Contribute
Register

How to build your own iMac Pro [Successful Build/Extended Guide]

Status
Not open for further replies.
Yes, I agree the first two links in the first post are not adequate. However the thread then later develops into a nice exploration of the correct config for the MSI X299. In particular, I got the hint of disabling almost everything on the Clover config from the thread.

I think you are right in not posting wrong guides here, so I have removed the link to the thread from my post. Sorry for that.



Well, I meant too say it was a bit too much information for my needs, and so it was difficult to find what I needed. For example, I know how to create the installation media, I don't need issues specific to RX cards, and so on. So I didn't mean to say your guide had problems, I just meant to say it was too much for what I need. And indeed I could not find everything there, but that is understandable, each system (Mobo+CPU) might have different issues.

Your guide was very helpful in teaching how to create a unique UUID so that I could have iMessage working, your guide was very helpful in having a consistent XPM (I knew of ssdtprgen, but I didn't know about Intel Power Gadget). Thanks to your guide, I didn't even try to use Toleda's Auidio kext and went straight to VoodooHDA. So your guide is full of very useful info for me.

All the best and thanks for the guide once more!

Thanks, man :thumbup:

More feedback and additional helpful comments and indications concerning the MSI X299 SLI Plus from your side would be highly appreciated! Don't hesitate to share your estimated experience within this thread! I will see, which parts of such additional information I could still implemented in my guide, of course with an adequate citation and reference to you as the originator of such kind of information. If you would be able to send me a successfully working board-specific MSI X299 SLI Plus XHC USB Kext with your initials in addition, I could implement the latter in the X299 XHC USB Kext-Library yet to be released and implemented for the X299 User Community.

With your experience and by following my detailed XHC USB Kext creation guidelines, it might be a small step and effort for you but a big step for MSI X299 SLI Plus mankind ;)!

Hope to keep in touch!

KGP
 
Last edited:
Question: Why not use MacPro6,1 and use Frequency Vectors of iMac17,1 with Piker Alpha script? You won't have to patch graphics or usb right? Just a thought.. correct me if I am wrong.
 
Question: Why not use MacPro6,1 and use Frequency Vectors of iMac17,1 with Piker Alpha script? You won't have to patch graphics or usb right? Just a thought.. correct me if I am wrong.

Because Skylake-X is a Skylake and not a Broadwell-E/EP or Haswell-E/EP Processor and a X299 board is a Series-200 and not a Series-9 Board! A macPro6,1 System is close to X99 and Broadwell-E/EP or Haswell-E/EP, while an iMac17,1 System is closer to Skylake, Skylake-X and series-100/series-200 board technology. You would not either use a Skylake processor and a Series-100 board with SMBIOS mac Pro6,1, would you?

Anyway, apart from the purely theoretical discussion above, it is simply fact that SMBIOS iMac17,1 is absolutely mandatory if one wants to run XCPM by means of Pike Alpha's ssdtPRGen.sh and ssdt.aml. With SMBIOS macPro 6,1, ssdtPRGen.sh does not even compile and it is therefore impossible to derive the ssdt.aml, which is absolutely vital for running XCPM in its most sophisticated way, unless you want to say that Pike Alpha's sophisticated XCPM approach is ********, which would be totally at odd with all the extended brilliant XCPM results gained with Boadwell-E/EP or HaswellE/EP Processor and X99 board technology and totally at odd with the tremendous success of Pike Alpha's XCPM approach over many many years!

Just tell me one reason why to use SMBIOS mac Pro6,1 instead of iMac17,1! The only transparent reason I would see would be the desperate intent of maintaining the USBInjectAll.kext approach or derivatives (e.g. X99_Injector USB 3.kext or, XHCI-200-series-injector.kext, etc.), which however apparently cannot be entirely and fully successfully implemented on X299 boards either, even despite the use of a SMBIOS macPro6,1 System Definition. In any case, please don't tell me seriously that USBInjectAll.kext and derivatives as well as related DSDT or Kext and Kernel patching distributed in the respective forums is not a USB patching! And please also do not tell me that SMBIOS iMac17,1 needs additional graphics patching when compared with SMBIOS macPro6,1 (just think on the black screen issue with Nvidia graphics cards). That's simply ridiculous, sorry!

It is true, that the SMBIOS iMac17,1 imposes the sudden dead of USBInjectAll.kext or any derivatives (e.g. X99_Injector USB 3.kext or, XHCI-200-series-injector.kext, etc.), as there is no iMac 17,1 identity neither in Apple's IOUSBHostFamily.kext nor in Apple's AppleUSBXHCIPCI.kext. However, the absolutely novel and innovative XHC USB Kext approach originally developed by Brumbaer and actually employed and further iterated in all detail in my guide, makes USBInjectAll.kext totally obsolete! Moreover, I am indeed totally happy about this latter fact, as I am honestly also very happy to finally get rid of the entire problematics and sometimes quite diffuse application strategies related to USBInjectAll.kext, spread over the years.

Thus in my opinion, the only excuse to stay with SMBIOS mac Pro6,1 is the personal laziness of creating board-specific XHC USB Kexts for mainboards not yet directly support by my guide. However, I implemented an explicit and very detailed XHC USB Kext Creation Guideline for all available X299 mainboards and by the way also for all other mainboards available on the market. The Guideline just needs to be applied once for each particular mainboard. The related effort does not extent more than a couple of hours of your time. Once a XHC USB Kext is available for a specific mainboard, all users of the same mainboard can directly employ the latter XHC USB Kext on their mainboard without any further personal efforts, like it is already the case for the ASUS Prime X299 Deluxe! Finally, the XHC USB Kext approach absolutely grants the full implementation of all available internal and external USB2.0 and USB3.0 ports, which furthermore deliver without any exception also all expected data rates! There is absolutely no need for additional DSDT, Kext or Kernel patching, like it is frequently the case when employing USBInjectAll.kext or derivatives, despite the OS X USB Port Limit Patch, which is anyway required in case of both USB approaches.

You asked me about my personal opinion. Saying the above, I hopefully satisfactorilly fulfilled your request and answered your question in the best possible way.

I don't want to launch once more an extended and never ending theoretical discussion with god and the world, which already totally messed up once this thread at the very beginning!

If you are content with the SMBIOS mac Pro6,1 System Definition, live in peace and be happy with this in my opinion however largely misleading approach. I will continue promoting and distributing the SMBIOS iMac17,1 approach, which in my opinion and given the excellent results, appears to be definitely the more transparent and intuitively correct approach for Skylake-X/X299.

Everybody is free to decide the direction of his Skylake-X/X299 journey. Nobody is obliged to follow my guide. The intention of my guide is to show and demonstrate to people how to derive a fully functional, excellently performing and absolutely stable Skylake-X/X299 System. All Users following guides different from mine have my blessing in their endevours!

Hurray to the diversity of nature!

Cheers,

KGP
 
Last edited:
Your are very welcome... Don't stop writing me until we succeed in successfully booting your rig :thumbup:

Have a sound sleep! :lol:

So here few other things I did try this morning (that X299 haunted me, I couldn't sleep so I woke at 4.30 and went to my office at 5, lol)

Did a Windows 10 clean install, everything went smooth, no issues, so I strongly confirm no hardware issues at all,

Little interesting that if I don't put the VoodooTSCSync.kext into the Kext Folder of my EFI partition, couple times I got the "Cross No go" with distor letters screen, will try to take a photo of it, but most of the time it still goes to the Restarting error

If I put the modify IOCPUNumber to 19 of the VoodooTSCSync.kext (or even keep at number 11 as your CPU 7800X), it also end up at the same screen, (make no sense to me ;))

I tried without the system disk (either SSD or Spinning HD) plug in, same dead-end message,

Here I do the screenshot of my USB Flash Drive, can you pls check if I need anything,
I double check the BIOS setup 100 times, same setup as you recommended, no OC yet,

If you need me to send any other info, photos, videos of other components of my build, pls let me know!
Thank you so much again for a huge help, :thumbup:

Screen Shot 2017-09-29 at 7.16.25 AM.png
 
So here few other things I did try this morning (that X299 haunted me, I couldn't sleep so I woke at 4.30 and went to my office at 5, lol)

Did a Windows 10 clean install, everything went smooth, no issues, so I strongly confirm no hardware issues at all,

Little interesting that if I don't put the VoodooTSCSync.kext into the Kext Folder of my EFI partition, couple times I got the "Cross No go" with distor letters screen, will try to take a photo of it, but most of the time it still goes to the Restarting error

If I put the modify IOCPUNumber to 19 of the VoodooTSCSync.kext (or even keep at number 11 as your CPU 7800X), it also end up at the same screen, (make no sense to me ;))

I tried without the system disk (either SSD or Spinning HD) plug in, same dead-end message,

Here I do the screenshot of my USB Flash Drive, can you pls check if I need anything,
I double check the BIOS setup 100 times, same setup as you recommended, no OC yet,

If you need me to send any other info, photos, videos of other components of my build, pls let me know!
Thank you so much again for a huge help, :thumbup:

View attachment 281785

Hmm... this screen shot does not say anything to me.. can you upload the EFI-Folder, which you use on the USB Flash Drive Installer? As the entire boot process is successfully apart from the very last step, I do not expect to find major deviations or errors anyway.

As it told you, I suddenly witnessed the identical issue to yours in my own rig yesterday. Just by experimenting with the EFI folder, BIOS settings and PCIe card implementations, I once more resolved the problem. The corresponding successful EFI-Folder has been uploaded immediately. All further knowledge gained along my intents has been asap implemented in my guide. I really don't know at present why you still witness this final boot process issue. But let's try to find the dirty bug. ;)

A complete listing of your build would be certainly of help in this issue...
 
Hmm... this screen shot does not say anything to me.. can you upload the EFI-Folder, which you use on the USB Flash Drive Installer? As the entire boot process is successfully apart from the very last step, I do not expect to find major deviations or errors anyway.

As it told you, I suddenly witnessed the identical issue to yours in my own rig yesterday. Just by experimenting with the EFI folder, BIOS settings and PCIe card implementations, I once more resolved the problem. The corresponding successful EFI-Folder has been uploaded immediately. All further knowledge gained along my intents has been asap implemented in my guide. I really don't know at present why you still witness this final boot process issue. But let's try to find the dirty bug. ;)

A complete listing of your build would be certainly of help in this issue...


I GOOOOOTTTTTT IIITTTT, OMG, HAHAHAHA

So basically I follow some setup from the guy here for core i9 7900X (I know you guys have some history, :)
https://www.tonymacx86.com/threads/...-i9-7900x-x299-ga-aorus-9-radeon-vega.229006/

And using some setup from your MB, lol, and NOW mine is boosted to Installation Screen,

IMG_6447.jpg
 
[....]
Everybody is free to decide the direction of his Skylake-X/X299 journey. Nobody is obliged to follow my guide. The intention of my guide is to show and demonstrate to people how to derive a fully functional, excellently performing and absolutely stable Skylake-X/X299 System. All Users following guides different from mine have my blessing in their endevours!

Hurray to the diversity of nature!

Cheers,

KGP

Hey no problem, just looking for some info to what exactly is loaded from iMac 17,1 apart from frequency vectors and usb. I am trying to understand the differences in what macOS will load with the different smbioses. I know the hardware is totally different, but your usb story sounds valid. Maybe I just need to search better. I know you can patch
X86PlatformPlugin.kext with frequency vectors from say iMac17,1 and use a MacPro6,1 smbios so you can load the XMCP stuff?. If this viable? (I don't know). Thanks for your opinion.
 
And what did you change, if I may ask?

The ACPI Drop OEM_DSM,
add some drop table
On CPU section select C2, C4, C6
Device add 0x0 on most parts.. I will send the EFI folder in a bit, the installation flash drive being used right now,
 
Hey no problem, just looking for some info to what exactly is loaded from iMac 17,1 apart from frequency vectors and usb. I am trying to understand the differences in what macOS will load with the different smbioses. I know the hardware is totally different, but your usb story sounds valid. Maybe I just need to search better. I know you can patch
X86PlatformPlugin.kext with frequency vectors from say iMac17,1 and use a MacPro6,1 smbios so you can load the XMCP stuff?. If this viable? (I don't know). Thanks for your opinion.

Yes, of course you can inject the iMac17,1 frequency vector, which simply consists in adding the latter (or by the way in adding the frequency vector of any other Mac with a frequency implementation in it's respective plist) to the macPro6,1 plist.

However, you cannot use Pike Alpha's ssdtPRGen.sh when using SMBIOS macPro6,1 on Skylake-X/X299, as you are therefore not able to derive the ssdt.aml, which is absolutely vital to successfully run Pike Alpha's XCPM approach. I don't know which method you guys apply to finally perform the speed stepping without Pike Alpha's ssdt.aml (and in fact I am not even interested). I also don't know if this latter approach results in the same brilliant XCPM-performance like in case of Pike Alpha's approach, which I successfully applied for many many years. I only know that with the latter approach and especially with the Gigabyte X299 AORUS Gaming 9, the XCPM performance is dramatically worse when compared with the one of Pike Alpha's approach and by employing the ASUS Prime X299 Deluxe. I herewith do not refer to the available C- and P-States, but rather to the effective XCPM CPU performance in IPG graphs when e.g. sequentially running Cinebench-benchmarks in parallel, which also dramatically affects the overall benchmark results of the respective Skylake-X processor.

However, all this does not affect my other statements and conclusions summarized in my previous reply to you above, which hopefully shed some light onto your yet open questions.

Cheers,

KGP
 
Last edited:
Status
Not open for further replies.
Back
Top