Re: master c1106b2ee7: Make sure native keypresses are received immediately on GTK

Re: master c1106b2ee7: Make sure native keypresses are received immediately on GTK

[Ken Brown]

On 1/8/2022 8:29 AM, Po Lu via Mailing list for Emacs changes wrote:
> diff --git a/src/gtkutil.c b/src/gtkutil.c
> index aaa2933f86..ff5c0f2751 100644
> --- a/src/gtkutil.c
> +++ b/src/gtkutil.c
> @@ -6250,6 +6250,8 @@ xg_widget_key_press_event_cb (GtkWidget *widget, GdkEvent *event,
>         xg_pending_quit_event.kind = NO_EVENT;
>         kbd_buffer_store_buffered_event (&inev, &xg_pending_quit_event);
>       }
> +
> +  raise (SIGIO);

Should this be guarded by "#ifdef USABLE_SIGIO"?  And I have the same question 
for the call to "raise (SIGIO)" in pgtkterm.c.

Ken

Re: master c1106b2ee7: Make sure native keypresses are received immediately on GTK

[Po Lu]

Ken Brown <kbrown@cornell.edu> writes:

> Should this be guarded by "#ifdef USABLE_SIGIO"?  And I have the same
> question for the call to "raise (SIGIO)" in pgtkterm.c.

The PGTK build is apparently not supposed to work in systems without
usable SIGIO, but this is.

Good catch, thanks!

Re: master c1106b2ee7: Make sure native keypresses are received immediately on GTK

[Ken Brown]

On 1/8/2022 7:51 PM, Po Lu wrote:
> The PGTK build is apparently not supposed to work in systems without
> usable SIGIO,

So should a test for this be added to configure.ac?  I just checked that a build 
--with-pgtk succeeds on Cygwin, which does not have usable SIGIO.

Ken

Re: master c1106b2ee7: Make sure native keypresses are received immediately on GTK

[Po Lu]

Ken Brown <kbrown@cornell.edu> writes:

> So should a test for this be added to configure.ac?  I just checked
> that a build --with-pgtk succeeds on Cygwin, which does not have
> usable SIGIO.

If SIGIO is defined and raising it works, it should work as well.
I don't know the details -- maybe Yuuki can chime in.

Thanks.