OpenE2K
/
binutils-gdb
12
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Remote non-stop mode support. 

* remote.c (pending_stop_reply): New.
	(struct remote_state) <non_stop_aware, support_vCont_t>: New
	fields.
	(remote_async_inferior_event_token)
	(remote_async_get_pending_events_token): New.
	(notice_new_inferiors): New, abstracted out from record_currthread.
	(record_currthread): Call it.
	(remote_threads_info): Default threads to running in non-stop
	mode.  In non-stop mode, only qfThreadInfo is supported.
	(remote_close): Discard all pending stop_replies.  Close the event
	sources.
	(set_stop_requested_callback): New.
	(remote_start_remote): Implement non-stop mode startup.  In
	all-stop, don't clear the thread list here.
	(remote_non_stop_feature): New.
	(remote_protocol_features): Add a "QNonStop" feature.
	(remote_open_1): Clear cached_wait_status and non_stop_aware.
	Clear the thread list here.
	(remote_detach_1): Discard pending stop replies of the process we
	detached from.
	(extended_remote_attach_1): Implement non-stop mode.
	(remote_vcont_probe): Recognize `vCont;t'.
	(remote_vcont_resume): Implement non-stop mode.
	(remote_resume): Don't set waiting_for_stop_reply in non-stop
	mode.
	(remote_stop_ns): New.
	(remote_stop): Rename to ...
	(remote_stop_as): ... this.  If we have a cached wait status,
	don't bother interrupting the remote.
	(remote_stop): Reimplement as wrapper around remote_stop_as and
	remote_stop_ns.
	(interrupt_query): Don't query in async mode.
	(struct cached_reg, cahed_reg_t): New.
	(struct stop_reply): New.
	(stop_reply_queue): New.
	(stop_reply_xmalloc, stop_reply_xfree)
	(discard_pending_stop_replies, do_stop_reply_xfree)
	(queued_stop_reply, push_stop_reply, peek_stop_reply)
	(remote_parse_stop_reply, remote_get_pending_stop_replies)
	(process_stop_reply): New.
	(remote_wait_ns): New.
	(remote_wait_as): Use remote_parse_stop_reply.  Invalidate the
	notion of current general thread is a process exit was reported.
	(remote_wait): Call remote_wait_ns in non-stop mode.
	(handle_notification): New.
	(putpkt_binary): Handle notifications.  Don't care for
	waiting_for_stop_reply in non-stop mode.
	(getpkt_sane): Rename to ...
	(getpkt_or_notif_sane_1): ... this.  Add `expecting_notif'
	argument.  Handle it.  Handle notifications.
	(getpkt_sane): Reimplement as wrapper around getpkt_or_notif_sane_1.
	(getpkt_or_notif_sane): New.
	(select_new_thread_callback): Check for exited state instead of
	comparing the ptid against minus_one_ptid.
	(extended_remote_create_inferior_1): Query the remote about the
	current thread.
	(remote_supports_non_stop): New.
	(init_remote_ops): Register it.
	(remote_async_inferior_event_handler): New.
	(remote_async_get_pending_events_handler): New.

	* infcmd.c (proceed_thread_callback): Comment.
	(proceed_after_attach_callback, proceed_after_attach): New.
	(attach_command_post_wait): In background attach, resume all
	threads, but only if they are unsignalled, and not explicitly
	stopped.  In foreground attach, in non-stop mode, make sure to
	stop all threads of the just attached to process.
	(attach_command): In non-stop mode: If doing a background attach,
	stop at least one thread.  If a foreground attach, stop all
	threads.
This commit is contained in:
Pedro Alves 2008-10-24 20:38:26 +00:00
parent 50d0174816
commit 74531fed1f
3 changed files with 1301 additions and 217 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

75
gdb/ChangeLog
View File

@ -1,3 +1,78 @@
2008-10-24 Pedro Alves <pedro@codesourcery.com>
Remote non-stop mode support.
* remote.c (pending_stop_reply): New.
(struct remote_state) <non_stop_aware, support_vCont_t>: New
fields.
(remote_async_inferior_event_token)
(remote_async_get_pending_events_token): New.
(notice_new_inferiors): New, abstracted out from record_currthread.
(record_currthread): Call it.
(remote_threads_info): Default threads to running in non-stop
mode. In non-stop mode, only qfThreadInfo is supported.
(remote_close): Discard all pending stop_replies. Close the event
sources.
(set_stop_requested_callback): New.
(remote_start_remote): Implement non-stop mode startup. In
all-stop, don't clear the thread list here.
(remote_non_stop_feature): New.
(remote_protocol_features): Add a "QNonStop" feature.
(remote_open_1): Clear cached_wait_status and non_stop_aware.
Clear the thread list here.
(remote_detach_1): Discard pending stop replies of the process we
detached from.
(extended_remote_attach_1): Implement non-stop mode.
(remote_vcont_probe): Recognize `vCont;t'.
(remote_vcont_resume): Implement non-stop mode.
(remote_resume): Don't set waiting_for_stop_reply in non-stop
mode.
(remote_stop_ns): New.
(remote_stop): Rename to ...
(remote_stop_as): ... this. If we have a cached wait status,
don't bother interrupting the remote.
(remote_stop): Reimplement as wrapper around remote_stop_as and
remote_stop_ns.
(interrupt_query): Don't query in async mode.
(struct cached_reg, cahed_reg_t): New.
(struct stop_reply): New.
(stop_reply_queue): New.
(stop_reply_xmalloc, stop_reply_xfree)
(discard_pending_stop_replies, do_stop_reply_xfree)
(queued_stop_reply, push_stop_reply, peek_stop_reply)
(remote_parse_stop_reply, remote_get_pending_stop_replies)
(process_stop_reply): New.
(remote_wait_ns): New.
(remote_wait_as): Use remote_parse_stop_reply. Invalidate the
notion of current general thread is a process exit was reported.
(remote_wait): Call remote_wait_ns in non-stop mode.
(handle_notification): New.
(putpkt_binary): Handle notifications. Don't care for
waiting_for_stop_reply in non-stop mode.
(getpkt_sane): Rename to ...
(getpkt_or_notif_sane_1): ... this. Add `expecting_notif'
argument. Handle it. Handle notifications.
(getpkt_sane): Reimplement as wrapper around getpkt_or_notif_sane_1.
(getpkt_or_notif_sane): New.
(select_new_thread_callback): Check for exited state instead of
comparing the ptid against minus_one_ptid.
(extended_remote_create_inferior_1): Query the remote about the
current thread.
(remote_supports_non_stop): New.
(init_remote_ops): Register it.
(remote_async_inferior_event_handler): New.
(remote_async_get_pending_events_handler): New.
* infcmd.c (proceed_thread_callback): Comment.
(proceed_after_attach_callback, proceed_after_attach): New.
(attach_command_post_wait): In background attach, resume all
threads, but only if they are unsignalled, and not explicitly
stopped. In foreground attach, in non-stop mode, make sure to
stop all threads of the just attached to process.
(attach_command): In non-stop mode: If doing a background attach,
stop at least one thread. If a foreground attach, stop all
threads.
2008-10-24 Pedro Alves <pedro@codesourcery.com>
* event-loop.h: Mention async_event_handlers.

101
gdb/infcmd.c
View File

@ -570,6 +570,15 @@ start_command (char *args, int from_tty)
static int
proceed_thread_callback (struct thread_info *thread, void *arg)
{
/* We go through all threads individually instead of compressing
into a single target `resume_all' request, because some threads
may be stopped in internal breakpoints/events, or stopped waiting
for its turn in the displaced stepping queue (that is, they are
running && !executing). The target side has no idea about why
the thread is stopped, so a `resume_all' command would resume too
much. If/when GDB gains a way to tell the target `hold this
thread stopped until I say otherwise', then we can optimize
this. */
if (!is_stopped (thread->ptid))
return 0;
@ -2011,6 +2020,48 @@ vector_info (char *args, int from_tty)
}
/* Used in `attach&' command. ARG is a point to an integer
representing a process id. Proceed threads of this process iff
they stopped due to debugger request, and when they did, they
reported a clean stop (TARGET_SIGNAL_0). Do not proceed threads
that have been explicitly been told to stop. */
static int
proceed_after_attach_callback (struct thread_info *thread,
void *arg)
{
int pid = * (int *) arg;
if (ptid_get_pid (thread->ptid) == pid
&& !is_exited (thread->ptid)
&& !is_executing (thread->ptid)
&& !thread->stop_requested
&& thread->stop_signal == TARGET_SIGNAL_0)
{
switch_to_thread (thread->ptid);
clear_proceed_status ();
proceed ((CORE_ADDR) -1, TARGET_SIGNAL_DEFAULT, 0);
}
return 0;
}
static void
proceed_after_attach (int pid)
{
/* Don't error out if the current thread is running, because
there may be other stopped threads. */
struct cleanup *old_chain;
/* Backup current thread and selected frame. */
old_chain = make_cleanup_restore_current_thread ();
iterate_over_threads (proceed_after_attach_callback, &pid);
/* Restore selected ptid. */
do_cleanups (old_chain);
}
/*
* TODO:
* Should save/restore the tty state since it might be that the
@ -2075,11 +2126,44 @@ attach_command_post_wait (char *args, int from_tty, int async_exec)
target_terminal_inferior ();
if (async_exec)
proceed ((CORE_ADDR) -1, TARGET_SIGNAL_0, 0);
{
/* The user requested an `attach&', so be sure to leave threads
that didn't get a signal running. */
/* Immediatelly resume all suspended threads of this inferior,
and this inferior only. This should have no effect on
already running threads. If a thread has been stopped with a
signal, leave it be. */
if (non_stop)
proceed_after_attach (inferior->pid);
else
{
if (inferior_thread ()->stop_signal == TARGET_SIGNAL_0)
{
clear_proceed_status ();
proceed ((CORE_ADDR) -1, TARGET_SIGNAL_DEFAULT, 0);
}
}
}
else
{
/* The user requested a plain `attach', so be sure to leave
the inferior stopped. */
if (target_can_async_p ())
async_enable_stdin ();
/* At least the current thread is already stopped. */
/* In all-stop, by definition, all threads have to be already
stopped at this point. In non-stop, however, although the
selected thread is stopped, others may still be executing.
Be sure to explicitly stop all threads of the process. This
should have no effect on already stopped threads. */
if (non_stop)
target_stop (pid_to_ptid (inferior->pid));
/* Tell the user/frontend where we're stopped. */
normal_stop ();
if (deprecated_attach_hook)
deprecated_attach_hook ();
@ -2161,6 +2245,21 @@ attach_command (char *args, int from_tty)
init_wait_for_inferior ();
clear_proceed_status ();
if (non_stop)
{
/* If we find that the current thread isn't stopped, explicitly
do so now, because we're going to install breakpoints and
poke at memory. */
if (async_exec)
/* The user requested an `attach&'; stop just one thread. */
target_stop (inferior_ptid);
else
/* The user requested an `attach', so stop all threads of this
inferior. */
target_stop (pid_to_ptid (ptid_get_pid (inferior_ptid)));
}
/* Some system don't generate traps when attaching to inferior.
E.g. Mach 3 or GNU hurd. */
if (!target_attach_no_wait)

1342
gdb/remote.c
View File

File diff suppressed because it is too large Load Diff