How to extend the iMac Pro to X99 [Successful Build/Extended Guide]

[Obecny]

From my own experience I can tell that I am using "ASUS MultiCore Enhancement" from almost the beginning of this board hackintosh story and didn't noticed anything bad about it! Also using VT-D without DART=0 until now and all was good in my case, virtualisation is a must in my case so I had to use this vt-d always and didn't know what the dart=0 is for.

I have tried reimplement nvidia again using only the AGDPFix and unfortunately it didn't work in my case (GTX970) as I am stuck at boot with gioscreenlockstate, uninstalling and reinstalling drivers didn't help so I still have to use the (2) nvidia fixup and libvalfix kexts in my efi

In my experience without any additional patches the performance in geekbench scores (24k) is about 10k lower than expected in my config, when I add an xcpm_assert_wrmsr patch to the config.plist the I am getting extra 10k (34k) without any impact on stability of the system, idle cpu speeds are much higher as it looks like it won't go below 2Ghz at all when without this patch the lowest I get is 1.2Ghz, temperature of the cpus changed only a small bit and still oscillating around 30C idle up to 85C under full load (rendering all cores) all of this tested with CPU overclocked to 4.2Ghz!

Would be great if someone here with 6900K overclocked to 4.2Ghz share with me a geekbench scores to compare.
Asus X99 Deluxe II, 32Gb ram @2666Mhz

All above is my personal experience and I don't encourage anyone to use any of my suggestion but would love to hear your opinion about it if someone of you did some of the above tests

 

[kgp]

From my own experience I can tell that I am using "ASUS MultiCore Enhancement" from almost the beginning of this board hackintosh story and didn't noticed anything bad about it! Also using VT-D without DART=0 until now and all was good in my case, virtualisation is a must in my case so I had to use this vt-d always and didn't know what the dart=0 is for.

I have tried reimplement nvidia again using only the AGDPFix and unfortunately it didn't work in my case (GTX970) as I am stuck at boot with gioscreenlockstate, uninstalling and reinstalling drivers didn't help so I still have to use the (2) nvidia fixup and libvalfix kexts in my efi

In my experience without any additional patches the performance in geekbench scores (24k) is about 10k lower than expected in my config, when I add an xcpm_assert_wrmsr patch to the config.plist the I am getting extra 10k (34k) without any impact on stability of the system, idle cpu speeds are much higher as it looks like it won't go below 2Ghz at all when without this patch the lowest I get is 1.2Ghz, temperature of the cpus changed only a small bit and still oscillating around 30C idle up to 85C under full load (rendering all cores) all of this tested with CPU overclocked to 4.2Ghz!

Would be great if someone here with 6900K overclocked to 4.2Ghz share with me a geekbench scores to compare.
Asus X99 Deluxe II, 32Gb ram @2666Mhz

All above is my personal experience and I don't encourage anyone to use any of my suggestion but would love to hear your opinion about it if someone of you did some of the above tests

Sorry, but NVWebDriverLibValFix.kext should be already part of NvidiaGraphicsFixup.kext v1.2.0! Thus, apparently in any case you seem to apply some overkill in your black screen prevention. Note that NvidiaGraphicsFixup.kext v1.2.0 also requires Lilu.kext v1.2.0!

You are free to include the xcpm_assert_wrmsr patch, why not? I will give it a try myself to see if there is any difference. By the way are you using PMDrvr.kext and EIST enabled?

Cheers,

KGP

 

[Obecny]

Sorry, but NVWebDriverLibValFix.kext should be already part of NvidiaGraphicsFixup.kext v1.2.0! Thus, apparently in any case you seem to apply some overkill in your black screen prevention. Note that NvidiaGraphicsFixup.kext v1.2.0 also requires Lilu.kext v1.2.0!

You are free to include the xcpm_assert_wrmsr patch, why not? I will give it a try myself to see if there is any difference. By the way are you using PMDrvr.kext and EIST enabled?

Cheers,

KGP

You absolutely right about the nvidiafixup, didn't know it so I have updated the NvidiaGaphicFixup.kext to 1.2 now and completely removed NVWebDriverLibValFix.kext and still everything works no problem!

I am not using PMDrvr.kext and EIST because sleep is important for me and as far as I know it will break the sleep, however didn't even tried it to confirm.

Thanks

 

[kgp]

You absolutely right about the nvidiafixup, didn't know it so I have updated the NvidiaGaphicFixup.kext to 1.2 now and completely removed NVWebDriverLibValFix.kext and still everything works no problem!

I am not using PMDrvr.kext and EIST because sleep is important for me and as far as I know it will break the sleep, however didn't even tried it to confirm.

Thanks

One question: Did you ever try to just apply the AGDPfix after a macOS clean install? I guess it will also work in your system... I don't think that everything you mention would be necessary.... Your deviating black screen prevention might be currently necessary on your system, because the latter might be already messed up to some extent... If you have some time, give it a try with the clean install on a test drive... it would be very interesting to know whether or not my suspicion is correct...

Cheers,

KGP

 

[Obecny]

One question: Did you ever try to just apply the AGDPfix after a macOS clean install? I guess it will also work in your system... I don't think that everything you mention would be necessary.... Your deviating black screen prevention might be currently necessary on your system, because the latter might be already messed up to some extent... If you have some time, give it a try with the clean install on a test drive... it would be very interesting to know whether or not my suspicion is correct...

Cheers,

KGP

I have tried it with my first install and didn't work but I am not so sure how pure the installation was so I will give a try some day to test it again, as you said it is possible something get broken in my system during all my tests, anyway this is not a big issue for me as long as I there is a way to get a screen working!

I am testing now your latest efi and after cleaning my nvram and caches I have no longer bluetooth available in my system, I believe the AirportBrcmFixup.kext is not enough in my case (deluxe II), wifi is still fully functional however I am not using it.

Thanks

 

[kgp]

I have tried it with my first install and didn't work but I am not so sure how pure the installation was so I will give a try some day to test it again, as you said it is possible something get broken in my system during all my tests, anyway this is not a big issue for me as long as I there is a way to get a screen working!

I am testing now your latest efi and after cleaning my nvram and caches I have no longer bluetooth available in my system, I believe the AirportBrcmFixup.kext is not enough in my case (deluxe II), wifi is still fully functional however I am not using it.

Thanks

This might be indeed the case. Please let me know with which additional approach you get your onboard bluetooth working again.
By the way did you try to boot your system disk without AirportBrcmFixup.kex? In my case the latter kext is only required for successfully booting the USB Flash Drive Installer....

 

[Obecny]

This might be indeed the case. Please let me know with which additional approach you get your onboard bluetooth working again.
By the way did you try to boot your system disk without AirportBrcmFixup.kex? In my case the latter kext is only required for successfully booting the USB Flash Drive Installer....

Just delated AirportBrcmFixup.kext, no changes at all, system still functional as before, wifi works fine, no bluetooth as expected

 

[Obecny]

I am temporary back to "BrcmFirmwareData.kext" and "BrcmPatchRAM2.kext" solutions for the bluetooth

@kgp if you get a chance would you please test for me this xcpm_assert_wrmsr patch as I would like to know your opinion about it and maybe you have some better solution in general, if you not get any better results with this patch it would mean there is something wrong with my configuration in general, thanks

 

[kgp]

I am temporary back to "BrcmFirmwareData.kext" and "BrcmPatchRAM2.kext" solutions for the bluetooth

@kgp if you get a chance would you please test for me this xcpm_assert_wrmsr patch as I would like to know your opinion about it and maybe you have some better solution in general, if you not get any better results with this patch it would mean there is something wrong with my configuration in general, thanks

Whenever I will find a little bit of time! No big deal!

 

[kgp]

Hi KGP
Can you know how to fix this yellow icon?
Thank you for your hard work.
Samsung SSD 960 PRO 512GB:

Please see detailed workaround in my guide! Thanks, man!

In any case it seems that the external NVMe icon bug has been fixed by Apple within SU! Did you check if this is also the case on X99 by upgrading to macOS High Sierra 10.13 SU?