[Solved] "End randomseed" with GA-Z370 Aorus UG 2.0: about to give up...

[ultrazone]

NullCPUPowerManagement

Try this in clover/kexts/other folder.

Hi there!

I tried with NullCPU but I got the same errors as before (lots of ACPI errors after passing the “End Randomseed” phase and the row of plus signs).

So I had an idea: I unchecked every option available under ACPI in Clover’s GUI, and the installer booted pretty quickly and installed the system in less than 10 minutes.

The installation boots from my USB stick without a hitch. Now I don’t know if NullCPU has helped booting the system, but tomorrow I’ll start installing Clover in my SSD and I’ll have to find out the best way to configure everything, so I’ll figure it out soon.

Thanks to everyone who has helped with this installation, especially to @Gigamaxx.

Time to move on to post-install phase! Thanks again!

 

[Gigamaxx]

Hi there!

I tried with NullCPU but I got the same errors as before (lots of ACPI errors after passing the “End Randomseed” phase and the row of plus signs).

So I had an idea: I unchecked every option available under ACPI in Clover’s GUI, and the installer booted pretty quickly and installed the system in less than 10 minutes.

The installation boots from my USB stick without a hitch. Now I don’t know if NullCPU has helped booting the system, but tomorrow I’ll start installing Clover in my SSD and I’ll have to find out the best way to configure everything, so I’ll figure it out soon.

Thanks to everyone who has helped with this installation, especially to @Gigamaxx.

Time to move on to post-install phase! Thanks again!

Good to hear, when you install the new clover to drive you can use clover configurator app to remove all those ACPI fixes. Most newer systems don’t really need those. You can probably remove the null cpu kext after install. It is needed for some unsupported cpus to prevent kernel panic but I find it can help during installation to bypass checks.

 

[ultrazone]

Good to hear, when you install the new clover to drive you can use clover configurator app to remove all those ACPI fixes. Most newer systems don’t really need those. You can probably remove the null cpu kext after install. It is needed for some unsupported cpus to prevent kernel panic but I find it can help during installation to bypass checks.

Hi again!

Today I've been doing post-install stuff, and I've found the following:

- If I remove the fake CPU id, the system gets stuck at "End randomseed".
- If I remove NullCPUPowerManagement.kext, I get a kernel panic and can't boot.

When I first posted this thread, the problem was install-related, but now it has become a post-install problem. What's more, someone marked this thread as "Solved" in the title, but the problem remains regarding post-install. Should I open another thread in "Post installation"?

Efi folder attached, just in case. Thanks again!

 

[P1LGRIM]

someone marked this thread as "Solved" in the title, but the problem remains regarding post-install.

It is your thread - You can edit the title if you wish.

This is wrong :

There should be no files here with a zip extension.
apfs.efi and ApfsDriverLoader.efi are mutually exclusive - They do the same job - One or the other but not both.

See if this config.plist makes any difference :

 

[ultrazone]

It is your thread - You can edit the title if you wish.

This is wrong :

View attachment 338783

There should be no files here with a zip extension.
apfs.efi and ApfsDriverLoader.efi are mutually exclusive - They do the same job - One or the other but not both.

See if this config.plist makes any difference :

Thanks a lot @P1LGRIM, it really made a difference!

It's the first time that the system passes the "End Randomseed" phase without using a fake CPUid. Now it has been possible to install Intel Power Gadget (before, the installer said it wasn't possible due to processor incompatibility, I believe it was due to the fake CPUid).

Then I tried booting without NullCPUPowerManagement.kext, but no dice: kernel panic after passing "End Randomseed". Several pages appear, and I'm able to read a mention to AppleIntelCPU...

So now using your plist as a base for future changes. Next step would be to get rid of NullCPUPowerManagement, but I'm not sure about how to accomplish it. Do I have to wait for an appropriate sysdef for my CPU? Maybe with VoodooTSCSync?

I've seen that in your plist you specify iMac18,3 as sysdef (the one Clover was choosing by itself). Should I try with different sysdefs?

The CPU is changing its speed during GeekBench's execution (from 0,8Ghz to 4,2Ghz), so not too bad. If it isn't possible to get rid of NullCPUPowerManagement.kext, would it be a good idea to just live with it till we have a proper sysdef for this CPU?

Again, thanks a lot!

 

[P1LGRIM]

Then I tried booting without NullCPUPowerManagement.kext, but no dice: kernel panic after passing "End Randomseed". Several pages appear, and I'm able to read a mention to AppleIntelCPU...

Try this (without NullCPUPowerManagement.kext) :

 

[ultrazone]

Try this (without NullCPUPowerManagement.kext) :

Worked! Thanks!

I have done some benchmarking now that I can boot without fake IDs and without NullCPUPowerManagement. Attached are the results. As you can see, HWMon shows the CPU and GPU throttling correctly (or at least that's what it seems to me, correct me if I'm wrong). In my previous build I remember that it was possible to see each core's frequency, but that doesn't seem to appear here (I don't know why).

What do you think guys? If stable enough, could this build move on to production? Should I wait for a more adequate sysdef?

Thanks a lot again to everybody!

 

[Gigamaxx]

Worked! Thanks!

I have done some benchmarking now that I can boot without fake IDs and without NullCPUPowerManagement. Attached are the results. As you can see, HWMon shows the CPU and GPU throttling correctly (or at least that's what it seems to me, correct me if I'm wrong). In my previous build I remember that it was possible to see each core's frequency, but that doesn't seem to appear here (I don't know why).

What do you think guys? If stable enough, could this build move on to production? Should I wait for a more adequate sysdef?

Thanks a lot again to everybody!

I don’t know if you could get it any better, those scores are great. If you were to run those same tests in windows it would confirm, but I think you have a great setup and windows may actually be a little lower on some tests.

 

[ultrazone]

I don’t know if you could get it any better, those scores are great. If you were to run those same tests in windows it would confirm, but I think you have a great setup and windows may actually be a little lower on some tests.

Hi there!

First of all, thanks for your support and your kind words.

Regarding my build, I'll run the tests in Windows to verify, since I have a spare installation of Windows 10 which will be of help with verifying.

Apart from performance issues, sleep doesn't work (monitors shut down but fans still spin) and I also have some problems with USB (some ports aren't working with USB3 devices and glitches like that). I'm following @RehabMan's guide to generate SSDTs for proper USB injection, and right now I'm finding out which name is of which port.

Maybe these problems are related to some ACPI errors I still have at boot, very similar to the ones mentioned in this post. These ACPI errors mention mainly [PXSX], AE_NOT_FOUND and RP followed by different numbers (I'll investigate with Windows to which devices these numbers are related). In the OP of that thread there is a method for patching DSDT and SSDT's for a very similar board, but I'm not sure if that will fix my errors and what's more, if that will conflict with my actually working config.plist.

So I'll keep you posted, thanks again and cheers from the scorching South of Spain!

 

[stevenkim]

haha.mabye you need this https://www.tonymacx86.com/threads/i7-8700k-gigabyte-aorus-ug-uad-arrow-dtm-build.250603/