nVidia driver claims I have no nVidia card

make · 09-28-2006, 01:11 PM

I have several computers here running SuSE 10.1, but only this one causes trouble. I cannot get the nvidia.ko-module to load. I provide output of several commands next. I have MSI NX7300GT PCI Express.

Code:

# uname -r
2.6.16.21-0.21-default

Code:

# rpm -qa|grep nvidia
x11-video-nvidia-1.0.8762-1
nvidia-gfx-kmp-default-1.0.8762_2.6.16.21_0.8-1
xorg-x11-driver-video-nvidia-6.9.0-46

Code:

# lspci
00:00.0 Host bridge: VIA Technologies, Inc. Unknown device 0336
00:00.1 Host bridge: VIA Technologies, Inc. Unknown device 1336
00:00.2 Host bridge: VIA Technologies, Inc. Unknown device 2336
00:00.3 Host bridge: VIA Technologies, Inc. Unknown device 3336
00:00.4 Host bridge: VIA Technologies, Inc. Unknown device 4336
00:00.5 PIC: VIA Technologies, Inc. Unknown device 5336
00:00.6 Host bridge: VIA Technologies, Inc. Unknown device 6290
00:00.7 Host bridge: VIA Technologies, Inc. Unknown device 7336
00:01.0 PCI bridge: VIA Technologies, Inc. VT8237 PCI bridge [K8T800/K8T890 South]
00:02.0 PCI bridge: VIA Technologies, Inc. K8T890 PCI to PCI Bridge Controller
00:0f.0 IDE interface: VIA Technologies, Inc. VT82C586A/B/VT82C686/A/B/VT823x/A/C PIPC Bus Master IDE (rev 07)
00:10.0 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 90)
00:10.1 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 90)
00:10.2 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 90)
00:10.3 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 90)
00:10.4 USB Controller: VIA Technologies, Inc. USB 2.0 (rev 90)
00:11.0 ISA bridge: VIA Technologies, Inc. VT8251 PCI to ISA Bridge
00:11.7 Host bridge: VIA Technologies, Inc. VT8251 Ultra VLINK Controller
00:13.0 PCI bridge: VIA Technologies, Inc. VT8251 PCI to PCIE Bridge
00:13.1 PCI bridge: VIA Technologies, Inc. VT8251 PCI to PCI Bridge
00:18.0 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] HyperTransport Technology Configuration
00:18.1 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Address Map
00:18.2 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] DRAM Controller
00:18.3 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Miscellaneous Control
02:00.0 VGA compatible controller: nVidia Corporation Unknown device 0393 (rev a1)
04:01.0 Audio device: VIA Technologies, Inc. VIA High Definition Audio Controller
05:09.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8169 Gigabit Ethernet (rev 10)

Code:

# cat /proc/interrupts 
           CPU0       
  0:     727812    IO-APIC-edge  timer
  7:          2    IO-APIC-edge  parport0
  8:          2    IO-APIC-edge  rtc
  9:          0   IO-APIC-level  acpi
 14:      58664    IO-APIC-edge  ide0
 15:      25890    IO-APIC-edge  ide1
193:       9829   IO-APIC-level  uhci_hcd:usb1
201:        126   IO-APIC-level  uhci_hcd:usb2, ehci_hcd:usb5
209:          0   IO-APIC-level  uhci_hcd:usb3
217:          0   IO-APIC-level  uhci_hcd:usb4
225:       6881   IO-APIC-level  HDA Intel, eth0
NMI:          0 
LOC:     727750 
ERR:          0
MIS:          0

Code:

# modprobe nvidia
FATAL: Error inserting nvidia (/lib/modules/2.6.16.21-0.21-default/weak-updates/nvidia.ko): No such device

Code:

# dmesg|tail
nvidia: module not supported by Novell, setting U taint flag.
NVRM: The NVIDIA probe routine was not called for 1 device(s).
NVRM: This can occur when a driver such as rivafb, nvidiafb or
NVRM: rivatv was loaded and obtained ownership of the NVIDIA
NVRM: device(s).
NVRM: Try unloading the rivafb, nvidiafb or rivatv kernel module
NVRM: (and/or reconfigure your kernel without rivafb/nvidiafb
NVRM: support), then try loading the NVIDIA kernel module again.
NVRM: No NVIDIA graphics adapter probed!

Code:

# lsmod|grep riva

Code:

# lsmod|grep nv

On all the other computers this software configuration (this kernel + this driver from the RPMs) works perfectly. With this one I don't even get a picture, if I try to use the nv-driver in xorg.conf (display refresh rate and resolution is right). I have tested 2 GeForce-cards: this 7300 and also a 6200LE.

I am forced to use the slow VESA-driver currently...

bigrigdriver · 09-28-2006, 05:09 PM

Try 'hwinfo --gfxcard'. What does it tell you about the graphics card?

make · 09-29-2006, 12:26 AM

Quote:

Originally Posted by bigrigdriver

Try 'hwinfo --gfxcard'. What does it tell you about the graphics card?

Code:

# hwinfo --gfxcard
34: PCI 200.0: 0300 VGA compatible controller (VGA)
  [Created at pci.300]
  UDI: /org/freedesktop/Hal/devices/pci_10de_393
  Unique ID: B35A.mVAfZDbyuT4
  Parent ID: _Znp.LUQ1QvXEEDB
  SysFS ID: /devices/pci0000:00/0000:00:02.0/0000:02:00.0
  SysFS BusID: 0000:02:00.0
  Hardware Class: graphics card
  Model: "Micro-Star International VGA compatible controller"
  Vendor: pci 0x10de "nVidia Corporation"
  Device: pci 0x0393 
  SubVendor: pci 0x1462 "Micro-Star International Co., Ltd."
  SubDevice: pci 0x0412 
  Revision: 0xa1
  Memory Range: 0xfa000000-0xfaffffff (rw,non-prefetchable)
  Memory Range: 0xd0000000-0xdfffffff (rw,prefetchable)
  Memory Range: 0xf9000000-0xf9ffffff (rw,non-prefetchable)
  I/O Ports: 0xd800-0xd87f (rw)
  Memory Range: 0xfbde0000-0xfbdfffff (ro,prefetchable,disabled)
  IRQ: 10 (no events)
  I/O Ports: 0x3c0-0x3df (rw)
  Module Alias: "pci:v000010DEd00000393sv00001462sd00000412bc03sc00i00"
  Config Status: cfg=new, avail=yes, need=no, active=unknown
  Attached to: #19 (PCI bridge)

Primary display adapter: #34

konsolebox · 09-29-2006, 01:01 AM

run this command

# dmesg | grep NVIDIA

If you see a message saying about nvidiafb loading, no doubt you need to recompile your kernel and exclude nvidiafb from the config.

There are two common drivers for nvidia cards. One is for the framebuffer type and one is for the agp type. Nvidiafb included in the kernel owns the framebuffer type and nvidia from nvidia's site controls the agp type. The two can never run together. I've also asked for more info about the two being able run together but found no answer.

Take a look at this thread if you want more info:
http://www.linuxquestions.org/questi...d.php?t=466176

make · 09-29-2006, 04:21 AM

Quote:

Originally Posted by konsolebox

If you see a message saying about nvidiafb loading, no doubt you need to recompile your kernel and exclude nvidiafb from the config.

No nvidiafb loading. I am using the kernel & RPM-drivers that Novell is recommending. So I highly doubt that I need a kernel recompile, as this is the kernel meant for the nVidia-drivers and this combination works perfectly on several other computers I have running SuSE 10.1

I wonder if this issue could be caused by some strange motherboard-issue?

raska · 09-29-2006, 09:32 AM

why don't you try the lastest nVidia driver? (don't ask me if there is a RPM)

make · 09-29-2006, 02:25 PM

Quote:

Originally Posted by raska

why don't you try the lastest nVidia driver?

Because the NVIDIA-installer almost always compiles a non-working driver. This has happened on about 15 different computers running SuSE 10.1. I only get a white screen when X.org is started, unless I comment out the line "Load glx" from xorg.conf, which makes installing the whole nVidia-driver a total waste of time in the first place. The latest driver does recognize this card, though. But the driver that the official nVidia installer-packet compiles, does not work on 90 % of the SuSE 10.1-computers I've tried it on. And these computers all have a totally different hardware configuration, so it is not a computer-specific problem.

make · 10-14-2006, 06:11 PM

Well, problem solved, kinda.

I switched to Mandriva 2007 and it works flawlessly. The card is detected perfectly and XGL is running very nicely.

tzbishop · 10-15-2006, 10:32 AM

strange. I have NVIDIA drivers working perfectly on SuSE 10.1 but I did not install via RPM. I installed via the binary file from nvidia.com for Linux.

Following the official tutorial, all is right!
http://www.novell.com/coolsolutions/feature/17174.html

make · 10-18-2006, 06:11 PM

Quote:

Originally Posted by tzbishop

I have NVIDIA drivers working perfectly on SuSE 10.1 but I did not install via RPM. I installed via the binary file from nvidia.com for Linux.

Installing the official binary from nVidia's site worked out for about 50 % of the computer I installed SuSE 10.1 on. And the other 50 % failed miserably, with just the white screen on X.org startup, unless 'Load "glx"' was commented out from xorg.conf. All in all, I have installed SuSE 10.1 on something between 30 - 40 computers.

I also noted that X.org- and Mesa-updates always broke my configuration and re-installing the nVidia-drivers never fixed it anymore - without a full system re-install. The issue was the same white screen-thing. I needed to lock specific packages from Smart, so that it did not upgrade them. This included the whole X.org and several other packages.

The advantage of using the RPM-packaged version of the driver is obvious: RPM won't let updates to overwrite the nVidia's files automatically, as they belong to another package. The kernel-update + RPM-version has always worked... until now on this particular computer. Using the official installer caused the white screen-issue also on this computer.

nVidia's drivers seem to be a huge issue on SuSE 10.1. Yes, they work for most computers, but when they don't, it requires hours of battle. I never had this issue on other distros.

BROse · 10-19-2006, 12:01 AM

Did you try "tiny-nvidia-installer -update" as root....

make · 11-04-2006, 04:56 AM

Quote:

Originally Posted by BROse

Did you try "tiny-nvidia-installer -update" as root....

Doesn't that just compile the driver from the sources? Meaning it might not work under SuSE 10.1. Would get me into the same "white screen at X.org startup"-issue.

Well, luckily Mandriva 2007 has a newer nVidia-driver as a RPM-packet.

BROse · 11-04-2006, 11:16 AM

Yes, but it will install he nvidia driver also......