Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
The first post of this thread is a WikiPost and can be edited by anyone with the appropiate permissions. Your edits will be public.

khronokernel

macrumors 6502
Sep 30, 2020
278
1,425
Alberta, Canada
What am I missing - is volume sealing or snapshotting involved? Can the hash be re-calculated and thus fixed somehow?
I assume you're referencing the _authenticate_root_hash patch, there shouldn't be any additional patches needed. We've been using it internally without much issue on several units (Macmini4,1, Macmini5,1, MacPro5,1, MacBookPro9,2, etc)

However regarding recalculating the hash, not possible without Apple's XBS system internally.

Could you verify the patch itself is enabled inside the config.plist under Kernel -> Patch? It should look similar to the one attached. Otherwise if that's still correct, I found _is_root_hash_authentication_required as another function you could try to patch to return zero (simply replace the Base with _is_root_hash_authentication_required)
Screen Shot 2022-06-19 at 2.46.14 PM.png
 

startergo

macrumors 603
Sep 20, 2018
5,022
2,283
  1. In a hex editor, replace all instances of Longavi with Berbice in the kext's executable
  2. Mount writable system volume
  3. Copy the modified kext along with AppleIntelSKL* from Monterey to writable volume's SLE. Skip AppleIntelSKLGraphicsMTLDriver. It doesn't work yet.
  4. Rebuild kernel collections with kmutil
  5. Create new snapshot with bless and reboot
Why not just creating SSDT with Berbice injected and adding AppleIntelSKL* from Monterey?
 
  • Like
Reactions: Larsvonhier

Nothistime

macrumors newbie
Nov 13, 2021
11
0
Installed on iMac 21.5 Late 2013 Iris Pro, everything works, only video acceleration is missing, no errors found OCLP 0.4.6
Great, but how? Did you spoofed your Mac model with a newer one (like in my case from MacBookPro12,1 to MacBookPro15,2)?
 

zzzippp

macrumors member
Jan 27, 2006
51
47
Portland, Oregon
I assume you're referencing the _authenticate_root_hash patch, there shouldn't be any additional patches needed. We've been using it internally without much issue on several units (Macmini4,1, Macmini5,1, MacPro5,1, MacBookPro9,2, etc)

Does anyone know if a Patching invalid size debug message is normal?

Code:
26:856 00:013 OCAK: Patching invalid size 18B000 with DC4000 for com.apple.filesystems.apfs
26:865 00:008 OC: Prelinked patcher result 4 for com.apple.filesystems.apfs (Disable Root Hash validation) - Success

Output from OC debug binary, v8.1
 
Last edited:

Larsvonhier

macrumors 68000
Aug 21, 2016
1,611
2,983
Germany, Black Forest
I assume you're referencing the _authenticate_root_hash patch, there shouldn't be any additional patches needed. We've been using it internally without much issue on several units (Macmini4,1, Macmini5,1, MacPro5,1, MacBookPro9,2, etc)

However regarding recalculating the hash, not possible without Apple's XBS system internally.

Could you verify the patch itself is enabled inside the config.plist under Kernel -> Patch? It should look similar to the one attached. Otherwise if that's still correct, I found _is_root_hash_authentication_required as another function you could try to patch to return zero (simply replace the Base with _is_root_hash_authentication_required)
View attachment 2021271
Yes, that was exactly what I was missing - thanks! Got it up and running on several machines now. Had to add the Kernel/Patch entry manually in the config.plist file after generating it with various OCLP nightlies.

Common denominator running Ventura was (aside from expectedly slow gfx) that on each machine the loginwindow process crashed once after login, then on 2nd attempt made it to the finder desktop.
Crash reporter blamed "Dispatch queue: com.apple.root.default-qos" as crashed thread.

Tested machines so far:
Mac Pro 4,1 (fw flashed to 5,1) - needed external USB hub for keyboard/mouse, ethernet not working but Wifi did (not sure if module was replaced by newer one, currently BCM43xx)
MacBook Pro 6,1 17" - internal USB/trackpad/keyboard working oob
MacBook 7,1 - needed ext. USB hub for keyboard/mouse as internal trackpad/kbd does not yet work
MacBook 6,1 - same as 7,1 except that Wifi works oob

tbc´d
 
Last edited:

0xCUBE

Suspended
Original poster
Jun 6, 2022
52
218
Planet Earth
Great, but how? Did you spoofed your Mac model with a newer one (like in my case from MacBookPro12,1 to MacBookPro15,2)?
since OCLP doesn't support spoofing to ventura models by default, if you can't figure something like this out then you should not be trying ventura. Sorry for sounding harsh but it's the truth. There also really are too many bugs for non-devs to even think about using it
 

Larsvonhier

macrumors 68000
Aug 21, 2016
1,611
2,983
Germany, Black Forest
To @0xCUBE and all: Can confirm Ventura booting to Finder on MacBookAir4,2 (13").
Trackpad (just as a single button mouse), keyboard and external USB already work.

Also nice: Teleport v1.3.5 works on Mx Ventura, so remote controlling of esp. MacBook4,1 (when the time is ripe) is still viable.
 
Last edited:

khronokernel

macrumors 6502
Sep 30, 2020
278
1,425
Alberta, Canada
Mac Pro 4,1 (fw flashed to 5,1) - needed external USB hub for keyboard/mouse, ethernet not working but Wifi did (not sure if module was replaced by newer one, currently BCM43xx)
MacBook Pro 6,1 17" - internal USB/trackpad/keyboard working oob
MacBook 7,1 - needed ext. USB hub for keyboard/mouse as internal trackpad/kbd does not yet work
MacBook 6,1 - same as 7,1 except that Wifi works oob
I assume you already know, but USB 1.1 and legacy Intel Ethernet support was stripped from Ventura. I believe the UHCI/OHCI portion is still broken even after injection through OC, so you'll need to downgrade the whole IOUSBHostFamily stack on disk (this has been confirmed to work)

This is why on Penny units, the internal keyboard and trackpad broke. Internally their wired through the OHCI controller, which now has no drivers.

Our GitHub issue does document this, and the ventura-alpha branch has patches (though as mentioned above, USB1.1 still needs to be researched):
after generating it with various OCLP nightlies.
Ensure you adjust the nightly link, currently nothing in mainline supports Ventura. We're trying to keep experimental support in it's own dedicated branch.
Trackpad (just as a single button mouse)
This is due to the removal of pre-Force Touch drivers, just need to reinject AppleUSBTopCase.kext and AppleUSBMultitouch.kext (performed in the ventura-alpha branch)
 

Aston441

macrumors 68030
Sep 16, 2014
2,607
3,948
To @0xCUBE and all: Can confirm Ventura booting to Finder on MacBookAir4,2 (13").
Trackpad (just as a single button mouse), keyboard and external USB already work.

Also nice: Teleport v1.3.5 works on Mx Ventura, so remote controlling of esp. MacBook4,1 (when the time is ripe) is still viable.
Wow. That’s good news, nice work! No vid acceleration I assume?
 
  • Like
Reactions: Larsvonhier

Larsvonhier

macrumors 68000
Aug 21, 2016
1,611
2,983
Germany, Black Forest
I assume you already know, but USB 1.1 and legacy Intel Ethernet support was stripped from Ventura. I believe the UHCI/OHCI portion is still broken even after injection through OC, so you'll need to downgrade the whole IOUSBHostFamily stack on disk (this has been confirmed to work)

This is why on Penny units, the internal keyboard and trackpad broke. Internally their wired through the OHCI controller, which now has no drivers.

Our GitHub issue does document this, and the ventura-alpha branch has patches (though as mentioned above, USB1.1 still needs to be researched):

Ensure you adjust the nightly link, currently nothing in mainline supports Ventura. We're trying to keep experimental support in it's own dedicated branch.

This is due to the removal of pre-Force Touch drivers, just need to reinject AppleUSBTopCase.kext and AppleUSBMultitouch.kext (performed in the ventura-alpha branch)
Thanks for clarification and infos on the USB kexts.
I might have phrased it a bit imprecise: I have not been able to use any of the Ventura alpha OCLP branch precompiled builds (TUI/GUI) to correctly set the _authenticate_root_hash patch in the OC plist. Had always to do it with OC config tool manually. Might be a usage error on my side or a faulty handling in the alpha release...
 
Last edited:

khronokernel

macrumors 6502
Sep 30, 2020
278
1,425
Alberta, Canada
Might be a usage error on my side or a faulty handling in the alpha release...
Are you certain you're using the right branch? The logic for enabling the patch is right here:
Python:
if smbios_data.smbios_dictionary[self.model]["CPU Generation"] <= cpu_data.cpu_data.ivy_bridge.value:
    print("- Allowing swapped dyld shared cache in Ventura")
    self.get_item_by_kv(self.config["Kernel"]["Patch"], "Comment", "Disable Root Hash validation")["Enabled"] = True

The CPU Generation value is defined in smbios_data.py and cpu_data.py.

I verified that the builds uploaded to Github actions are working as well. Maybe you're able to see the fault in my logic somewhere, I'm a bit stumped.

Screen Shot 2022-06-22 at 9.17.40 AM.png
 

Aston441

macrumors 68030
Sep 16, 2014
2,607
3,948
No video acceleration and more importantly no AVX2 emulation. Ventura beta 1 is currently full of such instructions and without a solution there will be definitively no Ventura on pre Haswell (pre Late 2013) Macs.
>no AVX2

Ouch. That’s bad. No way to recompile the kernel with that flag set huh?
 

khronokernel

macrumors 6502
Sep 30, 2020
278
1,425
Alberta, Canada
Can confirm Beta 2 does still work with the dyld swapping technique, though still nothing to report graphics wise with the community. Legacy Metal is still quite a challenge, and the lack of a KDK yet is quite concerning from a patcher perspective (as we're at the whim of Apple to seed them to the public)

Hopefully we'll see it around when the InstallAssistant.pkg is seeded.

Screenshot 2022-06-22 at 12.24.40 PM.png
 

BeatCrazy

macrumors 603
Jul 20, 2011
5,140
4,517
Ventura beta 1 is currently full of such instructions and without a solution there will be definitively no Ventura on pre Haswell (pre Late 2013) Macs.

I believe you, and understand. But then we have posts like this:

Tested machines so far:
Mac Pro 4,1 (fw flashed to 5,1) - needed external USB hub for keyboard/mouse, ethernet not working but Wifi did (not sure if module was replaced by newer one, currently BCM43xx)
MacBook Pro 6,1 17" - internal USB/trackpad/keyboard working oob
MacBook 7,1 - needed ext. USB hub for keyboard/mouse as internal trackpad/kbd does not yet work
MacBook 6,1 - same as 7,1 except that Wifi works oob

Seems like there should be some clarification on this in the Wiki? Anyone casually reading is either going to get confused (like me), or have incorrect expectations and start to litter this thread with questions about potential support.
 
  • Like
  • Sad
Reactions: olad and prefuse07

OnawaAfrica

Cancelled
Jul 26, 2019
470
377
Last os for intel macs probably ? (Leopard was the last for ppc macs)

Next year they start dropping m1 macs ? (like they did with the first intel chips that was not core2duo)

I know its a wild guess, but Apple .... ?
apple is not that aggressive in dropping new products usually they support their hardware at least for 5 or more major os releases
 

Heindijs

macrumors 6502
May 15, 2021
426
881
@trifero (I can't quote your post because there is no text)

Geekbench scores of course mean nothing. Even my 2012 Mac Mini outperforms even 2018 Macbook Airs. Yeah it still is ridiculous that Apple doesn't want to support the 2016 mbp

I mean come on..
2022-06-23 13_07_26-Mac Pro (Early 2008) Benchmarks - Geekbench Browser.png
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.