Ubuntu 20.04.3 LTS - power off button - problem turn off

I have 20.04.3 upgraded from 20.04.1 and problem still exists:

When i click “power off” > “power off” > and confirm, after pressing the button the shutdown procedure don’t start - i can wait and nothing. Next i click again power off and confirm and later it works and shutdown correctly.

The problem does not occur often, rather 2-3 times (out of 7) a week.

/var/log/syslog logs from example problem:


Sep  2 12:40:42 pw systemd[1]: Finished Ubuntu Advantage APT and MOTD Messages.
Sep  2 12:42:25 pw kernel: [ 2921.326768] pcieport 0000:00:1c.4: AER: Corrected error received: 0000:00:1c.4
Sep  2 12:42:25 pw kernel: [ 2921.326781] pcieport 0000:00:1c.4: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
Sep  2 12:42:25 pw kernel: [ 2921.326783] pcieport 0000:00:1c.4:   device [8086:9d14] error status/mask=00000001/00002000
Sep  2 12:42:25 pw kernel: [ 2921.326784] pcieport 0000:00:1c.4:    [ 0] RxErr                 
Sep  2 12:42:31 pw kernel: [ 2927.324405] pcieport 0000:00:1c.4: AER: Corrected error received: 0000:00:1c.4
Sep  2 12:42:31 pw kernel: [ 2927.324415] pcieport 0000:00:1c.4: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
Sep  2 12:42:31 pw kernel: [ 2927.324417] pcieport 0000:00:1c.4:   device [8086:9d14] error status/mask=00000001/00002000
Sep  2 12:42:31 pw kernel: [ 2927.324419] pcieport 0000:00:1c.4:    [ 0] RxErr

Hi Softwareman,

Welcome to our forums.

A strange issue you got there. I assume a shutdown -h now from the console would work every time, right? If so, the problem should be somewhere in the GUI layer. Does the console way work?

yes console work correctly, but the GUI doesn’t :confused:

I searched around a bit about the error messages you got, and it seems those aren’t relevant to the issue at hand.

Do you use the default Gnome desktop, or something else? Did you do any changes to the desktop environment? I’m thinking of advanced settings, not new background or shortcuts.

No, i didn’t make any specific changes, I read somewhere that new kernel updates should to resolve this problem but it didn’t.

I would look around in the logs related to Gnome with timestamp that matches the time when you try and not able to turn off the machine with the GUI. You should find a hidden file named .xsession-errors in your home directory, and you could also look into the gdm related logs under /var/log, and check the journal.

  1. i dont have .xsession-errors in my home directory
  2. i checked journal log and there are the same error logs:
   wrz 02 12:40:42 pw systemd[1]: Finished Ubuntu Advantage APT and MOTD Messages.
wrz 02 12:42:25 pw kernel: pcieport 0000:00:1c.4: AER: Corrected error received: 0000:00:1c.4
wrz 02 12:42:25 pw kernel: pcieport 0000:00:1c.4: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
wrz 02 12:42:25 pw kernel: pcieport 0000:00:1c.4:   device [8086:9d14] error status/mask=00000001/00002000
wrz 02 12:42:25 pw kernel: pcieport 0000:00:1c.4:    [ 0] RxErr                 
wrz 02 12:42:31 pw kernel: pcieport 0000:00:1c.4: AER: Corrected error received: 0000:00:1c.4
wrz 02 12:42:31 pw kernel: pcieport 0000:00:1c.4: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
wrz 02 12:42:31 pw kernel: pcieport 0000:00:1c.4:   device [8086:9d14] error status/mask=00000001/00002000
wrz 02 12:42:31 pw kernel: pcieport 0000:00:1c.4:    [ 0] RxErr                 
wrz 02 12:42:43 pw kernel: pcieport 0000:00:1c.4: AER: Corrected error received: 0000:00:1c.4
wrz 02 12:42:43 pw kernel: pcieport 0000:00:1c.4: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
wrz 02 12:42:43 pw kernel: pcieport 0000:00:1c.4:   device [8086:9d14] error status/mask=00000001/00002000
wrz 02 12:42:43 pw kernel: pcieport 0000:00:1c.4:    [ 0] RxErr                 
wrz 02 12:42:55 pw kernel: pcieport 0000:00:1c.4: AER: Corrected error received: 0000:00:1c.4
wrz 02 12:42:55 pw kernel: pcieport 0000:00:1c.4: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
wrz 02 12:42:55 pw kernel: pcieport 0000:00:1c.4:   device [8086:9d14] error status/mask=00000001/00002000
wrz 02 12:42:55 pw kernel: pcieport 0000:00:1c.4:    [ 0] RxErr                 
wrz 02 12:43:01 pw kernel: pcieport 0000:00:1c.4: AER: Corrected error received: 0000:00:1c.4
wrz 02 12:43:01 pw kernel: pcieport 0000:00:1c.4: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
wrz 02 12:43:01 pw kernel: pcieport 0000:00:1c.4:   device [8086:9d14] error status/mask=00000001/00002000
wrz 02 12:43:01 pw kernel: pcieport 0000:00:1c.4:    [ 0] RxErr                 
wrz 02 12:43:04 pw gnome-shell[2833]: ../clutter/clutter/clutter-actor.c:10558: The clutter_actor_set_allocation() function can only be called from within the implementation of the ClutterActor::allocate>
wrz 02 12:43:07 pw dbus-daemon[1290]: [system] Activating via systemd: service name='org.freedesktop.PackageKit' unit='packagekit.service' requested by ':1.85' (uid=1000 pid=2833 comm="/usr/bin/gnome-she>
wrz 02 12:43:07 pw systemd[1]: Starting PackageKit Daemon...
wrz 02 12:43:07 pw PackageKit[6232]: daemon start
wrz 02 12:43:07 pw dbus-daemon[1290]: [system] Successfully activated service 'org.freedesktop.PackageKit'
wrz 02 12:43:07 pw systemd[1]: Started PackageKit Daemon.
wrz 02 12:43:07 pw gnome-shell[2833]: endSessionDialog: No XDG_SESSION_ID, fetched from logind: 2
wrz 02 12:43:08 pw gnome-session[2819]: gnome-session-binary[2819]: WARNING: Client '/org/gnome/SessionManager/Client25' failed to reply before timeout
wrz 02 12:43:08 pw gnome-session-binary[2819]: WARNING: Client '/org/gnome/SessionManager/Client25' failed to reply before timeout
wrz 02 12:43:08 pw gnome-shell[2833]: endSessionDialog: No XDG_SESSION_ID, fetched from logind: 2
wrz 02 12:43:10 pw systemd[1]: unattended-upgrades.service: Succeeded.
wrz 02 12:43:19 pw systemd[2560]: Started Application launched by gnome-shell.
wrz 02 12:43:19 pw dbus-daemon[2571]: [session uid=1000 pid=2571] Activating via systemd: service name='org.gnome.Terminal' unit='gnome-terminal-server.service' requested by ':1.214' (uid=1000 pid=6243 c>
wrz 02 12:43:19 pw systemd[2560]: Created slice apps.slice.
wrz 02 12:43:19 pw systemd[2560]: Created slice apps-org.gnome.Terminal.slice.
wrz 02 12:43:19 pw systemd[2560]: Starting GNOME Terminal Server...
wrz 02 12:43:19 pw dbus-daemon[2571]: [session uid=1000 pid=2571] Successfully activated service 'org.gnome.Terminal'
wrz 02 12:43:19 pw systemd[2560]: Started GNOME Terminal Server.
wrz 02 12:43:19 pw at-spi-bus-launcher[2807]: dbus-daemon[2807]: Activating service name='org.a11y.atspi.Registry' requested by ':1.23' (uid=1000 pid=6246 comm="/usr/libexec/gnome-terminal-server " label>
wrz 02 12:43:19 pw at-spi-bus-launcher[2807]: dbus-daemon[2807]: Successfully activated service 'org.a11y.atspi.Registry'
wrz 02 12:43:19 pw at-spi-bus-launcher[6251]: SpiRegistry daemon is running with well-known name - org.a11y.atspi.Registry
wrz 02 12:43:20 pw systemd[2560]: Started VTE child process 6258 launched by gnome-terminal-server process 6246.
wrz 02 12:43:20 pw systemd[2560]: gnome-launched-org.gnome.Terminal.desktop-6242.scope: Succeeded.
wrz 02 12:43:31 pw kernel: pcieport 0000:00:1c.4: AER: Corrected error received: 0000:00:1c.4
wrz 02 12:43:31 pw kernel: pcieport 0000:00:1c.4: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
wrz 02 12:43:31 pw kernel: pcieport 0000:00:1c.4:   device [8086:9d14] error status/mask=00000001/00002000
wrz 02 12:43:31 pw kernel: pcieport 0000:00:1c.4:    [ 0] RxErr                 
wrz 02 12:43:41 pw systemd-logind[1325]: Delay lock is active (UID 1000/pawel, PID 2819/gnome-session-b) but inhibitor timeout is reached.
wrz 02 12:43:41 pw systemd-logind[1325]: System is powering down.

This message suggests the system goes to normal shutdown. Did you turn off your machine from the terminal, or the GUI power off was working in this case?

Tough call.

I can’t be sure, but maybe your PCI Express (PCIe) device can’t handle power management requests, yet the kernel is by default configured to enable PCIe power management support, hence the conflict between what the kernel “wants” and what the PCIe device can actually provide or do.

You may want to edit the GRUB default settings file with a text editor such as e.g. Nano:

sudo nano /etc/default/grub

…or Mousepad:

sudo mousepad /etc/default/grub

…or GNOME Edit:

sudo gedit /etc/default/grub

…or any other simple text editor available, and then add pcie_aspm=off as a parameter for GRUB_CMDLINE_LINUX_DEFAULT so GRUB “tells” the kernel that it must disable support for PCIe Architecture Power Management. For instance, replace:

GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"

…with:

GRUB_CMDLINE_LINUX_DEFAULT="quiet splash pcie_aspm=off"

…and then save the file and exit the text editor. Next, update GRUB:

sudo update-grub

Reboot the computer and then observe your system for a while, check if the kernel keeps reporting those error messages into the system logs and how the power off menu item behaves.

Also check if with a different DE (Desktop Environment) the power off call works: maybe this issue is DE-specific. A lightweight one is Xfce (I use XUbuntu and actually prefer it). If you decide to add the Xfce DE, the installation command is:

sudo apt install xubuntu-desktop -y

Then reboot and select Xfce before your next login.

Go to normal shutdown but from time to time after 20-30 seconds.
I turn off always in the GUI.

i’ll try with grub update