1999-05-11 22:29:07 +02:00
|
|
|
|
/* Top level stuff for GDB, the GNU debugger.
|
2005-01-17 17:17:36 +01:00
|
|
|
|
|
2016-01-01 05:33:14 +01:00
|
|
|
|
Copyright (C) 1999-2016 Free Software Foundation, Inc.
|
2005-01-17 17:17:36 +01:00
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
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
|
2007-08-23 20:08:50 +02:00
|
|
|
|
the Free Software Foundation; either version 3 of the License, or
|
1999-05-11 22:29:07 +02:00
|
|
|
|
(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 01:58:14 +01:00
|
|
|
|
along with this program. If not, see <http://www.gnu.org/licenses/>. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
#include "defs.h"
|
1999-06-14 20:08:47 +02:00
|
|
|
|
#include "top.h"
|
1999-05-11 22:29:07 +02:00
|
|
|
|
#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 13:29:11 +02:00
|
|
|
|
#include "infrun.h"
|
1999-10-26 05:43:48 +02:00
|
|
|
|
#include "target.h"
|
1999-07-07 22:19:36 +02:00
|
|
|
|
#include "terminal.h" /* for job_control */
|
1999-06-21 15:27:42 +02:00
|
|
|
|
#include "event-loop.h"
|
1999-09-22 05:28:34 +02:00
|
|
|
|
#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 02:19:12 +01:00
|
|
|
|
#include "interps.h"
|
2001-02-06 05:17:03 +01:00
|
|
|
|
#include <signal.h>
|
2006-07-21 16:46:56 +02:00
|
|
|
|
#include "cli/cli-script.h" /* for reset_command_nest_depth */
|
2008-05-02 15:58:38 +02:00
|
|
|
|
#include "main.h"
|
2008-07-10 00:16:15 +02:00
|
|
|
|
#include "gdbthread.h"
|
2011-07-21 13:03:48 +02:00
|
|
|
|
#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.
2011-05-27 20:28:18 +02:00
|
|
|
|
#include "continuations.h"
|
2010-12-29 01:58:14 +01:00
|
|
|
|
#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.
2013-01-22 21:17:10 +01:00
|
|
|
|
#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.
2013-03-21 18:37:30 +01:00
|
|
|
|
#include "maint.h"
|
2016-03-09 19:25:00 +01:00
|
|
|
|
#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 17:49:30 +02:00
|
|
|
|
#include "ser-event.h"
|
|
|
|
|
#include "gdb_select.h"
|
1999-08-31 03:14:27 +02:00
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* 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.
2004-02-28 19:04:37 +01:00
|
|
|
|
#include "readline/readline.h"
|
|
|
|
|
#include "readline/history.h"
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
/* readline defines this. */
|
|
|
|
|
#undef savestring
|
|
|
|
|
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void rl_callback_read_char_wrapper (gdb_client_data client_data);
|
|
|
|
|
static void command_line_handler (char *rl);
|
|
|
|
|
static void change_line_handler (void);
|
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 16:49:00 +02:00
|
|
|
|
static char *top_level_prompt (void);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Signal handlers. */
|
2005-03-16 18:05:31 +01:00
|
|
|
|
#ifdef SIGQUIT
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void handle_sigquit (int sig);
|
2005-03-16 18:05:31 +01:00
|
|
|
|
#endif
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#ifdef SIGHUP
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void handle_sighup (int sig);
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#endif
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void handle_sigfpe (int sig);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
/* Functions to be invoked by the event loop in response to
|
2010-12-29 01:58:14 +01:00
|
|
|
|
signals. */
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#if defined (SIGQUIT) || defined (SIGHUP)
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void async_do_nothing (gdb_client_data);
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#endif
|
|
|
|
|
#ifdef SIGHUP
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void async_disconnect (gdb_client_data);
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#endif
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void async_float_handler (gdb_client_data);
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#ifdef STOP_SIGNAL
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void async_stop_sig (gdb_client_data);
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#endif
|
2014-03-18 22:48:06 +01:00
|
|
|
|
static void async_sigterm_handler (gdb_client_data arg);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
/* Readline offers an alternate interface, via callback
|
2010-12-29 01:58:14 +01:00
|
|
|
|
functions. These are all included in the file callback.c in the
|
1999-05-11 22:29:07 +02:00
|
|
|
|
readline distribution. This file provides (mainly) a function, which
|
|
|
|
|
the event loop uses as callback (i.e. event handler) whenever an event
|
|
|
|
|
is detected on the standard input file descriptor.
|
|
|
|
|
readline_callback_read_char is called (by the GDB event loop) whenever
|
2010-12-29 01:58:14 +01:00
|
|
|
|
there is a new character ready on the input stream. This function
|
1999-05-11 22:29:07 +02:00
|
|
|
|
incrementally builds a buffer internal to readline where it
|
|
|
|
|
accumulates the line read up to the point of invocation. In the
|
|
|
|
|
special case in which the character read is newline, the function
|
|
|
|
|
invokes a GDB supplied callback routine, which does the processing of
|
|
|
|
|
a full command line. This latter routine is the asynchronous analog
|
2010-12-29 01:58:14 +01:00
|
|
|
|
of the old command_line_input in gdb. Instead of invoking (and waiting
|
1999-05-11 22:29:07 +02:00
|
|
|
|
for) readline to read the command line and pass it back to
|
|
|
|
|
command_loop for processing, the new command_line_handler function has
|
|
|
|
|
the command line already available as its parameter. INPUT_HANDLER is
|
|
|
|
|
to be set to the function that readline will invoke when a complete
|
|
|
|
|
line of input is ready. CALL_READLINE is to be set to the function
|
2010-12-29 01:58:14 +01:00
|
|
|
|
that readline offers as callback to the event_loop. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
1999-09-22 05:28:34 +02:00
|
|
|
|
void (*input_handler) (char *);
|
|
|
|
|
void (*call_readline) (gdb_client_data);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Important variables for the event loop. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
/* This is used to determine if GDB is using the readline library or
|
2010-12-29 01:58:14 +01:00
|
|
|
|
its own simplified form of readline. It is used by the asynchronous
|
1999-06-14 20:08:47 +02:00
|
|
|
|
form of the set editing command.
|
1999-05-25 20:09:09 +02:00
|
|
|
|
ezannoni: as of 1999-04-29 I expect that this
|
1999-05-11 22:29:07 +02:00
|
|
|
|
variable will not be used after gdb is changed to use the event
|
2010-12-29 01:58:14 +01:00
|
|
|
|
loop as default engine, and event-top.c is merged into top.c. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
int async_command_editing_p;
|
|
|
|
|
|
1999-08-31 03:14:27 +02:00
|
|
|
|
/* This is used to display the notification of the completion of an
|
2010-12-29 01:58:14 +01:00
|
|
|
|
asynchronous execution command. */
|
1999-08-31 03:14:27 +02:00
|
|
|
|
int exec_done_display_p = 0;
|
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
/* This is the file descriptor for the input stream that GDB uses to
|
2010-12-29 01:58:14 +01:00
|
|
|
|
read commands from. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
int input_fd;
|
|
|
|
|
|
2014-11-23 11:03:39 +01:00
|
|
|
|
/* 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 01:58:14 +01:00
|
|
|
|
/* Signal handling variables. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
/* Each of these is a pointer to a function that the event loop will
|
2010-12-29 01:58:14 +01:00
|
|
|
|
invoke if the corresponding signal has received. The real signal
|
1999-05-11 22:29:07 +02:00
|
|
|
|
handlers mark these functions as ready to be executed and the event
|
2010-12-29 01:58:14 +01:00
|
|
|
|
loop, in a later iteration, calls them. See the function
|
|
|
|
|
invoke_async_signal_handler. */
|
2012-10-23 08:12:04 +02:00
|
|
|
|
static struct async_signal_handler *sigint_token;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
#ifdef SIGHUP
|
2012-10-23 08:12:04 +02:00
|
|
|
|
static struct async_signal_handler *sighup_token;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
#endif
|
2005-03-16 18:05:31 +01:00
|
|
|
|
#ifdef SIGQUIT
|
2012-10-23 08:12:04 +02:00
|
|
|
|
static struct async_signal_handler *sigquit_token;
|
2005-03-16 18:05:31 +01:00
|
|
|
|
#endif
|
2012-10-23 08:12:04 +02:00
|
|
|
|
static struct async_signal_handler *sigfpe_token;
|
1999-06-14 20:08:47 +02:00
|
|
|
|
#ifdef STOP_SIGNAL
|
2012-10-23 08:12:04 +02:00
|
|
|
|
static struct async_signal_handler *sigtstp_token;
|
1999-06-14 20:08:47 +02:00
|
|
|
|
#endif
|
2014-03-18 22:48:06 +01:00
|
|
|
|
static struct async_signal_handler *async_sigterm_token;
|
1999-06-14 20:08:47 +02:00
|
|
|
|
|
2001-11-27 05:15:09 +01:00
|
|
|
|
/* This hook is called by rl_callback_read_char_wrapper after each
|
|
|
|
|
character is processed. */
|
2011-03-16 16:18:58 +01:00
|
|
|
|
void (*after_char_processing_hook) (void);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Wrapper function for calling into the readline library. The event
|
|
|
|
|
loop expects the callback function to have a paramter, while
|
|
|
|
|
readline expects none. */
|
1999-09-22 05:28:34 +02:00
|
|
|
|
static void
|
|
|
|
|
rl_callback_read_char_wrapper (gdb_client_data client_data)
|
|
|
|
|
{
|
|
|
|
|
rl_callback_read_char ();
|
2001-11-27 05:15:09 +01:00
|
|
|
|
if (after_char_processing_hook)
|
|
|
|
|
(*after_char_processing_hook) ();
|
1999-09-22 05:28:34 +02:00
|
|
|
|
}
|
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
/* Initialize all the necessary variables, start the event loop,
|
2013-09-06 10:53:09 +02:00
|
|
|
|
register readline, and stdin, start the loop. The DATA is the
|
|
|
|
|
interpreter data cookie, ignored for now. */
|
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
void
|
2013-09-06 10:53:09 +02:00
|
|
|
|
cli_command_loop (void *data)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2011-09-02 23:16:54 +02:00
|
|
|
|
display_gdb_prompt (0);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Now it's time to start the event loop. */
|
1999-06-28 18:06:02 +02:00
|
|
|
|
start_event_loop ();
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Change the function to be invoked every time there is a character
|
2010-12-29 01:58:14 +01:00
|
|
|
|
ready on stdin. This is used when the user sets the editing off,
|
1999-05-11 22:29:07 +02:00
|
|
|
|
therefore bypassing readline, and letting gdb handle the input
|
2016-03-09 19:24:59 +01:00
|
|
|
|
itself, via gdb_readline_no_editing_callback. Also it is used in
|
|
|
|
|
the opposite case in which the user sets editing on again, by
|
|
|
|
|
restoring readline handling of the input. */
|
1999-05-25 20:09:09 +02:00
|
|
|
|
static void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
change_line_handler (void)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* NOTE: this operates on input_fd, not instream. If we are reading
|
|
|
|
|
commands from a file, instream will point to the file. However in
|
1999-09-22 05:28:34 +02:00
|
|
|
|
async mode, we always read commands from a file with editing
|
2010-12-29 01:58:14 +01:00
|
|
|
|
off. This means that the 'set editing on/off' will have effect
|
|
|
|
|
only on the interactive session. */
|
1999-09-22 05:28:34 +02:00
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
if (async_command_editing_p)
|
|
|
|
|
{
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Turn on editing by using readline. */
|
1999-09-22 05:28:34 +02:00
|
|
|
|
call_readline = rl_callback_read_char_wrapper;
|
1999-06-14 20:08:47 +02:00
|
|
|
|
input_handler = command_line_handler;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
2016-03-09 19:24:59 +01:00
|
|
|
|
/* Turn off editing by using gdb_readline_no_editing_callback. */
|
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.
2014-10-23 18:13:35 +02:00
|
|
|
|
gdb_rl_callback_handler_remove ();
|
2016-03-09 19:24:59 +01:00
|
|
|
|
call_readline = gdb_readline_no_editing_callback;
|
1999-06-14 20:08:47 +02:00
|
|
|
|
|
|
|
|
|
/* Set up the command handler as well, in case we are called as
|
2010-12-29 01:58:14 +01:00
|
|
|
|
first thing from .gdbinit. */
|
1999-06-14 20:08:47 +02:00
|
|
|
|
input_handler = command_line_handler;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
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.
2014-10-23 18:13:35 +02:00
|
|
|
|
/* 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)
|
|
|
|
|
{
|
|
|
|
|
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)
|
|
|
|
|
{
|
|
|
|
|
/* 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);
|
|
|
|
|
|
|
|
|
|
rl_callback_handler_install (prompt, input_handler);
|
|
|
|
|
callback_handler_installed = 1;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See event-top.h, and above. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
gdb_rl_callback_handler_reinstall (void)
|
|
|
|
|
{
|
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
/* 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 01:58:14 +01:00
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
indicating that the command will continue on the next line. In
|
|
|
|
|
that case the prompt that is displayed is the empty string.
|
|
|
|
|
|
1999-06-14 20:08:47 +02:00
|
|
|
|
2. When the user is entering 'commands' for a breakpoint, or
|
2010-12-29 01:58:14 +01:00
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
|
|
|
|
|
3. On prompting for pagination. */
|
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
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>
2014-12-15 17:38:03 +01:00
|
|
|
|
display_gdb_prompt (const char *new_prompt)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2011-07-21 13:03:48 +02:00
|
|
|
|
char *actual_gdb_prompt = 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 16:49:00 +02:00
|
|
|
|
struct cleanup *old_chain;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
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.
2013-01-22 21:17:10 +01:00
|
|
|
|
annotate_display_prompt ();
|
|
|
|
|
|
2006-07-21 16:46:56 +02:00
|
|
|
|
/* 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 16:49:00 +02:00
|
|
|
|
old_chain = make_cleanup (free_current_contents, &actual_gdb_prompt);
|
2011-07-21 13:03:48 +02:00
|
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
/* 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 01:30:11 +02:00
|
|
|
|
{
|
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 16:49:00 +02:00
|
|
|
|
if (sync_execution)
|
2011-07-21 13:03:48 +02:00
|
|
|
|
{
|
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 16:49:00 +02:00
|
|
|
|
/* 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. */
|
|
|
|
|
|
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.
2014-10-23 18:13:35 +02:00
|
|
|
|
gdb_rl_callback_handler_remove ();
|
2013-05-05 18:54:26 +02:00
|
|
|
|
do_cleanups (old_chain);
|
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 16:49:00 +02:00
|
|
|
|
return;
|
2011-07-21 13:03:48 +02:00
|
|
|
|
}
|
|
|
|
|
else
|
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 16:49:00 +02:00
|
|
|
|
{
|
|
|
|
|
/* Display the top level prompt. */
|
|
|
|
|
actual_gdb_prompt = top_level_prompt ();
|
|
|
|
|
}
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
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 16:49:00 +02:00
|
|
|
|
else
|
|
|
|
|
actual_gdb_prompt = xstrdup (new_prompt);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
if (async_command_editing_p)
|
|
|
|
|
{
|
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.
2014-10-23 18:13:35 +02:00
|
|
|
|
gdb_rl_callback_handler_remove ();
|
|
|
|
|
gdb_rl_callback_handler_install (actual_gdb_prompt);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* new_prompt at this point can be the top of the stack or the one
|
2011-03-01 22:12:27 +01:00
|
|
|
|
passed in. It can't be NULL. */
|
|
|
|
|
else
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
|
|
|
|
/* 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. */
|
2011-07-21 13:03:48 +02:00
|
|
|
|
fputs_unfiltered (actual_gdb_prompt, gdb_stdout);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
|
|
|
|
|
do_cleanups (old_chain);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
/* Return the top level prompt, as specified by "set prompt", possibly
|
|
|
|
|
overriden by the python gdb.prompt_hook hook, and then composed
|
|
|
|
|
with the prompt prefix and suffix (annotations). The caller is
|
|
|
|
|
responsible for freeing the returned string. */
|
|
|
|
|
|
|
|
|
|
static char *
|
|
|
|
|
top_level_prompt (void)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2016-03-09 19:24:59 +01:00
|
|
|
|
char *prompt;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
/* 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 ());
|
|
|
|
|
|
2016-03-09 19:24:59 +01:00
|
|
|
|
prompt = get_prompt ();
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
if (annotation_level >= 2)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
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 16:49:00 +02:00
|
|
|
|
/* Prefix needs to have new line at end. */
|
2016-03-09 19:24:59 +01:00
|
|
|
|
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 16:49:00 +02:00
|
|
|
|
|
|
|
|
|
/* Suffix needs to have a new line at end and \032 \032 at
|
|
|
|
|
beginning. */
|
2016-03-09 19:24:59 +01:00
|
|
|
|
const char suffix[] = "\n\032\032prompt\n";
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:24:59 +01:00
|
|
|
|
return concat (prefix, prompt, suffix, 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 16:49:00 +02:00
|
|
|
|
|
2016-03-09 19:24:59 +01:00
|
|
|
|
return xstrdup (prompt);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
1999-09-22 05:28:34 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* Get a pointer to the command line buffer. This is used to
|
|
|
|
|
construct a whole line of input from partial input. */
|
|
|
|
|
|
|
|
|
|
static struct buffer *
|
|
|
|
|
get_command_line_buffer (void)
|
|
|
|
|
{
|
|
|
|
|
static struct buffer line_buffer;
|
|
|
|
|
static int line_buffer_initialized;
|
|
|
|
|
|
|
|
|
|
if (!line_buffer_initialized)
|
|
|
|
|
{
|
|
|
|
|
buffer_init (&line_buffer);
|
|
|
|
|
line_buffer_initialized = 1;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return &line_buffer;
|
|
|
|
|
}
|
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* When there is an event ready on the stdin file descriptor, instead
|
1999-09-22 05:28:34 +02:00
|
|
|
|
of calling readline directly throught the callback function, or
|
2016-03-09 19:24:59 +01:00
|
|
|
|
instead of calling gdb_readline_no_editing_callback, give gdb a
|
|
|
|
|
chance to detect errors and do something. */
|
|
|
|
|
|
1999-09-22 05:28:34 +02:00
|
|
|
|
void
|
1999-10-06 01:13:56 +02:00
|
|
|
|
stdin_event_handler (int error, gdb_client_data client_data)
|
1999-09-22 05:28:34 +02:00
|
|
|
|
{
|
|
|
|
|
if (error)
|
|
|
|
|
{
|
2005-02-11 Andrew Cagney <cagney@gnu.org>
Mark up some of printf_filtered and printf_unfiltered.
* ada-lang.c, annotate.c, arch-utils.c, breakpoint.c: Update.
* corelow.c, cp-namespace.c, cp-support.c, dcache.c: Update.
* demangle.c, dsrec.c, dwarf2read.c, dwarfread.c: Update.
* event-loop.c, event-top.c, exec.c, f-valprint.c: Update.
* gdbtypes.c, inf-loop.c, inf-ptrace.c, inf-ttrace.c: Update.
* infcmd.c, inflow.c, infrun.c, inftarg.c, language.c: Update.
* linespec.c, linux-nat.c, linux-thread-db.c, maint.c: Update.
* mdebugread.c, memattr.c, monitor.c, objc-lang.c: Update.
* ocd.c, osabi.c, printcmd.c, procfs.c, regcache.c: Update.
* remote.c, solib-som.c, solib.c, somsolib.c, source.c: Update.
* stack.c, symfile.c, symmisc.c, target.c, thread.c: Update.
* top.c, utils.c, valprint.c, value.c, cli/cli-cmds.c: Update.
* cli/cli-dump.c, cli/cli-logging.c, tui/tui-hooks.c: Update.
* tui/tui-regs.c, tui/tui-win.c: Update.
2005-02-12 01:39:24 +01:00
|
|
|
|
printf_unfiltered (_("error detected on stdin\n"));
|
1999-10-06 01:13:56 +02:00
|
|
|
|
delete_file_handler (input_fd);
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* If stdin died, we may as well kill gdb. */
|
2000-01-18 01:55:13 +01:00
|
|
|
|
quit_command ((char *) 0, stdin == instream);
|
1999-09-22 05:28:34 +02:00
|
|
|
|
}
|
|
|
|
|
else
|
2014-11-23 11:03:39 +01:00
|
|
|
|
{
|
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 17:49:30 +02:00
|
|
|
|
/* 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. */
|
|
|
|
|
QUIT;
|
|
|
|
|
|
2014-11-23 11:03:39 +01:00
|
|
|
|
do
|
|
|
|
|
{
|
|
|
|
|
call_stdin_event_handler_again_p = 0;
|
|
|
|
|
(*call_readline) (client_data);
|
|
|
|
|
} while (call_stdin_event_handler_again_p != 0);
|
|
|
|
|
}
|
1999-09-22 05:28:34 +02:00
|
|
|
|
}
|
|
|
|
|
|
1999-09-28 23:55:21 +02:00
|
|
|
|
/* 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 01:58:14 +01:00
|
|
|
|
the exec operation. */
|
1999-09-28 23:55:21 +02:00
|
|
|
|
|
|
|
|
|
void
|
2008-03-14 20:55:51 +01:00
|
|
|
|
async_enable_stdin (void)
|
1999-09-28 23:55:21 +02:00
|
|
|
|
{
|
2008-03-14 19:57:44 +01:00
|
|
|
|
if (sync_execution)
|
|
|
|
|
{
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* See NOTE in async_disable_stdin(). */
|
2008-03-14 19:57:44 +01:00
|
|
|
|
/* FIXME: cagney/1999-09-27: Call this before clearing
|
|
|
|
|
sync_execution. Current target_terminal_ours() implementations
|
2010-12-29 01:58:14 +01:00
|
|
|
|
check for sync_execution before switching the terminal. */
|
2008-03-14 19:57:44 +01:00
|
|
|
|
target_terminal_ours ();
|
|
|
|
|
sync_execution = 0;
|
|
|
|
|
}
|
1999-09-28 23:55:21 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Disable reads from stdin (the console) marking the command as
|
2010-12-29 01:58:14 +01:00
|
|
|
|
synchronous. */
|
1999-09-28 23:55:21 +02:00
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
async_disable_stdin (void)
|
|
|
|
|
{
|
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 16:49:00 +02:00
|
|
|
|
sync_execution = 1;
|
1999-09-28 23:55:21 +02:00
|
|
|
|
}
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
1999-09-28 23:55:21 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* 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 05:28:34 +02:00
|
|
|
|
command_handler (char *command)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2010-06-26 08:44:47 +02:00
|
|
|
|
struct cleanup *stat_chain;
|
2016-03-09 19:25:00 +01:00
|
|
|
|
char *c;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2015-12-19 02:55:26 +01:00
|
|
|
|
if (instream == stdin)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
reinitialize_more_filter ();
|
|
|
|
|
|
2010-06-26 08:44:47 +02:00
|
|
|
|
stat_chain = make_command_stats_cleanup (1);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* Do not execute commented lines. */
|
|
|
|
|
for (c = command; *c == ' ' || *c == '\t'; c++)
|
|
|
|
|
;
|
|
|
|
|
if (c[0] != '#')
|
|
|
|
|
{
|
|
|
|
|
execute_command (command, instream == stdin);
|
1999-07-07 22:19:36 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* Do any commands attached to breakpoint we stopped at. */
|
|
|
|
|
bpstat_do_actions ();
|
|
|
|
|
}
|
1999-07-07 22:19:36 +02:00
|
|
|
|
|
2010-06-26 08:44:47 +02:00
|
|
|
|
do_cleanups (stat_chain);
|
1999-07-05 19:58:44 +02:00
|
|
|
|
}
|
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* 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. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
static char *
|
|
|
|
|
command_line_append_input_line (struct buffer *cmd_line_buffer, char *rl)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
char *cmd;
|
|
|
|
|
size_t len;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
len = strlen (rl);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
if (len > 0 && rl[len - 1] == '\\')
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* Don't copy the backslash and wait for more. */
|
|
|
|
|
buffer_grow (cmd_line_buffer, rl, len - 1);
|
|
|
|
|
cmd = NULL;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
2016-03-09 19:25:00 +01:00
|
|
|
|
else
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* Copy whole line including terminating null, and we're
|
|
|
|
|
done. */
|
|
|
|
|
buffer_grow (cmd_line_buffer, rl, len + 1);
|
|
|
|
|
cmd = cmd_line_buffer->buffer;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* Allocated in readline. */
|
|
|
|
|
xfree (rl);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
return cmd;
|
|
|
|
|
}
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* Handle a line of input coming from readline.
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
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.
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
Returns EOF on end of file.
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
If REPEAT, handle command repetitions:
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
- 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.
|
2002-03-27 22:20:15 +01:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
- OTOH, if the input command line IS empty, return the previously
|
|
|
|
|
saved command instead of the empty input line.
|
|
|
|
|
*/
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
char *
|
|
|
|
|
handle_line_of_input (struct buffer *cmd_line_buffer,
|
|
|
|
|
char *rl, int repeat, char *annotation_suffix)
|
|
|
|
|
{
|
|
|
|
|
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;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* 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;
|
|
|
|
|
|
|
|
|
|
if (annotation_level > 1 && instream == stdin)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
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);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Do history expansion if that is wished. */
|
|
|
|
|
if (history_expansion_p && instream == stdin
|
|
|
|
|
&& ISATTY (instream))
|
|
|
|
|
{
|
|
|
|
|
char *history_value;
|
|
|
|
|
int expanded;
|
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
expanded = history_expand (cmd, &history_value);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
if (expanded)
|
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
size_t len;
|
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
/* Print the changes. */
|
|
|
|
|
printf_unfiltered ("%s\n", history_value);
|
|
|
|
|
|
|
|
|
|
/* If there was an error, call this function again. */
|
|
|
|
|
if (expanded < 0)
|
|
|
|
|
{
|
2000-12-15 02:01:51 +01:00
|
|
|
|
xfree (history_value);
|
2016-03-09 19:25:00 +01:00
|
|
|
|
return cmd;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
2016-03-09 19:25:00 +01:00
|
|
|
|
|
|
|
|
|
/* 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;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* If we just got an empty line, and that is supposed to repeat the
|
2016-03-09 19:25:00 +01:00
|
|
|
|
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. */
|
|
|
|
|
if (*cmd != '\0' && input_from_terminal_p ())
|
|
|
|
|
gdb_add_history (cmd);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* Save into global buffer if appropriate. */
|
|
|
|
|
if (repeat)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
xfree (saved_command_line);
|
|
|
|
|
saved_command_line = xstrdup (cmd);
|
|
|
|
|
return saved_command_line;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
2016-03-09 19:25:00 +01:00
|
|
|
|
else
|
|
|
|
|
return cmd;
|
|
|
|
|
}
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* 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.
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
NOTE: This is the asynchronous version of the command_line_input
|
|
|
|
|
function. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
void
|
|
|
|
|
command_line_handler (char *rl)
|
|
|
|
|
{
|
|
|
|
|
struct buffer *line_buffer = get_command_line_buffer ();
|
|
|
|
|
char *cmd;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
cmd = handle_line_of_input (line_buffer, rl, instream == stdin, "prompt");
|
|
|
|
|
if (cmd == (char *) EOF)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
/* 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");
|
|
|
|
|
execute_command ("quit", stdin == instream);
|
|
|
|
|
}
|
|
|
|
|
else if (cmd == NULL)
|
|
|
|
|
{
|
|
|
|
|
/* We don't have a full line yet. Print an empty prompt. */
|
|
|
|
|
display_gdb_prompt ("");
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
command_handler (cmd);
|
|
|
|
|
display_gdb_prompt (0);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Does reading of input from terminal w/o the editing features
|
2016-03-09 19:24:59 +01:00
|
|
|
|
provided by the readline library. Calls the line input handler
|
|
|
|
|
once we have a whole input line. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
1999-06-28 18:06:02 +02:00
|
|
|
|
void
|
2016-03-09 19:24:59 +01:00
|
|
|
|
gdb_readline_no_editing_callback (gdb_client_data client_data)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
|
|
|
|
int c;
|
|
|
|
|
char *result;
|
2016-03-09 19:25:00 +01:00
|
|
|
|
struct buffer line_buffer;
|
1999-08-16 21:57:19 +02:00
|
|
|
|
static int done_once = 0;
|
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
buffer_init (&line_buffer);
|
|
|
|
|
|
1999-08-16 21:57:19 +02:00
|
|
|
|
/* Unbuffer the input stream, so that, later on, the calls to fgetc
|
2010-12-29 01:58:14 +01:00
|
|
|
|
fetch only one char at the time from the stream. The fgetc's will
|
1999-08-16 21:57:19 +02:00
|
|
|
|
get up to the first newline, but there may be more chars in the
|
2010-12-29 01:58:14 +01:00
|
|
|
|
stream after '\n'. If we buffer the input and fgetc drains the
|
1999-08-16 21:57:19 +02:00
|
|
|
|
stream, getting stuff beyond the newline as well, a select, done
|
2010-12-29 01:58:14 +01:00
|
|
|
|
afterwards will not trigger. */
|
1999-08-16 21:57:19 +02:00
|
|
|
|
if (!done_once && !ISATTY (instream))
|
|
|
|
|
{
|
|
|
|
|
setbuf (instream, NULL);
|
|
|
|
|
done_once = 1;
|
|
|
|
|
}
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
/* We still need the while loop here, even though it would seem
|
2016-03-09 19:24:59 +01:00
|
|
|
|
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. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
while (1)
|
|
|
|
|
{
|
|
|
|
|
/* Read from stdin if we are executing a user defined command.
|
|
|
|
|
This is the right thing for prompt_for_continue, at least. */
|
|
|
|
|
c = fgetc (instream ? instream : stdin);
|
|
|
|
|
|
|
|
|
|
if (c == EOF)
|
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
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));
|
1999-06-14 20:08:47 +02:00
|
|
|
|
(*input_handler) (0);
|
2007-08-30 00:04:15 +02:00
|
|
|
|
return;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (c == '\n')
|
|
|
|
|
{
|
2016-03-09 19:25:00 +01:00
|
|
|
|
if (line_buffer.used_size > 0
|
|
|
|
|
&& line_buffer.buffer[line_buffer.used_size - 1] == '\r')
|
|
|
|
|
line_buffer.used_size--;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
buffer_grow_char (&line_buffer, c);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
2016-03-09 19:25:00 +01:00
|
|
|
|
buffer_grow_char (&line_buffer, '\0');
|
|
|
|
|
result = buffer_finish (&line_buffer);
|
1999-06-14 20:08:47 +02:00
|
|
|
|
(*input_handler) (result);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
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 17:49:30 +02:00
|
|
|
|
/* 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;
|
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
/* Initialization of signal handlers and tokens. There is a function
|
2010-12-29 01:58:14 +01:00
|
|
|
|
handle_sig* for each of the signals GDB cares about. Specifically:
|
1999-05-11 22:29:07 +02:00
|
|
|
|
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 01:58:14 +01:00
|
|
|
|
procedures are the old signal handlers. The event loop will take
|
1999-05-11 22:29:07 +02:00
|
|
|
|
care of invoking the queued procedures to perform the usual tasks
|
2010-12-29 01:58:14 +01:00
|
|
|
|
associated with the reception of the signal. */
|
1999-05-25 20:09:09 +02:00
|
|
|
|
/* NOTE: 1999-04-30 This is the asynchronous version of init_signals.
|
1999-05-11 22:29:07 +02:00
|
|
|
|
init_signals will become obsolete as we move to have to event loop
|
2010-12-29 01:58:14 +01:00
|
|
|
|
as the default for gdb. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
async_init_signals (void)
|
1999-07-07 22:19:36 +02:00
|
|
|
|
{
|
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 17:49:30 +02:00
|
|
|
|
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 17:49:30 +02:00
|
|
|
|
quit_serial_event = make_serial_event ();
|
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
signal (SIGINT, handle_sigint);
|
|
|
|
|
sigint_token =
|
1999-06-14 20:08:47 +02:00
|
|
|
|
create_async_signal_handler (async_request_quit, NULL);
|
2005-11-25 19:33:12 +01:00
|
|
|
|
signal (SIGTERM, handle_sigterm);
|
2014-03-18 22:48:06 +01:00
|
|
|
|
async_sigterm_token
|
|
|
|
|
= create_async_signal_handler (async_sigterm_handler, NULL);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
/* 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
|
|
|
|
|
|
2005-03-16 18:05:31 +01:00
|
|
|
|
#ifdef SIGQUIT
|
1999-05-11 22:29:07 +02:00
|
|
|
|
/* 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 =
|
1999-06-14 20:08:47 +02:00
|
|
|
|
create_async_signal_handler (async_do_nothing, NULL);
|
2005-03-16 18:05:31 +01:00
|
|
|
|
#endif
|
1999-05-11 22:29:07 +02:00
|
|
|
|
#ifdef SIGHUP
|
|
|
|
|
if (signal (SIGHUP, handle_sighup) != SIG_IGN)
|
|
|
|
|
sighup_token =
|
1999-06-14 20:08:47 +02:00
|
|
|
|
create_async_signal_handler (async_disconnect, NULL);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
else
|
|
|
|
|
sighup_token =
|
1999-06-14 20:08:47 +02:00
|
|
|
|
create_async_signal_handler (async_do_nothing, NULL);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
#endif
|
|
|
|
|
signal (SIGFPE, handle_sigfpe);
|
|
|
|
|
sigfpe_token =
|
1999-06-14 20:08:47 +02:00
|
|
|
|
create_async_signal_handler (async_float_handler, NULL);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
1999-06-14 20:08:47 +02:00
|
|
|
|
#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 17:49:30 +02:00
|
|
|
|
/* 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 17:49:32 +02:00
|
|
|
|
/* See defs.h. */
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
|
default_quit_handler (void)
|
|
|
|
|
{
|
|
|
|
|
if (check_quit_flag ())
|
|
|
|
|
{
|
|
|
|
|
if (target_terminal_is_ours ())
|
|
|
|
|
quit ();
|
|
|
|
|
else
|
|
|
|
|
target_pass_ctrlc ();
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See defs.h. */
|
|
|
|
|
quit_handler_ftype *quit_handler = default_quit_handler;
|
|
|
|
|
|
|
|
|
|
/* Data for make_cleanup_override_quit_handler. Wrap the previous
|
|
|
|
|
handler pointer in a data struct because it's not portable to cast
|
|
|
|
|
a function pointer to a data pointer, which is what make_cleanup
|
|
|
|
|
expects. */
|
|
|
|
|
struct quit_handler_cleanup_data
|
|
|
|
|
{
|
|
|
|
|
/* The previous quit handler. */
|
|
|
|
|
quit_handler_ftype *prev_handler;
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
/* Cleanup call that restores the previous quit handler. */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
restore_quit_handler (void *arg)
|
|
|
|
|
{
|
|
|
|
|
struct quit_handler_cleanup_data *data
|
|
|
|
|
= (struct quit_handler_cleanup_data *) arg;
|
|
|
|
|
|
|
|
|
|
quit_handler = data->prev_handler;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Destructor for the quit handler cleanup. */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
restore_quit_handler_dtor (void *arg)
|
|
|
|
|
{
|
|
|
|
|
xfree (arg);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See defs.h. */
|
|
|
|
|
|
|
|
|
|
struct cleanup *
|
|
|
|
|
make_cleanup_override_quit_handler (quit_handler_ftype *new_quit_handler)
|
|
|
|
|
{
|
|
|
|
|
struct cleanup *old_chain;
|
|
|
|
|
struct quit_handler_cleanup_data *data;
|
|
|
|
|
|
|
|
|
|
data = XNEW (struct quit_handler_cleanup_data);
|
|
|
|
|
data->prev_handler = quit_handler;
|
|
|
|
|
old_chain = make_cleanup_dtor (restore_quit_handler, data,
|
|
|
|
|
restore_quit_handler_dtor);
|
|
|
|
|
quit_handler = new_quit_handler;
|
|
|
|
|
return old_chain;
|
|
|
|
|
}
|
|
|
|
|
|
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 17:49:30 +02:00
|
|
|
|
/* Handle a SIGINT. */
|
|
|
|
|
|
1999-07-07 22:19:36 +02:00
|
|
|
|
void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
handle_sigint (int sig)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
|
|
|
|
signal (sig, handle_sigint);
|
|
|
|
|
|
2007-02-08 17:25:25 +01:00
|
|
|
|
/* 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 01:58:14 +01:00
|
|
|
|
set quit_flag to 1 here. Then if QUIT is called before we get to
|
2007-02-08 17:25:25 +01:00
|
|
|
|
the event loop, we will unwind as expected. */
|
|
|
|
|
|
2012-08-22 19:48:55 +02:00
|
|
|
|
set_quit_flag ();
|
2007-02-08 17:25:25 +01:00
|
|
|
|
|
1999-05-11 22:29:07 +02:00
|
|
|
|
/* If immediate_quit is set, we go ahead and process the SIGINT right
|
2010-12-29 01:58:14 +01:00
|
|
|
|
away, even if we usually would defer this to the event loop. The
|
1999-05-11 22:29:07 +02:00
|
|
|
|
assumption here is that it is safe to process ^C immediately if
|
2010-12-29 01:58:14 +01:00
|
|
|
|
immediate_quit is set. If we didn't, SIGINT would be really
|
1999-05-11 22:29:07 +02:00
|
|
|
|
processed only the next time through the event loop. To get to
|
|
|
|
|
that point, though, the command that we want to interrupt needs to
|
2008-03-05 18:21:10 +01:00
|
|
|
|
finish first, which is unacceptable. If immediate quit is not set,
|
2010-12-29 01:58:14 +01:00
|
|
|
|
we process SIGINT the next time through the loop, which is fine. */
|
2008-03-05 18:21:10 +01:00
|
|
|
|
gdb_call_async_signal_handler (sigint_token, immediate_quit);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
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 17:49:30 +02:00
|
|
|
|
/* 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;
|
|
|
|
|
}
|
|
|
|
|
|
2014-03-18 22:48:06 +01:00
|
|
|
|
/* Handle GDB exit upon receiving SIGTERM if target_can_async_p (). */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
async_sigterm_handler (gdb_client_data arg)
|
|
|
|
|
{
|
|
|
|
|
quit_force (NULL, stdin == instream);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See defs.h. */
|
|
|
|
|
volatile int sync_quit_force_run;
|
|
|
|
|
|
2005-11-25 19:33:12 +01:00
|
|
|
|
/* 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);
|
2014-03-18 22:48:06 +01:00
|
|
|
|
|
2015-07-24 19:58:47 +02:00
|
|
|
|
sync_quit_force_run = 1;
|
|
|
|
|
set_quit_flag ();
|
|
|
|
|
|
|
|
|
|
mark_async_signal_handler (async_sigterm_token);
|
2005-11-25 19:33:12 +01:00
|
|
|
|
}
|
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Do the quit. All the checks have been done by the caller. */
|
1999-07-07 22:19:36 +02:00
|
|
|
|
void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
async_request_quit (gdb_client_data arg)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2007-02-08 17:25:25 +01:00
|
|
|
|
/* If the quit_flag has gotten reset back to 0 by the time we get
|
2007-02-10 00:45:35 +01:00
|
|
|
|
back here, that means that an exception was thrown to unwind the
|
|
|
|
|
current command before we got back to the event loop. So there
|
2012-08-22 19:48:55 +02:00
|
|
|
|
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 17:49:32 +02:00
|
|
|
|
QUIT;
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
2005-03-16 18:05:31 +01:00
|
|
|
|
#ifdef SIGQUIT
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Tell the event loop what to do if SIGQUIT is received.
|
|
|
|
|
See event-signal.c. */
|
1999-07-07 22:19:36 +02:00
|
|
|
|
static void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
handle_sigquit (int sig)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2012-10-23 08:20:49 +02:00
|
|
|
|
mark_async_signal_handler (sigquit_token);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
signal (sig, handle_sigquit);
|
|
|
|
|
}
|
2005-03-16 18:05:31 +01:00
|
|
|
|
#endif
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#if defined (SIGQUIT) || defined (SIGHUP)
|
|
|
|
|
/* Called by the event loop in response to a SIGQUIT or an
|
|
|
|
|
ignored SIGHUP. */
|
1999-07-07 22:19:36 +02:00
|
|
|
|
static void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
async_do_nothing (gdb_client_data arg)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Empty function body. */
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
2006-02-23 19:37:42 +01:00
|
|
|
|
#endif
|
1999-05-11 22:29:07 +02:00
|
|
|
|
|
|
|
|
|
#ifdef SIGHUP
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Tell the event loop what to do if SIGHUP is received.
|
|
|
|
|
See event-signal.c. */
|
1999-07-07 22:19:36 +02:00
|
|
|
|
static void
|
2000-07-30 03:48:28 +02:00
|
|
|
|
handle_sighup (int sig)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2012-10-23 08:20:49 +02:00
|
|
|
|
mark_async_signal_handler (sighup_token);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
signal (sig, handle_sighup);
|
|
|
|
|
}
|
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Called by the event loop to process a SIGHUP. */
|
1999-07-07 22:19:36 +02:00
|
|
|
|
static void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
async_disconnect (gdb_client_data arg)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2011-09-21 17:21:28 +02:00
|
|
|
|
|
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 16:14:14 +01:00
|
|
|
|
TRY
|
2011-09-21 17:21:28 +02:00
|
|
|
|
{
|
|
|
|
|
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 16:14:14 +01:00
|
|
|
|
CATCH (exception, RETURN_MASK_ALL)
|
2011-09-21 17:21:28 +02:00
|
|
|
|
{
|
|
|
|
|
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 16:14:14 +01:00
|
|
|
|
END_CATCH
|
2011-09-21 17:21:28 +02:00
|
|
|
|
|
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 16:14:14 +01:00
|
|
|
|
TRY
|
2011-09-21 17:21:28 +02:00
|
|
|
|
{
|
2013-03-20 16:46:24 +01:00
|
|
|
|
pop_all_targets ();
|
2011-09-21 17:21:28 +02:00
|
|
|
|
}
|
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 16:14:14 +01:00
|
|
|
|
CATCH (exception, RETURN_MASK_ALL)
|
|
|
|
|
{
|
|
|
|
|
}
|
|
|
|
|
END_CATCH
|
2011-09-21 17:21:28 +02:00
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
signal (SIGHUP, SIG_DFL); /*FIXME: ??????????? */
|
2009-01-09 12:00:00 +01:00
|
|
|
|
raise (SIGHUP);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
1999-06-14 20:08:47 +02:00
|
|
|
|
#ifdef STOP_SIGNAL
|
1999-07-07 22:19:36 +02:00
|
|
|
|
void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
handle_stop_sig (int sig)
|
1999-06-14 20:08:47 +02:00
|
|
|
|
{
|
2012-10-23 08:20:49 +02:00
|
|
|
|
mark_async_signal_handler (sigtstp_token);
|
1999-07-07 22:19:36 +02:00
|
|
|
|
signal (sig, handle_stop_sig);
|
1999-06-14 20:08:47 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
async_stop_sig (gdb_client_data arg)
|
1999-06-14 20:08:47 +02:00
|
|
|
|
{
|
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 16:49:00 +02:00
|
|
|
|
char *prompt = get_prompt ();
|
2010-05-14 20:35:11 +02:00
|
|
|
|
|
1999-06-14 20:08:47 +02:00
|
|
|
|
#if STOP_SIGNAL == SIGTSTP
|
|
|
|
|
signal (SIGTSTP, SIG_DFL);
|
1999-10-06 01:13:56 +02:00
|
|
|
|
#if HAVE_SIGPROCMASK
|
|
|
|
|
{
|
|
|
|
|
sigset_t zero;
|
2001-08-28 00:39:56 +02:00
|
|
|
|
|
1999-10-06 01:13:56 +02:00
|
|
|
|
sigemptyset (&zero);
|
|
|
|
|
sigprocmask (SIG_SETMASK, &zero, 0);
|
|
|
|
|
}
|
2001-08-28 00:39:56 +02:00
|
|
|
|
#elif HAVE_SIGSETMASK
|
1999-06-14 20:08:47 +02:00
|
|
|
|
sigsetmask (0);
|
1999-10-06 01:13:56 +02:00
|
|
|
|
#endif
|
2009-01-09 12:00:00 +01:00
|
|
|
|
raise (SIGTSTP);
|
1999-06-14 20:08:47 +02:00
|
|
|
|
signal (SIGTSTP, handle_stop_sig);
|
|
|
|
|
#else
|
|
|
|
|
signal (STOP_SIGNAL, handle_stop_sig);
|
|
|
|
|
#endif
|
|
|
|
|
printf_unfiltered ("%s", prompt);
|
|
|
|
|
gdb_flush (gdb_stdout);
|
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Forget about any previous command -- null line now will do
|
|
|
|
|
nothing. */
|
1999-06-14 20:08:47 +02:00
|
|
|
|
dont_repeat ();
|
|
|
|
|
}
|
|
|
|
|
#endif /* STOP_SIGNAL */
|
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Tell the event loop what to do if SIGFPE is received.
|
|
|
|
|
See event-signal.c. */
|
1999-07-07 22:19:36 +02:00
|
|
|
|
static void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
handle_sigfpe (int sig)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2012-10-23 08:20:49 +02:00
|
|
|
|
mark_async_signal_handler (sigfpe_token);
|
1999-05-11 22:29:07 +02:00
|
|
|
|
signal (sig, handle_sigfpe);
|
|
|
|
|
}
|
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Event loop will call this functin to process a SIGFPE. */
|
1999-07-07 22:19:36 +02:00
|
|
|
|
static void
|
1999-09-22 05:28:34 +02:00
|
|
|
|
async_float_handler (gdb_client_data arg)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* 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 05:06:14 +01:00
|
|
|
|
error (_("Erroneous arithmetic operation."));
|
1999-05-11 22:29:07 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/* Called by do_setshow_command. */
|
|
|
|
|
void
|
2010-12-29 01:58:14 +01:00
|
|
|
|
set_async_editing_command (char *args, int from_tty,
|
|
|
|
|
struct cmd_list_element *c)
|
1999-05-11 22:29:07 +02:00
|
|
|
|
{
|
|
|
|
|
change_line_handler ();
|
|
|
|
|
}
|
|
|
|
|
|
1999-06-14 20:08:47 +02:00
|
|
|
|
/* Set things up for readline to be invoked via the alternate
|
|
|
|
|
interface, i.e. via a callback function (rl_callback_read_char),
|
2010-12-29 01:58:14 +01:00
|
|
|
|
and hook up instream to the event loop. */
|
1999-06-14 20:08:47 +02:00
|
|
|
|
void
|
2002-10-03 03:23:50 +02:00
|
|
|
|
gdb_setup_readline (void)
|
1999-06-14 20:08:47 +02:00
|
|
|
|
{
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
/* 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 12:09:18 +01:00
|
|
|
|
if (!batch_silent)
|
|
|
|
|
gdb_stdout = stdio_fileopen (stdout);
|
2013-08-28 14:25:05 +02:00
|
|
|
|
gdb_stderr = stderr_fileopen ();
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
gdb_stdlog = gdb_stderr; /* for moment */
|
|
|
|
|
gdb_stdtarg = gdb_stderr; /* for moment */
|
2010-09-11 18:00:27 +02:00
|
|
|
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
|
|
|
|
|
/* If the input stream is connected to a terminal, turn on
|
|
|
|
|
editing. */
|
|
|
|
|
if (ISATTY (instream))
|
1999-06-21 15:27:42 +02:00
|
|
|
|
{
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* 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.
2004-09-13 20:26:31 +02:00
|
|
|
|
could be overwritten by a command in .gdbinit like 'set
|
|
|
|
|
editing on' or 'off'. */
|
|
|
|
|
async_command_editing_p = 1;
|
2000-03-19 13:30:38 +01:00
|
|
|
|
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
/* When a character is detected on instream by select or poll,
|
|
|
|
|
readline will be invoked via this callback function. */
|
|
|
|
|
call_readline = rl_callback_read_char_wrapper;
|
1999-06-21 15:27:42 +02:00
|
|
|
|
}
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
async_command_editing_p = 0;
|
2016-03-09 19:24:59 +01:00
|
|
|
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* When readline has read an end-of-line character, it passes the
|
2010-12-29 01:58:14 +01:00
|
|
|
|
complete line to gdb for processing; command_line_handler is the
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
function that does this. */
|
|
|
|
|
input_handler = command_line_handler;
|
|
|
|
|
|
2010-12-29 01:58:14 +01:00
|
|
|
|
/* Tell readline to use the same input stream that gdb uses. */
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
rl_instream = instream;
|
|
|
|
|
|
|
|
|
|
/* Get a file descriptor for the input stream, so that we can
|
|
|
|
|
register it with the event loop. */
|
|
|
|
|
input_fd = fileno (instream);
|
|
|
|
|
|
|
|
|
|
/* Now we need to create the event sources for the input file
|
|
|
|
|
descriptor. */
|
|
|
|
|
/* At this point in time, this is the only event source that we
|
2010-12-29 01:58:14 +01:00
|
|
|
|
register with the even loop. Another source is going to be the
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
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. */
|
|
|
|
|
add_file_handler (input_fd, stdin_event_handler, 0);
|
1999-06-14 20:08:47 +02:00
|
|
|
|
}
|
2002-10-03 03:23:50 +02:00
|
|
|
|
|
2002-11-05 23:38:11 +01:00
|
|
|
|
/* 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. */
|
|
|
|
|
void
|
|
|
|
|
gdb_disable_readline (void)
|
|
|
|
|
{
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
/* 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. */
|
2002-11-05 23:38:11 +01:00
|
|
|
|
|
|
|
|
|
#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.
2004-09-13 20:26:31 +02:00
|
|
|
|
ui_file_delete (gdb_stdout);
|
|
|
|
|
ui_file_delete (gdb_stderr);
|
|
|
|
|
gdb_stdlog = NULL;
|
|
|
|
|
gdb_stdtarg = NULL;
|
2010-09-11 18:00:27 +02:00
|
|
|
|
gdb_stdtargerr = NULL;
|
2002-11-05 23:38:11 +01:00
|
|
|
|
#endif
|
|
|
|
|
|
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.
2014-10-23 18:13:35 +02:00
|
|
|
|
gdb_rl_callback_handler_remove ();
|
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.
2004-09-13 20:26:31 +02:00
|
|
|
|
delete_file_handler (input_fd);
|
2002-11-05 23:38:11 +01:00
|
|
|
|
}
|