Univention Bugzilla – Bug 44452
KDE login is not shown
Last modified: 2017-05-03 15:16:16 CEST
I installed UCS 4.2-0 VM on my local laptop as KVM VM (qemu-kmv 1:2.5+dfsg-5ubuntu10.11). The only software component selected during setup was the Desktop Environment (KDE). After the installation finished successful, the only thing I see is the console. Neither the KDE login nor the Welcome-Screen is shown. The bootsplash itself works, so I guess it's not Bug #44061. KDM complains about a dead X server: root@ucs-7125:~# systemctl status kdm.service ● kdm.service - LSB: X display manager for KDE Loaded: loaded (/etc/init.d/kdm) Active: active (running) since So 2017-04-23 14:47:38 CEST; 4min 5s ago Process: 907 ExecStart=/etc/init.d/kdm start (code=exited, status=0/SUCCESS) CGroup: /system.slice/kdm.service └─961 /usr/bin/kdm -config /var/run/kdm/kdmrc Apr 23 14:47:38 ucs-7125 kdm[907]: Starting K Display Manager: kdm. Apr 23 14:47:38 ucs-7125 systemd[1]: Started LSB: X display manager for KDE. Apr 23 14:47:38 ucs-7125 kdm[961]: X server died during startup Apr 23 14:47:38 ucs-7125 kdm[961]: X server for display :0 cannot be started, session disabled However, if I disable the boot splash the KDE login is displayed and KDM only complains about a missing default face file: root@ucs-7125:~# LANG=C systemctl status kdm.service ● kdm.service - LSB: X display manager for KDE Loaded: loaded (/etc/init.d/kdm) Active: active (running) since So 2017-04-23 14:52:55 CEST; 2min 59s ago Process: 902 ExecStart=/etc/init.d/kdm start (code=exited, status=0/SUCCESS) CGroup: /system.slice/kdm.service ├─ 942 /usr/bin/kdm -config /var/run/kdm/kdmrc ├─ 947 /usr/bin/X :0 vt7 -br -nolisten tcp -auth /var/run/xauth/A:0-0Gwjzb ├─1067 -:0 └─1070 /usr/lib/kde4/libexec/kdm_greet Apr 23 14:52:55 ucs-7125 kdm[902]: Starting K Display Manager: kdm. Apr 23 14:52:55 ucs-7125 systemd[1]: Started LSB: X display manager for KDE. Apr 23 14:52:59 ucs-7125 kdm_greet[1070]: Cannot load /usr/share/kde4/apps/kdm/faces/.default.face: Datei oder Verzeichnis nicht gefunden So, the workaround is: ucr set grub/bootsplash='nosplash' && reboot (but unfortunately: Bug #40650)
I could not reproduce your exact issue, i did test installs on hardware (lynxX), KVM (dvd + appliance installs), virtualbox. I have an idea though - Did you use a KVM Appliance image to setup UCS, or did you install from DVD? If so, does settings ucr set xorg/autodetect=yes help? A USI when the issue occurs would be helpful. That would be right after installation.
Created attachment 8803 [details] libvirt xml file of VM I used the 4.2-0 DVD/ISO (md5: d88b190e8d33d41101f89e10b3e64c70) and "Virtual Machine Manager" (virt-manager) as GUI. > $ dpkg -l | grep -E 'kvm|qemu|libvirt|virt-' > ii gir1.2-libvirt-glib-1.0:amd64 0.2.2-0.1ubuntu1 amd64 libvirt glib mainloop integration > ii ipxe-qemu 1.0.0+git-20150424.a25a16d-1ubuntu1 all PXE boot firmware - ROM images for qemu > ii libvirt-bin 1.3.1-1ubuntu10.8 amd64 programs for the libvirt library > ii libvirt-glib-1.0-0:amd64 0.2.2-0.1ubuntu1 amd64 libvirt glib mainloop integration > ii libvirt0:amd64 1.3.1-1ubuntu10.8 amd64 library for interfacing with different virtualization systems > ii python-libvirt 1.3.1-1ubuntu1 amd64 libvirt Python bindings > ii qemu-block-extra:amd64 1:2.5+dfsg-5ubuntu10.11 amd64 extra block backend modules for qemu-system and qemu-utils > ii qemu-kvm 1:2.5+dfsg-5ubuntu10.11 amd64 QEMU Full virtualization > ii qemu-system-common 1:2.5+dfsg-5ubuntu10.11 amd64 QEMU full system emulation binaries (common files) > ii qemu-system-x86 1:2.5+dfsg-5ubuntu10.11 amd64 QEMU full system emulation binaries (x86) > ii qemu-utils 1:2.5+dfsg-5ubuntu10.11 amd64 QEMU utilities > ii virt-manager 1:1.3.2-3ubuntu1.16.04.3 all desktop application for managing virtual machines > ii virt-viewer 1.0-1 amd64 Displaying the graphical console of a virtual machine $ cat /etc/lsb-release DISTRIB_ID=Ubuntu DISTRIB_RELEASE=16.04 DISTRIB_CODENAME=xenial DISTRIB_DESCRIPTION="Ubuntu 16.04.2 LTS" $ uname -smrvpio Linux 4.4.0-72-generic #93-Ubuntu SMP Fri Mar 31 14:07:41 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux I also attached the XML definition file of the VM. "ucr set xorg/autodetect=yes" doesn't help, unfortunately.
It is an issue with the stop of plymouth and the start of kdm. There was no implicit dependency, which lead to a race condition. If plymouth did not release the graphics card yet, the X server start by kdm would fail with /var/log/Xorg.0.log: (EE) /dev/dri/card0: failed to set DRM interface version 1.4: Permission denied Fixed by adding plymouth-quit in the kdm init script "Should-Start" dependency r17481 r17483 add and remove patch to/from wrong scope r17482 patch migration, commit by repo ng kde-workspace 4:4.11.13-2A~4.2.0.201704271220 r78969 yaml
r17484 depend on plymouth-quit-wait instead of plymouth-quit Check dependencies with "systemctl show kdm | grep ^After", and in /run/systemd/generator.late/kdm.service kde-workspace 4:4.11.13-2A~4.2.0.201704271339 r78970 yaml
OK - kdm patch OK - kdm init script sets "Should-Start: ... plymouth-quit-wait" which generates a systemd service with "After=... plymouth-quit-wait.service" OK - in my test kdm is now always started after plymouth-quit-wait OK - YAML
<http://errata.software-univention.de/ucs/4.2/7.html>