All programs in binutils+gdb git repo now support gABI compression
with the SHF_COMPRESSED bit. This patch makes the zlib-gabi option
as compression default for gas, gold, ld and objcopy, instead of the
zlib-gnu option whose outputs are incompatible with gABI.
binutils/
* objcopy.c (copy_file): Set BFD_COMPRESS_GABI if not
zlib-gnu.
* doc/binutils.texi: Change --compress-debug-sections and
--compress-debug-sections=zlib to zlib-gabi.
binutils/testsuite/
* binutils-all/compress.exp: Update.
gas/
* as.c (parse_args): Make --compress-debug-sections and
--compress-debug-sections=zlib the same as
--compress-debug-sections=zlib-gabi.
* doc/as.texinfo: Change --compress-debug-sections and
--compress-debug-sections=zlib to zlib-gabi.
gold/
* compressed_output.cc (Output_compressed_section::set_final_data_size):
Make --compress-debug-sections=zlib the same as
--compress-debug-sections=zlib-gabi.
* testsuite/Makefile.am (flagstest_compress_debug_sections.check):
Expect ".debug_.*" with the SHF_COMPRESSED bit, instead of
".zdebug_".
* testsuite/Makefile.in: Regenerated.
ld/
* emultempl/elf32.em (gld${EMULATION_NAME}_handle_option): Make
--compress-debug-sections=zlib the same as
--compress-debug-sections=zlib-gabi.
* ld.texinfo: Change --compress-debug-sections=zlib to zlib-gabi.
ld/testsuite/
* ld-elf/zlibbegin.rS: Updated to .debug_.* with the
SHF_COMPRESSED bit.
* ld-elf/zlibnormal.rS: Likewise.
Sync with GCC
2015-05-13 Eric Botcazou <ebotcazou@adacore.com>
* sjlj.m4: New file.
2015-05-04 Trevor Saunders <tbsaunde+gcc@tbsaunde.org>
* bitfields.m4: Change int to long long, and use bitfields of
width 1 instead of 0.
2015-05-01 Trevor Saunders <tbsaunde+gcc@tbsaunde.org>
* bitfields.m4: New file.
2015-04-14 H.J. Lu <hongjiu.lu@intel.com>
* bootstrap-mpx.mk: New file.
2015-03-25 Uros Bizjak <ubizjak@gmail.com>
PR bootstrap/65537
* bootstrap-lto-noplugin.mk: New build configuration.
2014-11-13 Kirill Yukhin <kirill.yukhin@intel.com>
* target-posix: New file.
2014-10-27 Tom Tromey <tromey@redhat.com>
* gcc-plugin.m4: New file.
* nat/linux-namespaces.c (setns): Rename from this ...
(do_setns): ... to this. Support calling setns if it exists.
(mnsh_handle_setns): Call do_setns.
Sync with GCC
2015-06-02 Jason Merrill <jason@redhat.com>
PR bootstrap/66319
* configure.ac: Use -std=gnu++98.
2015-05-28 Mike Frysinger <vapier@gentoo.org>
* configure.ac (--vtable-verify): Use AS_HELP_STRING for help.
* configure: Regenerate.
2015-05-11 Paulo Matos <paulo@matos-sorge.com>
* configure.ac: Fix typo.
* configure: Regenerate.
2015-05-03 Matthias Klose <doko@ubuntu.com>
* configure.ac: Match $host configured with triplets.
* configure: Regenerate.
2015-04-17 Jakub Jelinek <jakub@redhat.com>
PR bootstrap/62077
* configure.ac (--enable-stage1-checking): Default to
release,misc,gimple,rtlflag,tree,types if --disable-checking
or --enable-checking is not specified and DEV-PHASE is not
experimental.
* configure: Regenerated.
2015-03-27 Uros Bizjak <ubizjak@gmail.com>
Install back PR target/47230 fix (Revert the revert).
config/
Sync with GCC
* mh-alpha-linux: Restored.
This exercises the case of the inferior disappearing while GDB is
debugging it, such as something doing "kill -9 PID" while the program
is stopped under GDB or GDBserver. This triggered a set of internal
errors, fixed by previous patches.
gdb/testsuite/ChangeLog:
2015-07-14 Pedro Alves <palves@redhat.com>
* gdb.base/killed-outside.exp: New file.
* gdb.base/killed-outside.c: New file.
If the process disappears (e.g., killed with "kill -9" from the shell)
while it was stopped under GDBserver's control, and the GDBserver
tries to kill it, GDBserver asserts:
(gdb) shell kill -9 23084
(gdb) kill
...
Killing process(es): 23084
/home/pedro/gdb/mygit/src/gdb/gdbserver/linux-low.c:972: A problem internal to GDBserver has been detected.
kill_wait_lwp: Assertion `res > 0' failed.
...
gdb/gdbserver/ChangeLog:
2015-07-14 Pedro Alves <palves@redhat.com>
* linux-low.c (kill_wait_lwp): Don't assert if waitpid fails.
Instead, ignore ECHILD, and throw an error for other errnos.
When I examine the buildbot fails, I see this fail on
native-extended-gdbserver,
Attaching to process 13529^M
"target:/scratch/yao/gdb/build-git/x86_64/gdb/testsuite/gdb.base/attach-pie-noexec (deleted)": could not open as an executable file: No such file or directory^M
(gdb) FAIL: gdb.base/attach-pie-noexec.exp: attach
if I run tests with board file unix, it doesn't exist,
Attaching to process 13869^M
/scratch/yao/gdb/build-git/x86_64/gdb/testsuite/gdb.base/attach-pie-noexec (deleted): No such file or directory.^M
(gdb) PASS: gdb.base/attach-pie-noexec.exp: attach
the test expects to see the period at the end of the error message,
gdb_test "attach $testpid" "Attaching to process $testpid\r\n.*: No such file or directory\\." "attach"
however the period is missing when running with native-extended-gdbserver.
in exec.c:exec_file_attach, GDB has two places may throw errors [1] and [2],
if (load_via_target)
{
...
}
else
{
...
if (scratch_chan < 0)
perror_with_name (filename); <--- [1]
}
...
if (!exec_bfd)
{
error (_("\"%s\": could not open as an executable file: %s"), <-- [2]
scratch_pathname, bfd_errmsg (bfd_get_error ()));
}
perror_with_name [1] append a period at the end of error message,
but error [2] doesn't. This fix is to add a period at the end of the
error message. Note that this fail is shown up on 7.9 release as well.
gdb:
2015-07-13 Yao Qi <yao.qi@linaro.org>
* exec.c (exec_file_attach): Add period at the end of error
message.
Share the window name completion code from the focus command with the
winheight command, providing window name completion for the winheight
command.
gdb/ChangeLog:
* tui/tui-win.c (window_name_completer): New function.
(focus_completer): Call window_name_completer. All old content
moved into window_name_completer.
(winheight_completer): New function.
(_initialize_tui_win): Rename variable. Add completer to
winheight command. Update doc string on winheight.
This patch adds SHF_COMPRESSED section decompression to gold.
PR gold/18321
* compressed_output.h (decompress_input_section): Add arguments
for ELF class, big endian and sh_flags.
* compressed_output.cc (decompress_input_section): Likewise.
Support the SHF_COMPRESSED section.
* dynobj.h (Dynobj): Add elfsize and is_big_endian member
functions.
* plugin.h (Pluginobj): Likewise.
* layout.cc (Layout::get_output_section_flags): Also clear the
SHF_COMPRESSED bit.
* object.h (Compressed_section_info): Add flag to store sh_flags.
(Object): Add pure virtual elfsize and is_big_endian member
functions.
* object.cc (need_decompressed_section): Don't skip the ".zdebug"
prefix here.
(build_compressed_section_map): Check SHF_COMPRESSED for
uncompressed size. Store sh_flags in Compressed_section_info.
Pass size, big_endian and sh_flags to decompress_input_section.
Skip the ".debug"/".zdebug" prefix when passing section name to
need_decompressed_section.
(Sized_relobj_file<size, big_endian>::do_find_special_section):
Don't check ".zdebug_*" sections.
(Object::decompressed_section_contents): Pass ELF class, big
endian and sh_flags to decompress_input_section.
* reloc.cc (Sized_relobj_file<size, big_endian>::write_sections):
Likewise.
* testsuite/Makefile.am (check_DATA): Add
debug_msg_cdebug_gabi.err and gdb_index_test_2_gabi.stdout.
(MOSTLYCLEANFILES): Add debug_msg_cdebug_gabi.err and
gdb_index_test_2_gabi.stdout.
(debug_msg_cdebug_gabi.o): New.
(odr_violation1_cdebug_gabi.o): Likewise.
(odr_violation2_cdebug_gabi.o): Likewise.
(debug_msg_cdebug_gabi.err): Likewise.
(check_SCRIPTS): Add gdb_index_test_2_gabi.sh.
(gdb_index_test_cdebug_gabi.o): Likewise.
(gdb_index_test_2_gabi): Likewise.
(gdb_index_test_2_gabi.stdout): Likewise.
* testsuite/gdb_index_test_2_gabi.sh: New file.
* testsuite/Makefile.in: Regenerated.
The output SHF_COMPRESSED section size is different from input if
ELF classes of input and output aren't the same. We must adjust
the section sizes as well as the compression headers in
SHF_COMPRESSED sections when converting objects between different
ELF classes.
bfd/
PR binutils/18656
* bfd.c (bfd_convert_section_size): New function.
(bfd_convert_section_contents): Likewise.
* bfd-in2.h: Regenerated.
binutils/
2015-07-10 H.J. Lu <hongjiu.lu@intel.com>
PR binutils/18656
* objcopy.c (setup_section): Call bfd_convert_section_size
to get the output section size.
(copy_section): Get the section size from the output section
and call bfd_get_full_section_contents to convert section
contents for output.
binutils/testsuite/
PR binutils/18656
* binutils-all/compress.exp (convert_test): New proc.
Run conversion tests between x86-64 and x32.
This change causes the prologue scanner and the frame type scanner in
rx-tdep.c to use target_read_code() instead of target_read_memory().
This change allows these instruction scanners to operate much more
quickly due to the fact that target_read_code() can potentially read
from a cache maintained by GDB.
gdb/ChangeLog:
* rx-tdep.c (rx_get_opcode_byte): Use target_read_code instead
of target_read_memory.
Building with C++ catches a buglet here:
../../../src/gdb/gdbserver/event-loop.c:205:19: warning: invalid conversion from ‘gdb_client_data {aka void*}’ to ‘void**’ [-fpermissive]
event_ptr->data = data;
^
This works in practice because gdb_client_data is a pointer already
(hence in C we get an implicit conversion), and nothing deferences the
pointer. It's passed from client at event registration/creation time,
only to pass straight back to client callback.
Well, that and nothing in gdbserver uses the event data anyway.
gdb/gdbserver/ChangeLog:
2015-07-10 Pedro Alves <palves@redhat.com>
* event-loop.c (struct callback_event) <data>: Change type to
gdb_client_data instance instead of gdb_client_data pointer.
(append_callback_event): Adjust.
I have somehow missed gdb.asm/asm-source.exp PASS->FAIL even on x86_64.
It has no longer valid assumption that "break" breaks after the prologue even
in assembler. So I have changed this assumption of the testfile.
gdb/testsuite/ChangeLog
2015-07-10 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdb.asm/asm-source.exp (f at main): Stop at gdbasm_enter.
(n at main): New.
* gdb.asm/asmsrc1.s: Add comment "mark: main enter".
This patch cleans up the comments for each linux_target_ops methods. We
should mention which method each function implements but there is no
need to duplicate information already mentionned in the base target_ops
or linux_target_ops definitions.
gdb/gdbserver/ChangeLog:
* linux-aarch64-low.c: Add comments for each linux_target_ops
method. Remove comments already covered in target_ops and
linux_target_ops definitions.
(the_low_target): Add comments for each unimplemented method.
In parse_scrolling_args it is possible for a string copy to leak if an
error occurs. Switching to using a cleanup fixes this leak.
In tui_set_win_height the string can't be leaked, but switching to using
a cleanup guards against the possibility that a leak could be introduced
in the future (by adding an error somewhere in the call stack).
gdb/ChangeLog:
* tui/tui-win.c (tui_set_win_height): Use a cleanup to free the
string copy.
(parse_scrolling_args): Likewise.
Don't duplicate the window names inside the completion function.
Instead make use of the existing defines, and the tui_win_name function
to obtain the window names.
gdb/ChangeLog:
* tui/tui-win.c (focus_completer): Don't duplicate the tui window
names in this function.
This commit converts the window names for the TUI windows to lower case.
The windows names are already lower case in the documentation, and are
shown as lower case in some of the command completion results.
Given that all the commands that take a window name currently accept
upper or lower case, this commit just changes the window names to lower
case in the remaining places they are displayed by gdb.
gdb/ChangeLog:
* tui/tui-data.h (SRC_NAME): Convert to lower case.
(CMD_NAME): Likewise.
(DATA_NAME): Likewise.
(DISASSEM_NAME): Likewise.
* tui/tui-win.c (tui_set_focus): Window names are now lower case.
(tui_set_win_height): Likewise.
(parse_scrolling_args): Likewise.
Inside output sections, ALIGN(.,x) uses a section-relative value for
dot. The unary ALIGN always used the absolute value of dot.
* ldexp.c (align_dot_val): New function.
(fold_unary <ALIGN_K, NEXT>): Use it.
Since the new KFAILs/KPASSs for the infcall tests on x86 and x86_64
targets generated unnecessary noise, this change skips them with
UNSUPPORTED instead.
gdb/testsuite/ChangeLog:
* gdb.base/gnu_vector.exp: On x86 and x86_64 targets, skip the
infcall tests instead of setting up for KFAIL.
When navigating in the recorded execution trace via "record goto", we do not
set stop_pc. This may trigger an internal error in infrun.c when stepping
from that location. Set it.
(gdb) rec full
(gdb) c
Continuing.
Breakpoint 1, foo (void) at foo.c:42
42 x = y
(gdb) rn
foo (void)
at foo.c:41
41 y = x
(gdb) rec go end
Go forward to insn number 98724
at foo.c:42
42 x = y
(gdb) n
infrun.c:2382: internal-error: resume: Assertion `sig != GDB_SIGNAL_0' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n)
This happens because there's a breakpoint at PC when the "next"
is issued, so that breapoint should be immediately stepped over.
That should have been detected/done by proceed, here:
if (addr == (CORE_ADDR) -1)
{
if (pc == stop_pc
&& breakpoint_here_p (aspace, pc) == ordinary_breakpoint_here
&& execution_direction != EXEC_REVERSE)
/* There is a breakpoint at the address we will resume at,
step one instruction before inserting breakpoints so that
we do not stop right away (and report a second hit at this
breakpoint).
Note, we don't do this in reverse, because we won't
actually be executing the breakpoint insn anyway.
We'll be (un-)executing the previous instruction. */
tp->stepping_over_breakpoint = 1;
But since stop_pc was stale, the pc == stop_pc check failed, and left the
breakpont at PC inserted.
gdb/
* record-btrace.c (record_btrace_goto_begin, record_btrace_goto_end)
record_btrace_goto): Move call to print_stack_frame ...
(record_btrace_set_replay): ... here. Set stop_pc.
* record-full.c (record_full_goto_entry): Set stop_pc.
testsuite/
* gdb.btrace/record_goto-step.exp: New.
The psABI (AAELF) says this about mapping symbols:
4.6.5.1 Section-relative mapping symbols
Mapping symbols defined in a section define a sequence of
half-open address intervals that cover the address range of the
section. Each interval starts at the address defined by the
mapping symbol, and continues up to, but not including, the
address defined by the next (in address order) mapping symbol or
the end of the section. A section must have a mapping symbol
defined at the beginning of the section; however, if the section
contains only data then the mapping symbol may be omitted.
That makes it pretty clear that, in the absence of a mapping symbol,
the entire section can be treated as data.
gold/ChangeLog:
* aarch64.cc (AArch64_relobj::scan_errata): Drop missing symbol warning.
* arm.cc (Arm_relobj::scan_section_for_cortex_a8_erratum): Drop missing
symbol warning.
This patch adds support for AArch64 to the gdb.trace testsuite.
Note that it does not add support for testing fast tracepoint as it
isn't supported. Therefore the test cases with inline assembly are not
ported in this patch, as we do not know what the conditions for
inserting a fast tracepoint on AArch64 would be.
gdb/testsuite/ChangeLog:
* gdb.trace/backtrace.exp: Set registers for aarch64 target.
* gdb.trace/collection.exp: Likewise.
* gdb.trace/mi-trace-frame-collected.exp: Likewise.
* gdb.trace/mi-trace-unavailable.exp: Likewise.
* gdb.trace/report.exp: Likewise.
* gdb.trace/trace-break.exp: Likewise.
* gdb.trace/unavailable.exp: Likewise.
* gdb.trace/while-dyn.exp: Likewise.
This patch implements the 'collect $_ret' command to collect the return
address of a function in a tracepoint. It marks the LR register for
collection.
gdb/ChangeLog:
* aarch64-tdep.c: Add ax.h and ax-gdb.h includes.
(aarch64_gen_return_address): New function.
(aarch64_gdbarch_init): Hook it.
The stub unwinder is used on AArch64 if the target's memory is not
readable at the current PC. For example, the user could try to call at
an invalid address such as 0x0, as covered in the gdb.base/signull.exp
test case. Many GDB ports use a similar unwinder to handle this case
too.
If we purposely kill the inferior before examining the trace then we get
the following issue:
~~~
...
(gdb) trace f
Tracepoint 3 at 0x7fb7fc28c0
(gdb) tstart
(gdb) continue
...
(gdb) tstop
(gdb) tsave /tmp/trace
(gdb) kill
...
(gdb) target tfile /tmp/trace
...
(gdb) tfind
Register 31 is not available.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Found trace frame 0, tracepoint 3
#-1 0x0000007fb7fc28c0 in f () ...
^^^
~~~
This patch teaches the stub unwinder to report to the core frame code
with UNWIND_UNAVAILABLE when either the stack pointer of the return
address are unavailable to read from the target.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_make_stub_cache): Set available_p and
swallow NOT_AVAILABLE_ERROR.
(aarch64_stub_this_id): Call frame_id_build_unavailable_stack if
available_p is not set.
(aarch64_stub_frame_unwind_stop_reason): New function.
(aarch64_stub_unwind): Install it.
Without debugging information, we have the following issue when
examining a trace buffer:
~~~
...
(gdb) trace f
Tracepoint 3 at 0x7fb7fc28c0
(gdb) tstart
(gdb) continue
...
(gdb) tstop
(gdb) tfind start
Register 31 is not available.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Found trace frame 0, tracepoint 3
#-1 0x0000007fb7fc28c0 in f () ...
^^^
~~~
The reason for this is that the target's stack pointer is unavailable
when examining the trace buffer. What we are seeing is due to the
'tfind' command creating a sentinel frame and unwinding it. If an
exception is thrown, we are left with the sentinel frame being displayed
at level #-1. The exception is thrown when the prologue unwinder tries
to read the stack pointer to construct an ID for the frame.
This patch fixes this by making the prologue unwinder catch
NOT_AVAILABLE_ERROR exceptions when either registers or memory is
unreadable and report back to the frame core code with
UNWIND_UNAVAILABLE.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_prologue_cache) <available_p>: New
field.
(aarch64_make_prologue_cache_1): New function, factored out from
aarch64_make_prologue_cache. Do not allocate cache. Set
available_p.
(aarch64_make_prologue_cache): Reimplement wrapping
aarch64_make_prologue_cache_1, and swallowing
NOT_AVAILABLE_ERROR.
(aarch64_prologue_frame_unwind_stop_reason): New function.
Return UNWIND_UNAVAILABLE if available_p is not set.
(aarch64_prologue_unwind): Install it.
(aarch64_prologue_this_id): Move prev_pc and prev_sp limit
checks into aarch64_prologue_frame_unwind_stop_reason. Call
frame_id_build_unavailable_stack if available_p is not set.
This patch moves the address of the start of a function (func) and the
address from which it was called (prev_pc) into aarch64_prologue_cache.
The idea is to keep accesses to the inferior's registers into
aarch64_make_prologue_cache and aarch64_make_stub_cache.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_prologue_cache) <func, prev_pc>: New
fields.
(aarch64_scan_prologue): Set prev_pc.
(aarch64_make_prologue_cache): Set func.
(aarch64_make_stub_cache): Set prev_pc.
(aarch64_prologue_this_id): Remove local variables id, pc and
func. Read prev_pc and func from cache.
(aarch64_stub_this_id): Read prev_pc from cache.
