From: Patrick Palka <patrick@parcs.ath.cx>
To: Simon Marchi <simon.marchi@ericsson.com>
Cc: gdb-patches@sourceware.org
Subject: Re: [PATCH v2] Restore terminal state in mi_thread_exit (PR gdb/17627)
Date: Wed, 03 Dec 2014 00:09:00 -0000 [thread overview]
Message-ID: <CA+C-WL_7DR0Bj-Co4GhFV+=HuX4VYt1FpGUi-6BcZZGNAuOquA@mail.gmail.com> (raw)
In-Reply-To: <1417558223-27328-1-git-send-email-simon.marchi@ericsson.com>
On Tue, Dec 2, 2014 at 5:10 PM, Simon Marchi <simon.marchi@ericsson.com> wrote:
> When a thread exits, the terminal is left in mode "terminal_is_ours"
> while the target executes. This patch fixes that.
>
> From my understanding, a function calling target_terminal_ours expects
> that the terminal could be in any state at the moment it is called.
> Therefore, it should be its reponsibility to put back the terminal in
> whatever state it was before being called.
It seems to me that the other observer callbacks defined in
mi-interp.c are also affected by this issue, that is, the issue of
having to restore the original terminal state after altering it. So I
wonder if it would make sense to shift this responsibility to the
observer module itself (i.e. generic_observer_notify()), so that all
observers implicitly restore the original terminal state when they
return. That way this kind of pattern wouldn't have to be duplicated
for each individual observer.
>
> I find that this fits quite well the cleanup model, so I implemented a
> cleanup for that.
>
> New in v2:
>
> * Coding style fixes.
> * Use make_cleanup_dtor instead of make_cleanup.
>
> gdb/ChangeLog:
>
> PR gdb/17627
> * target.c (cleanup_restore_target_terminal): New function.
> (make_cleanup_restore_target_terminal): New function.
> * target.h (make_cleanup_restore_target_terminal): New
> declaration.
> * mi/mi-interp.c (mi_thread_exit): Use the new cleanup.
>
> Signed-off-by: Simon Marchi <simon.marchi@ericsson.com>
> ---
> gdb/mi/mi-interp.c | 4 ++++
> gdb/target.c | 34 ++++++++++++++++++++++++++++++++++
> gdb/target.h | 4 ++++
> 3 files changed, 42 insertions(+)
>
> diff --git a/gdb/mi/mi-interp.c b/gdb/mi/mi-interp.c
> index df2b558..60f0666 100644
> --- a/gdb/mi/mi-interp.c
> +++ b/gdb/mi/mi-interp.c
> @@ -386,6 +386,7 @@ mi_thread_exit (struct thread_info *t, int silent)
> {
> struct mi_interp *mi;
> struct inferior *inf;
> + struct cleanup *old_chain;
>
> if (silent)
> return;
> @@ -393,11 +394,14 @@ mi_thread_exit (struct thread_info *t, int silent)
> inf = find_inferior_pid (ptid_get_pid (t->ptid));
>
> mi = top_level_interpreter_data ();
> + old_chain = make_cleanup_restore_target_terminal ();
> target_terminal_ours ();
> fprintf_unfiltered (mi->event_channel,
> "thread-exited,id=\"%d\",group-id=\"i%d\"",
> t->num, inf->num);
> gdb_flush (mi->event_channel);
> +
> + do_cleanups (old_chain);
> }
>
> /* Emit notification on changing the state of record. */
> diff --git a/gdb/target.c b/gdb/target.c
> index ab5f2b9..7161e62 100644
> --- a/gdb/target.c
> +++ b/gdb/target.c
> @@ -528,6 +528,40 @@ target_supports_terminal_ours (void)
> return 0;
> }
>
> +/* Restore the terminal to its previous state (helper for
> + make_cleanup_restore_target_terminal). */
> +
> +static void
> +cleanup_restore_target_terminal (void *arg)
> +{
> + enum terminal_state *previous_state = arg;
> +
> + switch (*previous_state)
> + {
> + case terminal_is_ours:
> + target_terminal_ours ();
> + break;
> + case terminal_is_ours_for_output:
> + target_terminal_ours_for_output ();
> + break;
> + case terminal_is_inferior:
> + target_terminal_inferior ();
> + break;
> + }
> +}
> +
> +/* See target.h. */
> +
> +struct cleanup *
> +make_cleanup_restore_target_terminal (void)
> +{
> + enum terminal_state *ts = xmalloc (sizeof (*ts));
> +
> + *ts = terminal_state;
> +
> + return make_cleanup_dtor (cleanup_restore_target_terminal, ts, xfree);
> +}
> +
> static void
> tcomplain (void)
> {
> diff --git a/gdb/target.h b/gdb/target.h
> index d363b61..eb3220e 100644
> --- a/gdb/target.h
> +++ b/gdb/target.h
> @@ -1413,6 +1413,10 @@ extern void target_terminal_ours (void);
>
> extern int target_supports_terminal_ours (void);
>
> +/* Make a cleanup that restores the state of the terminal to the current
> + value. */
> +extern struct cleanup *make_cleanup_restore_target_terminal (void);
> +
> /* Print useful information about our terminal status, if such a thing
> exists. */
>
> --
> 2.1.3
>
next prev parent reply other threads:[~2014-12-03 0:09 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-12-02 22:10 Simon Marchi
2014-12-03 0:09 ` Patrick Palka [this message]
2014-12-03 14:29 ` Simon Marchi
2014-12-04 15:40 ` Pedro Alves
2014-12-04 15:35 ` Pedro Alves
2014-12-05 16:12 ` Simon Marchi
2014-12-10 17:20 ` Pedro Alves
2014-12-10 18:06 ` Simon Marchi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CA+C-WL_7DR0Bj-Co4GhFV+=HuX4VYt1FpGUi-6BcZZGNAuOquA@mail.gmail.com' \
--to=patrick@parcs.ath.cx \
--cc=gdb-patches@sourceware.org \
--cc=simon.marchi@ericsson.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).