[pre-release] macOS Ventura

[Middleman]

Just now, Cyberdevs said:

In my case (SkyLake, KabyLake on Z170X mainboards) the installation process starts ok and even completes the first stage after reboot it stops at:

 

See if you could add -disablegfxfirmware to your boot-arg. That should allow it to pass.

[Cyberdevs]

5 minutes ago, Middleman said:

See if you could add -disablegfxfirmware to your boot-arg. That should allow it to pass.

I'll give it a try and see what happens

[ndungu6678]

4 hours ago, Middleman said:

Judging by his EFI and System Profile, I would think he was using around OC 0.5.9-0.6.5. because Bootstrap was removed after 0.6.6.

Also he has spoofed what looks like a Comet Lake i9 chip to his i5 9400F setup.

I am actually using latest O.C 0.8.2! as you can confirm from Hackintool below- yes bootstrap was removed long ago but you can still enable it if you know how- I intentionally enabled it -using RequestBootVarRouting=YES in OC 0.8.2 in the UEFI settings,. My SMBIOS iMac 20,2 is detects my i5 9400F as a 6 core Comet Lake i9 on its own! from way back in Monterey and Big Sur without any spoofing- the only spoofing changes I made were to insert new Cupid1Mask and Cupid1Data values for Alder lake processors recently when experimenting-:

 

Edited June 8, 2022 by ndungu6678

[Cyberdevs]

@Middleman

didn't work. same error.

[Middleman]

Just now, eSaF said:

That is exactly where my rig is failing!!!! On the reboot to continue with the install, it reverts back to the beginning. 

What SMBIOS are you using?

[MorenoAv]

My EFI contains -no_compat_check in boot arguments and it fails all the same...

[Middleman]

Just now, Cyberdevs said:

@Middleman

didn't work. same error.

Did you look at aben's post? You have to set the IGPU to Kaby Lake for it to boot > 

 

[Middleman]

Just now, MorenoAv said:

My EFI contains -no_compat_check in boot arguments and it fails all the same...

There can be any number of reasons that it fails. It's obvious that Apple has made some changes to the underlying kernel

now that models up to Skylake have been deprecated.

 

Have you checked what are your SecureBoot settings btw? Is it set to default?

[Cyberdevs]

9 minutes ago, Middleman said:

What SMBIOS are you using?

I've tried many SMBIOSes and various settings. From iMac18,3 to MacPro7,1

yes I've seen his post.

[aben]

@Cyberdevs @miliuco @eSaF Hi guys, test with below settings: 
 

1. Kexts: Have only parent VirtualSMC kext enabled and all its fellow plug-ins (SMC/Processor and SuperIO) disabled. For safe measure include boot-arg: -vsmcbeta
2. Quirks (apart from AvoidRuntimeDefrag): 

• SyncRuntimePermissions - YES under Booter --> Quirks
• SetApfsTrimTimeout - 0 under Kernel --> Quirks
• SecureBootModel - Disabled under Misc --> Security
• AdviseFeatures - YES under PlatformInfo --> Generic
• SpoofVendor - NO under PlatformInfo --> Generic
• Min Date and Min Version to -1 under UEFI --> APFS
• DataHub - YES under UEFI --> ProtocolOverrides
• FirmwareVolume - YES under UEFI --> ProtocolOverrides

Settings from my side so hopefully one or more of the above should help.

Edited June 8, 2022 by aben

[Middleman]

24 minutes ago, eSaF said:

iMac (Retina 5K, 27-inch, 2020)

 

The crazy part is it takes me up to the point to choose Country and continues with the install right to the point for second reboot to carry on the installation and that's where it fails and falls back to the beginning.

 

BTH - Those extra boot-args don't work on my rig.

Right. If that’s the case it might be worthwhile to create a debug version of Opencore to try and see where it’s crashing. It sounds like a memory issue to me but likely also SecureBoot permissions playing up as well as SMBIOS/CPUID. You may have to also try test spoof Kaby Lake CPUID of EB060800 00000000 00000000 00000000 and use SMBIOS of iMac19,1 if it helps. Also change the 8 to a 9 in CPUID value for Coffee Lake if needed.

 

As it goes I’m currently using v8.2 of OC and to boot my Rocket Lake I used Comet Lake CPUID and CML UHD630 IGPU settings.

[ndungu6678]

1 hour ago, aben said:

@Cyberdevs @miliuco @eSaF Hi guys, test with below settings: 
 

1. Kexts: Have only parent VirtualSMC kext enabled and all its fellow plug-ins (SMC/Processor and SuperIO) disabled. For safe measure include boot-arg: -vsmcbeta
2. Quirks (apart from AvoidRuntimeDefrag): 

• SyncRuntimePermissions - YES under Booter --> Quirks
• SetApfsTrimTimeout - 0 under Kernel --> Quirks
• SecureBootModel - Disabled under Misc --> Security
• AdviseFeatures - YES under PlatformInfo --> Generic
• SpoofVendor - NO under PlatformInfo --> Generic
• Min Date and Min Version to -1 under UEFI --> APFS
• DataHub - YES under UEFI --> ProtocolOverrides
• FirmwareVolume - YES under UEFI --> ProtocolOverrides

Settings from my side so hopefully one or more of the above should help.

 

Thanks -confirmed- made all your changes above and my hack still boots fine!

 

Build 1: iMac20,2, i5-9400F(seen as 2.9 GHz 6-Core Intel Core i9 ?), Samsung 16GB DDR4 RAM, MSI Z390 Gaming Plus , Radeon RX580 8GB, OC 0.8.2

Edited June 8, 2022 by ndungu6678

[Cyberdevs]

@aben

I made all the changes but still the same error 😏

[ghost8282]

On 6/7/2022 at 3:07 AM, Pavo said:

Just a FYI, disabling the AvoidRuntimeDefrag quirk does not work for macOS as a VM on KVM+Qemu.

Hi Pavo, pinging directly as it seems we are a few people here discussing how to run mac os in kvm+qemu.

In my case AvoidRuntimeDefrag set to false made me proceed a little bit further, I was able to get to the graphical installer with iMacPro1,1 smbios, but I'm still not able to boot, the os reboots 2 times during installation, all ok, it creates and install files on the 'Ventura/Macintosh-hd' hd, then either hangs (with passed through gpu) with nothing useful in the log:

AmiShimTimerBoostExit changed current period to 100000

 

or with qxl/vnc only it bootloops at the same step instead of hanging.

I'm passing through 2 sandy bridge cpus, maybe the issue is there I really don't know, maybe some fixes to opencore are needed.

 

Did you make any progress?

 

Btw, for others, AvoidRuntimeDefrag=false is able to boot monterey disk, but not the ventura installer, nor complete the installation

Using latest ovmf from audk repo, sync with edk2 stable 202202

Edited June 8, 2022 by ghost8282

[ndungu6678]

5 hours ago, eSaF said:

Even if you compress and zip it? 

 

Ok I can see why its a problem because of the size. You've near enough got everything included and enabled in the config.plist i.e kexts and drivers even for Legacy installation so I dare say among all those files, kexts, and settings something stuck and allowed the installation. To be honest given all of that, I wouldn't know where to start sorting all that out for a comparison but thanks for the upload, maybe someone more knowledgeable than my humble self can glean something from your config.plist. Thanks.

 

Even after compression, full EFI is too big to post!-35MB!-LoL- have to admit I have a lot of unnecessary junk in it-too scared and too lazy to trim it down in case I remove or disable something and lock myself out from booting my hack!-as I said I'm no guru - have so much still left to learn!

 

Build 1: iMac20,2, i5-9400F(seen as 2.9 GHz 6-Core Intel Core i9 ?), Samsung 16GB DDR4 RAM, MSI Z390 Gaming Plus , Radeon RX580 8GB, OC 0.8.2

Edited June 8, 2022 by ndungu6678

[Middleman]

2 hours ago, ndungu6678 said:

I am actually using latest O.C 0.8.2! as you can confirm from Hackintool below- yes bootstrap was removed long ago but you can still enable it if you know how- I intentionally enabled it -using RequestBootVarRouting=YES in OC 0.8.2 in the UEFI settings,. My SMBIOS iMac 20,2 is detects my i5 9400F as a 6 core Comet Lake i9 on its own! from way back in Monterey and Big Sur without any spoofing- the only spoofing changes I made were to insert new Cupid1Mask and Cupid1Data values for Alder lake processors recently when experimenting-:

 

I know what you're saying here - but that is CPU spoofing you are doing right there.

Well value you have above for the so-called Álder Lake' processor is actually Comet Lake i9! I know as I was the one of the first few to find

that and post that particular value on my Z490 guides for High Sierra/Catalina. As for the CPU values the system does derive it from the SMBIOS.
Which is why when choosing a SMBIOS you have to find one that a) matches the processor to the best similar-specced Mac model and b) find

one that also matches the graphics card line-up as some of the more basic SMBIOSes doesn't have a DGPU (which is why they'll crash on boot).

[Common_Sense]

With my 4770k in sign I now have the following settings: AvoidRuntimeDefrag = true, iMac18,3 + spoof CPU with Clover FakeCPUID = 906E9 (7700K). 

 

Now I get an instant reboot with both Ventura and Monterey. With Monterey no longer hanging.

 

Hopefully this means that I have gone 1 step further, and now with a good CPU + SMBIOS combo, can try and fix this problem, and maybe then it can boot?

 

I don't know, am I trying something stupid? Should I try another SMBIOS + CPU fake combination? Anyone knows how to fix the reboot issue and/or why it happens? 

 

Lets see if we can solve the issue  

[aben]

2 hours ago, Cyberdevs said:

@aben

I made all the changes but still the same error 😏

Ok, try enabling one more quirk along with the others mentioned earlier: MaxBIOSVersion under PlatformInfo --> Generic. I had this enabled as well while getting to install macOS 13.

[Pavo]

1 hour ago, ghost8282 said:

Hi Pavo, pinging directly as it seems we are a few people here discussing how to run mac os in kvm+qemu.

In my case AvoidRuntimeDefrag set to false made me proceed a little bit further, I was able to get to the graphical installer with iMacPro1,1 smbios, but I'm still not able to boot, the os reboots 2 times during installation, all ok, it creates and install files on the 'Ventura/Macintosh-hd' hd, then either hangs (with passed through gpu) with nothing useful in the log:

AmiShimTimerBoostExit changed current period to 100000

 

or with qxl/vnc only it bootloops at the same step instead of hanging.

I'm passing through 2 sandy bridge cpus, maybe the issue is there I really don't know, maybe some fixes to opencore are needed.

 

Did you make any progress?

 

Btw, for others, AvoidRuntimeDefrag=false is able to boot monterey disk, but not the ventura installer, nor complete the installation

Using latest ovmf from audk repo, sync with edk2 stable 202202

Unfortunately I couldn’t get it to boot when using the normal OVMF, but when switched to OVMF TP version, it would boot but would reboot without a panic on drive selection. Still testing varies things without success right now.

[Cyberdevs]

I guess the issue is a combination of serval things and related to VirualSMC (Apple's SMC needs to be updated) and the nvram. The SMBIOS settings also need to be updated as well:

Here's the data from the 2017 iMac I tested yesterday:

 

Apple iMac 2017:

  System Firmware Version: 451.120.7.0.0

  OS Loader Version: 540.120.3~6

  SMC Version (system): 2.41f2

 

and here is the data from iMac18,3 SMBIOS generated by OCC:

 

Hackintosh:

  System Firmware Version: 447.80.3.0.0

  OS Loader Version: 540.120.3~6

  SMC Version (system): 2.41f2

 

[ghost8282]

41 minutes ago, Cyberdevs said:

and here is the data from iMac18,3 SMBIOS generated by OCC:

 

Hackintosh:

  System Firmware Version: 447.80.3.0.0

Hi, please see/use latest version of oc from the repo or dortania, databases were updated, in your case for imac18,3 are now:

BIOSVersion: "451.120.7.0.0"
BIOSLegacyVersion: "IM183.88Z.F000.B00.2203161719"
BIOSReleaseDate: "03/16/2022"

 

42 minutes ago, Pavo said:

but when switched to OVMF TP version

Nice thought!

Tried also the tpm version, compiled edk2 202205 (which I'm using in a windows 11 vm); there's no difference in my case compared to audk 202202 with no tpm support.

It always hangs in the log at:

12:145 00:043 OCSMC: SmcReadValue Key 4D535463 Size 1
12:151 00:005 OCSMC: SmcReadValue Key 4D534163 Size 2
12:158 00:007 AmiShimTimerBoostExit changed current period to 100000

Edited June 8, 2022 by ghost8282

[aben]

7 minutes ago, Cyberdevs said:

I guess the issue is a combination of serval things and related to VirualSMC (Apple's SMC needs to be updated) and the nvram. The SMBIOS settings also need to be updated as well:

Here's the data from the 2017 iMac I tested yesterday:

 

Apple iMac 2017:

  System Firmware Version: 451.120.7.0.0

  OS Loader Version: 540.120.3~6

  SMC Version (system): 2.41f2

 

and here is the data from iMac18,3 SMBIOS generated by OCC:

 

Hackintosh:

  System Firmware Version: 447.80.3.0.0

  OS Loader Version: 540.120.3~6

  SMC Version (system): 2.41f2

 

Do you possibly have other macOS volumes present on the drive you wish to install macOS 13 to? If so, I would suggest adding an APFS-container partition rather than a volume. You should be prompted with something like this to achieve this:

 

[Cyberdevs]

@ghost8282

Thanks for the info but I've already tried that.

 

@aben

Actually yes, I tried the clean install method and I even I ran the installer from an exiting macOS and once it was done the Ventura boot process nags about nvram and then stops at AppleSMC and then reboots.

[PropoFlexWX4]

macOS Ventura: The Unreachable Adventure LOL.

Watching this thread closely. I'm sure a solution will come by in the next few days.

 

That is -- if the solution is not something very simple, maybe just a setting we forgot to enable or disable in config plist -- and it's been hiding under our noses....

[prcmelo]

I'm also following this thread closely, anyone knows if still there are Kexts for Radeon R9 Series in macOS Ventura? I'm running a R9 380 and i7 4790 in Monterey, and I'm curious to know if maybe I'm on a "Dead End"... 😕