GNOME Bugzilla – Bug 780965
control-center hangs after clicking "connect to hidden network" and Wi-Fi gear button
Last modified: 2018-05-07 08:48:08 UTC
Already connected to a wifi network Steps to reproduce: 1. click the gear button on the wifi that has been connected 2. close the details dialog popup 3. click "connect to hidden network" button The following steps can also reproduce this bug: 1. click "connect to hidden network" button 2. close the dialog popped up 3. click the gear button on the wifi that has been connected Then the following errors show up in the command line: (gnome-control-center:9016): GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed (gnome-control-center:9016): GLib-CRITICAL **: g_once_init_leave: assertion 'result != 0' failed (gnome-control-center:9016): Gtk-WARNING **: /home/jonathan/jhbuild/checkout/gtk+-3/gtk/gtkliststore.c:434: Invalid type (null) (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_append: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_set_valist: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_append: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_set_valist: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_append: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_set_valist: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_append: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_set_valist: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed (gnome-control-center:9016): GLib-CRITICAL **: g_once_init_leave: assertion 'result != 0' failed (gnome-control-center:9016): Gtk-WARNING **: /home/jonathan/jhbuild/checkout/gtk+-3/gtk/gtkliststore.c:434: Invalid type (null) (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_append: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_set_valist: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_append: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_set_valist: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_append: assertion 'GTK_IS_LIST_STORE (list_store)' failed (gnome-control-center:9016): Gtk-CRITICAL **: gtk_list_store_set_valist: assertion 'GTK_IS_LIST_STORE (list_store)' failed And the program hangs. By using gdb, I get the following backtrace:
+ Trace 237335
This issue seems to be introduced by commit 0470c040b1e6f5f2b1eba852e15365579ce1dba5.
Works fine here. You'll want to check your downstream patches. This means that the type was already registered: " (gnome-control-center:9016): GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed " Check the backtrace of this warning, and find out what type is doubly registered.
(In reply to Bastien Nocera from comment #2) > Works fine here. You'll want to check your downstream patches. This issue happens in terms of the code base in master branch. So there are no downstream patches.
(In reply to Bastien Nocera from comment #2) > Works fine here. You'll want to check your downstream patches. > > This means that the type was already registered: > " > (gnome-control-center:9016): GLib-GObject-CRITICAL **: > g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed > " > > Check the backtrace of this warning, and find out what type is doubly > registered. EAPMethod is registered twice. See the backtrace below: Thread 1 "gnome-control-c" hit Breakpoint 1, g_boxed_type_register_static ( name=0x7f55b761b070 "EAPMethod", boxed_copy=0x55618d <eap_method_ref>, boxed_free=0x5561fd <eap_method_unref>) at /home/jonathan/jhbuild/checkout/glib/gobject/gboxed.c:294 294 GTypeInfo type_info = { (gdb) bt
+ Trace 237341
I'm using network-manager-applet and libnma 1.8.0, can you check whether this happens with this version, or are you using network-manager-applet from git master as well?
(In reply to Bastien Nocera from comment #5) > I'm using network-manager-applet and libnma 1.8.0, can you check whether > this happens with this version, or are you using network-manager-applet from > git master as well? I'm on the master branch of network-manager-applet, which is version 1.7.1.
Reverting commit 0470c040b1e6f5f2b1eba852e15365579ce1dba5 fixes the problem. Is it because EAPMethod has already been registered in network-manager-applet? I found that in commit 0470c040b1e6f5f2b1eba852e15365579ce1dba5, CcEAPMethod was renamed to EAPMethod, which is the same name as the type in network-manager-applet.
This issue still exist with the latest code in master branch.
*** This bug has been marked as a duplicate of bug 785099 ***