Z490 & Z590 - Will Z590 ever have macOS Support ?

[nhojb]

I am trying to diagnose and fix a kernel panic when sleeping my Gigabyte Z590i Vision D / i9-11900 Hackintosh.

The behaviour is as follows:
1. Everything works fine after booting.
2. The 1st time sleep works perfectly.
3. After waking everything is also fine.
4. The 2nd time sleep causes a kernel panic just before the S3 power down. Pressing the power button (or keyboard) causes the PC to restart.

My two clues are 1) "Sleep Wake Failure" log (see below) and 2) even after a "clean" shutdown and boot (no sleep) I see the "You shutdown your computer because of a problem" warning. So something's not quite right.

I've been through the Dortina "Fixing Sleep" guide multiple times and AFAIK everything is as it should be, including CPU PM. I've also tried various other things including disabling Thunderbolt and RC6 standby (iGPU is disabled however). I'm at a bit of a loss a) how to further diagnose this and b) how to fix it. Any help gratefully accepted

My EFI & boot log are attached.

Sleep Wake failure in EFI

Failure code:: 0x00000000 0x0000001f

Please IGNORE the below stackshot

================================================================
Date/Time: 2022-01-03 10:21:48.888 +0000
OS Version: ??? ??? (Build ???)
Architecture: x86_64
Report Version: 35
Incident Identifier: 239E3619-399B-4729-88B5-D418B9FE8709

Data Source: Stackshots
Shared Cache: 40432A03-88D3-305F-9C0C-E7549E71D927 slid base address 0x7ff813eb0000, slide 0x13eb0000

Event: Sleep Wake Failure
Duration: 0.00s
Steps: 1

Boot args: agdpmod=pikera -v keepsyms=1 swd_panic=1 chunklist-security-epoch=0 -chunklist-no-rev2-dev

Time Awake Since Boot: 12s



Process: swd [336]
Architecture: x86_64
Footprint: 476 KB
Time Since Fork: 3s
Num samples: 1 (1)

Thread 0x7ba 1 sample (1) priority 4 (base 4)
<thread QoS background (requested background), thread darwinbg, process darwinbg, IO tier 2>
1 ??? [0x112d984fe] 1
1 ??? [0x108fa1454] 1
1 ??? [0x108fa11e8] 1
1 __stack_snapshot_with_config + 10 (libsystem_kernel.dylib + 149174) [0x7ff8141b16b6] 1
*1 ??? [0xffffff8000228226] 1
*1 ??? [0xffffff800099e9d4] 1
*1 ??? [0xffffff8000894e84] 1
*1 ??? [0xffffff800024be78] 1
*1 ??? [0xffffff80002881ad] (running) 1

Binary Images:
0x7ff81418d000 - 0x7ff8141c3fff libsystem_kernel.dylib (8019.61.5) <5AA1E5BE-B5B8-3A02-9885-A8C99E0CA378> /usr/lib/system/libsystem_kernel.dylib

System Profile:
Network Service: Ethernet, Ethernet, en0
Thunderbolt Bus:
Boot Volume File System Type: apfs
PCI Card: Intel USB Controller, USB Controller, Built In
Memory Module: BANK 0/Controller0-ChannelA-DIMM0, 16 GB, DDR4, 3600 MHz, Essencore Limited, KD4AGU880-36A180U
Memory Module: BANK 1/Controller0-ChannelB-DIMM0, 16 GB, DDR4, 3600 MHz, Essencore Limited, KD4AGU880-36A180U
USB Device: USB31Bus
USB Device: USB32Bus
USB Device: bluetooth_device
USB Device: USB Audio
USB Device: USB2.0 Hub
USB Device: hub_device
USB Device: ALT Keyboard
Display: LG Ultra HD, 4608 x 2592, Main, MirrorOff, Online
Model: iMacPro1,1, BootROM 1715.40.15.0.0, 8 processors, 8-Core Intel Core i9, 2.5 GHz, 32 GB, SMC
Graphics: Radeon RX 580, Radeon RX 580, PCIe, 8 GB
AirPort: spairport_wireless_card_type_wifi (0x8086, 0x84), itlwm: 2.0.0 fw: 48.1335886879.0
Bluetooth: Version (null), 0 services, 0 devices, 0 incoming serial ports

 

[UtterDisbelief]

I am trying to diagnose and fix a kernel panic when sleeping my Gigabyte Z590i Vision D / i9-11900 Hackintosh.

The behaviour is as follows:
1. Everything works fine after booting.
2. The 1st time sleep works perfectly.
3. After waking everything is also fine.
4. The 2nd time sleep causes a kernel panic just before the S3 power down. Pressing the power button (or keyboard) causes the PC to restart.

My two clues are 1) "Sleep Wake Failure" log (see below) and 2) even after a "clean" shutdown and boot (no sleep) I see the "You shutdown your computer because of a problem" warning. So something's not quite right.

I've been through the Dortina "Fixing Sleep" guide multiple times and AFAIK everything is as it should be, including CPU PM. I've also tried various other things including disabling Thunderbolt and RC6 standby (iGPU is disabled however). I'm at a bit of a loss a) how to further diagnose this and b) how to fix it. Any help gratefully accepted

My EFI & boot log are attached.

Hi there.

Sorry my eyesight is not up to the task of reading the full bootlog...

What version macOS are you using ...?

 

[SchmockLord19]

@dehjomz
Gigabyte Z590i Vision D. I got hotplug working.

Attached is the SSDT. Thanks to @CaseySJ @Elias64Fr.

It is the SSDT-TB3HP-JHL8540-ASUS-Z590i-2.aml for the Asus Z590i Gaming Wifi changed to RP05 for the Vision D.

BIOS Settings:

BIOS version F7

• Above 4G Decoding to Enabled.
• CSM Support to Disabled.
• Legacy USB to Disabled.
• VT-D to Disabled.

Thunderbolt

• GPIO3 Force Pwr to Enabled.
• Native OS Security for TBT to Disabled.
• GPIO filter to Enabled.

And the attached Device properties. I am not sure if you really need them, haven't tried it out without them yet.

 

[CaseySJ]

I am trying to diagnose and fix a kernel panic when sleeping my Gigabyte Z590i Vision D / i9-11900 Hackintosh.

The behaviour is as follows:
1. Everything works fine after booting.
2. The 1st time sleep works perfectly.
3. After waking everything is also fine.
4. The 2nd time sleep causes a kernel panic just before the S3 power down. Pressing the power button (or keyboard) causes the PC to restart.

My two clues are 1) "Sleep Wake Failure" log (see below) and 2) even after a "clean" shutdown and boot (no sleep) I see the "You shutdown your computer because of a problem" warning. So something's not quite right.

I've been through the Dortina "Fixing Sleep" guide multiple times and AFAIK everything is as it should be, including CPU PM. I've also tried various other things including disabling Thunderbolt and RC6 standby (iGPU is disabled however). I'm at a bit of a loss a) how to further diagnose this and b) how to fix it. Any help gratefully accepted

My EFI & boot log are attached.

Based on your boot log, this should be fixable with the GSWApp ACPI patch. Details here:

[SUCCESS] Gigabyte Z590 VISION D + 11700K + RX 6600 XT

I'm stuck, following @CaseySJ description of how to drop DMAR tables from his post in Z490 build: https://www.tonymacx86.com/threads/gigabyte-z490-vision-d-thunderbolt-3-i5-10400-amd-rx-580.298642/post-2278034 I've Uncheck ACPI --> Add --> SSDT-DMAR.aml in config.plist Uncheck ACPI --> Delete...

www.tonymacx86.com

 

[UtterDisbelief]

Based on your boot log, this should be fixable with the GSWApp ACPI patch. Details here:

[SUCCESS] Gigabyte Z590 VISION D + 11700K + RX 6600 XT

I'm stuck, following @CaseySJ description of how to drop DMAR tables from his post in Z490 build: https://www.tonymacx86.com/threads/gigabyte-z490-vision-d-thunderbolt-3-i5-10400-amd-rx-580.298642/post-2278034 I've Uncheck ACPI --> Add --> SSDT-DMAR.aml in config.plist Uncheck ACPI --> Delete...

www.tonymacx86.com

Hi @CaseySJ

The reason I asked for the macOS version (as I could not see it in the post...) is because when using Monterey with the Intel wireless chip, Bluetooth will not work and the IntelbluetoothFirmware kext pair will either prevent booting or cause a crash - in my experience. BlueToolFixup has only worked for me with USB BT dongles.

YMMV etc. (or your eyesight better!) so let me know if your experience is different.

 

[SchmockLord19]

So, in the meantime I have upgraded Monterey. TB hotplug still works. So far so good.

Now I have the issue, that my W5500 is not used as main display anymore.

I have a 3090 which is inside my case in the PCIE and a W5500 which is in the eGPU Case Razer Core X.

Both are getting detected. But how can I disable the 3090 and tell MacOS to use the W5500?

Before hotplug was working, I just disabled the 3090 using the attached Device Properties and then macOS initialized the W5500 via Thunderbolt instead. But somehow they don't work anymore. When I use them the W5500 Display stays black. And without them, macOS is going into the VESA mode and displays via my 3090 in the low graphics mode.

I guess because the W5500 is detected as GFX1 instead of GFX0.

I think I need something like an SSDT which disabled the 3090 as my GFX0 and injects the W5500 as GFX0 instead.

I have already started the SSDT, but this only disables the 3090. Maybe someone can help me from there or has a better idea.

Attached are my PCIE devices. And these are the location paths from Windows:

3090:

\_SB.PC00.PEG1.PEGP


PCIROOT(0)#PCI(0100)#PCI(0000)
ACPI(_SB_)#ACPI(PC00)#ACPI(PEG1)#ACPI(PEGP)


W5500:
PCIROOT(0)#PCI(1C04)#PCI(0000)#PCI(0300)#PCI(0000)#PCI(0100)#PCI(0000)#PCI(0000)#PCI(0000)
ACPI(_SB_)#ACPI(PC00)#ACPI(RP05)#ACPI(PXSX)#PCI(0300)#PCI(0000)#PCI(0100)#PCI(0000)#PCI(0000)#PCI(0000)
ACPI(_SB_)#ACPI(PC00)#ACPI(RP05)#ACPI(HRUS)#PCI(0300)#PCI(0000)#PCI(0100)#PCI(0000)#PCI(0000)#PCI(0000)

IOService:/AppleACPIPlatformExpert/PC00@0/AppleACPIPCI/RP05@1C,4/IOPP/UPSB@0/IOPP/DSB3@3/IOPP/pci-bridge@0/IOPP/pci-bridge@1/IOPP/pci-bridge@0/IOPP/pci-bridge@0/IOPP/display@0

@CaseySJ Can you help me on this?

 

[CaseySJ]

@CaseySJ Can you help me on this?

Because the W5500 is an eGPU, we cannot modify its PCI device path. This was attempted unsuccessfully just a few weeks ago, so we know this is not viable.

The device properties for disabling your Nvidia 3090 are most likely still working. This can be checked by running IORegistryExplorer and looking at the device tree for PC00.PEG1. Feel free to post a screenshot of that section.

Suggestions:

• Try enabling CSM Support in BIOS.
• Try disconnecting and reconnecting the HDMI/DisplayPort cable from the W5500 after you see the black screen.
• Try power cycling the monitor after you see the black screen.

 

[dehjomz]

@dehjomz
Gigabyte Z590i Vision D. I got hotplug working.

Attached is the SSDT. Thanks to @CaseySJ @Elias64Fr.

It is the SSDT-TB3HP-JHL8540-ASUS-Z590i-2.aml for the Asus Z590i Gaming Wifi changed to RP05 for the Vision D.

BIOS Settings:

BIOS version F7

• Above 4G Decoding to Enabled.
• CSM Support to Disabled.
• Legacy USB to Disabled.
• VT-D to Disabled.

Thunderbolt

• GPIO3 Force Pwr to Enabled.
• Native OS Security for TBT to Disabled.
• GPIO filter to Enabled.

And the attached Device properties. I am not sure if you really need them, haven't tried it out without them yet.

So BIOS F7 fixed hotplug in macOS…?

 

[nhojb]

Based on your boot log, this should be fixable with the GSWApp ACPI patch

@CaseySJ thanks taking the time to look into this and for the tip. I'll check this out and report back.

The reason I asked for the macOS version (as I could not see it in the post...) is because when using Monterey with the Intel wireless chip, Bluetooth will not work and the IntelbluetoothFirmware kext pair will either prevent booting or cause a crash - in my experience. BlueToolFixup has only worked for me with USB BT dongles.

@UtterDisbelief sorry, I should have mentioned the macOS version: I'm on Monterey 12.1. FWIW I've been using the built-in Intel Bluetooth (with IntelBluetoothFirmware.kext and BlueToolFixup.kext) without any obvious problems.

 

[SchmockLord19]

Because the W5500 is an eGPU, we cannot modify its PCI device path. This was attempted unsuccessfully just a few weeks ago, so we know this is not viable.

The device properties for disabling your Nvidia 3090 are most likely still working. This can be checked by running IORegistryExplorer and looking at the device tree for PC00.PEG1. Feel free to post a screenshot of that section.

Suggestions:

• Try enabling CSM Support in BIOS.
• Try disconnecting and reconnecting the HDMI/DisplayPort cable from the W5500 after you see the black screen.
• Try power cycling the monitor after you see the black screen.

Thanks. I will try the first point you mention. Step 2/3 I have already tried with no success.

But maybe this helps you. I think there is only little left to solve this.

As soon as I remove this section from the SSDT, the W5500 is detected properly again, but this breaks hot plug again. No BIOS changes. Just removing this section and applying the DeviceProperties for the 3090 to be disabled:

Scope (PXSX)
{
Name (_STA, Zero) // _STA: Status
}

And then the W5500 is connected as in the attachments.