Gnome 3.14 problems with wayland

marzoul · 2014-10-17 08:50:25

Hi everybody,

Today I updated to repo testing to try the Gnome Wayland session. But I can't get it launched.

I followed what's advised on the wiki there:
https://wiki.archlinux.org/index.php/Wa … _libraries
https://wiki.archlinux.org/index.php/Wa … _libraries
https://wiki.gnome.org/Initiatives/Wayland/Applications

I tried to launch it as user (not from gdm) with the following script so it also sets some appropriate env vars:

#!/bin/sh

export GDK_BACKEND=wayland
export SDL_VIDEODRIVER=wayland
export CLUTTER_BACKEND=wayland

# WARNING: Something needed for QT backend? Recompile from git...
# https://wiki.archlinux.org/index.php/Wayland#GUI_libraries

export ECORE_EVAS_ENGINE=wayland_egl
export ELM_ENGINE=wayland_egl
#export ECORE_EVAS_ENGINE=wayland_shm
#export ELM_ENGINE=wayland_shm

gnome-session --session=gnome-wayland

I checked journalctl:

Oct 17 10:23:28 Boltzmann login[295]: LOGIN ON tty1 BY adrien
Oct 17 10:23:39 Boltzmann gnome-session[397]: gnome-session[397]: GLib-CRITICAL: g_environ_setenv: assertion 'value != NULL' failed
Oct 17 10:23:39 Boltzmann gnome-session[397]: GLib-CRITICAL: g_environ_setenv: assertion 'value != NULL' failed
Oct 17 10:23:42 Boltzmann systemd-logind[286]: kernel does not support evdev-revocation
Oct 17 10:23:42 Boltzmann kernel: gnome-shell[406]: segfault at 61 ip 00007f2a7d097455 sp 00007ffff0ab4178 error 4 in libgobject-2.0.so.0.4200.0[7f2a7d064000+4f000]
Oct 17 10:23:42 Boltzmann gnome-session[397]: gnome-session[397]: WARNING: Application 'gnome-shell-wayland.desktop' killed by signal 11
Oct 17 10:23:42 Boltzmann gnome-session[397]: WARNING: Application 'gnome-shell-wayland.desktop' killed by signal 11
Oct 17 10:23:42 Boltzmann gnome-session[397]: Unrecoverable failure in required component gnome-shell-wayland.desktop
Oct 17 10:23:43 Boltzmann kernel: fuse init (API version 7.23)
Oct 17 10:23:43 Boltzmann kernel: cfg80211: Calling CRDA to update world regulatory domain
Oct 17 10:23:45 Boltzmann org.gtk.vfs.Daemon[401]: A connection to the bus can't be made
Oct 17 10:23:45 Boltzmann org.gtk.vfs.Daemon[401]: g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream returned 0 bytes on an async read (g-io-erro
Oct 17 10:23:45 Boltzmann ca.desrt.dconf[401]: g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream returned 0 bytes on an async read (g-io-error-qu
Oct 17 10:23:50 Boltzmann systemd-coredump[410]: Process 406 (gnome-shell) of user 1000 dumped core.

And dmesg:

[   42.954209] gnome-shell[406]: segfault at 61 ip 00007f2a7d097455 sp 00007ffff0ab4178 error 4 in libgobject-2.0.so.0.4200.0[7f2a7d064000+4f000]

Any idea about what I did wrong?

bluebyt · 2014-10-17 18:53:31

Hi marzoul,
You have the same error as me with "journalctl", I have been looking for clue on the web, but I haven't found any solution yet.

Post #168

marzoul · 2014-10-17 20:23:34

HI bluebyt, thanks for the link to this thread.
So now we known launching it under gdm doesn't help.
My journalctl dump shows the source of the crash is that an assert that fails in gnome-session. What's sad is we don't know where in the code and what is was the test. So the next logical thing to try is gdb and probably the package gnome-session-git.

marzoul · 2014-10-17 23:31:31

Here is the story about me trying to launch gnome-session under gdb.

First compile it with debug symbols...
Get AUR package gnome-session-git

Modify PKGBUILD so the symbols are kept:
- Add this to options: '!strip'
- Override cflags: make the command autogen.sh launch with this env var:

CFLAGS="-pipe -fstack-protector-strong --param=ssp-buffer-size=4 -g"

Install that package, launch the command:

gdb --args gnome-session --session=gnome-wayland

Before running the program, set a breakpoint on the only function we know is involved in the assert failure:

b g_environ_setenv

Then run the program. GDB does not stop to indicate an assert failed, too bad. However the function g_environ_setenv is entered only 4 times, all of them from lines 1024, 1025 and 1028 of source file gnome-session/gsm-autostart-app.c.

A quick look at the line 1024:

g_app_launch_context_setenv (ctx, "DISPLAY", g_getenv ("DISPLAY"))

It seems to get the env var DISPLAY and save the value in a structure that contains all env vars for future app launch. On my machine, DISPLAY is not set so this line is a very good candidate for failure.

Firts test: set DISPLAY to something. I set :0.0 because I remember nothing about what we are supposed to set. Anyway, a crash still happens, but a completely different one, which seems to reference stuff from X... journalctl says this:

Oct 18 01:05:17 Boltzmann org.a11y.Bus[15403]: Activating service name='org.a11y.atspi.Registry'
Oct 18 01:05:17 Boltzmann org.a11y.Bus[15403]: Successfully activated service 'org.a11y.atspi.Registry'
Oct 18 01:05:17 Boltzmann org.a11y.Bus[15403]: Activating service name='org.a11y.atspi.Registry'
Oct 18 01:05:17 Boltzmann org.a11y.Bus[15403]: Successfully activated service 'org.a11y.atspi.Registry'
Oct 18 01:05:17 Boltzmann org.a11y.Bus[15403]: Activating service name='org.a11y.atspi.Registry'
Oct 18 01:05:17 Boltzmann org.a11y.Bus[15403]: Successfully activated service 'org.a11y.atspi.Registry'
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: SpiRegistry daemon is running with well-known name - org.a11y.atspi.Registry
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: ** (at-spi2-registryd:15413): WARNING **: Could not open X display
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: ** (at-spi2-registryd:15413): WARNING **: AT-SPI: Cannot open default display
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: SpiRegistry daemon is running with well-known name - org.a11y.atspi.Registry
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: ** (at-spi2-registryd:15416): WARNING **: Could not open X display
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: ** (at-spi2-registryd:15416): WARNING **: AT-SPI: Cannot open default display
Oct 18 01:05:17 Boltzmann gnome-session[15399]: ** (gnome-session-failed:15418): WARNING **: Could not open X display
Oct 18 01:05:17 Boltzmann gnome-session[15399]: Unable to init server: Could not connect: Connection refused
Oct 18 01:05:17 Boltzmann gnome-session[15399]: ** (gnome-session-failed:15418): WARNING **: Cannot open display:
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: SpiRegistry daemon is running with well-known name - org.a11y.atspi.Registry
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: ** (at-spi2-registryd:15420): WARNING **: Could not open X display
Oct 18 01:05:17 Boltzmann org.a11y.atspi.Registry[15410]: ** (at-spi2-registryd:15420): WARNING **: AT-SPI: Cannot open default display

Second test: comment the line 1024 just to see, recompile, and re-launch. Again, a crash, but journalctl simply says:

Oct 18 01:07:25 Boltzmann gnome-session[16991]: gnome-session[16991]: WARNING: Application 'gnome-shell-wayland.desktop' killed by signal 11
Oct 18 01:07:25 Boltzmann gnome-session[16991]: WARNING: Application 'gnome-shell-wayland.desktop' killed by signal 11
Oct 18 01:07:25 Boltzmann gnome-session[16991]: Unrecoverable failure in required component gnome-shell-wayland.desktop

Really looks like there is still some dirty stuff in gnome-session about wayland...

marzoul · 2015-04-17 22:49:47

Now the new Gnome version is released and also out of the testing repos, I tried again.
The issues I mentioned for Gnome 3.14 are solved, Gnome 3.16 starts.
So this topic can be considered closed.

nicolasdanelon · 2015-04-28 18:51:01

I have a related issue with 3.16. My computer restart it self

Apr 28 15:36:55 stark gnome-session[1423]: Conky: '/home/nda/.conkyrc' modified, reloading...
Apr 28 15:36:56 stark gnome-session[1423]: Conky: invalid num arg for top. Must be between 1 and 10.
Apr 28 15:36:56 stark gnome-session[1423]: Conky: invalid num arg for top. Must be between 1 and 10.
Apr 28 15:36:56 stark gnome-session[1423]: Conky: invalid num arg for top. Must be between 1 and 10.
Apr 28 15:36:56 stark gnome-session[1423]: Conky: invalid num arg for top. Must be between 1 and 10.
Apr 28 15:36:56 stark gnome-session[1423]: Conky: invalid num arg for top. Must be between 1 and 10.
Apr 28 15:36:56 stark gnome-session[1423]: Conky: invalid num arg for top. Must be between 1 and 10.
Apr 28 15:36:56 stark gnome-session[1423]: Conky: desktop window (a00016) is subwindow of root window (ac)
Apr 28 15:36:56 stark gnome-session[1423]: Conky: window type - desktop
Apr 28 15:36:56 stark gnome-session[1423]: Conky: drawing to created window (0xc00001)
Apr 28 15:36:56 stark gnome-session[1423]: Conky: drawing to double buffer
Apr 28 15:37:15 stark gnome-session[1423]: **
Apr 28 15:37:15 stark gnome-session[1423]: Clutter:ERROR:clutter-actor.c:1451:clutter_actor_update_map_state: assertion failed: (!(must_be_realized && !may_be_realized))
Apr 28 15:37:15 stark polkitd[332]: Unregistered Authentication Agent for unix-session:c2 (system bus name :1.48, object path /org/freedesktop/PolicyKit1/AuthenticationAgent, locale en_GB.UTF-8) (disconnected from bus)
Apr 28 15:37:15 stark gnome-session[1423]: gnome-session[1423]: WARNING: Application 'gnome-shell.desktop' killed by signal 6
Apr 28 15:37:15 stark gnome-session[1423]: WARNING: Application 'gnome-shell.desktop' killed by signal 6
Apr 28 15:37:15 stark gnome-session[1423]: gnome-session[1423]: GnomeDesktop-WARNING: Failed to acquire idle monitor proxy: GDBus.Error:org.freedesktop.DBus.Error.ServiceUnknown: The name :1.14 was not provided by any .service files
Apr 28 15:37:15 stark gnome-session[1423]: GnomeDesktop-WARNING: Failed to acquire idle monitor proxy: GDBus.Error:org.freedesktop.DBus.Error.ServiceUnknown: The name :1.14 was not provided by any .service files
Apr 28 15:37:15 stark gnome-session[1423]: Gjs-Message: JS LOG: Failed to launch ibus-daemon: Failed to execute child process "ibus-daemon" (No such file or directory)
Apr 28 15:37:15 stark /usr/lib/gdm/gdm-x-session[1413]: The XKEYBOARD keymap compiler (xkbcomp) reports:
Apr 28 15:37:15 stark /usr/lib/gdm/gdm-x-session[1413]: > Warning:          Type "ONE_LEVEL" has 1 levels, but <RALT> has 2 symbols
Apr 28 15:37:15 stark /usr/lib/gdm/gdm-x-session[1413]: >                   Ignoring extra symbols
Apr 28 15:37:15 stark /usr/lib/gdm/gdm-x-session[1413]: Errors from xkbcomp are not fatal to the X server
Apr 28 15:37:15 stark dbus[301]: [system] Activating via systemd: service name='org.freedesktop.GeoClue2' unit='geoclue.service'
Apr 28 15:37:15 stark systemd[1]: Starting Location Lookup Service...
Apr 28 15:37:15 stark dbus[301]: [system] Successfully activated service 'org.freedesktop.GeoClue2'
Apr 28 15:37:15 stark systemd[1]: Started Location Lookup Service.
Apr 28 15:37:15 stark polkitd[332]: Registered Authentication Agent for unix-session:c2 (system bus name :1.138 [/usr/bin/gnome-shell], object path /org/freedesktop/PolicyKit1/AuthenticationAgent, locale en_GB.UTF-8)
Apr 28 15:37:15 stark gnome-session[1423]: Gjs-Message: JS LOG: No permission to trigger offline updates: Polkit.Error: GDBus.Error:org.freedesktop.PolicyKit1.Error.Failed: Action org.freedesktop.packagekit.trigger-offline-update is not registered
Apr 28 15:37:15 stark gnome-session[1423]: (gnome-settings-daemon:1455): GnomeDesktop-WARNING **: Error setting property 'PowerSaveMode' on interface org.gnome.Mutter.DisplayConfig: GDBus.Error:org.freedesktop.DBus.Error.ServiceUnknown: The name org.gnome.Mutter.DisplayConfig was not provided by any .service files (g-dbus-error-quark, 2)
Apr 28 15:37:15 stark systemd-coredump[23166]: Process 1494 (gnome-shell) of user 1000 dumped core.
-- Reboot --

gnome-shell 3.16.1-2
gnome-session 3.16.0-1
lib32-wayland 1.7.0-1
wayland 1.7.0-1
xorg-server-xwayland 1.17.1-5
intel-tbb 4.3_20150209-1
libva-intel-driver 1.5.1-1
xf86-video-intel 2.99.917-5

blackout23 · 2015-04-28 22:42:44

If gnome-shell dumps the core you should install gdb and recover the backtrace.

"$ coredumpctl" will show you all coredumps
"$ coredumpctl gdb" will start gdb on the last coredump
"$ coredumpctl gdb <some PID>" will start gdb on the cordump that matchs the PID in the list of coredumps.

After that just run "bt" in gdb to get the traceback. There will probably be a lot of ??'s in there, because Arch doesn't ship the debug symbols so you could rebuild the packages in question with the debug option to get a better backtrace.

Last edited by blackout23 (2015-04-28 22:44:05)

nicolasdanelon · 2015-04-29 13:23:43

ok... now I'm running the coredumpctl... I fix conkyrc config.. I got 12 top proc argunments.. can this break my gnome ecosystem and crash the complete computer? WTF!?

$ coredumpctl:

TIME                            PID   UID   GID SIG PRESENT EXE
Fri 2015-04-17 11:48:27 ART     366   120   120   5   /usr/bin/gnome-shell
Fri 2015-04-17 11:48:30 ART     631  1000   100   5   /usr/bin/gnome-shell
Fri 2015-04-17 11:48:32 ART     315     0     0  11   /usr/bin/gdm
Fri 2015-04-17 11:50:48 ART   24675   120   120   5   /usr/bin/gnome-shell
Mon 2015-04-20 12:47:56 ART   19824  1000   100  11   /usr/lib/dleyna-renderer/dleyna-renderer-service
Mon 2015-04-20 15:21:22 ART    2872  1000   100  11   /usr/lib/dleyna-renderer/dleyna-renderer-service
Mon 2015-04-20 15:21:22 ART    2854  1000   100  11   /usr/lib/dleyna-server/dleyna-server-service
Mon 2015-04-20 15:21:22 ART     316     0     0  11   /usr/bin/gdm
Mon 2015-04-27 09:04:42 ART     891  1000   100   6 * /usr/bin/gnome-shell
Mon 2015-04-27 09:23:23 ART   20121  1000   100   6 * /usr/bin/gnome-shell
Mon 2015-04-27 11:12:49 ART     928  1000   100   6 * /usr/bin/gnome-shell
Tue 2015-04-28 15:37:15 ART    1494  1000   100   6 * /usr/bin/gnome-shell

$ coredumpctl gdb

           PID: 1494 (gnome-shell)
           UID: 1000 (nda)
           GID: 100 (users)
        Signal: 6 (ABRT)
     Timestamp: Tue 2015-04-28 15:37:15 ART (18h ago)
  Command Line: /usr/bin/gnome-shell
    Executable: /usr/bin/gnome-shell
 Control Group: /user.slice/user-1000.slice/session-c2.scope
          Unit: session-c2.scope
         Slice: user-1000.slice
       Session: c2
     Owner UID: 1000 (nda)
       Boot ID: 1eb332ef75204d2fadcb0d1a0f2dbbbc
    Machine ID: b8cc8df0efc04465b89bfdf1f898b762
      Hostname: stark
      Coredump: /var/lib/systemd/coredump/core.gnome-shell.1000.1eb332ef75204d2fadcb0d1a0f2dbbbc.1494.1430246235000000.lz4
       Message: Process 1494 (gnome-shell) of user 1000 dumped core.

GNU gdb (GDB) 7.9
Copyright (C) 2015 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-unknown-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/bin/gnome-shell...(no debugging symbols found)...done.
[New LWP 1494]
[New LWP 1498]
[New LWP 1502]
[New LWP 1504]
[New LWP 1501]
[New LWP 1499]
[New LWP 1503]

warning: Could not load shared library symbols for linux-vdso.so.1.
Do you need "set solib-search-path" or "set sysroot"?
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
Core was generated by `/usr/bin/gnome-shell'.
Program terminated with signal SIGABRT, Aborted.
#0  0x00007f5694b264b7 in raise () from /usr/lib/libc.so.6

$ coredumpctl gdb 1494

same as coredumpctl gdb

blackout23 · 2015-04-29 16:39:54

You should run "bt" in gdb to get the backtrace.

nicolasdanelon · 2015-04-29 18:26:21

here's the output:

$ coredumpctl gdb

           PID: 1223 (conky)
           UID: 1000 (nda)
           GID: 100 (users)
        Signal: 11 (SEGV)
     Timestamp: Wed 2015-04-29 10:25:19 ART (4h 59min ago)
  Command Line: conky -d
    Executable: /usr/bin/conky
 Control Group: /user.slice/user-1000.slice/session-c2.scope
          Unit: session-c2.scope
         Slice: user-1000.slice
       Session: c2
     Owner UID: 1000 (nda)
       Boot ID: f8f55ddd4b3d485d87c38dfcdcffc922
    Machine ID: b8cc8df0efc04465b89bfdf1f898b762
      Hostname: stark
      Coredump: /var/lib/systemd/coredump/core.conky.1000.f8f55ddd4b3d485d87c38dfcdcffc922.1223.1430313919000000.lz4
       Message: Process 1223 (conky) of user 1000 dumped core.

GNU gdb (GDB) 7.9
Copyright (C) 2015 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <[url]http://gnu.org/licenses/gpl.html[/url]>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-unknown-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<[url]http://www.gnu.org/software/gdb/bugs/[/url]>.
Find the GDB manual and other documentation resources online at:
<[url]http://www.gnu.org/software/gdb/documentation/[/url]>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/bin/conky...(no debugging symbols found)...done.
[New LWP 1223]
[New LWP 1242]
[New LWP 1241]
[New LWP 1240]
[New LWP 1244]
[New LWP 1246]
[New LWP 1245]
[New LWP 1243]

warning: Could not load shared library symbols for linux-vdso.so.1.
Do you need "set solib-search-path" or "set sysroot"?
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
Core was generated by `conky -d'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x00007fe924a10ca2 in vfprintf () from /usr/lib/libc.so.6
(gdb) bt
#0  0x00007fe924a10ca2 in vfprintf () from /usr/lib/libc.so.6
#1  0x00007fe924a38be9 in vsnprintf () from /usr/lib/libc.so.6
#2  0x00007fe924a171d2 in snprintf () from /usr/lib/libc.so.6
#3  0x0000000000443c42 in ?? ()
#4  0x0000000000424619 in ?? ()
#5  0x0000000000425cf5 in ?? ()
#6  0x0000000000407c38 in ?? ()
#7  0x00007fe9249e8800 in __libc_start_main () from /usr/lib/libc.so.6
#8  0x0000000000407f39 in ?? ()

Last edited by nicolasdanelon (2015-04-29 18:28:11)

blackout23 · 2015-04-29 21:38:23

Looks like the problem happens in the gnome-shell. So you need to rebuild that package with the debug option so that the backtrace shows more useful information instead of ??. This is described here:
https://wiki.archlinux.org/index.php/De … ing_Traces

Arch Linux

#1 2014-10-17 08:50:25

Gnome 3.14 problems with wayland

#2 2014-10-17 18:53:31

Re: Gnome 3.14 problems with wayland

#3 2014-10-17 20:23:34

Re: Gnome 3.14 problems with wayland

#4 2014-10-17 23:31:31

Re: Gnome 3.14 problems with wayland

#5 2015-04-17 22:49:47

Re: Gnome 3.14 problems with wayland

#6 2015-04-28 18:51:01

Re: Gnome 3.14 problems with wayland

#7 2015-04-28 22:42:44

Re: Gnome 3.14 problems with wayland

#8 2015-04-29 13:23:43

Re: Gnome 3.14 problems with wayland

#9 2015-04-29 16:39:54

Re: Gnome 3.14 problems with wayland

#10 2015-04-29 18:26:21

Re: Gnome 3.14 problems with wayland

#11 2015-04-29 21:38:23

Re: Gnome 3.14 problems with wayland

Board footer