opensuse 13.1, KDE -- switch user fails, returns to and locks current session

Hi,

I recently installed 13.1 on two machines (very different hardware). One machine is able to switch users without issue while the other does not switch users successfully. If I’m logged in and go to Leave -> Switch User, I get the normal dialog with open sessions and the ability to start a new session. Clicking on “Start New Session” briefly goes to a black screen (as though trying to start a new session), comes back to my original session, and then locks the session. I checked /var/log/messages and here is the output including the three events (i.e., original login, switch attempt, and unlocking the locked session):


2014-09-02T16:02:37.207327-06:00 my_comp kdm: :0[10555]: pam_unix(xdm:session): session opened for user me by (uid=0)
2014-09-02T16:02:37.210547-06:00 my_comp systemd[1]: Starting Session 509 of user me. 
2014-09-02T16:02:37.211192-06:00 my_comp systemd-logind[1115]: New session 509 of user me. 
2014-09-02T16:02:37.211802-06:00 my_comp systemd[1]: Started Session 509 of user me. 
2014-09-02T16:02:39.098598-06:00 my_comp dbus[1113]: [system] Activating service name='org.kde.powerdevil.backlighthelper' (using servicehelper)
2014-09-02T16:02:39.108452-06:00 my_comp dbus-daemon[1113]: no kernel backlight interface found
2014-09-02T16:02:39.128864-06:00 my_comp dbus[1113]: [system] Successfully activated service 'org.kde.powerdevil.backlighthelper' 
2014-09-02T16:02:44.296427-06:00 my_comp rtkit-daemon[3675]: Successfully made thread 21730 of process 21730 (/usr/bin/pulseaudio) owned by 'me' high priority at nice level -11.
2014-09-02T16:02:44.297203-06:00 my_comp rtkit-daemon[3675]: Supervising 1 threads of 1 processes of 1 users.
2014-09-02T16:02:44.529267-06:00 my_comp rtkit-daemon[3675]: Successfully made thread 21750 of process 21750 (/usr/bin/pulseaudio) owned by 'me' high priority at nice level -11.
2014-09-02T16:02:44.530186-06:00 my_comp rtkit-daemon[3675]: Supervising 2 threads of 2 processes of 1 users.
2014-09-02T16:02:44.530908-06:00 my_comp pulseaudio[21750]: [pulseaudio] pid.c: Daemon already running.
2014-09-02T16:02:44.537405-06:00 my_comp rtkit-daemon[3675]: Successfully made thread 21758 of process 21758 (/usr/bin/pulseaudio) owned by 'me' high priority at nice level -11.
2014-09-02T16:02:44.538093-06:00 my_comp rtkit-daemon[3675]: Supervising 2 threads of 2 processes of 1 users.
2014-09-02T16:02:44.538754-06:00 my_comp pulseaudio[21758]: [pulseaudio] pid.c: Daemon already running.
2014-09-02T16:02:46.764194-06:00 my_comp polkitd[1157]: Registered Authentication Agent for unix-session:509 (system bus name :1.1130 [/usr/lib64/kde4/libexec/polkit-kde-authentication-agent-1], object path /org/kde/PolicyKit1/Authentica
2014-09-02T16:02:57.643894-06:00 my_comp kdm_config[21884]: Multiple occurrences of section [General] in /usr/share/kde4/config/kdm/kdmrc. Consider merging them.
2014-09-02T16:02:57.644676-06:00 my_comp kdm_config[21884]: Multiple occurrences of section [Xdmcp] in /usr/share/kde4/config/kdm/kdmrc. Consider merging them.
2014-09-02T16:02:57.644967-06:00 my_comp kdm_config[21884]: Multiple occurrences of section [X-*-Core] in /usr/share/kde4/config/kdm/kdmrc. Consider merging them.
2014-09-02T16:02:57.645223-06:00 my_comp kdm_config[21884]: Multiple occurrences of section [X-*-Greeter] in /usr/share/kde4/config/kdm/kdmrc. Consider merging them.
2014-09-02T16:02:57.645491-06:00 my_comp kdm_config[21884]: Multiple occurrences of section [X-:*-Core] in /usr/share/kde4/config/kdm/kdmrc. Consider merging them.
2014-09-02T16:02:57.645765-06:00 my_comp kdm_config[21884]: Multiple occurrences of section [X-:0-Core] in /usr/share/kde4/config/kdm/kdmrc. Consider merging them.
2014-09-02T16:02:59.321681-06:00 my_comp kernel: [449846.156015] [drm] mga base 0
2014-09-02T16:02:59.375202-06:00 my_comp kdm[1187]: X server died during startup 
2014-09-02T16:02:59.375846-06:00 my_comp kdm[1187]: X server for display :1 cannot be started, session disabled
2014-09-02T16:03:00.937700-06:00 my_comp kernel: [449847.822013] [drm] mga base 0


I’m guessing that the “multiple occurrences” of kdmrc has something to do with it. I appreciate any advice.

Mark

No, that’s “normal”. It’s caused by the fact that on openSUSE an additional kdmrc is created from some values in /etc/sysconfig.

Your problem is this:

2014-09-02T16:02:59.375202-06:00 my_comp kdm[1187]: X server died during startup
2014-09-02T16:02:59.375846-06:00 my_comp kdm[1187]: X server for display :1 cannot be started, session disabled

I.e. the new Xserver instance crashes.

Please post the file /var/log/Xorg.1.log to see why.

Probably a problem with your graphics driver though. mga is not really the best supported one any more.

Thanks for your help! I copied the portion that includes MGA info and errors. Let me know if it’s insufficient!


[449668.708] (==) Matched mga as autoconfigured driver 0

...

[449668.708] (II) LoadModule: "mga"
[449668.708] (II) Loading /usr/lib64/xorg/modules/drivers/mga_drv.so
[449668.708] (II) Module mga: vendor="X.Org Foundation"
[449668.708]    compiled for 1.14.3, module version = 1.6.2
[449668.708]    Module class: X.Org Video Driver
[449668.708]    ABI class: X.Org Video Driver, version 14.1

...


[449668.710] (II) MGA: driver for Matrox chipsets: mga2064w, mga1064sg, mga2164w,
    mga2164w AGP, mgag100, mgag100 PCI, mgag200, mgag200 PCI,
    mgag200 SE A PCI, mgag200 SE B PCI, mgag200 EV Maxim,
    mgag200 ER SH7757, mgag200 eW Nuvoton, mgag200eH, mgag400, mgag550
[449668.710] (II) modesetting: Driver for Modesetting Kernel Drivers: kms
[449668.710] (II) FBDEV: driver for framebuffer: fbdev
[449668.710] (II) VESA: driver for VESA chipsets: vesa
[449668.710] (++) using VT number 8


[449670.362] (EE) mga: The PCI device 0x534 at 10@00:00:0 has a kernel module claiming it.
[449670.362] (EE) mga: This driver cannot operate until it has been unloaded.
[449670.362] (WW) Falling back to old probe method for modesetting
[449670.362] (II) modesetting(1): using default device
[449670.362] (II) Loading sub module "fbdevhw"
[449670.362] (II) LoadModule: "fbdevhw"
[449670.363] (II) Loading /usr/lib64/xorg/modules/libfbdevhw.so
[449670.363] (II) Module fbdevhw: vendor="X.Org Foundation"
[449670.363]    compiled for 1.14.3.901, module version = 0.0.2
[449670.363]    ABI class: X.Org Video Driver, version 14.1
[449670.363] (**) FBDEV(2): claimed PCI slot 10@0:0:0
[449670.363] (II) FBDEV(2): using default device
[449670.363] (WW) Falling back to old probe method for vesa
[449670.363] (EE) Screen 0 deleted because of no matching config section.
[449670.363] (II) UnloadModule: "modesetting"
[449670.363] (EE)
Fatal server error:
[449670.363] (EE) Cannot run in framebuffer mode. Please specify busIDs        for all framebuffer devices
[449670.363] (EE)
[449670.363] (EE)
Please consult the The X.Org Foundation support
     at http://wiki.x.org
 for help.
[449670.363] (EE) Please also check the log file at "/var/log/Xorg.1.log" for additional information.
[449670.363] (EE)
[449670.415] (EE) Server terminated with error (1). Closing log file.

Well, the problem apparently is this:


[449668.708][449670.362] (EE) mga: The PCI device 0x534 at 10@00:00:0 has a kernel module claiming it.
[449670.362] (EE) mga: This driver cannot operate until it has been unloaded.

So the mga driver refuses to load, as the graphics card is already in use by a kernel module.

I do remember the mga driver falsely reporting this, but that problem should be fixed and would already affect you during boot, i.e. X would fail to start completely.

In this case it’s probably because the driver is already running on another display.

Maybe mga just doesn’t support two or more X sessions at the same time?
Or this is a bug.
I have no idea.

You should probably file a bug report at http://bugs.freedesktop.org to get clarification.
But first you should maybe try the latest version of the driver.
I will create a package for 13.1 and then report back to tell you where to get it from.

The package is ready now and available here:
http://software.opensuse.org/download.html?project=home%3Awolfi323%3Abranches%3AopenSUSE%3A13.1%3AUpdate&package=xf86-video-mga

If that doesn’t help, you really should file a bug report.

A work-around would be to use the fbdev driver instead of mga, but then you would have absolutely no hardware acceleration or automatic detection of your screen’s resolution, and so on.

Rats, no dice. I’ll submit a bug report. I’ve included /var/log/Xorg.1.log again for educational purposes.


  2033.238] (II) LoadModule: "mga"
  2033.238] (II) Loading /usr/lib64/xorg/modules/drivers/mga_drv.so
  2033.238] (II) Module mga: vendor="X.Org Foundation"
  2033.238]    compiled for 1.14.3.901, module version = 1.6.3
  2033.238]    Module class: X.Org Video Driver
  2033.238]    ABI class: X.Org Video Driver, version 14.1

...

  2033.240] (II) MGA: driver for Matrox chipsets: mga2064w, mga1064sg, mga2164w,
    mga2164w AGP, mgag100, mgag100 PCI, mgag200, mgag200 PCI,
    mgag200 SE A PCI, mgag200 SE B PCI, mgag200 EV Maxim,
    mgag200 ER SH7757, mgag200 eW Nuvoton, mgag200eH, mgag400, mgag550

...

  2019.902] (EE) mga: The PCI device 0x534 at 10@00:00:0 has a kernel module claiming it.
132   2019.902] (EE) mga: This driver cannot operate until it has been unloaded.
133   2019.902] (WW) Falling back to old probe method for modesetting
134   2019.902] (II) modesetting(1): using default device
135   2019.902] (II) Loading sub module "fbdevhw"
136   2019.902] (II) LoadModule: "fbdevhw"
137   2019.902] (II) Loading /usr/lib64/xorg/modules/libfbdevhw.so
138   2019.902] (II) Module fbdevhw: vendor="X.Org Foundation"
139   2019.902]    compiled for 1.14.3.901, module version = 0.0.2
140   2019.902]    ABI class: X.Org Video Driver, version 14.1
141   2019.902] (**) FBDEV(2): claimed PCI slot 10@0:0:0
142   2019.903] (II) FBDEV(2): using default device
143   2019.903] (WW) Falling back to old probe method for vesa
144   2019.903] (EE) Screen 0 deleted because of no matching config section.
145   2019.903] (II) UnloadModule: "modesetting"
146   2019.903] (EE)
147 Fatal server error:
148   2019.903] (EE) Cannot run in framebuffer mode. Please specify busIDs        for all framebuffer devices
149   2019.903] (EE)
150   2019.903] (EE)
151 Please consult the The X.Org Foundation support
152      at http://wiki.x.org
153  for help.
154   2019.903] (EE) Please also check the log file at "/var/log/Xorg.1.log" for additional information.
155   2019.903] (EE)
156   2019.955] (EE) Server terminated with error (1). Closing log file.



I submitted the bug (#83458) at https://bugs.freedesktop.org/show_bug.cgi?id=83458. I tried to edit my last post, but the 10-minute time limit expired.

Should I expect a timely response from x.org folk?

Well, no idea.
I never reported a graphics driver bug there, I have to admit.

I emailed the individual automatically assigned to MGA bugs to see what timeframe to expect. He let me know that he no longer works on MGA and believes there is not any current development. His suggestion was to buy a different video card. I never bought a dedicated video card, so that’s probably my issue.

Thanks again for the help!