Some Nvidia driver fail to load in Kata container

[Jimmy-Xu]

@amshinde @Jimmy-Xu

Thank you for your kindly documentation for using Nvidia GPU with Kata-containers.
Also, I'm sorry to keep you inquiring about the completed PR.
I followed the guide at the documentation, and I'm successfully able to pass one of the GPU to the guest VM through vfio-pci.
It show up with lspci on container.

00:08.0 VGA compatible controller [0300]: NVIDIA Corporation GP108 [GeForce GT 1030] [10de:1d01] (rev a1) (prog-if 00 [VGA controller])
	Subsystem: NVIDIA Corporation GP108 [GeForce GT 1030] [10de:11c7]
	Physical Slot: 8
	Flags: bus master, fast devsel, latency 0, IRQ 10
	Memory at c0000000 (32-bit, non-prefetchable) [size=16M]
	Memory at 1ac0000000 (64-bit, prefetchable) [size=256M]
	Memory at 1ad0000000 (64-bit, prefetchable) [size=32M]
	I/O ports at 1000 [size=128]
	Capabilities: [60] Power Management version 3
	Capabilities: [68] MSI: Enable- Count=1/1 Maskable- 64bit+
	Capabilities: [78] Express Legacy Endpoint, MSI 00
	Kernel driver in use: nvidia

But, I have another problem.
When I check Nvidia device status with nvidia-smi, it is failed like as below.
And, the nvidia gpu information has an errors.

# nvidia-smi 
Unable to determine the device handle for GPU 0000:00:08.0: Unknown Error

# cat /proc/driver/nvidia/gpus/0000\:00\:08.0/information 
Model: 		 Unknown
IRQ:   		 35
GPU UUID: 	 GPU-????????-????-????-????-????????????
Video BIOS: 	 ??.??.??.??.??
Bus Type: 	 PCIe
DMA Size: 	 36 bits
DMA Mask: 	 0xfffffffff
Bus Location: 	 0000:00:08.0
Device Minor: 	 0
Blacklisted:	 No

Is there anything I missed that I should do?

Originally posted by @wansuyoo in #2410 (comment)