Contribute
Register

[Solved] Migrating from legacy to UEFI

Status
Not open for further replies.
Joined
Jan 30, 2011
Messages
9
Motherboard
GA-Z370XP SLI
CPU
i7 8700
Graphics
RX 580 4GB
Mac
  1. iMac
  2. Mac Pro
Reaching out to the experts community here, as I need help with converting to UEFI boot.
I’m not exactly a noob as I’ve been a happy hackintosher for over 10 years.

Anyway, I recently successfully and painlessly upgraded from a GA-Z68X-UD4-B3 to a GA-Z370XP-SLI with an i7-8700, running High Sierra 10.13.6.
Everything working just fine.

I have always been using the legacy type of boot, but since I now have a proper motherboard which supports UEFI, it’s time to make the switch.

Except no matter what I try (BOIS Vt-d and other settings, drivers64UEFI drivers, config.plist changes...), it simply won’t boot. It’s probably some silly thing I just overlooked, but I just can’t find what it is. I hope someone can guide me to the right direction.

I have 2 physical boot disks (cloned), so I can easily experiment things on one of them.

I am attaching data which should help the investigations, but if you need anything more, please just let me know. (not sure why the debug tool claims CLOVER is not installed, when clearly, it is - Using Clover release 4458 as per bootlog).

Whenever I try to boot in UEFI, I get a KP backtracing to AppleACPIPlatform kext.
 

Attachments

  • config.plist
    6.3 KB · Views: 221
  • debug_11590.zip
    330.9 KB · Views: 111
Thanks for moving this to the right thread! :)
 
This boot log is from legacy boot with Clover installed on an HFS plus partition. Can you boot from your Clover EFI installation on your other disk, pressing F2 to save /EFI/CLOVER/MISC/preboot.log and upload that file.

Disable VT-d in firmware settings for the time being.
 

Attachments

  • config.plist
    2.3 KB · Views: 236
Thanks for the reply Vulgo.

Since I cannot boot in UEFI, I had to resort to using the legacy boot. System from which I ran the debug script to provide you with as much information as possible on my config/setup. The systems are 99% the same since legacy (Macintosh HD) is cloned to another disk (Time Machine/UEFI attempt).

When I boot in UEFI from the Time Machine disk (Time Machine is the name of the volume; nothing to do with Apple TM), I get clover loaded from the right disk EFI and everything, picking up the right config and kexts, then starts loading MacOS and KP on some kext.


I have not yet converted to APFS: I didn't feel the need, and didn't want to introduce 2 major changes to the system at the same time (looks like that was a wise thing to do).

Anyway, attached is the preboot log, as requested. I hope this will be helpful to you. It wasn't to me :D
 

Attachments

  • preboot.log.txt
    35.1 KB · Views: 154
Anyway, attached is the preboot log, as requested. I hope this will be helpful to you. It wasn't to me :D
Some suggestions. You can remove APFS driver loader if you aren't booting from APFS, i wouldn't expect it to work with AptioFix2 which is from the past. Try enabling XHCI hand off if it isn't enabled, while there disabling CFG Lock if you have that setting under CPU. If you add 'debug=0x100' and '-v' to your boot arguments you may get more details of the boot. Remove anything in /EFI/CLOVER/ACPI/PATCHED or remove/rename empty DSDT->Name->String parameter in config.plist
 
I find it confusing as to which UEFI drivers to enable in drivers64UEFI: there is very little information found on what they do for some/most of them, while they are being recommended to be used in a number of cases (but never really explained why). Maybe I just missed that information somewhere... I only figured so much that I needed one of these OsxAptioFix to get to load something past clover.

Which OsxAptioFix will APFSDriverLoader work with (when I’m ready to make the conversion)?

XHCI: I confirm it’s always been enabled

CFG Lock:not an option on my mobo

debug=0x100' and '-v': don't return more information

/EFI/CLOVER/ACPI/PATCHED: I confirm it’s always been empty

DSDT: none specified, none picked up

Vt-d: confirmed disabled
 
The config.plist in use for the EFI preboot.log you posted is neither the one you attached to #1 nor the one attached to #3, maybe try the one from #3.
 
You're right. Actually, I have no idea where I got that config.plist in the first place!
It lacks all the custo I need for my system.
Anyway, I just gave it a try, and the system booted. Of course, I lack graphics acceleration, but it booted all the way.
Now I need to figure out what I can retrofit into the config.plist so everything works, without breaking it :D
If you have any leads on what in the config.plist #1 causes the issue, please let me know (it might save me a few 10s of reboots) :D

But many, many thanks again, and well done for spotting all this.
 
You're right. Actually, I have no idea where I got that config.plist in the first place!
It lacks all the custo I need for my system.
Anyway, I just gave it a try, and the system booted. Of course, I lack graphics acceleration, but it booted all the way.
Now I need to figure out what I can retrofit into the config.plist so everything works, without breaking it :D
If you have any leads on what in the config.plist #1 causes the issue, please let me know (it might save me a few 10s of reboots) :D

But many, many thanks again, and well done for spotting all this.
You could try dropping MATS and/or turning off FixHeaders and seeing if that causes the ACPI error. Otherwise enable Clover features one at a time, the defaults for patching and SMBIOS are usually better than incorrect options.
 
All fixed, thanks for all the help :)
 
Status
Not open for further replies.
Back
Top