Linux-x86-1.7184 compiled ok but loading failure

[neilyu]

I download the latest driver NVIDIA-Linux-x86-1.0-7184-pkg1.run for my Nvidia Geforce2 GTS (legacy) video card.
Compile procedure was passed with no-error, But loading the kernel module It reported:

NVRM: request_mem_region failed for 496M @ 0xe0ffffff. This can 
NVRM: occur when a driver such as rivatv is loaded and claims 
NVRM: ownership of the device's registers. 
nvidia: probe of 0000:01:00.0 failed with error -1 
NVRM: The NVIDIA probe routine failed for 1 device(s). 
NVRM: None of the NVIDIA graphics adapters were initialized! 
FATAL: Error inserting nvidia (/lib/modules/2.6.11-1.1369_FC4/kernel/drivers/video/nvidia.ko): No such device

The kernel I used is the offical FC4 release version ( 2.6.11-1.1369_FC4)

I try append

"video=riva:off vga=normal"

to kernel when boot, But it's still had this error.

"/var/log/nvidia-install.log" is passed also giving apart below

NVIDIA: left KBUILD. 
-> done. 
-> Kernel module compilation complete. 
-> Installing both new and classic TLS OpenGL libraries. 
-> Parsing log file: 
-> done. 
-> Validating previous installation: 
-> done. 
-> Uninstalling NVIDIA Accelerated Graphics Driver for Linux-x86 (1.0-7184): 
-> done. 
-> Uninstallation of existing driver: NVIDIA Accelerated Graphics Driver for Linux-x86 (1.0-7184) is complete. 
-> Searching for conflicting X files: 
-> done. 
-> Searching for conflicting OpenGL files: 
-> done. 
-> Installing 'NVIDIA Accelerated Graphics Driver for Linux-x86' (1.0-7184): executing: '/sbin/ldconfig'... /sbin/ldconfig: /usr/lib/libslang.so.1 is not an ELF file - it has the wrong magic bytes at the start. 

executing: '/sbin/depmod -aq 2.6.11-1.1369_FC4'... 
-> done. 
-> Driver file installation is complete. 
-> Running post-install sanity check: 
-> done. 
-> Post-install sanity check passed. 
-> Shared memory test passed. 
-> Running runtime sanity check: 
-> done. 
-> Runtime sanity check passed. 
-> Installation of the NVIDIA Accelerated Graphics Driver for Linux-x86 (version: 1.0-7184) is now complete. Please update your XF86Config or xorg.conf file as appropriate; see the file /usr/share/doc/NVIDIA_GLX-1.0/README for details.

Complete log {nvidia-installer.log, iomem.dump, dmesg.dump, Xorg.0.log} in attachment, Any advice...

[zander]

NVRM: request_mem_region failed for 496M @ 0xe0ffffff.

This looks like an SBIOS problem, the PCI resource assigned to the GPU looks to be invalid (both base and size).

[neilyu]

Code below is the section of video card in /proc/pci

Bus 1, device 0, function 0: 
     Class 0300: PCI device 10de:0150 (rev 163). 
        IRQ 10. 
        Master Capable. Latency=32. Min Gnt=5.Max Lat=1. 
        Non-prefetchable 32 bit memory at 0xe0000000 [0xe0ffffff]. 
        Prefetchable 32 bit memory at 0xd8000000 [0xdfffffff].

corresponding pcitweak output was:

PCI: 01:00:0: chip 10de,0150 card 1048,0c50 rev a3 class 03,00,00 hdr 00

complete dump of {/proc/pci, pcitweak, scanpci} were saved in attachment...

From your quote, How to set the SBIOS is right ? ( my System BIOS is Award`s )

Originally Posted by zander

NVRM: request_mem_region failed for 496M @ 0xe0ffffff.

This looks like an SBIOS problem, the PCI resource assigned to the GPU looks to be invalid (both base and size).

[netllama]

I think what Zander meant is that this is an SBIOS bug. You should verify that you're using the latest BIOS.

Thanks,
Lonni

[neilyu]

I updated the System BIOS from v1.3 to v3.6 (latest) today, It still report error while loading, output is:

nvidia: no version for "struct_module" found: kernel tainted. 
nvidia: module license 'NVIDIA' taints kernel. 
ACPI: PCI Interrupt 0000:01:00.0[A] -> Link [LNKA] -> GSI 10 (level, low) -> IRQ 10 
NVRM: request_mem_region failed for 496M @ 0xe0ffffff. This can 
NVRM: occur when a driver such as rivatv is loaded and claims 
NVRM: ownership of the device's registers. nvidia: probe of 0000:01:00.0 failed with error -1 
NVRM: The NVIDIA probe routine failed for 1 device(s). 
NVRM: None of the NVIDIA graphics adapters were initialized!

[neilyu]

I uploaded the nvidia-bug-report.sh generated nvidia-bug-report.log, Who can help me !

[Cambo]

One thing seems to leap out at me;

-> Installing 'NVIDIA Accelerated Graphics Driver for Linux-x86' (1.0-7184):
executing: '/sbin/ldconfig'...
/sbin/ldconfig: /usr/lib/libslang.so.1 is not an ELF file - it has the wrong
magic bytes at the start.

Looks like you possibly have a corrupt file there. You will probably need to track down which package that library file belongs to, and reinstall it.



Cambo

[netllama]

neilyu,
This looks like a BIOS bug. As MSI makes your motherboard, that's where you'd be best to turn for help in fixing this problem.

Thanks,
Lonni

[neilyu]

I contacted with motherboard's manufacturer, They represent no longer to suppot my old motherboard explicitly.

I ( I'm not a real hacker ) try to manully assign the correct PCI base and size to driver, that's let the dirver loaded sucessful ( checked by lsmod ), but errored on "NVRM: isr request failed 0xffffffea", for 0xffffffea is -EINVAL, I checked nv->interrupt_line, the data is a negative number, I consider It should be 10, So I try to set the dev->irq and nv->interrupt_line to 10 manully, but that cause the system to be crash ( black screen ). Instead of all, could I really do that manully ?