Error retrieving accessibility bus address: org.freedesktop.DBus.Error.NoReply: Did not receive a reply.

Bug #1217757 reported by dino99 on 2013-08-28
70
This bug affects 13 people
Affects Status Importance Assigned to Milestone
at-spi2-core (Ubuntu)
High
Luke Yelavich

Bug Description

Stock Saucy + gnome3-team ppa

ProblemType: Bug
DistroRelease: Ubuntu 13.10
Package: gedit 3.8.3-0ubuntu3
ProcVersionSignature: Ubuntu 3.11.0-4.9-generic 3.11.0-rc7
Uname: Linux 3.11.0-4-generic i686
NonfreeKernelModules: nvidia
ApportVersion: 2.12.1-0ubuntu2
Architecture: i386
Date: Wed Aug 28 10:24:40 2013
ExecutablePath: /usr/bin/gedit
MarkForUpload: True
SourcePackage: gedit
UpgradeStatus: No upgrade log present (probably fresh install)

dino99 (9d9) wrote :
dino99 (9d9) wrote :

oem@oem-desktop:~$ sudo gedit /etc/hosts
[sudo] password for oem:

** (gedit:4126): WARNING **: Error retrieving accessibility bus address: org.freedesktop.DBus.Error.NoReply: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken.

(gedit:4126): IBUS-WARNING **: The owner of /home/oem/.config/ibus/bus is not root!

(gedit:4126): Gtk-WARNING **: Calling Inhibit failed: GDBus.Error:org.freedesktop.DBus.Error.ServiceUnknown: The name org.gnome.SessionManager was not provided by any .service files

(gedit:4126): Gtk-WARNING **: Calling Inhibit failed: GDBus.Error:org.freedesktop.DBus.Error.ServiceUnknown: The name org.gnome.SessionManager was not provided by any .service files

That issue also affect all the app's loading : they are delayed by 5 to 10 seconds

Sebastien Bacher (seb128) wrote :

Thank you for your bug report

> dbus 1.6.12-0ubuntu4

you are using saucy-proposed it seems, you shouldn't (proposed is not meant for users but for automatic testing)

In any case that seems more likely an issue with the new dbus there than with gedit

affects: gedit (Ubuntu) → dbus (Ubuntu)
Changed in dbus (Ubuntu):
assignee: nobody → Tyler Hicks (tyhicks)
importance: Undecided → High
Sebastien Bacher (seb128) wrote :

@Tyler: the accessibility stack uses its own private bus and not the session one, could that create issue with the confinement changes?

Tyler Hicks (tyhicks) wrote :

That's possible. I'll be sure to test this configuration before reuploading dbus w/ apparmor mediation.

Tyler Hicks (tyhicks) wrote :

dino99 - Can you paste the output of `grep DENIED /var/log/syslog` after seeing these errors?

dino99 (9d9) wrote :

Thanks Sebastien,

i've downgrade dbus to 1.6.12.0ubuntu2 (main archive), but still get the same errors.

dino99 (9d9) wrote :

@Tyler

here is the DENIED output

Tyler Hicks (tyhicks) wrote :

While there are some apparmor denial messages in your syslog, none of them are related to this bug.

Did you reboot after downgrading dbus to 1.6.12.0ubuntu2 and reproducing the bug? Restarting your session may be enough, but a reboot would be best.

Changed in dbus (Ubuntu):
assignee: Tyler Hicks (tyhicks) → nobody
dino99 (9d9) wrote :

To resume:
- downgraded dbus to main archive (dbus / dbus-x11 / libdbus-1-3)
- logout/in with gnome-shell
- only get that error when running : sudo gedit ...

(gedit:1144): IBUS-WARNING **: The owner of /home/oem/.config/ibus/bus is not root!

the other errors are gone, and loading apps are no more delayed.

Tyler Hicks (tyhicks) wrote :

Ok, so the other errors were potentially caused by my dbus package. I'll look into this more tomorrow.

Changed in dbus (Ubuntu):
assignee: nobody → Tyler Hicks (tyhicks)
Tyler Hicks (tyhicks) wrote :

I was able to reproduce this in a saucy amd64 virtual machine

Changed in dbus (Ubuntu):
status: New → Confirmed
Tyler Hicks (tyhicks) wrote :

This bug is very finicky to reproduce. I've been testing with completely up to date and pristine Saucy virtual machines on both amd64 and i386, *without* the AppArmor mediation dbus package (so I'm testing dbus 1.6.12.0ubuntu2).

These are throw away VMs that I create and trash as needed. I'll create one and it will never hit this bug, hit this bug on some boots, or hit this bug on every boot. I haven't been able to see pattern.

However, I currently have an i386 VM that is hitting this bug on every boot. This is the contents of ~/.xsession-errors:

$ cat ~/.xsession-errors
Script for cjkv started at run_im.
Script for default started at run_im.
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd main process ended, respawning
init: at-spi2-registryd respawning too fast, stopped

I don't know much about at-spi2, but I noticed that this dbus method call hangs until a timeout is reached:

$ sudo dbus-send --session --print-reply --dest=org.a11y.Bus /org/a11y/bus org.a11y.Bus.GetAddress
Error org.freedesktop.DBus.Error.NoReply: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken.
Command exited with non-zero status 1
real 25.02
user 0.00
sys 0.00

There are no AppArmor denials in the syslog and dbus dbus 1.6.12-0ubuntu2 is running, so I think it is completely safe to say that AppArmor and AppArmor D-Bus mediation is not at fault here.

Changed in at-spi2-core (Ubuntu):
importance: Undecided → High
Sebastien Bacher (seb128) wrote :

Luke, could you look at that? The bug is recent, could be due to one of the updates you did this week?

Changed in at-spi2-core (Ubuntu):
assignee: nobody → Luke Yelavich (themuso)
Tyler Hicks (tyhicks) wrote :

Here's another data point from another VM:

* dbus 1.6.12-0ubuntu2
* at-spi2-core 2.9.90-0ubuntu1
* ~/.xsession-errors is clean
* at-spi2-core is running:
  $ ps aux | grep at-spi2
  tyhicks 1444 0.0 0.3 337552 3384 ? Ssl 15:14 0:00 /usr/lib/at-spi2-core/at-spi-bus-launcher --launch-immediately
  tyhicks 1450 0.0 0.1 28532 1792 ? S 15:14 0:00 /bin/dbus-daemon --config-file=/etc/at-spi2/accessibility.conf --nofork --print-address 3
  tyhicks 1458 0.0 0.3 124992 3328 ? Sl 15:14 0:00 /usr/lib/at-spi2-core/at-spi2-registryd --use-gnome-session
* `sudo dbus-send --session --print-reply --dest=org.a11y.Bus /org/a11y/bus org.a11y.Bus.GetAddress` still times out
* /var/log/lightdm/x-0-greeter.log contains the following messages:
  Successfully activated service 'org.a11y.atspi.Registry'

  ** (at-spi2-registryd:1114): WARNING **: Failed to register client: GDBus.Error:org.freedesktop.DBus.Error.ServiceUnknown: The name org.gnome.SessionManager was not provided by any .service files

  ** (at-spi2-registryd:1114): WARNING **: Unable to register client with session manager

Harry (harry33) wrote :

I have experienced this kind of slowing down when opening Synaptic.
Opening Synaptic is delayed about 10 sec. It will open finally, however.
I traced the bug to the package at-spi2-core (libatspi2.0-0 version 2.9.90-0ubuntu1)
and possibly also to at-spi2-atk (libatk-bridge2.0-0 version 2.9.90-0ubuntu1).

After downgrading the packages back to the previous ones (versions 2.9.5-1ubuntu1 and 2.9.5-1) all is well and Synaptic opens immediately without any delay.

I use the old d-bus (version 1.6.12-0ubuntu2) for the time being.
I haven't tried the latest version 1.6.12-0ubuntu5 depending on libapparmor1 yet.

I have a 64-bit setup with only Gnome-Shell DE.

dino99 (9d9) wrote :

Feedback after the upgraded dbus 1.6.12.0ubuntu5 and a lougout/in with gnome-shell on i386:

get the same #13 output :
 init: at-spi2-registryd respawning too fast, stopped

and :
 sudo dbus-send --session --print-reply --dest=org.a11y.Bus /org/a11y/bus org.a11y.Bus.GetAddress

ends up with a time out:
Error org.freedesktop.DBus.Error.NoReply: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken.

( its really a time out, because after validating the requested password, that command had seemed hanged for a few seconds before sending the error)

dino99 (9d9) wrote :

Also confirm the atspi 2.9.90 problem related in #16 above:

- some apps are largely delayed: for example, synaptic , firefox, midori
- while some others like chromium-browser are not

Launchpad Janitor (janitor) wrote :

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

Changed in at-spi2-core (Ubuntu):
status: New → Confirmed
Tyler Hicks (tyhicks) on 2013-08-29
Changed in dbus (Ubuntu):
assignee: Tyler Hicks (tyhicks) → nobody
Doug McMahon (mc3man) wrote :

The combo of latest dbus (1.6.12-0ubuntu5) & latest at-spi2-atk (2.9.90-0ubuntu1 ), specifically at-spi2-core, causes the slowdowns mentioned here plus some additional
gdebi
opening some apps from a root terminal prompt (sudo -i
opening gksu.desktop & using it

The old at-spi2-atk (2.9.5-1) with new dbus (1.6.12-0ubuntu5) is fine
The new at-spi2-atk (2.9.90-0ubuntu1 with old dbus (1.6.12-0ubuntu2) is also fine here in practice (non withstanding any logs

I am now running the latest packages, including the new dbus, and accessibility in general is not broken, but I can reproduce this problem. Whats more, I saw a thread about this on an upstrea a11y mailing list the other day, although I hadn't given it any thought till now.

The bug is not present in at-spi2-core 2.9.5, so I am in the process of bisecting the git tree to find the commit that causes this bug, and will act accordingly. I will probably revert the commit for now, and work with upstream to sort out a proper fix.

Thanks for the report.

Launchpad Janitor (janitor) wrote :

This bug was fixed in the package at-spi2-core - 2.9.90-0ubuntu2

---------------
at-spi2-core (2.9.90-0ubuntu2) saucy; urgency=low

  * debian/patches/revert_x11_code_separation.patch: Revert upstream git
    commit 7ec6f416a192b7ff86d4b90fe7f405de41165342, which was the
    beginning of separating out the X specific code. It seems this is
    breaking accessibility when used accross user boundaries, i.e sudo.
    (LP: #1217757)
  * Use dh-autoreconf for the above.
 -- Luke Yelavich <email address hidden> Fri, 30 Aug 2013 06:52:15 +1000

Changed in at-spi2-core (Ubuntu):
status: Confirmed → Fix Released
dino99 (9d9) wrote :

Feedback about atspi upgrade:
- no error logged
- midori & firefox load is still delayed (10 / 15 seconds)

Did you test to see whether you get the slower load times with firefox/midori if you used at-spi2-core 2.9.5?

dino99 (9d9) wrote :

Have found a workaround to the slowness load: build a fresh new .config folder

Pander (pander) wrote :

When starting for example gedit, gnome-character-map or any other GNOME applications in lxterminal on a fresh system with Lubuntu 17.10, the same warnings are printed:

WARNING **: Error retrieving accessibility bus address: org.freedesktop.DBus.Error.ServiceUnknown: The name org.a11y.Bus was not provided by any .service files

If this should be a separate bug, please let me know.

tags: added: artful lubuntu
dino99 (9d9) wrote :

That report is old, and the upstream version got the fix long ago.
Does not know about Lubuntu, but your system has been dist-upgraded many times, then clean it with gtkorphan & bleachbit (as root). That error should not be met nowadays.

no longer affects: dbus (Ubuntu)
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Duplicates of this bug

Other bug subscribers