LinuxQuestions.org
Visit Jeremy's Blog.
Go Back   LinuxQuestions.org > Linux Answers > Hardware
User Name
Password

Notices


By HolyLiaison at 2004-02-11 04:13
First off, go to the following site: http://www.nvidia.com/object/linux.html

Download the right Nvidia installer. More than likely it'll be the IA32 file, unless you have a 64bit processor, or FreeBSD. Also, if you're running any kernels released after January 26th 2004, make sure you have the kernel source downloaded and installed correctly. The installer will need it to compile the drivers correctly.

Once you have that all setup you'll want to login as root.

Step 1 - Editing Module section in XF86Config:
Open up /etc/X11/XF86Config in your favorite editor. Make sure to save a backup of the file before editing it. XF86Config.backup, or something easy to remember. Once you've done that, open up /etc/X11/XF86Config again then browse to the section that looks similar to this:

Section "Module"
Load "dbe"
Load "extmod"
Load "fbdevhw"
Load "dri"
Load "record"
Load "freetype"
Load "type1"
EndSection

If you see Load "dri" or Load "GLcore" you'll want to remove them and add Load "glx". After you've done that it should look similar to this:

Section "Module"
Load "dbe"
Load "extmod"
Load "fbdevhw"
Load "glx"
Load "record"
Load "freetype"
Load "type1"
EndSection

Step 2 - Editing Device section in XF86Config:
Once you've got the Modules section setup correctly you'll want to scroll down to the device section, which will look somewhat like this:

Section "Device"
Identifier "Videocard0"
Driver "nv"
VendorName "Videocard vendor"
BoardName "NVIDIA GeForce FX (generic)"
EndSection

All you have to change here is the Driver line. You'll want to change it to look like this:

Driver "nvidia"

Once you've changed the line save your XF86Config file and close it.

Step 3 - Shutting down X Server:
Now we have to shut down the X server so we can install the drivers. This will bring you to a black screen with a terminal prompt. To do this open a terminal and type:

[root@localhost *****]# init 3

Once you've done that you should have a black screen with a terminal prompt. Now browse to the directory you downloaded the Nvidia driver file.

Step 4 - Installing the drivers:
Once you're in the directory you downloaded the Nvidia drivers to
all you have to type is:

[root@localhost *****]# sh NVIDIA-Linux-x86-1.0-****-pkg1.run

The install should start, it'll run you through everything. If everything was setup and installed correctly it should remind you to edit your XF86Config. But you shouldn't have to do this, since we already did it. After that it should return you to the Terminal prompt.


Step 5 - Restarting X Server:
At the Terminal prompt type the following:

[root@localhost *****]# init 5 (note: this will be init 4 for some distros, such as Slackware)

It should now boot back in to X Server, greeting you with a Nvidia logo. If it doesn't, use your backup of XF86Config and boot back in to X and figure out what you did wrong.

If you find something wrong with this How-To please tell me so I can change it. All my contact info is available in my profile.

by General_Tso on Thu, 2004-06-03 21:48
Is this different with Fedora Core 2 and X.org? The "init 3" command didn't work on my machine, so I figured I'd just leave well enough alone until I read some more about it.

--Tso

by GufyMike on Fri, 2004-06-04 00:59
General when trying to boot into runlevel 3 in FC2 I find that hitting e and editting the kernel argument in GRUB before booting and adding 3 to the end of it it will boot to CLI without X

kernel /boot/vmlinuz-2.6.6 ro root=LABEL=\

add the 3 to the end so it looks like this:

kernel /boot/vmlinuz-2.6.6 ro root=LABEL=\ 3

Then boot.

by General_Tso on Fri, 2004-06-04 09:09
Oh, cool. Thanks! Any other divergences from the tutorial for those of us with X.org?

Thanks again,

--Tso

by GufyMike on Fri, 2004-06-04 09:18
Well for FC2 its not that simple, you need to recompile a non distro specific kernel and make sure #4KSTACKS and REGPARM are set to NO.

Then install the drivers as HolyLiaison says in his How-To.


For more on compil;ing and installing nvidia drivers on FC2 check this out.
http://www.linuxquestions.org/questi...hlight=REGPARM

by General_Tso on Fri, 2004-06-04 10:16
Excellent. That'll be my project when I get home from work today. Fedora Core 2 seems faster than FC1 to me, but the display seemed a bit sluggish so I thought the drivers would pep things up.

Thanks again!

--Tso

by penguin4 on Wed, 2004-06-09 13:12
HL,GT&GM, Thanks Guys, teaching & learning is what makes any linux distro
the only way to go. learned on W95-W98 wasted time, having to turn to
msw spport for assistance, even then not always correct. they were looking
at books with ms kernel and protocol. oh well 1 given did learn to operate
pc not all lost cause. But wasted time Yes!

by adonai on Sat, 2004-06-12 10:24
hi, I just got SuSE 9.0 and I installed the package from http://nvidia.com, it is a package updated on Feb 20, 2004.
I installed it and entered used the #sax2 -m 0=video command
my graphic card is nvidia geforce mx440 and it doesn't work, it is just as it was before, there is a message telling me to download and install the driver from nvidia.com...
I checked the XF86Config file and it's ok
still i can't enable the graphical acceleration..

by bootsy on Sat, 2004-06-26 01:26
Pls be nice - newbie having lotsa trouble.

Just been trying this install procedure (for about the last 5hours). When I run the NVIDIA-Linux-x86-1.0-4620-pkg1.run

ERROR: Unable to find the system utility `ld`; please make sure you have the
package 'binutils' installed. If you do have binutils installed,
then please check that `ld` is in your PATH.

Ok, so fair enough - I checked and can't find binutils. So I downloaded binutils-2.15.tar.gz and got to work on it. When I go to configure and make I get this error
[root@localhost binutils-2.15]# ./configure

loading cache ./config.cache
checking host system type... i686-pc-linux-gnuoldld
checking target system type... i686-pc-linux-gnuoldld
checking build system type... i686-pc-linux-gnuoldld
checking for a BSD compatible install... /usr/bin/install -c
./configure: line 2343: cc: command not found
*** The command 'cc -o conftest -g conftest.c' failed.
*** You must set the environment variable CC to a working compiler.

So being a good little noobie I went to the Readme and found something that said to try
"CC=gcc ./configure
make

A similar example using csh:

setenv CC gcc
./configure
make"

[root@localhost binutils-2.15]# CC=gcc ./configure
loading cache ./config.cache
checking host system type... i686-pc-linux-gnuoldld
checking target system type... i686-pc-linux-gnuoldld
checking build system type... i686-pc-linux-gnuoldld
checking for a BSD compatible install... /usr/bin/install -c
./configure: line 2232: gcc: command not found
./configure: line 2343: gcc: command not found
*** The command 'gcc -o conftest -g conftest.c' failed.
*** You must set the environment variable CC to a working compiler.


I'm dying here. Anyone who can help - it would be greatly appreciated. Does this mean I don't have something called gcc (yes, you can laugh at what a silly noobie I am) and if so how do I get it.


My system specs are:
Intel P4 3g, nvidia GeForce FX5700, M-audio audiophile 2496, 1g RAM running Fedora Core 1.

by penguin4 on Sat, 2004-06-26 15:34
adonai & bootsy; newbie too but have either or both tried the ldp site. this one:http://www.tldp.org/HOWTO/HOWTO-INDEX, scroll through till 4.3hardware then4.3.3 video cards/4.3.9 graphic tablets,on to #5 single list
of HowTos scroll through till finding help there. sorry lots of work but learning is valuable! not easy like Ws but this is not W!?

by GreenBanditAtWo on Tue, 2004-07-06 08:21
Just in case this helps anyone else, my system locked up & totally black screen froze when I tried to restart the GUI & the problem was due to AGPGART incompatabilities.

The solution was to edit the /etc/X11/XF86Config-4 file & add the line
Option "NvAgp" "0"
in the same section as the line
Driver "nvidia" (this line read Driver "nv" prior to editing it as per the nVidia instructions).

If at this point you can't get into the GUI, you might have to use non-GUI editirs such as emacs or vi.

On my system, the relevant part of the XF86Config-4 file is now

Driver "nvidia"
Option "DPMS
Option "NvAgp" "0"
Endsection

You could also try
Option "NvAgp" "1" to get some level of AGP support

I have a SiS741GX chipset & Option "NvAGP" "1" worked for me, but I understand that it might not work for all chipsets.


The rationale behind this is:-
Look at Appendix F of the nVidia README file (on the same page as where their driver is downloaded, currently
http://www.nvidia.com/object/linux_d..._1.0-5336.html
If your chipset is not listed, try what I described above.


  



All times are GMT -5. The time now is 01:23 PM.

Main Menu
Advertisement
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Open Source Consulting | Domain Registration