Mandrake 9 NVIDIA Problems

Hey Guys,

Let me start off by saying this is my first post to the board, hopefully the first of many. I know how annoying it is when people ask repetitive questions without searching and while mineis a repetitive question, I couldnt find much matching my circumstance.... mainly because of how new I am to this.

this is probably my third run in with linux but usually it was a dual boot system and I never really used the linux unless i was bored but this time my g/f accidentaly lost my windows partition so i decided its here to stay :)

My situation...

I am using Mandrake 9.0 with blackbox scheme deal, I started with kde but got bored quickly.

I am using the default driver for my Geforce3 Ti200, I used the NVchooser.sh and got those two rpms...

I logged out and went to a root command promt and ran the two rpms....

but from there I dont really know what to do... some of the other posts seem a little diffucult to understand...

but when I try to run X I get an error and a lockup so because of the lack of skills, I find myself reinstalling :)

now my question is, what exactly should i be doing after I install these two rpms and also...

how necessary are these drivers... my 2d is fine and I have yet to test my 3d... is that where the drivers are necessary?


thanks guys...

if someone knows a link I will not be offended with just a simple url reply so feel free :)

thanks
-greg

if you look on nvidia's website. www.nvidia.com

there is a txt file which explains what you need to do to your system before you rerun X.

basically you change your gfx driver within XF86CONFIG-4 to nvidia from nv

save and then rerun X

should come up with an nvidia splash screen which shows the nvidia driver works.

hth


dvdnut

ok so I am following the instructions to the T....

Im using NVchooser, downloading the specified files including the sources.

I am exiting and running nix at runlevel 3

I am rpm --rebuild ing the sources rpm's

then rpm -Uvh the drake specific kernel rpm

then rpm --rebuild ing the GLX source rpm

then -Uvh the source i38r rpm.

I then vi into the XF86confoig-4 and change the driver to nvidia, the load "glx" is already specified and the other two load's (for deletion) are not in there.

I then go and add the two alias's into the etc/modules.conf

then go run depmod -ae in my /lib/modules/<kernel version>

I get no visible errors at anytime... Im new to vi so I could be screwing up there... but when i exit and come back into vi, it looks like all my changes have been made.

I do all of this in root.

when i try to restart x in runlevel 5 it loads up again with no errors, the screen stays black and blinks like its changing res a few times.... about 4 minutes later it shows me the nvidia logo then flashes black a few times.... i left it like this for about 4 hours today to see if anything would happenn and nope... nothing.

athlon 2200
mandrake 9
full install

i really would like to get this working after prolly 6 attempts and 4 reinstalls

also.... if it screws up how can i go back to my old configuration.

all help is appreciated.

-greg

I would first advise you to change the default runlevel in /etc/inittab from '5' to '3' so you boot into text mode just for the Nvidia driver installation and testing purposes.

When that has been done and you reboot into text mode, run

startx

and see what errors you get.

Also

grep EE /var/log/XFree86.0.log

what do i do to revert back to my old settings so that when it breaks i can come back and post results?

also i was only in runlevel 3 for install and testing...

-g

change nvidia back to nv and then X will load with the nv generic driver.

okiedoke....

im gonna go give ti a shot...

hopefully not in for another headache.... if so i will be commenting later from the library :)

this time i am going to try and add alias char-major-195 nvidia to the etc/modules instead of the other two i was adding....

just an idea.

-g

good news is im back in X

bad news is, its with the old drivers still...

ok this time i -ivh'd my two rpm's then edited the XF86Config-4 to change nv to nvidia. then added

alias char-major-195 nvidia
alias char-major-10-175 agpgart

to the modules.conf

then did a depmod -ae (what does this do?) to the kernel directory.

tryied to boot and it just sits.... trying to chhange resolution or something.

Rebooted, went back into my XF86config-4 and changed the driver back to nv and the glc to glcore and im back.

now...

the errors...

when i did grep EE /var/log/XFree86.0.log

it didnt give me any errors or anything....

and because when i type startx it goes to a black screen super fast... there is no way to see any errors....

i think it is trying to work, is there anything i could possibly be leaving out?

should I try to install from the tarballs maybe?

thanks
-g

ok so

they sort of work now...

i rpmbuilded them and im in kde with the nvidia drivers...

however is it normal for the loadup process to take about 5 minutes?

is there something I can do to fix this...

its about 5 minutes before the nvidia splash screen comes up.... im thinking maybe this whole time i didnt wait around long enough.

thanks for all your help.


-greg

The popular response for Linux taking too long to boot or apps taking too long to startup would be to ensure you have localhost defined correctly in /etc/hosts

127.0.0.1 localhost.localdomain localhost <your hostname>

if you are using dhcp.

Since X is now starting with the nvidia drivers, again have a look for errors.

grep EE /var/log/XFree86.0.log

less /home/<user>/.xsession-errors

local hosts looks fine....

it just started doing this after yhe nvidia installation.

when i type in that grep command heres what it spits out at me....


(WW) Warning, (EE) Error, (NI) Not Implemented, (??)Unknown, (II) Loading Extension MIT-Screen Saver

and thats it.... sorry if im doing something wrong...

not too familiar with grep.

-greg

10
DCOPServer up and running.
20
25
kdecore (KLibLoader): WARNING: library=/usr/local/netscape/plugins/raclass.zip: file=/usr/local/netscape/plugins/raclass.zip: /usr/local/netscape/plugins/raclass.zip: invalid ELF
header
kdecore (KLibLoader): WARNING: library=/usr/local/netscape/plugins/rpnp.so: file=/usr/local/netscape/plugins/rpnp.so: /usr/local/netscape/plugins/rpnp.so: undefined symbol: __pur
e_virtual
30
36
41
kdecore (KLibLoader): WARNING: library=/usr/lib/netscape/plugins/raclass.zip: file=/usr/lib/netscape/plugins/raclass.zip: /usr/lib/netscape/plugins/raclass.zip: invalid ELF heade
r
kdecore (KLibLoader): WARNING: library=/usr/lib/netscape/plugins/rpnp.so: file=/usr/lib/netscape/plugins/rpnp.so: /usr/lib/netscape/plugins/rpnp.so: undefined symbol: __pure_virt
ual
47
52
58
63
69
75
80
85
90
ICE default IO error handler doing an exit(), pid = 2200, errno = 2
kdeinit: Communication error with launcher. Exiting!
ICE default IO error handler doing an exit(), pid = 2202, errno = 0
_KDE_IceTransmkdir: Owner of /tmp/.ICE-unix should be set to root
DCOPServer up and running.
ICE default IO error handler doing an exit(), pid = 2654, errno = 32
Mutex destroy failure: Device or resource busy
Mutex destroy failure: Device or resource busy
mcop warning: user defined signal handler found for SIG_PIPE, overriding
_IceTransmkdir: Owner of /tmp/.ICE-unix should be set to root
mcop warning: user defined signal handler found for SIG_PIPE, overriding
KDirWatch: /etc/security/fileshare.conf is a file. Use addFile!
WARNING: KDE detected X Error: BadDrawable (invalid Pixmap or Window parameter) 9
Major opcode: 14
QWidget::setMinimumSize: The smallest allowed size is (0,0)
QWidget::setMaximumSize: (unnamed/DigitalClock) Negative sizes (2,-4) are not possible
QWidget::setMinimumSize: The smallest allowed size is (0,0)
QWidget::setMaximumSize: (unnamed/DigitalClock) Negative sizes (2,-4) are not possible
WARNING: KDE detected X Error: BadDrawable (invalid Pixmap or Window parameter) 9
Major opcode: 14
Invalid entry (missing '=') at /usr/bin/evolution:1
Invalid entry (missing '=') at /usr/bin/evolution:2
Invalid entry (missing '=') at /usr/bin/evolution:3
Invalid entry (missing '=') at /usr/bin/evolution:4
Invalid entry (missing ']') at /usr/bin/evolution:7
Invalid entry (missing ']') at /usr/bin/evolution:10
Invalid entry (missing ']') at /usr/bin/evolution:11
Invalid entry (second locale!?) at /usr/bin/evolution:12
Invalid entry (missing '=') at /usr/bin/evolution:16
Invalid entry (missing '=') at /usr/bin/evolution:17
Invalid entry (missing ']') at /usr/bin/evolution:18
Invalid entry (missing ']') at /usr/bin/evolution:19
Invalid entry (missing '=') at /usr/bin/evolution:20
Invalid entry (missing '=') at /usr/bin/evolution:22
Invalid entry (missing '=') at /usr/bin/evolution:23
Invalid entry (missing '=') at /usr/bin/evolution:24
Invalid entry (missing '=') at /usr/bin/evolution:25
Invalid entry (missing '=') at /usr/bin/evolution:26
Invalid entry (missing '=') at /usr/bin/evolution:27
Invalid entry (missing '=') at /usr/bin/evolution:28
Invalid entry (missing '=') at /usr/bin/evolution:29
Invalid entry (missing '=') at /usr/bin/evolution:30
Invalid entry (missing '=') at /usr/bin/evolution:31
/home/gregory/.xsession-errors lines 1-63/2605 2%

thats what i get.

when i do the less command.