Latest Nvidia Edgy

From Ubuntu Document Storage Facility

file2.png Doc Info
Written By:
tseliot
Archived By:
tseliot
Original Thread:
Dapper section
Works with release(s):
Edgy 6.10
Works on:
(U,K,X)buntu

Contents

HOWTO: Latest NVIDIA drivers

This guide will guide you through the installation of the Nvidia drivers. Why should you install Nvidia proprietary drivers?

a) If you want to enable 3D acceleration.

b) If you want to make the most of your Graphic Card (also videoclips look better)

c) If your graphic card is not supported (or only partially supported) by the Opensource driver "nv"


REQUIREMENTS

1) an installed Ubuntu Edgy Eft 6.10 system (don't try it on a livecd)

2) an Nvidia graphic card

3) an Internet connection (better if broadband)

NOTEIf you do not have an Internet connection available in Ubuntu you can use Method1 (see the section about the Offline installation)


Warning: If you have an Intel integrated card and your Nvidia card makes your computer freeze if you try to use it you should follow this guide instead:

HOWTO: install an Nvidia card instead of your Intel integrated card


EXPLANATION OF THE 3 METHODS

METHOD 1It is the usual (and easiest) way to install the Nvidia drivers. It allows you to install nvidia driver 8774 OR 7184, i.e. the Legacy driver that you need ONLY IF your graphic card belongs to the list in the "NOTES SECTION", which you will find near the end of the guide. [WARMLY RECOMMENDED TO NEWBIES]. It works only with the kernels which you can find in Ubuntu's repository.

METHOD 1 - OFFLINE Like Method 1 but Offline.

METHOD 2If you need another driver version (e.g. the latest version if it is not available in the Ubuntu's repositories) or you have problems with Method 1 or you are using a kernel you have compiled yourself (thus lacking the restricted modules required for method one) or you want to try something different then Method 2 is for you. This Method works ONLY with driver 8762 or higher (e.g. 8776)!


Warning: Method 2 REQUIRES THE REMOVAL OF THE RESTRICTED MODULES. Therefore if you use a wireless card (or another device) which requires the restricted modules, please avoid using Method 2.


WHAT HAPPENS IF YOU CHANGE YOUR KERNEL OR IF YOUR KERNEL IS UPDATED

Every time you change or upgrade your kernel you will have to reinstall the Nvidia driver, no matter which method you used to install the driver.

* If you used Method 1, make sure you install also the restricted modules for the new kernel you decide to install:

file.pngCode:

sudo apt-get install linux-generic (or 386 according to your architecture, which you can check out by typing "uname -r")


After you do that the restricted modules will be installed every time your kernel is updated (unless you install a kernel for a different architecture, e.g. you have a generic kernel and install a 386 kernel).

* If you used Method 2, just follow Method 2 again.

WHAT TO DO IF THE METHOD YOU CHOSE DID NOT WORK OR WORKED PARTIALLY

If you have any problems with the methods you should have a look at the PROBLEMS SECTION which you can find at the end of the guide. If you can't find the answer to your problem in that section then you can post your problems in the thread the link to which you can find at the end of this page and I will try to help you.



METHOD 1

Make sure you have both the Universe and Multiverse repositories enabled.


1) Identify the architecture of your kernel (e.g. generic or 386,etc. )

Open Terminal or Konsole and type:

file.pngCode:

uname -r


(You will get something like this: 2.6.17-10-generic )


2) Now that you know the architecture (generic in my case but it could also be 386), type:

file.pngCode:

sudo apt-get install linux-generic


(of course you need to use your architecture instead of k7)

In this way you will install a dummy package which will always provide you with the latest kernel image, headers and restricted modules.

3) Install the Nvidia driver:

file.pngCode:

sudo apt-get install nvidia-glx


OR (If you need Legacy drivers, i.e. your graphic card is in the list at the end of the page or you need driver 7174)

file.pngCode:

sudo apt-get install nvidia-glx-legacy nvidia-xconfig nvidia-settings


4) Backup your xorg.conf

file.pngCode:

sudo cp /etc/X11/xorg.conf /etc/X11/xorg.conf_backup


NOTE: if you lost your old xorg.conf or that doesn't work type this:

file.pngCode:

dpkg-reconfigure -phigh xserver-xorg


5) Enable the driver in your xorg:

file.pngCode:

sudo nvidia-xconfig


NOTE: if you own a GeForce4 420/440 Go you should follow also point 7 of the PROBLEMS SECTION.

6) Create a link to the “Nvidia-Settings” Panel in your application menu:

file.pngCode:

sudo gedit /usr/share/applications/NVIDIA-Settings.desktop


OR (if you use KDE)

file.pngCode:

kdesu kate /usr/share/applications/NVIDIA-Settings.desktop


NOTE: do not worry if the file is empty

Insert the following lines into the new file:

file.pngCode:

[Desktop Entry]
Name=NVIDIA Settings
Comment=NVIDIA X Server Settings
Exec=nvidia-settings
Icon=
StartupNotify=true
Terminal=false
Type=Application
Categories=Application;System;


Save the edited file.

Log out and press CTRL+ALT+BACKSPACE (so as to restart the xserver).

In case of problems, please read the PROBLEMS SECTION


HOW TO UNINSTALL THE DRIVER

Restore your old xorg.conf

file.pngCode:

sudo cp /etc/X11/xorg.conf_backup /etc/X11/xorg.conf


NOTE: if the backup of your old xorg.conf doesn't work or you lost it try this command (so as to regenerate a clean xorg.conf): sudo dpkg -reconfigure -phigh xserver-xorg

If you used the nvidia driver 8774 (not the legacy one)

file.pngCode:

sudo apt-get --purge remove nvidia-glx linux-restricted-modules-`uname -r` linux-restricted-modules-common


OR (If you need Legacy drivers, i.e. your graphic card is in the list at the end of the page or you need driver 7184)

file.pngCode:

sudo apt-get --purge remove nvidia-glx-legacy nvidia-xconfig nvidia-settings linux-restricted-modules-`uname -r` linux-restricted-modules-common


Log out and press CTRL+ALT+BACKSPACE (so as to restart the xserver).


METHOD 1 - OFFLINE INSTALLATION

1) Install the restricted modules for your kernel:

Type:

file.pngCode:

sudo apt-cdrom add


then it should ask you to insert Ubuntu's cd.

Then (after the previous process finishes) type:

file.pngCode:

sudo apt-get update


file.pngCode:

sudo apt-get install linux-386


2) Download this package from another computer (in which you have an internet connection):

  • IF you use the NON-LEGACY DRIVER

Use this package if you run Ubuntu 32bit:

Nvidia_package_32

OR this package if you run Ubuntu 64bit:

Nvidia_package_64

  • OR IF you use the LEGACY DRIVER

Use this package if you run Ubuntu 32bit:

Nvidia_package_32

OR this package if you run Ubuntu 64bit:

Nvidia_package_64


Then move the package to your computer with Ubuntu (use a USB pen for example): Double click on the .deb file and install it.


3) Backup your xorg.conf

file.pngCode:

sudo cp /etc/X11/xorg.conf /etc/X11/xorg.conf_backup


NOTE: if you lost your old xorg.conf or that doesn't work type this:

file.pngCode:

dpkg-reconfigure -phigh xserver-xorg


4) Then IF AND ONLY IF the installation of the package went fine, set up your xorg.conf:

file.pngCode:

sudo nano -w /etc/X11/xorg.conf


Get to the Section "Module" and put a "#" before load dri and load glcore so that it looks like this:

file.pngCode:

Section "Module"
Load "i2c"
Load "bitmap"
#Load "dri"
#Load "glcore"
Load "ddc"
Load "extmod"
Load "freetype"
Load "glx"
Load "int10"
Load "type1"
Load "vbe"
EndSection


Then get to the Section device and set the driver to "nvidia" (instead of "nv" or "vesa"):

file.pngCode:

Driver "nvidia"


CTRL+X to exit (save the file)

5) Then log out and press CTRL+ALT+Backspace

NOTE: if you need to uninstall the driver you have to read the section "HOW TO UNINSTALL THE DRIVER" which you can find right after Method 1 (the online installation)


METHOD 2

If you follow this method you will use the nvidia installer, in this way you can install version 8774 or higher of the Nvidia driver (i.e. usually also the latest release of the driver)

Make sure you graphic card is not among the ones which are NOT SUPPORTED by looking at the list that you will find in the NOTES SECTION. If your card is among the ones listed you have to use driver 7174 (or lower) (you can install it by going to the archive if nvidia drivers in the Nvidia website)

REMEMBER: every time your kernel is upgraded or changed with another one you have to reinstall the drivers.

1) Download the installer from this page according to your architecture (32bit or 64bit) http://www.nvidia.com/object/unix.html

2) Open Terminal or Konsole and type:

file.pngCode:

sudo apt-get install linux-headers-`uname -r` build-essential gcc gcc-3.4 xserver-xorg-dev


NOTE: please make sure you use a ` instead of a common apostrophe '. If you don't know how to type a ` then you can use this $(uname -r) instead of `uname -r` (which is pretty much the same).

NOTE: if APT complains because it can't find any package to remove in the following steps it's not a problem at all and you can go ahead with the next steps. If there are other errors, please report to me.

3) Then open Terminal or Konsole and type:

file.pngCode:

sudo apt-get --purge remove nvidia-glx nvidia-settings nvidia-glx-legacy


NOTE: it doesn't matter if apt says that it can't find some of the packages


4) Press CTRL-ALT-F1 (so as to get to the command line, not a windowed terminal, but out of the graphical interface GUI)

login with your username and password (if required)

file.pngCode:

sudo /etc/init.d/gdm stop (OR "kdm stop" if you use KDE)



6) Make a backup of your xorg.conf

file.pngCode:

sudo cp /etc/X11/xorg.conf /etc/X11/xorg.conf_backup


[In this way if anything goes wrong and you have problems with the xserver (i.e. if it doesn't start) you can type:

sudo cp /etc/X11/xorg.conf_backup /etc/X11/xorg.conf ]

7) Get to the directory in which you have the nvidia installer

file.pngCode:

cd “directory_where_you_have_the_nvidia_installer”

NOTE: of course you have to replace “directory_where_you_have_the_nvidia_installer” with the directory in which you have the nvidia installer


8) Install the driver:

NOTE: in these examples I will NVIDIA-Linux-x86-1.0-8776-pkg1.run but you can replace it with any version of the nvidia installer

  • IF you are using the kernel that comes by default with Ubuntu (or if you don't know what a kernel is) then type:
file.pngCode:

sudo sh NVIDIA-Linux-x86-1.0-8776-pkg1.run -n -s --x-prefix=/usr/lib/xorg/ --kernel-source-path=/usr/src/linux-headers-`uname -r`

OR (if you use Ubuntu 64bit)

file.pngCode:

sudo sh NVIDIA-Linux-x86_64-1.0-8776-pkg2.run -n -s --x-prefix=/usr/lib64/xorg/ --kernel-source-path=/usr/src/linux-headers-`uname -r`


NOTE: if you own a GeForce4 420 Go or a GeForce4 440 Go you should follow also point 7 of the PROBLEMS SECTION.

Then go to Step 9


NOTE: please make sure you use a ` instead of a common apostrophe '. If you don't know how to type a ` then you can use this $(uname -r) instead of `uname -r` (which is pretty much the same).


  • OTHERWISE if you are using a kernel which you compiled from vanilla sources (i.e. from kernel.org) you should type these commands (and then go to Step 9):
file.pngCode:

sudo cp /usr/src/name_of_your_source/arch/i386/Makefile.cpu /usr/src/kernel-headers-`uname -r`/arch/i386/


NOTE: you have to replace "name_of_your_source" with the name of your source, such as linux-2.6.16, etc.)

file.pngCode:

cd /usr/src/kernel-headers-`uname -r`/


file.pngCode:

sudo make prepare


file.pngCode:

sudo make prepare scripts


file.pngCode:

cd $HOME


file.pngCode:

cd “directory_where_you_have_the_nvidia_installer”


file.pngCode:

sudo sh NVIDIA-Linux-x86-1.0-8776-pkg1.run -n -s --x-prefix=/usr/lib/xorg/ --kernel-source-path=/usr/src/kernel-headers-`uname -r`

OR (if you use Ubuntu 64bit)

file.pngCode:

sudo sh NVIDIA-Linux-x86_64-1.0-8776-pkg2.run -n -s --x-prefix=/usr/lib64/xorg/ --kernel-source-path=/usr/src/linux-headers-`uname -r`


NOTE: please make sure you use a ` instead of a common apostrophe '. If you don't know how to type a ` then you can use this $(uname -r) instead of `uname -r` (which is pretty much the same).


NOTE: if you own a GeForce4 420 Go, a GeForce4 440 Go or a GeForce4 MX 440 you should follow also point 7 of the PROBLEMS SECTION.


9) Set up your /etc/X11/xorg.conf by using nvidia-xconfig:

  • If you DO NOT use the legacy driver type:
file.pngCode:

sudo nvidia-xconfig --no-composite


Then go to point 10

  • OR, if you DO use the Legacy driver type:
file.pngCode:

sudo apt-get install nvidia-xconfig nvidia-settings


file.pngCode:

sudo nvidia-xconfig


Then add an entry in the Menu so as to be able to access nvidia-settings from there (this is required only by the legacy driver):

file.pngCode:

sudo nano -w /usr/share/applications/NVIDIA-Settings.desktop


Insert the following lines into the new file (which should be empty by default):

file.pngCode:

[Desktop Entry]
Name=NVIDIA Settings
Comment=NVIDIA X Server Settings
Exec=nvidia-settings
Icon=
StartupNotify=true
Terminal=false
Type=Application
Categories=Application;System;


Press CTRL+X to exit (say yes when it asks you whether to save the file or not)


10) Prevent Ubuntu's nvidia driver in the restricted modules from being loaded at boot:

file.pngCode:

sudo nano -w /etc/default/linux-restricted-modules-common


The file should look like this:

file.pngCode:

# This file is sourced from the linux-restricted-modules-common init
# script and is used to disable the link-on-boot feature, one module
# at a time. This can be useful if you want to use hand-compiled
# versions of one or more modules, but keep linux-restricted-modules
# installed on your system, or just to disable modules you don't use
# and speed up your boot process by a second or two.
#
# Use a space-separated list of modules you wish to not have linked
# on boot. The following example shows a (condensed) list of all
# modules shipped in the linux-restricted-modules packages:
#
# DISABLED_MODULES="ath_hal fc fglrx ltm nv"
#
# Note that disabling "fc" disables all fcdsl drivers, "ltm" disables
# ltmodem and ltserial, and "nv" disables both the nvidia drivers.
# You can also name each module individually, if you prefer a subset.
DISABLED_MODULES=""


You have to put "nv" in the inverted comas which you can find after DISABLED_MODULES so that your file looks like the following:

file.pngCode:

# This file is sourced from the linux-restricted-modules-common init
# script and is used to disable the link-on-boot feature, one module
# at a time. This can be useful if you want to use hand-compiled
# versions of one or more modules, but keep linux-restricted-modules
# installed on your system, or just to disable modules you don't use
# and speed up your boot process by a second or two.
#
# Use a space-separated list of modules you wish to not have linked
# on boot. The following example shows a (condensed) list of all
# modules shipped in the linux-restricted-modules packages:
#
# DISABLED_MODULES="ath_hal fc fglrx ltm nv"
#
# Note that disabling "fc" disables all fcdsl drivers, "ltm" disables
# ltmodem and ltserial, and "nv" disables both the nvidia drivers.
# You can also name each module individually, if you prefer a subset.
DISABLED_MODULES="nv"


Press CTRL+X to exit (say yes when it asks you whether to save the file or not)


11) Restart the login manager:

file.pngCode:

sudo /etc/init.d/gdm restart (or "kdm restart" if you use KDM)


Congratulations, you have installed the new nvidia driver!


REMEMBER: every time your kernel is upgraded or changed with another one you have to reinstall the drivers.


HOW TO UNINSTALL THE DRIVER (FROM METHOD 2)

1) Press CTRL-ALT-F1 (so as to get to the command line, not a windowed terminal, but out of the graphical interface GUI)

2) Login with your username and password (if required)

3) Stop the xserver:

file.pngCode:

sudo /etc/init.d/gdm stop (OR "kdm stop" if you use KDE)


4) Restore your old xorg.conf

file.pngCode:

sudo cp /etc/X11/xorg.conf_backup /etc/X11/xorg.conf


NOTE: if the backup of your old xorg.conf doesn't work or you lost it try this command (so as to regenerate a clean xorg.conf): sudo dpkg -reconfigure -phigh xserver-xorg


5) Get to the directory in which you have the nvidia installer

file.pngCode:

cd “directory_where_you_have_the_nvidia_installer”

NOTE: of course you have to replace “directory_where_you_have_the_nvidia_installer” with the directory in which you have the nvidia installer</nowiki>}}

6) Uninstall the driver NOTE: in these examples I will NVIDIA-Linux-x86-1.0-8776-pkg1.run but you can replace it with any version of the nvidia installer

file.pngCode:

sudo sh NVIDIA-Linux-x86-1.0-8776-pkg1.run --uninstall


IF AND ONLY IF you don't want to use the Nvidia driver any more you will have to reconfigure the Xserver in order to use either "nv", the opensource driver (which doesn't support all cards flawlessly), or "vesa", the more generic (but also more compatible) driver (OTHERWISE you can skip this step but you will have to reinstall the Nvidia driver). Type:

file.pngCode:

sudo dpkg-reconfigure xserver-xorg


and the program will ask you a few questions. If you don't know the answer just press ENTER to use the default one. Make sure you choose either "nv" or "vesa" when it asks you about which driver you want to use.

7) Remove Ubuntu's nvidia driver in the restricted modules from the "black list":

file.pngCode:

sudo nano -w /etc/default/linux-restricted-modules-common


The file should look like this:

file.pngCode:

# This file is sourced from the linux-restricted-modules-common init
# script and is used to disable the link-on-boot feature, one module
# at a time. This can be useful if you want to use hand-compiled
# versions of one or more modules, but keep linux-restricted-modules
# installed on your system, or just to disable modules you don't use
# and speed up your boot process by a second or two.
#
# Use a space-separated list of modules you wish to not have linked
# on boot. The following example shows a (condensed) list of all
# modules shipped in the linux-restricted-modules packages:
#
# DISABLED_MODULES="ath_hal fc fglrx ltm nv"
#
# Note that disabling "fc" disables all fcdsl drivers, "ltm" disables
# ltmodem and ltserial, and "nv" disables both the nvidia drivers.
# You can also name each module individually, if you prefer a subset.
DISABLED_MODULES="nv"


You have to remove "nv" from the inverted comas (if it is there) which you can find after DISABLED_MODULES so that your file looks like the following:

file.pngCode:

# This file is sourced from the linux-restricted-modules-common init
# script and is used to disable the link-on-boot feature, one module
# at a time. This can be useful if you want to use hand-compiled
# versions of one or more modules, but keep linux-restricted-modules
# installed on your system, or just to disable modules you don't use
# and speed up your boot process by a second or two.
#
# Use a space-separated list of modules you wish to not have linked
# on boot. The following example shows a (condensed) list of all
# modules shipped in the linux-restricted-modules packages:
#
# DISABLED_MODULES="ath_hal fc fglrx ltm nv"
#
# Note that disabling "fc" disables all fcdsl drivers, "ltm" disables
# ltmodem and ltserial, and "nv" disables both the nvidia drivers.
# You can also name each module individually, if you prefer a subset.
DISABLED_MODULES=""


Press CTRL+X to exit (say yes when it asks you whether to save the file or not)


8) Restart the xserver

file.pngCode:

sudo /etc/init.d/gdm restart (or "kdm restart" if you use KDE)



Enjoy!


NOTES SECTION

  • Below are the legacy GPUs that are no longer supported in the unified driver.

These GPUs will continue to be maintained through the special legacy NVIDIA GPU driver releases.

file.pngCode:

NVIDIA chip name Device PCI ID
------------------------------- -------------------------------
RIVA TNT 0x0020
RIVA TNT2/TNT2 Pro 0x0028
RIVA TNT2 Ultra 0x0029
Vanta/Vanta LT 0x002C
RIVA TNT2 Model 64/Model 64 Pro 0x002D
Aladdin TNT2 0x00A0
GeForce 256 0x0100
GeForce DDR 0x0101
Quadro 0x0103
GeForce2 GTS/GeForce2 Pro 0x0150
GeForce2 Ti 0x0151
GeForce2 Ultra 0x0152
Quadro2 Pro 0x0153


PROBLEMS SECTION

1) If the installer reports that the Framebuffer kernel module conflicts with the drivers you will have to recompile your kernel and disable this function

Here's a HOWTO for kernel compilation for newbies HOWTO: Kernel Compilation for Newbies

2) If the installer (in Method 2) complains in this way (this is an example of part of the error):

file.pngCode:

...
nvidia: version magic '2.6.10-5-386 preempt 386 gcc-3.4' should be
'2.6.10-5-386 preempt 386 gcc-3.3'
ERROR: Installation has failed. Please see the file
'/var/log/nvidia-installer.log' for details.
...


This means the installer tries to use gcc-3.4 instead of gcc-3.3(the right one in this case).Type this before launching NVIDIA installer:

file.pngCode:

CC=gcc-3.3
export CC


The number of the version of gcc has to be the same as the 2nd one reported in the error by nvidia installer.

Then run Nvidia installer again.

3) If the installer complains in this way and you are using a Breezy kernel or a version 2.6.12-x:

file.pngCode:

...
ERROR: Unable to find the development tool `cc` in your path; please make sure
that you have the package 'gcc' installed. If gcc is installed on your
system, then please check that `cc` is in your PATH.


Make sure you have gcc-3.4 installed

file.pngCode:

sudo apt-get install gcc-3.4


Then let's make the system use gcc-3.4 instead of gcc-4.0 by default:

file.pngCode:

sudo ln -sf /usr/bin/gcc-3.4 /usr/bin/gcc


Run the Nvidia installer again (this time it won't complain)

IF everything went fine you can switch back to your old gcc settings (so as to use gcc-4.0 by default)

file.pngCode:

sudo ln -sf /usr/bin/gcc-4.0 /usr/bin/gcc


4) If you have an AGP graphic card and your system freezes but you can still move the mouse pointer you will have to do this:

file.pngCode:

sudo nano /etc/X11/xorg.conf


Add the lines which begin with the word "Option" in this section of the file:

file.pngCode:

Section "Device"
Identifier "NVIDIA Corporation NV40 [GeForce 6200 TurboCache]"
Driver "nvidia"
BusID "PCI:1:0:0"
Option "NvAGP" "0"
Option "RenderAccel" "Off"
Option "IgnoreDisplayDevices" "DFP,TV"
Option "NoRenderExtension" "Off"
Option "AllowGLXWithComposite" "Off"
EndSection


Then save the file and exit. Press CTRL+ALT+Backspace in order to restart the xserver.

IF it doesn't work yet then you can add also the following line which will disable 3d acceleration:

file.pngCode:

Option "Accel" "Off"


If this doesn't work for you try asking on the Unofficial Nvidia Forum and you might be talking to some of the developers of the NVIDIA drivers (there's a Linux section) (it's very useful) " http://www.nvnews.net/vbulletin/forumdisplay.php?f=14


5) If you have problems with the Resolution and/or Refresh rate of your display, please do the following steps OR follow this excellent guide by Heimo HOWTO: change resolution/refresh rate in Xorg:

file.pngCode:

sudo dpkg-reconfigure xserver-xorg


When it asks you about your graphic card select it manually (don't do autodetect).

Select the "advanced" when it asks you about the refresh rate (make sure you know the horizontal and vertical refresh rate supported by your monitor (try in google or if you have a manual of the monitor) OR use "autodetect" (the easy one).

It will ask you your desired resolutions, select the ones you need by pressing the SPACEBAR.

If you don't know how to answer the other questions you can use the suggested answers (which will work) by pressing ENTER (without typing anything).

After you finish, log out and press CTRL+ALT+BACKSPACE

Log in and see if everything is displayed correctly and if your card is well detected.

6) If the installer says that it can't find the kernel sources

Make sure you have installed the headers for your kernel:

file.pngCode:

sudo apt-get install linux-headers-`uname -r`


Then run the installer again in the following way:

file.pngCode:

sudo sh NVIDIA-installer-that-you-used --kernel-source-path=/usr/src/linux-headers-`uname -r`


7) If you own a GeForce4 420/440 Go or a GeForce4 MX 440 you should follow these steps:

file.pngCode:

sudo nano /etc/modprobe.d/options


Add this option at the end of the file:

file.pngCode:

options nvidia NVreg_SoftEDIDs=0 NVreg_Mobile=1


CTRL+O to save CTRL+X to exit

Edit your xorg.conf:

file.pngCode:

sudo nano /etc/X11/xorg.conf


Get to the "Section Screen" and add the following options

file.pngCode:

Option "ExactModeTimingsDVI" "TRUE"
Option "ModeValidation" "DFP-0: NoEdidDFPMaxSizeCheck, NoVesaModes"


If you added the options that section should (approximately) look like the example below:

file.pngCode:

Section "Screen"
Identifier "Screen[0]"
Device "Device[0]"
Monitor "Monitor[0]"
DefaultDepth 24
Option "ExactModeTimingsDVI" "TRUE"
Option "ModeValidation" "DFP-0: NoEdidDFPMaxSizeCheck, NoVesaModes"
SubSection "Display"
Depth 16
Modes "1280x1024" "1024x768" "832x624" "800x600" "720x400" "640x480"
EndSubSection
SubSection "Display"
Depth 24
Modes "1280x1024" "1024x768" "832x624" "800x600" "720x400" "640x480"
EndSubSection
EndSection


If those options do not work then you may replace them with the following:

file.pngCode:

Option "ExactModeTimingsDVI"
Option "UseEdidDpi" "FALSE"
Option "ModeValidation" "NoEdidDFPMaxSizeCheck, NoVesaModes"


If that did not work for your card you can try setting the following option instead of the one you put before in your /etc/modprobe.d/options:

file.pngCode:

options nvidia NVreg_SoftEDIDs=0 NVreg_Mobile=4


NOTE: In this case you will have to specify my modes and metamodes for twinview to work (provided you use Twinview).


8) If texts in Firefox look bad after the installation of the Nvidia Driver try disabling RenderAccel (which is enabled by default) by adding the following option in your Section Device of your xorg.xconf:

file.pngCode:

Option "RenderAccel" "false"


Then restart the xserver (i.e. log out and press CTRL+ALT+Backspace)

9) If (after installing the Nvidia driver) you are not able to see your mouse cursor or it blinks and this bothers you try this:

Add the following options to the Section Device of your xorg.conf:

file.pngCode:

Option "HWcursor" "Disabled"
Option "SWcursor" "Enabled"


Then log out and press CTRL+ALT+Backspace

10) If the driver ignores the modelines (and therefore the Resolution and Refresh Rate are not right) which (worked with version 8178) you have to add the following option to the Section "Device" in your xorg.conf:

file.pngCode:

Option "UseEDID" "False"


Then get to the Section "Screen" and set you resolution and refresh rate in the following way:

Put the desired Refresh rate beside the resolution you need to use. For example if you want to set the resolution to 1280x1024 and the refresh rate to 75hz you have to write 1280x1024_75 . See the example below:

Section "Screen"
Identifier "Screen[0]"
Device "Device[0]"
Monitor "Monitor[0]"
DefaultDepth 24
SubSection "Display"
Depth 15
Modes "1280x1024" "1024x768" "832x624" "800x600" "720x400" "640x480"
EndSubSection
SubSection "Display"
Depth 16
Modes "1280x1024_75" "1024x768" "832x624" "800x600" "720x400" "640x480"
EndSubSection
SubSection "Display"
Depth 24
Modes "1280x1024_75" "1024x768" "832x624" "800x600" "720x400" "640x480"
EndSubSection
EndSection


11) If you want to use Method 2 or my script but you need the restricted modules for your wifi connection you shoud try this method suggested by chemist109:

Copy the modules you want to keep from

/lib/modules/$(uname -r)/volatile/ to /lib/modules/$(uname -r)/kernel/

before you uninstall the linux-restricted-modules package.

e.g. for the madwifi driver you should do this:

file.pngCode:

sudo cp /lib/modules/$(uname -r)/volatile/ath_hal.ko /lib/modules/$(uname -r)/kernel/


Then:

file.pngCode:

sudo depmod -a $(uname -r)


Uninstall the restricted modules package:

file.pngCode:

sudo apt-get --purge remove linux-restricted-modules-`uname -r`


Then, reboot and do the driver install.

You will have to do this each time you upgrade the kernel.

12) If u have problems with XGL after installing the latest driver

Like this:

file.pngCode:

$ thefuture
$ compiz.real: GLX_EXT_texture_from_pixmap is missing
compiz.real: Failed to manage screen: 0
compiz.real: No managable screens found on display :0.0


first try this:

file.pngCode:

sudo apt-get install libgl1-mesa --reinstall


If thats not working try:

file.pngCode:

sudo dpkg -i /var/cache/apt/archives/libgl1-mesa_6.5.0-0ubuntu5_i386.deb


and replace "libgl1-mesa_6.5.0-0ubuntu5_i386.deb" with your version.

this workaround was suggested by sYs^

13) If you own a Nvidia FX5900 or a 5700 card (but not only) you might be affected by a bug which prevents (or shows buggy graphics when) users from logging out, switching to another user, shutting down or changing to a console (ALT-(F1-6)).

Try this:

file.pngCode:

sudo nano -w /boot/grub/menu.lst


look for this line:

file.pngCode:

# defoptions=quiet splash


Remove the word "splash" from it.

Save and exit: press CTRL+X

Then type this in Terminal or Konsole:

file.pngCode:

sudo update-grub


and restart your computer

After that, you will have no pretty usplash screen to look at anymore but the bug should be gone.

14) If you have an Intel integrated card and your Nvidia card makes your computer freeze if you try to use it you should follow this guide instead: HOWTO: install an Nvidia card instead of your Intel integrated card

15) If "nvidia-settings" loses its settings every time you reboot Get to the following menus (if you use GNOME): System -> Preferences -> Sessions -> Startup Programs

Then click on "Add"

And insert this command:

file.pngCode:

nvidia-settings --load-config-only


Then click on "Close".

16) If you can hear drums (greeting) sound when your Ubuntu boots but there is no output on the monitor: See point 10 of this Problems Section.

Notes

  • Discussion here: [1]
  • Written by:tseliot
  • Archived by: tseliot
  • Last Revision: 22:58, October 21 2006 (GMT+1)