10.13.6 ACPI "AE Not Found" and BT USB failure

[thecythe]

Brand new build here of the following specs:

Gigabyte H370M-DS3H
EVGA GeForce 1050 Ti 4GB Single Fan
250MB 960 EVO nvme
16GB RAM
GMYLE 4.0 USB BT Adapter
TPLink Archer T9E
High Sierra 10.13.6
Clover 4630 from MultiBeast 10.4

Able to get a nearly perfect boot, less a few problems.

Initially installed 10.13.6 on the machine using another users EFI folder from a similar motherboard. Machine booted fine, and was able to install 10.13.6 no problem. Post installed with Multibeast 10.4 / Clover 4630 no problem. However after taking a closer look it was apparent a lot of fixing / pairing down of config.plist, kexts, SSDT etc was needed to get things right. I've done my best to get things working optimally but still have a few problems.

Using a BCM20702A0 based USB BT (GMYLE 4.0 USB Bluetooth Adapter)
Tried all methods in this thread https://www.tonymacx86.com/threads/bluetooth-not-working-after-update-to-10-13-6.255963/ Still cant get it to work under 10.13.6. When plugged in, system hangs on repeated output of:
AppleUSBHostController : lowerOnePowerStateTo : Session open : XHC > HS06 > BCM20702A0 > AppleUSBHostLegacyClient

Have done my best to properly patch USB via SSDT as per RehabMan's guide

Also noticing some ACPI errors at boot (see attached screenshot). Getting the old "AE NOT FOUND" error, probably USB related.

Problem files are attached

Any help you can provide would be much appreciated, thanks guys!

 

[RehabMan]

Brand new build here of the following specs:

Gigabyte H370M-DS3H
EVGA GeForce 1050 Ti 4GB Single Fan
250MB 960 EVO nvme
16GB RAM
GMYLE 4.0 USB BT Adapter
TPLink Archer T9E
High Sierra 10.13.6
Clover 4630 from MultiBeast 10.4

Able to get a nearly perfect boot, less a few problems.

Initially installed 10.13.6 on the machine using another users EFI folder from a similar motherboard. Machine booted fine, and was able to install 10.13.6 no problem. Post installed with Multibeast 10.4 / Clover 4630 no problem. However after taking a closer look it was apparent a lot of fixing / pairing down of config.plist, kexts, SSDT etc was needed to get things right. I've done my best to get things working optimally but still have a few problems.

Using a BCM20702A0 based USB BT (GMYLE 4.0 USB Bluetooth Adapter)
Tried all methods in this thread https://www.tonymacx86.com/threads/bluetooth-not-working-after-update-to-10-13-6.255963/ Still cant get it to work under 10.13.6. When plugged in, system hangs on repeated output of:
AppleUSBHostController : lowerOnePowerStateTo : Session open : XHC > HS06 > BCM20702A0 > AppleUSBHostLegacyClient

Have done my best to properly patch USB via SSDT as per RehabMan's guide

Also noticing some ACPI errors at boot (see attached screenshot). Getting the old "AE NOT FOUND" error, probably USB related.

Problem files are attached

Any help you can provide would be much appreciated, thanks guys!

USB is definitely not configured correctly. For example, you have no SSxx ports injected on XHC, which means USB3 is not going to work.

See guide:
https://www.tonymacx86.com/threads/guide-creating-a-custom-ssdt-for-usbinjectall-kext.211311/

 

[thecythe]

I figured something was strange about those SSxx ports not being injected.
Thing is, when discovering ports as per your guide, I'm not seeing those SSxx ports on the left hand side in IOreg like in your example.
Therefore, I'm not sure what SSxx ports I should keep or which ones I should lose?
Attached a screenshot of what I'm seeing with all ports injected with no patched SSDT.

Should I just go ahead and inject all 10 SSxx ports then?

Also, apparently XHCI handoff was disabled in my bios. Once enabled, my USB BT is now working as it should.

 

[RehabMan]

I figured something was strange about those SSxx ports not being injected.
Thing is, when discovering ports as per your guide, I'm not seeing those SSxx ports on the left hand side in IOreg like in your example.
Therefore, I'm not sure what SSxx ports I should keep or which ones I should lose?
Attached a screenshot of what I'm seeing with all ports injected with no patched SSDT.

You forgot to use the (correct) port limit patch.

 

[thecythe]

Thanks Rehab!
Things have changed quite a bit since my first hack build (45 series chipset first on Chameleon then on early Clover versions), hence my confusion with the way things are working now with these fancy new systems.
Had no actual USB3 devices on hand to properly test the ports hence my mistakes in my SSDT, doh.

Managed to fix all that now. Have USB injected properly, BT running perfectly from the USB dongle, sleep and power management working great. All "needed" kexts are also properly installed now to /L/E.

Still however getting the boot errors relevant to ACPI - PXSX. Running two Samsung 960 EVO Nvme drives here. Could this error be relevant to this post of yours from 2016:

https://www.tonymacx86.com/threads/...h-ionvmefamily-using-class-code-spoof.210316/

If so is there an easy fix for this in Clover (now running 4658), or do I need to generate a SSDT for my Nvme drives?

Problem files and screenshot attached...

 

[RehabMan]

Thanks Rehab!
Things have changed quite a bit since my first hack build (45 series chipset first on Chameleon then on early Clover versions), hence my confusion with the way things are working now with these fancy new systems.
Had no actual USB3 devices on hand to properly test the ports hence my mistakes in my SSDT, doh.

Managed to fix all that now. Have USB injected properly, BT running perfectly from the USB dongle, sleep and power management working great. All "needed" kexts are also properly installed now to /L/E.

Still however getting the boot errors relevant to ACPI - PXSX. Running two Samsung 960 EVO Nvme drives here. Could this error be relevant to this post of yours from 2016:

https://www.tonymacx86.com/threads/...h-ionvmefamily-using-class-code-spoof.210316/

If so is there an easy fix for this in Clover (now running 4658), or do I need to generate a SSDT for my Nvme drives?

Problem files and screenshot attached...

Your ACPI errors are caused by invalid use of module level CondRefOf tests.
Such as:

    If (CondRefOf (\_SB.PCI0.RP01.PXSX))
    {
        Scope (_SB.PCI0.RP01.PXSX)
        {
...
            Method (WIST, 0, Serialized)
        }
    }
...

I think that construct works in Windows ACPI runtime, but it seems not the official Intel ACPI runtime (which Apple uses).

You could get rid of the CondRefOf tests/blocks:

        Scope (_SB.PCI0.RP01.PXSX)
        {
...
            Method (WIST, 0, Serialized)
        }
...

 

[thecythe]

Thanks again Rehab, wow ok that's an interesting one.

Should I attempt to delete all references of If (CondRefOf (\_SB.PCI0.RPxx.PXSX))
or just the ones pertaining to Method (WIST, 0, Serialized) ?

 

[DavidTan]

I have IOGear GBU521(BCM20702A0) , Simple solution in Mojave 10.14.5 :1. Disable Intel Bluetooth(If you have) ;2.Remove BrcmFirmwareRepo.kext and BrcmPatchRAM2.kext from S/L/E and EFI/Clover/kexts/Other ;3.Use Kext Utility to reset the cache/permissions ; -- Perfect working for me .