Contribute
Register

[GUIDE] Catalina on HP EliteDesk 800 G4/G5 Mini - The Perfect MacMini8,1 Hackintosh - CLOVER & OC

*** Deleted ***
Superseded by this post.
 
Last edited:
I did some minor tweaks and got my G5 to have an even higher single-core score with a 65W i9-9900 than my previous Z390/i9 9900K at stock... Multi-Core remains still a good 20% lower and I can see some thermal throttling towards the end of the test. I am wondering if the fan profile can be changed as I don't hear the fan kick in until a few seconds after the CPU is throttling.
Also upgraded my G5 to the latest bios version 2.07.01 from 2.06.00. Got no problem.

I am facing something odd with my iwatch unlock. It seems to work right after I enable it but after sometime, the hack detects the watch, attempts to watch unlock but then times out.

Screen Shot 2020-11-07 at 18.49.17.png
 
I did some minor tweaks and got my G5 to have an even higher single-core score with a 65W i9-9900 than my previous Z390/i9 9900K at stock...
Would you mind posting your latest EFI?

By "tweaking," do you mean CPUFriend.kext data adjustments? Did you change your SystemProductName to iMac19,1? I'm comparing our OC EFIs to see if I can find a problem with mine. I am hoping to have a "vanilla" CPU power management like I do with CLOVER, but maybe not possible with OC? I don't see why our rigs should need CPUFriend.

Also, I noticed that your config.plist might be from a previous version of OC. You might want to look at the following:
  • There appears to be a new attribute Kernel > Quirks > DisableLinkeditJettison (something to do with Lilu stability)
  • You shouldn't need Kernel > Quirks > XhciPortLimit = 1 with USBInjectAll.kext / SSDT-UIAC. I think you should be able to set Kernel > Quirks > XhciPortLimit = 0
  • There appears to be a new config.plist block Kernel > Scheme. Not sure what it does.
  • There appear to be new attributes in Misc > Security
  • There appears to be a new attribute PlatformInfo > Generic > ProcessorType
 
Ok - my GeekBench 5 scores are back to normal with OC. The only change I made was the following attributes in my OC config.plist:
  • Misc > Debug > AppleDebug: Changed from 1 to 0
  • Misc > Debug > Target: Changed from 67 to 3
No surprise that the debug / logging level impacts performance. My updated config.plist is attached. If you use the attached config.plist (replace the one in the EFI posted here), replace the ***REPLACE*** values in PlatformInfo > Generic with your own values.

EDIT: So far, my rig has better CPU power management (lower CPU idle/temps and better GeekBench 5 scores) with CLOVER than with OC. I'd like to think I can achieve the same performance with OC and without using CPUFriend, so there must be something I'm missing. For now, CLOVER r5122 will remain my baseline.

Screen Shot 2020-11-09 at 9.47.36 PM.png
 

Attachments

  • config.plist.zip
    5.6 KB · Views: 113
Last edited:
Would you mind posting your latest EFI?

By "tweaking," do you mean CPUFriend.kext data adjustments? Did you change your SystemProductName to iMac19,1? I'm comparing our OC EFIs to see if I can find a problem with mine. I am hoping to have a "vanilla" CPU power management like I do with CLOVER, but maybe not possible with OC? I don't see why our rigs should need CPUFriend.

Also, I noticed that your config.plist might be from a previous version of OC. You might want to look at the following:
  • There appears to be a new attribute Kernel > Quirks > DisableLinkeditJettison (something to do with Lilu stability)
  • You shouldn't need Kernel > Quirks > XhciPortLimit = 1 with USBInjectAll.kext / SSDT-UIAC. I think you should be able to set Kernel > Quirks > XhciPortLimit = 0
  • There appears to be a new config.plist block Kernel > Scheme. Not sure what it does.
  • There appear to be new attributes in Misc > Security
  • There appears to be a new attribute PlatformInfo > Generic > ProcessorType

Attached my latest EFI. No I did not switch to the iMac19,1. I have not gotten around to test this but based on what I saw from my previous clover build (the 9900K on a Z370N motherboard), it did not do much.

Thanks for all the suggestions. I will look into them now.
Oddly I seem to be getting higher GeekBench scores with OC than what I was getting on clover.
 

Attachments

  • EFI.zip
    41.1 MB · Views: 184
Tried it but in my case, USB doesn’t work so I couldn’t login to the system. Is anything extra to be done to make the usb ports work?

EDIT: Mmmm, I can see this in the USBPorts.kext file:
<dict>
<key>Macmini8,1-XHC</key>
<dict>
<key>CFBundleIdentifier</key>
<string>com.apple.driver.AppleUSBMergeNub</string>

I have changed the SMBIOS to iMac19,1. Should the dictionary name be changed?

EDIT 2: Tried it but doesn't work :(
 
Last edited:
Tried it but in my case, USB doesn’t work so I couldn’t login to the system. Is anything extra to be done to make the usb ports work?

EDIT: Mmmm, I can see this in the USBPorts.kext file:
<dict>
<key>Macmini8,1-XHC</key>
<dict>
<key>CFBundleIdentifier</key>
<string>com.apple.driver.AppleUSBMergeNub</string>

I have changed the SMBIOS to iMac19,1. Should the dictionary name be changed?

EDIT 2: Tried it but doesn't work :(
Interesting. I didn't think that USBPorts.kext wouldn't be portable. Try reverting to USBInjectAll.kext / SSDT-UIAC.aml.

Also, you can always create your own USBPorts.kext using Hackintool. To do this:
  1. Remove USBPorts.kext, USBInjectAll.kext and SSDT-UIAC.aml from your EFI. You'll also need to disable these kext references in your config.plist (Kernel > Add).
  2. Set OC's Kernel > Quirks > XhciPortLimit = 1 in your OC config.plist
  3. Boot Catalina (with XhciPortLimit = 1)
  4. Run Hackintool and create your USBPorts.kext (which will also create SSDT-UIAC.aml). You'll need to learn how to use Hackintool to create USBPorts.kext. Read this.
  5. Copy USBPorts.kext to your OC Kexts folder
  6. Enable USBPorts.kext in your config.plist and set XhciPortLimit = 0
  7. Reboot and test USB.
 
Last edited:
hi, could you please check SSDT-PMC.aml, SSDT-PLUG-DRTNIA.aml?
i think, keepsyms=1 is no more needed in bootargs
you don´t need PciRoot(0x0)/Pci(0x1F,0x3) layout-id 14000000 because your bootargs alcid 20

WIN boot i found here: https://www.tonymacx86.com/threads/macos-win-dual-boot-at-same-disk-with-opencore.295892/
I imagine the alcid can be injected either way. I'm not using the alcid bootarg, but I don't see any reason not to. I'm injecting the ID via DeviceProperties since I have other properties there, too.

I did look at Dortania's SSDT-PMC and SSDT-PLUG, but will look at them again to be sure. Those Dortania pre-built SSDTs include conditions to make them "generic" so that they can be applied to the DSDTs of different motherboards. My equivalent SSDTs do the same thing, but are specific to the HP EliteDesk 800 G4 Mini DSDT.

It appears that keepsyms is still applicable to OpenCore: https://dortania.github.io/OpenCore-Install-Guide/config.plist/skylake.html#nvram


Attached my latest EFI. No I did not switch to the iMac19,1. I have not gotten around to test this but based on what I saw from my previous clover build (the 9900K on a Z370N motherboard), it did not do much.

Thanks for all the suggestions. I will look into them now.
Oddly I seem to be getting higher GeekBench scores with OC than what I was getting on clover.
I was hoping to at least match CLOVER performance with OC. Since this is my first attempt at OC, I'm certain that I have more to learn and that it's my mistake (not OC).

Thanks everyone for the suggestions and feedback. Keep it coming! :)
 
My CPU temps and clock have dropped after waiting a minute after booting. Since I performed a fresh Catalina installation for my OC testing, maybe macOS was doing something (including Spotlight indexing). With OC, my CPU idle clock / temps are now low as expected.

With OC, I'm still seeing GeekBench 5 benchmarks lower than I did with CLOVER. Note that I'm running without CPUFriend.kext and I'd rather not add it (would prefer to achieve optimal CPU performance with native Coffee Lake power management).
 
Back
Top