3

[UPDATE: I appreciate the suggestions you posted, but since I couldn't make anything work in the end, I ended up re-installing Ubuntu a little while ago.]

I am running Ubuntu 11.10 "Oneiric Ocelot", and in trying to install the nvidia CUDA developer drivers I switched to a virtual terminal (Ctrl-Alt-F5) and stopped lightdm (installation required that no X server instance be running) through sudo service lightdm stop.

Re-starting lightdm with sudo service lightdm start did not work: A couple of * Starting [...] lines where displayed, but the process hanged. (I do not remember at which point, but I think it was * Starting System V runlevel compatibility.

I manually rebooted my laptop, and ever since booting seems to hang, usually around the * Starting anac(h)ronistic cron [OK] log line (not consistently at that point, though).

From that point on, I seem to be able to interact with my system only through a tty session (Ctrl-Alt-F1).

I've tried purging and reinstalling both lightdm and gdm, as well as selecting both as the default display managers (through sudo dpkg-reconfigure [lightdm / gdm] or by manually editing /etc/X11/default-display-manager) through both apt-get and aptitude (that shouldn't make a difference anyway) after updating the packages, but the problem persists.

Some of the responses I'm getting are the following:

After running sudo dpkg-reconfigure lightdm (but not ... gdm) I get the following message:

dpkg-maintscript-helper:warning: environment variable DPKG_MATINSCRIPT_NAME missing
dpkg-maintscript-helper:warning: environment variable DPKG_MATINSCRIPT_PACKAGE missing

After trying sudo service lightdm start or sudo start lightdm I get to see the boot loading screen again but nothing changes. If I go back to the tty shell I see lightdm start/running, process <num> but ps -e | grep lightdm gives no output.

After trying sudo service gdm start or sudo starg gdm I get the gdm start/running, process <num> message, and gdm-binary is supposedly an active process, but all that happens is that the screen blinks a couple of times and nothing else.

Other candidate solutions that I'd found on the web included running startx but when I try that I get an error output [...] Fatal server error: no screens found [...]. Moreover, I made sure that lightdm-gtk-greeter is installed but that did not help either.

Please excuse my not including complete outputs/logs; I am writing this post from another computer and it's hard to manually copy the complete logs. Also, I've seen several posts that had to do with similar problems, but either there was no fix, or the one suggested did not work for me.

In closing: Please help! I very much hope to avoid re-installing Ubuntu from scratch! :)

Alex


@mosi

I did not manage to fix the NVIDIA kernel driver as per your instructions. I should perhaps mention that I'm on a Dell XPS15 laptop with an NVIDIA Optimus graphics card, and that I have bumblebee installed (which installs nvidia drivers during its installation, I believe). Issuing the mentioned commands I get the following:

~$uname -r
3.0.0-12-generic
~$lsmod | grep -i nvidia
nvidia               11713772  0
~$dmesg | grep -i nvidia
[    8.980041] nvidia: module license 'NVIDIA' taints kernel.
[    9.354860] nvidia 0000:01:00.0: power state changed by ACPI to D0
[    9.354864] nvidia 0000:01:00.0: power state changed by ACPI to D0
[    9.354868] nvidia 0000:01:00.0: enabling device (0006 -> 0007)
[    9.354873] nvidia 0000:01:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
[    9.354879] nvidia 0000:01:00.0: setting latency timer to 64
[    9.355052] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  280.13  Wed Jul 27 16:53:56 PDT 2011

Also, running aptitude search nvidia gives me the following:

p   nvidia-173                      - NVIDIA binary Xorg driver, kernel module a
p   nvidia-173-dev                  - NVIDIA binary Xorg driver development file
p   nvidia-173-updates              - NVIDIA binary Xorg driver, kernel module a
p   nvidia-173-updates-dev          - NVIDIA binary Xorg driver development file
p   nvidia-96                       - NVIDIA binary Xorg driver, kernel module a
p   nvidia-96-dev                   - NVIDIA binary Xorg driver development file
p   nvidia-96-updates               - NVIDIA binary Xorg driver, kernel module a
p   nvidia-96-updates-dev           - NVIDIA binary Xorg driver development file
p   nvidia-cg-toolkit               - Cg Toolkit - GPU Shader Authoring Language
p   nvidia-common                   - Find obsolete NVIDIA drivers              
i   nvidia-current                  - NVIDIA binary Xorg driver, kernel module a
p   nvidia-current-dev              - NVIDIA binary Xorg driver development file
c   nvidia-current-updates          - NVIDIA binary Xorg driver, kernel module a
p   nvidia-current-updates-dev      - NVIDIA binary Xorg driver development file
i   nvidia-settings                 - Tool of configuring the NVIDIA graphics dr
p   nvidia-settings-updates         - Tool of configuring the NVIDIA graphics dr
v   nvidia-va-driver                -                                           
v   nvidia-va-driver                -                                           

I've tried manually installing (sudo aptitude install <package>) packages nvidia-common and nvidia-settings-updates but to no avail. For example, sudo aptitude install nvidia-settings-updates returns the following log:

Reading package lists...
Building dependency tree...
Reading state information...
Reading extended state information...
Initializing package states...
Writing extended state information...
No packages will be installed, upgraded, or removed.
0 packages upgraded, 0 newly installed, 0 to remove and 83 not upgraded.
Need to get 0 B of archives. After unpacking 0 B will be used.
Writing extended state information...
Reading package lists...
Building dependency tree...
Reading state information...
Reading extended state information...
Initializing package states...
Writing extended state information...

The same happens with the Linux headers (i.e. I cannot seem to be able to install linux-headers-3.0.0-12-generic). The output of aptitude search linux-headers is as follows:

v   linux-headers                   -                                           
v   linux-headers                   -                                           
v   linux-headers-2.6               -                                           
i   linux-headers-2.6.38-11         - Header files related to Linux kernel versi
i   linux-headers-2.6.38-11-generic - Linux kernel headers for version 2.6.38 on
i A linux-headers-2.6.38-8          - Header files related to Linux kernel versi
i A linux-headers-2.6.38-8-generic  - Linux kernel headers for version 2.6.38 on
v   linux-headers-3                 -                                           
v   linux-headers-3.0               -                                           
v   linux-headers-3.0               -                                           
i A linux-headers-3.0.0-12          - Header files related to Linux kernel versi
p   linux-headers-3.0.0-12-generic  - Linux kernel headers for version 3.0.0 on 
p   linux-headers-3.0.0-12-generic- - Linux kernel headers for version 3.0.0 on 
p   linux-headers-3.0.0-12-server   - Linux kernel headers for version 3.0.0 on 
p   linux-headers-3.0.0-12-virtual  - Linux kernel headers for version 3.0.0 on 
p   linux-headers-generic           - Generic Linux kernel headers              
p   linux-headers-generic-pae       - Generic Linux kernel headers              
v   linux-headers-lbm               -                                           
v   linux-headers-lbm               -                                           
v   linux-headers-lbm-2.6           -                                           
v   linux-headers-lbm-2.6           -                                           
p   linux-headers-lbm-3.0.0-12-gene - Header files related to linux-backports-mo
p   linux-headers-lbm-3.0.0-12-gene - Header files related to linux-backports-mo
p   linux-headers-lbm-3.0.0-12-serv - Header files related to linux-backports-mo
p   linux-headers-server            - Linux kernel headers on Server Equipment. 
p   linux-headers-virtual           - Linux kernel headers for virtual machines 

@heartsmagic

I did try purging and reinstalling any nvidia driver packages, but it did not seem to make a difference,

My xorg.conf file contains the following:

# nvidia-xconfig: X configuration file generated by nvidia-xconfig
# nvidia-xconfig:  version 280.13  ([email protected])  Wed Jul 27 17:15:58 PDT 2011


Section "ServerLayout"
    Identifier     "Layout0"
    Screen      0  "Screen0" 0 0
    InputDevice    "Keyboard0" "CoreKeyboard"
    InputDevice    "Mouse0" "CorePointer"
EndSection

Section "Files"
EndSection

Section "InputDevice"

    # generated from default
    Identifier     "Mouse0"
    Driver         "mouse"
    Option         "Protocol" "auto"
    Option         "Device" "/dev/psaux"
    Option         "Emulate3Buttons" "no"
    Option         "ZAxisMapping" "4 5"
EndSection

Section "InputDevice"

    # generated from default
    Identifier     "Keyboard0"
    Driver         "kbd"
EndSection

Section "Monitor"
    Identifier     "Monitor0"
    VendorName     "Unknown"
    ModelName      "Unknown"
    HorizSync       28.0 - 33.0
    VertRefresh     43.0 - 72.0
    Option         "DPMS"
EndSection

Section "Device"
    Identifier     "Device0"
    Driver         "nvidia"
    VendorName     "NVIDIA Corporation"
EndSection

Section "Screen"
    Identifier     "Screen0"
    Device         "Device0"
    Monitor        "Monitor0"
    DefaultDepth    24
    SubSection     "Display"
        Depth       24
    EndSubSection
EndSection
2
  • 1
    Could'nt be the problem Nvidia driver installation? Have you tried uninstalling them? What about your xorg.conf file content? Nov 9, 2011 at 2:02
  • WoW, a very big problem(question) Nov 26, 2011 at 22:13

3 Answers 3

3

I had this issue with 12.04. The fix is to change /etc/X11/default-display-manager to say /usr/sbin/lightdm instead of just lightdm as dpkg-reconfigure sets incorrectly.

Why? in /etc/init/lightdm.conf the script is checking that file says /usr/sbin/lightdm

This is a bug. A simple fix would be to change lightdm.conf to accept other variations including just lightdm without the full path. Unless there is some standard that should be used purely for consistency.

1
  • +1 even though this is buried at the bottom of lots of text in the OP and other answers, this simple fix solved my issue. (would not go from splash screen to login page after upgrade to 12.04) May 4, 2012 at 22:22
1

I had a similar issue. It seems that the nvidia kernel driver is not loaded. If you don`t have aptitude, install by:

apt-get install aptitude

First, login as root.

(1. see which nvidia kernel modules are loaded:

dmesg | grep -i nvidia
[   77.226200] nvidia: module license 'NVIDIA' taints kernel.
[   78.066167] nvidia 0000:01:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
[   78.066175] nvidia 0000:01:00.0: setting latency timer to 64
[   78.066794] NVRM: loading NVIDIA UNIX x86 Kernel Module  285.05.09

(2. Ubuntu nvidia drivers must be removed, otherwise ubuntu might still use them even after reboot, remove all nvidia-XX packages except these:

aptitude search nvidia | grep -i "i "
i   nvidia-common                   - Find obsolete NVIDIA drivers              
i   nvidia-current                  - NVIDIA binary Xorg driver, kernel module a
i   nvidia-settings                 - Tool of configuring the NVIDIA graphics dr
i   nvidia-settings-updates         - Tool of configuring the NVIDIA graphics dr

(3. download correct nvidia drivers for your architecture from nvidia.com make sure you have the linux-headers installed.

aptitude install linux-headers-$(uname -r)

(4. run as root: NVIDIA-Linux-x86-XXXXX.run if xserver is running, stop with:

/etc/init.d/x11-common stop

(5. generate new nvidia xconfig configuration

nvidia-xconfig

(6. start the X server

/etc/init.d/x11-common start
/etc/init.d/lightdm start

If something goes wrong, post the output of following commands:

uname -r
lsmod | grep -i nvidia
dmesg | grep -i nvidia
2
  • Would it be possible to do this with apt-get? apitude and apt-get are incompatible. :)
    – jrg
    Nov 9, 2011 at 15:48
  • Thanks for the reply! I'm afraid, however, that I couldn't get that to work. I've updated my post to reflect what went on while I tried the steps you posted.
    – Sweeters
    Nov 9, 2011 at 16:36
0

Environment: - XPS M 1530, 8600M GT - Ubuntu 11.10 64 Bit fresh install - Additional Nvidia driver-173 installed(From "Additional Drivers" section)

My external monitor was not being recognised with the additional driver i.e. nvidia-173. So I tried to install the nvidia driver found at nvidia.com for 8600M GT card and followed the instruction. To exit x server, I did

sudo stop lightdm

and installed the nvidia driver. After that I faced the same issues as the OP i.e. not able to successfully bring up the graphical system. Could interact using terminal. Starting/re starting lightdm did not help.

So I had to purge everything nvidia and use mesa-gl. With the help of the following links and commands I was able to.

https://askubuntu.com/a/68222 https://askubuntu.com/a/36936

The commands:

sudo apt-get purge nvidia*
sudo rm /etc/X11/xorg.conf
sudo apt-get install --reinstall xserver-xorg-core libgl1-mesa-glx:amd64 libgl1-mesa-dri:amd64
sudo dpkg-reconfigure xserver-xorg
sudo reboot

And bingo, I am back in action, plus the external monitors work flawlessly(or as flawless as it could be. While connecting the monitor through HDMI cable the screen jumbles up.)

I still haven't/don't want to try any more drivers. May be later.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .