NVIDIA Releases Alternate Graphics Drivers for macOS High Sierra 10.13.0 (378.10.10.10.15)

[Stork]

fosworn,
Since we're not telepathic, we need to know your system components. See the tonymacx86 Forum Rules & Policies on signature block components.

Additionally, please provide complete details in your Profile.

Mobo: motherboard manufacturer/model
CPU: detailed CPU model (e.g., i7-4770K)
Graphics: Video card manufacturer/model and/or Intel Integrated Graphics version (e.g., HD4600)

See my Profile items to the left of my post, under my name.

By completing your Profile and signature block, you help us help you better.

 

[iLikeHackintosh]

In the Terminal, remove the Nvidia drivers with the following Terminal commands:

There is no 100% guarantee but you maybe don't need this approach implement fully actually. On case of failure to load Nvidia Web Drivers, just removing NVDAStartupWeb.kext from /Library/Extensions might solve issue to get into GUI to fix problem with kexts et to install NVIDIA Web Drivers properlu.

 

[vulgo]

I have no idea why, but EmuVariableUefi-64.efi might have actually made a difference.

You'll be booting with nvram variables from nvram.plist. Which won't exist. So its like booting without persistent nvram until you install the rc.scripts and have shut down at least once.

 

[djlaauol]

Hi, im installed https://images.nvidia.com/mac/pkg/378/WebDriver-378.10.10.10.15.117.pkg + Kext in clover (lilu,nvidiafix,nvidiawebdriverifx,) + DriverUEFI64 (afhs,csmvideo,osxaptiofix2drv-64,emuvariableuefi-64) on OS X 10.13 (405)

Nvidia GTX 750 with TV Led 24'>HDMI
Intel G2030 3.0 Ghz
Gigabyte b75m-d3h
ram 12 GB ddr3
hd 120 gb
wifi tp-link N900 (change ap wifi channel to 64 for access 450mbps)

work audio, video, wifi,...

but no work cuda last version (9), any try ? ... ( we will have to wait for the release of the driver version with cuda??)

 

[fosworn]

You'll be booting with nvram variables from nvram.plist. Which won't exist. So its like booting without persistent nvram until you install the rc.scripts and have shut down at least once.

I'm new to setting up a Hackintosh, so please excuse my newbie question.
So I assume the Nvidia web drivers require some NVRAM settings to be recognized properly, is that correct?

What do the rc.scripts exactly do? Save NVRAM on shutdown and load them on boot?
How do the scripts differ from EmuVariableUefi-64.efi? Is it safe to leave EmuVariableUefi-64.efi or should I remove it and use the rc.scripts instead?

 

[iLikeHackintosh]

One important note about NVIDIA Web Drivers.

Usage of Lilu.kext, NVWebDriverLibValFix.kext and NvidiaGraphicsFixup.kext together can cause your HS to loop reboot. Latest NvidiaGraphicsFixup.kext already has built in the NVWebDriverLibValFix.kext!

Use only latest:

• Lilu.kext (>=1.20)
• NvidiaGraphicsFixup.kext (>=1.20) (requires Lilu.kext)

Do not use NVWebDriverLibValFix.kext with them!

 

[iLikeHackintosh]

but no work cuda last version (9), any try ? ... ( we will have to wait for the release of the driver version with cuda??)

You have to patch CUDA!

See instructions here: #563

 

[kobbygh]

finally got it to work now need help for my audio

thx

 

[vulgo]

So I assume the Nvidia web drivers require some NVRAM settings to be recognized properly, is that correct?

It shouldn't need anything to be recognised, but from looking through EFI variable dumps i believe it saves some sort of configuration there. If that is true and for what purpose i cannot confirm.

What do the rc.scripts exactly do? Save NVRAM on shutdown and load them on boot?

They save the variables on shutdown and Clover loads them back at boot.

How do the scripts differ from EmuVariableUefi-64.efi? Is it safe to leave EmuVariableUefi-64.efi or should I remove it and use the rc.scripts instead?

The two work together to emulate hardware NVRAM. Without the scripts nvram.plist will not change and variables will be the same every boot. Without the RC scripts and also without a plist, variables will be initialised every boot. Without the EFI driver, hardware NVRAM will be used if available.

 

[fosworn]

It shouldn't need anything to be recognised, but from looking through EFI variable dumps i believe it saves some sort of configuration there. If that is true and for what purpose i cannot confirm.They save the variables on shutdown and Clover loads them back at boot.The two work together to emulate hardware NVRAM. Without the scripts nvram.plist will not change and variables will be the same every boot. Without the RC scripts and also without a plist, variables will be initialised every boot. Without the EFI driver, hardware NVRAM will be used if available.

Thanks a lot for the explanation!
I should be able to remove EmuVariableUefi-64.efi then without problems. Why it was necessary to get me out of the boot loop is still unclear, though. Maybe instead of adding the file resetting the NVRAM would have worked as well.