Contribute
Register

[SUCCESS] Gigabyte Designare Z390 (Thunderbolt 3) + i7-9700K + AMD RX 580

Hi , i would like to desactivate my wifi card CNVI for my z390 aorus master, but how can I modify the SSDT-DESIGNARE-Z390-NO-CNVW.aml file, i try with CotEditor but i think its not really for this type of file extention because of the 'ÿÿÿÿ' and ¤¤' symbol


Capture d’écran 2019-10-14 à 23.32.01.png
 
Hi , i would like to desactivate my wifi card CNVI for my z390 aorus master, but how can I modify the SSDT-DESIGNARE-Z390-NO-CNVW.aml file, i try with CotEditor but i think its not really for this type of file extention because of the 'ÿÿÿÿ' and ¤¤' symbol


View attachment 430612
use macIASL
 
Hi , i would like to desactivate my wifi card CNVI for my z390 aorus master, but how can I modify the SSDT-DESIGNARE-Z390-NO-CNVW.aml file, i try with CotEditor but i think its not really for this type of file extention because of the 'ÿÿÿÿ' and ¤¤' symbol


View attachment 430612
Because you're using a Gigabyte Z390 motherboard, you can use this SSDT without changing it. But if you really want to change something about it, change just the file name to SSDT-GIGABYTE-Z390-NO-CNVW.aml :)
 
Casey, Since getting my Clover install and kexts all caught up I am having terrible issues with slow disk access. Loads of spinning beach balls wether its from local disks or network storage.
I'm attaching my Clover folder if someone could take a look and let me know if I borked something along the way.
Do you really need the following kexts in CLOVER/kexts/Other?
  • VegaTab_64.kext
  • CPUFriend.kext
  • dummy141.kext
  • NoVPAJpeg.kext
  • RadeonMonitor.kext
If these kexts are also copied to /Library/Extensions, consider removing them and rebuilding the kernel cache with Kext Utility.

In CLOVER/ACPI/patched, please delete the VEGA SSDT.

Also copy everything from CLOVER/drivers64UEFI into CLOVER/drivers/UEFI and delete the old drivers64UEFI folder.

In Clover Configurator --> ACPI, do you need Fix Darwin?
 
Let's check a few things first:
  • What are the makes/models of all of the SSDs in question (both the original and the clone)?
  • Because you're upgrading from High Sierra (10.13), what version of Clover is installed on the original disk?
  • In the CLOVER drivers folder (either drivers64UEFI or drivers/UEFI) what files are present?
    • Is there only 1 memory driver?
    • If so, is it OsxAptioFix2Drv-free2000.efi? I continue to recommend this driver at least for initial installation. When the system is fully set up and backed-up, then it's okay to experiment with other drivers such as AptioMemoryFix and OsxAptioFix3Drv.
    • If unsure, please post screenshot of that folder.
  • If you're on Clover 5xxx, has the old drivers64UEFIfolder been renamed or deleted?
    • I usually rename mine to OLD-drivers64UEFI.
    • We can also copy all files from that folder into the new drivers/off folder and then delete drivers64UEFI.
Thanks CaseySJ.

Looks like it may be an Aptio issue. I replaced AptioMemoryFix-64.efi on one of my backups with OsxAptioFix2Drv-free2000.efi and it booted. The kicker was all 4 drives produced the runtime area issue in a row and had to use a spare to reintsall the OS and repair each 10.13 and 10.14.6 EFI folder.

After replacing AptioMemoryFix in each EFI with OsxAptioFix2Drv-free2000.efi and each volume booted again. I guess I'll have to get used to living on the edge. :p

Although I wasn't able to track the runtime area trigger down to a specific event the most common event before boot failure was moving all 3rd party kexts to L/E with hackintool. Seems like when it repairs and rebuilds the kernel cache AptioMemoryFix goes sideways :?:

There were other issues like my stable build having an older version of Clover than the Mojave version, but each time I switched drives I:
  • Shut the machine down and pulled the plug
  • Never had more than a single drive plugged in at the same time unelss strictly cloning.
  • Never attempted to boot from another drive without performing the power cycle above.

Eventually all 4 drives returned the runtime area error in a row, (despite the steps above).
This is why I had said perhaps there's something about Clover and/or NVRAM I'm not grasping... I.E. If having two separate EFI folders can be confuse Clover even though I'm booting from a single drive with a version of Clover that has successfully booted that drive multiple times. (Hope that's explained clearly.)

Finally all EFIs now have the same version of Clover, r4961.

Requested Screenshots...

Probably not a bad idea to upload them anyway and see if everything looks correct...

  • I also had some boot flags suggested when I first set the machine up that might not be needed. Probably not a bad idea to post these and see if any of it's useless cruft...
Attached below...

Also:

  • I added all of the post-install files to the Mojave drive and it works fine except Jpeg preview... NoVPAJpeg fixes the issue. Is this the only workaround or have newer versions of Lilu/WEG resolved this?

  • The EFI folder with post install files set up for 10.14.6 seems to work ok on a spare 10.13 clone. Any reason I might want to avoid using this EFI folder in 10.13?

  • Finally if I were hoping to eventually have a 10.13 and 10.14 drive inside the same machine, is the correct way to set this up is by each drive having its own Clover EFI with RC scripts installed on each one?

Other Stuff:

I tried iMac 19,1 SMBIOS on a clone with 1005 set as the CPU in Clover. Although the numbers in GB are a little better, I top out at the exact same track count in Logic using the New Logic Benchmark. (Linked below in case anyone else wants to give it a go...) At least in terms of Logic, it looks like iMac 18,3 gives me the same real-world performance as 19,1.

New Logic Benchmark:

NewLogicBenchmarkTest Link
NewLogicBenchmarkTest
Thread on gearslutz

[My machine = 154 tracks seamless playback. 10.13/10.14 identical results...
  • Apple 9900k iMac 19,1 = 131 Tracks --- Hello throttling Apple---
  • Apple iMac Pro 8 Core = 166 Tracks. ]

Clover and L/E Screenshots:

10.13 EFI:

10.13 EFI Folder.png


10.13 L/E:
10.13 L-E Installed Kexts.png


10.14.6 EFI:
10.14 EFI Folder.png


10.14.6 L/E:

10.14 L-E Installed Kexts.png


BOOT ARGUMENTS:

10.13-10.14 Boot Arguments.png
 
Last edited:
Hi,

Ah ok. I was just confused when you said " refined minimal config", and correlated it with Clover.
Can't find the reference to "minimal" meaning that default will be used. in the documentation. Perhaps it is not a general specific word for the documentations. Or, it might be that I myself haven't read the documentations well enough. :p

What settings did you feel you didn't need, and what did you feel you needed?

I am not entirely sure how your expectations are of the boot options.

AGPM is highly correlated with AGDP as well. Which in this case, you don't really need to worry, as you most likely won't breech AGDP.

Some examples of different settings between us:
DisableVariableWrite: Unlike you I have no native nvram so I need this on. I can't boot past entering password in FileVault without it either.

ProvideCustomSlide: My testing is still up for debate on this. I started with it on, and when I got the sleep crash, a theory was that since this hides slide=0 from the OS, maybe it was doing something else. I didn't need it, but now that I just fixed sleep a few minutes ago I may enable it, if it does not conflict with me explicitly passing slide=0. While I have not done the memmap method you posted for OC, I have tried it with Clover and nothing actually booted. So far I've only gotten slide=0 to work. Currently it's off but I may turn it back on now if for no other reason than the "This also ensures that slide= argument is never passed to the operating system for security reasons."

SetupVirtualMap: I did not need to enable this to boot and run. I initally started with it on and turned it off. The documentation says it is a quirk for early boot crashes, but I have not had any early boot issues, so I disabled it and have not noticed any side effects so far.

AppleXcpmCfgLock: I unlocked MSR CFG Lock and the documentation says to avoid enabling this setting if possible, and patching CFG Lock allows not enabling it.

DisableIoMapper: I turned this on (in Clover I have dart=0 and drop DMAR table) so that if I leave VT-d enabled I won't have any problems.

Anyway, it looks like I fixed sleep, as I've put my PC to sleep twice and woke without the panic and restart on wake. OC doesn't enable plugin-type=1 on it's own, and my initial SSDT attempt was botched. My latest attempt fixes that. So I have to load an SSDT file to get this working, but it's using native power management now, which seems to fix sleep. I did not load the AGPM kext I made, as it wasn't necessary, but the reading there led me to figure out I didn't have native CPU power management working properly. Without loading the AGPMInjector kext, I haven't noticed any problems and it looked the same as it did in Clover in IORegistryExplorer.

For the boot options, I may just be projecting my experiences with other boot managers, and OC just doesn't behave as I expect, but I would assume that if I enable the boot picker, that should I not choose an option and let the time expire, that it will boot from NVME. But it lists a USB drive for Catalina installer and then my USB SSD bootable backup above my NVME drive. Yes, I can exclude USB from ScanPolicy, but then I'd never be able to boot my bootable backup. I tried RequestBootVarRouting on and off, and I went as far as to setup LegacySchema to accept any variables in case I was excluding something. In your build you don't even use the picker at all and your ScanPolicy filters out the USB devices.
 
Do you really need the following kexts in CLOVER/kexts/Other?
  • VegaTab_64.kext
  • CPUFriend.kext
  • dummy141.kext
  • NoVPAJpeg.kext
  • RadeonMonitor.kext
If these kexts are also copied to /Library/Extensions, consider removing them and rebuilding the kernel cache with Kext Utility.

In CLOVER/ACPI/patched, please delete the VEGA SSDT.

Also copy everything from CLOVER/drivers64UEFI into CLOVER/drivers/UEFI and delete the old drivers64UEFI folder.

In Clover Configurator --> ACPI, do you need Fix Darwin?
Thanks...
Isn't the Vega SSDT required wit the Vega cards?
I thought I had already deleted the old drivers directory.
As far as the FixDarwin... No idea how that got checked.
Restarting now!
 
Some examples of different settings between us:
DisableVariableWrite: Unlike you I have no native nvram so I need this on. I can't boot past entering password in FileVault without it either.

ProvideCustomSlide: My testing is still up for debate on this. I started with it on, and when I got the sleep crash, a theory was that since this hides slide=0 from the OS, maybe it was doing something else. I didn't need it, but now that I just fixed sleep a few minutes ago I may enable it, if it does not conflict with me explicitly passing slide=0. While I have not done the memmap method you posted for OC, I have tried it with Clover and nothing actually booted. So far I've only gotten slide=0 to work. Currently it's off but I may turn it back on now if for no other reason than the "This also ensures that slide= argument is never passed to the operating system for security reasons."

SetupVirtualMap: I did not need to enable this to boot and run. I initally started with it on and turned it off. The documentation says it is a quirk for early boot crashes, but I have not had any early boot issues, so I disabled it and have not noticed any side effects so far.

AppleXcpmCfgLock: I unlocked MSR CFG Lock and the documentation says to avoid enabling this setting if possible, and patching CFG Lock allows not enabling it.

DisableIoMapper: I turned this on (in Clover I have dart=0 and drop DMAR table) so that if I leave VT-d enabled I won't have any problems.

Anyway, it looks like I fixed sleep, as I've put my PC to sleep twice and woke without the panic and restart on wake. OC doesn't enable plugin-type=1 on it's own, and my initial SSDT attempt was botched. My latest attempt fixes that. So I have to load an SSDT file to get this working, but it's using native power management now, which seems to fix sleep. I did not load the AGPM kext I made, as it wasn't necessary, but the reading there led me to figure out I didn't have native CPU power management working properly. Without loading the AGPMInjector kext, I haven't noticed any problems and it looked the same as it did in Clover in IORegistryExplorer.

For the boot options, I may just be projecting my experiences with other boot managers, and OC just doesn't behave as I expect, but I would assume that if I enable the boot picker, that should I not choose an option and let the time expire, that it will boot from NVME. But it lists a USB drive for Catalina installer and then my USB SSD bootable backup above my NVME drive. Yes, I can exclude USB from ScanPolicy, but then I'd never be able to boot my bootable backup. I tried RequestBootVarRouting on and off, and I went as far as to setup LegacySchema to accept any variables in case I was excluding something. In your build you don't even use the picker at all and your ScanPolicy filters out the USB devices.


Sounds good. It gives me a smile to my face when I find someone dedicated and eager to know how MacOS is running on their computer. The entire experience just seems more like an art work.

I don't have CFG unlocked, unfortunately. Though I once modified the UEFI fw for it.But I never used it. :/

When it comes to slide calculation. There are actually two ways, depending on Clover version.

Method 1: [(available start value) - 100000] ÷ 200000 in base-16 to base-10.

or

Method 2: [(available start value) - 100000] ÷ 200000 +1 in base-16 to base-10.

When you do this, you're always suppose to keep it in base-16, till you get the final result out of it.


example:

0000000001A11000 is the available start value.
S0 0x1A11000

Method 1: [0x1A11000 - 0x100000] ÷ 0x200000 = 0xC
Convert to decimal from base-16 of 0xC = 12
slide=12

Method 2: 0x1A11000 - 0x100000 + 0x1 = 0xE
Convert to decimal from base-16 of 0xE = 14
slide=14

Regarding the boot option. Yes, I mentioned it in my post with my uploaded EFI, that it won't show the boot picker.
I just find it much convenient to boot to MacOS that way. And if someone uses that OC-EFI, they can gladly help themselves to show the boot picker. The thing is, I am actually fully blind. So I have to memorise things such as functions in the UEFI firmware GUI, and boot options etc. Memorising how many arrow keys it takes to come to a certain place, what number the right boot volume is and so on. That is the reason why I find it convenient to boot instantly without selecting. A lot of people are unaware of blind people using computers, but we do. I use screen readers such as VoiceOver on Apple platform, and JAWS on Windows. I also use Braille display and avail the chance to OCR often.

I digress...

I take backups, but I generally don't have any backups to boot to. I don't have any bootable OS to boot to, via USB. But I completely understand what you mean. I can easily get how inconvenient that might be.

The main focus of OpenCore is to make it as close as Apple MacOS booting experience is. Even when you have windows, it is suggested to have the bootable media installer made from BootCamp the "non-legacy way". Whilst Apple doesn't really use UEFI, but rather EFI with the extensibility being used more, and CSM support for bootcamp from 2006. So there are perks of using OpenCore, with advantages and disadvantages. Boot options being a downside. Though OpenCore is still beta, so it is quite understandable.
 
Last edited:
What I meant is the driver doesn’t seem to be present in the beta. I guess it will be in the final 10.15.1 release.
Yep. @CaseySJ they aren't in there or my card isn't supported idk, I'm kinda a noob. But after updating still no acceleration. Decided to return the card and get a Vega 56.
 
Back
Top