Contribute
Register

iMac Pro X99 - Live the Future now with macOS 10.14 Mojave [Successful Build/Extended Guide]

Status
Not open for further replies.
Hi @kgp I got the USB3 working following your guide and most of the PCI devices implemented.
Screen Shot 2018-09-30 at 11.38.06 PM.png
Some were easy other had to be lightly modified. But I hit a wall trying to implement SSDT-X99-XHC2.aml and SSDT-X99-TB3HP-APR.aml. Think you can share some wisdom with me based on my Ioreg if you have a chance?

Greatly appriciated.
 

Attachments

  • Mac Pro IOREG.ioreg
    26.2 MB · Views: 102
  • EFI.zip
    31.9 MB · Views: 93
Hi @kgp I got the USB3 working following your guide and most of the PCI devices implemented.
View attachment 354337
Some were easy other had to be lightly modified. But I hit a wall trying to implement SSDT-X99-XHC2.aml and SSDT-X99-TB3HP-APR.aml. Think you can share some wisdom with me based on my Ioreg if you have a chance?

Greatly appriciated.

Seems that you have two TB adopters?

1.) PCI0/BR3A/H000

2.) PCI0/RP01/

ad.1) Adopt and modify ACPI path in the SSDT accordingly. Analyse your DSDT.aml which device replaces PCI0/BR3A/H000 if nulled. Null also the replacing device ???? in the SSDT and implement it (????) also in the definition block..

Code:
    Scope (_SB.PCI0.BR3A)
    {
        Scope (H000)
        {
            Name (_STA, Zero)  // _STA: Status
        }

        Scope (????)
        {
            Name (_STA, Zero)  // _STA: Status
        }

ad .2) Adopt and modify ACPI path in the respective SSDT accordingly. Seems the nulling of any device can be skipped in this case.

Code:
    Scope (_SB.PCI0.RP01)
    {
        Method (_RMV, 0, NotSerialized)  // _RMV: Removal Status
        {
            Return (One)
        }

        Device (UPSB)
        {

You do not have any USB3.1 onboard controller? The only XHC apart from XHCI I see in your IOREG is the one of the TB controller in PCI0.RP01! What for you want to use SSDT-X99-XHC2.aml in this case?

Hope that helps..

Good luck,

KGP
 
Last edited:
I enclose the settings of the bios do not make sure that some settings are for overclocking and cpu no, I simply set the car on the CPU to show you and even so the temperatures are high (43 °) should be at 30-32 °

I enclose the settings of the bios do not make sure that some settings are for overclocking and cpu no, I simply set the car on the CPU to show you and even so the temperatures are high (43 °) should be at 30-32 °
I set the standard Haswell-E / EP FakeCPUID: "0x0306F2"
also put voodo tsc sync with 6core values (5820k)
Verify with the terminal command "sysctl machdep.xcpm.mode val.0 and not 1 (is not active)

Many thanks
 

Attachments

  • EFI.zip
    31.9 MB · Views: 91
  • Like
Reactions: kgp
I enclose the settings of the bios do not make sure that some settings are for overclocking and cpu no, I simply set the car on the CPU to show you and even so the temperatures are high (43 °) should be at 30-32 °

I enclose the settings of the bios do not make sure that some settings are for overclocking and cpu no, I simply set the car on the CPU to show you and even so the temperatures are high (43 °) should be at 30-32 °
I set the standard Haswell-E / EP FakeCPUID: "0x0306F2"
also put voodo tsc sync with 6core values (5820k)
Verify with the terminal command "sysctl machdep.xcpm.mode val.0 and not 1 (is not active)

Many thanks

Please upload screenshots of your BIOS settings. The BIOS.txt is useless, unfortunately.
 
Please upload screenshots of your BIOS settings. The BIOS.txt is useless, unfortunately.
here are the most important parts
 

Attachments

  • bios setting.zip
    2.9 MB · Views: 63
  • Like
Reactions: kgp
Could you please also answer remaining question in post 259?

1.) Can you enable CPU SVID Support?
2.) Your memory really just works @ 2400MHz?
3.) It seems that you use a CPU Load Line Calibration, wich I don't use at all and where I also don't know what you did in particular. This might be the source for your actual temps and likely also the source of temp differences between High Sierra and Mohave. What happens if you set everything under Ai Tweaker/External Digi+ Power Control to its default value?
 
I enclose the settings of the bios do not make sure that some settings are for overclocking and cpu no, I simply set the car on the CPU to show you and even so the temperatures are high (43 °) should be at 30-32 °

I enclose the settings of the bios do not make sure that some settings are for overclocking and cpu no, I simply set the car on the CPU to show you and even so the temperatures are high (43 °) should be at 30-32 °
I set the standard Haswell-E / EP FakeCPUID: "0x0306F2"
also put voodo tsc sync with 6core values (5820k)
Verify with the terminal command "sysctl machdep.xcpm.mode val.0 and not 1 (is not active)

Many thanks

What happens if you just enable the xcpm_pkg_scope_msrs Kernel patch in your config.plist? Does XCPM continues working at reasonable performance or do you also need the _xcpm_performance Kernel patch and the _xcpm_SMT_scope_msrs Kernel patches to achieve a reasonable CPU performance?

I see that you also enable the _xcpm_core_scope_msrsm Kernel patch, thus, the MSR Register is locked in your BIOS? If so.. wouldn't it be necessary to also check KernelPM in this case?
 
Seems that you have two TB adopters?

1.) PCI0/BR3A/H000

2.) PCI0/RP01/
Thank you @kgp . ill try again tonight.
So "H000" acts as nulling when no sub device is present?
(That explains why when i did it to the RX Vega a "D084" popped after the restart.)

PCI0/PR01/ is my USB 3.1.
I do not know why it shows Thunderbolt device under it as well maybe the Type-C connector?
 
Thank you @kgp . ill try again tonight.
So "H000" acts as nulling when no sub device is present?
(That explains why when i did it to the RX Vega a "D084" popped after the restart.)

PCI0/PR01/ is my USB 3.1.
I do not know why it shows Thunderbolt device under it as well maybe the Type-C connector?

I don't understand the question. You have to null H000. It's replacement is defined in your DSDT.aml. It can be that there are even two replacements defined in the DSDT.aml, like e.g. H001 and D077. In this case you would have to also null H001 and D077, before introducing the new UPSB device.

PCI0/PR01/ definitely shows another TB adopter.. why.. only you might know.. ;)
 
Status
Not open for further replies.
Back
Top