public inbox for gdb@sourceware.org
 help / color / mirror / Atom feed
From: Jonah Graham <jonah@kichwacoders.com>
To: Jan Vrany <jan.vrany@fit.cvut.cz>
Cc: "Joel Brobecker" <brobecker@adacore.com>,
	"Tom Tromey" <tom@tromey.com>,
	"gdb@sourceware.org" <gdb@sourceware.org>,
	"André Pönitz" <apoenitz@t-online.de>
Subject: Re: MI3 and async notifications
Date: Mon, 17 Jun 2019 21:50:00 -0000	[thread overview]
Message-ID: <CAPmGMvgEgO7rkNF8fORh5ndvMyhNd-pLS1CQRW11GSdWWe8jXw@mail.gmail.com> (raw)
In-Reply-To: <a6128e1d7f56d0f5cfcb21c3cb88d8f73c10883a.camel@fit.cvut.cz>

On Mon, 17 Jun 2019 at 16:58, Jan Vrany <jan.vrany@fit.cvut.cz> wrote:

> On Mon, 2019-06-17 at 16:45 -0400, Joel Brobecker wrote:
> > > > Jonah, I was about to ask the same. I understand that you need to
> know
> > > > which breakpoint has been inserted by given command, but this
> > > > if we respond with something like
> > > >
> > > > 1-break-insert main
> > > > =breakpoint-created,bkpt={number="1",type=...}
> > > > 1^done,bkpt-number=1
> > > >
> > > > then you just search for breakpoint with that id, no? Given that MI
> > > > guarantees that =breakpoint-created arrives before ^done reply to
> command.
> > > > Am I missing something?
> > > >
> > >
> > > No you aren't missing anything. That would be a perfectly acceptable
> > > solution for CDT.
> > >
> > > There would still be some other new logic needed for CDT, we would
> > > still have to store all the =breakpoint-created if there is a
> > > -break-insert active and then process all of them when the ^done is
> > > received. However that seems fairly reasonable.
> >
> > Do we even need the bkpt-number=1 attribute in the "done" command?
> > The notification includes that information, so the GUI should have
> > enough info from there to determine which UI element to update, right?
> >
>
> I think so. Imagine you have a UI with breakpoint list and there's
> an "Add Breakpoint" button. When "Add Breakpoint" is pressed, user
> fills location, press "OK", new breakpoint is added and
> *UI pre-selects just added breakpoint*. This is IMO sensible UI behavior.
>
> Now, when you press "OK" to add the breakpoint on a location, behind the
> scenes
> frontend issues -break-insert command and waits for ^done confirmation.
> But as Jonah mentioned, another breakpoint could be inserted meanwhile so
> you'd get something like this on MI channel.
>
> 10-break-insert main
> =breakpoint-created,bkpt={number="1",type=...<some unrelated breakpoint
> from CLI or script>}
> =breakpoint-created,bkpt={number="2",type=...<the breakpoint on main you
> just requested>}
> =breakpoint-created,bkpt={number="3",type=...<some unrelated breakpoint
> from CLI or script>}
> 10^done
>
> Now, if ^done response would not contain bkpt-number=2 attribute, how
> would you know which
> is the breakpoint you need to pre-select in the UI?
>
>
>
Thanks Jan, that looks like a good example to show the problem.

  reply	other threads:[~2019-06-17 21:50 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-06-10 21:19 Jan Vrany
2019-06-10 23:23 ` Jonah Graham
2019-06-11  8:50   ` Jan Vrany
2019-06-11 13:37     ` Jonah Graham
2019-07-05 20:00       ` Pedro Alves
2019-07-05 21:58         ` Jonah Graham
2019-06-15 14:34 ` Tom Tromey
2019-06-17 10:53   ` Jan Vrany
2019-06-17 12:11     ` Jonah Graham
2019-06-17 12:14     ` Joel Brobecker
2019-06-17 12:26       ` Jonah Graham
2019-06-17 12:56         ` Joel Brobecker
2019-06-17 13:12           ` Jan Vrany
2019-06-17 13:23             ` Jonah Graham
2019-06-17 20:45               ` Joel Brobecker
2019-06-17 20:58                 ` Jan Vrany
2019-06-17 21:50                   ` Jonah Graham [this message]
2019-06-17 13:12           ` Jonah Graham
2019-06-17 19:52     ` André Pönitz
2019-06-18  3:14 ` Simon Marchi
2019-06-18 20:38   ` Jan Vrany
2019-06-19 15:29     ` Simon Marchi
2019-06-19 20:58       ` Jan Vrany
2019-06-20 15:31         ` Simon Marchi
2019-06-20 20:46           ` Jan Vrany
2019-07-05 19:35           ` Pedro Alves

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=CAPmGMvgEgO7rkNF8fORh5ndvMyhNd-pLS1CQRW11GSdWWe8jXw@mail.gmail.com \
    --to=jonah@kichwacoders.com \
    --cc=apoenitz@t-online.de \
    --cc=brobecker@adacore.com \
    --cc=gdb@sourceware.org \
    --cc=jan.vrany@fit.cvut.cz \
    --cc=tom@tromey.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).