Re: mouse-face and help echo support for xterm mouse

[Eli Zaretskii <eliz@gnu.org>]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Cc: Eli Zaretskii <eliz@gnu.org>,  emacs-devel@gnu.org
> Date: Wed, 04 Nov 2020 13:47:29 -0500
> 
> `gpm_hold_quit->kind` could hold a different value from `NO_EVENT` on the
> second iteration (and subsequent ones) of the loop.
> 
> > Because it is the first time passed, we can safely deduce that
> >  hold_quit->kind will still be NO_EVENT *unless* a previous loop iteration
> > changed it.
> 
> That's right.
> 
> > Because it is the only time passed, we can safely deduce that
> >  hold_quit->kind can only changed in a previous loop iteration by this
> >  function storing a quit event.  However, this can not happen.  hold_quit
> >  would get set by kbd_buffer_store_buffered_event which
> >  kbd_buffer_store_event_hold calls.  To get set, the following must all be
> > true:
> >
> > 1. An event of kind ASCII_KEYSTROKE_EVENT must be processed.
> > 2. And that event's character code must be quit_char.
> > 3. And that event must be on the current frame's KBOARD (not quite sure what
> >   this means).
> >
> > However, 1 will never be true as the only events that ever reach this point
> > are of kind GPM_CLICK_EVENT, as deduced from the above asserts.
> 
> Oh, I see, indeed.

Are you guys taking into consideration the fact that this is a TTY,
where C-g causes a SIGINT, so it can come in at any time?