GeForce4 Ti 4200 Redhat 8.0 fresh install question

[Tom1976]

I have a problem.

First, the facts:
Athlon XP 2100+
ASUS A7V333 mobo
PNY GeForce4 Ti 4200 64MB
1GB PC2700 DDR SDRAM
Redhat 8.0
Kernel 2.4.18-14

I ran NVChooser.sh which returned:
NVIDIA_kernel-1.0-4191.rh80up.athlon.rpm
NVIDIA_GLX-1.0-4191.i386.rpm

So I downloaded them both. Followed the README.
rpm -ivh NVIDIA_kernel-1.0-4191.rh80up.athlon.rpm
rpm -ivh NVIDIA_GLX-1.0-4191.i386.rpm

Changed the /etc/X11/XF86Config
deleted "Load dri"
deleted "Driver vesa"
added "Driver nvidia"
(Attached XF86Config file below for reference)

restarted X. Everything came up as expected. Saw the nVidia logo screen. Started tuxracer..... everything was flashing then it crashed and froze X.

Tried it a couple times same thing. So rpm -e the two NVIDIA rpm's. Downloaded the source rpm's. Again, followed the instructions in the readme and built and installed from the source RPM's. Didn't change the XF86Config file.

restarted X. Same thing. Went into tuxracer, all the rendered objects flickered, crash, froozen X.

Here is the kicker.... this was working fine yesterday on the same machine. I've been playing ut2k3 on linux for days with the 4191 drivers. This morning I reinstalled RH8 because I wanted to get rid of my Windows partition.

What changed?!?!?!?!

Here is what my XF86Config looks like:
------------------------------------------------------

# File generated by anaconda.

Section "ServerLayout"
Identifier "Anaconda Configured"
Screen 0 "Screen0" 0 0
InputDevice "Mouse0" "CorePointer"
InputDevice "Mouse1" "SendCoreEvents"
InputDevice "Keyboard0" "CoreKeyboard"
EndSection

Section "Files"
# The location of the RGB database. Note, this is the name of the
# file minus the extension (like ".txt" or ".db"). There is normally
# no need to change the default.
# Multiple FontPath entries are allowed (they are concatenated together)
# By default, Red Hat 6.0 and later now use a font server independent of
# the X server to render fonts.
RgbPath "/usr/X11R6/lib/X11/rgb"
FontPath "unix/:7100"
EndSection

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

Section "InputDevice"
# Option "AutoRepeat" "500 5"
# when using XQUEUE, comment out the above line, and uncomment the
# following line
# Option "Protocol" "Xqueue"
# Specify which keyboard LEDs can be user-controlled (eg, with xset(1))
# Option "Xleds" "1 2 3"
# To disable the XKEYBOARD extension, uncomment XkbDisable.
# Option "XkbDisable"
# To customise the XKB settings to suit your keyboard, modify the
# lines below (which are the defaults). For example, for a non-U.S.
# keyboard, you will probably want to use:
# Option "XkbModel" "pc102"
# If you have a US Microsoft Natural keyboard, you can use:
# Option "XkbModel" "microsoft"
#
# Then to change the language, change the Layout setting.
# For example, a german layout can be obtained with:
# Option "XkbLayout" "de"
# or:
# Option "XkbLayout" "de"
# Option "XkbVariant" "nodeadkeys"
#
# If you'd like to switch the positions of your capslock and
# control keys, use:
# Option "XkbOptions" "ctrl:swapcaps"
#Option "XkbOptions" ""
Identifier "Keyboard0"
Driver "keyboard"
Option "XkbRules" "xfree86"
Option "XkbModel" "pc105"
Option "XkbLayout" "us" #Option "XkbVariant" ""
EndSection

Section "InputDevice"
Identifier "Mouse0"
Driver "mouse"
Option "Protocol" "IMPS/2"
Option "Device" "/dev/psaux"
Option "ZAxisMapping" "4 5"
Option "Emulate3Buttons" "no"
EndSection

Section "InputDevice"
Identifier "Mouse1"
Driver "mouse"
Option "Device" "/dev/input/mice"
Option "Protocol" "IMPS/2"
Option "Emulate3Buttons" "no"
Option "ZAxisMapping" "4 5"
EndSection

Section "Monitor"
Identifier "Monitor0"
VendorName "Monitor Vendor"
ModelName "Dell P990"
HorizSync 30.0 - 96.0
VertRefresh 48.0 - 120.0
Option "dpms"
EndSection

Section "Device"
# no known options
#BusID
Identifier "NVIDIA GeForce 4 (generic)"
Driver "nvidia"
VendorName "NVIDIA GeForce 4 (generic)"
BoardName "NVIDIA GeForce 4 (generic)"
EndSection

Section "Screen"
Identifier "Screen0"
Device "NVIDIA GeForce 4 (generic)"
Monitor "Monitor0"
DefaultDepth 16
SubSection "Display"
Depth 16
Modes "1024x768" "800x600" "640x480"
EndSubSection
EndSection

#Section "DRI"
# Mode 0666
#EndSection

[Tom1976]

UPDATE:

I upgraded my RH kernel to 2.4.18-19.8.0.

Built the kernel and GLX packages from the source RPM's. And still.... when I test the 3D acceleration in tuxracer, it flickers, then freezes X.

I can't be the first person to encounter this problem.....

[junkieclown]

Questions: What brand of RAM are you using?
(the asus manual mentions 3 kinds that are tested and verified by them. I am not using one of those 3)
Do you have a Windose system you can try the card with to see if it's OK?
(the asus manual mentions that putting the video card in or taking it out when the system is plugged in at all can damage the card. I think that's what the green diode on the MB is for. If the crashing behavior follows the card when used in a windows system, then there's been damage to the card.)
Is the Nvidia card sharing an interrupt with anything else?
I disabled USB2.0 on the MB to free the IRq for Nvidia's exclusive use -- still no joy. But maybe you have more stuff attached to your system than I do.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

My setup:
Asus A7v333 (non-raid)
AMD 2100+ cpu
512 mb Corsair XMS pc2700

Gainward ti4200 gf4 64mb ddr
-or-
Aopen MX200 gf2 64mb sdram

intel etherexpress 100 pro nic
enermax whisper 350 watt PS
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
With the GF4 ti4200, any OpenGL game crashes a few seconds after starting. The screen freezes with large blue and greenish blocky chuncks like a giant pixelized version from whatever frame it was just rendering. "Algae-vision".

I am using RH8.0 --have upgraded kernel to 2.4.20-ac2 to get latest chipset support.
Latest Nvidia driver + glx files (1.0-4.x). Haven't tried any previous versions like 1.0-3.x

I have updated my Asus BIOS to the latest stable version, with no change.

Using the Gforce 2 card, everything is fine. But slow rendering. Drop the Gforce4 in and no matter what I try I get the same crashing error-- chunky frozen screen in OpenGL, but X is otherwise ok in 2d. I planned to test the card with Windose tomorrow on a different system to see if it was defective. That is looking less likely, as you seem to have the same problem.

I Cannot find any BIOS options or driver options to stabilize the OpenGL operation. There is Agp drive strength in the BIOS but I don't know what I'd be doing there at all and I read that you can make your system unbootable that way. Hairy.

Actually I haven't tested turing off AGP altogether with the GF4. (What would be the point of a GF4 that didn't work on a AGP bus?)

Any Nvidia people want to comment about this?

[Tom1976]

I'm using 2 sticks of 512MB Samsung RAM.

http://usa.asus.com/mb/socketa/a7v333/specification.htm

This page has a table at the bottom that lists qualified RAM. The 128 and 256 Samsung RAM is listed, but the 512 is not.

I'm using the most up to date BIOS, version 1015.

I'm going to install Windows 2000 on this box now and see what happens.

I don't know if the card is sharing an interrupt. Where can I check? I have the G4 in the AGP slot and I have a modem in the bottom PCI slot, and that's it as far as cards go in this box.

My mobo is also the non-RAID version.

Tom

[junkieclown]

See interrupts quickly:
# cat /proc/interrupts

Here's mine:
CPU0
0: 42568985 XT-PIC timer
1: 225415 XT-PIC keyboard
2: 0 XT-PIC cascade
4: 1581644 XT-PIC usb-uhci, usb-uhci
8: 1 XT-PIC rtc
10: 9045138 XT-PIC cmpci
11: 36467541 XT-PIC nvidia
12: 462082 XT-PIC eth0
14: 209595 XT-PIC ide0
15: 7 XT-PIC ide1
NMI: 0
LOC: 42569165
ERR: 28856
MIS: 0

Before I disabled USB2.0 jumpers, I would see 'ehci' as sharing irq11 with nvidia. Sharing the interrupt between the nvidia card and something else supposedly is one thing that can cause crashes, but when using the Ti-4200 GF4 card I have what seems to be an identical crash problem, whether the irq is shared or not.

one other thing about my setup: I have tried turning the apic interrupt controller off in my bios, but it doesn't seem to affect the crashing. gf4 crashes, gf2 does not. My present kernel is built to support local apics, I don't know whether RH packaged kernels do, but either way I can't tell if it makes a difference. What Bios settings variations have you tried?

I would really appreciate some input from Nvidia, since they have my money, several times now.

PS: did i read you correctly as saying that you initially installed rh8.0 with the ti-4200 in place and opengl worked with the nvidia 1.0-4xxx driver?

[Tom1976]

Here is what I get from running cat /proc/interrupts

CPU0
0: 1259321 XT-PIC timer
1: 1984 XT-PIC keyboard
2: 0 XT-PIC cascade
5: 117573 XT-PIC usb-uhci, serial
6: 0 XT-PIC usb-uhci, usb-uhci
8: 1 XT-PIC rtc
10: 0 XT-PIC ehci-hcd, cmpci
11: 207302 XT-PIC usb-uhci, nvidia
12: 82134 XT-PIC PS/2 Mouse
14: 11672 XT-PIC ide0
15: 47917 XT-PIC ide1
NMI: 0
ERR: 0


I haven't touched anything in my BIOS other the intial prompting for choosing the FBS speed (I'm not OCing), and changing my drive boot order from it's default.

[Tom1976]

PS: did i read you correctly as saying that you initially installed rh8.0 with the ti-4200 in place and opengl worked with the nvidia 1.0-4xxx driver?

This is what happened. I originally had an Athlon XP 1700+ with 256MB of Crucial PC2100 DDR SDRAM in this rig. Everything was working fine with the 4191 drivers. On Friday I upgraded to the 2100+ and the 1G of Samsung, didn't change anything in software, rebooted, and everything came up fine. Linux saw and reported in the proc directory the new chip and ram. Was able to play UT2003 just fine all friday night and saturday morning.

Then saturday afternoon I decided to blow away my windows partition and install RH8 over my entire HD. That's how I we came to be where we are now....

[junkieclown]

and presumably you followed the same procedures as before to install the Nvidia drivers and configure your XF86Config.....

I am impressed. I had a problem that merely seemed the same no matter what options I tried. You have a similar problem that has appeared where once there was working hardware & setup...

I am so stumped now, I don't know whether to **** or go blind.

You were using RH8.0 before as well, right, when it was working for UT2003?

Did you ever have a boot option like mem=nopentium for grub or lilo before your reinstall?
I don't think that's supposed to be necessary with recent kernels but...


Here's another thing: there's no NvAGP option set in your XF86Config.
You could try it different ways to see if one avoids the crashing.
"NvAGP" "0" #equals no agp of any kind
.............."1" #equals , try nvagp if possible
.............."2" # equals, try Linux agpgart if possible
.............."3" #equals, try agpgart and fallback to nvagp if unsuccessful.

default is "3" if nothing is specified.
Since you aren't specifying anything, then it should try agpgart. But I don't know if there is agpgart support for Kt333 chipsets in the kernels RH ships. I can't really say if there is any in the Nvidia driver either. What I've seen in Nvidia's documentation says Kt266. However, my 2.4.20-ac2 kernel identifies the 333 chipset as Kt266Pro at least for purposes of AGP.
cat /proc/driver/nvidia/agp/host-bridge
Host Bridge: Via Apollo Pro KT266
Fast Writes: Supported
SBA: Supported
AGP Rates: 4x 2x 1x
Registers: 0x1f000217:0x00000104


It seems to work with Nvidia's agp module :::

cat /proc/driver/nvidia/agp/status
Status: Enabled
Driver: NVIDIA
AGP Rate: 4x
Fast Writes: Disabled
SBA: Disabled

(snipped from /var/log/XFree86.0.log)
(II) NVIDIA(0): Setting mode "1600x1200"
SwitchToMode - Succeeded
(II) Open APM successful
(II) NVIDIA(0): AGP 4X successfully initialized
(II) NVIDIA(0): Setting mode "1600x1200"
(II) NVIDIA(0): Using the NVIDIA 2D acceleration architecture
(**) NVIDIA(0): DPMS enabled
(II) Loading extension NV-GLX
(II) [GLX]: Initializing GLX extension
(II) Open APM successful
(II) NVIDIA(0): AGP 4X successfully initialized
(II) NVIDIA(0): Setting mode "1600x1200"
(II) NVIDIA(0): Using the NVIDIA 2D acceleration architecture
(**) NVIDIA(0): DPMS enabled
(II) Loading extension NV-GLX
(II) [GLX]: Initializing GLX extension
(II) Open APM successful
(II) NVIDIA(0): AGP 4X successfully initialized
(II) NVIDIA(0): Setting mode "1600x1200"
(II) NVIDIA(0): Using the NVIDIA 2D acceleration architecture
(**) NVIDIA(0): DPMS enabled
(II) Loading extension NV-GLX
(II) [GLX]: Initializing GLX extension
(II) Open APM successful
(II) NVIDIA(0): AGP 4X successfully initialized
(II) NVIDIA(0): Setting mode "1600x1200"
(II) NVIDIA(0): Using the NVIDIA 2D acceleration architecture
(**) NVIDIA(0): DPMS enabled
(II) Loading extension NV-GLX
(II) [GLX]: Initializing GLX extension

It's very perplexing and i'm getting the distinct feeling that neither Linux kernel people nor Nvidia give much of a damn about AMD cpus and VIA boards, as they're always having trouble with them, and are typically very late getting things working right.
Maybe it won't crash on NvAGP "0" (won't that be fun?)

[Tom1976]

Installed Windows 2000 Pro on this box.
Installed the latest windows nvidia drivers from nvidia's site.

Ran 3DMark2001 SE successfully (scored 10233).
Ran UT2003 under Win2000 successfully (played for about an hour).

This leads me to believe it's not a hardware problem.

So now I'm back to square one. I just reinstalled RH8.0 again (kernel 2.4.18-14).



I'm going to try again with the 4191 driver. If that doesn't work, I'll try with older drivers.

nd presumably you followed the same procedures as before to install the Nvidia drivers and configure your XF86Config.....

I didn't actually do anything between hardware upgrades. That is, the 4191 driver was working fine with the old proc and ram. I popped in the new proc and ram and just started the machine. Redhat didn't seem to mind, and I just kept playing. It wasn't until after I reinstalled after the hardware upgrade that the problems came about.

[junkieclown]

Well, Option "NvAGP" "0" crashes on my setup, same as always.
If it works for you though, that might be significant: like maybe I have a damaged card and you don't. I kinda doubt it, but still...
(People talk about these cards overheating, but a couple days ago I tried opening the case and directing a very powerful fan right onto the mobo... I could see a major diff in cpu/mobo temp but no difference in the crashing. If anything it crashed quicker.)

The one major difference between our systems is: my Ti-4200 has never worked in OpenGL without crashes seconds into launching the opengl app.
Yours used to work. I can't account for that.

2morrow I will try a GF3 on this system, and my ti4200 on win98 and an sis735 just to make sure it's alright. I already know in my heart it will work fine --just like yours does in W2K.

I'm getting pretty fed up with Nvidia Linux "support".


BTW: what kind of powersupply are you using?
Not that it really matters since it's clear that the card and motherboard work just fine together under Windows.

[Tom1976]

Antec 350W

[Tom1976]

RH8's default game package comes with a game called chromium. According to the faq on the chromium website, it's hardware accelerated

http://www.reptilelabour.com/softwar...ium/faq.htm#q5

I'm able to play that game fine. yet, start tuxracer... as we all know... crashes and freezes X.

Just thought I'd throw that out there....