Description:
A patch to nvidia's header kernel files was created due to a bug with nullptr de-referencing in https://bugs.archlinux.org/task/62142, however the module has been fixed upstream for some time.

I have tested the change on an SLI system with two Quadro P4000 and the `1_Utilities/p2pBandwidthLatencyTest` mentioned in that bug and confirmed it works without the patch. Patch is attached to this bug :)

Additional info:
upstream bug is internal only, sorry.

Steps to reproduce:
- have 2 identical nvidia gpu's
```
~ jshap@exodia
> nvidia-smi
Tue Oct 29 11:28:40 2019
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 435.21 Driver Version: 435.21 CUDA Version: 10.1 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 Quadro P4000 Off | 00000000:17:00.0 Off | N/A |
| 46% 32C P8 5W / 105W | 2MiB / 8119MiB | 0% Default |
+-------------------------------+----------------------+----------------------+
| 1 Quadro P4000 Off | 00000000:65:00.0 Off | N/A |
| 46% 35C P8 9W / 105W | 2MiB / 8116MiB | 0% Default |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+
```
- package driver with this patch applied to build (probably `rmmod` just to be sure)
- build and run p2p test
```
cd /opt/cuda/samples/1_Utilities/p2pBandwidthLatencyTest
make
./p2pBandwidthLatencyTest
```
- don't crash :)