MIPS/GAS: Use the module level ISA setting for R6 relaxation
Use the module level ISA setting rather than the last ISA selected with
a `.set' directive in the source file in determination as to whether to
keep PC-relative relocations and then with the original symbol referred,
for the purpose of R6 linker relaxation.
This is so that with e.g. code like this:
b foo
.set mips32r2
...
it's the command line options or any `.module' directive that decides
how to encode any relocation for `foo' rather than the presence of `.set
mips32r2'.
gas/
* config/tc-mips.c (mips_force_relocation): Use `file_mips_opts'
rather than `mips_opts' for the R6 ISA check.
(mips_fix_adjustable): Likewise.
* testsuite/gas/mips/pcrel-reloc-1.d: New test.
* testsuite/gas/mips/pcrel-reloc-1-r6.d: New test.
* testsuite/gas/mips/pcrel-reloc-2.d: New test.
* testsuite/gas/mips/pcrel-reloc-2-r6.d: New test.
* testsuite/gas/mips/pcrel-reloc-3.d: New test.
* testsuite/gas/mips/pcrel-reloc-3-r6.d: New test.
* testsuite/gas/mips/pcrel-reloc-1.s: New test source.
* testsuite/gas/mips/mips.exp: Run the new tests.
MIPS/BFD: Don't stop processing on a cross-mode jump conversion error
As with commit ed53407eec9e ("MIPS/BFD: Don't stop processing on
`bfd_reloc_outofrange'") don't bail out right away and instead continue
processing on a cross-mode jump conversion error, so that any further
issues are also reported. Adjust message formatting accordingly, using
`%X' to abort processing at conclusion. Remove the full stop from the
end of the message, for consistency across error reporting.
Adjust the corresponding test case accordingly and make it trigger the
error twice.
bfd/
* elfxx-mips.c (mips_elf_perform_relocation): Call
`info->callbacks->einfo' rather than `*_bfd_error_handler' and
use the `%X%H' format for the cross-mode jump conversion error
message. Remove the full stop from the end of the message.
Continue processing rather than returning failure.
ld/
* testsuite/ld-mips-elf/mode-change-error-1a.s: Trigger an error
twice rather than once.
* testsuite/ld-mips-elf/mode-change-error-1.d: Adjust
accordingly. Remove the full stop from the end of the message.
Graham Markall [Tue, 21 Jun 2016 13:03:08 +0000 (14:03 +0100)]
Arc assembler: Convert nps400 from a machine type to an extension.
gas * config/tc-arc.c (check_cpu_feature, md_parse_option):
Add nps400 option and feature. Add check for nps400
feature. Refactor existing checks to check subclass before
feature enablement.
(md_show_usage): Document flags for NPS-400 and add some other
undocumented flags.
(cpu_type): Remove nps400 CPU type entry
(check_zol): Remove bfd_mach_arc_nps400 case.
(md_show_usage): Add help on -mcpu=nps400.
(cpu_types): Add entry for nps400 as arc700 plus nps400 extension
set.
* doc/c-arc.texi: Document the -mnps400, -mspfp, -mdpfp, and
-fpuda flags. Document -mcpu=nps400.
* testsuite/gas/arc/nps-400-0.d: Use -mcpu=arc700 -mnps400. Change
expected flags to match ARC700 instead of NPS400.
* testsuite/gas/arc/nps-400-1.d: Use -mcpu=arc700 -mnps400.
* testsuite/gas/arc/nps-400-2.d: Likewise.
* testsuite/gas/arc/nps-400-3.d: Likewise.
* testsuite/gas/arc/nps-400-4.d: Likewise.
* testsuite/gas/arc/nps-400-5.d: Likewise.
* testsuite/gas/arc/nps-400-6.d: Likewise.
* testsuite/gas/arc/nps-400-7.d: Likewise.
* testsuite/gas/arc/textinsn2op01.s: Change opcode of myinsn to
avoid clash with cbba instruction.
* testsuite/gas/arc/textinsn2op01.d: Likewise.
* testsuite/gas/arc/textinsn3op.d: Likewise.
* testsuite/gas/arc/textinsn3op.s: Likewise.
* testsuite/gas/arc/nps-400-0.d: Test using NPS-400 using
-mcpu=nps400 as an alternative to -mcpu=arc700 -mnps400 flags.
ld * testsuite/ld-arc/nps-1a.d: Use -mcpu=arc700 -mnps400.
* testsuite/ld-arc/nps-1b.d: Likewise.
include * opcode/arc.h: Add nps400 extension and instruction
subclass.
Remove ARC_OPCODE_NPS400
* elf/arc.h: Remove E_ARC_MACH_NPS400
opcodes * arc-dis.c (arc_insn_length): Add comment on instruction length.
Use same method for determining instruction length on ARC700 and
NPS-400.
(arc_insn_length, print_insn_arc): Remove bfd_mach_arc_nps400.
* arc-nps400-tbl.h: Make all nps400 instructions ARC700 instructions
with the NPS400 subclass.
* arc-opc.c: Likewise.
Pedro Alves [Tue, 21 Jun 2016 00:11:57 +0000 (01:11 +0100)]
Always switch fork child to the main UI
The following scenario:
- gdb started in normal CLI mode.
- separate MI channel created with new-ui
- inferior output redirected with the "set inferior-tty" command.
- use -exec-run in the MI channel to run the inferior
is presently mishandled.
When we create the inferior, in fork-child.c, right after vfork, we'll
close all the file descriptors in the vfork child, and then dup the
tty to file descriptors 0/1/2, create a session, etc. Note that when
we close all descriptors, we close the file descriptors behind
gdb_stdin/gdb_stdout/gdb_stderr of all secondary UIs... So if
anything goes wrong in the child and it calls warning/error, it'll end
up writting to the current UI's stdout/stderr streams, which are
backed by file descriptors that have since been closed. Because this
happens in a vfork region, the corresponding stdin/stdout/stderr in
the parent/gdb end up corrupted.
The fix is to switch to the main UI right after the vfork, so that
gdb_stdin/gdb_stdout/gdb_stderr are correctly mapped to
stdin/stdout/stderr (and thus to file descriptors 0/1/2), so this code
works as it has always worked.
(Technically, we're doing a lot of stuff we shouldn't be doing after a
vfork, while we should only be calling async-signal-safe functions.)
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* fork-child.c (fork_inferior): Switch the child to the main UI
right after vfork. Save/restore the current UI in the parent.
Flush outputs of the main UI instead of the current UI.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
Pedro Alves [Tue, 21 Jun 2016 00:11:56 +0000 (01:11 +0100)]
Make mi-break.exp always expect breakpoint commands output on the main UI
mi-break.exp regresses when tested with MI running on a secondary UI,
with RUNTESTFLAGS="FORCE_SEPARATE_MI_TTY=1".
The problem is simply that the test sets a breakpoint, and attaches
"print" commands to the breakpoint. Since breakpoint commands always
run with the main UI as current UI, the breakpoint command's output
goes to the main UI. So we need to tweak the test to expect it there.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.mi/mi-break.exp (test_breakpoint_commands): Always expect
breakpoint command's output on the main UI.
(test_break): New procedure, factored out from calls in the top
level.
(top level): Use foreach_with_prefix to test MI as main UI and as
separate UI.
Pedro Alves [Tue, 21 Jun 2016 00:11:55 +0000 (01:11 +0100)]
Add testing infrastruture bits for running with MI on a separate UI
With this, a specific test may can start GDB with MI on a separate UI
by using:
mi_gdb_start separate-mi-tty
In addition, it's also possible to run the whole testsuite with MI on
a separate tty, with:
make check RUNTESTFLAGS="FORCE_SEPARATE_MI_TTY=1"
gdb_main_spawn_id and mi_spawn_id are added so that tests may expect
output from either channel.
While at it, inferior_spawn_id was not being cleared when gdb exits,
unlike the other spawn ids, thus a test that starts gdb more than once
would end up using a stale spawn id.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* README (Testsuite Parameters): Document FORCE_SEPARATE_MI_TTY.
* lib/gdb.exp (default_gdb_exit): Clear inferior_spawn_id.
* lib/mi-support.exp (mi_uncatched_gdb_exit): Unset
gdb_main_spawn_id, mi_spawn_id, unset inferior_spawn_id.
(gdb_main_spawn_id, mi_spawn_id): Declare and
comment.
(mi_create_inferior_pty): New procedure,
factored out from default_mi_gdb_start.
(switch_gdb_spawn_id, mi_gdb_start_separate_mi_tty): New
procedures.
(default_mi_gdb_start): Call mi_gdb_start_separate_mi_tty if the
separate-mi-tty option is specified, or SEPARATE_MI_TTY is set.
Use mi_create_inferior_pty.
(mi_gdb_start): Use eval to pass down args list.
Pedro Alves [Tue, 21 Jun 2016 00:11:55 +0000 (01:11 +0100)]
Add new command to create extra console/mi UIs
With all the previous plumbing in place, it's now easy to add a
command that actually creates a new console/mi UI.
The intended use case is to make it possible and easy for MI frontends
to provide a fully featured GDB console to users, with readline
support, command line editing, history, etc., just like if gdb was
started on the command line. Currently MI frontends have to try to
implement all of that theirselves and make use of "-interpreter-exec
console ...", which is far from perfect. If you ever tried Eclipse's
gdb console window, you'll know what I mean...
Instead of trying to multiplex console through MI, this command let's
just leverage all the built in readline/editing support already inside
gdb.
The plan is for the MI frontend to start GDB in regular console mode,
running inside a terminal emulator widget embedded in Eclipse (which
already exists, for supporting the shell widget; other frontends have
similar widgets), and then tell GDB to run a full MI interpreter on an
specified input/output device, independent of the console.
My original prototype planned to do things the other way around --
start GDB in MI mode, and then start an extra CLI console on separate
tty. I handed over that prototype to Marc Khouzam @ Eclipse CDT, and
after experimentation and discussion, we ended up concluding that
starting GDB in CLI mode instead was both easier and actually also
supported an interesting use case -- connect an Eclipse frontend to a
GDB that is already running outside Eclipse.
The current usage is "new-ui <interpreter> <tty>".
E.g., on a terminal run this scriplet:
$ cat gdb-client
#!/bin/bash
reset
tty
tail -f /dev/null
$ gdb-client
/dev/pts/15
Now run gdb on another terminal, and tell it to start a MI interpreter
on the tty of the other terminal:
...
(gdb) new-ui mi /dev/pts/15
New UI allocated
Now back to the the gdb-client terminal, we'll get an MI prompt, ready
for MI input:
/dev/pts/15
=thread-group-added,id="i1"
(gdb)
You can also start a new UI running a CLI, with:
(gdb) new-ui console /dev/pts/15
Though note that this console won't support readline command editing.
It works as if "set editing off" was entered.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* interps.c (set_top_level_interpreter): New function, factored
out from captured_main.
(interpreter_completer): Make extern.
* interps.h (set_top_level_interpreter, interpreter_completer):
New declarations.
(captured_main): Use set_top_level_interpreter.
* top.c [!O_NOCTTY] (O_NOCTTY): Define as 0.
(open_terminal_stream, new_ui_command): New functions.
(init_main): Install the "new-ui" command.
Pedro Alves [Tue, 21 Jun 2016 00:11:54 +0000 (01:11 +0100)]
Make stdin be per UI
This commit makes each UI have its own "stdin" stream pointer. This
is used to determine whether the "from_tty" argument to
execute_command, etc. should be true.
Related, this commit makes input_from_terminal_p take an UI parameter,
and then avoids the gdb_has_a_terminal in it. gdb_has_a_terminal only
returns info on gdb's own main/primary terminal (the real stdin).
However, the places that call input_from_terminal_p really want to
know is whether the command came from an interactive tty. This patch
thus renames input_from_terminal_p to input_interactive_p for clarity,
and then makes input_interactive_p check for "set interactive" itself,
along with ISATTY, instead of calling gdb_has_a_terminal. Actually,
quit_force wants to call input_interactive_p _after_ stdin is closed,
we can't call ISATTY that late. So instead we save the result of
ISATTY in a field of the UI.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-script.c (read_next_line): Adjust to per-UI stdin.
(read_command_lines): Use input_interactive_p instead of
input_from_terminal_p.
* defs.h (struct ui): Forward declare.
(input_from_terminal_p): Rename to ...
(input_interactive_p): ... this.
* event-top.c (stdin_event_handler): Pass 0 as from_tty argument
to quit_command.
(command_handler): Adjust to per-UI stdin.
(handle_line_of_input): Adjust to per-UI stdin and use
input_interactive_p instead of ISATTY and input_from_terminal_p.
(gdb_readline_no_editing_callback): Adjust to per-UI stdin.
(command_line_handler): Always pass true as "from_tty" parameter
of handle_line_of_input and execute_command.
(async_sigterm_handler): Pass 0 as from_tty argument to
quit_command.
* inflow.c (interactive_mode, show_interactive_mode): Moved to ...
(gdb_has_a_terminal): Don't check interactive_mode here.
(_initialize_inflow): Don't install "set interactive-mode" here.
* main.c (captured_command_loop): Adjust to per-UI stdin.
* mi/mi-interp.c (mi_execute_command_wrapper): Adjust to per-UI
stdin.
* top.c (new_ui): Save the stdin stream and whether it's a tty.
(dont_repeat): Adjust to per-UI stdin.
(command_line_input): Adjust to per-UI stdin and to use
input_interactive_p.
(quit_force): Write history if any UI supports interactive input.
(interactive_mode, show_interactive_mode): Move here, from
inflow.c.
(input_from_terminal_p): Rename to ...
(input_interactive_p): ... this, and check the "interactive_mode"
global instead of calling gdb_has_a_terminal.
(_initialize_top): Install "set interactive-mode" here.
* top.h (struct ui) <stdin_stream, input_interactive_p>: New
fields.
* utils.c (quit): Pass 0 as from_tty argument to quit_force.
(defaulted_query): Adjust to per-UI stdin and to use
input_interactive_p.
Pedro Alves [Tue, 21 Jun 2016 00:11:53 +0000 (01:11 +0100)]
Make main_ui be heap allocated
This is preparation for being able to create more than one UI object.
The change to gdb_main to stop using catch_errors is necessary because
catch_errors references current_uiout, which expands to
current_ui->m_current_ui, which would crash because current_ui is not
initialized yet at that point. It didn't trigger earlier in the
series because before this patch, main_ui/current_ui always start out
non-NULL.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* event-top.c (main_ui_): Delete.
(main_ui, current_ui, ui_list): No longer initialize here.
* main.c (captured_main): UI initialization code factored out to
new new_ui function.
(gdb_main): Wrap captured_main with TRY/CATCH instead of
catch_errors.
* top.c (highest_ui_num): New global.
(new_ui): New function.
* top.h (struct ui) <num>: New field.
(new_ui): New declaration.
Pedro Alves [Tue, 21 Jun 2016 00:11:53 +0000 (01:11 +0100)]
Only send sync execution command output to the UI that ran the command
Currently when a "step", "next", etc. finishes, the current source
line is printed on all console UIs.
This patch makes the CLI and TUI interpreters reuse MI's logic to only
emit console output related to a synchronous command on the
console-like interpreter that started the command in the first place.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-interp.c (cli_on_normal_stop): Bail out early if there's
nothing to print. Use should_print_stop_to_console.
* tui/tui-interp.c (tui_on_normal_stop): Likewise.
Pedro Alves [Tue, 21 Jun 2016 00:11:53 +0000 (01:11 +0100)]
Push thread->control.command_interp to the struct thread_fsm
I noticed that if we step into an inline function, step_1 never
reaches proceed, and thus nevers sets the thread's
tp->control.command_interp. Because of that,
should_print_stop_to_console fails to determine that is should print
stop output to the console.
The fix is to set the thread's command_interp earlier. However, I
realized that we can move that field to the thread_fsm, given that its
lifetime is exactly the same as thread_fsm. So the patch plumbs all
fsms constructors to take the command interp and store it in the
thread_fsm.
We can see the fix in action, with e.g., the gdb.opt/inline-cmds.exp
test, and issuing a step when stopped at line 67:
&"s\n"
^running
*running,thread-id="all"
(gdb)
~"67\t result = func2 ();\n"
*stopped,reason="end-stepping-range",frame={addr="0x00000000004004d0",func="main",args=[],file="/home/pedro/gdb/mygit/src/gdb/testsuite/gdb.opt/inline-cmds.c",fullname="/home/pedro/gdb/mygit/src/gdb/testsuite/gdb.opt/inline-cmds.c",line="67"},thread-id="1",stopped-threads="all",core="0"
(gdb)
s
&"s\n"
^running
*running,thread-id="all"
(gdb)
+ ~"func2 () at /home/pedro/gdb/mygit/src/gdb/testsuite/gdb.opt/inline-cmds.c:67\n"
+ ~"67\t result = func2 ();\n"
*stopped,reason="end-stepping-range",frame={addr="0x00000000004004d0",func="func2",args=[],file="/home/pedro/gdb/mygit/src/gdb/testsuite/gdb.opt/inline-cmds.c",fullname="/home/pedro/gdb/mygit/src/gdb/testsuite/gdb.opt/inline-cmds.c",line="67"},thread-id="1",stopped-threads="all",core="0"
(gdb)
(The inline-cmds.exp command is adjusted to exercise this.)
(Due to the follow_fork change, this also fixes "next N" across a fork
with "set follow-fork child" with "set detach-on-fork on". Commands
that rely on internal breakpoints, like "finish" will still require
more work to migrate breakpoints etc. to the child thread.)
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* breakpoint.c (new_until_break_fsm): Add 'cmd_interp' parameter.
(until_break_fsm_should_stop, until_break_fsm_clean_up): Add
thread parameter.
(until_break_command): Pass command interpreter to thread fsm
ctor.
* cli/cli-interp.c (should_print_stop_to_console): Adjust.
* gdbthread.h (struct thread_control_state) <command_interp>:
Delete field.
* infcall.c (new_call_thread_fsm): Add 'cmd_interp' parameter.
Pass it down.
(call_thread_fsm_should_stop): Add thread parameter.
(call_function_by_hand_dummy): Pass command interpreter to thread
fsm ctor. Pass thread pointer to fsm clean up method.
* infcmd.c: Include interps.h.
(struct step_command_fsm) <thread>: Delete field.
(new_step_command_fsm): Add 'cmd_interp' parameter. Pass it down.
(step_command_fsm_prepare): Remove references to fsm's thread
field.
(step_1): Pass command interpreter to thread
fsm ctor. Pass thread pointer to fsm clean up method.
(step_command_fsm_should_stop, step_command_fsm_clean_up): Add
thread parameter and use it.
(new_until_next_fsm): Add 'cmd_interp' parameter. Pass it down.
(until_next_fsm_should_stop, until_next_fsm_clean_up): Add thread
parameter and use it.
(until_next_command): Pass command interpreter to thread fsm ctor.
(struct finish_command_fsm) <thread>: Delete field.
(finish_command_fsm_ops): Add NULL slot for should_notify_stop.
(new_finish_command_fsm): Add 'cmd_interp' parameter and pass it
down. Remove thread parameter and adjust.
(finish_command_fsm_should_stop, finish_command_fsm_clean_up): Add
thread parameter and use it.
(finish_command): Pass command interpreter to thread fsm ctor.
Don't pass thread.
* infrun.c (follow_fork): Move thread fsm to child fork instead of
command interpreter, only.
(clear_proceed_status_thread): Remove reference to command_interp.
(proceed): Don't record the thread's command interpreter.
(clean_up_just_stopped_threads_fsms): Pass thread to fsm clean_up
method.
(fetch_inferior_event): Pass thread to fsm should_stop method.
* thread-fsm.c (thread_fsm_ctor): Add 'cmd_interp' parameter.
Store it.
(thread_fsm_clean_up, thread_fsm_should_stop): Add thread
parameter and pass it down.
* thread-fsm.h (struct thread_fsm) <command_interp>: New field.
(struct thread_fsm_ops) <clean_up, should_stop>: Add thread
parameter.
(thread_fsm_ctor): Add 'cmd_interp' parameter.
(thread_fsm_clean_up, thread_fsm_should_stop): Add thread
parameter.
* thread.c (thread_cancel_execution_command): Pass thread to
thread fsm clean_up method.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.opt/inline-cmds.c: Add "set mi break here" marker.
* gdb.opt/inline-cmds.exp: Add MI tests.
Pedro Alves [Tue, 21 Jun 2016 00:11:52 +0000 (01:11 +0100)]
New function should_print_stop_to_console
There's code in the MI interpreter that decides whether a stop should
be sent to MI's console stream. Move this check to the CLI
interpreter code, so that we can reuse it in both the CLI and TUI
interpreters.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-interp.c: Include gdbthread.h and thread-fsm.h.
(should_print_stop_to_console): New function, factored out from
mi_on_normal_stop_1.
* cli/cli-interp.h (should_print_stop_to_console): Declare.
* mi/mi-interp.c (mi_on_normal_stop_1): Use
should_print_stop_to_console. Pass it the current UI's console
interpreter.
* mi/mi-main.c (captured_mi_execute_command): Use the
INTERP_CONSOLE symbol rather than explicit "console".
The test sets a breakpoint that runs "continue", a foreground command.
When we get to run the "continue", we've already emitted the *stopped
event on the MI UI, and set its prompt state to PROMPT_NEEDED (this is
done from within normal_stop). Since inferior events are always
handled with the main UI as current UI, breakpoint commands always run
with the main UI as current UI too. This means that the "continue"
ends up always disabling the prompt on the main UI, instead of the UI
that had just been done with synchronous execution.
I think we'll want to extend this with a concept of "set of
threads/inferiors a UI/interpreter is blocked waiting on", but I'm
leaving that for a separate series.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* infcmd.c (prepare_execution_command): Use
all_uis_on_sync_execution_starting.
* infrun.c (all_uis_on_sync_execution_starting): New function.
* infrun.h (all_uis_on_sync_execution_starting): Declare.
Pedro Alves [Tue, 21 Jun 2016 00:11:51 +0000 (01:11 +0100)]
Replace the sync_execution global with a new enum prompt_state tristate
When sync_execution (a boolean) is true, it means we're running a
foreground command -- we hide the prompt stop listening to input, give
the inferior the terminal, then go to the event loop waiting for the
target to stop.
With multiple independent UIs, we need to track whether each UI is
synchronously blocked waiting for the target. IOW, if you do
"continue" in one console, that console stops accepting commands, but
you should still be free to type other commands in the others
consoles.
Just simply making sync_execution be per-UI alone not sufficient,
because of this in fetch_inferior_event:
/* If the inferior was in sync execution mode, and now isn't,
restore the prompt (a synchronous execution command has finished,
and we're ready for input). */
if (current_ui->async && was_sync && !sync_execution)
observer_notify_sync_execution_done ();
We'd have to record at entry the "was_sync" state for each UI, not
just of the current UI.
This patch instead replaces the sync_execution flag by a per-UI
tristate flag indicating the command line prompt state:
enum prompt_state
{
/* The command line is blocked simulating synchronous execution.
This is used to implement the foreground execution commands
('run', 'continue', etc.). We won't display the prompt and
accept further commands until the execution is actually over. */
PROMPT_BLOCKED,
/* The command finished; display the prompt before returning back to
the top level. */
PROMPT_NEEDED,
/* We've displayed the prompt already, ready for input. */
PROMPTED,
;
I think the end result is _much_ clearer than the current code, and,
it addresses the original motivation too.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* annotate.c: Include top.h.
(async_background_execution_p): Delete.
(print_value_flags): Check the UI's prompt state rather then
async_background_execution_p.
* event-loop.c (start_event_loop): Set the prompt state to
PROMPT_NEEDED.
* event-top.c (display_gdb_prompt, async_enable_stdin)
(async_disable_stdin): Check the current UI's prompt state instead
of the sync_execution global.
(command_line_handler): Set the prompt state to PROMPT_NEEDED
before running a command, and display the prompt if still needed
afterwards.
* infcall.c (struct call_thread_fsm) <waiting_ui>: New field.
(new_call_thread_fsm): New parameter 'waiting_ui'. Store it.
(call_thread_fsm_should_stop): Set the prompt state to
PROMPT_NEEDED.
(run_inferior_call): Adjust to temporarily set the prompt state to
PROMPT_BLOCKED instead of using the sync_execution global.
(call_function_by_hand_dummy): Pass the current UI to
new_call_thread_fsm.
* infcmd.c: Include top.h.
(continue_1): Check the current UI's prompt state instead of the
sync_execution global.
(continue_command): Validate global execution state before calling
prepare_execution_command.
(step_1): Call all_uis_check_sync_execution_done.
(attach_post_wait): Don't call async_enable_stdin here. Remove
reference to sync_execution.
* infrun.c (sync_execution): Delete global.
(follow_fork_inferior)
(reinstall_readline_callback_handler_cleanup): Check the current
UI's prompt state instead of the sync_execution global.
(check_curr_ui_sync_execution_done)
(all_uis_check_sync_execution_done): New functions.
(fetch_inferior_event): Call all_uis_check_sync_execution_done
instead of trying to determine whether the global sync execution
changed.
(handle_no_resumed): Check the prompt state of all UIs.
(normal_stop): Emit the no unwait-for even to all PROMPT_BLOCKED
UIs. Emit the "Switching to" notification to all UIs. Enable
stdin in all UIs.
* infrun.h (sync_execution): Delete.
(all_uis_check_sync_execution_done): Declare.
* main.c (captured_command_loop): Don't call
interp_pre_command_loop if the prompt is blocked.
(catch_command_errors, catch_command_errors_const): Adjust.
(captured_main): Set the initial prompt state to PROMPT_NEEDED.
* mi/mi-interp.c (display_mi_prompt): Set the prompt state to
PROMPTED.
(mi_interpreter_resume): Don't clear sync_execution. Remove hack
comment.
(mi_execute_command_input_handler): Set the prompt state to
PROMPT_NEEDED before executing the command, and only display the
prompt if the prompt state is PROMPT_NEEDED afterwards.
(mi_on_resume_1): Adjust to check the prompt state.
* target.c (target_terminal_inferior): Adjust to check the prompt
state.
* top.c (wait_sync_command_done, maybe_wait_sync_command_done)
(execute_command): Check the current UI's prompt state instead of
sync_execution.
* top.h (enum prompt_state): New.
(struct ui) <prompt_state>: New field.
(ALL_UIS): New macro.
Pedro Alves [Tue, 21 Jun 2016 00:11:51 +0000 (01:11 +0100)]
Simplify starting the command event loop
All interpreter types (CLI/TUI/MI) print the prompt, and then call
start_event_loop.
Because we'll need an interpreter hook to display the
interpreter-specific prompt before going back to the event loop,
without actually starting an event loop, this patch moves the
start_event_loop call to common code, and replaces the command_loop
hook with a pre_command_look hook, that now just prints the prompt.
Turns out to be a cleanup on its own right anyway.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-interp.c (cli_interpreter_pre_command_loop): New
function.
(cli_interp_procs): Install it instead of cli_command_loop.
* cli/cli-interp.h (cli_interpreter_pre_command_loop): Declare.
* event-top.c (cli_command_loop): Delete.
* interps.c (interp_new): Remove reference to command_loop_proc.
(current_interp_command_loop): Delete.
(interp_pre_command_loop): New function.
(interp_command_loop_ftype): Delete.
* interps.h (interp_pre_command_loop_ftype): New typedef.
(struct interp_procs) <command_loop_proc>: Delele field.
<pre_command_loop_proc>: New field.
(current_interp_command_loop): Delete declaration.
(interp_pre_command_loop): New declaration.
* main.c (captured_command_loop): Call interp_pre_command_loop
instead of current_interp_command_loop and start an event loop.
* mi/mi-interp.c (mi_command_loop): Delete.
(mi_interpreter_pre_command_loop): New.
(mi_interp_procs): Update.
* tui/tui-interp.c (tui_interp_procs): Install
cli_interpreter_pre_command_loop instead of cli_command_loop.
Pedro Alves [Tue, 21 Jun 2016 00:11:49 +0000 (01:11 +0100)]
Make target_terminal_inferior/ours almost nops on non-main UIs
Since we always run the inferior in the main console (unless "set
inferior-tty" is in effect), when some UI other than the main one
calls target_terminal_inferior/target_terminal_inferior, then we only
register/unregister the UI's input from the event loop, but leave the
main UI's terminal settings as is.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* target.c (target_terminal_inferior): Bail out after
unregistering input_fd if not on the main UI.
(target_terminal_ours): Bail out after registering input_fd if not
on the main UI.
(target_terminal_ours_for_output): Bail out if not on the main UI.
Pedro Alves [Tue, 21 Jun 2016 00:11:49 +0000 (01:11 +0100)]
Always process target events in the main UI
This makes target events always be always processed with the main UI
as current UI. This way, warnings, debug output, etc. are always
consistently sent to the main console.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* event-top.c (restore_ui_cleanup): Make extern.
* infrun.c (fetch_inferior_event): Always switch to the main UI.
* top.h (restore_ui_cleanup): Declare.
Pedro Alves [Tue, 21 Jun 2016 00:11:48 +0000 (01:11 +0100)]
Make command line editing (use of readline) be per UI
Due to the way that readline's API works (based on globals), we can
only have one instance of readline in a process. So the goal of this
patch is to only allow editing in the main UI, and make sure that only
one UI calls into readline. Some MI paths touch readline variables
currently, which is bad as that is changing variables that matter for
the main console UI. This patch fixes those.
This actually fixes a nasty bug -- starting gdb in MI mode ("gdb
-i=mi"), and then doing "set editing on" crashes GDB, because MI is
not prepared to use readline:
set editing on
&"set editing on\n"
=cmd-param-changed,param="editing",value="on"
^done
(gdb)
p 1
readline: readline_callback_read_char() called with no handler!
Aborted (core dumped)
The fix for that was to add an interp_proc method to query the
interpreter whether it actually supports editing. New test included.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
PR mi/20034
* cli/cli-interp.c: Include cli-interp.h and event-top.h.
(cli_interpreter_resume): Pass 1 to gdb_setup_readline. Set the
UI's input_handler here.
(cli_interpreter_supports_command_editing): New function.
(cli_interp_procs): Install it.
* cli/cli-interp.h: New file.
* event-top.c (async_command_editing_p): Rename to ...
(set_editing_cmd_var): ... this.
(change_line_handler): Add parameter 'editing', and use it. Bail
early if the interpreter doesn't support editing. Don't touch
readline state if editing is off.
(gdb_rl_callback_handler_remove, gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): Assert the current UI is the
main UI.
(display_gdb_prompt): Don't call gdb_rl_callback_handler_remove if
not using readline. Check whether the current UI is using command
editing instead of checking the async_command_editing_p global.
(set_async_editing_command): Delete.
(gdb_setup_readline): Add 'editing' parameter. Only allow editing
on the main UI. Don't touch readline state if editing is off.
(gdb_disable_readline): Don't touch readline state if editing is
off.
* event-top.h (gdb_setup_readline): Add 'int' parameter.
(set_async_editing_command): Delete declaration.
(change_line_handler, command_line_handler): Declare.
(async_command_editing_p): Rename to ...
(set_editing_cmd_var): ... this.
* infrun.c (reinstall_readline_callback_handler_cleanup): Check
whether the current UI has editing enabled rather than checking
the async_command_editing_p global.
* interps.c (interp_supports_command_editing): New function.
* interps.h (interp_supports_command_editing_ftype): New typedef.
(struct interp_procs) <supports_command_editing_proc>: New field.
(interp_supports_command_editing): Declare.
* mi/mi-interp.c (mi_interpreter_resume): Pass 0 to
gdb_setup_readline. Don't clear the async_command_editing_p
global. Update comments.
* top.c (gdb_readline_wrapper_line, gdb_readline_wrapper): Check
whether the current UI has editing enabled rather than checking
the async_command_editing_p global. Don't touch readline state if
editing is off.
(undo_terminal_modifications_before_exit): Switch to the main UI.
Unconditionally call gdb_disable_readline.
(set_editing): New function.
(show_async_command_editing_p): Rename to ...
(show_editing): ... this. Show the state of the current UI.
(_initialize_top): Adjust.
* top.h (struct ui) <command_editing>: New field.
* tui/tui-interp.c: Include cli/cli-interp.h.
(tui_resume): Pass 1 to gdb_setup_readline. Set the UI's
input_handler.
(tui_interp_procs): Install
cli_interpreter_supports_command_editing.
* tui/tui-io.c (tui_getc): Check whether the current UI has
editing enabled rather than checking the async_command_editing_p
global.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
Pedro Alves [Tue, 21 Jun 2016 00:11:47 +0000 (01:11 +0100)]
Make out and error streams be per UI
stderr_fileopen () references stderr directly, which doesn't work when
we have a separate UI with its own stderr-like stream. So this also
adds a "errstream" to "struct ui", and plumbs stderr_fileopen to take
a stream parameter.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* event-top.c (gdb_setup_readline): Pass the UI's outstream and
errstream to stdout_fileopen and stderr_fileopen.
* exceptions.c: Include top.h.
(print_flush): Open the current UI's outstream file descriptor,
instead of hardcoding file descriptor 1.
* main.c (captured_main): Save the main UI's out and error
streams. Adjust stderr_fileopen call.
* top.h (struct ui) <outstream, errstream>: New fields.
* ui-file.c (stderr_fileopen): Add stream parameter. Use it
instead of stderr.
* ui-file.h (stderr_fileopen): Add stream parameter and update
comment.
Pedro Alves [Tue, 21 Jun 2016 00:11:47 +0000 (01:11 +0100)]
Make input_fd be per UI
And with that, we can switch the current UI to the UI whose input
descriptor woke up the event loop. IOW, if the user types in UI 2,
the event loop wakes up, switches to UI 2, and processes the input.
Next the user types in UI 3, the event loop wakes up and switches to
UI 3, etc.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* event-top.c (input_fd): Delete.
(stdin_event_handler): Switch to the UI whose input descriptor got
the event. Adjust to per-UI input_fd.
(gdb_setup_readline): Don't set the input_fd global. Adjust to
per-UI input_fd.
(gdb_disable_readline): Adjust to per-UI input_fd.
* event-top.h (input_fd): Delete declaration.
* linux-nat.c (linux_nat_terminal_inferior): Don't remove input_fd
from the event-loop here.
(linux_nat_terminal_ours): Don't register input_fd in the
event-loop here.
* main.c (captured_main): Adjust to per-UI input_fd.
* remote.c (remote_terminal_inferior): Don't remove input_fd from
the event-loop here.
(remote_terminal_ours): Don't register input_fd in the event-loop
here.
* target.c: Include top.h and event-top.h.
(target_terminal_inferior): Remove input_fd from the event-loop
here.
(target_terminal_ours): Register input_fd in the event-loop.
* top.h (struct ui) <input_fd>: New field.
Pedro Alves [Tue, 21 Jun 2016 00:11:46 +0000 (01:11 +0100)]
Always run async signal handlers in the main UI
Async signal handlers have no connection to whichever was the current
UI, and thus always run on the main one.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* event-loop.c: Include top.h.
(invoke_async_signal_handlers): Switch to the main UI.
* event-top.c (main_ui_): Update comment.
(main_ui): New global.
* top.h (main_ui): Declare.
Pedro Alves [Tue, 21 Jun 2016 00:11:45 +0000 (01:11 +0100)]
Make the intepreters output to all UIs
When we have multiple consoles, MI channels, etc., then we need to
broadcast breakpoint hits, etc. to all UIs. In the past, I've
adjusted most of the run control to communicate events to the
interpreters through observer notifications, so events would be
properly sent to console and MI streams, in sync and async modes.
This patch does the next logical step -- have each interpreter's
observers output interpreter-specific info to _all_ UIs.
Note that when we have multiple instances of active cli/tui
interpreters, then the cli_interp and tui_interp globals no longer
work. This is addressed by this patch.
Also, the interpreters currently register some observers when resumed
and remove them when suspended. If we have multiple instances of the
interpreters, and they can be suspended/resumed at different,
independent times, that no longer works. What we instead do is always
install the observers, and then have the observers themselves know
when to do nothing.
An earlier prototype of this series did the looping over struct UIs in
common code, and then dispatched events to the interpreters through a
matching interp_on_foo method for each observer. That turned out a
lot more complicated than the present solution, as we'd end up with
having to create a new interp method every time some interpreter
wanted to listen to some observer notification, resulting in a lot of
duplicated make-work and more coupling than desirable.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-interp.c (cli_interp): Delete.
(as_cli_interp): New function.
(cli_on_normal_stop, cli_on_signal_received)
(cli_on_end_stepping_range, cli_on_signal_exited, cli_on_exited)
(cli_on_no_history): Send output to all CLI UIs.
(cli_on_sync_execution_done, cli_on_command_error): Skip output if
the top level interpreter is not a CLI.
(cli_interpreter_init): Don't set cli_interp or install observers
here.
(_initialize_cli_interp): Install observers here.
* event-top.c (main_ui_, ui_list): New globals.
(current_ui): Point to main_ui_.
(restore_ui_cleanup, switch_thru_all_uis_init)
(switch_thru_all_uis_cond, switch_thru_all_uis_next): New
functions.
* mi/mi-interp.c (as_mi_interp): New function.
(mi_interpreter_init): Don't install observers here.
(mi_on_sync_execution_done): Skip output if the top level
interpreter is not a MI.
(mi_new_thread, mi_thread_exit, mi_record_changed)
(mi_inferior_added, mi_inferior_appeared, mi_inferior_exit)
(mi_inferior_removed): Send output to all MI UIs.
(find_mi_interpreter, mi_interp_data): Delete.
(find_mi_interp): New function.
(mi_on_signal_received, mi_on_end_stepping_range)
(mi_on_signal_exited, mi_on_exited, mi_on_no_history): Send output
to all MI UIs.
(mi_on_normal_stop): Rename to ...
(mi_on_normal_stop_1): ... this.
(mi_on_normal_stop): Reimplement, sending output to all MI UIs.
(mi_traceframe_changed, mi_tsv_created, mi_tsv_deleted)
(mi_tsv_modified, mi_breakpoint_created, mi_breakpoint_deleted)
(mi_breakpoint_modified, mi_output_running_pid): Send output to
all MI UIs.
(mi_on_resume): Rename to ...
(mi_on_resume_1): ... this. Don't handle infcalls here.
(mi_on_resume): Reimplement, sending output to all MI UIs.
(mi_solib_loaded, mi_solib_unloaded, mi_command_param_changed)
(mi_memory_changed): Send output to all MI UIs.
(report_initial_inferior): Install observers here.
* top.h (struct ui) <next>: New field.
(ui_list): Declare.
(struct switch_thru_all_uis): New.
(switch_thru_all_uis_init, switch_thru_all_uis_cond)
(switch_thru_all_uis_next): Declare.
(SWITCH_THRU_ALL_UIS): New macro.
* tui/tui-interp.c (tui_interp): Delete global.
(as_tui_interp): New function.
(tui_on_normal_stop, tui_on_signal_received)
(tui_on_end_stepping_range, tui_on_signal_exited, tui_on_exited)
(tui_on_no_history): Send output to all TUI UIs.
(tui_on_sync_execution_done, tui_on_command_error): Skip output if
the top level interpreter is not a TUI.
(tui_init): Don't set tui_interp or install observers here.
(_initialize_tui_interp): Install observers here.
Pedro Alves [Tue, 21 Jun 2016 00:11:45 +0000 (01:11 +0100)]
Introduce interpreter factories
If every UI instance has its own set of interpreters, then the current
scheme of creating the interpreters at GDB initialization time no
longer works. We need to create them whenever a new UI instance is
created.
The scheme implemented here has each interpreter register a factory
callback that when called creates a new instance of a specific
interpreter type. Then, when some code in gdb looks up an interpreter
(always by name), if there's none yet, the factory method is called to
construct one.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-interp.c (cli_uiout): Delete, moved into ...
(struct cli_interp): ... this new structure.
(cli_on_normal_stop, cli_on_signal_received)
(cli_on_end_stepping_range, cli_on_signal_exited, cli_on_exited)
(cli_on_no_history): Use interp_ui_out.
(cli_interpreter_init): If top level, set the cli_interp global.
(cli_interpreter_init): Return the interp's data instead of NULL.
(cli_interpreter_resume, cli_interpreter_exec, cli_ui_out): Adjust
to cli_uiout being in the interpreter's data.
(cli_interp_procs): New, factored out from _initialize_cli_interp.
(cli_interp_factory): New function.
(_initialize_cli_interp): Call interp_factory_register.
* interps.c (get_interp_info): New, factored out from ...
(get_current_interp_info): ... this.
(interp_new): Add parameter 'data'. Store it.
(struct interp_factory): New function.
(interp_factory_p): New typedef. Define a VEC_P.
(interpreter_factories): New global.
(interp_factory_register): New function.
(interp_add): Add 'ui' parameter. Use get_interp_info and
interp_lookup_existing.
(interp_lookup): Rename to ...
(interp_lookup_existing): ... this. Add 'ui' parameter. Don't
check for NULL or empty name here.
(interp_lookup): Add 'ui' parameter and reimplement.
(interp_set_temp, interpreter_exec_cmd): Adjust.
(interpreter_completer): Complete on registered interpreter
factories instead of interpreters.
* interps.h (interp_factory_func): New typedef.
(interp_factory_register): Declare.
(interp_new, interp_add): Adjust.
(interp_lookup): Declare.
* main.c (captured_main): Adjust.
* mi/mi-interp.c (mi_cmd_interpreter_exec): Adjust.
(mi_interp_procs): New, factored out from
_initialize_mi_interp.
(mi_interp_factory): New function.
* python/python.c (execute_gdb_command): Adjust.
* tui/tui-interp.c (tui_init): If top level, set the tui_interp
global.
(tui_interp_procs): New.
(tui_interp_factory): New function.
(_initialize_tui_interp): Call interp_factory_register.
Pedro Alves [Tue, 21 Jun 2016 00:11:45 +0000 (01:11 +0100)]
Make the interpreters be per UI
Make each UI have its own interpreter list, top level interpreter,
current interpreter, etc. The "interpreter_async" global is not
really specific to an struct interp (it crosses interpreter-exec ...),
so I moved it to "struct ui" directly, while the other globals were
left hidden in interps.c, opaque to the rest of GDB.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* breakpoint.c (bpstat_do_actions_1): Access the current UI's
async field instead of the interpreter_async global.
* cli/cli-script.c (execute_user_command, while_command)
(if_command, script_from_file): Likewise.
* compile/compile.c: Include top.h instead of interps.h.
(compile_file_command, compile_code_command)
(compile_print_command): Access the current UI's async field
instead of the interpreter_async global.
* guile/guile.c: Include top.h instead of interps.h.
(guile_repl_command, guile_command, gdbscm_execute_gdb_command):
Access the current UI's async field instead of the
interpreter_async global.
* guile/scm-ports.c: Include top.h instead of interps.h.
(ioscm_with_output_to_port_worker): Access the current UI's async
field instead of the interpreter_async global.
* inf-loop.c (inferior_event_handler): Likewise.
* infcall.c (run_inferior_call): Likewise.
* infrun.c (reinstall_readline_callback_handler_cleanup)
(fetch_inferior_event): Likewise.
* interps.c (interpreter_async): Delete.
(struct ui_interp_info): New.
(get_current_interp_info): New function.
(interp_list, current_interpreter, top_level_interpreter_ptr):
Delete.
(interp_add, interp_set, interp_lookup, interp_ui_out)
(current_interp_set_logging, interp_set_temp)
(current_interp_named_p): Adjust to per-UI interpreters.
(command_interpreter): Delete.
(command_interp, current_interp_command_loop, interp_quiet_p)
(interp_exec, interpreter_exec_cmd, interpreter_completer)
(top_level_interpreter, top_level_interpreter_data): Adjust to
per-UI interpreters.
* interps.h (interpreter_async): Delete.
* main.c (captured_command_loop): Access the current UI's async
field instead of the interpreter_async global.
* python/python.c (python_interactive_command, python_command)
(execute_gdb_command): Likewise.
* top.c (maybe_wait_sync_command_done, execute_command_to_string):
Access the current UI's async field instead of the
interpreter_async global.
* top.h (struct tl_interp_info): Forward declare.
(struct ui) <interp_info, async>: New fields.
Pedro Alves [Tue, 21 Jun 2016 00:11:44 +0000 (01:11 +0100)]
Make gdb_stdout&co be per UI
We need to have these send output to the proper UI.
However, this patch still make them look like globals. Kind of like
__thread variables, if you will. Changing everything throughout to
write something like current_ui->gdb_stdout instead would be massive
overkill, IMNSHO.
This leaves gdb_stdtargin/stdtarg/stdtargerr global, but maybe that was a
mistake, I'm not sure -- IIRC, MI formats target I/O differently, so
if we have a separate MI channel, then I guess target output should go
there instead of to gdb's stdout. OTOH, maybe GDB should send that
instead to "set inferior-tty", instead of multiplexing it over MI. We
can always fix those later when it gets clearer where they should go.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* main.c (gdb_stdout, gdb_stderr, gdb_stdlog, gdb_stdin): Delete
globals.
(gen_ret_current_ui_field_ptr): New macro. Use it to generate
wrappers for gdb_stdout, gdb_stderr, gdb_stdlog and gdb_stdin.
* top.h (struct ui) <m_gdb_stdout, m_gdb_stdin, m_gdb_stderr,
m_gdb_stdlog>: New fields.
(current_ui_gdb_stdout_ptr, current_ui_gdb_stdin_ptr)
(current_ui_gdb_stderr_ptr, current_ui_gdb_stdlog_ptr): Declare.
(gdb_stdout, gdb_stdin, gdb_stderr, gdb_stdlog): Reimplement as
macros.
Pedro Alves [Tue, 21 Jun 2016 00:11:44 +0000 (01:11 +0100)]
Introduce "struct ui"
This is a step towards supporting multiple consoles/MIs, each on its
own stdio streams / terminal.
See intro comment in top.h.
(I've had trouble picking a name for this object. I've started out
with "struct console" originally. But then this is about MI as well,
and there's "interpreter-exec console", which is specifically about
the CLI...
So I changed to "struct terminal", but, then we have a terminal object
that works when the input is not a terminal as well ...
Then I sort of gave up and renamed it to "struct top_level". But it
then gets horribly confusing when we talk about the "top level
interpreter that's running on the current top level".
In the end, I realized we're already sort of calling this "ui", in
struct ui_out, struct ui_file, and a few coments here and there.)
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* event-top.c: Update readline-related comments.
(input_handler, call_readline): Delete globals.
(gdb_rl_callback_handler): Call the current UI's input_handler
method.
(change_line_handler): Adjust to set current UI's properties
instead of globals.
(current_ui_, current_ui): New globals.
(get_command_line_buffer): Rewrite to refer to the current UI.
(stdin_event_handler): Adjust to call the call_readline method of
the current UI.
(gdb_readline_no_editing_callback): Adjust to call the current UI's
input_handler method.
(gdb_setup_readline): Adjust to set current UI's properties
instead of globals.
* event-top.h (call_readline, input_handler): Delete declarations.
* mi/mi-interp.c (mi_interpreter_resume): Adjust to set current
UI's properties instead of globals.
* top.c (gdb_readline_wrapper_cleanup): Adjust to set current UI's
properties instead of globals.
(gdb_readline_wrapper): Adjust to call and set current UI's
methods instead of globals.
* top.h: Include buffer.h and event-loop.h.
(struct ui): New struct.
(current_ui): New declaration.
Pedro Alves [Tue, 21 Jun 2016 00:11:43 +0000 (01:11 +0100)]
[Ada catchpoints] Fix "warning: failed to get exception name: No definition of \"e.full_name\" in current context"
Looking at testsuite results, I noticed this warning in an MI test:
~"\nCatchpoint "
~"2, "
&"warning: failed to get exception name: No definition of \"e.full_name\" in current context.\n"
~"exception at 0x000000000040192d in foo () at /home/pedro/brno/pedro/gdb/mygit/src/gdb/testsuite/gdb.ada/mi_catch_ex/foo.adb:20\n"
~"20\t raise Constraint_Error; -- SPOT1\n"
*stopped,reason="breakpoint-hit",disp="keep",bkptno="2",exception-name="CONSTRAINT_ERROR",frame={addr="0x000000000040192d",func="foo",args=[],file="/home/pedro/brno/pedro/gdb/mygit/src/gdb/testsuite/gdb.ada/mi_catch_ex/foo.adb",fullname="/home/pedro/brno/pedro/gdb/mygit/src/gdb/testsuite/gdb.ada/mi_catch_ex/foo.adb",line="20"},thread-id="1",stopped-threads="all",core="5"
(gdb)
PASS: gdb.ada/mi_catch_ex.exp: continue until CE caught by all-exceptions catchpoint
The problem is that:
- MI prints the breakpoint hit twice: once on the MI stream;
another time on the console stream.
- After printing the Ada catchpoint hit, gdb selects a non-current
frame, from within the catchpoint's print_it routine.
So the second time the breakpoint is printed, the selected frame is no
longer the current frame, and then evaluating e.full_name in
ada_exception_name_addr fails.
This commit fixes the problem and enhances the gdb.ada/mi_catch_ex.exp
test to make sure the catchpoint hit is printed correctly on the
console stream too.
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* ada-lang.c (ada_exception_name_addr_1): Add comment.
(print_it_exception): Select the current frame.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.ada/mi_catch_ex.exp (continue_to_exception): New procedure.
(top level): Use it instead of mi_execute_to.
Pedro Alves [Tue, 21 Jun 2016 00:11:43 +0000 (01:11 +0100)]
Prepare gdb.python/mi-py-events.exp for Python/MI in separate channels
Similarly to 5068630ad34dce5fefbe68d70d3a50cd8b92f71e
(gdb.python/py-events.exp and normal_stop observers ordering) [1],
this commit makes the gdb.python/py-mi-events.exp test not rely on
order in which MI and Python observers run, or even on where each
observer sends its output to.
This shows up as a problem when testing with MI running as a separate
terminal, for example, where Python event output and MI output go to
different channels, even. But in any case, relying on the order in
which observers run is always going to be fragile.
The fix is to save the string output in the handlers in some variables
and then having MI print them explicitly, instead of printing them
directly from the Python events.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.python/py-mi-events-gdb.py (stop_handler_str)
(cont_handler_str): New.
(signal_stop_handler): Set stop_handler_str instead of printing to
stdout.
(continue_handler): Set cont_handler_str instead of printing to
stdout.
* gdb.python/py-mi-events.exp: Ues mi_execute_to instead of
mi_send_resuming_command. Print stop_handler_str and
cont_handler_str instead of expecting the python events print
directly.
Complement commit 44d3da233815 ("MIPS/GAS: Treat local jump relocs the
same no matter if REL or RELA") and update and clarify the comment on
jump reloc conversion.
gas/
* config/tc-mips.c (mips_fix_adjustable): Update comment on jump
reloc conversion.
Includes DT_MIPS_RLD_MAP and DT_MIPS_RLD_MAP_REL dynamic tags and
__RLD_MAP symbol.
2016-06-20 Vladimir Radosavljevic <Vladimir.Radosavljevic@imgtec.com>
elfcpp/
* elfcpp.h (DT_MIPS_RLD_MAP_REL): New enum constant.
gold/
* mips.cc (Target_mips::Target_mips): Initialize rld_map_.
(Target_mips::rld_map_): New data member.
(Target_mips::do_finalize_sections): Add support for
DT_MIPS_RLD_MAP and DT_MIPS_RLD_MAP_REL dynamic tags,
.rld_map section, and __RLD_MAP symbol.
(Target_mips::do_dynamic_tag_custom_value): Add support for
DT_MIPS_RLD_MAP_REL dynamic tag.
* output.cc (Output_data_dynamic::get_entry_offset): New method
definition.
* output.h (Output_data_dynamic::get_entry_offset): New method
declaration.
elfcpp/
* mips.h (R_MIPS_PC21_S2, R_MIPS_PC26_S2, R_MIPS_PC18_S3,
R_MIPS_PC19_S2, R_MIPS_PCHI16, R_MIPS_PCLO16): New enums for
Mips32r6 and Mips64r6 relocations.
(r6_isa): New function.
gold/
* mips.cc (relocation_needs_la25_stub): Add support for relocs:
R_MIPS_PC21_S2 and R_MIPS_PC26_S2.
(hi16_reloc): Add support for R_MIPS_PCHI16 relocation.
(is_matching_lo16_reloc): Likewise.
(lo16_reloc): Add support for R_MIPS_PCLO16 relocation.
(Mips_output_data_plt::plt_entry_r6): New static data member for
R6 PLT entry.
(Target_mips::is_output_r6): New method.
(Target_mips::Mips_mach): Add new enum constants.
(Mips_relocate_functions::Status): Likewise.
(Mips_relocate_functions::pchi16_relocs): New static data member.
(Mips_relocate_functions::relpc21): New method.
(Mips_relocate_functions::relpc26): Likewise.
(Mips_relocate_functions::relpc18): Likewise.
(Mips_relocate_functions::relpc19): Likewise.
(Mips_relocate_functions::relpchi16): Likewise.
(Mips_relocate_functions::do_relpchi16): Likewise.
(Mips_relocate_functions::relpclo16): Likewise.
(Mips_output_data_plt::do_write): Add support for Mips r6 plt
entry.
(Target_mips::mips_32bit_flags): Add E_MIPS_ARCH_32R6 support.
(Target_mips::elf_mips_mach): Add E_MIPS_ARCH_32R6 and
E_MIPS_ARCH_64R6 support.
(Target_mips::update_abiflags_isa): Likewise.
(mips_get_size_for_reloc): Add support for relocs: R_MIPS_PCHI16,
R_MIPS_PCLO16, R_MIPS_PC21_S2, R_MIPS_PC26_S2, R_MIPS_PC18_S3 and
R_MIPS_PC19_S2.
(Target_mips::Scan::local): Add support for relocs: R_MIPS_PCHI16
and R_MIPS_PCLO16.
(Target_mips::Scan::global): Add support for relocs:
R_MIPS_PC21_S2 and R_MIPS_PC26_S2.
(Target_mips::Relocate::relocate): Call functions for resolving
Mips32r6 and Mips64r6 relocations, and print error message for
STATUS_PCREL_UNALIGNED.
(Target_mips::Scan::get_reference_flags): Add support for relocs:
R_MIPS_PCHI16, R_MIPS_PCLO16, R_MIPS_PC21_S2, R_MIPS_PC26_S2,
R_MIPS_PC18_S3 and R_MIPS_PC19_S2.
(Target_mips::elf_mips_mach_name): Add E_MIPS_ARCH_32R6 and
E_MIPS_ARCH_64R6 support.
H.J. Lu [Mon, 20 Jun 2016 12:10:29 +0000 (05:10 -0700)]
Use the IR symbol table for the IR input object
ELF linker shouldn't skip the IR object when searching the symbol table
of an archive element. If linker doesn't know if the object file is an
IR object, it should give LTO plugin a chance to get the correct symbol
table and use the IR symbol table if the input is an IR object.
bfd/
PR ld/18250
PR ld/20267
* elflink.c: Include plugin.h if BFD_SUPPORTS_PLUGINS is
defined.
(elf_link_is_defined_archive_symbol): Call
bfd_link_plugin_object_p on unknown plugin object and use the
IR symbol table if the input is an IR object.
* plugin.c (bfd_link_plugin_object_p): New function.
* plugin.h (bfd_link_plugin_object_p): New prototype.
ld/
PR ld/20267
* testsuite/ld-plugin/lto.exp (lto_link_tests): Add test for
PR ld/20267.
(lto_run_tests): Likewise.
* testsuite/ld-plugin/pr20267a.c: New file.
* testsuite/ld-plugin/pr20267b.c: Likewise.
H.J. Lu [Sat, 18 Jun 2016 16:16:52 +0000 (09:16 -0700)]
Don't generate PLT for IFUNC GOT/pointer reference
If a backend supports it, PLT entry isn't needed when all references
to a STT_GNU_IFUNC symbols are done via GOT or static function pointers.
For GOT entries, We generate dynamic R_*_GLOB_DAT relocations for
preemptable symbols and R_*_IRELATIVE relocations for non-preemptable
symbols to update them with real function address. For static pointer
pointers, we generate dynamic pointer relocations and store them in:
1. .rel[a].ifunc section in PIC object.
2. .rel[a].got section in dynamic executable.
3. .rel[a].iplt section in static executable.
We don't allocate GOT entry if it isn't used.
bfd/
PR ld/20253
* elf-bfd.h (_bfd_elf_allocate_ifunc_dyn_relocs): Add an
bfd_boolean argument.
* elf-ifunc.c (_bfd_elf_create_ifunc_sections): Replace
"shared object" with "PIC object" in comments.
(_bfd_elf_allocate_ifunc_dyn_relocs): Updated. Replace
"shared object" with "PIC object" in comments. Avoid PLT if
requested. Generate dynamic relocations for non-GOT references.
Make room for the special first entry in PLT and allocate PLT
entry only for PLT and PC-relative references. Store dynamic
GOT relocations in .rel[a].iplt section for static executables.
If PLT isn't used, always use GOT for symbol value. Don't
allocate GOT entry if it isn't used.
* elf32-i386.c (elf_i386_check_relocs): Increment PLT reference
count only in the code section. Allocate dynamic pointer
relocation against STT_GNU_IFUNC symbol in the non-code section.
(elf_i386_adjust_dynamic_symbol): Increment PLT reference count
only for PC-relative references.
(elf_i386_allocate_dynrelocs): Pass TRUE to
_bfd_elf_allocate_ifunc_dyn_relocs.
(elf_i386_relocate_section): Allow R_386_GOT32/R_386_GOT32X
relocations against STT_GNU_IFUNC symbols without PLT. Generate
dynamic pointer relocation against STT_GNU_IFUNC symbol in
the non-code section and store it in the proper REL section.
Don't allow non-pointer relocation against STT_GNU_IFUNC symbol
without PLT.
(elf_i386_finish_dynamic_symbol): Generate dynamic
R_386_IRELATIVE and R_386_GLOB_DAT GOT relocations against
STT_GNU_IFUNC symbols without PLT.
(elf_i386_finish_dynamic_sections): Don't handle local
STT_GNU_IFUNC symbols here.
(elf_i386_output_arch_local_syms): Handle local STT_GNU_IFUNC
symbols here.
(elf_backend_output_arch_local_syms): New.
* elf32-x86-64.c (elf_i386_check_relocs): Increment PLT reference
count only in the code section. Allocate dynamic pointer
relocation against STT_GNU_IFUNC symbol in the non-code section.
(elf_x86_64_adjust_dynamic_symbol): Increment PLT reference
count only for PC-relative references.
(elf_x86_64_allocate_dynrelocs): Pass TRUE to
_bfd_elf_allocate_ifunc_dyn_relocs.
(elf_x86_64_relocate_section): Allow R_X86_64_GOTPCREL,
R_X86_64_GOTPCRELX, R_X86_64_REX_GOTPCRELX and
R_X86_64_GOTPCREL64 relocations against STT_GNU_IFUNC symbols
without PLT. Generate dynamic pointer relocation against
STT_GNU_IFUNC symbol in the non-code section and store it in
the proper RELA section. Don't allow non-pointer relocation
against STT_GNU_IFUNC symbol without PLT.
(elf_x86_64_finish_dynamic_symbol): Generate dynamic
R_X86_64_IRELATIVE and R_X86_64_GLOB_DAT GOT relocations against
STT_GNU_IFUNC symbols without PLT.
(elf_x86_64_finish_dynamic_sections): Don't handle local
STT_GNU_IFUNC symbols here.
(elf_x86_64_output_arch_local_syms): Handle local STT_GNU_IFUNC
symbols here.
(elf_backend_output_arch_local_syms): New.
* elfnn-aarch64.c (elfNN_aarch64_allocate_ifunc_dynrelocs):
Pass FALSE to _bfd_elf_allocate_ifunc_dyn_relocs.
Sanjoy Das [Fri, 17 Jun 2016 18:24:08 +0000 (19:24 +0100)]
Add a test case for the jit-reader interface
Originally intended to be committed on 2013-01-17 in 675921c059dbaddd02ab2eb8a1eaf77b3ac727dd (Test case for the
jit-reader), but by mistake the files were not added. Fortunately
they still work.
gdb/testsuite/ChangeLog:
2016-06-17 Sanjoy Das <sanjoy@playingwithpointers.com>
* gdb.base/jit-reader.exp: New file.
* gdb.base/jithost.c: New file.
* gdb.base/jithost.h: New file.
* gdb.base/jitreader.c : New file.
* gdb.base/jit-protocol.h: New file.
2016-06-17 Thomas Preud'homme <thomas.preudhomme@arm.com>
Tony Wang <tony.wang@arm.com>
bfd/
* elf32-arm.c (elf32_arm_stub_long_branch_thumb2_only): Define stub
sequence.
(stub_long_branch_thumb2_only): Define stub.
(arm_stub_is_thumb): Add case for arm_stub_long_branch_thumb2_only.
(arm_stub_long_branch_thumb2_only): Likewise.
(arm_type_of_stub): Use arm_stub_long_branch_thumb2_only for Thumb-2
capable targets.
ld/
* testsuite/ld-arm/arm-elf.exp (Thumb-Thumb farcall M profile):
Assemble for ARMv6-M.
(Thumb2-Thumb2 farcall M profile): New testcase.
* testsuite/ld-arm/farcall-thumb2-thumb2-m.d: New file.
* testsuite/ld-arm/jump-reloc-veneers-cond-long-backward.d: Update to
reflect the use of Thumb-2 veneers for Thumb-2 capable targets.
* testsuite/ld-arm/jump-reloc-veneers-cond-long.d: Likewise.
Yao Qi [Fri, 17 Jun 2016 09:25:13 +0000 (10:25 +0100)]
Extend step-over-syscall.exp with different detach-on-fork and follow-fork modes
This patch extends step-over-syscall.exp by setting different values to
detach-on-fork and follow-fork.
gdb/testsuite:
2016-06-17 Yao Qi <yao.qi@linaro.org>
* gdb.base/step-over-syscall.exp (break_cond_on_syscall): New
parameters follow_fork and detach_on_fork. Set follow-fork-mode
and detach-on-fork. Adjust tests.
(top level): Invoke break_cond_on_syscall with combinations of
syscall, follow-fork-mode and detach-on-fork.
Yao Qi [Fri, 17 Jun 2016 09:25:13 +0000 (10:25 +0100)]
Handle reinsert breakpoints for vforked child
When a thread is doing step-over with reinsert breakpoint, and the
instruction executed is a syscall doing vfork, both parent and child
share the memory, so the reinsert breakpoint in the space is visible
to both of them. Also, removing the reinsert breakpoints from the
child will effectively remove them from the parent. We should
carefully manipulate reinsert breakpoints for both processes.
What we are doing here is that
- uninsert reinsert breakpoints from the parent before cloning the
breakpoint list. We use "uninsert" instead of "remove", because
we need to "reinsert" them back after vfork is done. In fact,
"uninsert" removes them from both child and parent process space.
- reinsert breakpoints in parent process are still copied to child's
breakpoint list,
- remove them from child's breakpoint list as what we did for fork,
at this point, reinsert breakpoints are removed from the child and
the parent, but they are still tracked by the parent's breakpoint
list,
- once vfork is done, "reinsert" them back to the parent,
gdb/gdbserver:
2016-06-17 Yao Qi <yao.qi@linaro.org>
* linux-low.c (handle_extended_wait): Call
uninsert_reinsert_breakpoints for the parent process. Remove
reinsert breakpoints from the child process. Reinsert them to
the parent process when vfork is done.
* mem-break.c (uninsert_reinsert_breakpoints): New function.
(reinsert_reinsert_breakpoints): New function.
* mem-break.h (uninsert_reinsert_breakpoints): Declare
(reinsert_reinsert_breakpoints): Declare.
Yao Qi [Fri, 17 Jun 2016 09:25:13 +0000 (10:25 +0100)]
Delete reinsert breakpoints from forked child
When a thread is stepping over a syscall instruction with software
single step, GDBserver inserts reinsert breakpoints at the next pcs.
If the syscall call is fork, the forked child has reinsert breakpoint
in its space, and GDBserver clones parent's breakpoint list to child's.
When GDBserver resumes the child, its bp_reinsert is zero, but has
reinsert breakpoints, so the following assert is triggered if I apply
the patch extending step-over-syscall.exp.
gdb/gdbserver/linux-low.c:4292: A problem internal to GDBserver has been detected.^M
void linux_resume_one_lwp_throw(lwp_info*, int, int, siginfo_t*): Assertion `!has_reinsert_breakpoints (proc)' failed.
gdb/gdbserver:
2016-06-17 Yao Qi <yao.qi@linaro.org>
* linux-low.c (handle_extended_wait): If the parent is doing
step-over, remove the reinsert breakpoints from the forked child.
Yao Qi [Fri, 17 Jun 2016 09:25:12 +0000 (10:25 +0100)]
Step over exit with reinsert breakpoints
This patch fixes a GDBserver crash when one thread is stepping over
a syscall instruction which is exit. Step-over isn't finished due
to the exit, but GDBserver doesn't clean up the state of step-over,
so in the wait next time, GDBserver will wait on step_over_bkpt,
which is already exited, and GDBserver crashes because
'requested_child' is NULL. See gdbserver logs below,
Need step over [LWP 14858]? yes, found breakpoint at 0x2aaaaad91307^M
proceed_all_lwps: found thread 14858 needing a step-over^M
Starting step-over on LWP 14858. Stopping all threads^M
>>>> entering void stop_all_lwps(int, lwp_info*)
....
<<<< exiting void stop_all_lwps(int, lwp_info*)^M
Done stopping all threads for step-over.^M
pc is 0x2aaaaad91307^M
Writing 0f to 0x2aaaaad91307 in process 14858^M
Could not find fast tracepoint jump at 0x2aaaaad91307 in list (uninserting).^M
pending reinsert at 0x2aaaaad91307^M
step from pc 0x2aaaaad91307^M
Resuming lwp 14858 (step, signal 0, stop not expected)^M
# but step_over_bkpt is still LWP 14858, which is wrong
The fix is to finish step-over if it is ongoing, and unsuspend other
threads. Without the fix in linux-low.c, GDBserver will crash in
with running gdb.base/step-over-exit.exp.
gdb/gdbserver:
2016-06-17 Yao Qi <yao.qi@linaro.org>
* linux-low.c (unsuspend_all_lwps): Declare.
(linux_low_filter_event): If thread exited, call finish_step_over.
If step-over is finished, unsuspend other threads.
Yao Qi [Fri, 17 Jun 2016 09:25:12 +0000 (10:25 +0100)]
More assert checks on reinsert breakpoint
This patch adds more asserts, so the incorrect or sub-optimal
reinsert breakpoints manipulations (from the tests in the following
patches) can trigger them.
gdb/gdbserver:
2016-06-17 Yao Qi <yao.qi@linaro.org>
* linux-low.c (linux_resume_one_lwp_throw): Assert
has_reinsert_breakpoints returns false.
* mem-break.c (delete_disabled_breakpoints): Assert
bp type isn't reinsert_breakpoint.
Yao Qi [Fri, 17 Jun 2016 09:25:12 +0000 (10:25 +0100)]
Switch to current thread in finish_step_over
This patch adds some sanity check that reinsert breakpoints must be
there when doing step-over on software single step target. The check
triggers an assert when running forking-threads-plus-breakpoint.exp
on arm-linux target,
gdb/gdbserver/linux-low.c:4714: A problem internal to GDBserver has been detected.^M
int finish_step_over(lwp_info*): Assertion `has_reinsert_breakpoints ()' failed.
the error happens when GDBserver has already resumed a thread of
process A for step-over (and wait for it hitting reinsert breakpoint),
but receives detach request for process B from GDB, which is shown in
the backtrace below,
(gdb) bt
#2 0x000228aa in finish_step_over (lwp=0x12bbd98) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/linux-low.c:4703
#3 0x00025a50 in finish_step_over (lwp=0x12bbd98) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/linux-low.c:4749
#4 complete_ongoing_step_over () at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/linux-low.c:4760
#5 linux_detach (pid=25228) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/linux-low.c:1503
#6 0x00012bae in process_serial_event () at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/server.c:3974
#7 handle_serial_event (err=<optimized out>, client_data=<optimized out>) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/server.c:4347
#8 0x00016d68 in handle_file_event (event_file_desc=<optimized out>) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/event-loop.c:429
#9 0x000173ea in process_event () at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/event-loop.c:184
#10 start_event_loop () at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/event-loop.c:547
#11 0x0000aa2c in captured_main (argv=<optimized out>, argc=<optimized out>) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/server.c:3719
#12 main (argc=<optimized out>, argv=<optimized out>) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/server.c:3804
the sanity check tries to find the reinsert breakpoint from process B,
but nothing is found. It is wrong, we need to search in process A,
since we started step-over of a thread of process A.
(gdb) p lwp->thread->entry.id
$3 = {pid = 25120, lwp = 25131, tid = 0}
(gdb) p current_thread->entry.id
$4 = {pid = 25228, lwp = 25228, tid = 0}
This patch switched current_thread to the thread we are doing step-over
in finish_step_over.
gdb/gdbserver:
2016-06-17 Yao Qi <yao.qi@linaro.org>
* linux-low.c (maybe_hw_step): New function.
(linux_resume_one_lwp_throw): Call maybe_hw_step.
(finish_step_over): Switch current_thread to lwp temporarily,
and assert has_reinsert_breakpoints returns true.
(proceed_one_lwp): Call maybe_hw_step.
* mem-break.c (has_reinsert_breakpoints): New function.
* mem-break.h (has_reinsert_breakpoints): Declare.
This patch fixes and expands the definition of the read/write
instructions for ancillary-state, privileged and hyperprivileged
registers in opcodes.
It also adds support for three new v9m hyperprivileged registers:
%hmcdper, %hmcddfr and %hva_mask_nz.
Finally, the patch expands existing tests (and adds several new ones) in
order to cover all the read/write instructions in all its variants.
opcodes/ChangeLog:
2016-06-17 Jose E. Marchesi <jose.marchesi@oracle.com>
* sparc-opc.c (rdasr): New macro.
(wrasr): Likewise.
(rdpr): Likewise.
(wrpr): Likewise.
(rdhpr): Likewise.
(wrhpr): Likewise.
(sparc_opcodes): Use the macros above to fix and expand the
definition of read/write instructions from/to
asr/privileged/hyperprivileged instructions.
* sparc-dis.c (v9_hpriv_reg_names): Add %hmcdper, %hmcddfr and
%hva_mask_nz. Prefer softint_set and softint_clear over
set_softint and clear_softint.
(print_insn_sparc): Support %ver in Rd.
gas/ChangeLog:
2016-06-17 Jose E. Marchesi <jose.marchesi@oracle.com>
* config/tc-sparc.c (hpriv_reg_table): Add registers %hmcdper,
%hmcddfr and %hva_mask_nz.
(sparc_ip): New handling of asr/privileged/hyperprivileged
registers, adapted to the new form of the sparc opcodes table.
* testsuite/gas/sparc/rdasr.s: New file.
* testsuite/gas/sparc/rdasr.d: Likewise.
* testsuite/gas/sparc/wrasr.s: Likewise.
* testsuite/gas/sparc/wrasr.d: Likewise.
* testsuite/gas/sparc/sparc.exp (sparc_elf_setup): Add rdasr and
wrasr tests.
* testsuite/gas/sparc/rdpr.d: Use -Av9m, as some privileged
registers require it.
* testsuite/gas/sparc/wrpr.s: Complete to cover all privileged
registers and write instruction modalities.
* testsuite/gas/sparc/wrpr.d: Likewise.
* testsuite/gas/sparc/rdhpr.s: Likewise for hyperprivileged
registers.
* testsuite/gas/sparc/rdhpr.d: Likewise.
* testsuite/gas/sparc/wrhpr.s: Likewise.
* testsuite/gas/sparc/wrhpr.d: Likewise.
Jose E. Marchesi [Fri, 17 Jun 2016 09:14:18 +0000 (02:14 -0700)]
opcodes,gas: adjust sparc insns and make GAS aware of it
This patch marks the SPARC instructions in the opcodes table with their
proper opcode architectures, and makes the assembler aware of them.
This allows the assembler to properly realize when a new instruction
needs a higher architecture (after v9b) and to react accordingly
emitting an error message or bumping the architecture.
It also expands architecture mismatch tests to cover architectures
higher than v9b, and fixes a couple of minor bugs in the GAS testsuite.
opcodes/ChangeLog:
2016-06-17 Jose E. Marchesi <jose.marchesi@oracle.com>
* sparc-opc.c (sparc_opcodes): Adjust instructions opcode
architecture according to the hardware capabilities they require.
(sparc_priv_regs): New table.
(sparc_hpriv_regs): Likewise.
(sparc_asr_regs): Likewise.
(v9anotv9m): Define.
gas/ChangeLog:
2016-06-17 Jose E. Marchesi <jose.marchesi@oracle.com>
* config/tc-sparc.c (sparc_arch_table): adjust the GAS
architectures to use the right opcode architecture.
(sparc_md_end): Handle v9{c,d,e,v,m}.
(sparc_ip): Fix some comments.
* testsuite/gas/sparc/ldx_efsr.d: Fix the architecture of this
instruction, which is v9d.
* testsuite/gas/sparc/mwait.s: Remove the `rd %mwait,%g1'
instruction from the test, as %mwait is not readable.
* testsuite/gas/sparc/mwait.d: Likewise.
* testsuite/gas/sparc/mism-1.s: Expand to check v9b and v9e
mismatch architecture errors.
* testsuite/gas/sparc/mism-2.s: New file.
Jose E. Marchesi [Fri, 17 Jun 2016 09:13:30 +0000 (02:13 -0700)]
gas: sparc: fix collision of registers and pseudo-ops.
The current sparc assembler breaks when the name of an ancillary-state
register, privileged register or hyperprivileged register has a
%-pseudo-operation name as a prefix. For example, %hmcdper and %hm(),
or %hintp and %hi().
This patch fixes it by introducing a new table `perc_table' (for
%-table) that contains an entry for every %name supported by the
assembler, other than the general registers. This table is used to
detect name collisions when the assembler tries to detect a %-pseudo-op.
This patch also fixes a related bug, making sure that v9a_asr_table and
hpriv_reg_table are sorted in reverse lexicographic order, as otherwise
the search code may fail.
gas/ChangeLog:
2016-06-17 Jose E. Marchesi <jose.marchesi@oracle.com>
* config/tc-sparc.c (priv_reg_table): Use NULL instead of the
empty string to mark the end of the array.
(hpriv_reg_table): Likewise.
(v9a_asr_table): Likewise.
(cmp_reg_entry): Handle entries with NULL names.
(F_POP_V9): Define.
(F_POP_PCREL): Likewise.
(F_POP_TLS_CALL): Likewise.
(F_POP_POSTFIX): Likewise.
(struct pop_entry): New type.
(pop_table): New variable.
(enum pop_entry_type): New type.
(struct perc_entry): Likewise.
(NUM_PERC_ENTRIES): Define.
(perc_table): New variable.
(cmp_perc_entry): New function.
(md_begin): Sort hpriv_reg_table and v9a_asr_table, and initialize
perc_table.
(sparc_ip): Handle entries with NULL names in priv_reg_table,
hpriv_reg_table and v9a_asr_table. Use perc_table to handle
%-pseudo-ops.
Jose E. Marchesi [Fri, 17 Jun 2016 09:12:48 +0000 (02:12 -0700)]
bfd,opcodes: sparc: new opcode v9{c,d,e,v,m} architectures and bfd machine numbers.
This patch adds support for the opcode architectures
SPARC_OPCODE_ARCH_V9{C,D,E,V,M} and its associated BFD machine numbers
bfd_mach_sparc_v9{c,d,e,v,m} and bfd_mach_sparc_v8plus{c,d,e,v,m}.
Note that for arches up to v9b (UltraSPARC III), the detection of the
BFD machine type was based on the bits in the e_machine field of the ELF
header. However, there are no more available bits in that field, so
this patch takes the approach of using the hardware capabilities stored
in the object attributes HWCAPS/HWCAPS2 in order to characterize the
machine the object was built for.
bfd/ChangeLog:
2016-06-17 Jose E. Marchesi <jose.marchesi@oracle.com>
H.J. Lu [Wed, 15 Jun 2016 17:35:38 +0000 (10:35 -0700)]
Don't check undefined symbol for IFUNC reloc
Since x86 elf_*_check_relocs is called after all symbols have been
resolved, there is no need to check undefined symbols for relocations
against IFUNC symbols.
bfd/
* elf32-i386.c (elf_i386_check_relocs): Don't check undefined
symbols for relocations against IFUNC symbols.
* elf64-x86-64.c (elf_x86_64_check_relocs): Likewise.
ld/
* testsuite/ld-i386/i386.exp: Run pr19636-2e-nacl.
* testsuite/ld-i386/pr19636-2e.d: Skip for NaCl targets.
Remove .rel.plt section.
* testsuite/ld-i386/pr19636-2e-nacl.d: New file.
bfd/s390: Fix DT_PLTRELSZ in presence of R_390_IRELATIVE.
This was broken by 4ade44b727ee77adaa9c22719935d012e253a5e6,
which changed the calculation to use the .rela.plt linker section
instead of its output section - thus skipping .rela.iplt .
Fix the calculations to include it.
bfd/ChangeLog:
* elf32-s390.c (elf_s390_finish_dynamic_sections): Include
.rela.iplt in DT_PLTRELSZ.
* elf64-s390.c (elf_s390_finish_dynamic_sections): Likewise,
for DT_PLTRELSZ and DT_RELASZ as well.
H.J. Lu [Thu, 16 Jun 2016 12:52:34 +0000 (05:52 -0700)]
Skip relocations in non-loaded, non-alloced sections
Don't do anything special with non-loaded, non-alloced sections.
In particular, any relocs in such sections should not affect GOT
and PLT reference counting (ie. we don't allow them to create GOT
or PLT entries), there's no possibility or desire to optimize TLS
relocs, and there's not much point in propagating relocs to shared
libs that the dynamic linker won't relocate.
Nick Clifton [Wed, 15 Jun 2016 15:25:34 +0000 (16:25 +0100)]
Fix simple gas testsuite failures.
binutils* readelf.c (is_24bit_abs_reloc): Add support for R_FT32_20
reloc.
gas * config/tc-ft32.c (md_assemble): Call dwarf2_emit_insn with the
instruction size.
* config/tc-mcore.c (md_assemble): Likewise.
* config/tc-mn10200.c (md_assemble): Likewise.
* config/tc-moxie.c (md_assemble): Likewise.
* config/tc-pj.c (md_apply_fix): Handle BFD_RELOC_PJ_CODE_REL32.
* testsuite/gas/all/gas.exp (diff1 test): Alpha sort list of
exception targets. Add alpha, hppa, microblaze and rl78 to list
of exceptions.
(forward): Add microblaze to list of exceptions.
(fwdexp): Add alpha to list of exceptions.
(redef2): Add arm-epoc-pe and rl78 to list of exceptions.
(redef3): Add rl78 and x86_64 cygwin to list of exceptions.
(do_930509a): Alpha sort list of exception targets. Add h8300 and
mn10200 to list of exceptions.
(align2): Expect to fail for nds32.
(cond): Add alpha and rl78 to list of exceptions.
* testsuite/gas/all/none.d: Skip for ft32 and hppa.
* testsuite/gas/all/string.d: Skip for tic4x.
* testsuite/gas/alpha/alpha.exp: Note that the alpha-linuxecoff
target does not support ELF.
* testsuite/gas/arm/blx-bl-convert.dL Skip for the nto target.
* testsuite/gas/cfi/cfi-alpha-2.d: All extended format names.
* testsuite/gas/cfi/cfi.exp: Alpha sort list of targets. Skip SH
tests for sh-pe and sh-rtemscoff targets.
* testsuite/gas/elf/elf.exp (redef): Add rl78, xgate and vax to
list of exceptions.
(type): Run the noifunc version for alpha-freebsd and visium.
* testsuite/gas/elf/warn-2.s: Do not expect to fail on the mcore,
mn10200 or moxie targets.
* testsuite/gas/ft32/insn.d: Update expected disassembly.
* testsuite/gas/i386/i386.exp (x86-64-pcrel): Skip for cygwin
targets.
* testsuite/gas/lns/lns.exp (lns-common-1): No longer skip for
mcore and rx targets.
* testsuite/gas/macros/macros.exp (dot): Add exceptions for ns32k,
rl78 and vax.
(purge): Expect to fail on the ns32k and vax.
* testsuite/gas/nds32/alu-2.d: Update expected disassembly.
* testsuite/gas/nds32/ls.d: Likewise.
* testsuite/gas/nds32/sys-reg.d: Likewise.
* testsuite/gas/nds32/usr-spe-reg.d: Likewise.
* testsuite/gas/pe/aligncomm-d.d: Skip for the sh.
* testsuite/gas/pe/section-align-3.d: Likewise.
* testsuite/gas/pe/section-exclude.d: Likewise.
* testsuite/gas/ppc/test2xcoff32.d: Pass once all the required
data has been seen.
* testsuite/gas/ppc/textalign-xcoff-001.d: Fix up regexp to allow
for variations in whitespace.
* testsuite/gas/tilepro/t_constants.d: Pass once all the required
data has been seen.
* testsuite/gas/tilepro/t_constants.s (.safe_word): New macro.
Installs a 32-bit value without generating warnings on 64-bit
hosts.
Use the new macro to replace the .word directives.
opcodes * nds32-dis.c (nds32_parse_audio_ext): Change printing of integer
constants to match expected behaviour.
(nds32_parse_opcode): Likewise. Also for whitespace.
This patch fixes another edge case related to alignment property
records - reloc offsets adjacent to property record offsets were not
getting adjusted during relaxation.
bfd/
PR ld/20254
* elf32-avr.c (elf32_avr_relax_delete_bytes): Adjust reloc
offsets until reloc_toaddr.
ld/
PR ld/20254
* testsuite/ld-avr/avr-prop-6.d: New test.
* testsuite/ld-avr/avr-prop-6.s: New test.
John Baldwin [Sat, 11 Jun 2016 20:10:48 +0000 (13:10 -0700)]
Remove unneeded checks on type lengths.
Type lengths are unsigned, so they are always greater than or equal to
zero. A check against the length of 'tgt_type' is retained to prevent
dividing by zero.
gdb/ChangeLog:
* v850-tdep.c (v850_use_struct_convention): Trim type length checks.
John Baldwin [Sat, 11 Jun 2016 14:26:09 +0000 (07:26 -0700)]
Initialize 'ra' to zero to avoid uninitialized use.
If the instruction in this case does not include an RA field, then 'ra'
is used uninitialized. Use the same idiom used elsewhere in this file of
initializing ra to zero before check for an RA field.
H.J. Lu [Tue, 14 Jun 2016 17:18:26 +0000 (10:18 -0700)]
Check R_*_IRELATIVE in x86 reloc_type_class
elf_{i386|x86_64}_reloc_type_class should return reloc_class_ifunc for
R_386_IRELATIVE/R_X86_64_IRELATIVE relocations. There is no need to
check symbol type for STN_UNDEF symbol index.
* elf32-i386.c (elf_i386_reloc_type_class): Check R_386_IRELATIVE.
Don't check symbol type for STN_UNDEF symbol index.
* elf64-x86-64.c (elf_x86_64_reloc_type_class): Check
R_X86_64_IRELATIVE. Don't check symbol type for STN_UNDEF symbol
index.
Graham Markall [Mon, 13 Jun 2016 08:03:05 +0000 (09:03 +0100)]
[ARC] Add ldbit for nps
This commit adds the ldbit instruction for the NPS-400. The ldbit
instruction uses the same encoding as the ld instruction, but sets
the ZZ field to 11 (which is a reserved setting), and sets the AA
field to 1 or 2 for the x2 and x4 flags respectively.
Fix feature checks based on ARM architecture value
2016-06-14 Thomas Preud'homme <thomas.preudhomme@arm.com>
bfd/
* elf32-arm.c (using_thumb_only): Force review of arch check logic for
new architecture.
(using_thumb2): Try Tag_THUMB_ISA_use first and check
for exact arch value then. Force review of arch check logic for new
architecture.
(arch_has_arm_nop): Update and fix arch check logic. Force review of
that logic for new architecture.
(arch_has_thumb2_nop): Remove.
(elf32_arm_tls_relax): Use using_thumb2 instead of above function.
(elf32_arm_final_link_relocate): Likewise but using thumb2.
Alan Modra [Tue, 14 Jun 2016 03:42:00 +0000 (13:12 +0930)]
Set my_archive for thin archives
LTO plugin support in plugin_maybe_claim wants to close the IR bfd
after replacing it with the recompiled object, but can't do so for
archive elements due to various pointers that access the archive bfd.
Thin archives have the same problem. They too cannot have their
element bfds closed.
PR ld/20241
bfd/
* archive.c (open_nested_file): Set my_archive.
* bfd.c (_bfd_default_error_handler <%B>): Exclude archive file name
for thin archives.
* bfdio.c (bfd_tell): Don't adjust origin for thin archives.
(bfd_seek): Likewise.
* bfdwin.c (bfd_get_file_window): Likewise.
* cache.c (cache_bmmap): Likewise.
(bfd_cache_lookup_worker): Don't look in my_archive for thin archives.
* mach-o.c (bfd_mach_o_follow_dsym): Don't open my_archive for
thin archives.
* plugin.c (try_claim): Likewise.
* xcofflink.c (xcoff_link_add_dynamic_symbols): Use import path of
file within thin archive, not the archive.
binutils/
* bucomm.c (bfd_get_archive_filename): Return file name within thin
archive.
ld/
* ldmain.c (add_archive_element): Just print file name of file within
thin archives.
* ldmisc.c (vfinfo): Likewise.
* plugin.c (plugin_object_p): Open file within thin archives.
(plugin_maybe_claim): Expand comment.
projects / thirdparty / binutils-gdb.git / log
