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

[nmano]

But I know how to get sleep with SSDT

Gorgeous, man! Let me see all details and come back to you!

Thanks!

Question: Where do you rename H000 -> GFX1 and H001 -> HDAU ?

Good Question HDAU
Try SSDT-HDAU make other one.
I don't know how to rename HDAU with clover config.

 

[kgp]

Good Question HDAU
Try SSDT-HDAU make other one.
I don't know how to rename HDAU with clover config.

H000 -> GFX1      48303030    47465831
H001 -> HDAU      48303031    48444155

Anything else to be replaced to make HDMI/DP work?

Problem: also NVMe figures under H000 and after the ACPI DSDT Replacement Patch also as GFX1

 

[nmano]

H000 -> GFX1      48303030    47465831
H001 -> HDAU      48303031    48444155

Anything else to be replaced to make HDMI/DP work?

Problem: also NVMe figures under H000 and after the ACPI DSDT Replacement Patch also as GFX1

X99 board has lots H000 devices.This is not good idea.

 

[kgp]

X99 board has lots H000 devices.This is not good idea.

That's true! One can easily do all required replacements with your DSDT patches... But my question was if we could do everything, also the necessary replacements, within one single SSDT..

At least all H000 replacements cannot be performed within a ACPI Table, as there are several and not only one...

 

[nmano]

ALL IN ONE SSDT
# X99_SSDT_Maintained and disassembled by:N.Mano
# Credit: RehabMan
#Worked ALL
#Enjoy

 

[athre0z]

Finally, got my USB to be fully functional! I had to disable "Above 4G decoding" to prevent AppleUSBXHCI from erroring.

Situation with >4G decoding enabled:

• Without any kexts:
  • Only ASMedia USB ports working
  • Remaining ports: dead
• With USBInjectAll:
  • ASMedia USB working
  • 2 USB ports in the second row from above working
  • 2 internal USB2 ports working
  • 1 (maybe 2, didn't test) internal USB3 ports working
  • Remaining ports: dead
• With DSM2's driver:
  • Same situation like without any kext

Relevant kernel log entry that lead me to the workaround of disabling >4G decoding:

2017-10-28 00:08:42.509323+0200 0x1f9      Default     0x0                  0      0    kernel: (AppleUSBXHCI) 000002.509315 XHCI@14000000: AppleUSBXHCI::start: unexpected PAGESIZE 0xffffffff

Without >4G decoding:
What seems to be fully functional USB2, 3 and 3.1 support, without any kext (neither USBInjectAll, nor @DSM2's kext).

(ASUS X99-A II, i7-6850k, latest UEFI, MSR unlock patch applied)

Is there any reason why the guide suggests to enable this feature?

 

[kgp]

Finally, got my USB to be fully functional! I had to disable "Above 4G decoding" to prevent AppleUSBXHCI from erroring.

Situation with >4G decoding enabled:

• Without any kexts:
  • Only ASMedia USB ports working
  • Remaining ports: dead
• With USBInjectAll:
  • ASMedia USB working
  • 2 USB ports in the second row from above working
  • 2 internal USB2 ports working
  • 1 (maybe 2, didn't test) internal USB3 ports working
  • Remaining ports: dead
• With DSM2's driver:
  • Same situation like without any kext

Relevant kernel log entry that lead me to the workaround of disabling >4G decoding:

2017-10-28 00:08:42.509323+0200 0x1f9      Default     0x0                  0      0    kernel: (AppleUSBXHCI) 000002.509315 XHCI@14000000: AppleUSBXHCI::start: unexpected PAGESIZE 0xffffffff

Without >4G decoding:
What seems to be fully functional USB2, 3 and 3.1 support, without any kext (neither USBInjectAll, nor @DSM2's kext).

(ASUS X99-A II, i7-6850k, latest UEFI, MSR unlock patch applied)

Is there any reason why the guide suggests to enable this feature?

Strange, my USB functionality is not affected by >4G decoding at all! If you do not use the ThunderboltEX 3, you can disable it in any case... I ma not really sure if >4G decoding is really mandatory in case of the ThunderboltEX 3....

We have reached native XHC USB implementation on X299 boards since 10.13 SU.

Are you saying that also on the ASUS X99-A II, XHC USB2.0 and USB3.0 is now natively implemented? I did not find the time yet to check on that! But I can imagine that native XHC implementation also has been established for X99 mobos with 10.13 SU. Please report back if this is the case! The first signs of native XHC USB implementation are, when all other non-native XHC USB implementations stop working. Thus, what you witness and report above might be just due to the fact that indeed XHC USB is now natively implemented, and no XHC USB Kext implementation are further necessary. But I don't hink that your USB issues are related to >4G decoding...

By the way, USB3.1 is implemented on different controllers then XHC and already natively worked during the 10.13 beta testing.

 

[kgp]

ALL IN ONE SSDT
# X99_SSDT_Maintained and disassembled by:N.Mano
# Credit: RehabMan
#Worked ALL
#Enjoy

Does not work in my case.. Only HDMI/DP SSDT, that works in my case is the one attached. However no HDAU PCI entry in System Report...

Can you modify? I have a GTX 1080 Ti in Slot 0

 

[kgp]

@nmano, @mm2margaret
I did it!

ASUS X99-A II Analogue and Nvidia GeForce GTX 1080 Ti 11Gb Extreme Edition HDMI/DP Audio successfully implemented!

 

[athre0z]

Strange, my USB functionality is not affected by >4G decoding at all! If you do not use the ThunderboltEX 3, you can disable it in any case... I ma not really sure if >4G decoding is really mandatory in case of the ThunderboltEX 3....

We have reached native XHC USB implementation on X299 boards since 10.13 SU.

Are you saying that also on the ASUS X99-A II, XHC USB2.0 and USB3.0 is now natively implemented? I did not find the time yet to check on that! But I can imagine that native XHC implementation also has been established for X99 mobos with 10.13 SU. Please report back if this is the case! The first signs of native XHC USB implementation are, when all other non-native XHC USB implementations stop working. Thus, what you witness and report above might be just due to the fact that indeed XHC USB is now natively implemented, and no XHC USB Kext implementation are further necessary. But I don't hink that your USB issues are related to >4G decoding...

By the way, USB3.1 is implemented on different controllers then XHC and already natively worked during the 10.13 beta testing.

It certainly looks like it. I found the clue towards disabling >4G decoding on Pike R Alpha's blog.
https://pikeralpha.wordpress.com/2016/07/14/asus-x99e-with-i7-6850k/
CTRL+F for "unexpected PAGESIZE 0xffffffff". I'm not an expert with neither XHCI, PCI-e or Apple's implementation of those -- all I know is that this certainly did fix my issue after previously trying every possible permutation of drivers and USB UEFI settings.