Contribute
Register

[SUCCESS] Ongoing Status of Designare Z390 with i7-9700K

Status
Not open for further replies.
pastrychef said:
It sounds like you did more hacking on your Mac to run High Sierra than on your new build... And, a 9th gen CPU will run circles around that 3,1.

If the Wi-Fi/Bluetooth card you updated on the MacPro3,1 is using a BRCM94360xxx card, you can probably use that card in your new build. You just need a different adaptor.
Click to expand...
Yes you're right! I spent weeks trying to get 4K60 to work on the MacPro with various graphics cards. Even DosDude1's hack to install High Sierra took two or three weekends because my RAID-0 boot drive did not have a Recovery Partition, which is needed in order to disable SIP.

The Hackintosh installation, by comparison, took very little time and was much more satisfying. I will, however, continue to iron out the last couple of issues, namely:
  • Try to get Shutdown working.
  • Move my Broadcom WiFi/BT card from the MacPro (as you suggested) to the Socket 1 on the Designare via an adapter (or find a native Socket 1 Broadcom card).
  • Try to see if the second Ethernet port can be made to work (device ID 1539; Intel PRO/1000; Intel i211 chip type).
I also need to run GeekBench...

(Correction: In my previous [lengthy] post which I cannot edit because of a 'spam' warning, I should have said that El Capitan, not Sierra, was the last officially supported version of MacOS on the MacPro 3,1.)

Update: Just ordered the following for same-day delivery...
I prefer the Socket 1 (NGFF M.2) approach versus the use of WiFi/BT PCIe cards because:
  • External antenna connections are already provided on the I/O shield.
  • PCIe cards with on-board Bluetooth require a cable to be connected between the card and a USB header on the motherboard. The Socket 1 solution is fully self-contained.
Finally, some photos:
  • Stack of components.
  • Cable management 1.
  • Cable management 2.
  • RGB overkill (just because I can :))

Edit 12-05-2018: NGFF M.2 adapter cannot be used on this motherboard because of insufficient clearance between the Socket-1 port and a PCIe x1 slot located behind it. I've switched to this instead.
 

Attachments

  • Hack Components.jpg
    Hack Components.jpg
    302.7 KB · Views: 411
  • Case1.jpg
    Case1.jpg
    500.7 KB · Views: 422
  • Case2.jpg
    Case2.jpg
    317.7 KB · Views: 424
  • Case3.jpg
    Case3.jpg
    518.8 KB · Views: 412
Last edited:
johnnypig said:
Apollo x6 here, same mobo and cpu as OP. Updated to F4 bios before installing the Apollo so I can't say if it works or not because now it won't boot into high sierra. Runs fine in win10 though. Helluva job UA did this time.
Click to expand...
You probably did this already, but it doesn't hurt to ask: After the F4 BIOS upgrade, all BIOS settings are reset to defaults. Did you subsequently apply the correct settings?
 
CaseySJ said:
The Hackintosh installation, by comparison, took very little time and was much more satisfying. I will, however, continue to iron out the last couple of issues, namely:
  • Try to get Shutdown working.
  • Move my Broadcom WiFi/BT card from the MacPro (as you suggested) to the Socket 1 on the Designare via an adapter (or find a native Socket 1 Broadcom card).
  • Try to see if the second Ethernet port can be made to work (device ID 1539; Intel PRO/1000; Intel i211 chip type).
Click to expand...

  • Try installing EmuVariableUEFI-64 and RC scripts to see if it helps your shutdown issues.
  • The BCM94360CS2 with adaptor that you ordered should work in the M.2 slot.
  • Try searching around for the Smalltree drivers for your second NIC.
 
CaseySJ said:
Yes you're right! I spent weeks trying to get 4K60 to work on the MacPro with various graphics cards. Even DosDude1's hack to install High Sierra took two or three weekends because my RAID-0 boot drive did not have a Recovery Partition, which is needed in order to disable SIP.

The Hackintosh installation, by comparison, took very little time and was much more satisfying. I will, however, continue to iron out the last couple of issues, namely:
  • Try to get Shutdown working.
  • Move my Broadcom WiFi/BT card from the MacPro (as you suggested) to the Socket 1 on the Designare via an adapter (or find a native Socket 1 Broadcom card).
  • Try to see if the second Ethernet port can be made to work (device ID 1539; Intel PRO/1000; Intel i211 chip type).
I also need to run GeekBench...

(Correction: In my previous [lengthy] post which I cannot edit because of a 'spam' warning, I should have said that El Capitan, not Sierra, was the last officially supported version of MacOS on the MacPro 3,1.)

Update: Just ordered the following for same-day delivery...
I prefer the Socket 1 (NGFF M.2) approach versus the use of WiFi/BT PCIe cards because:
  • External antenna connections are already provided on the I/O shield.
  • PCIe cards with on-board Bluetooth require a cable to be connected between the card and a USB header on the motherboard. The Socket 1 solution is fully self-contained.
Finally, some photos:
  • Stack of components.
  • Cable management 1.
  • Cable management 2.
  • RGB overkill (just because I can :))
Click to expand...

Your machine looks fantastic. I'm slightly worried about the thunderbolt issue, but I feel like i'll go for something very similar very soon.
 
pastrychef said:
  • Try installing EmuVariableUEFI-64 and RC scripts to see if it helps your shutdown issues.
  • The BCM94360CS2 with adaptor that you ordered should work in the M.2 slot.
  • Try searching around for the Smalltree drivers for your second NIC.
Click to expand...
Wow, thanks @pastrychef! I just downloaded the Smalltree driver, which I'll test when I get home this evening, but found this in the info.plist:
Code: 
    <dict>
        <key>SmallTreeIntel82576</key>
        <dict>
            <key>CFBundleIdentifier</key>
            <string>com.smalltree.driver.SmallTreeIntel82576</string>
            <key>IOClass</key>
            <string>SmallTreeIntel82576</string>
            <key>IOPCIPrimaryMatch</key>
            <string>0x15218086 0x15228086 0x15398086</string>
            <key>IOPCITunnelCompatible</key>
            <true/>
            <key>IOProviderClass</key>
            <string>IOPCIDevice</string>
            <key>IOResourceMatch</key>
            <string>IOKit</string>
            <key>IOUserClientClass</key>
            <string>STECUserClient</string>
        </dict>
    </dict>

The Device ID 0x1539 and Vendor ID 0x8086 of the i211 are listed in the kext (IOPCIPrimaryMatch)!

Getting the second NIC to function is important because I have a managed switch in my home office. My MacPro 3,1, ReadyNAS, and QNAP NAS all have dual Gigabit NICs and all of these devices are connected to the switch with 802.3ad LACP link aggregation. Actually, this is another reason I chose the Designare motherboard! I'd like to get the Hackintosh onto LACP as well without using the gigabit Ethernet port on the OWC Thunderbolt 3 dock.

By the way, on this motherboard, EmuVariableUEFI-64 has always produced fatal "Cannot allocate memory" errors at boot with every one of the memory fix drivers. Nevertheless, I can try this again.
 
Last edited:
Ethernet Port 2 (Intel I211 PRO/1000) is finally working! This means both of the Gigabit ports on the motherboard are now operational.

The code segment I posted above is actually from a modified version of the Smalltree 82576 kext posted here by @the-darkvoid. Perhaps they took the standard 82576-1.2.5 kext and added the "0x15398086" device/vendor ID because this version is not available from the Smalltree website.

Getting the port to work required a couple of extra steps based on this post by @slawa:
  • Copy Rehabman's FakePCIID.kext to CLOVER/kexts/Other.
  • Because the Smalltree driver looks for its own "subsystem vendor" ID (0x000A), we have to replace Intel's subsystem vendor ID with Smalltree's ID. This bit of fakery is performed by the modified FakePCIID_Intel_GbX.kext which must also be copied to CLOVER/kexts/Other. Because the Device ID (0x1539) is already supported by the Smalltree kext, I did not have to add a RM,device entry to this kext. Only the RM,subsystem-id had to be changed to 0x000A. Additionally, I added the device ID 0x15398086 to the IOPCIPrimaryMatch field so that the spoofing takes place on this particular device. Lastly, I changed the model field to Intel I211 Gigabit Ethernet. The modified <dict> section of the kext is shown below.
Code: 
<dict>
        <key>Intel GbX</key>
        <dict>
            <key>CFBundleIdentifier</key>
            <string>org.rehabman.driver.FakePCIID</string>
            <key>FakeProperties</key>
            <dict>
                <key>model</key>
                <string>Intel I211 Gigabit Ethernet</string>
                <key>RM,subsystem-id</key>
                <data>
                CgAAAA==
                </data>
            </dict>
            <key>IOClass</key>
            <string>FakePCIID</string>
            <key>IOMatchCategory</key>
            <string>FakePCIID</string>
            <key>IOPCIPrimaryMatch</key>
            <string>0x15398086 0x15d18086 0x15638086</string>
            <key>IOProviderClass</key>
            <string>IOPCIDevice</string>
        </dict>
    </dict>
  • Finally, we have to copy the attached SmallTreeIntel82576.kext to the CLOVER/kexts/10.14 folder. Apparently, this is necessary to ensure correct load order. Kexts in the Other folder are loaded first, then those in OS-specific folders.
A screenshot of the System Report --> Ethernet Cards screen is attached as well.

Posted from Intel I211 Ethernet port :)

Edit: Moved all attachments to Post #1.
 
Last edited:
Update on Broadcom WiFi/BT module with adapter (BCM 94360CS2):
  • Unfortunately there is a PCIe x1 slot located just behind the Socket 1 port (NGFF M.2) on this motherboard (see photo).
  • Although the Broadcom card itself is short enough to fit, the adapter increases the total length by a few millimeters (even if we chop off the perforated section), causing the back end of the adapter to overlap the PCIe slot.
  • So do not use adapters on this motherboard!
 

Attachments

  • Designare-Z390b.jpg
    Designare-Z390b.jpg
    121.4 KB · Views: 900
CaseySJ said:
You probably did this already, but it doesn't hurt to ask: After the F4 BIOS upgrade, all BIOS settings are reset to defaults. Did you subsequently apply the correct settings?
Click to expand...

Indeed I did but thanks for the reply. It won't book off the usb drive either so I'll dig into it when I have some free time. On a side note, the F4 update seems to have reduced dpc latency for audio in windows. I'd rather not go back.
 
Build 1
Z390 DESIGNARE |  i7 9700K |  GeForce Other
Could you please answer a few questions for me. Im running a similar build as you.
  1. Which video drivers are you referring to? Im having a problem with my wake/sleep
  2. What do you have darkwake set to?
  3. What are your bios settings?
 
mroverst said:
Could you please answer a few questions for me. Im running a similar build as you.
  1. Which video drivers are you referring to? Im having a problem with my wake/sleep
  2. What do you have darkwake set to?
  3. What are your bios settings?
Click to expand...
  1. Because I have the RX580, no video drivers are needed. For Vega GPUs -- particularly for power management, I believe -- there are a couple of extra recommended steps. In this link, see Installation of the OS step 2.
  2. I have just started playing with darkwake. I've tried 02 and 10, but according to a post I came across (somewhere?), both of these values are identical (binary 00000010 and 00001010) because that 4th bit flag is no longer used. Because I'm still actively testing sleep/wake, I'll provide a status update soon, probably in Post #1.
  3. Shutdown and Reboot are not working. Removing the debug flag in boot section just allows the system to reboot rather than halt on a kernel panic, but does not in itself prevent the kernel panic. So those panics are still happening on Reboot and Shutdown. Any attempt to install EmuVariableUEFI-64 on this motherboard leads to severe global catastrophe. The only reliable recovery procedure requires re-flashing the motherboard BIOS, which clears out the built-in NVRAM. (If there is a way to clear motherboard NVRAM directly, I'd certainly like to know...) I have to examine appropriate logs and core files to see what's causing the kernel panic.
  4. My current BIOS settings are (only the ones I've changed after choosing "Load Optimized Defaults"):
    • Memory XMP Profile 1
    • Intel Vt-d disabled
    • Enable IGPU always
    • XHCI Hand-off enabled
    • Operating System other
    • CSM Support disabled (this also disables Secure Boot)
    • Trusted Computing disabled
    • Network Stack Configuration disabled
    • Thunderbolt: please see TB3 BIOS screenshots in Post #61.
    • I'll post a more complete list soon. These are off the top of my head...
 
Last edited:
Status
Not open for further replies.
Copyright © 2010 - 2024 tonymacx86 LLC
Back
Top