Jump to content

Help installing Mojave on Xeon W-2175 and Asus WS C422 mobo

obus

By obus,
in OSx86 10.14 (Mojave)

 Share
852 posts in this topic

Recommended Posts

yapan4
yapan4
Posted
Posted (edited)
On 4/15/2019 at 11:46 AM, skyflying5 said:

everything is perfect in my windows 10

 

I'm not sure about BIOS setting  whether is satisfied for the os or not

because my motherboard is not the same as Asus WS C422

I just can make it as same setting as possible

 

So，as you all see it cannot change the frequence at all

You can try kill AppleIntelCPUPowerManagement.kext in /S/L/E  -  just rename it to .kext.bak

Edited by yapan4
Link to comment
Share on other sites
apianti
apianti
Posted
Posted (edited)

No, don't do that with the AppleIntelCPUPowerManagement.kext. Basically you are all using the same solutions but incorrectly. I have given you the two and only cases that will give you working power management. Whoever posted that pick with the W-2150b, you don't need all those patches is the problem. And you need an SSDT.

 

EDIT: Make sure that xcpm is enabled and that you have frequency vector count of right amount.

Edited by apianti
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted (edited)
8 hours ago, apianti said:

No, don't do that with the AppleIntelCPUPowerManagement.kext. Basically you are all using the same solutions but incorrectly. I have given you the two and only cases that will give you working power management. Whoever posted that pick with the W-2150b, you don't need all those patches is the problem. And you need an SSDT.

 

EDIT: Make sure that xcpm is enabled and that you have frequency vector count of right amount.

I use this solutions temporary to avoid CPU overhead. Is AppleIntelCPUPowerManagement.kext still necessary for xcpm?

OK,  will be try with AppleIntelCPUPoxxxx.kext enabled.

 

Update

Now:

AppleIntelCPUPowerManagement.kext enabled

SSDTs.aml generated by ssdtPRGen.sh(with some Warnings) - in /Clover/ACPI/patched

freqVectorsEdit.sh - passed

X86PlatformPlugin.kext and X86PlatformShim.kext - loaded 

 

no success, still full trottling, sysctl -n machdep.xcpm.mode 0, many errors in Kernel Message Log(attached)

 

Kernel Message Log.rtf

Edited by yapan4
Update
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted
11 hours ago, apianti said:

EDIT: Make sure that xcpm is enabled and that you have frequency vector count of right amount.

I'm not sure, but xcpm=1 available in ssdt.aml.

How to check frequency vectors - lack of knowledge

Link to comment
Share on other sites
Guest
Guest
Posted
Posted (edited)
27 minutes ago, yapan4 said:

I'm not sure, but xcpm=1 available in ssdt.aml.

How to check frequency vectors - lack of knowledge

 

open terminal windows and type this

sysctl -a | grep xcpm

 

you will see many useful things...

check for vector count value for your question..if 1 it is OK, if 0 no

 

Edited by Guest
Link to comment
Share on other sites
Guest
Guest
Posted
Posted

Hi also XCPM=0 so it is an half disaster :)

 

If I can say I will try to start with msr0xe2 disabled (Bios section related)

then a minimal patching as we did some time ago with @obus Imean boot strap patch

then I will try different Smbios and when all of this fails I will try to inject FV with Pike's script

I think it is possible to have full step for your cpu in some way ..but only you can try ;)

 

Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted (edited)
28 minutes ago, fabiosun said:

Hi also XCPM=0 so it is an half disaster :)

 

If I can say I will try to start with msr0xe2 disabled (Bios section related)

then a minimal patching as we did some time ago with @obus Imean boot strap patch

then I will try different Smbios and when all of this fails I will try to inject FV with Pike's script

I think it is possible to have full step for your cpu in some way ..but only you can try ;)

 

If you mean a patch for BIOS, then the answer is NO. Sorry but any interference with the hardware of the motherboard, gold rule:)

Software only.

Edited by yapan4
Link to comment
Share on other sites
Guest
Guest
Posted
Posted (edited)
5 minutes ago, yapan4 said:

If you mean a patch for BIOS, then the answer is NO. Sorry but any interference with the hardware of the motherboard, gold rule:)

No patch..it is an option you should have if you are using latest Asus Bios

 

180815204713.thumb.jpg.c5d3ec73aba027262bc6a416fbdcd0ae.jpg

Edited by Guest
Link to comment
Share on other sites
Guest
Guest
Posted
Posted (edited)
6 minutes ago, yapan4 said:

What kind?

Look at my previous edited by now message

start from there try boot strap patch (try always without firstly) only MSR 0xe2 is useful if cpu is seen by OSX as we hope

 

Edited by Guest
Link to comment
Share on other sites
Guest
Guest
Posted
Posted (edited)

maybe you patch msr0xe2 in other way?

 

I have an ASUS Sage x299 10G WS server mb..and I have to disable it (unlock from bios option to boot)

I am using no kernel PM or other kernel patches in my config

I will start from there..

Edited by Guest
Link to comment
Share on other sites
Guest
Guest
Posted
Posted

Ciao yapan4

I will try to explain better

I will test things in a way I expose here:

1) use less patch as it possible 

2) if your system is compatible with OSX try to unlock MSR0xe2 on bios settings and see where your system hangs

3) if it hangs try only boot strap and see

4) full throttling because for now no XCPM working in your system, so try to inject Frequency vector in your SMBios and check if value become 1

5) if not try a fakecpuid and see

 

in the past I have had a 2696 V4 xeon on a X99 system and I have used OSX using boot strap and fakecpuid patch happily with zero problem...

your system should work more vanilla..but if not possible using a fakecpuid and some patches for XCPM should work well ;)

 

Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted (edited)

That's great!

 

But while we were talking, my processor was overheated and I was forced to return to the first state (with the deactivated AppleIntelCPUPowerManagement.kext) until the correct solution for cpupm was found.

Edited by yapan4
Link to comment
Share on other sites
apianti
apianti
Posted
Posted

Leave your MSRs unlocked. Also, you are correct clover automatically patches when MSR 0xE2 is locked, so this test will do nothing. There is a problem with XCPM being enabled. What are the error messages you are getting in the kernel log?

Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted
15 hours ago, fabiosun said:

Ciao yapan4

I will try to explain better

I will test things in a way I expose here:

1) use less patch as it possible 

2) if your system is compatible with OSX try to unlock MSR0xe2 on bios settings and see where your system hangs

3) if it hangs try only boot strap and see

4) full throttling because for now no XCPM working in your system, so try to inject Frequency vector in your SMBios and check if value become 1

5) if not try a fakecpuid and see

 

in the past I have had a 2696 V4 xeon on a X99 system and I have used OSX using boot strap and fakecpuid patch happily with zero problem...

your system should work more vanilla..but if not possible using a fakecpuid and some patches for XCPM should work well ;)

 

I will continue the tests and report the results

Link to comment
Share on other sites
apianti
apianti
Posted
Posted

Are these the only messages? Those messages mean your SSDT is incorrect. If I had to guess, you should decompile the SSDT, and you will have CP00 through CP07, you should change CP04 through CP07 to CP08 to CP0B.

  • Like 1
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted (edited)
On 4/16/2019 at 8:24 PM, fabiosun said:

1) use less patch as it possible 

2) if your system is compatible with OSX try to unlock MSR0xe2 on bios settings and see where your system hangs

3) if it hangs try only boot strap and see

4) full throttling because for now no XCPM working in your system, so try to inject Frequency vector in your SMBios and check if value become 1

5) if not try a fakecpuid and see

Clover v.4920 did not allow me to turn off KernelPM patch. I tried to do it in Clover's GUI before macOS boot and in in config.plist. The result is always the same:

MSR 0xE2 is locked, PM patches will be turned on
MSR 0xCE              00070C2C_F3812400
MSR 0x1B0             00000000

 

Tested with 

xcpm_bootstrap (c) Pike R. Alpha

Find 89D804C4 3C22 

Replace 89D80455 3C22

and "Configure lock - Enabled" in BIOS (MSR 0xE2 is locked)

 

The system starts up normally. Later I'll try the same with FakeCPUID + _xcpm_pkg_scope_msrs_PMhart

 

Logfiles and efi folder attached

Log Files-MSR 0xE2 locked.zip

Edited by yapan4
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted
14 minutes ago, fabiosun said:

@yapan4

you have misunderstood msr0xe2 part

in bios you have to set it ti disable 

so you can pass OS X hang 

 

We will return to the beginning again:)

 

Ok, now i will continue with SSDT and XCPM...

Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted (edited)
On 4/17/2019 at 3:23 PM, apianti said:

Are these the only messages? Those messages mean your SSDT is incorrect. If I had to guess, you should decompile the SSDT, and you will have CP00 through CP07, you should change CP04 through CP07 to CP08 to CP0B.

A little progress, but still not successful.

What I did:

1) in ssdtPRGen.sh run command

sudo ./ssdtPRGen.sh -p W-2123 -b Mac-7BA5B2D9E42DDD94 -target 4 -m iMacPro1,1 -mode normal -a CPU0 -x 1  | tee -a file.txt

Passed with 2 warning and 1 error, created files ssdt.aml and ssdt.dsl

then

2) File ssdt.dsl opened in MaciASL.app, all PR04 - PR07 changed to PR08 - PR0B, compiled w/o errors, saved as .aml on /Desktop and copied to /ACPI/patched

then

3) Reboot

then

4) run freqVectorsEdit.sh , passed w/o errors

then

5) Reboot

 

...still full trotlling and machdep.xcpm.mode: 0

 

Useful files and log-files -- in attach

Files.zip

 

update:added bootlog

bootlog.txt.zip

Edited by yapan4
Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...