bug#54656: 29.0.50; [menu] key is not read anymore

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

Hi,

After compiling from master I found my [menu] key not usable anymore
within emacs.

It don't show with C-h c
It show in xev


In GNU Emacs 29.0.50 (build 1, x86_64-pc-linux-gnu, GTK+ Version 3.24.24, cairo version 1.16.0)
 of 2022-03-31 built on localhost
Repository revision: 948181df9cbdcc8845fc3662e2007d8e09f48c71
Repository branch: master
Windowing system distributor 'The X.Org Foundation', version 11.0.12011000
System Description: Debian GNU/Linux 11 (bullseye)

Configured using:
 'configure --build x86_64-linux-gnu --prefix=/opt/emacs
 --with-mailutils --with-sound=yes --without-gconf --without-gsettings
 --with-x=yes --without-toolkit-scroll-bars --with-x-toolkit=gtk3
 --with-json --with-native-compilation --with-xwidgets
 build_alias=x86_64-linux-gnu 'CFLAGS=-O2 -Wall ''

Configured features:
CAIRO DBUS FREETYPE GIF GLIB GMP GNUTLS HARFBUZZ JPEG JSON LIBSELINUX
LIBXML2 MODULES NATIVE_COMP NOTIFY INOTIFY PDUMPER PNG SECCOMP SOUND
SQLITE3 THREADS TIFF X11 XDBE XIM XPM XWIDGETS GTK3 ZLIB

Important settings:
  value of $LANG: en_US
  locale-coding-system: utf-8

Load-path shadows:
/home/user/.emacs.d/elpa/transient-0.3.7/transient hides /opt/emacs/share/emacs/29.0.50/lisp/transient

bug#54656: 29.0.50; [menu] key is not read anymore

[Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors]

dal-blazej@onenetbeyond.org writes:

> Hi,
>
> After compiling from master I found my [menu] key not usable anymore
> within emacs.
>
> It don't show with C-h c
> It show in xev
>

What happens if you run Emacs like this?

  emacs -q -xrm 'Emacs.useXIM: false'

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

> What happens if you run Emacs like this?
>
>   emacs -q -xrm 'Emacs.useXIM: false'

The key is still not read.

bug#54656: 29.0.50; [menu] key is not read anymore

[Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors]

dal-blazej@onenetbeyond.org writes:

> The key is still not read.

Okay, please put a breakpoint here:

 	  /* Common for all keysym input events.  */
 	  XSETFRAME (inev.ie.frame_or_window, f);
=======>  inev.ie.modifiers
 	    = x_x_to_emacs_modifiers (FRAME_DISPLAY_INFO (f), modifiers);
 	  inev.ie.timestamp = xkey.time;


and show the value of `keysym' when you press the menu key.

Thanks.

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

Po Lu <luangruo@yahoo.com> writes:

> Okay, please put a breakpoint here:
>
>  	  /* Common for all keysym input events.  */
>  	  XSETFRAME (inev.ie.frame_or_window, f);
> =======>  inev.ie.modifiers
>  	    = x_x_to_emacs_modifiers (FRAME_DISPLAY_INFO (f), modifiers);
>  	  inev.ie.timestamp = xkey.time;
>
>
> and show the value of `keysym' when you press the menu key.

Nothing : gdb use the break point for any other key, but not for menu.

bug#54656: 29.0.50; [menu] key is not read anymore

[Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors]

dal-blazej@onenetbeyond.org writes:

> Nothing : gdb use the break point for any other key, but not for menu.

Okay, can you place a breakpoint on `x_display_set_last_user_time' and
see if it gets called when you press the menu key?

Thanks in advance.

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

> Okay, can you place a breakpoint on `x_display_set_last_user_time' and
> see if it gets called when you press the menu key?

Sorry I lack comprehension of the gdb and C.

If I place a breakpoint on :

> static void
> x_display_set_last_user_time (struct x_display_info *dpyinfo, Time time)
> {
> ...

It is called to many time while initiating Emacs and I cannot verify the
keypress.  How do I do that the right way ?

bug#54656: 29.0.50; [menu] key is not read anymore

[Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors]

dal-blazej@onenetbeyond.org writes:

> Sorry I lack comprehension of the gdb and C.
>
> If I place a breakpoint on :
>
>> static void
>> x_display_set_last_user_time (struct x_display_info *dpyinfo, Time time)
>> {
>> ...
>
> It is called to many time while initiating Emacs and I cannot verify the
> keypress.  How do I do that the right way ?

Ah, sorry.  Place the breakpoint here instead, inside the function
`handle_one_xevent':

    case KeyPress:
==>   x_display_set_last_user_time (dpyinfo, event->xkey.time);
      ignore_next_mouse_click_timeout = 0;
      coding = Qlatin_1;

Thanks.

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

It is triggered :

> Thread 1 "emacs" hit Breakpoint 1, handle_one_xevent (dpyinfo=0x55555774f800, event=0x7fffffffd700, finish=0x555555beabb8 <current_finish>, hold_quit=0x7fffffffd970) at xterm.c:5240
> 5240	  dpyinfo->last_user_time = time;
> (gdb) s
> 13844	      x_display_set_last_user_time (dpyinfo, event->xkey.time);
> (gdb) s
> 0x000055555568a997 in x_display_set_last_user_time (time=119383756, dpyinfo=0x55555774f800) at xterm.c:5240
> 5240	  dpyinfo->last_user_time = time;

bug#54656: 29.0.50; [menu] key is not read anymore

[Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors]

dal-blazej@onenetbeyond.org writes:

> It is triggered :
>
>> Thread 1 "emacs" hit Breakpoint 1, handle_one_xevent
>> (dpyinfo=0x55555774f800, event=0x7fffffffd700, finish=0x555555beabb8
>> <current_finish>, hold_quit=0x7fffffffd970) at xterm.c:5240
>> 5240	  dpyinfo->last_user_time = time;
>> (gdb) s
>> 13844	      x_display_set_last_user_time (dpyinfo, event->xkey.time);
>> (gdb) s
>> 0x000055555568a997 in x_display_set_last_user_time (time=119383756, dpyinfo=0x55555774f800) at xterm.c:5240
>> 5240	  dpyinfo->last_user_time = time;

I guess what you have to do is to step through each statement until you
find out why it isn't reaching this piece of handle_one_xevent:

#if defined XK_ISO_Lock && defined XK_ISO_Last_Group_Lock
		    || (XK_ISO_Lock <= orig_keysym
			&& orig_keysym <= XK_ISO_Last_Group_Lock)
#endif
		    ))
	    {
	      STORE_KEYSYM_FOR_DEBUG (keysym);
	      /* make_lispy_event will convert this to a symbolic
		 key.  */
	      inev.ie.kind = NON_ASCII_KEYSTROKE_EVENT;
	      inev.ie.code = keysym;
====>	      goto done_keysym;
	    }

Thanks.

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

> I guess what you have to do is to step through each statement until you
> find out why it isn't reaching this piece of handle_one_xevent:

It seems the key is ... initialized to 0 ?

(gdb) p xkey
$13 = {type = 0, serial = 0, send_event = 1466608144,
  display = 0x898f22a9e227e100, window = 140737488343872, root = 93824994679155,
  subwindow = 0, time = 93824993826130, x = 2, y = 0, x_root = 0, y_root = 0,
  state = 0, keycode = 0, same_screen = 0}

(gdb) s

(gdb) p xkey.state
$14 = 0

(gdb) s
x_emacs_to_x_modifiers (dpyinfo=0x555557727a00, state=0) at lisp.h:1155

(gdb) p a
$15 = {i = {0, 1045149306}, d = 1.2904777690891933e-08}

(gdb) s
Fget (symbol=0x0, propname=0xaf80) at lisp.h:747

[...]

(gdb) u
x_emacs_to_x_modifiers (dpyinfo=0x555557727a00, state=0) at xterm.c:9546

(gdb) p tem
$21 = (Lisp_Object) 0x0

[...]

(gdb) n
handle_one_xevent (dpyinfo=0x555557727a00, event=0x7fffffffd5e0,
    finish=0x555555beabb8 <current_finish>, hold_quit=0x7fffffffd850)
    at xterm.c:13926

(gdb) p xkey.state
$30 = 0

At some point gdb prints
(gpb) n
0x00007ffff760623f in ?? () from /lib/x86_64-linux-gnu/libgdk-3.so.0

and I am unable to continue my inspection.

bug#54656: 29.0.50; [menu] key is not read anymore

[Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors]

dal-blazej@onenetbeyond.org writes:

>> I guess what you have to do is to step through each statement until you
>> find out why it isn't reaching this piece of handle_one_xevent:
>
> It seems the key is ... initialized to 0 ?
>
> (gdb) p xkey
> $13 = {type = 0, serial = 0, send_event = 1466608144,
>   display = 0x898f22a9e227e100, window = 140737488343872, root = 93824994679155,
>   subwindow = 0, time = 93824993826130, x = 2, y = 0, x_root = 0, y_root = 0,
>   state = 0, keycode = 0, same_screen = 0}

I think your Emacs is out of date, since this shouldn't be possible if
the breakpoint was set to the location I asked.

> [...]
>
> (gdb) u
> x_emacs_to_x_modifiers (dpyinfo=0x555557727a00, state=0) at xterm.c:9546
>
> (gdb) p tem
> $21 = (Lisp_Object) 0x0
>
> [...]
>
> (gdb) n
> handle_one_xevent (dpyinfo=0x555557727a00, event=0x7fffffffd5e0,
>     finish=0x555555beabb8 <current_finish>, hold_quit=0x7fffffffd850)
>     at xterm.c:13926

I think your copy of Emacs is very out of date.  Please rebuild and try
again.

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

Hi,

Po Lu <luangruo@yahoo.com> writes:

> dal-blazej@onenetbeyond.org writes:
>
>>> I guess what you have to do is to step through each statement until you
>>> find out why it isn't reaching this piece of handle_one_xevent:
>>
>> It seems the key is ... initialized to 0 ?
>>
>> (gdb) p xkey
>> $13 = {type = 0, serial = 0, send_event = 1466608144,
>>   display = 0x898f22a9e227e100, window = 140737488343872, root =
>> 93824994679155,
>>   subwindow = 0, time = 93824993826130, x = 2, y = 0, x_root = 0, y_root = 0,
>>   state = 0, keycode = 0, same_screen = 0}
>
> I think your Emacs is out of date, since this shouldn't be possible if
> the breakpoint was set to the location I asked.

You asked me to find out why this position was not reachable, so I set
breakpoints ahead... Again, setting at the position asked does not trigger.

I have rebuilt today. I have still at some point :

(gdb) n
0x00007ffff760623f in ?? () from /lib/x86_64-linux-gnu/libgdk-3.so.0
(gdb) n
Cannot find bounds of current function.

...

Setting it to

      if (f != 0)
        {
==>       KeySym keysym, orig_keysym;
          /* al%imercury@uunet.uu.net says that making this 81
             instead of 80 fixed a bug whereby meta chars made
             his Emacs hang.

(gdb) p keysym
$1 = 4596353178100193889

(gdb) p orig_keysym
$2 = <optimized out>

(gdb) p xkey
$6 = {type = 2, serial = 2027, send_event = 0, display = 0x0, window = 0,
  root = 0, subwindow = 2, time = 9223372036854775822, x = 0, y = 0, x_root = 0,
  y_root = 0, state = 0, keycode = 0, same_screen = 0}

Eventually you may try to reproduce that issue with
$ xdotool key Menu
?

Thanks

bug#54656: 29.0.50; [menu] key is not read anymore

[Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors]

dal-blazej@onenetbeyond.org writes:

> You asked me to find out why this position was not reachable, so I set
> breakpoints ahead... Again, setting at the position asked does not trigger.
>
> I have rebuilt today. I have still at some point :
>
> (gdb) n
> 0x00007ffff760623f in ?? () from /lib/x86_64-linux-gnu/libgdk-3.so.0
> (gdb) n
> Cannot find bounds of current function.

That isn't a problem, though you might want to install debug info for
GTK.  You don't have to look in any function other than
handle_one_xevent.

> ...
>
> Setting it to
>
>       if (f != 0)
>         {
> ==>       KeySym keysym, orig_keysym;
>           /* al%imercury@uunet.uu.net says that making this 81
>              instead of 80 fixed a bug whereby meta chars made
>              his Emacs hang.
>
> (gdb) p keysym
> $1 = 4596353178100193889
>
> (gdb) p orig_keysym
> $2 = <optimized out>

You will have to proceed some more, I guess.

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

I read "strange behaviour: Emacs ignoring one key on my keyboard," on
emacs-help today and tried.

$ xmodmap -e 'clear Mod4'

The key is now again responsive.

bug#54656: 29.0.50; [menu] key is not read anymore

[Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors]

dal-blazej@onenetbeyond.org writes:

> I read "strange behaviour: Emacs ignoring one key on my keyboard," on
> emacs-help today and tried.
>
> $ xmodmap -e 'clear Mod4'
>
> The key is now again responsive.

Then that's intended behavior, not a bug, so I'm closing this bug
report.

Thanks.

bug#54656: 29.0.50; [menu] key is not read anymore

[dal-blazej--- via Bug reports for GNU Emacs, the Swiss army knife of text editors]

> What happens if you run Emacs like this?
>
>  emacs -q -xrm 'Emacs.useXIM: false'

Still no joy. 

Is this issue reproductible on another computer ?