View Single Post
Old 04-29-09, 03:05 PM   #1
RAMChYLD
Registered User
 
Join Date: Feb 2005
Location: Malaysia
Posts: 95
Default Problem with Kernel 2.6.29 and Legacy 96.43.11 drivers

Hi, I'm not sure if this has been reported in the past since my search turned up nothing, but I've been having a hard time with the 96.43.11 drivers since Kernel 2.6.29 was released. I've been receiving the following error messages:

Code:
[   64.850635] nvidia: module license 'NVIDIA' taints kernel.
[   65.114192] ACPI: PCI Interrupt Link [APC4] enabled at IRQ 19
[   65.114212] nvidia 0000:02:00.0: PCI INT A -> Link[APC4] -> GSI 19 (level, high) -> IRQ 19
[   65.117276] NVRM: loading NVIDIA Linux x86 Kernel Module  96.43.11  Mon Feb 23 15:29:05 PST 2009
[   65.118420] IRQ 19/nvidia: IRQF_DISABLED is not guaranteed on shared IRQs
[   65.225800] agpgart-nvidia 0000:00:00.0: AGP 3.0 bridge
[   65.225827] agpgart-nvidia 0000:00:00.0: putting AGP V3 device into 8x mode
[   65.225905] nvidia 0000:02:00.0: putting AGP V3 device into 8x mode
[   65.809497] Xorg:5600 conflicting memory types e0000000-e0600000 uncached-minus<->write-combining
[   65.809508] reserve_memtype failed 0xe0000000-0xe0600000, track uncached-minus, req write-combining
[   65.810485] Xorg:5600 conflicting memory types e0000000-e0600000 uncached-minus<->write-combining
[   65.810492] reserve_memtype failed 0xe0000000-0xe0600000, track uncached-minus, req write-combining
Under kernel 2.6.28.7 I don't have this error, X starts up as it should.

I've finally found a workaround for it which involves adding "nopat" to the kernel boot line. However, I'm guessing that the hack is a bad thing and could potentially open up a can of worms.

So I would like to know if there's a long-term fix for this. I'm convinced that it's a kernel-module problem since the module works fine under 2.6.28.7.

Thanks.
RAMChYLD is offline   Reply With Quote