1999-05-12 04:29:07 +08:00
|
|
|
|
/* Top level stuff for GDB, the GNU debugger.
|
2005-01-18 00:17:36 +08:00
|
|
|
|
|
2017-01-01 14:50:51 +08:00
|
|
|
|
Copyright (C) 1999-2017 Free Software Foundation, Inc.
|
2005-01-18 00:17:36 +08:00
|
|
|
|
|
1999-05-12 04:29:07 +08:00
|
|
|
|
Written by Elena Zannoni <ezannoni@cygnus.com> of Cygnus Solutions.
|
|
|
|
|
|
|
|
|
|
This file is part of GDB.
|
|
|
|
|
|
|
|
|
|
This program is free software; you can redistribute it and/or modify
|
|
|
|
|
it under the terms of the GNU General Public License as published by
|
2007-08-24 02:08:50 +08:00
|
|
|
|
the Free Software Foundation; either version 3 of the License, or
|
1999-05-12 04:29:07 +08:00
|
|
|
|
(at your option) any later version.
|
|
|
|
|
|
|
|
|
|
This program is distributed in the hope that it will be useful,
|
|
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
|
GNU General Public License for more details.
|
|
|
|
|
|
|
|
|
|
You should have received a copy of the GNU General Public License
|
2010-12-29 08:58:14 +08:00
|
|
|
|
along with this program. If not, see <http://www.gnu.org/licenses/>. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
#include "defs.h"
|
1999-06-15 02:08:47 +08:00
|
|
|
|
#include "top.h"
|
1999-05-12 04:29:07 +08:00
|
|
|
|
#include "inferior.h"
|
Add new infrun.h header.
Move infrun.c declarations out of inferior.h to a new infrun.h file.
Tested by building on:
i686-w64-mingw32, enable-targets=all
x86_64-linux, enable-targets=all
i586-pc-msdosdjgpp
And also grepped the whole tree for each symbol moved to find where
infrun.h might be necessary.
gdb/
2014-05-22 Pedro Alves <palves@redhat.com>
* inferior.h (debug_infrun, debug_displaced, stop_on_solib_events)
(sync_execution, sched_multi, step_stop_if_no_debug, non_stop)
(disable_randomization, enum exec_direction_kind)
(execution_direction, stop_registers, start_remote)
(clear_proceed_status, proceed, resume, user_visible_resume_ptid)
(wait_for_inferior, normal_stop, get_last_target_status)
(prepare_for_detach, fetch_inferior_event, init_wait_for_inferior)
(insert_step_resume_breakpoint_at_sal)
(follow_inferior_reset_breakpoints, stepping_past_instruction_at)
(set_step_info, print_stop_event, signal_stop_state)
(signal_print_state, signal_pass_state, signal_stop_update)
(signal_print_update, signal_pass_update)
(update_signals_program_target, clear_exit_convenience_vars)
(displaced_step_dump_bytes, update_observer_mode)
(signal_catch_update, gdb_signal_from_command): Move
declarations ...
* infrun.h: ... to this new file.
* amd64-tdep.c: Include infrun.h.
* annotate.c: Include infrun.h.
* arch-utils.c: Include infrun.h.
* arm-linux-tdep.c: Include infrun.h.
* arm-tdep.c: Include infrun.h.
* break-catch-sig.c: Include infrun.h.
* breakpoint.c: Include infrun.h.
* common/agent.c: Include infrun.h instead of inferior.h.
* corelow.c: Include infrun.h.
* event-top.c: Include infrun.h.
* go32-nat.c: Include infrun.h.
* i386-tdep.c: Include infrun.h.
* inf-loop.c: Include infrun.h.
* infcall.c: Include infrun.h.
* infcmd.c: Include infrun.h.
* infrun.c: Include infrun.h.
* linux-fork.c: Include infrun.h.
* linux-nat.c: Include infrun.h.
* linux-thread-db.c: Include infrun.h.
* monitor.c: Include infrun.h.
* nto-tdep.c: Include infrun.h.
* procfs.c: Include infrun.h.
* record-btrace.c: Include infrun.h.
* record-full.c: Include infrun.h.
* remote-m32r-sdi.c: Include infrun.h.
* remote-mips.c: Include infrun.h.
* remote-notif.c: Include infrun.h.
* remote-sim.c: Include infrun.h.
* remote.c: Include infrun.h.
* reverse.c: Include infrun.h.
* rs6000-tdep.c: Include infrun.h.
* s390-linux-tdep.c: Include infrun.h.
* solib-irix.c: Include infrun.h.
* solib-osf.c: Include infrun.h.
* solib-svr4.c: Include infrun.h.
* target.c: Include infrun.h.
* top.c: Include infrun.h.
* windows-nat.c: Include infrun.h.
* mi/mi-interp.c: Include infrun.h.
* mi/mi-main.c: Include infrun.h.
* python/py-threadevent.c: Include infrun.h.
2014-05-22 19:29:11 +08:00
|
|
|
|
#include "infrun.h"
|
1999-10-26 11:43:48 +08:00
|
|
|
|
#include "target.h"
|
1999-07-08 04:19:36 +08:00
|
|
|
|
#include "terminal.h" /* for job_control */
|
1999-06-21 21:27:42 +08:00
|
|
|
|
#include "event-loop.h"
|
1999-09-22 11:28:34 +08:00
|
|
|
|
#include "event-top.h"
|
2003-02-05 Jim Ingham <jingham@apple.com>
Keith Seitz <keiths@redhat.com>
Elena Zannoni <ezannoni@redhat.com>
Andrew Cagney <ac131313@redhat.com>
* Makefile.in (SUBDIR_CLI_OBS): Add "cli-interp.o".
(SUBDIR_CLI_SRCS): Add "cli/cli-interp.c".
(SUBDIR_MI_OBS): Add "mi-interp.o".
(SUBDIR_MI_SRCS): Add "mi/mi-interp.c".
(SFILES): Add "interps.c".
(COMMON_OBS): Add "interps.o".
(interps_h, mi_main_h): Define.
(interps.o, cli-interp.o, mi-interp.o): Add dependencies.
(mi-main.o, main.o, event-top.o): Update dependencies.
* cli/cli-interp.c: New file.
* interps.h, interps.c: New files.
* top.c: (gdb_init): Don't install the default interpreter, handed
by captured_main.
* main.c: Include "interps.h".
(interpreter_p): Note that it should malloc'ed.
(captured_command_loop): Call current_interp_command_loop.
(captured_main): Initialize interpreter_p to INTERP_CONSOLE. Use
xfree and xstrdup when updating interpreter_p. Install the
default interpreter. Add hack to stop mi1's copyright notice
being encoded.
* event-top.h (gdb_setup_readline): Declare.
(gdb_disable_readline): Declare.
* event-top.c: Include "interps.h".
(display_gdb_prompt): Call current_interp_display_prompt_p.
(gdb_setup_readline): Initialize gdb_stdout, gdb_stderr,
gdb_stdlog, and gdb_stdtarg.
(_initialize_event_loop): Don't call gdb_setup_readline.
* cli-out.c (cli_out_set_stream): New function.
* cli-out.h (cli_out_set_stream): Declare.
2003-02-06 09:19:12 +08:00
|
|
|
|
#include "interps.h"
|
2001-02-06 12:17:03 +08:00
|
|
|
|
#include <signal.h>
|
2006-07-21 22:46:56 +08:00
|
|
|
|
#include "cli/cli-script.h" /* for reset_command_nest_depth */
|
2008-05-02 21:58:38 +08:00
|
|
|
|
#include "main.h"
|
2008-07-10 06:16:15 +08:00
|
|
|
|
#include "gdbthread.h"
|
2011-07-21 19:03:48 +08:00
|
|
|
|
#include "observer.h"
|
2011-05-27 Pedro Alves <pedro@codesourcery.com>
* defs.h (struct continuation, continuation_ftype)
(continuation_free_arg_ftype, add_continuation)
(do_all_continuations, do_all_continuations_thread)
(discard_all_continuations, discard_all_continuations_thread)
(add_intermediate_continuation, do_all_intermediate_continuations)
(do_all_intermediate_continuations_thread)
(discard_all_intermediate_continuations)
(discard_all_intermediate_continuations_thread)
(add_inferior_continuation, do_all_inferior_continuations)
(discard_all_inferior_continuations): Move to ...
* continuations.h: ... this new file.
* breakpoint.c, continuations.c, event-top.c, inf-loop.c,
infcmd.c, inferior.c, infrun.c, interps.c: Include
continuations.h.
2011-05-28 02:28:18 +08:00
|
|
|
|
#include "continuations.h"
|
2010-12-29 08:58:14 +08:00
|
|
|
|
#include "gdbcmd.h" /* for dont_repeat() */
|
With some changes to how software single-step (SSS) breakpoints are
handled, one of those being to place SSS breakpoints on the breakpoint
chain as all other breakpoints, annota1.exp times out with lots and
lots of breakpoint-invalid and frame-changed annotations. All those
extra annotations are actually unnecessary. For one, SSS breakpoints
are internal breakpoints, so the frontend shouldn't care if they were
added, removed or changed. Then, there's really no point in emitting
"breakpoints-invalid" or "frames-invalid" more than once between times
the frontend/user can actually issues GDB commands; the frontend will
have to wait for the GDB prompt to refresh its state, so emitting
those annotations at most once between prompts is enough. Non-stop or
async would complicate this, but no frontend will be using annotations
in those modes (one of goes of emacs switching to MI was non-stop mode
support, AFAIK). The previous patch reveals there has been an
intention in the past to suppress multiple breakpoints-invalid
annotations caused by ignore count changes. As the previous patch
shows, that's always been broken, but in any case, this patch actually
makes it work. The next patch will remove several annotation-specific
calls in breakpoint.c in favor of always using the breakpoint modified
& friends observers, and that causes yet more of these annotations,
because several calls to the corresponding annotate_* functions in
breakpoint.c are missing, particularly in newer code.
So all in all, here's a simple mechanism that avoids sending the same
annotation to the frontend more than once until gdb is ready to accept
further commands.
Tested on x86_64 Fedora 17.
2013-01-22 Pedro Alves <palves@redhat.com>
* annotate.c: Include "inferior.h".
(frames_invalid_emitted)
(breakpoints_invalid_emitted): New globals.
(async_background_execution_p): New function.
(annotate_breakpoints_changed, annotate_frames_invalid): Skip
emitting the annotation if it has already been emitted.
(annotate_display_prompt): New function.
* annotate.h (annotate_display_prompt): New declaration.
* event-top.c: Include annotate.h.
(display_gdb_prompt): Call annotate_display_prompt.
2013-01-23 04:17:10 +08:00
|
|
|
|
#include "annotate.h"
|
New commands "mt set per-command {space,time,symtab} {on,off}".
* NEWS: Add entry.
* event-top.c: #include "maint.h".
* main.c: #include "maint.h".
* maint.c: #include <sys/time.h>, <time.h>, block.h, top.h,
timeval-utils.h, maint.h, cli/cli-setshow.h.
(per_command_time, per_command_space): New static globals.
(per_command_symtab): New static global.
(per_command_setlist, per_command_showlist): New static globals.
(struct cmd_stats): Move here from utils.c.
(set_per_command_time): Renamed from set_display_time in utils.c
and moved here. All callers updated.
(set_per_command_space): Renamed from set_display_space in utils.c
and moved here. All callers updated.
(count_symtabs_and_blocks): New function.
(report_command_stats): Moved here from utils.c. Add support for
printing symtab stats. Only print data if enabled before command
executed.
(make_command_stats_cleanup): Ditto.
(sert_per_command_cmd, show_per_command_cmd): New functions.
(_initialize_maint_cmds): Add new commands
mt set per-command {space,time,symtab} {on,off}.
* maint.h: New file.
* top.c: #include "maint.h".
* utils.c (reset_prompt_for_continue_wait_time): New function.
(get_prompt_for_continue_wait_time): New function.
* utils.h (reset_prompt_for_continue_wait_time): Declare
(get_prompt_for_continue_wait_time): Declare.
(make_command_stats_cleanup): Moved to maint.h.
(set_display_time, set_display_space): Moved to maint.h and renamed
to set_per_command_time, set_per_command_space.
* cli/cli-setshow.c (parse_cli_boolean_value): Renamed from
parse_binary_operation and made non-static. Don't call error,
just return an error marker. All callers updated.
* cli/cli-setshow.h (parse_cli_boolean_value): Declare.
doc/
* gdb.texinfo (Maintenance Commands): Add docs for
"mt set per-command {space,time,symtab} {on,off}".
testsuite/
* gdb.base/maint.exp: Update tests for per-command stats.
2013-03-22 01:37:30 +08:00
|
|
|
|
#include "maint.h"
|
2016-03-10 02:25:00 +08:00
|
|
|
|
#include "buffer.h"
|
Introduce interruptible_select
We have places where we call a blocking gdb_select expecting that a
Ctrl-C will unblock it. However, if the Ctrl-C is pressed just before
gdb_select, the SIGINT handler runs before gdb_select, and thus
gdb_select won't return.
For example gdb_readline_no_editing:
QUIT;
/* Wait until at least one byte of data is available. Control-C
can interrupt gdb_select, but not fgetc. */
FD_ZERO (&readfds);
FD_SET (fd, &readfds);
if (gdb_select (fd + 1, &readfds, NULL, NULL, NULL) == -1)
and stdio_file_read:
/* For the benefit of Windows, call gdb_select before reading from
the file. Wait until at least one byte of data is available.
Control-C can interrupt gdb_select, but not read. */
{
fd_set readfds;
FD_ZERO (&readfds);
FD_SET (stdio->fd, &readfds);
if (gdb_select (stdio->fd + 1, &readfds, NULL, NULL, NULL) == -1)
return -1;
}
return read (stdio->fd, buf, length_buf);
This is a race classically fixed with either the self-pipe trick, or
by blocking SIGINT and then using pselect instead of select.
Blocking SIGINT most of the time would mean that check_quit_flag (and
thus QUIT) would need to do a syscall every time it is called, which
sounds best avoided, since QUIT is called in many loops. Thus we take
the self-pipe trick route (wrapped in a serial event).
Instead of having all places that need this manually add an extra file
descriptor to the set of gdb_select's watched file descriptors, we
introduce a wrapper, interruptible_select, that does that.
The Windows version of gdb_select actually does not suffer from this,
because mingw-hdep.c:gdb_call_async_signal_handler sets a Windows
event that gdb_select always waits on. So this patch can be seen as
generalization of that technique. We can't remove that extra event
from mingw-hdep.c until we get rid of immediate_quit though.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h: Extend QUIT-related comments to mention
interruptible_select.
(quit_serial_event_set, quit_serial_event_clear): Declare.
* event-top.c: Include "ser-event.h" and "gdb_select.h".
(quit_serial_event): New global.
(async_init_signals): Make quit_serial_event.
(quit_serial_event_set, quit_serial_event_clear)
(quit_serial_event_fd, interruptible_select): New functions.
* extension.c (set_quit_flag): Set the quit serial event.
(check_quit_flag): Clear the quit serial event.
* gdb_select.h (interruptible_select): New declaration.
* guile/scm-ports.c (ioscm_input_waiting): Use
interruptible_select instead of gdb_select.
* top.c (gdb_readline_no_editing): Likewise.
* ui-file.c (stdio_file_read): Likewise.
2016-04-12 23:49:30 +08:00
|
|
|
|
#include "ser-event.h"
|
|
|
|
|
#include "gdb_select.h"
|
1999-08-31 09:14:27 +08:00
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* readline include files. */
|
2004-02-28 Andrew Cagney <cagney@redhat.com>
* utils.c: Use "", instead of <>, to include readline.
tui/tui-win.c, tui/tui.c, tui/tui-hooks.c: Ditto.
* tracepoint.c, top.c, symmisc.c, symfile.c: Ditto.
* source.c, solib.c, exec.c, event-top.c: Ditto.
* corelow.c, completer.c, cli/cli-setshow.c: Ditto.
* cli/cli-dump.c, cli/cli-cmds.c: Ditto.
* Makefile.in: Update all dependencies.
(readline_tilde_h, readline_history_h): Define.
(readline_headers): Delete.
2004-02-29 02:04:37 +08:00
|
|
|
|
#include "readline/readline.h"
|
|
|
|
|
#include "readline/history.h"
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
/* readline defines this. */
|
|
|
|
|
#undef savestring
|
|
|
|
|
|
2017-09-28 10:30:19 +08:00
|
|
|
|
static std::string top_level_prompt ();
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Signal handlers. */
|
2005-03-17 01:05:31 +08:00
|
|
|
|
#ifdef SIGQUIT
|
1999-09-22 11:28:34 +08:00
|
|
|
|
static void handle_sigquit (int sig);
|
2005-03-17 01:05:31 +08:00
|
|
|
|
#endif
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#ifdef SIGHUP
|
1999-09-22 11:28:34 +08:00
|
|
|
|
static void handle_sighup (int sig);
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#endif
|
1999-09-22 11:28:34 +08:00
|
|
|
|
static void handle_sigfpe (int sig);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
/* Functions to be invoked by the event loop in response to
|
2010-12-29 08:58:14 +08:00
|
|
|
|
signals. */
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#if defined (SIGQUIT) || defined (SIGHUP)
|
1999-09-22 11:28:34 +08:00
|
|
|
|
static void async_do_nothing (gdb_client_data);
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#endif
|
|
|
|
|
#ifdef SIGHUP
|
1999-09-22 11:28:34 +08:00
|
|
|
|
static void async_disconnect (gdb_client_data);
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#endif
|
1999-09-22 11:28:34 +08:00
|
|
|
|
static void async_float_handler (gdb_client_data);
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#ifdef STOP_SIGNAL
|
1999-09-22 11:28:34 +08:00
|
|
|
|
static void async_stop_sig (gdb_client_data);
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#endif
|
2014-03-19 05:48:06 +08:00
|
|
|
|
static void async_sigterm_handler (gdb_client_data arg);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
/* Instead of invoking (and waiting for) readline to read the command
|
|
|
|
|
line and pass it back for processing, we use readline's alternate
|
|
|
|
|
interface, via callback functions, so that the event loop can react
|
|
|
|
|
to other event sources while we wait for input. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Important variables for the event loop. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
/* This is used to determine if GDB is using the readline library or
|
2010-12-29 08:58:14 +08:00
|
|
|
|
its own simplified form of readline. It is used by the asynchronous
|
1999-06-15 02:08:47 +08:00
|
|
|
|
form of the set editing command.
|
1999-05-26 02:09:09 +08:00
|
|
|
|
ezannoni: as of 1999-04-29 I expect that this
|
1999-05-12 04:29:07 +08:00
|
|
|
|
variable will not be used after gdb is changed to use the event
|
2010-12-29 08:58:14 +08:00
|
|
|
|
loop as default engine, and event-top.c is merged into top.c. */
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
int set_editing_cmd_var;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
1999-08-31 09:14:27 +08:00
|
|
|
|
/* This is used to display the notification of the completion of an
|
2010-12-29 08:58:14 +08:00
|
|
|
|
asynchronous execution command. */
|
1999-08-31 09:14:27 +08:00
|
|
|
|
int exec_done_display_p = 0;
|
|
|
|
|
|
2014-11-23 18:03:39 +08:00
|
|
|
|
/* Used by the stdin event handler to compensate for missed stdin events.
|
|
|
|
|
Setting this to a non-zero value inside an stdin callback makes the callback
|
|
|
|
|
run again. */
|
|
|
|
|
int call_stdin_event_handler_again_p;
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Signal handling variables. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
/* Each of these is a pointer to a function that the event loop will
|
2010-12-29 08:58:14 +08:00
|
|
|
|
invoke if the corresponding signal has received. The real signal
|
1999-05-12 04:29:07 +08:00
|
|
|
|
handlers mark these functions as ready to be executed and the event
|
2010-12-29 08:58:14 +08:00
|
|
|
|
loop, in a later iteration, calls them. See the function
|
|
|
|
|
invoke_async_signal_handler. */
|
2012-10-23 14:12:04 +08:00
|
|
|
|
static struct async_signal_handler *sigint_token;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
#ifdef SIGHUP
|
2012-10-23 14:12:04 +08:00
|
|
|
|
static struct async_signal_handler *sighup_token;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
#endif
|
2005-03-17 01:05:31 +08:00
|
|
|
|
#ifdef SIGQUIT
|
2012-10-23 14:12:04 +08:00
|
|
|
|
static struct async_signal_handler *sigquit_token;
|
2005-03-17 01:05:31 +08:00
|
|
|
|
#endif
|
2012-10-23 14:12:04 +08:00
|
|
|
|
static struct async_signal_handler *sigfpe_token;
|
1999-06-15 02:08:47 +08:00
|
|
|
|
#ifdef STOP_SIGNAL
|
2012-10-23 14:12:04 +08:00
|
|
|
|
static struct async_signal_handler *sigtstp_token;
|
1999-06-15 02:08:47 +08:00
|
|
|
|
#endif
|
2014-03-19 05:48:06 +08:00
|
|
|
|
static struct async_signal_handler *async_sigterm_token;
|
1999-06-15 02:08:47 +08:00
|
|
|
|
|
2016-04-22 23:18:33 +08:00
|
|
|
|
/* This hook is called by gdb_rl_callback_read_char_wrapper after each
|
2001-11-27 12:15:09 +08:00
|
|
|
|
character is processed. */
|
2011-03-16 23:18:58 +08:00
|
|
|
|
void (*after_char_processing_hook) (void);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
/* Wrapper function for calling into the readline library. This takes
|
|
|
|
|
care of a couple things:
|
|
|
|
|
|
|
|
|
|
- The event loop expects the callback function to have a parameter,
|
|
|
|
|
while readline expects none.
|
|
|
|
|
|
|
|
|
|
- Propagation of GDB exceptions/errors thrown from INPUT_HANDLER
|
|
|
|
|
across readline requires special handling.
|
|
|
|
|
|
|
|
|
|
On the exceptions issue:
|
|
|
|
|
|
|
|
|
|
DWARF-based unwinding cannot cross code built without -fexceptions.
|
|
|
|
|
Any exception that tries to propagate through such code will fail
|
|
|
|
|
and the result is a call to std::terminate. While some ABIs, such
|
|
|
|
|
as x86-64, require all code to be built with exception tables,
|
|
|
|
|
others don't.
|
|
|
|
|
|
|
|
|
|
This is a problem when GDB calls some non-EH-aware C library code,
|
|
|
|
|
that calls into GDB again through a callback, and that GDB callback
|
|
|
|
|
code throws a C++ exception. Turns out this is exactly what
|
|
|
|
|
happens with GDB's readline callback.
|
|
|
|
|
|
|
|
|
|
In such cases, we must catch and save any C++ exception that might
|
|
|
|
|
be thrown from the GDB callback before returning to the
|
|
|
|
|
non-EH-aware code. When the non-EH-aware function itself returns
|
|
|
|
|
back to GDB, we then rethrow the original C++ exception.
|
|
|
|
|
|
|
|
|
|
In the readline case however, the right thing to do is to longjmp
|
|
|
|
|
out of the callback, rather than do a normal return -- there's no
|
|
|
|
|
way for the callback to return to readline an indication that an
|
|
|
|
|
error happened, so a normal return would have rl_callback_read_char
|
|
|
|
|
potentially continue processing further input, redisplay the
|
|
|
|
|
prompt, etc. Instead of raw setjmp/longjmp however, we use our
|
|
|
|
|
sjlj-based TRY/CATCH mechanism, which knows to handle multiple
|
|
|
|
|
levels of active setjmp/longjmp frames, needed in order to handle
|
|
|
|
|
the readline callback recursing, as happens with e.g., secondary
|
Fix longjmp across readline w/ --enable-sjlj-exceptions toolchains
Nowadays, GDB propagates C++ exceptions across readline using
setjmp/longjmp 89525768cd08 ("Propagate GDB/C++ exceptions across
readline using sj/lj-based TRY/CATCH") because DWARF-based unwinding
can't cross C functions compiled without -fexceptions (see details
from the commit above).
Unfortunately, toolchains that use SjLj-based C++ exceptions got
broken with that fix, because _Unwind_SjLj_Unregister, which is put at
the exit of a function, is not executed due to the longjmp added by
that commit.
(gdb) [New Thread 2936.0xb80]
kill
Thread 1 received signal SIGSEGV, Segmentation fault.
0x03ff662b in ?? ()
top?bt 15
#0 0x03ff662b in ?? ()
#1 0x00526b92 in stdin_event_handler (error=0, client_data=0x172ed8)
at ../../binutils-gdb/gdb/event-top.c:555
#2 0x00525a94 in handle_file_event (ready_mask=<optimized out>,
file_ptr=0x3ff5cb8) at ../../binutils-gdb/gdb/event-loop.c:733
#3 gdb_wait_for_event (block=block@entry=1)
at ../../binutils-gdb/gdb/event-loop.c:884
#4 0x00525bfb in gdb_do_one_event ()
at ../../binutils-gdb/gdb/event-loop.c:347
#5 0x00525ce5 in start_event_loop ()
at ../../binutils-gdb/gdb/event-loop.c:371
#6 0x0051fada in captured_command_loop (data=0x0)
at ../../binutils-gdb/gdb/main.c:324
#7 0x0051cf5d in catch_errors (
func=func@entry=0x51fab0 <captured_command_loop(void*)>,
func_args=func_args@entry=0x0,
errstring=errstring@entry=0x7922bf <VEC_interp_factory_p_quick_push(VEC_inte rp_factory_p*, interp_factory*, char const*, unsigned int)::__PRETTY_FUNCTION__+351> "", mask=mask@entry=RETURN_MASK_ALL)
at ../../binutils-gdb/gdb/exceptions.c:236
#8 0x00520f0c in captured_main (data=0x328feb4)
at ../../binutils-gdb/gdb/main.c:1149
#9 gdb_main (args=args@entry=0x328feb4) at ../../binutils-gdb/gdb/main.c:1159
#10 0x0071e400 in main (argc=1, argv=0x171220)
at ../../binutils-gdb/gdb/gdb.c:32
Fix this by making the functions involved in setjmp/longjmp as
noexcept, so that the compiler knows it doesn't need to emit the
_Unwind_SjLj_Register / _Unwind_SjLj_Unregister calls for C++
exceptions.
Tested on x86_64 Fedora 23 with:
- GCC 5.3.1 w/ DWARF-based exceptions.
- GCC 7 built with --enable-sjlj-exceptions.
gdb/ChangeLog:
2016-12-20 Pedro Alves <palves@redhat.com>
Yao Qi <yao.qi@linaro.org>
PR gdb/20977
* event-top.c (gdb_rl_callback_read_char_wrapper_noexcept): New
noexcept function, factored out from ...
(gdb_rl_callback_read_char_wrapper): ... this.
(gdb_rl_callback_handler): Mark noexcept.
2016-12-20 23:46:44 +08:00
|
|
|
|
prompts / queries, through gdb_readline_wrapper. This must be
|
|
|
|
|
noexcept in order to avoid problems with mixing sjlj and
|
|
|
|
|
(sjlj-based) C++ exceptions. */
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
|
Fix longjmp across readline w/ --enable-sjlj-exceptions toolchains
Nowadays, GDB propagates C++ exceptions across readline using
setjmp/longjmp 89525768cd08 ("Propagate GDB/C++ exceptions across
readline using sj/lj-based TRY/CATCH") because DWARF-based unwinding
can't cross C functions compiled without -fexceptions (see details
from the commit above).
Unfortunately, toolchains that use SjLj-based C++ exceptions got
broken with that fix, because _Unwind_SjLj_Unregister, which is put at
the exit of a function, is not executed due to the longjmp added by
that commit.
(gdb) [New Thread 2936.0xb80]
kill
Thread 1 received signal SIGSEGV, Segmentation fault.
0x03ff662b in ?? ()
top?bt 15
#0 0x03ff662b in ?? ()
#1 0x00526b92 in stdin_event_handler (error=0, client_data=0x172ed8)
at ../../binutils-gdb/gdb/event-top.c:555
#2 0x00525a94 in handle_file_event (ready_mask=<optimized out>,
file_ptr=0x3ff5cb8) at ../../binutils-gdb/gdb/event-loop.c:733
#3 gdb_wait_for_event (block=block@entry=1)
at ../../binutils-gdb/gdb/event-loop.c:884
#4 0x00525bfb in gdb_do_one_event ()
at ../../binutils-gdb/gdb/event-loop.c:347
#5 0x00525ce5 in start_event_loop ()
at ../../binutils-gdb/gdb/event-loop.c:371
#6 0x0051fada in captured_command_loop (data=0x0)
at ../../binutils-gdb/gdb/main.c:324
#7 0x0051cf5d in catch_errors (
func=func@entry=0x51fab0 <captured_command_loop(void*)>,
func_args=func_args@entry=0x0,
errstring=errstring@entry=0x7922bf <VEC_interp_factory_p_quick_push(VEC_inte rp_factory_p*, interp_factory*, char const*, unsigned int)::__PRETTY_FUNCTION__+351> "", mask=mask@entry=RETURN_MASK_ALL)
at ../../binutils-gdb/gdb/exceptions.c:236
#8 0x00520f0c in captured_main (data=0x328feb4)
at ../../binutils-gdb/gdb/main.c:1149
#9 gdb_main (args=args@entry=0x328feb4) at ../../binutils-gdb/gdb/main.c:1159
#10 0x0071e400 in main (argc=1, argv=0x171220)
at ../../binutils-gdb/gdb/gdb.c:32
Fix this by making the functions involved in setjmp/longjmp as
noexcept, so that the compiler knows it doesn't need to emit the
_Unwind_SjLj_Register / _Unwind_SjLj_Unregister calls for C++
exceptions.
Tested on x86_64 Fedora 23 with:
- GCC 5.3.1 w/ DWARF-based exceptions.
- GCC 7 built with --enable-sjlj-exceptions.
gdb/ChangeLog:
2016-12-20 Pedro Alves <palves@redhat.com>
Yao Qi <yao.qi@linaro.org>
PR gdb/20977
* event-top.c (gdb_rl_callback_read_char_wrapper_noexcept): New
noexcept function, factored out from ...
(gdb_rl_callback_read_char_wrapper): ... this.
(gdb_rl_callback_handler): Mark noexcept.
2016-12-20 23:46:44 +08:00
|
|
|
|
static struct gdb_exception
|
|
|
|
|
gdb_rl_callback_read_char_wrapper_noexcept () noexcept
|
1999-09-22 11:28:34 +08:00
|
|
|
|
{
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
struct gdb_exception gdb_expt = exception_none;
|
|
|
|
|
|
|
|
|
|
/* C++ exceptions can't normally be thrown across readline (unless
|
|
|
|
|
it is built with -fexceptions, but it won't by default on many
|
|
|
|
|
ABIs). So we instead wrap the readline call with a sjlj-based
|
|
|
|
|
TRY/CATCH, and rethrow the GDB exception once back in GDB. */
|
|
|
|
|
TRY_SJLJ
|
|
|
|
|
{
|
|
|
|
|
rl_callback_read_char ();
|
|
|
|
|
if (after_char_processing_hook)
|
|
|
|
|
(*after_char_processing_hook) ();
|
|
|
|
|
}
|
|
|
|
|
CATCH_SJLJ (ex, RETURN_MASK_ALL)
|
|
|
|
|
{
|
|
|
|
|
gdb_expt = ex;
|
|
|
|
|
}
|
|
|
|
|
END_CATCH_SJLJ
|
|
|
|
|
|
Fix longjmp across readline w/ --enable-sjlj-exceptions toolchains
Nowadays, GDB propagates C++ exceptions across readline using
setjmp/longjmp 89525768cd08 ("Propagate GDB/C++ exceptions across
readline using sj/lj-based TRY/CATCH") because DWARF-based unwinding
can't cross C functions compiled without -fexceptions (see details
from the commit above).
Unfortunately, toolchains that use SjLj-based C++ exceptions got
broken with that fix, because _Unwind_SjLj_Unregister, which is put at
the exit of a function, is not executed due to the longjmp added by
that commit.
(gdb) [New Thread 2936.0xb80]
kill
Thread 1 received signal SIGSEGV, Segmentation fault.
0x03ff662b in ?? ()
top?bt 15
#0 0x03ff662b in ?? ()
#1 0x00526b92 in stdin_event_handler (error=0, client_data=0x172ed8)
at ../../binutils-gdb/gdb/event-top.c:555
#2 0x00525a94 in handle_file_event (ready_mask=<optimized out>,
file_ptr=0x3ff5cb8) at ../../binutils-gdb/gdb/event-loop.c:733
#3 gdb_wait_for_event (block=block@entry=1)
at ../../binutils-gdb/gdb/event-loop.c:884
#4 0x00525bfb in gdb_do_one_event ()
at ../../binutils-gdb/gdb/event-loop.c:347
#5 0x00525ce5 in start_event_loop ()
at ../../binutils-gdb/gdb/event-loop.c:371
#6 0x0051fada in captured_command_loop (data=0x0)
at ../../binutils-gdb/gdb/main.c:324
#7 0x0051cf5d in catch_errors (
func=func@entry=0x51fab0 <captured_command_loop(void*)>,
func_args=func_args@entry=0x0,
errstring=errstring@entry=0x7922bf <VEC_interp_factory_p_quick_push(VEC_inte rp_factory_p*, interp_factory*, char const*, unsigned int)::__PRETTY_FUNCTION__+351> "", mask=mask@entry=RETURN_MASK_ALL)
at ../../binutils-gdb/gdb/exceptions.c:236
#8 0x00520f0c in captured_main (data=0x328feb4)
at ../../binutils-gdb/gdb/main.c:1149
#9 gdb_main (args=args@entry=0x328feb4) at ../../binutils-gdb/gdb/main.c:1159
#10 0x0071e400 in main (argc=1, argv=0x171220)
at ../../binutils-gdb/gdb/gdb.c:32
Fix this by making the functions involved in setjmp/longjmp as
noexcept, so that the compiler knows it doesn't need to emit the
_Unwind_SjLj_Register / _Unwind_SjLj_Unregister calls for C++
exceptions.
Tested on x86_64 Fedora 23 with:
- GCC 5.3.1 w/ DWARF-based exceptions.
- GCC 7 built with --enable-sjlj-exceptions.
gdb/ChangeLog:
2016-12-20 Pedro Alves <palves@redhat.com>
Yao Qi <yao.qi@linaro.org>
PR gdb/20977
* event-top.c (gdb_rl_callback_read_char_wrapper_noexcept): New
noexcept function, factored out from ...
(gdb_rl_callback_read_char_wrapper): ... this.
(gdb_rl_callback_handler): Mark noexcept.
2016-12-20 23:46:44 +08:00
|
|
|
|
return gdb_expt;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
gdb_rl_callback_read_char_wrapper (gdb_client_data client_data)
|
|
|
|
|
{
|
|
|
|
|
struct gdb_exception gdb_expt
|
|
|
|
|
= gdb_rl_callback_read_char_wrapper_noexcept ();
|
|
|
|
|
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
/* Rethrow using the normal EH mechanism. */
|
|
|
|
|
if (gdb_expt.reason < 0)
|
|
|
|
|
throw_exception (gdb_expt);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* GDB's readline callback handler. Calls the current INPUT_HANDLER,
|
|
|
|
|
and propagates GDB exceptions/errors thrown from INPUT_HANDLER back
|
Fix longjmp across readline w/ --enable-sjlj-exceptions toolchains
Nowadays, GDB propagates C++ exceptions across readline using
setjmp/longjmp 89525768cd08 ("Propagate GDB/C++ exceptions across
readline using sj/lj-based TRY/CATCH") because DWARF-based unwinding
can't cross C functions compiled without -fexceptions (see details
from the commit above).
Unfortunately, toolchains that use SjLj-based C++ exceptions got
broken with that fix, because _Unwind_SjLj_Unregister, which is put at
the exit of a function, is not executed due to the longjmp added by
that commit.
(gdb) [New Thread 2936.0xb80]
kill
Thread 1 received signal SIGSEGV, Segmentation fault.
0x03ff662b in ?? ()
top?bt 15
#0 0x03ff662b in ?? ()
#1 0x00526b92 in stdin_event_handler (error=0, client_data=0x172ed8)
at ../../binutils-gdb/gdb/event-top.c:555
#2 0x00525a94 in handle_file_event (ready_mask=<optimized out>,
file_ptr=0x3ff5cb8) at ../../binutils-gdb/gdb/event-loop.c:733
#3 gdb_wait_for_event (block=block@entry=1)
at ../../binutils-gdb/gdb/event-loop.c:884
#4 0x00525bfb in gdb_do_one_event ()
at ../../binutils-gdb/gdb/event-loop.c:347
#5 0x00525ce5 in start_event_loop ()
at ../../binutils-gdb/gdb/event-loop.c:371
#6 0x0051fada in captured_command_loop (data=0x0)
at ../../binutils-gdb/gdb/main.c:324
#7 0x0051cf5d in catch_errors (
func=func@entry=0x51fab0 <captured_command_loop(void*)>,
func_args=func_args@entry=0x0,
errstring=errstring@entry=0x7922bf <VEC_interp_factory_p_quick_push(VEC_inte rp_factory_p*, interp_factory*, char const*, unsigned int)::__PRETTY_FUNCTION__+351> "", mask=mask@entry=RETURN_MASK_ALL)
at ../../binutils-gdb/gdb/exceptions.c:236
#8 0x00520f0c in captured_main (data=0x328feb4)
at ../../binutils-gdb/gdb/main.c:1149
#9 gdb_main (args=args@entry=0x328feb4) at ../../binutils-gdb/gdb/main.c:1159
#10 0x0071e400 in main (argc=1, argv=0x171220)
at ../../binutils-gdb/gdb/gdb.c:32
Fix this by making the functions involved in setjmp/longjmp as
noexcept, so that the compiler knows it doesn't need to emit the
_Unwind_SjLj_Register / _Unwind_SjLj_Unregister calls for C++
exceptions.
Tested on x86_64 Fedora 23 with:
- GCC 5.3.1 w/ DWARF-based exceptions.
- GCC 7 built with --enable-sjlj-exceptions.
gdb/ChangeLog:
2016-12-20 Pedro Alves <palves@redhat.com>
Yao Qi <yao.qi@linaro.org>
PR gdb/20977
* event-top.c (gdb_rl_callback_read_char_wrapper_noexcept): New
noexcept function, factored out from ...
(gdb_rl_callback_read_char_wrapper): ... this.
(gdb_rl_callback_handler): Mark noexcept.
2016-12-20 23:46:44 +08:00
|
|
|
|
across readline. See gdb_rl_callback_read_char_wrapper. This must
|
|
|
|
|
be noexcept in order to avoid problems with mixing sjlj and
|
|
|
|
|
(sjlj-based) C++ exceptions. */
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
|
|
|
|
|
static void
|
Fix longjmp across readline w/ --enable-sjlj-exceptions toolchains
Nowadays, GDB propagates C++ exceptions across readline using
setjmp/longjmp 89525768cd08 ("Propagate GDB/C++ exceptions across
readline using sj/lj-based TRY/CATCH") because DWARF-based unwinding
can't cross C functions compiled without -fexceptions (see details
from the commit above).
Unfortunately, toolchains that use SjLj-based C++ exceptions got
broken with that fix, because _Unwind_SjLj_Unregister, which is put at
the exit of a function, is not executed due to the longjmp added by
that commit.
(gdb) [New Thread 2936.0xb80]
kill
Thread 1 received signal SIGSEGV, Segmentation fault.
0x03ff662b in ?? ()
top?bt 15
#0 0x03ff662b in ?? ()
#1 0x00526b92 in stdin_event_handler (error=0, client_data=0x172ed8)
at ../../binutils-gdb/gdb/event-top.c:555
#2 0x00525a94 in handle_file_event (ready_mask=<optimized out>,
file_ptr=0x3ff5cb8) at ../../binutils-gdb/gdb/event-loop.c:733
#3 gdb_wait_for_event (block=block@entry=1)
at ../../binutils-gdb/gdb/event-loop.c:884
#4 0x00525bfb in gdb_do_one_event ()
at ../../binutils-gdb/gdb/event-loop.c:347
#5 0x00525ce5 in start_event_loop ()
at ../../binutils-gdb/gdb/event-loop.c:371
#6 0x0051fada in captured_command_loop (data=0x0)
at ../../binutils-gdb/gdb/main.c:324
#7 0x0051cf5d in catch_errors (
func=func@entry=0x51fab0 <captured_command_loop(void*)>,
func_args=func_args@entry=0x0,
errstring=errstring@entry=0x7922bf <VEC_interp_factory_p_quick_push(VEC_inte rp_factory_p*, interp_factory*, char const*, unsigned int)::__PRETTY_FUNCTION__+351> "", mask=mask@entry=RETURN_MASK_ALL)
at ../../binutils-gdb/gdb/exceptions.c:236
#8 0x00520f0c in captured_main (data=0x328feb4)
at ../../binutils-gdb/gdb/main.c:1149
#9 gdb_main (args=args@entry=0x328feb4) at ../../binutils-gdb/gdb/main.c:1159
#10 0x0071e400 in main (argc=1, argv=0x171220)
at ../../binutils-gdb/gdb/gdb.c:32
Fix this by making the functions involved in setjmp/longjmp as
noexcept, so that the compiler knows it doesn't need to emit the
_Unwind_SjLj_Register / _Unwind_SjLj_Unregister calls for C++
exceptions.
Tested on x86_64 Fedora 23 with:
- GCC 5.3.1 w/ DWARF-based exceptions.
- GCC 7 built with --enable-sjlj-exceptions.
gdb/ChangeLog:
2016-12-20 Pedro Alves <palves@redhat.com>
Yao Qi <yao.qi@linaro.org>
PR gdb/20977
* event-top.c (gdb_rl_callback_read_char_wrapper_noexcept): New
noexcept function, factored out from ...
(gdb_rl_callback_read_char_wrapper): ... this.
(gdb_rl_callback_handler): Mark noexcept.
2016-12-20 23:46:44 +08:00
|
|
|
|
gdb_rl_callback_handler (char *rl) noexcept
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
{
|
|
|
|
|
struct gdb_exception gdb_rl_expt = exception_none;
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
|
|
|
|
|
TRY
|
|
|
|
|
{
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
ui->input_handler (rl);
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
}
|
|
|
|
|
CATCH (ex, RETURN_MASK_ALL)
|
|
|
|
|
{
|
|
|
|
|
gdb_rl_expt = ex;
|
|
|
|
|
}
|
|
|
|
|
END_CATCH
|
|
|
|
|
|
|
|
|
|
/* If we caught a GDB exception, longjmp out of the readline
|
|
|
|
|
callback. There's no other way for the callback to signal to
|
|
|
|
|
readline that an error happened. A normal return would have
|
|
|
|
|
readline potentially continue processing further input, redisplay
|
|
|
|
|
the prompt, etc. (This is what GDB historically did when it was
|
|
|
|
|
a C program.) Note that since we're long jumping, local variable
|
|
|
|
|
dtors are NOT run automatically. */
|
|
|
|
|
if (gdb_rl_expt.reason < 0)
|
|
|
|
|
throw_exception_sjlj (gdb_rl_expt);
|
1999-09-22 11:28:34 +08:00
|
|
|
|
}
|
|
|
|
|
|
1999-05-12 04:29:07 +08:00
|
|
|
|
/* Change the function to be invoked every time there is a character
|
2010-12-29 08:58:14 +08:00
|
|
|
|
ready on stdin. This is used when the user sets the editing off,
|
1999-05-12 04:29:07 +08:00
|
|
|
|
therefore bypassing readline, and letting gdb handle the input
|
2016-03-10 02:24:59 +08:00
|
|
|
|
itself, via gdb_readline_no_editing_callback. Also it is used in
|
|
|
|
|
the opposite case in which the user sets editing on again, by
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
restoring readline handling of the input.
|
|
|
|
|
|
|
|
|
|
NOTE: this operates on input_fd, not instream. If we are reading
|
|
|
|
|
commands from a file, instream will point to the file. However, we
|
|
|
|
|
always read commands from a file with editing off. This means that
|
|
|
|
|
the 'set editing on/off' will have effect only on the interactive
|
|
|
|
|
session. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
change_line_handler (int editing)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
|
|
|
|
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
/* We can only have one instance of readline, so we only allow
|
|
|
|
|
editing on the main UI. */
|
|
|
|
|
if (ui != main_ui)
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
/* Don't try enabling editing if the interpreter doesn't support it
|
|
|
|
|
(e.g., MI). */
|
|
|
|
|
if (!interp_supports_command_editing (top_level_interpreter ())
|
|
|
|
|
|| !interp_supports_command_editing (command_interp ()))
|
|
|
|
|
return;
|
1999-09-22 11:28:34 +08:00
|
|
|
|
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
if (editing)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
gdb_assert (ui == main_ui);
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Turn on editing by using readline. */
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
ui->call_readline = gdb_rl_callback_read_char_wrapper;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
2016-03-10 02:24:59 +08:00
|
|
|
|
/* Turn off editing by using gdb_readline_no_editing_callback. */
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
if (ui->command_editing)
|
|
|
|
|
gdb_rl_callback_handler_remove ();
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
ui->call_readline = gdb_readline_no_editing_callback;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
ui->command_editing = editing;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
PR python/17372 - Python hangs when displaying help()
This is more of a readline/terminal issue than a Python one.
PR17372 is a regression in 7.8 caused by the fix for PR17072:
commit 0017922d0292d8c374584f6100874580659c9973
Author: Pedro Alves <palves@redhat.com>
Date: Mon Jul 14 19:55:32 2014 +0100
Background execution + pagination aborts readline/gdb
gdb_readline_wrapper_line removes the handler after a line is
processed. Usually, we'll end up re-displaying the prompt, and that
reinstalls the handler. But if the output is coming out of handling
a stop event, we don't re-display the prompt, and nothing restores the
handler. So the next input wakes up the event loop and calls into
readline, which aborts.
...
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* top.c (gdb_readline_wrapper_line): Tweak comment.
(gdb_readline_wrapper_cleanup): If readline is enabled, reinstall
the input handler callback.
The problem is that installing the input handler callback also preps
the terminal, putting it in raw mode and with echo disabled, which is
bad if we're going to call a command that assumes cooked/canonical
mode, and echo enabled, like in the case of the PR, Python's
interactive shell. Another example I came up with that doesn't depend
on Python is starting a subshell with "(gdb) shell /bin/sh" from a
multi-line command. Tests covering both these examples are added.
The fix is to revert the original fix for PR gdb/17072, and instead
restore the callback handler after processing an asynchronous target
event.
Furthermore, calling rl_callback_handler_install when we already have
some input in readline's line buffer discards that input, which is
obviously a bad thing to do while the user is typing. No specific
test is added for that, because I first tried calling it even if the
callback handler was still installed and that resulted in hundreds of
failures in the testsuite.
gdb/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* event-top.c (change_line_handler): Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(callback_handler_installed): New global.
(gdb_rl_callback_handler_remove, gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New functions.
(display_gdb_prompt): Call gdb_rl_callback_handler_remove and
gdb_rl_callback_handler_install instead of
rl_callback_handler_remove and rl_callback_handler_install.
(gdb_disable_readline): Call gdb_rl_callback_handler_remove
instead of rl_callback_handler_remove.
* event-top.h (gdb_rl_callback_handler_remove)
(gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New declarations.
* infrun.c (reinstall_readline_callback_handler_cleanup): New
cleanup function.
(fetch_inferior_event): Install it.
* top.c (gdb_readline_wrapper_line) Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(gdb_readline_wrapper_cleanup): Don't call
rl_callback_handler_install.
gdb/testsuite/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* gdb.python/python.exp: Test a multi-line command that spawns
interactive Python.
* gdb.base/multi-line-starts-subshell.exp: New file.
2014-10-24 00:13:35 +08:00
|
|
|
|
/* The functions below are wrappers for rl_callback_handler_remove and
|
|
|
|
|
rl_callback_handler_install that keep track of whether the callback
|
|
|
|
|
handler is installed in readline. This is necessary because after
|
|
|
|
|
handling a target event of a background execution command, we may
|
|
|
|
|
need to reinstall the callback handler if it was removed due to a
|
|
|
|
|
secondary prompt. See gdb_readline_wrapper_line. We don't
|
|
|
|
|
unconditionally install the handler for every target event because
|
|
|
|
|
that also clears the line buffer, thus installing it while the user
|
|
|
|
|
is typing would lose input. */
|
|
|
|
|
|
|
|
|
|
/* Whether we've registered a callback handler with readline. */
|
|
|
|
|
static int callback_handler_installed;
|
|
|
|
|
|
|
|
|
|
/* See event-top.h, and above. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
gdb_rl_callback_handler_remove (void)
|
|
|
|
|
{
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
gdb_assert (current_ui == main_ui);
|
|
|
|
|
|
PR python/17372 - Python hangs when displaying help()
This is more of a readline/terminal issue than a Python one.
PR17372 is a regression in 7.8 caused by the fix for PR17072:
commit 0017922d0292d8c374584f6100874580659c9973
Author: Pedro Alves <palves@redhat.com>
Date: Mon Jul 14 19:55:32 2014 +0100
Background execution + pagination aborts readline/gdb
gdb_readline_wrapper_line removes the handler after a line is
processed. Usually, we'll end up re-displaying the prompt, and that
reinstalls the handler. But if the output is coming out of handling
a stop event, we don't re-display the prompt, and nothing restores the
handler. So the next input wakes up the event loop and calls into
readline, which aborts.
...
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* top.c (gdb_readline_wrapper_line): Tweak comment.
(gdb_readline_wrapper_cleanup): If readline is enabled, reinstall
the input handler callback.
The problem is that installing the input handler callback also preps
the terminal, putting it in raw mode and with echo disabled, which is
bad if we're going to call a command that assumes cooked/canonical
mode, and echo enabled, like in the case of the PR, Python's
interactive shell. Another example I came up with that doesn't depend
on Python is starting a subshell with "(gdb) shell /bin/sh" from a
multi-line command. Tests covering both these examples are added.
The fix is to revert the original fix for PR gdb/17072, and instead
restore the callback handler after processing an asynchronous target
event.
Furthermore, calling rl_callback_handler_install when we already have
some input in readline's line buffer discards that input, which is
obviously a bad thing to do while the user is typing. No specific
test is added for that, because I first tried calling it even if the
callback handler was still installed and that resulted in hundreds of
failures in the testsuite.
gdb/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* event-top.c (change_line_handler): Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(callback_handler_installed): New global.
(gdb_rl_callback_handler_remove, gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New functions.
(display_gdb_prompt): Call gdb_rl_callback_handler_remove and
gdb_rl_callback_handler_install instead of
rl_callback_handler_remove and rl_callback_handler_install.
(gdb_disable_readline): Call gdb_rl_callback_handler_remove
instead of rl_callback_handler_remove.
* event-top.h (gdb_rl_callback_handler_remove)
(gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New declarations.
* infrun.c (reinstall_readline_callback_handler_cleanup): New
cleanup function.
(fetch_inferior_event): Install it.
* top.c (gdb_readline_wrapper_line) Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(gdb_readline_wrapper_cleanup): Don't call
rl_callback_handler_install.
gdb/testsuite/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* gdb.python/python.exp: Test a multi-line command that spawns
interactive Python.
* gdb.base/multi-line-starts-subshell.exp: New file.
2014-10-24 00:13:35 +08:00
|
|
|
|
rl_callback_handler_remove ();
|
|
|
|
|
callback_handler_installed = 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See event-top.h, and above. Note this wrapper doesn't have an
|
|
|
|
|
actual callback parameter because we always install
|
|
|
|
|
INPUT_HANDLER. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
gdb_rl_callback_handler_install (const char *prompt)
|
|
|
|
|
{
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
gdb_assert (current_ui == main_ui);
|
|
|
|
|
|
PR python/17372 - Python hangs when displaying help()
This is more of a readline/terminal issue than a Python one.
PR17372 is a regression in 7.8 caused by the fix for PR17072:
commit 0017922d0292d8c374584f6100874580659c9973
Author: Pedro Alves <palves@redhat.com>
Date: Mon Jul 14 19:55:32 2014 +0100
Background execution + pagination aborts readline/gdb
gdb_readline_wrapper_line removes the handler after a line is
processed. Usually, we'll end up re-displaying the prompt, and that
reinstalls the handler. But if the output is coming out of handling
a stop event, we don't re-display the prompt, and nothing restores the
handler. So the next input wakes up the event loop and calls into
readline, which aborts.
...
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* top.c (gdb_readline_wrapper_line): Tweak comment.
(gdb_readline_wrapper_cleanup): If readline is enabled, reinstall
the input handler callback.
The problem is that installing the input handler callback also preps
the terminal, putting it in raw mode and with echo disabled, which is
bad if we're going to call a command that assumes cooked/canonical
mode, and echo enabled, like in the case of the PR, Python's
interactive shell. Another example I came up with that doesn't depend
on Python is starting a subshell with "(gdb) shell /bin/sh" from a
multi-line command. Tests covering both these examples are added.
The fix is to revert the original fix for PR gdb/17072, and instead
restore the callback handler after processing an asynchronous target
event.
Furthermore, calling rl_callback_handler_install when we already have
some input in readline's line buffer discards that input, which is
obviously a bad thing to do while the user is typing. No specific
test is added for that, because I first tried calling it even if the
callback handler was still installed and that resulted in hundreds of
failures in the testsuite.
gdb/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* event-top.c (change_line_handler): Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(callback_handler_installed): New global.
(gdb_rl_callback_handler_remove, gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New functions.
(display_gdb_prompt): Call gdb_rl_callback_handler_remove and
gdb_rl_callback_handler_install instead of
rl_callback_handler_remove and rl_callback_handler_install.
(gdb_disable_readline): Call gdb_rl_callback_handler_remove
instead of rl_callback_handler_remove.
* event-top.h (gdb_rl_callback_handler_remove)
(gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New declarations.
* infrun.c (reinstall_readline_callback_handler_cleanup): New
cleanup function.
(fetch_inferior_event): Install it.
* top.c (gdb_readline_wrapper_line) Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(gdb_readline_wrapper_cleanup): Don't call
rl_callback_handler_install.
gdb/testsuite/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* gdb.python/python.exp: Test a multi-line command that spawns
interactive Python.
* gdb.base/multi-line-starts-subshell.exp: New file.
2014-10-24 00:13:35 +08:00
|
|
|
|
/* Calling rl_callback_handler_install resets readline's input
|
|
|
|
|
buffer. Calling this when we were already processing input
|
|
|
|
|
therefore loses input. */
|
|
|
|
|
gdb_assert (!callback_handler_installed);
|
|
|
|
|
|
Propagate GDB/C++ exceptions across readline using sj/lj-based TRY/CATCH
If we map GDB'S TRY/CATCH macros to C++ try/catch, GDB breaks on
systems where readline isn't built with exceptions support. The
problem is that readline calls into GDB through the callback
interface, and if GDB's callback throws a C++ exception/error, the
system unwinder won't manage to unwind past the readline frame, and
ends up calling std::terminate(), which aborts the process:
(gdb) whatever-command-that-causes-an-error
terminate called after throwing an instance of 'gdb_exception_RETURN_MASK_ERROR'
Aborted
$
This went unnoticed for so long because:
- the x86-64 ABI requires -fasynchronous-unwind-tables, making it
possible for exceptions to cross readline with no special handling.
But e.g., on ARM or AIX, unless you build readline with
-fexceptions, you trip on the problem.
- TRY/CATCH was mapped to setjmp/longjmp, even in C++ mode, until
quite recently.
The fix is to catch and save any GDB exception that is thrown inside
the GDB readline callback, and then once the callback returns back to
the GDB code that called into readline in the first place, rethrow the
saved GDB exception.
This is similar in spirit to how we catch/map GDB exceptions at the
GDB/Python and GDB/Guile API boundaries.
The next question is then: if we intercept all exceptions within GDB's
readline callback, should we simply return normally to readline? The
callback prototype has no way to signal an error back to readline (*).
The answer is no -- if we return normally, we'll be returning to a
loop inside rl_callback_read_char that continues processing pending
input, calling into GDB again, redisplaying the prompt, etc. Thus if
we want to error out of rl_callback_read_char, we need to long jump
across it, just like we always did before TRY/CATCH were ever mapped
to C++ exceptions.
My first approach built a specialized API to handle this, with a
couple macros to hide the setjmp/longjmp and the struct gdb_exception
saving/rethrowing.
However, I realized that we need to:
- Handle multiple active rl_callback_read_char invocations. If,
while processing input something triggers a secondary prompt, we
end up in a nested rl_callback_read_char call, through
gdb_readline_wrapper.
- Propagate a struct gdb_exception along with the longjmp.
... and that this is exactly what the setjmp/longjmp-based TRY/CATCH
does.
So the fix makes the setjmp/longjmp TRY/CATCH always available under
new TRY_SJLJ/CATCH_SJLJ aliases, even when TRY/CATCH is mapped to C++
try/catch, and then uses TRY_SJLJ/CATCH_SJLJ to propagate GDB
exceptions across the readline callback.
This turns out to be a much better looking fix than my bespoke API
attempt, even. We'll probably be able to simplify TRY_SJLJ/CATCH_SJLJ
when we finally get rid of TRY/CATCH all over the tree, but until
then, this reuse seems quite nice for avoiding a second parallel
setjmp/longjmp mechanism.
(*) - maybe we could propose a readline API change, but we still need
to handle current readline, anyway.
gdb/ChangeLog:
2016-04-22 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (enum catcher_state, struct catcher)
(current_catcher): Define in C++ mode too.
(exceptions_state_mc_catch): Call throw_exception_sjlj instead of
throw_exception.
(throw_exception_sjlj, throw_exception_cxx): New functions,
factored out from throw_exception.
(throw_exception): Reimplement.
* common/common-exceptions.h (exceptions_state_mc_init)
(exceptions_state_mc_action_iter)
(exceptions_state_mc_action_iter_1, exceptions_state_mc_catch):
Declare in C++ mode too.
(TRY): Rename to ...
(TRY_SJLJ): ... this.
(CATCH): Rename to ...
(CATCH_SJLJ): ... this.
(END_CATCH): Rename to ...
(END_CATCH_SJLJ): ... this.
[GDB_XCPT == GDB_XCPT_SJMP] (TRY, CATCH, END_CATCH): Map to SJLJ
equivalents.
(throw_exception): Update comments.
(throw_exception_sjlj): Declare.
* event-top.c (gdb_rl_callback_read_char_wrapper): Extend intro
comment. Wrap body in TRY_SJLJ/CATCH_SJLJ and rethrow any
intercepted exception.
(gdb_rl_callback_handler): New function.
(gdb_rl_callback_handler_install): Always install
gdb_rl_callback_handler as readline callback.
2016-04-22 23:18:33 +08:00
|
|
|
|
rl_callback_handler_install (prompt, gdb_rl_callback_handler);
|
PR python/17372 - Python hangs when displaying help()
This is more of a readline/terminal issue than a Python one.
PR17372 is a regression in 7.8 caused by the fix for PR17072:
commit 0017922d0292d8c374584f6100874580659c9973
Author: Pedro Alves <palves@redhat.com>
Date: Mon Jul 14 19:55:32 2014 +0100
Background execution + pagination aborts readline/gdb
gdb_readline_wrapper_line removes the handler after a line is
processed. Usually, we'll end up re-displaying the prompt, and that
reinstalls the handler. But if the output is coming out of handling
a stop event, we don't re-display the prompt, and nothing restores the
handler. So the next input wakes up the event loop and calls into
readline, which aborts.
...
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* top.c (gdb_readline_wrapper_line): Tweak comment.
(gdb_readline_wrapper_cleanup): If readline is enabled, reinstall
the input handler callback.
The problem is that installing the input handler callback also preps
the terminal, putting it in raw mode and with echo disabled, which is
bad if we're going to call a command that assumes cooked/canonical
mode, and echo enabled, like in the case of the PR, Python's
interactive shell. Another example I came up with that doesn't depend
on Python is starting a subshell with "(gdb) shell /bin/sh" from a
multi-line command. Tests covering both these examples are added.
The fix is to revert the original fix for PR gdb/17072, and instead
restore the callback handler after processing an asynchronous target
event.
Furthermore, calling rl_callback_handler_install when we already have
some input in readline's line buffer discards that input, which is
obviously a bad thing to do while the user is typing. No specific
test is added for that, because I first tried calling it even if the
callback handler was still installed and that resulted in hundreds of
failures in the testsuite.
gdb/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* event-top.c (change_line_handler): Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(callback_handler_installed): New global.
(gdb_rl_callback_handler_remove, gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New functions.
(display_gdb_prompt): Call gdb_rl_callback_handler_remove and
gdb_rl_callback_handler_install instead of
rl_callback_handler_remove and rl_callback_handler_install.
(gdb_disable_readline): Call gdb_rl_callback_handler_remove
instead of rl_callback_handler_remove.
* event-top.h (gdb_rl_callback_handler_remove)
(gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New declarations.
* infrun.c (reinstall_readline_callback_handler_cleanup): New
cleanup function.
(fetch_inferior_event): Install it.
* top.c (gdb_readline_wrapper_line) Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(gdb_readline_wrapper_cleanup): Don't call
rl_callback_handler_install.
gdb/testsuite/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* gdb.python/python.exp: Test a multi-line command that spawns
interactive Python.
* gdb.base/multi-line-starts-subshell.exp: New file.
2014-10-24 00:13:35 +08:00
|
|
|
|
callback_handler_installed = 1;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See event-top.h, and above. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
gdb_rl_callback_handler_reinstall (void)
|
|
|
|
|
{
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
gdb_assert (current_ui == main_ui);
|
|
|
|
|
|
PR python/17372 - Python hangs when displaying help()
This is more of a readline/terminal issue than a Python one.
PR17372 is a regression in 7.8 caused by the fix for PR17072:
commit 0017922d0292d8c374584f6100874580659c9973
Author: Pedro Alves <palves@redhat.com>
Date: Mon Jul 14 19:55:32 2014 +0100
Background execution + pagination aborts readline/gdb
gdb_readline_wrapper_line removes the handler after a line is
processed. Usually, we'll end up re-displaying the prompt, and that
reinstalls the handler. But if the output is coming out of handling
a stop event, we don't re-display the prompt, and nothing restores the
handler. So the next input wakes up the event loop and calls into
readline, which aborts.
...
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* top.c (gdb_readline_wrapper_line): Tweak comment.
(gdb_readline_wrapper_cleanup): If readline is enabled, reinstall
the input handler callback.
The problem is that installing the input handler callback also preps
the terminal, putting it in raw mode and with echo disabled, which is
bad if we're going to call a command that assumes cooked/canonical
mode, and echo enabled, like in the case of the PR, Python's
interactive shell. Another example I came up with that doesn't depend
on Python is starting a subshell with "(gdb) shell /bin/sh" from a
multi-line command. Tests covering both these examples are added.
The fix is to revert the original fix for PR gdb/17072, and instead
restore the callback handler after processing an asynchronous target
event.
Furthermore, calling rl_callback_handler_install when we already have
some input in readline's line buffer discards that input, which is
obviously a bad thing to do while the user is typing. No specific
test is added for that, because I first tried calling it even if the
callback handler was still installed and that resulted in hundreds of
failures in the testsuite.
gdb/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* event-top.c (change_line_handler): Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(callback_handler_installed): New global.
(gdb_rl_callback_handler_remove, gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New functions.
(display_gdb_prompt): Call gdb_rl_callback_handler_remove and
gdb_rl_callback_handler_install instead of
rl_callback_handler_remove and rl_callback_handler_install.
(gdb_disable_readline): Call gdb_rl_callback_handler_remove
instead of rl_callback_handler_remove.
* event-top.h (gdb_rl_callback_handler_remove)
(gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New declarations.
* infrun.c (reinstall_readline_callback_handler_cleanup): New
cleanup function.
(fetch_inferior_event): Install it.
* top.c (gdb_readline_wrapper_line) Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(gdb_readline_wrapper_cleanup): Don't call
rl_callback_handler_install.
gdb/testsuite/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* gdb.python/python.exp: Test a multi-line command that spawns
interactive Python.
* gdb.base/multi-line-starts-subshell.exp: New file.
2014-10-24 00:13:35 +08:00
|
|
|
|
if (!callback_handler_installed)
|
|
|
|
|
{
|
|
|
|
|
/* Passing NULL as prompt argument tells readline to not display
|
|
|
|
|
a prompt. */
|
|
|
|
|
gdb_rl_callback_handler_install (NULL);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
/* Displays the prompt. If the argument NEW_PROMPT is NULL, the
|
|
|
|
|
prompt that is displayed is the current top level prompt.
|
|
|
|
|
Otherwise, it displays whatever NEW_PROMPT is as a local/secondary
|
|
|
|
|
prompt.
|
|
|
|
|
|
|
|
|
|
This is used after each gdb command has completed, and in the
|
|
|
|
|
following cases:
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
1. When the user enters a command line which is ended by '\'
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
indicating that the command will continue on the next line. In
|
|
|
|
|
that case the prompt that is displayed is the empty string.
|
|
|
|
|
|
1999-06-15 02:08:47 +08:00
|
|
|
|
2. When the user is entering 'commands' for a breakpoint, or
|
2010-12-29 08:58:14 +08:00
|
|
|
|
actions for a tracepoint. In this case the prompt will be '>'
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
|
|
|
|
|
3. On prompting for pagination. */
|
|
|
|
|
|
1999-05-12 04:29:07 +08:00
|
|
|
|
void
|
Fix build with Python 3.4 (PR python/16784)
The type of the function pointer PyOS_ReadlineFunctionPointer (part of the
Python C API), which we use, slightly changed starting with Python 3.4. The
signature went from
PyAPI_DATA(char) *(*PyOS_ReadlineFunctionPointer)(FILE *, FILE *, char *);
to
PyAPI_DATA(char) *(*PyOS_ReadlineFunctionPointer)(FILE *, FILE *, const char *);
The parameter that changed is the prompt text.
This commits adjust gdb accordingly by making the prompt_arg parameter
const, as well as the fallouts of that. I needed to rework how
annotations are added to the prompt, since the it is now const. If
annotations are enabled, it will make a copy of the prompt overwrite the
prompt variable that is used throughout the function. Otherwise, no copy
is done and the original prompt_arg value is passed.
I changed the signature of deprecated_readline_hook. I would've changed any
user of it, but it seems like nothing is using it,
Built-tested with python 2.7.x, 3.3.y and 3.4.z.
gdb/ChangeLog:
* defs.h (gdb_readline): Constify argument.
(gdb_readline_wrapper): Same.
(command_line_input): Same.
(deprecated_readline_hook): Same.
* top.c (deprecated_readline_hook): Same.
(gdb_readline): Same.
(gdb_readline_wrapper): Same.
(command_line_input): Constify argument. Pass prompt to
called functions instead of local_prompt, overwriting prompt
if using annotations.
* event-top.h (display_gdb_prompt): Constify argument.
* event-top.c (display_gdb_prompt): Same.
* python/py-gdb-readline.c (gdbpy_readline_wrapper): Constify
argument if building with Python 3.4 and up.
Signed-off-by: Simon Marchi <simon.marchi@ericsson.com>
2014-12-16 00:38:03 +08:00
|
|
|
|
display_gdb_prompt (const char *new_prompt)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2017-09-28 10:30:19 +08:00
|
|
|
|
std::string actual_gdb_prompt;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
With some changes to how software single-step (SSS) breakpoints are
handled, one of those being to place SSS breakpoints on the breakpoint
chain as all other breakpoints, annota1.exp times out with lots and
lots of breakpoint-invalid and frame-changed annotations. All those
extra annotations are actually unnecessary. For one, SSS breakpoints
are internal breakpoints, so the frontend shouldn't care if they were
added, removed or changed. Then, there's really no point in emitting
"breakpoints-invalid" or "frames-invalid" more than once between times
the frontend/user can actually issues GDB commands; the frontend will
have to wait for the GDB prompt to refresh its state, so emitting
those annotations at most once between prompts is enough. Non-stop or
async would complicate this, but no frontend will be using annotations
in those modes (one of goes of emacs switching to MI was non-stop mode
support, AFAIK). The previous patch reveals there has been an
intention in the past to suppress multiple breakpoints-invalid
annotations caused by ignore count changes. As the previous patch
shows, that's always been broken, but in any case, this patch actually
makes it work. The next patch will remove several annotation-specific
calls in breakpoint.c in favor of always using the breakpoint modified
& friends observers, and that causes yet more of these annotations,
because several calls to the corresponding annotate_* functions in
breakpoint.c are missing, particularly in newer code.
So all in all, here's a simple mechanism that avoids sending the same
annotation to the frontend more than once until gdb is ready to accept
further commands.
Tested on x86_64 Fedora 17.
2013-01-22 Pedro Alves <palves@redhat.com>
* annotate.c: Include "inferior.h".
(frames_invalid_emitted)
(breakpoints_invalid_emitted): New globals.
(async_background_execution_p): New function.
(annotate_breakpoints_changed, annotate_frames_invalid): Skip
emitting the annotation if it has already been emitted.
(annotate_display_prompt): New function.
* annotate.h (annotate_display_prompt): New declaration.
* event-top.c: Include annotate.h.
(display_gdb_prompt): Call annotate_display_prompt.
2013-01-23 04:17:10 +08:00
|
|
|
|
annotate_display_prompt ();
|
|
|
|
|
|
2006-07-21 22:46:56 +08:00
|
|
|
|
/* Reset the nesting depth used when trace-commands is set. */
|
|
|
|
|
reset_command_nest_depth ();
|
|
|
|
|
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
/* Do not call the python hook on an explicit prompt change as
|
|
|
|
|
passed to this function, as this forms a secondary/local prompt,
|
|
|
|
|
IE, displayed but not set. */
|
|
|
|
|
if (! new_prompt)
|
1999-07-20 07:30:11 +08:00
|
|
|
|
{
|
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.
2016-06-21 08:11:51 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
|
|
|
|
|
|
|
|
|
if (ui->prompt_state == PROMPTED)
|
|
|
|
|
internal_error (__FILE__, __LINE__, _("double prompt"));
|
|
|
|
|
else if (ui->prompt_state == PROMPT_BLOCKED)
|
2011-07-21 19:03:48 +08:00
|
|
|
|
{
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
/* This is to trick readline into not trying to display the
|
|
|
|
|
prompt. Even though we display the prompt using this
|
|
|
|
|
function, readline still tries to do its own display if
|
|
|
|
|
we don't call rl_callback_handler_install and
|
|
|
|
|
rl_callback_handler_remove (which readline detects
|
|
|
|
|
because a global variable is not set). If readline did
|
|
|
|
|
that, it could mess up gdb signal handlers for SIGINT.
|
|
|
|
|
Readline assumes that between calls to rl_set_signals and
|
|
|
|
|
rl_clear_signals gdb doesn't do anything with the signal
|
|
|
|
|
handlers. Well, that's not the case, because when the
|
|
|
|
|
target executes we change the SIGINT signal handler. If
|
|
|
|
|
we allowed readline to display the prompt, the signal
|
|
|
|
|
handler change would happen exactly between the calls to
|
|
|
|
|
the above two functions. Calling
|
|
|
|
|
rl_callback_handler_remove(), does the job. */
|
|
|
|
|
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
if (current_ui->command_editing)
|
|
|
|
|
gdb_rl_callback_handler_remove ();
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
return;
|
2011-07-21 19:03:48 +08:00
|
|
|
|
}
|
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.
2016-06-21 08:11:51 +08:00
|
|
|
|
else if (ui->prompt_state == PROMPT_NEEDED)
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
{
|
|
|
|
|
/* Display the top level prompt. */
|
|
|
|
|
actual_gdb_prompt = top_level_prompt ();
|
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.
2016-06-21 08:11:51 +08:00
|
|
|
|
ui->prompt_state = PROMPTED;
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
}
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
else
|
2017-09-28 10:30:19 +08:00
|
|
|
|
actual_gdb_prompt = new_prompt;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
if (current_ui->command_editing)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
PR python/17372 - Python hangs when displaying help()
This is more of a readline/terminal issue than a Python one.
PR17372 is a regression in 7.8 caused by the fix for PR17072:
commit 0017922d0292d8c374584f6100874580659c9973
Author: Pedro Alves <palves@redhat.com>
Date: Mon Jul 14 19:55:32 2014 +0100
Background execution + pagination aborts readline/gdb
gdb_readline_wrapper_line removes the handler after a line is
processed. Usually, we'll end up re-displaying the prompt, and that
reinstalls the handler. But if the output is coming out of handling
a stop event, we don't re-display the prompt, and nothing restores the
handler. So the next input wakes up the event loop and calls into
readline, which aborts.
...
gdb/
2014-07-14 Pedro Alves <palves@redhat.com>
PR gdb/17072
* top.c (gdb_readline_wrapper_line): Tweak comment.
(gdb_readline_wrapper_cleanup): If readline is enabled, reinstall
the input handler callback.
The problem is that installing the input handler callback also preps
the terminal, putting it in raw mode and with echo disabled, which is
bad if we're going to call a command that assumes cooked/canonical
mode, and echo enabled, like in the case of the PR, Python's
interactive shell. Another example I came up with that doesn't depend
on Python is starting a subshell with "(gdb) shell /bin/sh" from a
multi-line command. Tests covering both these examples are added.
The fix is to revert the original fix for PR gdb/17072, and instead
restore the callback handler after processing an asynchronous target
event.
Furthermore, calling rl_callback_handler_install when we already have
some input in readline's line buffer discards that input, which is
obviously a bad thing to do while the user is typing. No specific
test is added for that, because I first tried calling it even if the
callback handler was still installed and that resulted in hundreds of
failures in the testsuite.
gdb/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* event-top.c (change_line_handler): Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(callback_handler_installed): New global.
(gdb_rl_callback_handler_remove, gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New functions.
(display_gdb_prompt): Call gdb_rl_callback_handler_remove and
gdb_rl_callback_handler_install instead of
rl_callback_handler_remove and rl_callback_handler_install.
(gdb_disable_readline): Call gdb_rl_callback_handler_remove
instead of rl_callback_handler_remove.
* event-top.h (gdb_rl_callback_handler_remove)
(gdb_rl_callback_handler_install)
(gdb_rl_callback_handler_reinstall): New declarations.
* infrun.c (reinstall_readline_callback_handler_cleanup): New
cleanup function.
(fetch_inferior_event): Install it.
* top.c (gdb_readline_wrapper_line) Call
gdb_rl_callback_handler_remove instead of
rl_callback_handler_remove.
(gdb_readline_wrapper_cleanup): Don't call
rl_callback_handler_install.
gdb/testsuite/
2014-10-29 Pedro Alves <palves@redhat.com>
PR python/17372
* gdb.python/python.exp: Test a multi-line command that spawns
interactive Python.
* gdb.base/multi-line-starts-subshell.exp: New file.
2014-10-24 00:13:35 +08:00
|
|
|
|
gdb_rl_callback_handler_remove ();
|
2017-09-28 10:30:19 +08:00
|
|
|
|
gdb_rl_callback_handler_install (actual_gdb_prompt.c_str ());
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* new_prompt at this point can be the top of the stack or the one
|
2011-03-02 05:12:27 +08:00
|
|
|
|
passed in. It can't be NULL. */
|
|
|
|
|
else
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
|
|
|
|
/* Don't use a _filtered function here. It causes the assumed
|
|
|
|
|
character position to be off, since the newline we read from
|
|
|
|
|
the user is not accounted for. */
|
2017-09-28 10:30:19 +08:00
|
|
|
|
fputs_unfiltered (actual_gdb_prompt.c_str (), gdb_stdout);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
gdb_flush (gdb_stdout);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
/* Return the top level prompt, as specified by "set prompt", possibly
|
|
|
|
|
overriden by the python gdb.prompt_hook hook, and then composed
|
2017-09-28 10:30:19 +08:00
|
|
|
|
with the prompt prefix and suffix (annotations). */
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
|
2017-09-28 10:30:19 +08:00
|
|
|
|
static std::string
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
top_level_prompt (void)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2016-03-10 02:24:59 +08:00
|
|
|
|
char *prompt;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
/* Give observers a chance of changing the prompt. E.g., the python
|
|
|
|
|
`gdb.prompt_hook' is installed as an observer. */
|
|
|
|
|
observer_notify_before_prompt (get_prompt ());
|
|
|
|
|
|
2016-03-10 02:24:59 +08:00
|
|
|
|
prompt = get_prompt ();
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
if (annotation_level >= 2)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
/* Prefix needs to have new line at end. */
|
2016-03-10 02:24:59 +08:00
|
|
|
|
const char prefix[] = "\n\032\032pre-prompt\n";
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
|
|
|
|
|
/* Suffix needs to have a new line at end and \032 \032 at
|
|
|
|
|
beginning. */
|
2016-03-10 02:24:59 +08:00
|
|
|
|
const char suffix[] = "\n\032\032prompt\n";
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2017-09-28 10:30:19 +08:00
|
|
|
|
return std::string (prefix) + prompt + suffix;
|
2016-03-10 02:24:59 +08:00
|
|
|
|
}
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
|
2017-09-28 10:30:19 +08:00
|
|
|
|
return prompt;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
1999-09-22 11:28:34 +08:00
|
|
|
|
|
2016-06-21 08:11:53 +08:00
|
|
|
|
/* See top.h. */
|
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.
2016-06-21 08:11:45 +08:00
|
|
|
|
|
2016-06-21 08:11:53 +08:00
|
|
|
|
struct ui *main_ui;
|
|
|
|
|
struct ui *current_ui;
|
|
|
|
|
struct ui *ui_list;
|
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.
2016-06-21 08:11:45 +08:00
|
|
|
|
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
/* Get a pointer to the current UI's line buffer. This is used to
|
2016-03-10 02:25:00 +08:00
|
|
|
|
construct a whole line of input from partial input. */
|
|
|
|
|
|
|
|
|
|
static struct buffer *
|
|
|
|
|
get_command_line_buffer (void)
|
|
|
|
|
{
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
return ¤t_ui->line_buffer;
|
2016-03-10 02:25:00 +08:00
|
|
|
|
}
|
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* When there is an event ready on the stdin file descriptor, instead
|
1999-09-22 11:28:34 +08:00
|
|
|
|
of calling readline directly throught the callback function, or
|
2016-03-10 02:24:59 +08:00
|
|
|
|
instead of calling gdb_readline_no_editing_callback, give gdb a
|
|
|
|
|
chance to detect errors and do something. */
|
|
|
|
|
|
1999-09-22 11:28:34 +08:00
|
|
|
|
void
|
1999-10-06 07:13:56 +08:00
|
|
|
|
stdin_event_handler (int error, gdb_client_data client_data)
|
1999-09-22 11:28:34 +08:00
|
|
|
|
{
|
2016-06-21 08:11:47 +08:00
|
|
|
|
struct ui *ui = (struct ui *) client_data;
|
|
|
|
|
|
1999-09-22 11:28:34 +08:00
|
|
|
|
if (error)
|
|
|
|
|
{
|
2016-06-21 08:11:54 +08:00
|
|
|
|
/* Switch to the main UI, so diagnostics always go there. */
|
|
|
|
|
current_ui = main_ui;
|
|
|
|
|
|
2016-06-21 08:11:47 +08:00
|
|
|
|
delete_file_handler (ui->input_fd);
|
2016-06-21 08:11:54 +08:00
|
|
|
|
if (main_ui == ui)
|
|
|
|
|
{
|
|
|
|
|
/* If stdin died, we may as well kill gdb. */
|
|
|
|
|
printf_unfiltered (_("error detected on stdin\n"));
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
quit_command ((char *) 0, 0);
|
2016-06-21 08:11:54 +08:00
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
/* Simply delete the UI. */
|
2017-09-30 12:35:58 +08:00
|
|
|
|
delete ui;
|
2016-06-21 08:11:54 +08:00
|
|
|
|
}
|
1999-09-22 11:28:34 +08:00
|
|
|
|
}
|
|
|
|
|
else
|
2014-11-23 18:03:39 +08:00
|
|
|
|
{
|
2016-06-21 08:11:54 +08:00
|
|
|
|
/* Switch to the UI whose input descriptor woke up the event
|
|
|
|
|
loop. */
|
|
|
|
|
current_ui = ui;
|
|
|
|
|
|
|
|
|
|
/* This makes sure a ^C immediately followed by further input is
|
|
|
|
|
always processed in that order. E.g,. with input like
|
|
|
|
|
"^Cprint 1\n", the SIGINT handler runs, marks the async
|
|
|
|
|
signal handler, and then select/poll may return with stdin
|
|
|
|
|
ready, instead of -1/EINTR. The
|
|
|
|
|
gdb.base/double-prompt-target-event-error.exp test exercises
|
|
|
|
|
this. */
|
Don't set immediate_quit in prompt_for_continue
immediate_quit used to be necessary back when prompt_for_continue used
blocking fread, but nowadays it uses gdb_readline_wrapper, which is
implemented in terms of a nested event loop, which already knows how
to react to SIGINT:
#0 throw_it (reason=RETURN_QUIT, error=GDB_NO_ERROR, fmt=0x9d6d7e "Quit", ap=0x7fffffffcb88)
at .../src/gdb/common/common-exceptions.c:324
#1 0x00000000007bab5d in throw_vquit (fmt=0x9d6d7e "Quit", ap=0x7fffffffcb88) at .../src/gdb/common/common-exceptions.c:366
#2 0x00000000007bac9f in throw_quit (fmt=0x9d6d7e "Quit") at .../src/gdb/common/common-exceptions.c:385
#3 0x0000000000773a2d in quit () at .../src/gdb/utils.c:1039
#4 0x000000000065d81b in async_request_quit (arg=0x0) at .../src/gdb/event-top.c:893
#5 0x000000000065c27b in invoke_async_signal_handlers () at .../src/gdb/event-loop.c:949
#6 0x000000000065aeef in gdb_do_one_event () at .../src/gdb/event-loop.c:280
#7 0x0000000000770838 in gdb_readline_wrapper (prompt=0x7fffffffcd40 "---Type <return> to continue, or q <return> to quit---")
at .../src/gdb/top.c:873
The need for the QUIT in stdin_event_handler is then exposed by the
gdb.base/double-prompt-target-event-error.exp test, which has:
# We're now stopped in a pagination query while handling a
# target event (printing where the program stopped). Quitting
# the pagination should result in only one prompt being
# output.
send_gdb "\003p 1\n"
Without that change we'd get:
Continuing.
---Type <return> to continue, or q <return> to quit---PASS: gdb.base/double-prompt-target-event-error.exp: ctrlc target event: continue: continue to pagination
^CpQuit
(gdb) 1
Undefined command: "1". Try "help".
(gdb) PASS: gdb.base/double-prompt-target-event-error.exp: ctrlc target event: continue: first prompt
ERROR: Undefined command "".
UNRESOLVED: gdb.base/double-prompt-target-event-error.exp: ctrlc target event: continue: no double prompt
Vs:
Continuing.
---Type <return> to continue, or q <return> to quit---PASS: gdb.base/double-prompt-target-event-error.exp: ctrlc target event: continue: continue to pagination
^CQuit
(gdb) p 1
$1 = 1
(gdb) PASS: gdb.base/double-prompt-target-event-error.exp: ctrlc target event: continue: first prompt
PASS: gdb.base/double-prompt-target-event-error.exp: ctrlc target event: continue: no double prompt
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* event-top.c (stdin_event_handler): Call QUIT;
(prompt_for_continue): Don't run with immediate_quit set.
2016-04-12 23:49:30 +08:00
|
|
|
|
QUIT;
|
|
|
|
|
|
2014-11-23 18:03:39 +08:00
|
|
|
|
do
|
|
|
|
|
{
|
|
|
|
|
call_stdin_event_handler_again_p = 0;
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
ui->call_readline (client_data);
|
2016-06-21 08:11:54 +08:00
|
|
|
|
}
|
|
|
|
|
while (call_stdin_event_handler_again_p != 0);
|
2014-11-23 18:03:39 +08:00
|
|
|
|
}
|
1999-09-22 11:28:34 +08:00
|
|
|
|
}
|
|
|
|
|
|
Fix PR gdb/20418 - Problems with synchronous commands and new-ui
When executing commands on a secondary UI running the MI interpreter,
some commands that should be synchronous are not. MI incorrectly
continues processing input right after the synchronous command is
sent, before the target stops.
The problem happens when we emit MI async events (=library-loaded,
etc.), and we go about restoring the previous terminal state, we end
up calling target_terminal_ours, which incorrectly always installs the
current UI's input_fd in the event loop... That is, code like this:
old_chain = make_cleanup_restore_target_terminal ();
target_terminal_ours_for_output ();
fprintf_unfiltered (mi->event_channel, "library-loaded");
...
do_cleanups (old_chain);
The fix is to move the add_file_handler/delete_file_handler calls out
of target_terminal_$foo, making these completely no-ops unless called
with the main UI as current UI.
gdb/ChangeLog:
2016-08-09 Pedro Alves <palves@redhat.com>
PR gdb/20418
* event-top.c (ui_register_input_event_handler)
(ui_unregister_input_event_handler): New functions.
(async_enable_stdin): Register input in the event loop.
(async_disable_stdin): Unregister input from the event loop.
(gdb_setup_readline): Register input in the event loop.
* infrun.c (check_curr_ui_sync_execution_done): Register input in
the event loop.
* target.c (target_terminal_inferior): Don't unregister input from
the event loop.
(target_terminal_ours): Don't register input in the event loop.
* target.h (target_terminal_inferior)
(target_terminal_ours_for_output, target_terminal_ours): Update
comments.
* top.h (ui_register_input_event_handler)
(ui_unregister_input_event_handler): New declarations.
* utils.c (ui_unregister_input_event_handler_cleanup)
(prepare_to_handle_input): New functions.
(defaulted_query, prompt_for_continue): Use
prepare_to_handle_input.
gdb/testsuite/ChangeLog:
2016-08-09 Pedro Alves <palves@redhat.com>
Simon Marchi <simon.marchi@ericsson.com>
PR gdb/20418
* gdb.mi/new-ui-mi-sync.c, gdb.mi/new-ui-mi-sync.exp: New files.
* lib/mi-support.exp (mi_expect_interrupt): Remove anchors.
2016-08-10 05:45:40 +08:00
|
|
|
|
/* See top.h. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
ui_register_input_event_handler (struct ui *ui)
|
|
|
|
|
{
|
|
|
|
|
add_file_handler (ui->input_fd, stdin_event_handler, ui);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See top.h. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
ui_unregister_input_event_handler (struct ui *ui)
|
|
|
|
|
{
|
|
|
|
|
delete_file_handler (ui->input_fd);
|
|
|
|
|
}
|
|
|
|
|
|
1999-09-29 05:55:21 +08:00
|
|
|
|
/* Re-enable stdin after the end of an execution command in
|
|
|
|
|
synchronous mode, or after an error from the target, and we aborted
|
2010-12-29 08:58:14 +08:00
|
|
|
|
the exec operation. */
|
1999-09-29 05:55:21 +08:00
|
|
|
|
|
|
|
|
|
void
|
2008-03-15 03:55:51 +08:00
|
|
|
|
async_enable_stdin (void)
|
1999-09-29 05:55:21 +08:00
|
|
|
|
{
|
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.
2016-06-21 08:11:51 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
|
|
|
|
|
|
|
|
|
if (ui->prompt_state == PROMPT_BLOCKED)
|
2008-03-15 02:57:44 +08:00
|
|
|
|
{
|
Remove make_cleanup_restore_target_terminal
This removes make_cleanup_restore_target_terminal and generally
C++-ifies target terminal handling. It changes all target_terminal_*
functions to be static members of a new target_terminal class and
changes the cleanup to be a scoped_* class.
make_cleanup_override_quit_handler is also removed in favor of simply
using scoped_restore.
Note that there are some files in this patch that I could not compile.
Considering that some of the rewrites were automated, and that none of
these files involed cleanups, I feel that this is relatively safe.
Regression tested by the buildbot.
gdb/ChangeLog
2017-09-20 Tom Tromey <tom@tromey.com>
* windows-nat.c (get_windows_debug_event, windows_wait)
(do_initial_windows_stuff, windows_attach): Update.
* utils.c (vwarning, internal_vproblem): Update.
(ui_unregister_input_event_handler_cleanup)
(prepare_to_handle_input): Remove.
(class scoped_input_handler): New.
(defaulted_query, prompt_for_continue): Update.
* tui/tui-hooks.c (tui_refresh_frame_and_register_information):
Update.
* top.c (undo_terminal_modifications_before_exit): Update.
* target/target.h (target_terminal_init, target_terminal_inferior)
(target_terminal_ours): Don't declare.
(class target_terminal): New.
* target.h (target_terminal_is_inferior, target_terminal_is_ours)
(target_terminal_ours_for_output)
(make_cleanup_restore_target_terminal): Don't declare.
(target_terminal_info): Remove.
* target.c (enum terminal_state, terminal_state): Remove.
(target_terminal::terminal_state): Define.
(target_terminal::init): Rename from target_terminal_init.
(target_terminal::inferior): Rename from
target_terminal_inferior.
(target_terminal::ours): Rename from target_terminal_ours.
(target_terminal::ours_for_output): Rename from
target_terminal_ours_for_output.
(target_terminal::info): New method.
(cleanup_restore_target_terminal)
(make_cleanup_restore_target_terminal): Remove.
* solib.c (handle_solib_event): Update.
* remote.c (remote_serial_quit_handler): Update.
(remote_terminal_inferior, remote_wait_as): Update.
* record-full.c (record_full_wait_1): Update.
* nto-procfs.c (procfs_create_inferior): Update.
* nat/fork-inferior.c (startup_inferior): Update.
* mi/mi-interp.c (mi_new_thread, mi_thread_exit)
(mi_record_changed, mi_inferior_added, mi_inferior_appeared)
(mi_inferior_exit, mi_inferior_removed, mi_traceframe_changed)
(mi_tsv_created, mi_tsv_deleted, mi_tsv_modified)
(mi_breakpoint_created, mi_breakpoint_deleted)
(mi_breakpoint_modified, mi_on_resume, mi_solib_loaded)
(mi_solib_unloaded, mi_command_param_changed, mi_memory_changed)
(mi_user_selected_context_changed, report_initial_inferior):
Update.
* linux-nat.c (linux_nat_attach, linux_nat_terminal_ours)
(linux_nat_terminal_inferior): Update.
* infrun.c (follow_fork_inferior)
(handle_vfork_child_exec_or_exit, do_target_resume)
(check_curr_ui_sync_execution_done, handle_inferior_event_1)
(handle_signal_stop, maybe_remove_breakpoints, normal_stop):
Update.
* inflow.c (child_terminal_init, info_terminal_command): Update.
* infcmd.c (post_create_inferior, continue_1, prepare_one_step)
(attach_command): Update.
* infcall.c (call_thread_fsm_should_stop): Update.
* gnu-nat.c (gnu_attach): Update.
* extension.c (struct active_ext_lang_state)
(restore_active_ext_lang): Update.
* exceptions.c (print_flush): Update.
* event-top.c (async_enable_stdin, default_quit_handler): Update.
(struct quit_handler_cleanup_data, restore_quit_handler)
(restore_quit_handler_dtor, make_cleanup_override_quit_handler):
Remove.
* cp-support.c (gdb_demangle): Update.
* breakpoint.c (update_inserted_breakpoint_locations)
(insert_breakpoint_locations, handle_jit_event)
(disable_breakpoints_in_unloaded_shlib): Update.
* annotate.c (annotate_breakpoints_invalid)
(annotate_frames_invalid): Update.
gdb/gdbserver/ChangeLog
2017-09-20 Tom Tromey <tom@tromey.com>
* target.c (target_terminal::terminal_state): Define.
(target_terminal::init): Rename from target_terminal_init.
(target_terminal::inferior): Rename from
target_terminal_inferior.
(target_terminal::ours): Rename from target_terminal_ours.
(target_terminal::ours_for_output, target_terminal::info): New.
2017-09-20 11:56:36 +08:00
|
|
|
|
target_terminal::ours ();
|
Fix PR gdb/20418 - Problems with synchronous commands and new-ui
When executing commands on a secondary UI running the MI interpreter,
some commands that should be synchronous are not. MI incorrectly
continues processing input right after the synchronous command is
sent, before the target stops.
The problem happens when we emit MI async events (=library-loaded,
etc.), and we go about restoring the previous terminal state, we end
up calling target_terminal_ours, which incorrectly always installs the
current UI's input_fd in the event loop... That is, code like this:
old_chain = make_cleanup_restore_target_terminal ();
target_terminal_ours_for_output ();
fprintf_unfiltered (mi->event_channel, "library-loaded");
...
do_cleanups (old_chain);
The fix is to move the add_file_handler/delete_file_handler calls out
of target_terminal_$foo, making these completely no-ops unless called
with the main UI as current UI.
gdb/ChangeLog:
2016-08-09 Pedro Alves <palves@redhat.com>
PR gdb/20418
* event-top.c (ui_register_input_event_handler)
(ui_unregister_input_event_handler): New functions.
(async_enable_stdin): Register input in the event loop.
(async_disable_stdin): Unregister input from the event loop.
(gdb_setup_readline): Register input in the event loop.
* infrun.c (check_curr_ui_sync_execution_done): Register input in
the event loop.
* target.c (target_terminal_inferior): Don't unregister input from
the event loop.
(target_terminal_ours): Don't register input in the event loop.
* target.h (target_terminal_inferior)
(target_terminal_ours_for_output, target_terminal_ours): Update
comments.
* top.h (ui_register_input_event_handler)
(ui_unregister_input_event_handler): New declarations.
* utils.c (ui_unregister_input_event_handler_cleanup)
(prepare_to_handle_input): New functions.
(defaulted_query, prompt_for_continue): Use
prepare_to_handle_input.
gdb/testsuite/ChangeLog:
2016-08-09 Pedro Alves <palves@redhat.com>
Simon Marchi <simon.marchi@ericsson.com>
PR gdb/20418
* gdb.mi/new-ui-mi-sync.c, gdb.mi/new-ui-mi-sync.exp: New files.
* lib/mi-support.exp (mi_expect_interrupt): Remove anchors.
2016-08-10 05:45:40 +08:00
|
|
|
|
ui_register_input_event_handler (ui);
|
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.
2016-06-21 08:11:51 +08:00
|
|
|
|
ui->prompt_state = PROMPT_NEEDED;
|
2008-03-15 02:57:44 +08:00
|
|
|
|
}
|
1999-09-29 05:55:21 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Disable reads from stdin (the console) marking the command as
|
2010-12-29 08:58:14 +08:00
|
|
|
|
synchronous. */
|
1999-09-29 05:55:21 +08:00
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
async_disable_stdin (void)
|
|
|
|
|
{
|
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.
2016-06-21 08:11:51 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
|
|
|
|
|
|
|
|
|
ui->prompt_state = PROMPT_BLOCKED;
|
Fix PR gdb/20418 - Problems with synchronous commands and new-ui
When executing commands on a secondary UI running the MI interpreter,
some commands that should be synchronous are not. MI incorrectly
continues processing input right after the synchronous command is
sent, before the target stops.
The problem happens when we emit MI async events (=library-loaded,
etc.), and we go about restoring the previous terminal state, we end
up calling target_terminal_ours, which incorrectly always installs the
current UI's input_fd in the event loop... That is, code like this:
old_chain = make_cleanup_restore_target_terminal ();
target_terminal_ours_for_output ();
fprintf_unfiltered (mi->event_channel, "library-loaded");
...
do_cleanups (old_chain);
The fix is to move the add_file_handler/delete_file_handler calls out
of target_terminal_$foo, making these completely no-ops unless called
with the main UI as current UI.
gdb/ChangeLog:
2016-08-09 Pedro Alves <palves@redhat.com>
PR gdb/20418
* event-top.c (ui_register_input_event_handler)
(ui_unregister_input_event_handler): New functions.
(async_enable_stdin): Register input in the event loop.
(async_disable_stdin): Unregister input from the event loop.
(gdb_setup_readline): Register input in the event loop.
* infrun.c (check_curr_ui_sync_execution_done): Register input in
the event loop.
* target.c (target_terminal_inferior): Don't unregister input from
the event loop.
(target_terminal_ours): Don't register input in the event loop.
* target.h (target_terminal_inferior)
(target_terminal_ours_for_output, target_terminal_ours): Update
comments.
* top.h (ui_register_input_event_handler)
(ui_unregister_input_event_handler): New declarations.
* utils.c (ui_unregister_input_event_handler_cleanup)
(prepare_to_handle_input): New functions.
(defaulted_query, prompt_for_continue): Use
prepare_to_handle_input.
gdb/testsuite/ChangeLog:
2016-08-09 Pedro Alves <palves@redhat.com>
Simon Marchi <simon.marchi@ericsson.com>
PR gdb/20418
* gdb.mi/new-ui-mi-sync.c, gdb.mi/new-ui-mi-sync.exp: New files.
* lib/mi-support.exp (mi_expect_interrupt): Remove anchors.
2016-08-10 05:45:40 +08:00
|
|
|
|
delete_file_handler (ui->input_fd);
|
1999-09-29 05:55:21 +08:00
|
|
|
|
}
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
1999-09-29 05:55:21 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Handle a gdb command line. This function is called when
|
|
|
|
|
handle_line_of_input has concatenated one or more input lines into
|
|
|
|
|
a whole command. */
|
|
|
|
|
|
|
|
|
|
void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
command_handler (char *command)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
Make instream be per UI
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-script.c (execute_user_command, read_next_line)
(read_next_line): Adjust to per-UI instream.
* event-top.c (stdin_event_handler, command_handler)
(handle_line_of_input, command_line_handler)
(gdb_readline_no_editing_callback, async_sigterm_handler)
(gdb_setup_readline): Likewise.
* inflow.c: Include top.h.
(gdb_has_a_terminal, child_terminal_init_with_pgrp)
(gdb_save_tty_state, child_terminal_inferior)
(child_terminal_ours_1, copy_terminal_info): Use the main UI.
(initialize_stdin_serial): Adjust to per-UI instream.
* main.c (captured_command_loop, captured_main): Adjust to per-UI
instream.
* mi/mi-interp.c (mi_execute_command_wrapper): Likewise.
* python/python.c (python_interactive_command): Likewise.
* terminal.h (struct ui): Forward declare.
(initialize_stdin_serial): Add struct ui parameter.
* top.c (instream): Delete.
(do_restore_instream_cleanup, read_command_file, dont_repeat)
(gdb_readline_no_editing, command_line_input)
(input_from_terminal_p, gdb_init): Adjust to per-UI instream.
* top.h (struct ui) <instream>: New field.
(instream): Delete declaration.
(quit): Adjust to per-UI instream.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.gdb/selftest.exp (do_steps_and_nexts): Add new regexp.
2016-06-21 08:11:46 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
2016-03-10 02:25:00 +08:00
|
|
|
|
char *c;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
if (ui->instream == ui->stdin_stream)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
reinitialize_more_filter ();
|
|
|
|
|
|
2016-09-25 11:10:45 +08:00
|
|
|
|
scoped_command_stats stat_reporter (true);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Do not execute commented lines. */
|
|
|
|
|
for (c = command; *c == ' ' || *c == '\t'; c++)
|
|
|
|
|
;
|
|
|
|
|
if (c[0] != '#')
|
|
|
|
|
{
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
execute_command (command, ui->instream == ui->stdin_stream);
|
1999-07-08 04:19:36 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Do any commands attached to breakpoint we stopped at. */
|
|
|
|
|
bpstat_do_actions ();
|
|
|
|
|
}
|
1999-07-06 01:58:44 +08:00
|
|
|
|
}
|
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Append RL, an input line returned by readline or one of its
|
|
|
|
|
emulations, to CMD_LINE_BUFFER. Returns the command line if we
|
|
|
|
|
have a whole command line ready to be processed by the command
|
|
|
|
|
interpreter or NULL if the command line isn't complete yet (input
|
|
|
|
|
line ends in a backslash). Takes ownership of RL. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
static char *
|
|
|
|
|
command_line_append_input_line (struct buffer *cmd_line_buffer, char *rl)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
char *cmd;
|
|
|
|
|
size_t len;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
len = strlen (rl);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
if (len > 0 && rl[len - 1] == '\\')
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Don't copy the backslash and wait for more. */
|
|
|
|
|
buffer_grow (cmd_line_buffer, rl, len - 1);
|
|
|
|
|
cmd = NULL;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
2016-03-10 02:25:00 +08:00
|
|
|
|
else
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Copy whole line including terminating null, and we're
|
|
|
|
|
done. */
|
|
|
|
|
buffer_grow (cmd_line_buffer, rl, len + 1);
|
|
|
|
|
cmd = cmd_line_buffer->buffer;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Allocated in readline. */
|
|
|
|
|
xfree (rl);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
return cmd;
|
|
|
|
|
}
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Handle a line of input coming from readline.
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
If the read line ends with a continuation character (backslash),
|
|
|
|
|
save the partial input in CMD_LINE_BUFFER (except the backslash),
|
|
|
|
|
and return NULL. Otherwise, save the partial input and return a
|
|
|
|
|
pointer to CMD_LINE_BUFFER's buffer (null terminated), indicating a
|
|
|
|
|
whole command line is ready to be executed.
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
Returns EOF on end of file.
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
If REPEAT, handle command repetitions:
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
- If the input command line is NOT empty, the command returned is
|
|
|
|
|
copied into the global 'saved_command_line' var so that it can
|
|
|
|
|
be repeated later.
|
2002-03-28 05:20:15 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
- OTOH, if the input command line IS empty, return the previously
|
|
|
|
|
saved command instead of the empty input line.
|
|
|
|
|
*/
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
char *
|
|
|
|
|
handle_line_of_input (struct buffer *cmd_line_buffer,
|
-Wwrite-strings: The Rest
This is the remainder boring constification that all looks more of less
borderline obvious IMO.
gdb/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* ada-exp.y (yyerror): Constify.
* ada-lang.c (bound_name, get_selections)
(ada_variant_discrim_type)
(ada_variant_discrim_name, ada_value_struct_elt)
(ada_lookup_struct_elt_type, is_unchecked_variant)
(ada_which_variant_applies, standard_exc, ada_get_next_arg)
(catch_ada_exception_command_split)
(catch_ada_assert_command_split, catch_assert_command)
(ada_op_name): Constify.
* ada-lang.h (ada_yyerror, get_selections)
(ada_variant_discrim_name, ada_value_struct_elt): Constify.
* arc-tdep.c (arc_print_frame_cache): Constify.
* arm-tdep.c (arm_skip_stub): Constify.
* ax-gdb.c (gen_binop, gen_struct_ref_recursive, gen_struct_ref)
(gen_aggregate_elt_ref): Constify.
* bcache.c (print_bcache_statistics): Constify.
* bcache.h (print_bcache_statistics): Constify.
* break-catch-throw.c (catch_exception_command_1):
* breakpoint.c (struct ep_type_description::description):
Constify.
(add_solib_catchpoint): Constify.
(catch_fork_command_1): Add cast.
(add_catch_command): Constify.
* breakpoint.h (add_catch_command, add_solib_catchpoint):
Constify.
* bsd-uthread.c (bsd_uthread_state): Constify.
* buildsym.c (patch_subfile_names): Constify.
* buildsym.h (next_symbol_text_func, patch_subfile_names):
Constify.
* c-exp.y (yyerror): Constify.
(token::oper): Constify.
* c-lang.h (c_yyerror, cp_print_class_member): Constify.
* c-varobj.c (cplus_describe_child): Constify.
* charset.c (find_charset_names): Add cast.
(find_charset_names): Constify array and add const_cast.
* cli/cli-cmds.c (complete_command, cd_command): Constify.
(edit_command): Constify.
* cli/cli-decode.c (lookup_cmd): Constify.
* cli/cli-dump.c (dump_memory_command, dump_value_command):
Constify.
(struct dump_context): Constify.
(add_dump_command, restore_command): Constify.
* cli/cli-script.c (get_command_line): Constify.
* cli/cli-script.h (get_command_line): Constify.
* cli/cli-utils.c (check_for_argument): Constify.
* cli/cli-utils.h (check_for_argument): Constify.
* coff-pe-read.c (struct read_pe_section_data): Constify.
* command.h (lookup_cmd): Constify.
* common/print-utils.c (decimal2str): Constify.
* completer.c (gdb_print_filename): Constify.
* corefile.c (set_gnutarget): Constify.
* cp-name-parser.y (yyerror): Constify.
* cp-valprint.c (cp_print_class_member): Constify.
* cris-tdep.c (cris_register_name, crisv32_register_name):
Constify.
* d-exp.y (yyerror): Constify.
(struct token::oper): Constify.
* d-lang.h (d_yyerror): Constify.
* dbxread.c (struct header_file_location::name): Constify.
(add_old_header_file, add_new_header_file, last_function_name)
(dbx_next_symbol_text, add_bincl_to_list)
(find_corresponding_bincl_psymtab, set_namestring)
(find_stab_function_addr, read_dbx_symtab, start_psymtab)
(dbx_end_psymtab, read_ofile_symtab, process_one_symbol):
* defs.h (command_line_input, print_address_symbolic)
(deprecated_readline_begin_hook): Constify.
* dwarf2read.c (anonymous_struct_prefix, dwarf_bool_name):
Constify.
* event-top.c (handle_line_of_input): Constify and add cast.
* exceptions.c (catch_errors): Constify.
* exceptions.h (catch_errors): Constify.
* expprint.c (print_subexp_standard, op_string, op_name)
(op_name_standard, dump_raw_expression, dump_raw_expression):
* expression.h (op_name, op_string, dump_raw_expression):
Constify.
* f-exp.y (yyerror): Constify.
(struct token::oper): Constify.
(struct f77_boolean_val::name): Constify.
* f-lang.c (f_word_break_characters): Constify.
* f-lang.h (f_yyerror): Constify.
* fork-child.c (fork_inferior): Add cast.
* frv-tdep.c (struct gdbarch_tdep::register_names): Constify.
(new_variant): Constify.
* gdbarch.sh (pstring_ptr, pstring_list): Constify.
* gdbarch.c: Regenerate.
* gdbcore.h (set_gnutarget): Constify.
* go-exp.y (yyerror): Constify.
(token::oper): Constify.
* go-lang.h (go_yyerror): Constify.
* go32-nat.c (go32_sysinfo): Constify.
* guile/scm-breakpoint.c (gdbscm_breakpoint_expression): Constify.
* guile/scm-cmd.c (cmdscm_function): Constify.
* guile/scm-param.c (pascm_param_value): Constify.
* h8300-tdep.c (h8300_register_name, h8300s_register_name)
(h8300sx_register_name): Constify.
* hppa-tdep.c (hppa32_register_name, hppa64_register_name):
Constify.
* ia64-tdep.c (ia64_register_names): Constify.
* infcmd.c (construct_inferior_arguments): Constify.
(path_command, attach_post_wait): Constify.
* language.c (show_range_command, show_case_command)
(unk_lang_error): Constify.
* language.h (language_defn::la_error)
(language_defn::la_name_of_this): Constify.
* linespec.c (decode_line_2): Constify.
* linux-thread-db.c (thread_db_err_str): Constify.
* lm32-tdep.c (lm32_register_name): Constify.
* m2-exp.y (yyerror): Constify.
* m2-lang.h (m2_yyerror): Constify.
* m32r-tdep.c (m32r_register_names): Constify and make static.
* m68hc11-tdep.c (m68hc11_register_names): Constify.
* m88k-tdep.c (m88k_register_name): Constify.
* macroexp.c (appendmem): Constify.
* mdebugread.c (fdr_name, add_data_symbol, parse_type)
(upgrade_type, parse_external, parse_partial_symbols)
(mdebug_next_symbol_text, cross_ref, mylookup_symbol, new_psymtab)
(new_symbol): Constify.
* memattr.c (mem_info_command): Constify.
* mep-tdep.c (register_name_from_keyword): Constify.
* mi/mi-cmd-env.c (mi_cmd_env_path, _initialize_mi_cmd_env):
Constify.
* mi/mi-cmd-stack.c (list_args_or_locals): Constify.
* mi/mi-cmd-var.c (mi_cmd_var_show_attributes): Constify.
* mi/mi-main.c (captured_mi_execute_command): Constify and add
cast.
(mi_execute_async_cli_command): Constify.
* mips-tdep.c (mips_register_name): Constify.
* mn10300-tdep.c (register_name, mn10300_generic_register_name)
(am33_register_name, am33_2_register_name)
* moxie-tdep.c (moxie_register_names): Constify.
* nat/linux-osdata.c (osdata_type): Constify fields.
* nto-tdep.c (nto_parse_redirection): Constify.
* objc-lang.c (lookup_struct_typedef, lookup_objc_class)
(lookup_child_selector): Constify.
(objc_methcall::name): Constify.
* objc-lang.h (lookup_objc_class, lookup_child_selector)
(lookup_struct_typedef): Constify.
* objfiles.c (pc_in_section): Constify.
* objfiles.h (pc_in_section): Constify.
* p-exp.y (struct token::oper): Constify.
(yyerror): Constify.
* p-lang.h (pascal_yyerror): Constify.
* parser-defs.h (op_name_standard): Constify.
(op_print::string): Constify.
(exp_descriptor::op_name): Constify.
* printcmd.c (print_address_symbolic): Constify.
* psymtab.c (print_partial_symbols): Constify.
* python/py-breakpoint.c (stop_func): Constify.
(bppy_get_expression): Constify.
* python/py-cmd.c (cmdpy_completer::name): Constify.
(cmdpy_function): Constify.
* python/py-event.c (evpy_add_attribute)
(gdbpy_initialize_event_generic): Constify.
* python/py-event.h (evpy_add_attribute)
(gdbpy_initialize_event_generic): Constify.
* python/py-evts.c (add_new_registry): Constify.
* python/py-finishbreakpoint.c (outofscope_func): Constify.
* python/py-framefilter.c (get_py_iter_from_func): Constify.
* python/py-inferior.c (get_buffer): Add cast.
* python/py-param.c (parm_constant::name): Constify.
* python/py-unwind.c (fprint_frame_id): Constify.
* python/python.c (gdbpy_parameter_value): Constify.
* remote-fileio.c (remote_fio_func_map): Make 'name' const.
* remote.c (memory_packet_config::name): Constify.
(show_packet_config_cmd, remote_write_bytes)
(remote_buffer_add_string):
* reverse.c (exec_reverse_once): Constify.
* rs6000-tdep.c (variant::name, variant::description): Constify.
* rust-exp.y (rustyyerror): Constify.
* rust-lang.c (rust_op_name): Constify.
* rust-lang.h (rustyyerror): Constify.
* serial.h (serial_ops::name): Constify.
* sh-tdep.c (sh_sh_register_name, sh_sh3_register_name)
(sh_sh3e_register_name, sh_sh2e_register_name)
(sh_sh2a_register_name, sh_sh2a_nofpu_register_name)
(sh_sh_dsp_register_name, sh_sh3_dsp_register_name)
(sh_sh4_register_name, sh_sh4_nofpu_register_name)
(sh_sh4al_dsp_register_name): Constify.
* sh64-tdep.c (sh64_register_name): Constify.
* solib-darwin.c (lookup_symbol_from_bfd): Constify.
* spu-tdep.c (spu_register_name, info_spu_dma_cmdlist): Constify.
* stabsread.c (patch_block_stabs, read_type_number)
(ref_map::stabs, ref_add, process_reference)
(symbol_reference_defined, define_symbol, define_symbol)
(error_type, read_type, read_member_functions, read_cpp_abbrev)
(read_one_struct_field, read_struct_fields, read_baseclasses)
(read_tilde_fields, read_struct_type, read_array_type)
(read_enum_type, read_sun_builtin_type, read_sun_floating_type)
(read_huge_number, read_range_type, read_args, common_block_start)
(find_name_end): Constify.
* stabsread.h (common_block_start, define_symbol)
(process_one_symbol, symbol_reference_defined, ref_add):
* symfile.c (get_section_index, add_symbol_file_command):
* symfile.h (get_section_index): Constify.
* target-descriptions.c (tdesc_type::name): Constify.
(tdesc_free_type): Add cast.
* target.c (find_default_run_target):
(add_deprecated_target_alias, find_default_run_target)
(target_announce_detach): Constify.
(do_option): Constify.
* target.h (add_deprecated_target_alias): Constify.
* thread.c (print_thread_info_1): Constify.
* top.c (deprecated_readline_begin_hook, command_line_input):
Constify.
(init_main): Add casts.
* top.h (handle_line_of_input): Constify.
* tracefile-tfile.c (tfile_write_uploaded_tsv): Constify.
* tracepoint.c (tvariables_info_1, trace_status_mi): Constify.
(tfind_command): Rename to ...
(tfind_command_1): ... this and constify.
(tfind_command): New function.
(tfind_end_command, tfind_start_command): Adjust.
(encode_source_string): Constify.
* tracepoint.h (encode_source_string): Constify.
* tui/tui-data.c (tui_partial_win_by_name): Constify.
* tui/tui-data.h (tui_partial_win_by_name): Constify.
* tui/tui-source.c (tui_set_source_content_nil): Constify.
* tui/tui-source.h (tui_set_source_content_nil): Constify.
* tui/tui-win.c (parse_scrolling_args): Constify.
* tui/tui-windata.c (tui_erase_data_content): Constify.
* tui/tui-windata.h (tui_erase_data_content): Constify.
* tui/tui-winsource.c (tui_erase_source_content): Constify.
* tui/tui.c (tui_enable): Add cast.
* utils.c (defaulted_query): Constify.
(init_page_info): Add cast.
(puts_debug, subset_compare): Constify.
* utils.h (subset_compare): Constify.
* varobj.c (varobj_format_string): Constify.
* varobj.h (varobj_format_string): Constify.
* vax-tdep.c (vax_register_name): Constify.
* windows-nat.c (windows_detach): Constify.
* xcoffread.c (process_linenos, xcoff_next_symbol_text): Constify.
* xml-support.c (gdb_xml_end_element): Constify.
* xml-tdesc.c (tdesc_start_reg): Constify.
* xstormy16-tdep.c (xstormy16_register_name): Constify.
* xtensa-tdep.c (xtensa_find_register_by_name): Constify.
* xtensa-tdep.h (xtensa_register_t::name): Constify.
gdb/gdbserver/ChangeLog:
2017-04-05 Pedro Alves <palves@redhat.com>
* gdbreplay.c (sync_error): Constify.
* linux-x86-low.c (push_opcode): Constify.
2017-04-06 02:21:37 +08:00
|
|
|
|
char *rl, int repeat, const char *annotation_suffix)
|
2016-03-10 02:25:00 +08:00
|
|
|
|
{
|
Make instream be per UI
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-script.c (execute_user_command, read_next_line)
(read_next_line): Adjust to per-UI instream.
* event-top.c (stdin_event_handler, command_handler)
(handle_line_of_input, command_line_handler)
(gdb_readline_no_editing_callback, async_sigterm_handler)
(gdb_setup_readline): Likewise.
* inflow.c: Include top.h.
(gdb_has_a_terminal, child_terminal_init_with_pgrp)
(gdb_save_tty_state, child_terminal_inferior)
(child_terminal_ours_1, copy_terminal_info): Use the main UI.
(initialize_stdin_serial): Adjust to per-UI instream.
* main.c (captured_command_loop, captured_main): Adjust to per-UI
instream.
* mi/mi-interp.c (mi_execute_command_wrapper): Likewise.
* python/python.c (python_interactive_command): Likewise.
* terminal.h (struct ui): Forward declare.
(initialize_stdin_serial): Add struct ui parameter.
* top.c (instream): Delete.
(do_restore_instream_cleanup, read_command_file, dont_repeat)
(gdb_readline_no_editing, command_line_input)
(input_from_terminal_p, gdb_init): Adjust to per-UI instream.
* top.h (struct ui) <instream>: New field.
(instream): Delete declaration.
(quit): Adjust to per-UI instream.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.gdb/selftest.exp (do_steps_and_nexts): Add new regexp.
2016-06-21 08:11:46 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
int from_tty = ui->instream == ui->stdin_stream;
|
2016-03-10 02:25:00 +08:00
|
|
|
|
char *p1;
|
|
|
|
|
char *cmd;
|
|
|
|
|
|
|
|
|
|
if (rl == NULL)
|
|
|
|
|
return (char *) EOF;
|
|
|
|
|
|
|
|
|
|
cmd = command_line_append_input_line (cmd_line_buffer, rl);
|
|
|
|
|
if (cmd == NULL)
|
|
|
|
|
return NULL;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* We have a complete command line now. Prepare for the next
|
|
|
|
|
command, but leave ownership of memory to the buffer . */
|
|
|
|
|
cmd_line_buffer->used_size = 0;
|
|
|
|
|
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
if (from_tty && annotation_level > 1)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
printf_unfiltered (("\n\032\032post-"));
|
|
|
|
|
puts_unfiltered (annotation_suffix);
|
|
|
|
|
printf_unfiltered (("\n"));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#define SERVER_COMMAND_PREFIX "server "
|
|
|
|
|
if (startswith (cmd, SERVER_COMMAND_PREFIX))
|
|
|
|
|
{
|
|
|
|
|
/* Note that we don't set `saved_command_line'. Between this
|
|
|
|
|
and the check in dont_repeat, this insures that repeating
|
|
|
|
|
will still do the right thing. */
|
|
|
|
|
return cmd + strlen (SERVER_COMMAND_PREFIX);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Do history expansion if that is wished. */
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
if (history_expansion_p && from_tty && input_interactive_p (current_ui))
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
|
|
|
|
char *history_value;
|
|
|
|
|
int expanded;
|
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
expanded = history_expand (cmd, &history_value);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
if (expanded)
|
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
size_t len;
|
|
|
|
|
|
1999-05-12 04:29:07 +08:00
|
|
|
|
/* Print the changes. */
|
|
|
|
|
printf_unfiltered ("%s\n", history_value);
|
|
|
|
|
|
|
|
|
|
/* If there was an error, call this function again. */
|
|
|
|
|
if (expanded < 0)
|
|
|
|
|
{
|
2000-12-15 09:01:51 +08:00
|
|
|
|
xfree (history_value);
|
2016-03-10 02:25:00 +08:00
|
|
|
|
return cmd;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
2016-03-10 02:25:00 +08:00
|
|
|
|
|
|
|
|
|
/* history_expand returns an allocated string. Just replace
|
|
|
|
|
our buffer with it. */
|
|
|
|
|
len = strlen (history_value);
|
|
|
|
|
xfree (buffer_finish (cmd_line_buffer));
|
|
|
|
|
cmd_line_buffer->buffer = history_value;
|
|
|
|
|
cmd_line_buffer->buffer_size = len + 1;
|
|
|
|
|
cmd = history_value;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* If we just got an empty line, and that is supposed to repeat the
|
2016-03-10 02:25:00 +08:00
|
|
|
|
previous command, return the previously saved command. */
|
|
|
|
|
for (p1 = cmd; *p1 == ' ' || *p1 == '\t'; p1++)
|
|
|
|
|
;
|
|
|
|
|
if (repeat && *p1 == '\0')
|
|
|
|
|
return saved_command_line;
|
|
|
|
|
|
|
|
|
|
/* Add command to history if appropriate. Note: lines consisting
|
|
|
|
|
solely of comments are also added to the command history. This
|
|
|
|
|
is useful when you type a command, and then realize you don't
|
|
|
|
|
want to execute it quite yet. You can comment out the command
|
|
|
|
|
and then later fetch it from the value history and remove the
|
|
|
|
|
'#'. The kill ring is probably better, but some people are in
|
|
|
|
|
the habit of commenting things out. */
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
if (*cmd != '\0' && from_tty && input_interactive_p (current_ui))
|
2016-03-10 02:25:00 +08:00
|
|
|
|
gdb_add_history (cmd);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Save into global buffer if appropriate. */
|
|
|
|
|
if (repeat)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
xfree (saved_command_line);
|
|
|
|
|
saved_command_line = xstrdup (cmd);
|
|
|
|
|
return saved_command_line;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
2016-03-10 02:25:00 +08:00
|
|
|
|
else
|
|
|
|
|
return cmd;
|
|
|
|
|
}
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* Handle a complete line of input. This is called by the callback
|
|
|
|
|
mechanism within the readline library. Deal with incomplete
|
|
|
|
|
commands as well, by saving the partial input in a global
|
|
|
|
|
buffer.
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
NOTE: This is the asynchronous version of the command_line_input
|
|
|
|
|
function. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
void
|
|
|
|
|
command_line_handler (char *rl)
|
|
|
|
|
{
|
|
|
|
|
struct buffer *line_buffer = get_command_line_buffer ();
|
Make instream be per UI
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-script.c (execute_user_command, read_next_line)
(read_next_line): Adjust to per-UI instream.
* event-top.c (stdin_event_handler, command_handler)
(handle_line_of_input, command_line_handler)
(gdb_readline_no_editing_callback, async_sigterm_handler)
(gdb_setup_readline): Likewise.
* inflow.c: Include top.h.
(gdb_has_a_terminal, child_terminal_init_with_pgrp)
(gdb_save_tty_state, child_terminal_inferior)
(child_terminal_ours_1, copy_terminal_info): Use the main UI.
(initialize_stdin_serial): Adjust to per-UI instream.
* main.c (captured_command_loop, captured_main): Adjust to per-UI
instream.
* mi/mi-interp.c (mi_execute_command_wrapper): Likewise.
* python/python.c (python_interactive_command): Likewise.
* terminal.h (struct ui): Forward declare.
(initialize_stdin_serial): Add struct ui parameter.
* top.c (instream): Delete.
(do_restore_instream_cleanup, read_command_file, dont_repeat)
(gdb_readline_no_editing, command_line_input)
(input_from_terminal_p, gdb_init): Adjust to per-UI instream.
* top.h (struct ui) <instream>: New field.
(instream): Delete declaration.
(quit): Adjust to per-UI instream.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.gdb/selftest.exp (do_steps_and_nexts): Add new regexp.
2016-06-21 08:11:46 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
2016-03-10 02:25:00 +08:00
|
|
|
|
char *cmd;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
cmd = handle_line_of_input (line_buffer, rl, 1, "prompt");
|
2016-03-10 02:25:00 +08:00
|
|
|
|
if (cmd == (char *) EOF)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
/* stdin closed. The connection with the terminal is gone.
|
|
|
|
|
This happens at the end of a testsuite run, after Expect has
|
|
|
|
|
hung up but GDB is still alive. In such a case, we just quit
|
|
|
|
|
gdb killing the inferior program too. */
|
|
|
|
|
printf_unfiltered ("quit\n");
|
2017-04-06 02:21:36 +08:00
|
|
|
|
execute_command ((char *) "quit", 1);
|
2016-03-10 02:25:00 +08:00
|
|
|
|
}
|
|
|
|
|
else if (cmd == NULL)
|
|
|
|
|
{
|
|
|
|
|
/* We don't have a full line yet. Print an empty prompt. */
|
|
|
|
|
display_gdb_prompt ("");
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
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.
2016-06-21 08:11:51 +08:00
|
|
|
|
ui->prompt_state = PROMPT_NEEDED;
|
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
command_handler (cmd);
|
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.
2016-06-21 08:11:51 +08:00
|
|
|
|
|
|
|
|
|
if (ui->prompt_state != PROMPTED)
|
|
|
|
|
display_gdb_prompt (0);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Does reading of input from terminal w/o the editing features
|
2016-03-10 02:24:59 +08:00
|
|
|
|
provided by the readline library. Calls the line input handler
|
|
|
|
|
once we have a whole input line. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
1999-06-29 00:06:02 +08:00
|
|
|
|
void
|
2016-03-10 02:24:59 +08:00
|
|
|
|
gdb_readline_no_editing_callback (gdb_client_data client_data)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
|
|
|
|
int c;
|
|
|
|
|
char *result;
|
2016-03-10 02:25:00 +08:00
|
|
|
|
struct buffer line_buffer;
|
1999-08-17 03:57:19 +08:00
|
|
|
|
static int done_once = 0;
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
1999-08-17 03:57:19 +08:00
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
buffer_init (&line_buffer);
|
|
|
|
|
|
1999-08-17 03:57:19 +08:00
|
|
|
|
/* Unbuffer the input stream, so that, later on, the calls to fgetc
|
2010-12-29 08:58:14 +08:00
|
|
|
|
fetch only one char at the time from the stream. The fgetc's will
|
1999-08-17 03:57:19 +08:00
|
|
|
|
get up to the first newline, but there may be more chars in the
|
2010-12-29 08:58:14 +08:00
|
|
|
|
stream after '\n'. If we buffer the input and fgetc drains the
|
1999-08-17 03:57:19 +08:00
|
|
|
|
stream, getting stuff beyond the newline as well, a select, done
|
2010-12-29 08:58:14 +08:00
|
|
|
|
afterwards will not trigger. */
|
Make instream be per UI
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-script.c (execute_user_command, read_next_line)
(read_next_line): Adjust to per-UI instream.
* event-top.c (stdin_event_handler, command_handler)
(handle_line_of_input, command_line_handler)
(gdb_readline_no_editing_callback, async_sigterm_handler)
(gdb_setup_readline): Likewise.
* inflow.c: Include top.h.
(gdb_has_a_terminal, child_terminal_init_with_pgrp)
(gdb_save_tty_state, child_terminal_inferior)
(child_terminal_ours_1, copy_terminal_info): Use the main UI.
(initialize_stdin_serial): Adjust to per-UI instream.
* main.c (captured_command_loop, captured_main): Adjust to per-UI
instream.
* mi/mi-interp.c (mi_execute_command_wrapper): Likewise.
* python/python.c (python_interactive_command): Likewise.
* terminal.h (struct ui): Forward declare.
(initialize_stdin_serial): Add struct ui parameter.
* top.c (instream): Delete.
(do_restore_instream_cleanup, read_command_file, dont_repeat)
(gdb_readline_no_editing, command_line_input)
(input_from_terminal_p, gdb_init): Adjust to per-UI instream.
* top.h (struct ui) <instream>: New field.
(instream): Delete declaration.
(quit): Adjust to per-UI instream.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.gdb/selftest.exp (do_steps_and_nexts): Add new regexp.
2016-06-21 08:11:46 +08:00
|
|
|
|
if (!done_once && !ISATTY (ui->instream))
|
1999-08-17 03:57:19 +08:00
|
|
|
|
{
|
Make instream be per UI
gdb/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* cli/cli-script.c (execute_user_command, read_next_line)
(read_next_line): Adjust to per-UI instream.
* event-top.c (stdin_event_handler, command_handler)
(handle_line_of_input, command_line_handler)
(gdb_readline_no_editing_callback, async_sigterm_handler)
(gdb_setup_readline): Likewise.
* inflow.c: Include top.h.
(gdb_has_a_terminal, child_terminal_init_with_pgrp)
(gdb_save_tty_state, child_terminal_inferior)
(child_terminal_ours_1, copy_terminal_info): Use the main UI.
(initialize_stdin_serial): Adjust to per-UI instream.
* main.c (captured_command_loop, captured_main): Adjust to per-UI
instream.
* mi/mi-interp.c (mi_execute_command_wrapper): Likewise.
* python/python.c (python_interactive_command): Likewise.
* terminal.h (struct ui): Forward declare.
(initialize_stdin_serial): Add struct ui parameter.
* top.c (instream): Delete.
(do_restore_instream_cleanup, read_command_file, dont_repeat)
(gdb_readline_no_editing, command_line_input)
(input_from_terminal_p, gdb_init): Adjust to per-UI instream.
* top.h (struct ui) <instream>: New field.
(instream): Delete declaration.
(quit): Adjust to per-UI instream.
gdb/testsuite/ChangeLog:
2016-06-21 Pedro Alves <palves@redhat.com>
* gdb.gdb/selftest.exp (do_steps_and_nexts): Add new regexp.
2016-06-21 08:11:46 +08:00
|
|
|
|
setbuf (ui->instream, NULL);
|
1999-08-17 03:57:19 +08:00
|
|
|
|
done_once = 1;
|
|
|
|
|
}
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
/* We still need the while loop here, even though it would seem
|
2016-03-10 02:24:59 +08:00
|
|
|
|
obvious to invoke gdb_readline_no_editing_callback at every
|
|
|
|
|
character entered. If not using the readline library, the
|
|
|
|
|
terminal is in cooked mode, which sends the characters all at
|
|
|
|
|
once. Poll will notice that the input fd has changed state only
|
|
|
|
|
after enter is pressed. At this point we still need to fetch all
|
|
|
|
|
the chars entered. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
while (1)
|
|
|
|
|
{
|
|
|
|
|
/* Read from stdin if we are executing a user defined command.
|
|
|
|
|
This is the right thing for prompt_for_continue, at least. */
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
c = fgetc (ui->instream != NULL ? ui->instream : ui->stdin_stream);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
if (c == EOF)
|
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
if (line_buffer.used_size > 0)
|
|
|
|
|
{
|
|
|
|
|
/* The last line does not end with a newline. Return it, and
|
|
|
|
|
if we are called again fgetc will still return EOF and
|
|
|
|
|
we'll return NULL then. */
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
xfree (buffer_finish (&line_buffer));
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
ui->input_handler (NULL);
|
2007-08-30 06:04:15 +08:00
|
|
|
|
return;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (c == '\n')
|
|
|
|
|
{
|
2016-03-10 02:25:00 +08:00
|
|
|
|
if (line_buffer.used_size > 0
|
|
|
|
|
&& line_buffer.buffer[line_buffer.used_size - 1] == '\r')
|
|
|
|
|
line_buffer.used_size--;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
buffer_grow_char (&line_buffer, c);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
2016-03-10 02:25:00 +08:00
|
|
|
|
buffer_grow_char (&line_buffer, '\0');
|
|
|
|
|
result = buffer_finish (&line_buffer);
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
ui->input_handler (result);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
Introduce interruptible_select
We have places where we call a blocking gdb_select expecting that a
Ctrl-C will unblock it. However, if the Ctrl-C is pressed just before
gdb_select, the SIGINT handler runs before gdb_select, and thus
gdb_select won't return.
For example gdb_readline_no_editing:
QUIT;
/* Wait until at least one byte of data is available. Control-C
can interrupt gdb_select, but not fgetc. */
FD_ZERO (&readfds);
FD_SET (fd, &readfds);
if (gdb_select (fd + 1, &readfds, NULL, NULL, NULL) == -1)
and stdio_file_read:
/* For the benefit of Windows, call gdb_select before reading from
the file. Wait until at least one byte of data is available.
Control-C can interrupt gdb_select, but not read. */
{
fd_set readfds;
FD_ZERO (&readfds);
FD_SET (stdio->fd, &readfds);
if (gdb_select (stdio->fd + 1, &readfds, NULL, NULL, NULL) == -1)
return -1;
}
return read (stdio->fd, buf, length_buf);
This is a race classically fixed with either the self-pipe trick, or
by blocking SIGINT and then using pselect instead of select.
Blocking SIGINT most of the time would mean that check_quit_flag (and
thus QUIT) would need to do a syscall every time it is called, which
sounds best avoided, since QUIT is called in many loops. Thus we take
the self-pipe trick route (wrapped in a serial event).
Instead of having all places that need this manually add an extra file
descriptor to the set of gdb_select's watched file descriptors, we
introduce a wrapper, interruptible_select, that does that.
The Windows version of gdb_select actually does not suffer from this,
because mingw-hdep.c:gdb_call_async_signal_handler sets a Windows
event that gdb_select always waits on. So this patch can be seen as
generalization of that technique. We can't remove that extra event
from mingw-hdep.c until we get rid of immediate_quit though.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h: Extend QUIT-related comments to mention
interruptible_select.
(quit_serial_event_set, quit_serial_event_clear): Declare.
* event-top.c: Include "ser-event.h" and "gdb_select.h".
(quit_serial_event): New global.
(async_init_signals): Make quit_serial_event.
(quit_serial_event_set, quit_serial_event_clear)
(quit_serial_event_fd, interruptible_select): New functions.
* extension.c (set_quit_flag): Set the quit serial event.
(check_quit_flag): Clear the quit serial event.
* gdb_select.h (interruptible_select): New declaration.
* guile/scm-ports.c (ioscm_input_waiting): Use
interruptible_select instead of gdb_select.
* top.c (gdb_readline_no_editing): Likewise.
* ui-file.c (stdio_file_read): Likewise.
2016-04-12 23:49:30 +08:00
|
|
|
|
/* The serial event associated with the QUIT flag. set_quit_flag sets
|
|
|
|
|
this, and check_quit_flag clears it. Used by interruptible_select
|
|
|
|
|
to be able to do interruptible I/O with no race with the SIGINT
|
|
|
|
|
handler. */
|
|
|
|
|
static struct serial_event *quit_serial_event;
|
|
|
|
|
|
1999-05-12 04:29:07 +08:00
|
|
|
|
/* Initialization of signal handlers and tokens. There is a function
|
2010-12-29 08:58:14 +08:00
|
|
|
|
handle_sig* for each of the signals GDB cares about. Specifically:
|
1999-05-12 04:29:07 +08:00
|
|
|
|
SIGINT, SIGFPE, SIGQUIT, SIGTSTP, SIGHUP, SIGWINCH. These
|
|
|
|
|
functions are the actual signal handlers associated to the signals
|
|
|
|
|
via calls to signal(). The only job for these functions is to
|
|
|
|
|
enqueue the appropriate event/procedure with the event loop. Such
|
2010-12-29 08:58:14 +08:00
|
|
|
|
procedures are the old signal handlers. The event loop will take
|
1999-05-12 04:29:07 +08:00
|
|
|
|
care of invoking the queued procedures to perform the usual tasks
|
2010-12-29 08:58:14 +08:00
|
|
|
|
associated with the reception of the signal. */
|
1999-05-26 02:09:09 +08:00
|
|
|
|
/* NOTE: 1999-04-30 This is the asynchronous version of init_signals.
|
1999-05-12 04:29:07 +08:00
|
|
|
|
init_signals will become obsolete as we move to have to event loop
|
2010-12-29 08:58:14 +08:00
|
|
|
|
as the default for gdb. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
async_init_signals (void)
|
1999-07-08 04:19:36 +08:00
|
|
|
|
{
|
Fix signal handler/event-loop races
GDB's core signal handling suffers from a classical signal handler /
mainline code race:
int
gdb_do_one_event (void)
{
...
/* First let's see if there are any asynchronous signal handlers
that are ready. These would be the result of invoking any of the
signal handlers. */
if (invoke_async_signal_handlers ())
return 1;
...
/* Block waiting for a new event. (...). */
if (gdb_wait_for_event (1) < 0)
return -1;
...
}
If a signal is delivered while gdb is blocked in the poll/select
inside gdb_wait_for_event, then the select/poll breaks with EINTR,
we'll loop back around and call invoke_async_signal_handlers.
However, if the signal handler runs between
invoke_async_signal_handlers and gdb_wait_for_event,
gdb_wait_for_event will block, until the next unrelated event...
The fix is to a struct serial_event, and register it in the set of
files that select/poll in gdb_wait_for_event waits on. The signal
handlers that defer work to invoke_async_signal_handlers call
mark_async_signal_handler, which is adjusted to also set the new
serial event in addition to setting a flag, and is thus now is
garanteed to immediately unblock the next gdb_select/poll call, up
until invoke_async_signal_handlers is called and the event is cleared.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* event-loop.c: Include "ser-event.h".
(async_signal_handlers_serial_event): New global.
(async_signals_handler, initialize_async_signal_handlers): New
functions.
(mark_async_signal_handler): Set
async_signal_handlers_serial_event.
(invoke_async_signal_handlers): Clear
async_signal_handlers_serial_event.
* event-top.c (async_init_signals): Call
initialize_async_signal_handlers.
2016-04-12 23:49:30 +08:00
|
|
|
|
initialize_async_signal_handlers ();
|
|
|
|
|
|
Introduce interruptible_select
We have places where we call a blocking gdb_select expecting that a
Ctrl-C will unblock it. However, if the Ctrl-C is pressed just before
gdb_select, the SIGINT handler runs before gdb_select, and thus
gdb_select won't return.
For example gdb_readline_no_editing:
QUIT;
/* Wait until at least one byte of data is available. Control-C
can interrupt gdb_select, but not fgetc. */
FD_ZERO (&readfds);
FD_SET (fd, &readfds);
if (gdb_select (fd + 1, &readfds, NULL, NULL, NULL) == -1)
and stdio_file_read:
/* For the benefit of Windows, call gdb_select before reading from
the file. Wait until at least one byte of data is available.
Control-C can interrupt gdb_select, but not read. */
{
fd_set readfds;
FD_ZERO (&readfds);
FD_SET (stdio->fd, &readfds);
if (gdb_select (stdio->fd + 1, &readfds, NULL, NULL, NULL) == -1)
return -1;
}
return read (stdio->fd, buf, length_buf);
This is a race classically fixed with either the self-pipe trick, or
by blocking SIGINT and then using pselect instead of select.
Blocking SIGINT most of the time would mean that check_quit_flag (and
thus QUIT) would need to do a syscall every time it is called, which
sounds best avoided, since QUIT is called in many loops. Thus we take
the self-pipe trick route (wrapped in a serial event).
Instead of having all places that need this manually add an extra file
descriptor to the set of gdb_select's watched file descriptors, we
introduce a wrapper, interruptible_select, that does that.
The Windows version of gdb_select actually does not suffer from this,
because mingw-hdep.c:gdb_call_async_signal_handler sets a Windows
event that gdb_select always waits on. So this patch can be seen as
generalization of that technique. We can't remove that extra event
from mingw-hdep.c until we get rid of immediate_quit though.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h: Extend QUIT-related comments to mention
interruptible_select.
(quit_serial_event_set, quit_serial_event_clear): Declare.
* event-top.c: Include "ser-event.h" and "gdb_select.h".
(quit_serial_event): New global.
(async_init_signals): Make quit_serial_event.
(quit_serial_event_set, quit_serial_event_clear)
(quit_serial_event_fd, interruptible_select): New functions.
* extension.c (set_quit_flag): Set the quit serial event.
(check_quit_flag): Clear the quit serial event.
* gdb_select.h (interruptible_select): New declaration.
* guile/scm-ports.c (ioscm_input_waiting): Use
interruptible_select instead of gdb_select.
* top.c (gdb_readline_no_editing): Likewise.
* ui-file.c (stdio_file_read): Likewise.
2016-04-12 23:49:30 +08:00
|
|
|
|
quit_serial_event = make_serial_event ();
|
|
|
|
|
|
1999-05-12 04:29:07 +08:00
|
|
|
|
signal (SIGINT, handle_sigint);
|
|
|
|
|
sigint_token =
|
1999-06-15 02:08:47 +08:00
|
|
|
|
create_async_signal_handler (async_request_quit, NULL);
|
2005-11-26 02:33:12 +08:00
|
|
|
|
signal (SIGTERM, handle_sigterm);
|
2014-03-19 05:48:06 +08:00
|
|
|
|
async_sigterm_token
|
|
|
|
|
= create_async_signal_handler (async_sigterm_handler, NULL);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
/* If SIGTRAP was set to SIG_IGN, then the SIG_IGN will get passed
|
|
|
|
|
to the inferior and breakpoints will be ignored. */
|
|
|
|
|
#ifdef SIGTRAP
|
|
|
|
|
signal (SIGTRAP, SIG_DFL);
|
|
|
|
|
#endif
|
|
|
|
|
|
2005-03-17 01:05:31 +08:00
|
|
|
|
#ifdef SIGQUIT
|
1999-05-12 04:29:07 +08:00
|
|
|
|
/* If we initialize SIGQUIT to SIG_IGN, then the SIG_IGN will get
|
|
|
|
|
passed to the inferior, which we don't want. It would be
|
|
|
|
|
possible to do a "signal (SIGQUIT, SIG_DFL)" after we fork, but
|
|
|
|
|
on BSD4.3 systems using vfork, that can affect the
|
|
|
|
|
GDB process as well as the inferior (the signal handling tables
|
|
|
|
|
might be in memory, shared between the two). Since we establish
|
|
|
|
|
a handler for SIGQUIT, when we call exec it will set the signal
|
|
|
|
|
to SIG_DFL for us. */
|
|
|
|
|
signal (SIGQUIT, handle_sigquit);
|
|
|
|
|
sigquit_token =
|
1999-06-15 02:08:47 +08:00
|
|
|
|
create_async_signal_handler (async_do_nothing, NULL);
|
2005-03-17 01:05:31 +08:00
|
|
|
|
#endif
|
1999-05-12 04:29:07 +08:00
|
|
|
|
#ifdef SIGHUP
|
|
|
|
|
if (signal (SIGHUP, handle_sighup) != SIG_IGN)
|
|
|
|
|
sighup_token =
|
1999-06-15 02:08:47 +08:00
|
|
|
|
create_async_signal_handler (async_disconnect, NULL);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
else
|
|
|
|
|
sighup_token =
|
1999-06-15 02:08:47 +08:00
|
|
|
|
create_async_signal_handler (async_do_nothing, NULL);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
#endif
|
|
|
|
|
signal (SIGFPE, handle_sigfpe);
|
|
|
|
|
sigfpe_token =
|
1999-06-15 02:08:47 +08:00
|
|
|
|
create_async_signal_handler (async_float_handler, NULL);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
1999-06-15 02:08:47 +08:00
|
|
|
|
#ifdef STOP_SIGNAL
|
|
|
|
|
sigtstp_token =
|
|
|
|
|
create_async_signal_handler (async_stop_sig, NULL);
|
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
Introduce interruptible_select
We have places where we call a blocking gdb_select expecting that a
Ctrl-C will unblock it. However, if the Ctrl-C is pressed just before
gdb_select, the SIGINT handler runs before gdb_select, and thus
gdb_select won't return.
For example gdb_readline_no_editing:
QUIT;
/* Wait until at least one byte of data is available. Control-C
can interrupt gdb_select, but not fgetc. */
FD_ZERO (&readfds);
FD_SET (fd, &readfds);
if (gdb_select (fd + 1, &readfds, NULL, NULL, NULL) == -1)
and stdio_file_read:
/* For the benefit of Windows, call gdb_select before reading from
the file. Wait until at least one byte of data is available.
Control-C can interrupt gdb_select, but not read. */
{
fd_set readfds;
FD_ZERO (&readfds);
FD_SET (stdio->fd, &readfds);
if (gdb_select (stdio->fd + 1, &readfds, NULL, NULL, NULL) == -1)
return -1;
}
return read (stdio->fd, buf, length_buf);
This is a race classically fixed with either the self-pipe trick, or
by blocking SIGINT and then using pselect instead of select.
Blocking SIGINT most of the time would mean that check_quit_flag (and
thus QUIT) would need to do a syscall every time it is called, which
sounds best avoided, since QUIT is called in many loops. Thus we take
the self-pipe trick route (wrapped in a serial event).
Instead of having all places that need this manually add an extra file
descriptor to the set of gdb_select's watched file descriptors, we
introduce a wrapper, interruptible_select, that does that.
The Windows version of gdb_select actually does not suffer from this,
because mingw-hdep.c:gdb_call_async_signal_handler sets a Windows
event that gdb_select always waits on. So this patch can be seen as
generalization of that technique. We can't remove that extra event
from mingw-hdep.c until we get rid of immediate_quit though.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h: Extend QUIT-related comments to mention
interruptible_select.
(quit_serial_event_set, quit_serial_event_clear): Declare.
* event-top.c: Include "ser-event.h" and "gdb_select.h".
(quit_serial_event): New global.
(async_init_signals): Make quit_serial_event.
(quit_serial_event_set, quit_serial_event_clear)
(quit_serial_event_fd, interruptible_select): New functions.
* extension.c (set_quit_flag): Set the quit serial event.
(check_quit_flag): Clear the quit serial event.
* gdb_select.h (interruptible_select): New declaration.
* guile/scm-ports.c (ioscm_input_waiting): Use
interruptible_select instead of gdb_select.
* top.c (gdb_readline_no_editing): Likewise.
* ui-file.c (stdio_file_read): Likewise.
2016-04-12 23:49:30 +08:00
|
|
|
|
/* See defs.h. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
quit_serial_event_set (void)
|
|
|
|
|
{
|
|
|
|
|
serial_event_set (quit_serial_event);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See defs.h. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
quit_serial_event_clear (void)
|
|
|
|
|
{
|
|
|
|
|
serial_event_clear (quit_serial_event);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Return the selectable file descriptor of the serial event
|
|
|
|
|
associated with the quit flag. */
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
quit_serial_event_fd (void)
|
|
|
|
|
{
|
|
|
|
|
return serial_event_fd (quit_serial_event);
|
|
|
|
|
}
|
|
|
|
|
|
target remote: Don't rely on immediate_quit (introduce quit handlers)
remote.c is the last user of immediate_quit. It's relied on to
immediately break the initial remote connection sync up, if the user
does Ctrl-C, assuming that was because the target isn't responding.
At that stage, since the connection isn't synced yet, disconnecting is
the only safe thing to do. This commit reworks that, to not rely on
throwing from the SIGINT signal handler.
So, this commit:
- Introduces the concept of a "quit handler". This is used to
override what does the QUIT macro do when the quit flag is set.
- Makes the "struct serial" reachar / write code call QUIT in the
partial read/write loops, so the current quit handler is invoked
whenever a serial->read_prim / serial->write_prim returns EINTR.
- Makes the "struct serial" reachar / write code call
interruptible_select instead of gdb_select, so that QUITs are
detected in a race-free manner.
- Stops remote.c from setting immediate_quit during the initial
connection.
- Instead, we install a custom quit handler whenever we're calling
into the serial code. This custom quit handler knows to immediately
throw a quit when we're in the initial connection setup, and
otherwise defer handling the quit/Ctrl-C request to later, when
we're safely out of a packet command/response sequence. This also
is what is now responsible for handling "double Ctrl-C because
target connection is stuck/wedged."
- remote.c no longer installs a specialized SIGINT handlers, and
instead re-uses the quit flag. Since we want to rely on the QUIT
macro, the SIGINT handler must also set the quit. And the easiest
is just to not install custom SIGINT handler in remote.c. Let the
standard SIGINT handler do its job of setting the quit flag.
Centralizing SIGINT handlers seems like a good thing to me, anyway.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h (quit_handler_ftype, quit_handler)
(make_cleanup_override_quit_handler, default_quit_handler): New.
(QUIT): Adjust comments.
* event-top.c (default_quit_handler): New function.
(quit_handler): New global.
(struct quit_handler_cleanup_data): New.
(restore_quit_handler, restore_quit_handler_dtor)
(make_cleanup_override_quit_handler): New.
(async_request_quit): Call QUIT.
* remote.c (struct remote_state) <got_ctrlc_during_io>: New field.
(async_sigint_remote_twice_token, async_sigint_remote_token):
Delete.
(remote_close): Update comments.
(remote_start_remote): Don't set immediate_quit. Set starting_up
earlier.
(remote_serial_quit_handler, remote_unpush_and_throw): New
functions.
(remote_open_1): Clear got_ctrlc_during_io. Set
remote_async_terminal_ours_p unconditionally.
(async_initialize_sigint_signal_handler)
(async_handle_remote_sigint, async_handle_remote_sigint_twice)
(remote_check_pending_interrupt, async_remote_interrupt)
(async_remote_interrupt_twice)
(async_cleanup_sigint_signal_handler, ofunc)
(sync_remote_interrupt, sync_remote_interrupt_twice): Delete.
(remote_terminal_inferior, remote_terminal_ours): Remove async
checks.
(remote_wait_as): Don't install a SIGINT handler in sync mode.
(readchar, remote_serial_write): Override the quit handler with
remote_serial_quit_handler.
(getpkt_or_notif_sane_1): Don't call QUIT.
(initialize_remote_ops): Don't install
remote_check_pending_interrupt.
(_initialize_remote): Don't create async_sigint_remote_token and
async_sigint_remote_twice_token.
* ser-base.c (ser_base_wait_for): Call QUIT and use
interruptible_select.
(ser_base_write): Call QUIT.
* ser-go32.c (dos_readchar, dos_write): Call QUIT.
* ser-unix.c (wait_for): Don't use VTIME. Always take the
gdb_select path, but call QUIT and interruptible_select.
* utils.c (maybe_quit): Call the current quit handler. Don't call
target_check_pending_interrupt.
(defaulted_query, prompt_for_continue): Override the quit handler
with the default quit handler.
2016-04-12 23:49:32 +08:00
|
|
|
|
/* See defs.h. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
default_quit_handler (void)
|
|
|
|
|
{
|
|
|
|
|
if (check_quit_flag ())
|
|
|
|
|
{
|
Remove make_cleanup_restore_target_terminal
This removes make_cleanup_restore_target_terminal and generally
C++-ifies target terminal handling. It changes all target_terminal_*
functions to be static members of a new target_terminal class and
changes the cleanup to be a scoped_* class.
make_cleanup_override_quit_handler is also removed in favor of simply
using scoped_restore.
Note that there are some files in this patch that I could not compile.
Considering that some of the rewrites were automated, and that none of
these files involed cleanups, I feel that this is relatively safe.
Regression tested by the buildbot.
gdb/ChangeLog
2017-09-20 Tom Tromey <tom@tromey.com>
* windows-nat.c (get_windows_debug_event, windows_wait)
(do_initial_windows_stuff, windows_attach): Update.
* utils.c (vwarning, internal_vproblem): Update.
(ui_unregister_input_event_handler_cleanup)
(prepare_to_handle_input): Remove.
(class scoped_input_handler): New.
(defaulted_query, prompt_for_continue): Update.
* tui/tui-hooks.c (tui_refresh_frame_and_register_information):
Update.
* top.c (undo_terminal_modifications_before_exit): Update.
* target/target.h (target_terminal_init, target_terminal_inferior)
(target_terminal_ours): Don't declare.
(class target_terminal): New.
* target.h (target_terminal_is_inferior, target_terminal_is_ours)
(target_terminal_ours_for_output)
(make_cleanup_restore_target_terminal): Don't declare.
(target_terminal_info): Remove.
* target.c (enum terminal_state, terminal_state): Remove.
(target_terminal::terminal_state): Define.
(target_terminal::init): Rename from target_terminal_init.
(target_terminal::inferior): Rename from
target_terminal_inferior.
(target_terminal::ours): Rename from target_terminal_ours.
(target_terminal::ours_for_output): Rename from
target_terminal_ours_for_output.
(target_terminal::info): New method.
(cleanup_restore_target_terminal)
(make_cleanup_restore_target_terminal): Remove.
* solib.c (handle_solib_event): Update.
* remote.c (remote_serial_quit_handler): Update.
(remote_terminal_inferior, remote_wait_as): Update.
* record-full.c (record_full_wait_1): Update.
* nto-procfs.c (procfs_create_inferior): Update.
* nat/fork-inferior.c (startup_inferior): Update.
* mi/mi-interp.c (mi_new_thread, mi_thread_exit)
(mi_record_changed, mi_inferior_added, mi_inferior_appeared)
(mi_inferior_exit, mi_inferior_removed, mi_traceframe_changed)
(mi_tsv_created, mi_tsv_deleted, mi_tsv_modified)
(mi_breakpoint_created, mi_breakpoint_deleted)
(mi_breakpoint_modified, mi_on_resume, mi_solib_loaded)
(mi_solib_unloaded, mi_command_param_changed, mi_memory_changed)
(mi_user_selected_context_changed, report_initial_inferior):
Update.
* linux-nat.c (linux_nat_attach, linux_nat_terminal_ours)
(linux_nat_terminal_inferior): Update.
* infrun.c (follow_fork_inferior)
(handle_vfork_child_exec_or_exit, do_target_resume)
(check_curr_ui_sync_execution_done, handle_inferior_event_1)
(handle_signal_stop, maybe_remove_breakpoints, normal_stop):
Update.
* inflow.c (child_terminal_init, info_terminal_command): Update.
* infcmd.c (post_create_inferior, continue_1, prepare_one_step)
(attach_command): Update.
* infcall.c (call_thread_fsm_should_stop): Update.
* gnu-nat.c (gnu_attach): Update.
* extension.c (struct active_ext_lang_state)
(restore_active_ext_lang): Update.
* exceptions.c (print_flush): Update.
* event-top.c (async_enable_stdin, default_quit_handler): Update.
(struct quit_handler_cleanup_data, restore_quit_handler)
(restore_quit_handler_dtor, make_cleanup_override_quit_handler):
Remove.
* cp-support.c (gdb_demangle): Update.
* breakpoint.c (update_inserted_breakpoint_locations)
(insert_breakpoint_locations, handle_jit_event)
(disable_breakpoints_in_unloaded_shlib): Update.
* annotate.c (annotate_breakpoints_invalid)
(annotate_frames_invalid): Update.
gdb/gdbserver/ChangeLog
2017-09-20 Tom Tromey <tom@tromey.com>
* target.c (target_terminal::terminal_state): Define.
(target_terminal::init): Rename from target_terminal_init.
(target_terminal::inferior): Rename from
target_terminal_inferior.
(target_terminal::ours): Rename from target_terminal_ours.
(target_terminal::ours_for_output, target_terminal::info): New.
2017-09-20 11:56:36 +08:00
|
|
|
|
if (target_terminal::is_ours ())
|
target remote: Don't rely on immediate_quit (introduce quit handlers)
remote.c is the last user of immediate_quit. It's relied on to
immediately break the initial remote connection sync up, if the user
does Ctrl-C, assuming that was because the target isn't responding.
At that stage, since the connection isn't synced yet, disconnecting is
the only safe thing to do. This commit reworks that, to not rely on
throwing from the SIGINT signal handler.
So, this commit:
- Introduces the concept of a "quit handler". This is used to
override what does the QUIT macro do when the quit flag is set.
- Makes the "struct serial" reachar / write code call QUIT in the
partial read/write loops, so the current quit handler is invoked
whenever a serial->read_prim / serial->write_prim returns EINTR.
- Makes the "struct serial" reachar / write code call
interruptible_select instead of gdb_select, so that QUITs are
detected in a race-free manner.
- Stops remote.c from setting immediate_quit during the initial
connection.
- Instead, we install a custom quit handler whenever we're calling
into the serial code. This custom quit handler knows to immediately
throw a quit when we're in the initial connection setup, and
otherwise defer handling the quit/Ctrl-C request to later, when
we're safely out of a packet command/response sequence. This also
is what is now responsible for handling "double Ctrl-C because
target connection is stuck/wedged."
- remote.c no longer installs a specialized SIGINT handlers, and
instead re-uses the quit flag. Since we want to rely on the QUIT
macro, the SIGINT handler must also set the quit. And the easiest
is just to not install custom SIGINT handler in remote.c. Let the
standard SIGINT handler do its job of setting the quit flag.
Centralizing SIGINT handlers seems like a good thing to me, anyway.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h (quit_handler_ftype, quit_handler)
(make_cleanup_override_quit_handler, default_quit_handler): New.
(QUIT): Adjust comments.
* event-top.c (default_quit_handler): New function.
(quit_handler): New global.
(struct quit_handler_cleanup_data): New.
(restore_quit_handler, restore_quit_handler_dtor)
(make_cleanup_override_quit_handler): New.
(async_request_quit): Call QUIT.
* remote.c (struct remote_state) <got_ctrlc_during_io>: New field.
(async_sigint_remote_twice_token, async_sigint_remote_token):
Delete.
(remote_close): Update comments.
(remote_start_remote): Don't set immediate_quit. Set starting_up
earlier.
(remote_serial_quit_handler, remote_unpush_and_throw): New
functions.
(remote_open_1): Clear got_ctrlc_during_io. Set
remote_async_terminal_ours_p unconditionally.
(async_initialize_sigint_signal_handler)
(async_handle_remote_sigint, async_handle_remote_sigint_twice)
(remote_check_pending_interrupt, async_remote_interrupt)
(async_remote_interrupt_twice)
(async_cleanup_sigint_signal_handler, ofunc)
(sync_remote_interrupt, sync_remote_interrupt_twice): Delete.
(remote_terminal_inferior, remote_terminal_ours): Remove async
checks.
(remote_wait_as): Don't install a SIGINT handler in sync mode.
(readchar, remote_serial_write): Override the quit handler with
remote_serial_quit_handler.
(getpkt_or_notif_sane_1): Don't call QUIT.
(initialize_remote_ops): Don't install
remote_check_pending_interrupt.
(_initialize_remote): Don't create async_sigint_remote_token and
async_sigint_remote_twice_token.
* ser-base.c (ser_base_wait_for): Call QUIT and use
interruptible_select.
(ser_base_write): Call QUIT.
* ser-go32.c (dos_readchar, dos_write): Call QUIT.
* ser-unix.c (wait_for): Don't use VTIME. Always take the
gdb_select path, but call QUIT and interruptible_select.
* utils.c (maybe_quit): Call the current quit handler. Don't call
target_check_pending_interrupt.
(defaulted_query, prompt_for_continue): Override the quit handler
with the default quit handler.
2016-04-12 23:49:32 +08:00
|
|
|
|
quit ();
|
|
|
|
|
else
|
|
|
|
|
target_pass_ctrlc ();
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See defs.h. */
|
|
|
|
|
quit_handler_ftype *quit_handler = default_quit_handler;
|
|
|
|
|
|
Introduce interruptible_select
We have places where we call a blocking gdb_select expecting that a
Ctrl-C will unblock it. However, if the Ctrl-C is pressed just before
gdb_select, the SIGINT handler runs before gdb_select, and thus
gdb_select won't return.
For example gdb_readline_no_editing:
QUIT;
/* Wait until at least one byte of data is available. Control-C
can interrupt gdb_select, but not fgetc. */
FD_ZERO (&readfds);
FD_SET (fd, &readfds);
if (gdb_select (fd + 1, &readfds, NULL, NULL, NULL) == -1)
and stdio_file_read:
/* For the benefit of Windows, call gdb_select before reading from
the file. Wait until at least one byte of data is available.
Control-C can interrupt gdb_select, but not read. */
{
fd_set readfds;
FD_ZERO (&readfds);
FD_SET (stdio->fd, &readfds);
if (gdb_select (stdio->fd + 1, &readfds, NULL, NULL, NULL) == -1)
return -1;
}
return read (stdio->fd, buf, length_buf);
This is a race classically fixed with either the self-pipe trick, or
by blocking SIGINT and then using pselect instead of select.
Blocking SIGINT most of the time would mean that check_quit_flag (and
thus QUIT) would need to do a syscall every time it is called, which
sounds best avoided, since QUIT is called in many loops. Thus we take
the self-pipe trick route (wrapped in a serial event).
Instead of having all places that need this manually add an extra file
descriptor to the set of gdb_select's watched file descriptors, we
introduce a wrapper, interruptible_select, that does that.
The Windows version of gdb_select actually does not suffer from this,
because mingw-hdep.c:gdb_call_async_signal_handler sets a Windows
event that gdb_select always waits on. So this patch can be seen as
generalization of that technique. We can't remove that extra event
from mingw-hdep.c until we get rid of immediate_quit though.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h: Extend QUIT-related comments to mention
interruptible_select.
(quit_serial_event_set, quit_serial_event_clear): Declare.
* event-top.c: Include "ser-event.h" and "gdb_select.h".
(quit_serial_event): New global.
(async_init_signals): Make quit_serial_event.
(quit_serial_event_set, quit_serial_event_clear)
(quit_serial_event_fd, interruptible_select): New functions.
* extension.c (set_quit_flag): Set the quit serial event.
(check_quit_flag): Clear the quit serial event.
* gdb_select.h (interruptible_select): New declaration.
* guile/scm-ports.c (ioscm_input_waiting): Use
interruptible_select instead of gdb_select.
* top.c (gdb_readline_no_editing): Likewise.
* ui-file.c (stdio_file_read): Likewise.
2016-04-12 23:49:30 +08:00
|
|
|
|
/* Handle a SIGINT. */
|
|
|
|
|
|
1999-07-08 04:19:36 +08:00
|
|
|
|
void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
handle_sigint (int sig)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
|
|
|
|
signal (sig, handle_sigint);
|
|
|
|
|
|
2007-02-09 00:25:25 +08:00
|
|
|
|
/* We could be running in a loop reading in symfiles or something so
|
|
|
|
|
it may be quite a while before we get back to the event loop. So
|
2010-12-29 08:58:14 +08:00
|
|
|
|
set quit_flag to 1 here. Then if QUIT is called before we get to
|
2007-02-09 00:25:25 +08:00
|
|
|
|
the event loop, we will unwind as expected. */
|
2012-08-23 01:48:55 +08:00
|
|
|
|
set_quit_flag ();
|
2007-02-09 00:25:25 +08:00
|
|
|
|
|
2016-04-12 23:49:32 +08:00
|
|
|
|
/* In case nothing calls QUIT before the event loop is reached, the
|
|
|
|
|
event loop handles it. */
|
|
|
|
|
mark_async_signal_handler (sigint_token);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
Introduce interruptible_select
We have places where we call a blocking gdb_select expecting that a
Ctrl-C will unblock it. However, if the Ctrl-C is pressed just before
gdb_select, the SIGINT handler runs before gdb_select, and thus
gdb_select won't return.
For example gdb_readline_no_editing:
QUIT;
/* Wait until at least one byte of data is available. Control-C
can interrupt gdb_select, but not fgetc. */
FD_ZERO (&readfds);
FD_SET (fd, &readfds);
if (gdb_select (fd + 1, &readfds, NULL, NULL, NULL) == -1)
and stdio_file_read:
/* For the benefit of Windows, call gdb_select before reading from
the file. Wait until at least one byte of data is available.
Control-C can interrupt gdb_select, but not read. */
{
fd_set readfds;
FD_ZERO (&readfds);
FD_SET (stdio->fd, &readfds);
if (gdb_select (stdio->fd + 1, &readfds, NULL, NULL, NULL) == -1)
return -1;
}
return read (stdio->fd, buf, length_buf);
This is a race classically fixed with either the self-pipe trick, or
by blocking SIGINT and then using pselect instead of select.
Blocking SIGINT most of the time would mean that check_quit_flag (and
thus QUIT) would need to do a syscall every time it is called, which
sounds best avoided, since QUIT is called in many loops. Thus we take
the self-pipe trick route (wrapped in a serial event).
Instead of having all places that need this manually add an extra file
descriptor to the set of gdb_select's watched file descriptors, we
introduce a wrapper, interruptible_select, that does that.
The Windows version of gdb_select actually does not suffer from this,
because mingw-hdep.c:gdb_call_async_signal_handler sets a Windows
event that gdb_select always waits on. So this patch can be seen as
generalization of that technique. We can't remove that extra event
from mingw-hdep.c until we get rid of immediate_quit though.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h: Extend QUIT-related comments to mention
interruptible_select.
(quit_serial_event_set, quit_serial_event_clear): Declare.
* event-top.c: Include "ser-event.h" and "gdb_select.h".
(quit_serial_event): New global.
(async_init_signals): Make quit_serial_event.
(quit_serial_event_set, quit_serial_event_clear)
(quit_serial_event_fd, interruptible_select): New functions.
* extension.c (set_quit_flag): Set the quit serial event.
(check_quit_flag): Clear the quit serial event.
* gdb_select.h (interruptible_select): New declaration.
* guile/scm-ports.c (ioscm_input_waiting): Use
interruptible_select instead of gdb_select.
* top.c (gdb_readline_no_editing): Likewise.
* ui-file.c (stdio_file_read): Likewise.
2016-04-12 23:49:30 +08:00
|
|
|
|
/* See gdb_select.h. */
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
interruptible_select (int n,
|
|
|
|
|
fd_set *readfds, fd_set *writefds, fd_set *exceptfds,
|
|
|
|
|
struct timeval *timeout)
|
|
|
|
|
{
|
|
|
|
|
fd_set my_readfds;
|
|
|
|
|
int fd;
|
|
|
|
|
int res;
|
|
|
|
|
|
|
|
|
|
if (readfds == NULL)
|
|
|
|
|
{
|
|
|
|
|
readfds = &my_readfds;
|
|
|
|
|
FD_ZERO (&my_readfds);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
fd = quit_serial_event_fd ();
|
|
|
|
|
FD_SET (fd, readfds);
|
|
|
|
|
if (n <= fd)
|
|
|
|
|
n = fd + 1;
|
|
|
|
|
|
|
|
|
|
do
|
|
|
|
|
{
|
|
|
|
|
res = gdb_select (n, readfds, writefds, exceptfds, timeout);
|
|
|
|
|
}
|
|
|
|
|
while (res == -1 && errno == EINTR);
|
|
|
|
|
|
|
|
|
|
if (res == 1 && FD_ISSET (fd, readfds))
|
|
|
|
|
{
|
|
|
|
|
errno = EINTR;
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
return res;
|
|
|
|
|
}
|
|
|
|
|
|
2014-03-19 05:48:06 +08:00
|
|
|
|
/* Handle GDB exit upon receiving SIGTERM if target_can_async_p (). */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
async_sigterm_handler (gdb_client_data arg)
|
|
|
|
|
{
|
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.
2016-06-21 08:11:54 +08:00
|
|
|
|
quit_force (NULL, 0);
|
2014-03-19 05:48:06 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See defs.h. */
|
|
|
|
|
volatile int sync_quit_force_run;
|
|
|
|
|
|
2005-11-26 02:33:12 +08:00
|
|
|
|
/* Quit GDB if SIGTERM is received.
|
|
|
|
|
GDB would quit anyway, but this way it will clean up properly. */
|
|
|
|
|
void
|
|
|
|
|
handle_sigterm (int sig)
|
|
|
|
|
{
|
|
|
|
|
signal (sig, handle_sigterm);
|
2014-03-19 05:48:06 +08:00
|
|
|
|
|
2015-07-25 01:58:47 +08:00
|
|
|
|
sync_quit_force_run = 1;
|
|
|
|
|
set_quit_flag ();
|
|
|
|
|
|
|
|
|
|
mark_async_signal_handler (async_sigterm_token);
|
2005-11-26 02:33:12 +08:00
|
|
|
|
}
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Do the quit. All the checks have been done by the caller. */
|
1999-07-08 04:19:36 +08:00
|
|
|
|
void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
async_request_quit (gdb_client_data arg)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2007-02-09 00:25:25 +08:00
|
|
|
|
/* If the quit_flag has gotten reset back to 0 by the time we get
|
2007-02-10 07:45:35 +08:00
|
|
|
|
back here, that means that an exception was thrown to unwind the
|
|
|
|
|
current command before we got back to the event loop. So there
|
2012-08-23 01:48:55 +08:00
|
|
|
|
is no reason to call quit again here. */
|
target remote: Don't rely on immediate_quit (introduce quit handlers)
remote.c is the last user of immediate_quit. It's relied on to
immediately break the initial remote connection sync up, if the user
does Ctrl-C, assuming that was because the target isn't responding.
At that stage, since the connection isn't synced yet, disconnecting is
the only safe thing to do. This commit reworks that, to not rely on
throwing from the SIGINT signal handler.
So, this commit:
- Introduces the concept of a "quit handler". This is used to
override what does the QUIT macro do when the quit flag is set.
- Makes the "struct serial" reachar / write code call QUIT in the
partial read/write loops, so the current quit handler is invoked
whenever a serial->read_prim / serial->write_prim returns EINTR.
- Makes the "struct serial" reachar / write code call
interruptible_select instead of gdb_select, so that QUITs are
detected in a race-free manner.
- Stops remote.c from setting immediate_quit during the initial
connection.
- Instead, we install a custom quit handler whenever we're calling
into the serial code. This custom quit handler knows to immediately
throw a quit when we're in the initial connection setup, and
otherwise defer handling the quit/Ctrl-C request to later, when
we're safely out of a packet command/response sequence. This also
is what is now responsible for handling "double Ctrl-C because
target connection is stuck/wedged."
- remote.c no longer installs a specialized SIGINT handlers, and
instead re-uses the quit flag. Since we want to rely on the QUIT
macro, the SIGINT handler must also set the quit. And the easiest
is just to not install custom SIGINT handler in remote.c. Let the
standard SIGINT handler do its job of setting the quit flag.
Centralizing SIGINT handlers seems like a good thing to me, anyway.
gdb/ChangeLog:
2016-04-12 Pedro Alves <palves@redhat.com>
* defs.h (quit_handler_ftype, quit_handler)
(make_cleanup_override_quit_handler, default_quit_handler): New.
(QUIT): Adjust comments.
* event-top.c (default_quit_handler): New function.
(quit_handler): New global.
(struct quit_handler_cleanup_data): New.
(restore_quit_handler, restore_quit_handler_dtor)
(make_cleanup_override_quit_handler): New.
(async_request_quit): Call QUIT.
* remote.c (struct remote_state) <got_ctrlc_during_io>: New field.
(async_sigint_remote_twice_token, async_sigint_remote_token):
Delete.
(remote_close): Update comments.
(remote_start_remote): Don't set immediate_quit. Set starting_up
earlier.
(remote_serial_quit_handler, remote_unpush_and_throw): New
functions.
(remote_open_1): Clear got_ctrlc_during_io. Set
remote_async_terminal_ours_p unconditionally.
(async_initialize_sigint_signal_handler)
(async_handle_remote_sigint, async_handle_remote_sigint_twice)
(remote_check_pending_interrupt, async_remote_interrupt)
(async_remote_interrupt_twice)
(async_cleanup_sigint_signal_handler, ofunc)
(sync_remote_interrupt, sync_remote_interrupt_twice): Delete.
(remote_terminal_inferior, remote_terminal_ours): Remove async
checks.
(remote_wait_as): Don't install a SIGINT handler in sync mode.
(readchar, remote_serial_write): Override the quit handler with
remote_serial_quit_handler.
(getpkt_or_notif_sane_1): Don't call QUIT.
(initialize_remote_ops): Don't install
remote_check_pending_interrupt.
(_initialize_remote): Don't create async_sigint_remote_token and
async_sigint_remote_twice_token.
* ser-base.c (ser_base_wait_for): Call QUIT and use
interruptible_select.
(ser_base_write): Call QUIT.
* ser-go32.c (dos_readchar, dos_write): Call QUIT.
* ser-unix.c (wait_for): Don't use VTIME. Always take the
gdb_select path, but call QUIT and interruptible_select.
* utils.c (maybe_quit): Call the current quit handler. Don't call
target_check_pending_interrupt.
(defaulted_query, prompt_for_continue): Override the quit handler
with the default quit handler.
2016-04-12 23:49:32 +08:00
|
|
|
|
QUIT;
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
2005-03-17 01:05:31 +08:00
|
|
|
|
#ifdef SIGQUIT
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Tell the event loop what to do if SIGQUIT is received.
|
|
|
|
|
See event-signal.c. */
|
1999-07-08 04:19:36 +08:00
|
|
|
|
static void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
handle_sigquit (int sig)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2012-10-23 14:20:49 +08:00
|
|
|
|
mark_async_signal_handler (sigquit_token);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
signal (sig, handle_sigquit);
|
|
|
|
|
}
|
2005-03-17 01:05:31 +08:00
|
|
|
|
#endif
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#if defined (SIGQUIT) || defined (SIGHUP)
|
|
|
|
|
/* Called by the event loop in response to a SIGQUIT or an
|
|
|
|
|
ignored SIGHUP. */
|
1999-07-08 04:19:36 +08:00
|
|
|
|
static void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
async_do_nothing (gdb_client_data arg)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Empty function body. */
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
2006-02-24 02:37:42 +08:00
|
|
|
|
#endif
|
1999-05-12 04:29:07 +08:00
|
|
|
|
|
|
|
|
|
#ifdef SIGHUP
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Tell the event loop what to do if SIGHUP is received.
|
|
|
|
|
See event-signal.c. */
|
1999-07-08 04:19:36 +08:00
|
|
|
|
static void
|
2000-07-30 09:48:28 +08:00
|
|
|
|
handle_sighup (int sig)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2012-10-23 14:20:49 +08:00
|
|
|
|
mark_async_signal_handler (sighup_token);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
signal (sig, handle_sighup);
|
|
|
|
|
}
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Called by the event loop to process a SIGHUP. */
|
1999-07-08 04:19:36 +08:00
|
|
|
|
static void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
async_disconnect (gdb_client_data arg)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2011-09-21 23:21:28 +08:00
|
|
|
|
|
Split TRY_CATCH into TRY + CATCH
This patch splits the TRY_CATCH macro into three, so that we go from
this:
~~~
volatile gdb_exception ex;
TRY_CATCH (ex, RETURN_MASK_ERROR)
{
}
if (ex.reason < 0)
{
}
~~~
to this:
~~~
TRY
{
}
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
~~~
Thus, we'll be getting rid of the local volatile exception object, and
declaring the caught exception in the catch block.
This allows reimplementing TRY/CATCH in terms of C++ exceptions when
building in C++ mode, while still allowing to build GDB in C mode
(using setjmp/longjmp), as a transition step.
TBC, after this patch, is it _not_ valid to have code between the TRY
and the CATCH blocks, like:
TRY
{
}
// some code here.
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
Just like it isn't valid to do that with C++'s native try/catch.
By switching to creating the exception object inside the CATCH block
scope, we can get rid of all the explicitly allocated volatile
exception objects all over the tree, and map the CATCH block more
directly to C++'s catch blocks.
The majority of the TRY_CATCH -> TRY+CATCH+END_CATCH conversion was
done with a script, rerun from scratch at every rebase, no manual
editing involved. After the mechanical conversion, a few places
needed manual intervention, to fix preexisting cases where we were
using the exception object outside of the TRY_CATCH block, and cases
where we were using "else" after a 'if (ex.reason) < 0)' [a CATCH
after this patch]. The result was folded into this patch so that GDB
still builds at each incremental step.
END_CATCH is necessary for two reasons:
First, because we name the exception object in the CATCH block, which
requires creating a scope, which in turn must be closed somewhere.
Declaring the exception variable in the initializer field of a for
block, like:
#define CATCH(EXCEPTION, mask) \
for (struct gdb_exception EXCEPTION; \
exceptions_state_mc_catch (&EXCEPTION, MASK); \
EXCEPTION = exception_none)
would avoid needing END_CATCH, but alas, in C mode, we build with C90,
which doesn't allow mixed declarations and code.
Second, because when TRY/CATCH are wired to real C++ try/catch, as
long as we need to handle cleanup chains, even if there's no CATCH
block that wants to catch the exception, we need for stop at every
frame in the unwind chain and run cleanups, then rethrow. That will
be done in END_CATCH.
After we require C++, we'll still need TRY/CATCH/END_CATCH until
cleanups are completely phased out -- TRY/CATCH in C++ mode will
save/restore the current cleanup chain, like in C mode, and END_CATCH
catches otherwise uncaugh exceptions, runs cleanups and rethrows, so
that C++ cleanups and exceptions can coexist.
IMO, this still makes the TRY/CATCH code look a bit more like a
newcomer would expect, so IMO worth it even if we weren't considering
C++.
gdb/ChangeLog.
2015-03-07 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (struct catcher) <exception>: No
longer a pointer to volatile exception. Now an exception value.
<mask>: Delete field.
(exceptions_state_mc_init): Remove all parameters. Adjust.
(exceptions_state_mc): No longer pop the catcher here.
(exceptions_state_mc_catch): New function.
(throw_exception): Adjust.
* common/common-exceptions.h (exceptions_state_mc_init): Remove
all parameters.
(exceptions_state_mc_catch): Declare.
(TRY_CATCH): Rename to ...
(TRY): ... this. Remove EXCEPTION and MASK parameters.
(CATCH, END_CATCH): New.
All callers adjusted.
gdb/gdbserver/ChangeLog:
2015-03-07 Pedro Alves <palves@redhat.com>
Adjust all callers of TRY_CATCH to use TRY/CATCH/END_CATCH
instead.
2015-03-07 23:14:14 +08:00
|
|
|
|
TRY
|
2011-09-21 23:21:28 +08:00
|
|
|
|
{
|
|
|
|
|
quit_cover ();
|
|
|
|
|
}
|
|
|
|
|
|
Split TRY_CATCH into TRY + CATCH
This patch splits the TRY_CATCH macro into three, so that we go from
this:
~~~
volatile gdb_exception ex;
TRY_CATCH (ex, RETURN_MASK_ERROR)
{
}
if (ex.reason < 0)
{
}
~~~
to this:
~~~
TRY
{
}
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
~~~
Thus, we'll be getting rid of the local volatile exception object, and
declaring the caught exception in the catch block.
This allows reimplementing TRY/CATCH in terms of C++ exceptions when
building in C++ mode, while still allowing to build GDB in C mode
(using setjmp/longjmp), as a transition step.
TBC, after this patch, is it _not_ valid to have code between the TRY
and the CATCH blocks, like:
TRY
{
}
// some code here.
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
Just like it isn't valid to do that with C++'s native try/catch.
By switching to creating the exception object inside the CATCH block
scope, we can get rid of all the explicitly allocated volatile
exception objects all over the tree, and map the CATCH block more
directly to C++'s catch blocks.
The majority of the TRY_CATCH -> TRY+CATCH+END_CATCH conversion was
done with a script, rerun from scratch at every rebase, no manual
editing involved. After the mechanical conversion, a few places
needed manual intervention, to fix preexisting cases where we were
using the exception object outside of the TRY_CATCH block, and cases
where we were using "else" after a 'if (ex.reason) < 0)' [a CATCH
after this patch]. The result was folded into this patch so that GDB
still builds at each incremental step.
END_CATCH is necessary for two reasons:
First, because we name the exception object in the CATCH block, which
requires creating a scope, which in turn must be closed somewhere.
Declaring the exception variable in the initializer field of a for
block, like:
#define CATCH(EXCEPTION, mask) \
for (struct gdb_exception EXCEPTION; \
exceptions_state_mc_catch (&EXCEPTION, MASK); \
EXCEPTION = exception_none)
would avoid needing END_CATCH, but alas, in C mode, we build with C90,
which doesn't allow mixed declarations and code.
Second, because when TRY/CATCH are wired to real C++ try/catch, as
long as we need to handle cleanup chains, even if there's no CATCH
block that wants to catch the exception, we need for stop at every
frame in the unwind chain and run cleanups, then rethrow. That will
be done in END_CATCH.
After we require C++, we'll still need TRY/CATCH/END_CATCH until
cleanups are completely phased out -- TRY/CATCH in C++ mode will
save/restore the current cleanup chain, like in C mode, and END_CATCH
catches otherwise uncaugh exceptions, runs cleanups and rethrows, so
that C++ cleanups and exceptions can coexist.
IMO, this still makes the TRY/CATCH code look a bit more like a
newcomer would expect, so IMO worth it even if we weren't considering
C++.
gdb/ChangeLog.
2015-03-07 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (struct catcher) <exception>: No
longer a pointer to volatile exception. Now an exception value.
<mask>: Delete field.
(exceptions_state_mc_init): Remove all parameters. Adjust.
(exceptions_state_mc): No longer pop the catcher here.
(exceptions_state_mc_catch): New function.
(throw_exception): Adjust.
* common/common-exceptions.h (exceptions_state_mc_init): Remove
all parameters.
(exceptions_state_mc_catch): Declare.
(TRY_CATCH): Rename to ...
(TRY): ... this. Remove EXCEPTION and MASK parameters.
(CATCH, END_CATCH): New.
All callers adjusted.
gdb/gdbserver/ChangeLog:
2015-03-07 Pedro Alves <palves@redhat.com>
Adjust all callers of TRY_CATCH to use TRY/CATCH/END_CATCH
instead.
2015-03-07 23:14:14 +08:00
|
|
|
|
CATCH (exception, RETURN_MASK_ALL)
|
2011-09-21 23:21:28 +08:00
|
|
|
|
{
|
|
|
|
|
fputs_filtered ("Could not kill the program being debugged",
|
|
|
|
|
gdb_stderr);
|
|
|
|
|
exception_print (gdb_stderr, exception);
|
|
|
|
|
}
|
Split TRY_CATCH into TRY + CATCH
This patch splits the TRY_CATCH macro into three, so that we go from
this:
~~~
volatile gdb_exception ex;
TRY_CATCH (ex, RETURN_MASK_ERROR)
{
}
if (ex.reason < 0)
{
}
~~~
to this:
~~~
TRY
{
}
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
~~~
Thus, we'll be getting rid of the local volatile exception object, and
declaring the caught exception in the catch block.
This allows reimplementing TRY/CATCH in terms of C++ exceptions when
building in C++ mode, while still allowing to build GDB in C mode
(using setjmp/longjmp), as a transition step.
TBC, after this patch, is it _not_ valid to have code between the TRY
and the CATCH blocks, like:
TRY
{
}
// some code here.
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
Just like it isn't valid to do that with C++'s native try/catch.
By switching to creating the exception object inside the CATCH block
scope, we can get rid of all the explicitly allocated volatile
exception objects all over the tree, and map the CATCH block more
directly to C++'s catch blocks.
The majority of the TRY_CATCH -> TRY+CATCH+END_CATCH conversion was
done with a script, rerun from scratch at every rebase, no manual
editing involved. After the mechanical conversion, a few places
needed manual intervention, to fix preexisting cases where we were
using the exception object outside of the TRY_CATCH block, and cases
where we were using "else" after a 'if (ex.reason) < 0)' [a CATCH
after this patch]. The result was folded into this patch so that GDB
still builds at each incremental step.
END_CATCH is necessary for two reasons:
First, because we name the exception object in the CATCH block, which
requires creating a scope, which in turn must be closed somewhere.
Declaring the exception variable in the initializer field of a for
block, like:
#define CATCH(EXCEPTION, mask) \
for (struct gdb_exception EXCEPTION; \
exceptions_state_mc_catch (&EXCEPTION, MASK); \
EXCEPTION = exception_none)
would avoid needing END_CATCH, but alas, in C mode, we build with C90,
which doesn't allow mixed declarations and code.
Second, because when TRY/CATCH are wired to real C++ try/catch, as
long as we need to handle cleanup chains, even if there's no CATCH
block that wants to catch the exception, we need for stop at every
frame in the unwind chain and run cleanups, then rethrow. That will
be done in END_CATCH.
After we require C++, we'll still need TRY/CATCH/END_CATCH until
cleanups are completely phased out -- TRY/CATCH in C++ mode will
save/restore the current cleanup chain, like in C mode, and END_CATCH
catches otherwise uncaugh exceptions, runs cleanups and rethrows, so
that C++ cleanups and exceptions can coexist.
IMO, this still makes the TRY/CATCH code look a bit more like a
newcomer would expect, so IMO worth it even if we weren't considering
C++.
gdb/ChangeLog.
2015-03-07 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (struct catcher) <exception>: No
longer a pointer to volatile exception. Now an exception value.
<mask>: Delete field.
(exceptions_state_mc_init): Remove all parameters. Adjust.
(exceptions_state_mc): No longer pop the catcher here.
(exceptions_state_mc_catch): New function.
(throw_exception): Adjust.
* common/common-exceptions.h (exceptions_state_mc_init): Remove
all parameters.
(exceptions_state_mc_catch): Declare.
(TRY_CATCH): Rename to ...
(TRY): ... this. Remove EXCEPTION and MASK parameters.
(CATCH, END_CATCH): New.
All callers adjusted.
gdb/gdbserver/ChangeLog:
2015-03-07 Pedro Alves <palves@redhat.com>
Adjust all callers of TRY_CATCH to use TRY/CATCH/END_CATCH
instead.
2015-03-07 23:14:14 +08:00
|
|
|
|
END_CATCH
|
2011-09-21 23:21:28 +08:00
|
|
|
|
|
Split TRY_CATCH into TRY + CATCH
This patch splits the TRY_CATCH macro into three, so that we go from
this:
~~~
volatile gdb_exception ex;
TRY_CATCH (ex, RETURN_MASK_ERROR)
{
}
if (ex.reason < 0)
{
}
~~~
to this:
~~~
TRY
{
}
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
~~~
Thus, we'll be getting rid of the local volatile exception object, and
declaring the caught exception in the catch block.
This allows reimplementing TRY/CATCH in terms of C++ exceptions when
building in C++ mode, while still allowing to build GDB in C mode
(using setjmp/longjmp), as a transition step.
TBC, after this patch, is it _not_ valid to have code between the TRY
and the CATCH blocks, like:
TRY
{
}
// some code here.
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
Just like it isn't valid to do that with C++'s native try/catch.
By switching to creating the exception object inside the CATCH block
scope, we can get rid of all the explicitly allocated volatile
exception objects all over the tree, and map the CATCH block more
directly to C++'s catch blocks.
The majority of the TRY_CATCH -> TRY+CATCH+END_CATCH conversion was
done with a script, rerun from scratch at every rebase, no manual
editing involved. After the mechanical conversion, a few places
needed manual intervention, to fix preexisting cases where we were
using the exception object outside of the TRY_CATCH block, and cases
where we were using "else" after a 'if (ex.reason) < 0)' [a CATCH
after this patch]. The result was folded into this patch so that GDB
still builds at each incremental step.
END_CATCH is necessary for two reasons:
First, because we name the exception object in the CATCH block, which
requires creating a scope, which in turn must be closed somewhere.
Declaring the exception variable in the initializer field of a for
block, like:
#define CATCH(EXCEPTION, mask) \
for (struct gdb_exception EXCEPTION; \
exceptions_state_mc_catch (&EXCEPTION, MASK); \
EXCEPTION = exception_none)
would avoid needing END_CATCH, but alas, in C mode, we build with C90,
which doesn't allow mixed declarations and code.
Second, because when TRY/CATCH are wired to real C++ try/catch, as
long as we need to handle cleanup chains, even if there's no CATCH
block that wants to catch the exception, we need for stop at every
frame in the unwind chain and run cleanups, then rethrow. That will
be done in END_CATCH.
After we require C++, we'll still need TRY/CATCH/END_CATCH until
cleanups are completely phased out -- TRY/CATCH in C++ mode will
save/restore the current cleanup chain, like in C mode, and END_CATCH
catches otherwise uncaugh exceptions, runs cleanups and rethrows, so
that C++ cleanups and exceptions can coexist.
IMO, this still makes the TRY/CATCH code look a bit more like a
newcomer would expect, so IMO worth it even if we weren't considering
C++.
gdb/ChangeLog.
2015-03-07 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (struct catcher) <exception>: No
longer a pointer to volatile exception. Now an exception value.
<mask>: Delete field.
(exceptions_state_mc_init): Remove all parameters. Adjust.
(exceptions_state_mc): No longer pop the catcher here.
(exceptions_state_mc_catch): New function.
(throw_exception): Adjust.
* common/common-exceptions.h (exceptions_state_mc_init): Remove
all parameters.
(exceptions_state_mc_catch): Declare.
(TRY_CATCH): Rename to ...
(TRY): ... this. Remove EXCEPTION and MASK parameters.
(CATCH, END_CATCH): New.
All callers adjusted.
gdb/gdbserver/ChangeLog:
2015-03-07 Pedro Alves <palves@redhat.com>
Adjust all callers of TRY_CATCH to use TRY/CATCH/END_CATCH
instead.
2015-03-07 23:14:14 +08:00
|
|
|
|
TRY
|
2011-09-21 23:21:28 +08:00
|
|
|
|
{
|
2013-03-20 23:46:24 +08:00
|
|
|
|
pop_all_targets ();
|
2011-09-21 23:21:28 +08:00
|
|
|
|
}
|
Split TRY_CATCH into TRY + CATCH
This patch splits the TRY_CATCH macro into three, so that we go from
this:
~~~
volatile gdb_exception ex;
TRY_CATCH (ex, RETURN_MASK_ERROR)
{
}
if (ex.reason < 0)
{
}
~~~
to this:
~~~
TRY
{
}
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
~~~
Thus, we'll be getting rid of the local volatile exception object, and
declaring the caught exception in the catch block.
This allows reimplementing TRY/CATCH in terms of C++ exceptions when
building in C++ mode, while still allowing to build GDB in C mode
(using setjmp/longjmp), as a transition step.
TBC, after this patch, is it _not_ valid to have code between the TRY
and the CATCH blocks, like:
TRY
{
}
// some code here.
CATCH (ex, RETURN_MASK_ERROR)
{
}
END_CATCH
Just like it isn't valid to do that with C++'s native try/catch.
By switching to creating the exception object inside the CATCH block
scope, we can get rid of all the explicitly allocated volatile
exception objects all over the tree, and map the CATCH block more
directly to C++'s catch blocks.
The majority of the TRY_CATCH -> TRY+CATCH+END_CATCH conversion was
done with a script, rerun from scratch at every rebase, no manual
editing involved. After the mechanical conversion, a few places
needed manual intervention, to fix preexisting cases where we were
using the exception object outside of the TRY_CATCH block, and cases
where we were using "else" after a 'if (ex.reason) < 0)' [a CATCH
after this patch]. The result was folded into this patch so that GDB
still builds at each incremental step.
END_CATCH is necessary for two reasons:
First, because we name the exception object in the CATCH block, which
requires creating a scope, which in turn must be closed somewhere.
Declaring the exception variable in the initializer field of a for
block, like:
#define CATCH(EXCEPTION, mask) \
for (struct gdb_exception EXCEPTION; \
exceptions_state_mc_catch (&EXCEPTION, MASK); \
EXCEPTION = exception_none)
would avoid needing END_CATCH, but alas, in C mode, we build with C90,
which doesn't allow mixed declarations and code.
Second, because when TRY/CATCH are wired to real C++ try/catch, as
long as we need to handle cleanup chains, even if there's no CATCH
block that wants to catch the exception, we need for stop at every
frame in the unwind chain and run cleanups, then rethrow. That will
be done in END_CATCH.
After we require C++, we'll still need TRY/CATCH/END_CATCH until
cleanups are completely phased out -- TRY/CATCH in C++ mode will
save/restore the current cleanup chain, like in C mode, and END_CATCH
catches otherwise uncaugh exceptions, runs cleanups and rethrows, so
that C++ cleanups and exceptions can coexist.
IMO, this still makes the TRY/CATCH code look a bit more like a
newcomer would expect, so IMO worth it even if we weren't considering
C++.
gdb/ChangeLog.
2015-03-07 Pedro Alves <palves@redhat.com>
* common/common-exceptions.c (struct catcher) <exception>: No
longer a pointer to volatile exception. Now an exception value.
<mask>: Delete field.
(exceptions_state_mc_init): Remove all parameters. Adjust.
(exceptions_state_mc): No longer pop the catcher here.
(exceptions_state_mc_catch): New function.
(throw_exception): Adjust.
* common/common-exceptions.h (exceptions_state_mc_init): Remove
all parameters.
(exceptions_state_mc_catch): Declare.
(TRY_CATCH): Rename to ...
(TRY): ... this. Remove EXCEPTION and MASK parameters.
(CATCH, END_CATCH): New.
All callers adjusted.
gdb/gdbserver/ChangeLog:
2015-03-07 Pedro Alves <palves@redhat.com>
Adjust all callers of TRY_CATCH to use TRY/CATCH/END_CATCH
instead.
2015-03-07 23:14:14 +08:00
|
|
|
|
CATCH (exception, RETURN_MASK_ALL)
|
|
|
|
|
{
|
|
|
|
|
}
|
|
|
|
|
END_CATCH
|
2011-09-21 23:21:28 +08:00
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
signal (SIGHUP, SIG_DFL); /*FIXME: ??????????? */
|
2009-01-09 19:00:00 +08:00
|
|
|
|
raise (SIGHUP);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
1999-06-15 02:08:47 +08:00
|
|
|
|
#ifdef STOP_SIGNAL
|
1999-07-08 04:19:36 +08:00
|
|
|
|
void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
handle_stop_sig (int sig)
|
1999-06-15 02:08:47 +08:00
|
|
|
|
{
|
2012-10-23 14:20:49 +08:00
|
|
|
|
mark_async_signal_handler (sigtstp_token);
|
1999-07-08 04:19:36 +08:00
|
|
|
|
signal (sig, handle_stop_sig);
|
1999-06-15 02:08:47 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
async_stop_sig (gdb_client_data arg)
|
1999-06-15 02:08:47 +08:00
|
|
|
|
{
|
2011-09-06 Pedro Alves <pedro@codesourcery.com>
* event-top.h (MAXPROMPTS, struct prompts): Delete.
(set_async_annotation_level, set_async_prompt, pop_prompt)
(push_prompt, new_async_prompt): Delete declarations.
* top.h (get_prompt, set_prompt): Change prototype.
(get_prefix, set_prefix, get_suffix, set_suffix): Delete
declarations.
* top.c (command_loop):
(top_prompt): New global.
(get_prefix, set_prefix, get_suffix, ): Delete.
(get_prompt, set_prompt): Rewrite.
(show_new_async_prompt): Rename to ...
(show_prompt): ... this.
(init_main): Adjust. Don't handle --annotate=2 here.
* event-top.c (new_async_prompt): Delete.
(the_prompts): Delete.
(more_to_come): Make static.
(display_gdb_prompt): Use top_level_prompt() to compute the top
level prompt, and don't notify the before_prompt observers
directly here. Always trick readline into not trying to display
the prompt if sync_execution and displaying the primary prompt.
If displaying a local/secondary prompt, always show it, even if
sync_execution is set.
(change_annotation_level): Delete.
(top_level_prompt): New, based on change_annotation_level.
(push_prompt, pop_prompt): Delete.
(async_disable_stdin): No longer pushes prompt.
(command_line_handler): No longer pushes or pops prompt. If more
input is expected, call display_gdb_prompt with an explicit empty
prompt.
(async_stop_sig): Adjust.
(set_async_annotation_level, set_async_prompt): Delete.
* python/python.c (before_prompt_hook): Adjust.
2011-09-06 22:49:00 +08:00
|
|
|
|
char *prompt = get_prompt ();
|
2010-05-15 02:35:11 +08:00
|
|
|
|
|
1999-06-15 02:08:47 +08:00
|
|
|
|
#if STOP_SIGNAL == SIGTSTP
|
|
|
|
|
signal (SIGTSTP, SIG_DFL);
|
1999-10-06 07:13:56 +08:00
|
|
|
|
#if HAVE_SIGPROCMASK
|
|
|
|
|
{
|
|
|
|
|
sigset_t zero;
|
2001-08-28 06:39:56 +08:00
|
|
|
|
|
1999-10-06 07:13:56 +08:00
|
|
|
|
sigemptyset (&zero);
|
|
|
|
|
sigprocmask (SIG_SETMASK, &zero, 0);
|
|
|
|
|
}
|
2001-08-28 06:39:56 +08:00
|
|
|
|
#elif HAVE_SIGSETMASK
|
1999-06-15 02:08:47 +08:00
|
|
|
|
sigsetmask (0);
|
1999-10-06 07:13:56 +08:00
|
|
|
|
#endif
|
2009-01-09 19:00:00 +08:00
|
|
|
|
raise (SIGTSTP);
|
1999-06-15 02:08:47 +08:00
|
|
|
|
signal (SIGTSTP, handle_stop_sig);
|
|
|
|
|
#else
|
|
|
|
|
signal (STOP_SIGNAL, handle_stop_sig);
|
|
|
|
|
#endif
|
|
|
|
|
printf_unfiltered ("%s", prompt);
|
|
|
|
|
gdb_flush (gdb_stdout);
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Forget about any previous command -- null line now will do
|
|
|
|
|
nothing. */
|
1999-06-15 02:08:47 +08:00
|
|
|
|
dont_repeat ();
|
|
|
|
|
}
|
|
|
|
|
#endif /* STOP_SIGNAL */
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Tell the event loop what to do if SIGFPE is received.
|
|
|
|
|
See event-signal.c. */
|
1999-07-08 04:19:36 +08:00
|
|
|
|
static void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
handle_sigfpe (int sig)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2012-10-23 14:20:49 +08:00
|
|
|
|
mark_async_signal_handler (sigfpe_token);
|
1999-05-12 04:29:07 +08:00
|
|
|
|
signal (sig, handle_sigfpe);
|
|
|
|
|
}
|
|
|
|
|
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Event loop will call this functin to process a SIGFPE. */
|
1999-07-08 04:19:36 +08:00
|
|
|
|
static void
|
1999-09-22 11:28:34 +08:00
|
|
|
|
async_float_handler (gdb_client_data arg)
|
1999-05-12 04:29:07 +08:00
|
|
|
|
{
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* This message is based on ANSI C, section 4.7. Note that integer
|
|
|
|
|
divide by zero causes this, so "float" is a misnomer. */
|
2005-02-10 Andrew Cagney <cagney@gnu.org>
Mark up all error and warning messages.
* ada-lang.c, amd64-tdep.c, arch-utils.c, breakpoint.c: Update.
* bsd-kvm.c, bsd-uthread.c, coff-solib.h, coffread.c: Update.
* core-aout.c, core-regset.c, corefile.c, corelow.c: Update.
* cp-abi.c, cp-support.c, cp-valprint.c, cris-tdep.c: Update.
* dbxread.c, demangle.c, doublest.c, dsrec.c: Update.
* dve3900-rom.c, dwarf2expr.c, dwarf2loc.c: Update.
* dwarf2read.c, dwarfread.c, elfread.c, eval.c: Update.
* event-top.c, exec.c, expprint.c, f-lang.c: Update.
* f-typeprint.c, f-valprint.c, fbsd-nat.c, findvar.c: Update.
* frame.c, frv-linux-tdep.c, gcore.c, gdbtypes.c: Update.
* gnu-nat.c, gnu-v2-abi.c, gnu-v3-abi.c, go32-nat.c: Update.
* hpacc-abi.c, hppa-hpux-nat.c, hppa-hpux-tdep.c: Update.
* hppa-linux-nat.c, hppa-linux-tdep.c, hppa-tdep.c: Update.
* hpread.c, hpux-thread.c, i386-linux-nat.c: Update.
* i386-linux-tdep.c, i386-tdep.c, i386bsd-nat.c: Update.
* i386gnu-nat.c, i387-tdep.c, ia64-linux-nat.c: Update.
* ia64-tdep.c, inf-child.c, inf-ptrace.c, inf-ttrace.c: Update.
* infcall.c, infcmd.c, inflow.c, infptrace.c, infrun.c: Update.
* inftarg.c, interps.c, irix5-nat.c, jv-lang.c: Update.
* kod-cisco.c, kod.c, language.c, libunwind-frame.c: Update.
* linespec.c, linux-nat.c, linux-thread-db.c, m2-lang.c: Update.
* m32r-rom.c, m68hc11-tdep.c, m68k-tdep.c: Update.
* m68klinux-nat.c, macrocmd.c, macroexp.c, main.c: Update.
* maint.c, mdebugread.c, mem-break.c, memattr.c: Update.
* mips-linux-tdep.c, mips-tdep.c, mipsread.c, monitor.c: Update.
* nlmread.c, nto-procfs.c, objc-lang.c, objfiles.c: Update.
* observer.c, ocd.c, p-lang.c, p-typeprint.c: Update.
* p-valprint.c, pa64solib.c, parse.c, ppc-linux-tdep.c: Update.
* ppcnbsd-tdep.c, printcmd.c, procfs.c, remote-e7000.c: Update.
* remote-fileio.c, remote-m32r-sdi.c, remote-rdi.c: Update.
* remote-rdp.c, remote-sim.c, remote-st.c: Update.
* remote-utils.c, remote-utils.h, remote.c: Update.
* rom68k-rom.c, rs6000-nat.c, s390-tdep.c, scm-lang.c: Update.
* ser-e7kpc.c, ser-tcp.c, ser-unix.c, sh-tdep.c: Update.
* sh3-rom.c, shnbsd-tdep.c, sol-thread.c, solib-aix5.c: Update.
* solib-frv.c, solib-irix.c, solib-osf.c, solib-pa64.c: Update.
* solib-som.c, solib-sunos.c, solib-svr4.c, solib.c: Update.
* somread.c, somsolib.c, source.c, stabsread.c: Update.
* stack.c, std-regs.c, symfile-mem.c, symfile.c: Update.
* symmisc.c, symtab.c, target.c, thread.c, top.c: Update.
* tracepoint.c, trad-frame.c, typeprint.c, utils.c: Update.
* uw-thread.c, valarith.c, valops.c, valprint.c: Update.
* value.c, varobj.c, version.in, win32-nat.c, wince.c: Update.
* xcoffread.c, xcoffsolib.c, cli/cli-cmds.c: Update.
* cli/cli-decode.c, cli/cli-dump.c, cli/cli-logging.c: Update.
* cli/cli-script.c, cli/cli-setshow.c, mi/mi-cmd-break.c: Update.
* mi/mi-cmd-disas.c, mi/mi-cmd-env.c, mi/mi-cmd-file.c: Update.
* mi/mi-cmd-stack.c, mi/mi-cmd-var.c, mi/mi-getopt.c: Update.
* mi/mi-symbol-cmds.c, tui/tui-layout.c, tui/tui-stack.c: Update.
* tui/tui-win.c: Update.
2005-02-11 12:06:14 +08:00
|
|
|
|
error (_("Erroneous arithmetic operation."));
|
1999-05-12 04:29:07 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
1999-06-15 02:08:47 +08:00
|
|
|
|
/* Set things up for readline to be invoked via the alternate
|
2016-04-22 23:18:33 +08:00
|
|
|
|
interface, i.e. via a callback function
|
|
|
|
|
(gdb_rl_callback_read_char), and hook up instream to the event
|
|
|
|
|
loop. */
|
|
|
|
|
|
1999-06-15 02:08:47 +08:00
|
|
|
|
void
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
gdb_setup_readline (int editing)
|
1999-06-15 02:08:47 +08:00
|
|
|
|
{
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
|
|
|
|
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
/* This function is a noop for the sync case. The assumption is
|
|
|
|
|
that the sync setup is ALL done in gdb_init, and we would only
|
|
|
|
|
mess it up here. The sync stuff should really go away over
|
|
|
|
|
time. */
|
2005-11-01 19:09:18 +08:00
|
|
|
|
if (!batch_silent)
|
Eliminate make_cleanup_ui_file_delete / make ui_file a class hierarchy
This patch starts from the desire to eliminate
make_cleanup_ui_file_delete, but then goes beyond. It makes ui_file &
friends a real C++ class hierarchy, and switches temporary
ui_file-like objects to stack-based allocation.
- mem_fileopen -> string_file
mem_fileopen is replaced with a new string_file class that is treated
as a value class created on the stack. This alone eliminates most
make_cleanup_ui_file_delete calls, and, simplifies code a whole lot
(diffstat shows around 1k loc dropped.)
string_file's internal buffer is a std::string, thus the "string" in
the name. This simplifies the implementation much, compared to
mem_fileopen, which managed growing its internal buffer manually.
- ui_file_as_string, ui_file_strdup, ui_file_obsavestring all gone
The new string_file class has a string() method that provides direct
writable access to the internal std::string buffer. This replaced
ui_file_as_string, which forced a copy of the same data the stream had
inside. With direct access via a writable reference, we can instead
move the string out of the string_stream, avoiding deep string
copying.
Related, ui_file_xstrdup calls are replaced with xstrdup'ping the
stream's string, and ui_file_obsavestring is replaced by
obstack_copy0.
With all those out of the way, getting rid of the weird ui_file_put
mechanism was possible.
- New ui_file::printf, ui_file::puts, etc. methods
These simplify / clarify client code. I considered splitting
client-code changes, like these, e.g.:
- stb = mem_fileopen ();
- fprintf_unfiltered (stb, "%s%s%s",
- _("The valid values are:\n"),
- regdesc,
- _("The default is \"std\"."));
+ string_file stb;
+ stb.printf ("%s%s%s",
+ _("The valid values are:\n"),
+ regdesc,
+ _("The default is \"std\"."));
In two steps, with the first step leaving fprintf_unfiltered (etc.)
calls in place, and only afterwards do a pass to change all those to
call stb.printf etc.. I didn't do that split, because (when I tried),
it turned out to be pointless make-work: the first pass would have to
touch the fprintf_unfiltered line anyway, to replace "stb" with
"&stb".
- gdb_fopen replaced with stack-based objects
This avoids the need for cleanups or unique_ptr's. I.e., this:
struct ui_file *file = gdb_fopen (filename, "w");
if (filename == NULL)
perror_with_name (filename);
cleanups = make_cleanup_ui_file_delete (file);
// use file.
do_cleanups (cleanups);
is replaced with this:
stdio_file file;
if (!file.open (filename, "w"))
perror_with_name (filename);
// use file.
- odd contorsions in null_file_write / null_file_fputs around when to
call to_fputs / to_write eliminated.
- Global null_stream object
A few places that were allocating a ui_file in order to print to
"nowhere" are adjusted to instead refer to a new 'null_stream' global
stream.
- TUI's tui_sfileopen eliminated. TUI's ui_file much simplified
The TUI's ui_file was serving a dual purpose. It supported being used
as string buffer, and supported being backed by a stdio FILE. The
string buffer part is gone, replaced by using of string_file. The
'FILE *' support is now much simplified, by making the TUI's ui_file
inherit from stdio_file.
gdb/ChangeLog:
2017-02-02 Pedro Alves <palves@redhat.com>
* ada-lang.c (type_as_string): Use string_file.
* ada-valprint.c (ada_print_floating): Use string_file.
* ada-varobj.c (ada_varobj_scalar_image)
(ada_varobj_get_value_image): Use string_file.
* aix-thread.c (aix_thread_extra_thread_info): Use string_file.
* arm-tdep.c (_initialize_arm_tdep): Use string_printf.
* breakpoint.c (update_inserted_breakpoint_locations)
(insert_breakpoint_locations, reattach_breakpoints)
(print_breakpoint_location, print_one_detail_ranged_breakpoint)
(print_it_watchpoint): Use string_file.
(save_breakpoints): Use stdio_file.
* c-exp.y (oper): Use string_file.
* cli/cli-logging.c (set_logging_redirect): Use ui_file_up and
tee_file.
(pop_output_files): Use delete.
(handle_redirections): Use stdio_file and tee_file.
* cli/cli-setshow.c (do_show_command): Use string_file.
* compile/compile-c-support.c (c_compute_program): Use
string_file.
* compile/compile-c-symbols.c (generate_vla_size): Take a
'string_file &' instead of a 'ui_file *'.
(generate_c_for_for_one_variable): Take a 'string_file &' instead
of a 'ui_file *'. Use string_file.
(generate_c_for_variable_locations): Take a 'string_file &'
instead of a 'ui_file *'.
* compile/compile-internal.h (generate_c_for_for_one_variable):
Take a 'string_file &' instead of a 'ui_file *'.
* compile/compile-loc2c.c (push, pushf, unary, binary)
(print_label, pushf_register_address, pushf_register)
(do_compile_dwarf_expr_to_c): Take a 'string_file &' instead of a
'ui_file *'. Adjust.
* compile/compile.c (compile_to_object): Use string_file.
* compile/compile.h (compile_dwarf_expr_to_c)
(compile_dwarf_bounds_to_c): Take a 'string_file &' instead of a
'ui_file *'.
* cp-support.c (inspect_type): Use string_file and obstack_copy0.
(replace_typedefs_qualified_name): Use string_file and
obstack_copy0.
* disasm.c (gdb_pretty_print_insn): Use string_file.
(gdb_disassembly): Adjust reference the null_stream global.
(do_ui_file_delete): Delete.
(gdb_insn_length): Use null_stream.
* dummy-frame.c (maintenance_print_dummy_frames): Use stdio_file.
* dwarf2loc.c (dwarf2_compile_property_to_c)
(locexpr_generate_c_location, loclist_generate_c_location): Take a
'string_file &' instead of a 'ui_file *'.
* dwarf2loc.h (dwarf2_compile_property_to_c): Likewise.
* dwarf2read.c (do_ui_file_peek_last): Delete.
(dwarf2_compute_name): Use string_file.
* event-top.c (gdb_setup_readline): Use stdio_file.
* gdbarch.sh (verify_gdbarch): Use string_file.
* gdbtypes.c (safe_parse_type): Use null_stream.
* guile/scm-breakpoint.c (gdbscm_breakpoint_commands): Use
string_file.
* guile/scm-disasm.c (gdbscm_print_insn_from_port): Take a
'string_file *' instead of a 'ui_file *'.
(gdbscm_arch_disassemble): Use string_file.
* guile/scm-frame.c (frscm_print_frame_smob): Use string_file.
* guile/scm-ports.c (class ioscm_file_port): Now a class that
inherits from ui_file.
(ioscm_file_port_delete, ioscm_file_port_rewind)
(ioscm_file_port_put): Delete.
(ioscm_file_port_write): Rename to ...
(ioscm_file_port::write): ... this. Remove file_port_magic
checks.
(ioscm_file_port_new): Delete.
(ioscm_with_output_to_port_worker): Use ioscm_file_port and
ui_file_up.
* guile/scm-type.c (tyscm_type_name): Use string_file.
* guile/scm-value.c (vlscm_print_value_smob, gdbscm_value_print):
Use string_file.
* infcmd.c (print_return_value_1): Use string_file.
* infrun.c (print_target_wait_results): Use string_file.
* language.c (add_language): Use string_file.
* location.c (explicit_to_string_internal): Use string_file.
* main.c (captured_main_1): Use null_file.
* maint.c (maintenance_print_architecture): Use stdio_file.
* mi/mi-cmd-stack.c (list_arg_or_local): Use string_file.
* mi/mi-common.h (struct mi_interp) <out, err, log, targ,
event_channel>: Change type to mi_console_file pointer.
* mi/mi-console.c (mi_console_file_fputs, mi_console_file_flush)
(mi_console_file_delete): Delete.
(struct mi_console_file): Delete.
(mi_console_file_magic): Delete.
(mi_console_file_new): Delete.
(mi_console_file::mi_console_file): New.
(mi_console_file_delete): Delete.
(mi_console_file_fputs): Delete.
(mi_console_file::write): New.
(mi_console_raw_packet): Delete.
(mi_console_file::flush): New.
(mi_console_file_flush): Delete.
(mi_console_set_raw): Rename to ...
(mi_console_file::set_raw): ... this.
* mi/mi-console.h (class mi_console_file): New class.
(mi_console_file_new, mi_console_set_raw): Delete.
* mi/mi-interp.c (mi_interpreter_init): Use mi_console_file.
(mi_set_logging): Use delete and tee_file. Adjust.
* mi/mi-main.c (output_register): Use string_file.
(mi_cmd_data_evaluate_expression): Use string_file.
(mi_cmd_data_read_memory): Use string_file.
(mi_cmd_execute, print_variable_or_computed): Use string_file.
* mi/mi-out.c (mi_ui_out::main_stream): New.
(mi_ui_out::rewind): Use main_stream and
string_file.
(mi_ui_out::put): Use main_stream and string_file.
(mi_ui_out::mi_ui_out): Remove 'stream' parameter.
Allocate a 'string_file' instead.
(mi_out_new): Don't allocate a mem_fileopen stream here.
* mi/mi-out.h (mi_ui_out::mi_ui_out): Remove 'stream' parameter.
(mi_ui_out::main_stream): Declare method.
* printcmd.c (eval_command): Use string_file.
* psymtab.c (maintenance_print_psymbols): Use stdio_file.
* python/py-arch.c (archpy_disassemble): Use string_file.
* python/py-breakpoint.c (bppy_get_commands): Use string_file.
* python/py-frame.c (frapy_str): Use string_file.
* python/py-framefilter.c (py_print_type, py_print_single_arg):
Use string_file.
* python/py-type.c (typy_str): Use string_file.
* python/py-unwind.c (unwind_infopy_str): Use string_file.
* python/py-value.c (valpy_str): Use string_file.
* record-btrace.c (btrace_insn_history): Use string_file.
* regcache.c (regcache_print): Use stdio_file.
* reggroups.c (maintenance_print_reggroups): Use stdio_file.
* remote.c (escape_buffer): Use string_file.
* rust-lang.c (rust_get_disr_info): Use string_file.
* serial.c (serial_open_ops_1): Use stdio_file.
(do_serial_close): Use delete.
* stack.c (print_frame_arg): Use string_file.
(print_frame_args): Remove local mem_fileopen stream, not used.
(print_frame): Use string_file.
* symmisc.c (maintenance_print_symbols): Use stdio_file.
* symtab.h (struct symbol_computed_ops) <generate_c_location>:
Take a 'string_file *' instead of a 'ui_file *'.
* top.c (new_ui): Use stdio_file and stderr_file.
(free_ui): Use delete.
(execute_command_to_string): Use string_file.
(quit_confirm): Use string_file.
* tracepoint.c (collection_list::append_exp): Use string_file.
* tui/tui-disasm.c (tui_disassemble): Use string_file.
* tui/tui-file.c: Don't include "ui-file.h".
(enum streamtype, struct tui_stream): Delete.
(tui_file_new, tui_file_delete, tui_fileopen, tui_sfileopen)
(tui_file_isatty, tui_file_rewind, tui_file_put): Delete.
(tui_file::tui_file): New method.
(tui_file_fputs): Delete.
(tui_file_get_strbuf): Delete.
(tui_file::puts): New method.
(tui_file_adjust_strbuf): Delete.
(tui_file_flush): Delete.
(tui_file::flush): New method.
* tui/tui-file.h: Tweak intro comment.
Include ui-file.h.
(tui_fileopen, tui_sfileopen, tui_file_get_strbuf)
(tui_file_adjust_strbuf): Delete declarations.
(class tui_file): New class.
* tui/tui-io.c (tui_initialize_io): Use tui_file.
* tui/tui-regs.c (tui_restore_gdbout): Use delete.
(tui_register_format): Use string_stream.
* tui/tui-stack.c (tui_make_status_line): Use string_file.
(tui_get_function_from_frame): Use string_file.
* typeprint.c (type_to_string): Use string_file.
* ui-file.c (struct ui_file, ui_file_magic, ui_file_new): Delete.
(null_stream): New global.
(ui_file_delete): Delete.
(ui_file::ui_file): New.
(null_file_isatty): Delete.
(ui_file::~ui_file): New.
(null_file_rewind): Delete.
(ui_file::printf): New.
(null_file_put): Delete.
(null_file_flush): Delete.
(ui_file::putstr): New.
(null_file_write): Delete.
(ui_file::putstrn): New.
(null_file_read): Delete.
(ui_file::putc): New.
(null_file_fputs): Delete.
(null_file_write_async_safe): Delete.
(ui_file::vprintf): New.
(null_file_delete): Delete.
(null_file::write): New.
(null_file_fseek): Delete.
(null_file::puts): New.
(ui_file_data): Delete.
(null_file::write_async_safe): New.
(gdb_flush, ui_file_isatty): Adjust.
(ui_file_put, ui_file_rewind): Delete.
(ui_file_write): Adjust.
(ui_file_write_for_put): Delete.
(ui_file_write_async_safe, ui_file_read): Adjust.
(ui_file_fseek): Delete.
(fputs_unfiltered): Adjust.
(set_ui_file_flush, set_ui_file_isatty, set_ui_file_rewind)
(set_ui_file_put, set_ui_file_write, set_ui_file_write_async_safe)
(set_ui_file_read, set_ui_file_fputs, set_ui_file_fseek)
(set_ui_file_data): Delete.
(string_file::~string_file, string_file::write)
(struct accumulated_ui_file, do_ui_file_xstrdup, ui_file_xstrdup)
(do_ui_file_as_string, ui_file_as_string): Delete.
(do_ui_file_obsavestring, ui_file_obsavestring): Delete.
(struct mem_file): Delete.
(mem_file_new): Delete.
(stdio_file::stdio_file): New.
(mem_file_delete): Delete.
(stdio_file::stdio_file): New.
(mem_fileopen): Delete.
(stdio_file::~stdio_file): New.
(mem_file_rewind): Delete.
(stdio_file::set_stream): New.
(mem_file_put): Delete.
(stdio_file::open): New.
(mem_file_write): Delete.
(stdio_file_magic, struct stdio_file): Delete.
(stdio_file_new, stdio_file_delete, stdio_file_flush): Delete.
(stdio_file::flush): New.
(stdio_file_read): Rename to ...
(stdio_file::read): ... this. Adjust.
(stdio_file_write): Rename to ...
(stdio_file::write): ... this. Adjust.
(stdio_file_write_async_safe): Rename to ...
(stdio_file::write_async_safe) ... this. Adjust.
(stdio_file_fputs): Rename to ...
(stdio_file::puts) ... this. Adjust.
(stdio_file_isatty): Delete.
(stdio_file_fseek): Delete.
(stdio_file::isatty): New.
(stderr_file_write): Rename to ...
(stderr_file::write) ... this. Adjust.
(stderr_file_fputs): Rename to ...
(stderr_file::puts) ... this. Adjust.
(stderr_fileopen, stdio_fileopen, gdb_fopen): Delete.
(stderr_file::stderr_file): New.
(tee_file_magic): Delete.
(struct tee_file): Delete.
(tee_file::tee_file): New.
(tee_file_new): Delete.
(tee_file::~tee_file): New.
(tee_file_delete): Delete.
(tee_file_flush): Rename to ...
(tee_file::flush): ... this. Adjust.
(tee_file_write): Rename to ...
(tee_file::write): ... this. Adjust.
(tee_file::write_async_safe): New.
(tee_file_fputs): Rename to ...
(tee_file::puts): ... this. Adjust.
(tee_file_isatty): Rename to ...
(tee_file::isatty): ... this. Adjust.
* ui-file.h (struct obstack, struct ui_file): Don't
forward-declare.
(ui_file_new, ui_file_flush_ftype, set_ui_file_flush)
(ui_file_write_ftype)
(set_ui_file_write, ui_file_fputs_ftype, set_ui_file_fputs)
(ui_file_write_async_safe_ftype, set_ui_file_write_async_safe)
(ui_file_read_ftype, set_ui_file_read, ui_file_isatty_ftype)
(set_ui_file_isatty, ui_file_rewind_ftype, set_ui_file_rewind)
(ui_file_put_method_ftype, ui_file_put_ftype, set_ui_file_put)
(ui_file_delete_ftype, set_ui_file_data, ui_file_fseek_ftype)
(set_ui_file_fseek): Delete.
(ui_file_data, ui_file_delete, ui_file_rewind)
(struct ui_file): New.
(ui_file_up): New.
(class null_file): New.
(null_stream): Declare.
(ui_file_write_for_put, ui_file_put): Delete.
(ui_file_xstrdup, ui_file_as_string, ui_file_obsavestring):
Delete.
(ui_file_fseek, mem_fileopen, stdio_fileopen, stderr_fileopen)
(gdb_fopen, tee_file_new): Delete.
(struct string_file): New.
(struct stdio_file): New.
(stdio_file_up): New.
(struct stderr_file): New.
(class tee_file): New.
* ui-out.c (ui_out::field_stream): Take a 'string_file &' instead
of a 'ui_file *'. Adjust.
* ui-out.h (class ui_out) <field_stream>: Likewise.
* utils.c (do_ui_file_delete, make_cleanup_ui_file_delete)
(null_stream): Delete.
(error_stream): Take a 'string_file &' instead of a 'ui_file *'.
Adjust.
* utils.h (struct ui_file): Delete forward declaration..
(make_cleanup_ui_file_delete, null_stream): Delete declarations.
(error_stream): Take a 'string_file &' instead of a
'ui_file *'.
* varobj.c (varobj_value_get_print_value): Use string_file.
* xtensa-tdep.c (xtensa_verify_config): Use string_file.
* gdbarch.c: Regenerate.
2017-02-02 19:11:47 +08:00
|
|
|
|
gdb_stdout = new stdio_file (ui->outstream);
|
|
|
|
|
gdb_stderr = new stderr_file (ui->errstream);
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
gdb_stdlog = gdb_stderr; /* for moment */
|
|
|
|
|
gdb_stdtarg = gdb_stderr; /* for moment */
|
2010-09-12 00:00:27 +08:00
|
|
|
|
gdb_stdtargerr = gdb_stderr; /* for moment */
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
/* If the input stream is connected to a terminal, turn on editing.
|
|
|
|
|
However, that is only allowed on the main UI, as we can only have
|
|
|
|
|
one instance of readline. */
|
|
|
|
|
if (ISATTY (ui->instream) && editing && ui == main_ui)
|
1999-06-21 21:27:42 +08:00
|
|
|
|
{
|
2010-12-29 08:58:14 +08:00
|
|
|
|
/* Tell gdb that we will be using the readline library. This
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
could be overwritten by a command in .gdbinit like 'set
|
|
|
|
|
editing on' or 'off'. */
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
ui->command_editing = 1;
|
|
|
|
|
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
/* When a character is detected on instream by select or poll,
|
|
|
|
|
readline will be invoked via this callback function. */
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
ui->call_readline = gdb_rl_callback_read_char_wrapper;
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
|
|
|
|
|
/* Tell readline to use the same input stream that gdb uses. */
|
|
|
|
|
rl_instream = ui->instream;
|
1999-06-21 21:27:42 +08:00
|
|
|
|
}
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
else
|
|
|
|
|
{
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
ui->command_editing = 0;
|
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.
2016-06-21 08:11:44 +08:00
|
|
|
|
ui->call_readline = gdb_readline_no_editing_callback;
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
}
|
|
|
|
|
|
2016-06-21 08:11:47 +08:00
|
|
|
|
/* Now create the event source for this UI's input file descriptor.
|
|
|
|
|
Another source is going to be the target program (inferior), but
|
|
|
|
|
that must be registered only when it actually exists (I.e. after
|
|
|
|
|
we say 'run' or after we connect to a remote target. */
|
Fix PR gdb/20418 - Problems with synchronous commands and new-ui
When executing commands on a secondary UI running the MI interpreter,
some commands that should be synchronous are not. MI incorrectly
continues processing input right after the synchronous command is
sent, before the target stops.
The problem happens when we emit MI async events (=library-loaded,
etc.), and we go about restoring the previous terminal state, we end
up calling target_terminal_ours, which incorrectly always installs the
current UI's input_fd in the event loop... That is, code like this:
old_chain = make_cleanup_restore_target_terminal ();
target_terminal_ours_for_output ();
fprintf_unfiltered (mi->event_channel, "library-loaded");
...
do_cleanups (old_chain);
The fix is to move the add_file_handler/delete_file_handler calls out
of target_terminal_$foo, making these completely no-ops unless called
with the main UI as current UI.
gdb/ChangeLog:
2016-08-09 Pedro Alves <palves@redhat.com>
PR gdb/20418
* event-top.c (ui_register_input_event_handler)
(ui_unregister_input_event_handler): New functions.
(async_enable_stdin): Register input in the event loop.
(async_disable_stdin): Unregister input from the event loop.
(gdb_setup_readline): Register input in the event loop.
* infrun.c (check_curr_ui_sync_execution_done): Register input in
the event loop.
* target.c (target_terminal_inferior): Don't unregister input from
the event loop.
(target_terminal_ours): Don't register input in the event loop.
* target.h (target_terminal_inferior)
(target_terminal_ours_for_output, target_terminal_ours): Update
comments.
* top.h (ui_register_input_event_handler)
(ui_unregister_input_event_handler): New declarations.
* utils.c (ui_unregister_input_event_handler_cleanup)
(prepare_to_handle_input): New functions.
(defaulted_query, prompt_for_continue): Use
prepare_to_handle_input.
gdb/testsuite/ChangeLog:
2016-08-09 Pedro Alves <palves@redhat.com>
Simon Marchi <simon.marchi@ericsson.com>
PR gdb/20418
* gdb.mi/new-ui-mi-sync.c, gdb.mi/new-ui-mi-sync.exp: New files.
* lib/mi-support.exp (mi_expect_interrupt): Remove anchors.
2016-08-10 05:45:40 +08:00
|
|
|
|
ui_register_input_event_handler (ui);
|
1999-06-15 02:08:47 +08:00
|
|
|
|
}
|
2002-10-03 09:23:50 +08:00
|
|
|
|
|
2002-11-06 06:38:11 +08:00
|
|
|
|
/* Disable command input through the standard CLI channels. Used in
|
|
|
|
|
the suspend proc for interpreters that use the standard gdb readline
|
|
|
|
|
interface, like the cli & the mi. */
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
|
2002-11-06 06:38:11 +08:00
|
|
|
|
void
|
|
|
|
|
gdb_disable_readline (void)
|
|
|
|
|
{
|
2016-06-21 08:11:47 +08:00
|
|
|
|
struct ui *ui = current_ui;
|
|
|
|
|
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
/* FIXME - It is too heavyweight to delete and remake these every
|
|
|
|
|
time you run an interpreter that needs readline. It is probably
|
|
|
|
|
better to have the interpreters cache these, which in turn means
|
|
|
|
|
that this needs to be moved into interpreter specific code. */
|
2002-11-06 06:38:11 +08:00
|
|
|
|
|
|
|
|
|
#if 0
|
2004-09-13 Andrew Cagney <cagney@gnu.org>
Eliminate event_loop_p, always has the value 1.
* defs.h (event_loop_p): Delete macro.
* breakpoint.c (until_break_command): Simplify.
* utils.c (prompt_for_continue): Simplify.
* tracepoint.c (read_actions): Simplify.
* top.c (throw_exception, execute_command, gdb_readline_wrapper)
(gdb_rl_operate_and_get_next, command_line_input, get_prompt)
(set_prompt, init_main): Simplify.
(init_signals, disconnect): Delete, unused.
* remote.c (remote_async_resume)
(extended_remote_async_create_inferior): Simplify.
* mi/mi-interp.c (mi_input): Delete, unused.
(mi_interpreter_resume, mi_command_loop): Simplify.
* interps.c (current_interp_command_loop): Simplify.
* infrun.c (proceed): Simplify.
* infcmd.c (run_command, continue_command, step_1, jump_command)
(until_command, advance_command, finish_command)
(interrupt_target_command): Simplify.
* event-top.c (gdb_setup_readline, gdb_disable_readline): Simplify.
2004-09-14 02:26:31 +08:00
|
|
|
|
ui_file_delete (gdb_stdout);
|
|
|
|
|
ui_file_delete (gdb_stderr);
|
|
|
|
|
gdb_stdlog = NULL;
|
|
|
|
|
gdb_stdtarg = NULL;
|
2010-09-12 00:00:27 +08:00
|
|
|
|
gdb_stdtargerr = NULL;
|
2002-11-06 06:38:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
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>
PR mi/20034
* gdb.mi/mi-editing.exp: New file.
2016-06-21 08:11:48 +08:00
|
|
|
|
if (ui->command_editing)
|
|
|
|
|
gdb_rl_callback_handler_remove ();
|
2016-06-21 08:11:47 +08:00
|
|
|
|
delete_file_handler (ui->input_fd);
|
2002-11-06 06:38:11 +08:00
|
|
|
|
}
|