DIGIT-ARC read-only archive · snapshot: July 23, 2026

nvidia driver in fc3

8 posts · originally from https://geek.digit.in/community/threads/nvidia-driver-in-fc3.5124/

V
venky@linux Original post #1
Jan 27, 2005 at 7:35 PM
Hi,
When i installed the nvidia driver in fc3, and reboot my machine, it says that it cannot load the nvidia kernel module; and x doesnt start. when i login as root and do modprobe nvidia, then startx, it starts up. i however cant do this every time. how to avoid this?
I have a geforce fx 5200.
Thanks in advance.
GNUrag
GNUrag #2
Jan 27, 2005 at 7:51 PM
I dunno what commands you're giving to get X started, but you might want to have a look at your /etc/modules.conf file. This file controls to behavior of loadable kernel modules.

$ man modules.conf for more help.
P
Jan 27, 2005 at 9:50 PM
I too have a gainward 5200 card on FC3. To install drivers just download the latest script file from the nvidia web site turn of x server and run the shell script file it will build the kernel drivers and all modify the x config file as told in the read me file and start x run glxgears to see the effects
V
venky@linux #4
Jan 27, 2005 at 10:34 PM
i used to boot into runlevel 3, and i used to give startx command to run Xwindows

well, i dont know exactly what these two lines do, but they fixed my problem:
cp -a /dev/nvidia* /udev/devices/
chown root.root /etc/udev/devices/nvidia*
(googled for this)

But only partially. X started , but now, on trying to run tuxracer, it gave the error about permissions restricted or something when logged in as normal user. a consultation of the nvidia faq told me to remove a line from console.perms which said <dri> something. now everythings running.

would anyone be kind enough to explain what the above lines did?

Thanks.....
I
indro #5
Feb 15, 2005 at 2:40 AM
Nvidia drivers for FC3 needs to be recompiled ,assuming that you passed that , all you need to do is update the conf file which is located in etc\X11\xfree86.conf or something like that . If you are not sure which one is your conf file , check the system log and look for the lines starting with ==
Then in the section where it says "VESA" or "NV" edit it with "nvidia" and in the modules sections you have to change some settings as well , you will get the precise information from the nvidia website only , and all you need to do is init 3 and then startx and you should be ready to rock with opengl power :)
P
Feb 15, 2005 at 8:02 AM
intro get your "slash" right it pains me to see a "windows" guy trying his hand at linux. Linux used the forward slash not the reverse slash to seperate the path remember.... Also the configration file in fedora core 3 is /etc/X11/xorg.conf as Xfree86 is no longer a freeware. Well the settings in detail are as under
Section "Module"
# Load "GLcore"
Load "glx"
# Load "dri"
Load "nvidia"
# 'There are certainly more modules booked here !'
EndSection

Note: Make sure you commented out the 'dri' module, because nvidia already offers this module.
Device Section

Now Xserver loads the libraries and the driver into it's memory. But we need more. It should also access the nvidia hardware. So search the Device Section in your configfile and change:

Load nv

to

Load nvidia

The whole section should now look like this:

Section "Device"
Identifier "Here might come something ident name !"
Driver "nvidia"
# Maybe coming more (optional) parts
EndSection
GNUrag
GNUrag #7
Feb 15, 2005 at 1:34 PM
pradeep_chauhan said:
intro get your "slash" right it pains me to see a "windows" guy trying his hand at linux.
Forward slash is the right thing. URLs necessarily need to use a forward slash everywhere, even if the URL is for a local file. Windows does not follow standards. (or do they have their own standards?)
Z
zeo #8
Feb 24, 2005 at 2:51 PM
pl suggest me exactly how to disable my x-server.
where do i need to make the changes for that .
also is there anywhere i can download the kernel source rpm for fc3 as downloading the entire srpm disk image is not possible for me because of bandwidth constraints