Welcome screen clock missing

[Gema Gomez]

Today I have noticed the welcome screen clock missing.

The only thing have done that is time related is changing the timezone to london on the system settings.

ProblemType: Bug
DistroRelease: Ubuntu 13.10
Package: unity8 7.82+13.10.20131011.2-0ubuntu1
Uname: Linux 3.4.0-3-mako armv7l
ApportVersion: 2.12.5-0ubuntu2
Architecture: armhf
Date: Mon Oct 14 14:50:07 2013
InstallationDate: Installed on 2013-10-14 (0 days ago)
InstallationMedia: Ubuntu 13.10 - armhf (20131014)
MarkForUpload: True
ProcEnviron:
 TERM=linux
 PATH=(custom, no user)
SourcePackage: unity8
UpgradeStatus: No upgrade log present (probably fresh install)

[Nick Dedekind (nick-dedekind)]

Do you know if the datetime indicator was showing in the Panel?

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Alan Pope 🍺🐧🐱 🦄 (popey)]

I just updated my mako from 97 to 98. When it first booted there was no clock in the indicator but there was a clock on the welcome screen.

I unlocked and faffed about a little bit, and then when the screen locked (after 2 or 3 times of unlocking and playing) the time was gone from the welcome screen.

To clarify, on first boot there was NO indicator clock, but there WAS a clock on the welcome screen.

See screenshot.

[Vincent Ladeuil (vila)]

Seen the missing date/time on welcome screen and the indicator was missing at the same time. That was with 97. Not seen with 98 so far. No change in the time zone. Encountered a freeze later and had to reboot. Date/time was back in both places.

[Nick Dedekind (nick-dedekind)]

I've finally been able to randomly reproduce this with some effort.
Following is from upstart logs.

phablet@ubuntu-phablet:~$ cat ~/.cache/upstart/dbus.log | grep indicator
Activating service name='com.canonical.indicator.bluetooth'
Activating service name='com.canonical.indicator.datetime'
Activating service name='com.canonical.indicator.location'
Activating service name='com.canonical.indicator.messages'
Activating service name='com.canonical.indicator.power'
Activating service name='com.canonical.indicator.sound'
Successfully activated service 'com.canonical.indicator.bluetooth'
Successfully activated service 'com.canonical.indicator.messages'
Successfully activated service 'com.canonical.indicator.power'
Successfully activated service 'com.canonical.indicator.location'
Successfully activated service 'com.canonical.indicator.sound'
Failed to activate service 'com.canonical.indicator.datetime': timed out // This takes a while to appear.

Added some debugging to unitymenumodel to log out dbus name watching:
INDICATORS @ "15:55:36" : nameAppeared => "com.canonical.indicator.network"
INDICATORS @ "15:55:36" : nameAppeared => "com.canonical.indicator.bluetooth"
INDICATORS @ "15:55:36" : nameAppeared => "com.canonical.indicator.messages"
INDICATORS @ "15:55:36" : nameAppeared => "com.canonical.indicator.power"
INDICATORS @ "15:55:36" : nameAppeared => "com.canonical.indicator.location"
INDICATORS @ "15:55:36" : nameAppeared => "com.canonical.indicator.sound"
INDICATORS @ "15:55:59" : nameVanished => "com.canonical.indicator.datetime"

I've attached the dbus upstart log.

[Nick Dedekind (nick-dedekind)]

Attached to the indicator-datetime process while it was in the "failed" state (seems to abort after awhile sometimes).
Backtrace:

#0 0x40067fe4 in __libc_do_syscall () from /lib/arm-linux-gnueabihf/libpthread.so.0
#1 0x40065c40 in __lll_lock_wait () from /lib/arm-linux-gnueabihf/libpthread.so.0
#2 0x400625e0 in pthread_mutex_lock () from /lib/arm-linux-gnueabihf/libpthread.so.0
#3 0x401aef8e in g_type_add_interface_static () from /usr/lib/arm-linux-gnueabihf/libgobject-2.0.so.0
#4 0x402e2032 in g_dbus_proxy_get_type () from /usr/lib/arm-linux-gnueabihf/libgio-2.0.so.0
#5 0x402e2fb2 in g_dbus_proxy_new_for_bus () from /usr/lib/arm-linux-gnueabihf/libgio-2.0.so.0
#6 0x0000ea30 in ?? ()
#7 0x0000ea30 in ?? ()

I've been able to reproduce this trace by continually restarting the indicator-datetime-service process

[Lars Karlitski (larsu)]

Looks like this is the same glib deadlock bug as in bug #1239710.