bug#1708: 23.0.60; g_main_context_prepare() called recursively

bug#1708: 23.0.60; g_main_context_prepare() called recursively

[Markus Triska]

From OSX 10.4, I log in via ssh to a remote machine R running Ubuntu
jaunty, and there do:

   $ emacs -nw -Q -f server-start

In another terminal in OSX, I again ssh to R, and there do:

   $ emacsclient -c

As expected, I get an X frame on the local machine (OSX). Now I close
the terminal in which I started the remote emacsclient, thus killing
the client. The remote server at this point shows the message:

   server: Connection lost to X server `localhost:11.0'

Now I again ssh to R, and again do there:

   $ emacsclient -c

Now, only a blank frame shows up, and the remote server repeatedly
shows the following message:

 (emacs:11624): GLib-WARNING **: g_main_context_prepare() called
   recursively from within a source's check() or prepare() member.

Information about Emacs on the remote machine R:

   In GNU Emacs 23.0.60.1 (i686-pc-linux-gnu, GTK+ Version 2.14.5)
    of 2008-12-24 on R
   Windowing system distributor `The XFree86 Project, Inc', version 11.0.40400000
   Important settings:
     value of $LC_ALL: nil
     value of $LC_COLLATE: nil
     value of $LC_CTYPE: nil
     value of $LC_MESSAGES: nil
     value of $LC_MONETARY: nil
     value of $LC_NUMERIC: nil
     value of $LC_TIME: nil
     value of $LANG: en_US.UTF-8
     value of $XMODIFIERS: nil
     locale-coding-system: utf-8-unix
     default-enable-multibyte-characters: t

This also works when I ssh from R to R itself, i.e., I can reproduce
the problem in Ubuntu jaunty alone as well.

In GNU Emacs 23.0.60.1 (i386-apple-darwin8.11.1, GTK+ Version 2.12.9)
 of 2008-12-25 on mt-computer.local
Windowing system distributor `The XFree86 Project, Inc', version 11.0.40400000
Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: en_GB.UTF-8
  value of $XMODIFIERS: nil
  locale-coding-system: nil
  default-enable-multibyte-characters: t

bug#1708: 23.0.60; g_main_context_prepare() called recursively

[Dan Nicolaescu]

Markus Triska <markus.triska@gmx.at> writes:

  > From OSX 10.4, I log in via ssh to a remote machine R running Ubuntu
  > jaunty, and there do:
  > 
  >    $ emacs -nw -Q -f server-start
  > 
  > In another terminal in OSX, I again ssh to R, and there do:
  > 
  >    $ emacsclient -c
  > 
  > As expected, I get an X frame on the local machine (OSX). Now I close
  > the terminal in which I started the remote emacsclient, thus killing
  > the client. The remote server at this point shows the message:
  > 
  >    server: Connection lost to X server `localhost:11.0'
  > 
  > Now I again ssh to R, and again do there:
  > 
  >    $ emacsclient -c
  > 
  > Now, only a blank frame shows up, and the remote server repeatedly
  > shows the following message:
  > 
  >  (emacs:11624): GLib-WARNING **: g_main_context_prepare() called
  >    recursively from within a source's check() or prepare() member.
  > 
  > Information about Emacs on the remote machine R:
  > 
  >    In GNU Emacs 23.0.60.1 (i686-pc-linux-gnu, GTK+ Version 2.14.5)
  >     of 2008-12-24 on R

Can you reproduce this when configuring without GTK+?
If not, this and your other disconnect bug are probably just instances
of the same very old Gtk+ bug: http://bugzilla.gnome.org/show_bug.cgi?id=85715.
Gtk+ does deal well with disconnecting... 

bug#1708: 23.0.60; g_main_context_prepare() called recursively

[Markus Triska]

Dan Nicolaescu <dann@ics.uci.edu> writes:

> Can you reproduce this when configuring without GTK+?

With --with-x-toolkit=lucid, the second "emacsclient -c" shows:

   Waiting for Emacs...
   *ERROR*: X protocol error: BadLength (poly request too large or internal
   Xlib length error) on protocol request 19

After a few further $ emacsclient -c, the server always crashes with:

#0 0x0003392d in redisplay_internal (preserve_echo_area=0) at
 xdisp.c:11800#1 0x000342f6 in redisplay_preserve_echo_area
 (from_where=12) at xdisp.c:12040#2 0x001920a7 in
 wait_reading_process_output (time_limit=30, microsecs=0, read_kbd=-1,
 do_display=1, wait_for_cell=50332681, wait_proc=0x0, just_wait_proc=0)
 at process.c:4997#3 0x0000c01e in sit_for (timeout=240, reading=1,
 do_display=1) at dispnew.c:6637#4 0x000e8c9a in read_char
 (commandflag=1, nmaps=2, maps=0xbffff3d0, prev_event=50332681,
 used_mouse_menu=0xbffff4c8, end_time=0x0) at keyboard.c:2903#5
 0x000eaa0a in read_key_sequence (keybuf=0xbffff588, bufsize=30,
 prompt=50332681, dont_downcase_last=0, can_return_switch_frame=1,
 fix_current_buffer=1) at keyboard.c:9354#6 0x000ece2b in command_loop_1
 () at keyboard.c:1632#7 0x0014c94c in internal_condition_case
 (bfun=0xecc0d <command_loop_1>, handlers=50372345, hfun=0xe59a8
 <cmd_error>) at eval.c:1511#8 0x000ded84 in command_loop_2 () at
 keyboard.c:1349#9 0x0014c59e in internal_catch (tag=50368417,
 func=0xded40 <command_loop_2>, arg=50332681) at eval.c:1247#10
 0x000deb26 in command_loop () at keyboard.c:1328#11 0x000debdf in
 recursive_edit_1 () at keyboard.c:942#12 0x000ded27 in Frecursive_edit
 () at keyboard.c:1004#13 0x000ddb3c in main (argc=5, argv=0xbffff9c0)
 at emacs.c:1786

bug#1708: 23.0.60; g_main_context_prepare() called recursively

[Dan Nicolaescu]

Markus Triska <markus.triska@gmx.at> writes:

  > Dan Nicolaescu <dann@ics.uci.edu> writes:
  > 
  > > Can you reproduce this when configuring without GTK+?
  > 
  > With --with-x-toolkit=lucid, the second "emacsclient -c" shows:
  > 
  >    Waiting for Emacs...
  >    *ERROR*: X protocol error: BadLength (poly request too large or internal
  >    Xlib length error) on protocol request 19
  > 
  > After a few further $ emacsclient -c, the server always crashes with:

I can't reproduce this on an Fedora 10 machine...

bug#1708: 23.0.60; g_main_context_prepare() called recursively

[Lars Ingebrigtsen]

Markus Triska <markus.triska@gmx.at> writes:

>> Can you reproduce this when configuring without GTK+?
>
> With --with-x-toolkit=lucid, the second "emacsclient -c" shows:
>
>    Waiting for Emacs...
>    *ERROR*: X protocol error: BadLength (poly request too large or internal
>    Xlib length error) on protocol request 19
>
> After a few further $ emacsclient -c, the server always crashes with:
>
> #0 0x0003392d in redisplay_internal (preserve_echo_area=0) at
>  xdisp.c:11800#1 0x000342f6 in redisplay_preserve_echo_area

Are you still seeing this in Emacs 24.3?

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

bug#1708: 23.0.60; g_main_context_prepare() called recursively

[Markus Triska]

Lars Ingebrigtsen <larsi@gnus.org> writes:

> Are you still seeing this in Emacs 24.3?

No, since I am no longer using emacsclient due to its memory leaks (#1504).