, REV16,
and REVSH instructions.
I look at some fails in gdb.reverse/solib-precsave.exp in -mthumb,
they are caused by some bugs on decoding these three instructions,
uxtb, ldr and mrc. This patch adds unit tests against these three
instructions, and fix these bugs by re-organizing the code to match
the table in ARM ARM.
gdb:
2017-03-16 Yao Qi <yao.qi@linaro.org>
* arm-tdep.c [GDB_SELF_TEST]: include "selftests.h".
(arm_record_test): Declare.
(_initialize_arm_tdep) [GDB_SELF_TEST]: call register_self_test.
(thumb_record_ld_st_reg_offset): Rewrite the opcode matching to
align with the manual.
(thumb_record_misc): Adjust the code order to align with the
manual.
(thumb2_record_decode_insn_handler): Fix instruction matching.
(instruction_reader_thumb): New class.
(arm_record_test): New function.
This patch adds an abstract class abstract_memory_reader a
and pass it to the code reading instructions in arm process record,
rather than using target_read_memory to read from real target. This
paves the way for adding more unit tests to arm process record.
gdb:
2017-03-16 Yao Qi <yao.qi@linaro.org>
* arm-tdep.c (abstract_memory_reader): New class.
(instruction_reader): New class.
(extract_arm_insn): Add argument 'reader'. Callers updated.
(decode_insn): Likewise.
This patch keeps the Scheme side of lazy string handling in sync
with the python size, bringing over fixes for
PRs python/17728, python/18439, python/18779.
gdb/ChangeLog:
* guile/scm-lazy-string.c (lazy_string_smob): Clarify use of LENGTH
member. Change type of TYPE member to SCM. All uses updated.
(lsscm_make_lazy_string_smob): Add assert.
(lsscm_make_lazy_string): Flag bad length values.
(lsscm_elt_type): New function.
(gdbscm_lazy_string_to_value): Rewrite to use
lsscm_safe_lazy_string_to_value.
(lsscm_safe_lazy_string_to_value): Fix handling of TYPE_CODE_PTR.
* guile/scm-value.c (gdbscm_value_to_lazy_string): Flag bad length
values. Fix TYPE_CODE_PTR. Handle TYPE_CODE_ARRAY. Handle typedefs
in incoming type.
* guile/guile-internal.h (tyscm_scm_to_type): Declare.
* guile/scm-type.c (tyscm_scm_to_type): New function.
gdb/testsuite/ChangeLog:
* gdb.guile/scm-value.c (main) Delete locals sptr, sn.
* gdb.guile/scm-lazy-string.c: New file.
* gdb.guile/scm-value.exp: Move lazy string tests to ...
* gdb.guile/scm-lazy-string.exp: ... here, new file. Add more tests
for pointer, array, typedef lazy strings.
gdb/ChangeLog:
PR python/17728, python/18439, python/18779
* python/py-lazy-string.c (lazy_string_object): Clarify use of LENGTH
member. Change type of TYPE member to PyObject *. All uses updated.
(stpy_convert_to_value): Fix handling of TYPE_CODE_PTR.
(gdbpy_create_lazy_string_object): Flag bad length values.
Handle TYPE_CODE_ARRAY with possibly different user-provided length.
Handle typedefs in incoming type.
(stpy_lazy_string_elt_type): New function.
(gdbpy_extract_lazy_string): Call it.
* python/py-value.c (valpy_lazy_string): Flag bad length values.
Fix handling of TYPE_CODE_PTR. Handle TYPE_CODE_ARRAY. Handle
typedefs in incoming type.
gdb/testsuite/ChangeLog:
PR python/17728, python/18439, python/18779
* gdb.python/py-value.c (main) Delete locals sptr, sn.
* gdb.python/py-lazy-string.c (pointer): New typedef.
(main): New locals ptr, array, typedef_ptr.
* gdb.python/py-value.exp: Move lazy string tests to ...
* gdb.python/py-lazy-string.exp: ... here. Add more tests for pointer,
array, typedef lazy strings.
The expectation in gdb.cp/m-static.exp for the ptype of
single_constructor is to get in the result of destructor with the
following prototype: ~single_constructor(int).
Yet, m-static.cc declares the destructor as ~single_constructor(). This
commit fixes the expectation.
2017-03-16 Thomas Preud'homme <thomas.preudhomme@arm.com>
gdb/testsuite/
* gdb.cp/m-static.exp: Fix expectation for prototype of
test5.single_constructor and single_constructor::single_constructor.
When inf_ptrace_xfer_partial performs a memory transfer via ptrace with
PT_READ_I, PT_WRITE_I (aka PTRACE_PEEKTEXT, PTRACE_POKETEXT), etc., then
it currently transfers at most one word. This behavior yields degraded
performance, particularly if the caller has significant preparation work
for each invocation. And indeed it has for writing, in
memory_xfer_partial in target.c, where all of the remaining data to be
transferred is copied to a temporary buffer each time, for breakpoint
shadow handling. Thus large writes have quadratic runtime and can take
hours.
Note: On GNU/Linux targets GDB usually does not use
inf_ptrace_xfer_partial for large memory transfers, but attempts a single
read/write from/to /proc/<pid>/mem instead. However, the kernel may
reject writes to /proc/<pid>/mem (such as kernels prior to 2.6.39), or
/proc may not be mounted. In both cases GDB falls back to the ptrace
mechanism.
This patch fixes the performance issue by attempting to fulfill the whole
transfer request in inf_ptrace_xfer_partial, using a loop around the
ptrace call.
gdb/ChangeLog:
PR gdb/21220
* inf-ptrace.c (inf_ptrace_xfer_partial): In "case
TARGET_OBJECT_MEMORY", extract the logic for ptrace peek/poke...
(inf_ptrace_peek_poke): ...here. New function. Now also loop
over ptrace peek/poke until end of buffer or error.
It isn't used anywhere else than the file it's defined in.
gdb/ChangeLog:
* parse.c (length_of_subexp): Make static.
* parser-defs.h (length_of_subexp): Remove.
An optional parameter TEST has been added to get_hexadecimal_valueof in commit:
https://sourceware.org/ml/gdb-patches/2016-06/msg00469.html
This patch adds a similar optional parameter to other related methods that
retrieve expression values: get_valueof, get_integer_valueof and get_sizeof.
Thus tests that evaluate same expression multiple times can provide custom
test names, ensuring that test names will be unique.
gdb/testsuite/ChangeLog:
2017-03-14 Anton Kolesov <anton.kolesov@synopsys.com>
* lib/gdb.exp (get_valueof, get_integer_valueof, get_sizeof):
Add optional 'test' parameter.
So far linux_proc_xfer_partial refused to handle write requests. This is
still based on the assumption that the Linux kernel does not support
writes to /proc/<pid>/mem. That used to be true, but has changed with
Linux 2.6.39 released in May 2011.
This patch lifts this restriction and now exploits /proc/<pid>/mem for
writing to inferior memory as well, if possible.
gdb/ChangeLog:
* linux-nat.c (linux_proc_xfer_partial): Handle write operations
as well.
Commit c8b23b3f89 ("Add constructor and destructor to
demangle_parse_info") a while ago broke the "test-cp-name-parser"
build:
$ make test-cp-name-parser
[...]
src/gdb/cp-name-parser.y: In function ‘int main(int, char**)’:
src/gdb/cp-name-parser.y:2190:9: error: cannot convert ‘std::unique_ptr<demangle_parse_info>’ to ‘demangle_parse_info*’ in assignment
result = cp_demangled_name_to_comp (str2, &errmsg);
^
src/gdb/cp-name-parser.y:2199:38: error: ‘cp_demangled_name_parse_free’ was not declared in this scope
cp_demangled_name_parse_free (result);
^
src/gdb/cp-name-parser.y:2211:14: error: cannot convert ‘std::unique_ptr<demangle_parse_info>’ to ‘demangle_parse_info*’ in assignment
result = cp_demangled_name_to_comp (argv[arg], &errmsg);
^
src/gdb/cp-name-parser.y:2219:43: error: ‘cp_demangled_name_parse_free’ was not declared in this scope
cp_demangled_name_parse_free (result);
^
Makefile:2107: recipe for target 'test-cp-name-parser.o' failed
make: *** [test-cp-name-parser.o] Error 1
This commit restores it.
gdb/ChangeLog:
2017-03-14 Pedro Alves <palves@redhat.com>
* cp-name-parser.y (cp_demangled_name_to_comp): Update comment.
(main): Use std::unique_ptr. Remove calls to
cp_demangled_name_parse_free.
gdb/ChangeLog:
* amd64-linux-nat.c (amd64_linux_fetch_inferior_registers,
amd64_linux_fetch_inferior_registers): Use regcache->ptid
instead of inferior_ptid.
We are currently assuming that regcache->ptid is equal to inferior_ptid
when we call target_fetch/store_registers. These asserts just validate
that assumption. Also, since the following patches will change target
code to use regcache->ptid instead of inferior_ptid, asserting that they
are the same should ensure that our changes don't have any unintended
consequences.
gdb/ChangeLog:
* target.c (target_fetch_registers, target_store_registers): Add
assert.
This patch introduces the regcache_get_ptid function, which can be used
to retrieve the ptid a regcache is connected to. It is used in
subsequent patches.
gdb/ChangeLog:
* regcache.h (regcache_get_ptid): New function.
* regcache.c (regcache_get_ptid): New function.
I noticed that backslash_in_multi_line_command_test in
gdb.base/commands.exp failed on our RHEL6 servers. I traced it to the
old version of DejaGnu (1.4.4). I have found that instead of receiving
the expected:
"print \\\nargc\n"
gdb received:
"print argc\n"
thus breaking the test and its purpose. Versionof DejaGnu < 1.5 mess
up sending "\\\n", it somehow gets replaced with a space. I found that
the following commit in DejaGnu fixed the issue:
http://git.savannah.gnu.org/cgit/dejagnu.git/commit/lib/remote.exp?id=3f39294f5cd6802858838d3bcc0ccce847ae17f2
Even though the commit is almost 10 years old, the following release of
DejaGnu was only in 2013, which is why we still have systems with the
old code.
If the DejaGnu version is < 1.5, we just skip the test.
gdb/testsuite/ChangeLog:
* gdb.base/commands.exp (backslash_in_multi_line_command_test):
Skip for versions of DejaGnu < 1.5.
The next patch will require checking the DejaGnu version. There is
already a test that does this,
gdb.threads/attach-many-short-lived-threads.exp. This patch introduces
a new procedure, dejagnu_version, and makes that test use it.
The version number is "right-padded" with zeroes, to make sure that we
always return a triplet (major, minor, patch).
The procedure does not consider the DejaGnu versions from git. For
example, if you used DejaGnu from its current master branch, the version
would be "1.6.1-git", meaning that 1.6.1 will be the next release. I
figured we'll cross that bridge when (and if) we get there.
gdb/testsuite/ChangeLog:
* lib/gdb.exp (dejagnu_version): New proc.
* gdb.threads/attach-many-short-lived-threads.exp (bad_dejagnu):
Use dejagnu_version.
While integrating the d_printing recursion guard change into gdb I
noticed we forgot to initialize the demangle_component d_printing
field in cplus_demangle_fill_{name,extended_operator,ctor,dtor}.
As is done in cplus_demangle_fill_{component,builtin_type,operator}.
It happened to work because in gcc all demangle_components were
allocated through d_make_empty. But gdb has its own allocation
mechanism (as might other users).
libiberty/ChangeLog:
* cp-demangle.c (cplus_demangle_fill_name): Initialize
demangle_component d_printing.
(cplus_demangle_fill_extended_operator): Likewise.
(cplus_demangle_fill_ctor): Likewise.
(cplus_demangle_fill_dtor): Likewise.
gdb/ChangeLog:
* cp-name-parser.y (make_empty): Initialize d_printing to zero.
For a long time now, c++/8218 has noted that GDB is printing argument types
for destructors:
(gdb) ptype A
type = class A {
public:
~A(int);
}
This happens because cp_type_print_method_args doesn't ignore artificial
arguments. [It ignores the first `this' pointer because it simply skips
the first argument for any non-static function.]
This patch fixes this:
(gdb) ptype A
type = class A {
public:
~A();
}
I've adjusted gdb.cp/templates.exp to account for this and added a new
passing regexp.
gdb/ChangeLog
PR c++/8218
* c-typeprint.c (cp_type_print_method_args): Skip artificial arguments.
gdb/testsuite/ChangeLog
PR c++/8128
* gdb.cp/templates.exp (test_ptype_of_templates): Remove argument
type from destructor regexps.
Add a branch which actually passes the test.
Adjust "ptype t5i" test names.
Currently diffing testrun results shows:
-PASS: gdb.base/step-over-exit.exp: break *0x7ffff77e18c6 if main == 0
+PASS: gdb.base/step-over-exit.exp: break *0x2aaaab0988c6 if main == 0
gdb/testsuite/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
* gdb.base/step-over-exit.exp: Add explicit test message.
If you do "interrupt -a" just while some thread is stepping over a
breakpoint, gdb trips on an internal error.
The test added by this patch manages to trigger this consistently by
spawning a few threads that are constantly tripping on a conditional
breakpoint whose condition always evaluates to false. With current
gdb, you get:
~~~
interrupt -a
.../src/gdb/inline-frame.c:343: internal-error: void skip_inline_frames(ptid_t): Assertion `find_inline_frame_state (ptid) == NULL' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) FAIL: gdb.threads/interrupt-while-step-over.exp: displaced-stepping=on: iter=0: interrupt -a (GDB internal error)
[...]
.../src/gdb/inline-frame.c:343: internal-error: void skip_inline_frames(ptid_t): Assertion `find_inline_frame_state (ptid) == NULL' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) FAIL: gdb.threads/interrupt-while-step-over.exp: displaced-stepping=off: iter=0: wait for stops (GDB internal error)
~~~
The assertion triggers because we're processing a stop for a thread
that had already stopped before and thus had already its inline-frame
state filled in.
Calling handle_inferior_event_1 directly within a
"thread_stop_requested" observer is something that I've wanted to get
rid of before, for being fragile. Nowadays, infrun is aware of
threads with pending events, so we can use that instead, and let the
normal fetch_inferior_event -> handle_inferior_event code path handle
the forced stop.
The change to finish_step_over is necessary because sometimes a thread
that was told to PTRACE_SINGLESTEP reports back a SIGSTOP instead of a
SIGTRAP (i.e., we tell it to single-step, and then interrupt it quick
enough that on the kernel side the thread dequeues the SIGTOP before
ever having had a chance of executing the instruction to be stepped).
SIGSTOP gets translated to a GDB_SIGNAL_0. And then finish_step_over
would miss calling clear_step_over_info, and thus miss restarting the
other threads (which in this case of threads with pending events,
means setting their "resumed" flag, so their pending events can be
consumed).
And now that we always restart threads in finish_step_over, we no
longer need to do that in handle_signal_stop.
Tested on x86_64 Fedora 23, native and gdbserver.
gdb/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
PR gdb/18360
* infrun.c (start_step_over, do_target_resume, resume)
(restart_threads): Assert we're not resuming a thread that is
meant to be stopped.
(infrun_thread_stop_requested_callback): Delete.
(infrun_thread_stop_requested): If the thread is internally
stopped, queue a pending stop event and clear the thread's
inline-frame state.
(handle_stop_requested): New function.
(handle_syscall_event, handle_inferior_event_1): Use
handle_stop_requested.
(handle_stop_requested): New function.
(handle_signal_stop): Set the thread's stop_signal here instead of
at caller.
(finish_step_over): Clear step over info unconditionally.
(handle_signal_stop): If the user had interrupted the event
thread, consider the stop a random signal.
(handle_signal_stop) <signal arrived while stepping over
breakpoint>: Don't restart threads here.
(stop_waiting): Don't clear step-over info here.
gdb/testsuite/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
PR gdb/18360
* gdb.threads/interrupt-while-step-over.c: New file.
* gdb.threads/interrupt-while-step-over.exp: New file.
Should fix the build failure with Clang mentioned at
<https://sourceware.org/bugzilla/show_bug.cgi?id=21206#c2>:
In file included from ../../binutils-gdb/gdb/dwarf2read.c:72:
../../binutils-gdb/gdb/common/gdb_unlinker.h:35:35: error: '__nonnull__' attribute is invalid for the implicit this argument
unlinker (const char *filename) ATTRIBUTE_NONNULL (1)
^ ~
../../binutils-gdb/gdb/../include/ansidecl.h:169:48: note: expanded from macro 'ATTRIBUTE_NONNULL'
# define ATTRIBUTE_NONNULL(m) __attribute__ ((__nonnull__ (m)))
gdb/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
PR 21206
* common/gdb_unlinker.h (unlinker::unlinker): Attribute nonnull
goes to argument 2, not 1.
gdb/testsuite/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
* gdb.arch/amd64-entry-value-param-dwarf5.exp: Use with_test_prefix.
* gdb.arch/amd64-entry-value-param.exp: Use with_test_prefix.
Currently I get:
(gdb) print have_pkru()
$1 = 0
(gdb) FAIL: gdb.arch/i386-pkru.exp: probe PKRU support
UNSUPPORTED: gdb.arch/i386-pkru.exp: processor does not support protection key feature.
Probing suceeded, so that should be a PASS -> UNSUPPORTED.
gdb/testsuite/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
* gdb.arch/i386-pkru.exp (probe PKRU support): Handle detecting
PKRU as not supported as a PASS.
Avoid putting unstable path names in test messages, in order to avoid
spurious testrun result diffs like:
[....]
-PASS: gdb.base/break-fun-addr.exp: /home/pedro/gdb/test-build1/gdb/testsuite/outputs/gdb.base/break-fun-addr/break-fun-addr1: break *main
+PASS: gdb.base/break-fun-addr.exp: /home/pedro/gdb/test-build2/gdb/testsuite/outputs/gdb.base/break-fun-addr/break-fun-addr1: break *main
[....]
gdb/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
* gdb.base/break-fun-addr.exp: Use $testfile1/$testfile2 for test
prefix instead of $binfile1/$binfile2.
* gdb.btrace/gcore.exp: Use "core" instead of unstable path name
in test message.
* gdb.python/py-completion.exp: Use "load python file" as test
messages instead of unstable path names.
With commit 3b12939dfc ("Replace the sync_execution global with a
new enum prompt_state tristate"), GDB started aborting if you try
splitting an input line with a continuation char (backslash) while in
a multi-line command:
(gdb) commands
Type commands for breakpoint(s) 1, one per line.
End with a line saying just "end".
>print \
(gdb) 1 # note "(gdb)" incorrectly printed here.
>end
readline: readline_callback_read_char() called with no handler!
$
That abort is actually a symptom of an old problem introduced when
gdb_readline_wrapper was rewritten to use asynchronous readline, back
in 2007. Note how the "(gdb)" prompt is printed above in the "(gdb)
1" line. Clearly it shouldn't be there, but it already was before the
commit mentioned above. Fixing that also fixes the readline abort
shown above.
The problem starts when command_line_input passes a NULL prompt to
gdb_readline_wrapper when it finds previous incomplete input due to a
backslash, trying to fetch more input without printing another ">"
secondary prompt. That itself should not be a problem, because
passing NULL to gdb_readline_wrapper has the same meaning as passing a
pointer to empty string, since gdb_readline_wrapper exposes the same
interface as 'readline(char *)'. However, gdb_readline_wrapper passes
the prompt argument directly to display_gdb_prompt, and for the
latter, a NULL prompt argument has a different meaning - it requests
printing the primary prompt.
Before commit 782a7b8ef9c096 (which rewrote gdb_readline_wrapper to
use asynchronous readline), GDB behaved like this:
(gdb) commands
[....]
>print \
1
>end
(gdb)
The above is what this commit restores GDB back to.
New test included.
gdb/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
PR cli/21218
* top.c (gdb_readline_wrapper): Avoid passing NULL to
display_gdb_prompt.
(command_line_input): Add comment.
gdb/testsuite/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
Jan Kratochvil <jan.kratochvil@redhat.com>
PR cli/21218
* gdb.base/commands.exp (backslash_in_multi_line_command_test):
New proc.
(top level): Call it.
Commit d7e747318f ("Eliminate make_cleanup_ui_file_delete / make
ui_file a class hierarchy") regressed the TUI's command window.
Newlines miss doing a "carriage return", resulting in output like:
~~~~~~~~~~~~~~~~~~
(gdb) helpList of classes of commands:
aliases -- Aliases of other commands
breakpoints -- Making program stop at certain points
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Before the commit mentioned above, the default ui_file->to_write
implementation had a hack that would defer into the ui_file->to_fputs
method. The TUI's ui_file did not implement the to_write method, so
all writes would end up going to the ncurses window via tui_file_fputs
-> tui_puts.
After the commit above, the hack is gone, but the TUI's ui_file still
does not implement the ui_file::write method. Since tui_file inherits
from stdio_file, writing to a tui_file ends up doing fwrite on the
FILE stream the TUI is "associated" with, via stdio_file::write,
instead of writing to the ncurses window.
The fix is to have tui_file override the "write" method.
New test included.
gdb/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
PR tui/21216
* tui/tui-file.c (tui_file::write): New.
* tui/tui-file.h (tui_file): Override "write".
* tui/tui-io.c (do_tui_putc, update_start_line): New functions,
factored out from ...
(tui_puts): ... here.
(tui_putc): Use them.
(tui_write): New function.
* tui/tui-io.h (tui_write): Declare.
gdb/testsuite/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
PR tui/21216
* gdb.tui/tui-nl-filtered-output.exp: New file.
gdb/testsuite/ChangeLog:
2017-03-08 Pedro Alves <palves@redhat.com>
* gdb.base/completion.exp: Move TUI completion tests to ...
* gdb.tui/completion.exp: ... this new file.
Let's start putting TUI tests in their own dir.
gdb/testsuite/
2017-03-08 Pedro Alves <palves@redhat.com>
* gdb.base/tui-disasm-long-lines.c,
gdb.base/tui-disasm-long-lines.exp, gdb.base/tui-layout.c,
gdb.base/tui-layout.exp: Move to ...
* gdb.tui/: ... this new directory.
We will need access to the environment functions when we share
fork_inferior between GDB and gdbserver, therefore we simply make the
API on gdb/environ.[ch] available on common/. No extra adjustments
are needed to make it compile on gdbserver.
gdb/ChangeLog:
2017-03-07 Sergio Durigan Junior <sergiodj@redhat.com>
* Makefile.in (SFILES): Replace "environ.c" with
"common/environ.c".
(HFILES_NO_SRCDIR): Likewise, for "environ.h".
* environ.c: Include "common-defs.h" instead of "defs.h. Moved
to...
* common/environ.c: ... here.
* environ.h: Moved to...
* common/environ.h: ... here.
gdb/gdbserver/ChangeLog:
2017-03-07 Sergio Durigan Junior <sergiodj@redhat.com>
* Makefile.in (SFILES): Add "common/environ.c".
(OBJS): Add "common/environ.h".
gdb/
* gdbarch.sh (pstring_ptr): New static function.
(gdbarch_disassembler_options): Use it.
(gdbarch_verify_disassembler_options): Print valid_disassembler_options,
not valid_disassembler_option->name.
* gdbarch.c: Regenerate.
Commit d7e747318f ("Eliminate make_cleanup_ui_file_delete / make
ui_file a class hierarchy") introduced a problem when using "layout
regs", that leads gdb to crash when issuing:
./gdb ./a.out -ex 'layout regs' -ex start
From the backtrace, it's caused by this 'delete' on tui_restore_gdbout():
(gdb) bt
#0 0x00007ffff6b962b2 in free () from /lib64/libc.so.6
#1 0x000000000059fa47 in tui_restore_gdbout (ui=0x22997b0) at ../../gdb/tui/tui-regs.c:714
#2 0x0000000000619996 in do_my_cleanups (pmy_chain=pmy_chain@entry=0x1e08320 <cleanup_chain>, old_chain=old_chain@entry=0x235b4b0) at ../../gdb/common/cleanups.c:154
#3 0x0000000000619b1d in do_cleanups (old_chain=old_chain@entry=0x235b4b0) at ../../gdb/common/cleanups.c:176
#4 0x000000000059fb0d in tui_register_format (frame=frame@entry=0x22564e0, regnum=regnum@entry=0) at ../../gdb/tui/tui-regs.c:747
#5 0x000000000059ffeb in tui_get_register (data=0x2434d18, changedp=0x0, regnum=0, frame=0x22564e0) at ../../gdb/tui/tui-regs.c:768
#6 tui_show_register_group (refresh_values_only=<optimized out>, frame=0x22564e0, group=0x1e09250 <general_group>) at ../../gdb/tui/tui-regs.c:287
#7 tui_show_registers (group=0x1e09250 <general_group>) at ../../gdb/tui/tui-regs.c:156
#8 0x00000000005a07cf in tui_check_register_values (frame=frame@entry=0x22564e0) at ../../gdb/tui/tui-regs.c:496
#9 0x00000000005a3e65 in tui_check_data_values (frame=frame@entry=0x22564e0) at ../../gdb/tui/tui-windata.c:232
#10 0x000000000059cf65 in tui_refresh_frame_and_register_information (registers_too_p=1) at ../../gdb/tui/tui-hooks.c:156
#11 0x00000000006d5c05 in generic_observer_notify (args=0x7fffffffdbe0, subject=<optimized out>) at ../../gdb/observer.c:167
#12 observer_notify_normal_stop (bs=<optimized out>, print_frame=print_frame@entry=1) at ./observer.inc:61
#13 0x00000000006a6409 in normal_stop () at ../../gdb/infrun.c:8364
#14 0x00000000006af8f5 in fetch_inferior_event (client_data=<optimized out>) at ../../gdb/infrun.c:3990
#15 0x000000000066f0fd in gdb_wait_for_event (block=block@entry=0) at ../../gdb/event-loop.c:859
#16 0x000000000066f237 in gdb_do_one_event () at ../../gdb/event-loop.c:322
#17 0x000000000066f386 in gdb_do_one_event () at ../../gdb/event-loop.c:353
#18 0x00000000007411bc in wait_sync_command_done () at ../../gdb/top.c:570
#19 0x0000000000741426 in maybe_wait_sync_command_done (was_sync=0) at ../../gdb/top.c:587
#20 execute_command (p=<optimized out>, p@entry=0x7fffffffe43a "start", from_tty=from_tty@entry=1) at ../../gdb/top.c:676
#21 0x00000000006c2048 in catch_command_errors (command=0x741200 <execute_command(char*, int)>, arg=0x7fffffffe43a "start", from_tty=1) at ../../gdb/main.c:376
#22 0x00000000006c2b60 in captured_main_1 (context=0x7fffffffde70) at ../../gdb/main.c:1119
#23 captured_main (data=0x7fffffffde70) at ../../gdb/main.c:1140
#24 gdb_main (args=args@entry=0x7fffffffdf90) at ../../gdb/main.c:1158
#25 0x0000000000408cf5 in main (argc=<optimized out>, argv=<optimized out>) at ../../gdb/gdb.c:32
(gdb) f 1
#1 0x000000000059fa47 in tui_restore_gdbout (ui=0x22997b0) at ../../gdb/tui/tui-regs.c:714
714 delete gdb_stdout;
The problem is simply that the commit mentioned above made the ui_file
that gdb_stdout is temporarily set to be a stack-allocated
string_file, while before it used to be a heap-allocated ui_file. The
fix is simply to remove the now-incorrect delete.
New test included, which exercises enabling all TUI layouts, with and
without execution. (This particular crash only triggers with
execution.)
gdb/ChangeLog:
2017-03-07 Pedro Alves <palves@redhat.com>
* tui/tui-regs.c (tui_restore_gdbout): Don't delete gdb_stdout.
gdb/testsuite/ChangeLog:
2017-03-07 Pedro Alves <palves@redhat.com>
* gdb.base/tui-layout.c: New file.
* gdb.base/tui-layout.exp: New file.
To better reflect what the testcase is about, and to make room for a
different testcase.
gdb/testsuite/ChangeLog:
2017-03-07 Pedro Alves <palves@redhat.com>
* gdb.base/tui-layout.c: Rename to ...
* gdb.base/tui-disasm-long-lines.c: ... this.
* gdb.base/tui-layout.exp: Rename to ...
* gdb.base/tui-disasm-long-lines.exp: ... this.
This patch initializes the BND registers before executing the inferior
call. BND registers can be in arbitrary values at the moment of the
inferior call. In case the function being called uses as part of the
parameters BND register, e.g. when passing a pointer as parameter, the
current value of the register will be used. This can cause boundary
violations that are not due to a real bug or even desired by the user.
In this sense the best to be done is set the BND registers to allow
access to the whole memory, i.e. initialized state, before pushing the
inferior call.
2017-03-07 Walfred Tedeschi <walfred.tedeschi@intel.com>
gdb/ChangeLog:
* i387-tdep.h (i387_reset_bnd_regs): Add function definition.
* i387-tdep.c (i387_reset_bnd_regs): Add function implementation.
* i386-tdep.c (i386_push_dummy_call): Call i387_reset_bnd_regs.
* amd64-tdep (amd64_push_dummy_call): Call i387_reset_bnd_regs.
gdb/testsuite/ChangeLog:
* i386-mpx-call.c: New file.
* i386-mpx-call.exp: New file.
gdb/doc/ChangeLog:
* Memory Protection Extensions: Add information about inferior
calls.
As reported in PR 21165,
(gdb) info locals^M
gv = /home/yao/SourceCode/gnu/gdb/git/gdb/value.c:372: internal-error: int value_bits_any_optimized_out(const value*, int, int): Assertion `!value->lazy' failed.^M
A problem internal to GDB has been detected,^M
further debugging may prove unreliable.^M
Quit this debugging session? (y or n) FAIL: gdb.ada/info_locals_renaming.exp: info locals (GDB internal error)
Resyncing due to internal error.
This internal error is caused by e8b24d9 (Remove parameter valaddr from
la_val_print). Commit e8b24d9 removes some calls to
value_contents_for_printing, but value_fetch_lazy is not called, so the
internal error above is triggered. This patch adds value_fetch_lazy
call before val_print.
gdb:
2017-03-03 Yao Qi <yao.qi@linaro.org>
PR gdb/21165
* ada-valprint.c (ada_val_print_ref): Call value_fetch_lazy if
value is lazy.
* valprint.c (common_val_print): Likewise.
This commit adds support to GDB so that it can modify the disassembler-options
value that is passed to the disassembler, similar to objdump's -M option.
Currently, the only supported targets are ARM, PowerPC and S/390, but
adding support for a new target(s) is not difficult.
include/
* dis-asm.h (disasm_options_t): New typedef.
(parse_arm_disassembler_option): Remove prototype.
(set_arm_regname_option): Likewise.
(get_arm_regnames): Likewise.
(get_arm_regname_num_options): Likewise.
(disassemble_init_s390): New prototype.
(disassembler_options_powerpc): Likewise.
(disassembler_options_arm): Likewise.
(disassembler_options_s390): Likewise.
(remove_whitespace_and_extra_commas): Likewise.
(disassembler_options_cmp): Likewise.
(next_disassembler_option): New inline function.
(FOR_EACH_DISASSEMBLER_OPTION): New macro.
opcodes/
* disassemble.c Include "safe-ctype.h".
(disassemble_init_for_target): Handle s390 init.
(remove_whitespace_and_extra_commas): New function.
(disassembler_options_cmp): Likewise.
* arm-dis.c: Include "libiberty.h".
(NUM_ELEM): Delete.
(regnames): Use long disassembler style names.
Add force-thumb and no-force-thumb options.
(NUM_ARM_REGNAMES): Rename from this...
(NUM_ARM_OPTIONS): ...to this. Use ARRAY_SIZE.
(get_arm_regname_num_options): Delete.
(set_arm_regname_option): Likewise.
(get_arm_regnames): Likewise.
(parse_disassembler_options): Likewise.
(parse_arm_disassembler_option): Rename from this...
(parse_arm_disassembler_options): ...to this. Make static.
Use new FOR_EACH_DISASSEMBLER_OPTION macro to scan over options.
(print_insn): Use parse_arm_disassembler_options.
(disassembler_options_arm): New function.
(print_arm_disassembler_options): Handle updated regnames.
* ppc-dis.c: Include "libiberty.h".
(ppc_opts): Add "32" and "64" entries.
(ppc_parse_cpu): Use ARRAY_SIZE and disassembler_options_cmp.
(powerpc_init_dialect): Add break to switch statement.
Use new FOR_EACH_DISASSEMBLER_OPTION macro.
(disassembler_options_powerpc): New function.
(print_ppc_disassembler_options): Use ARRAY_SIZE.
Remove printing of "32" and "64".
* s390-dis.c: Include "libiberty.h".
(init_flag): Remove unneeded variable.
(struct s390_options_t): New structure type.
(options): New structure.
(init_disasm): Rename from this...
(disassemble_init_s390): ...to this. Add initializations for
current_arch_mask and option_use_insn_len_bits_p. Remove init_flag.
(print_insn_s390): Delete call to init_disasm.
(disassembler_options_s390): New function.
(print_s390_disassembler_options): Print using information from
struct 'options'.
* po/opcodes.pot: Regenerate.
binutils/
* objdump.c (main): Use remove_whitespace_and_extra_commas.
gdb/
* NEWS: Mention new set/show disassembler-options commands.
* doc/gdb.texinfo: Document new set/show disassembler-options commands.
* disasm.c: Include "arch-utils.h", "gdbcmd.h" and "safe-ctype.h".
(prospective_options): New static variable.
(gdb_disassembler::gdb_disassembler): Initialize
m_di.disassembler_options.
(gdb_buffered_insn_length_init_dis): Initilize di->disassembler_options.
(get_disassembler_options): New function.
(set_disassembler_options): Likewise.
(set_disassembler_options_sfunc): Likewise.
(show_disassembler_options_sfunc): Likewise.
(disassembler_options_completer): Likewise.
(_initialize_disasm): Likewise.
* disasm.h (get_disassembler_options): New prototype.
(set_disassembler_options): Likewise.
* gdbarch.sh (gdbarch_disassembler_options): New variable.
(gdbarch_verify_disassembler_options): Likewise.
* gdbarch.c: Regenerate.
* gdbarch.h: Likewise.
* arm-tdep.c (num_disassembly_options): Delete.
(set_disassembly_style): Likewise.
(arm_disassembler_options): New static variable.
(set_disassembly_style_sfunc): Convert short style name into long
option name. Call set_disassembler_options.
(show_disassembly_style_sfunc): New function.
(arm_gdbarch_init): Call set_gdbarch_disassembler_options and
set_gdbarch_verify_disassembler_options.
(_initialize_arm_tdep): Delete regnames variable and update callers.
(arm_disassembler_options): Initialize.
(disasm_options): New variable.
(num_disassembly_options): Rename from this...
(num_disassembly_styles): ...to this. Compute by scanning through
disasm_options.
(valid_disassembly_styles): Initialize using disasm_options.
Remove calls to parse_arm_disassembler_option, get_arm_regnames and
set_arm_regname_option.
Pass show_disassembly_style_sfunc to the "disassembler" setshow command.
* rs6000-tdep.c (powerpc_disassembler_options): New static variable.
(rs6000_gdbarch_init): Call set_gdbarch_disassembler_options and
set_gdbarch_verify_disassembler_options.
* s390-tdep.c (s390_disassembler_options): New static variable.
(s390_gdbarch_init):all set_gdbarch_disassembler_options and
set_gdbarch_verify_disassembler_options.
gdb/testsuite/
* gdb.arch/powerpc-power.exp: Delete test.
* gdb.arch/powerpc-power.s: Likewise.
* gdb.disasm/disassembler-options.exp: New test.
* gdb.arch/powerpc-altivec.exp: Likewise.
* gdb.arch/powerpc-altivec.s: Likewise.
* gdb.arch/powerpc-altivec2.exp: Likewise.
* gdb.arch/powerpc-altivec2.s: Likewise.
* gdb.arch/powerpc-altivec3.exp: Likewise.
* gdb.arch/powerpc-altivec3.s: Likewise.
* gdb.arch/powerpc-power7.exp: Likewise.
* gdb.arch/powerpc-power7.s: Likewise.
* gdb.arch/powerpc-power8.exp: Likewise.
* gdb.arch/powerpc-power8.s: Likewise.
* gdb.arch/powerpc-power9.exp: Likewise.
* gdb.arch/powerpc-power9.s: Likewise.
* gdb.arch/powerpc-vsx.exp: Likewise.
* gdb.arch/powerpc-vsx.s: Likewise.
* gdb.arch/powerpc-vsx2.exp: Likewise.
* gdb.arch/powerpc-vsx2.s: Likewise.
* gdb.arch/powerpc-vsx3.exp: Likewise.
* gdb.arch/powerpc-vsx3.s: Likewise.
* gdb.arch/arm-disassembler-options.exp: Likewise.
* gdb.arch/powerpc-disassembler-options.exp: Likewise.
* gdb.arch/s390-disassembler-options.exp: Likewise.
The previous patch introduced this error with recent-ish GCCs:
../../binutils-gdb/gdb/remote.c: In function ‘int remote_add_target_side_condition(gdbarch*, bp_target_info*, char*, char*)’:
../../binutils-gdb/gdb/remote.c:9668:8: error: types may not be defined in a for-range-declaration [-Werror]
for (struct agent_expr *aexpr : bp_tgt->conditions)
^~~~~~
Removing the struct keyword fixes the error.
gdb/ChangeLog:
* remote.c (remote_add_target_side_condition): Remove "struct"
keyword from range-based for loop.
Using a range based for loop makes this code a bit more clean and
readable.
The comment above is clearly erroneous, so I've updated it.
gdb/ChangeLog:
* remote.c (remote_add_target_side_condition): Use range-based
for loop. Update comment.
ASAN reports the following error,
(gdb) PASS: gdb.fortran/vla-ptr-info.exp: continue to breakpoint: pvla-associated
print &pvla^M
=================================================================^M
^[[1m^[[31m==14331==ERROR: AddressSanitizer: global-buffer-overflow on address 0x000000ea569f at pc 0x0000008eb546 bp 0x7ffde0c1dc70 sp 0x7ffde0c1dc60^M
^[[1m^[[0m^[[1m^[[34mREAD of size 1 at 0x000000ea569f thread T0^[[1m^[[0m^M
#0 0x8eb545 in f_print_type(type*, char const*, ui_file*, int, int, type_print_options const*) ../../binutils-gdb/gdb/f-typeprint.c:89^M
#1 0xb611e2 in type_print(type*, char const*, ui_file*, int) ../../binutils-gdb/gdb/typeprint.c:365^M
#2 0x7b3471 in c_value_print(value*, ui_file*, value_print_options const*) ../../binutils-gdb/gdb/c-valprint.c:650^M
#3 0xb99517 in value_print(value*, ui_file*, value_print_options const*) ../../binutils-gdb/gdb/valprint.c:1233^M
#4 0xa42be8 in print_formatted ../../binutils-gdb/gdb/printcmd.c:321^M
#5 0xa46ac9 in print_value(value*, format_data const*) ../../binutils-gdb/gdb/printcmd.c:1233^M
#6 0xa46d82 in print_command_1 ../../binutils-gdb/gdb/printcmd.c:1261^M
#7 0xa46e3e in print_command ../../binutils-gdb/gdb/printcmd.c:1267
on this line of code
demangled_args = varstring[strlen (varstring) - 1] == ')';
because varstring is an empty string and strlen () is 0, so "strlen () - 1"
is definitely out of the bound of "varstring",
(gdb) bt 10
at /home/yao/SourceCode/gnu/gdb/git/gdb/f-typeprint.c:56
at /home/yao/SourceCode/gnu/gdb/git/gdb/typeprint.c:365
at /home/yao/SourceCode/gnu/gdb/git/gdb/c-valprint.c:650
at /home/yao/SourceCode/gnu/gdb/git/gdb/valprint.c:1236
This patch adds a pre-check that varstring is empty or not.
gdb:
2017-02-27 Yao Qi <yao.qi@linaro.org>
* f-typeprint.c (f_print_type): Check "varstring" is empty first.
gcc-6.3.1-1.fc25.x86_64
dwarf2read.c: In function ‘void create_debug_type_hash_table(dwo_file*, dwarf2_section_info*, htab*&, rcuh_kind)’:
dwarf2read.c:4776:32: error: ‘header.comp_unit_head::type_offset_in_tu.cu_offset::cu_off’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
dwarf2read.c:4816:21: error: ‘header.comp_unit_head::signature’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
cc1plus: all warnings being treated as errors
gdb/ChangeLog
2017-02-26 Jan Kratochvil <jan.kratochvil@redhat.com>
* dwarf2read.c (create_debug_type_hash_table): Initialize
header.signature and header.type_offset_in_tu.
add_symtab_completions does the exact same as the code that it is
replacing.
gdb/ChangeLog:
2017-02-24 Pedro Alves <palves@redhat.com>
* symtab.c (make_file_symbol_completion_list_1): Use
add_symtab_completions.
As reported in PR21166, there are Intel processors out there that support
rdrand but not rdseed. The fix is to verify both features separately and only
run rdrand/rdseed tests if supported.
gdb/testsuite/ChangeLog:
2017-02-23 Luis Machado <lgustavo@codesourcery.com>
* gdb.reverse/insn-reverse.x86.c (check_rdrand_support): Renamed to ...
(check_supported_features): ... this. Changed return type to void.
(supports_rdseed): New static global.
(rdseed): Check supports_rdseed.
(initialize): Call check_supported_features.
Tom Tromey
Pedro Alves
Simon Marchi
Alan Hayward
Sergio Durigan Junior
Philipp Rudo
Yao Qi
Simon Marchi
Thomas Preud'homme
John Baldwin
Iain Buclaw
Jan Kratochvil
Anton Kolesov
Max Filippov
Pádraig Brady
Andreas Arnez
Jonah Graham
Ivo Raisr
Tim Wiederhake
Artemiy Volkov
Marc-Andre Laperle
Doug Evans
Jiong Wang
mark
Keith Seitz
Peter Bergner
Walfred Tedeschi
Luis Machado