Bug 4708 - failure to register before timeout
Summary: failure to register before timeout
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: Cauldron
Hardware: i586 Linux
Priority: Normal normal
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL: https://bugzilla.gnome.org/show_bug.c...
Whiteboard:
Keywords: UPSTREAM
Depends on:
Blocks:
 
Reported: 2012-02-26 22:12 CET by Guillaume Rousse
Modified: 2012-08-07 21:08 CEST (History)
2 users (show)

See Also:
Source RPM: gnome-settings-daemon
CVE:
Status comment:


Attachments

Description Guillaume Rousse 2012-02-26 22:12:31 CET
Since latest update of gnome-settings-daemon (3.3.90.1-3.mga2 currently), my gnome session hangs until the fatal error screen: something wrong happened, please try again, ... Killing this process allow to run the session, but with a blank background image.

In system logs, and also in ~/.xession-errors,there is error message:
gnome-session[2404]: WARNING: Application 'gnome-settings-daemon.desktop' failed to register before timeout
Comment 1 Guillaume Rousse 2012-02-26 22:13:47 CET
I also find one of the error signaled in bug #4667:
Avertissement du gestionnaire de fenêtres : Log level 16: Unable to register authentication agent: GDBus.Error:org.freedesktop.PolicyKit1.Error.Failed: An authentication agent already exists for the given subject

I'm using systemd, BTW.
Comment 2 Colin Guthrie 2012-02-26 22:24:34 CET
Hmmm.... I think I remember this bug from a while ago with e.g. gdm.

The failed to register thing is basically caused by a missing build dep... lemme take a look.

CC: (none) => mageia

Comment 3 Colin Guthrie 2012-02-26 22:36:30 CET
Scratch that. I thought it was maybe missing a libsm require, but I don't see any linkage against it, optional or otherwise in the code.

I guess it's just a consolekit vs. systemd user session thing?
Comment 4 Olav Vitters 2012-02-26 23:45:56 CET
I didn't enable any systemd things. Perhaps we need to explicitly disable it to be really sure it doesn't get enabled by accident?

Bit late today, I want to see if I can reproduce locally. Starting a session is ok, big problem is the memory usage of mutter/gnome-shell (eventually using 1.2GB).

CC: (none) => olav

Comment 5 Colin Guthrie 2012-02-27 00:27:08 CET
FWIW, things seem to work OK for me too. I can even see gnome-settings-daemon listed in dbus introspection of org.gnome.SessionManager (I have it as Client 4).

What does systemctl status console-kit-daemon.service say?

(I see the g-s memory hog bug too - urk!)
Comment 6 Guillaume Rousse 2012-02-27 09:36:09 CET
OK, I found the issue by comparing warnings in ~/.xsession-errors with a new user (which had no problems): I was missing at-spi2-atk package. I don't know how it was uninstalled, and it's also quite painful than missing an optional accessibility stuff leads to such problems.

Status: NEW => RESOLVED
Resolution: (none) => FIXED

Comment 7 Colin Guthrie 2012-02-27 09:40:23 CET
Strange. I do not have that package installed here either... will keep my eyes open at the next reboot!
Comment 8 Olav Vitters 2012-02-27 09:42:13 CET
Maybe at-spi2-atk is a suggests and it should be a require? If it happens to you, might happen to others. Checking, it seems I do not have it installed (bit weird).
Comment 9 Colin Guthrie 2012-02-27 12:06:23 CET
Just did a reboot. Can't say I'm missing the at-spi2-atk package. Everything seems fine without it....

I do get the same error you noticed:

Window manager warning: Log level 16: Unable to register authentication agent: GDBus.Error:org.freedesktop.PolicyKit1.Error.Failed: An authentication agent already exists for the given subject
Window manager warning: Log level 16: Error registering polkit authentication agent: GDBus.Error:org.freedesktop.PolicyKit1.Error.Failed: An authentication agent already exists for the given subject (polkit-error-quark 0)

But it seems harmless enough... :s
Comment 10 Guillaume Rousse 2012-02-27 22:44:37 CET
Unfortunatly, the issue is back, with a more frightening error messages:

** CRITICAL **: DBus AddMatch for dconf path '/org/gnome/settings-daemon/plugins/housekeeping/': La connexion est fermée

GLib-GIO-CRITICAL **: g_dbus_connection_signal_subscribe: assertion `G_IS_DBUS_CONNECTION (connection)' failed

** WARNING **: Could not connect: Connexion refusée

GLib-GIO-CRITICAL **: Error while sending AddMatch() message: La connexion est fermée

g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream returned 0 bytes on an async read (g-io-error-quark, 0). Exiting.

Status: RESOLVED => REOPENED
Resolution: FIXED => (none)

Comment 11 Olav Vitters 2012-02-27 23:22:27 CET
And I don't have any issues (aside from gdm not starting in the proper mode).
Comment 12 Guillaume Rousse 2012-03-01 18:59:41 CET
And the culprit is... cups trying to contact the cups server declared in ~/.cups/client.conf. If the network is not available, or if there is no such server declared, everything is OK.
Comment 13 Guillaume Rousse 2012-03-27 19:58:03 CEST
The problem still exists...
Comment 14 Olav Vitters 2012-03-28 10:21:34 CEST
Could you file a bug upstream maybe? This is a bit too detailed for me..
Comment 15 Colin Guthrie 2012-03-28 10:59:55 CEST
To reproduce would all I need to do is put some invalid data in ~/.cups/client.conf?
Comment 16 Guillaume Rousse 2012-03-28 12:28:00 CEST
colin: in theory, yes.
Comment 17 Marja Van Waes 2012-05-26 13:05:50 CEST
Hi,

This bug was filed against cauldron, but we do not have cauldron at the moment.

Please report whether this bug is still valid for Mageia 2.

Thanks :)

Cheers,
marja

Keywords: (none) => NEEDINFO

Comment 18 Guillaume Rousse 2012-05-30 12:55:20 CEST
Bug reported upstream: https://bugzilla.gnome.org/show_bug.cgi?id=677081

Keywords: NEEDINFO => UPSTREAM
URL: (none) => https://bugzilla.gnome.org/show_bug.cgi?id=677081

Comment 19 Guillaume Rousse 2012-08-07 21:08:50 CEST
This was fixed upstream, closing.

Status: REOPENED => RESOLVED
Resolution: (none) => FIXED


Note You need to log in before you can comment on or make changes to this bug.