Contribute
Register

[Success] AMD RX6000 Series working in macOS

Let me know if you're able open it, so we can let Gigamaxx & CaseySJ know.

I can open it with IORegistryExplorere 3.0.1 that I downloaded from Softpedia

Well it looks like the SSDT named the ACPI path, but the spoof isn't loading. The device-id is <"@s">

Can you get the DeviceProperties path using either Hackintool or gfxutil while the SSDT-EGFX0.aml is loaded? It might have changed that path.
 
I can open it with IORegistryExplorere 3.0.1 that I downloaded from Softpedia

Well it looks like the SSDT named the ACPI path, but the spoof isn't loading. The device-id is <"@s">

Can you get the DeviceProperties path using either Hackintool or gfxutil while the SSDT-EGFX0.aml is loaded? It might have changed that path.
You're looking at PEG0 which is my dGPU (5500M). My eGPU is PEG2->UPSB->DSB1->UPS0->pci-bridge->pci-bridge->pci-bridge->display... there you can see my device-id: <af 73 00 00>.
 
Scope (\_SB.PCI0.PEG0.PEG1.PEG2.IOPP.DSB1.IOPP.UPS0)

I will definitely defer to Gigamaxx here, but shouldn't it be

Scope (\_SB.PCI0.PEG0.PEG1.PEG2.UPSB.DSB1.UPS0)
 
Scope (\_SB.PCI0.PEG0.PEG1.PEG2.IOPP.DSB1.IOPP.UPS0)

I will definitely defer to Gigamaxx here, but shouldn't it be

Scope (\_SB.PCI0.PEG0.PEG1.PEG2.UPSB.DSB1.UPS0)
In the aml Gigamaxx told me to try, he had the former, but I thought it was the latter too.
 
In the aml Gigamaxx told me to try, he had the former, but I thought it was the latter too.
Please edit and try new formulas. It’s a really long path, you may also see if you can assign BRG0 to the 5500 to confirm you’re on the right path.
 
@WrathOfThePast

Try this; it's with the edit
Please edit and try new formulas.
@Gigamaxx, @tedyun, I already tried that after we talked about it last night... didn't work.
It’s a really long path, you may also see if you can assign BRG0 to the 5500 to confirm you’re on the right path.
The 5500's device-id is listed for "ATY,DeviceID" on GFX0 while "device-id" is "<"@s">." Bizarre, but that confirms that I'm on the right path.

I found the debug logs for Lilu & WEG and I think I may have found the issue on the WEG log:
"WhateverGreen unfair: @ (DBG) disabling unfair gva due to missing boot argument"
This is bizarre considering that "...weg: @ (DBG) agdpmod using config pikera" appears in the log long before that error occurs. Is there another boot-arg I'm missing for WEG? Note: there's no errors in the Lilu debug log.

Another interesting item: System Information lists that my 6900 XT supports metal when it didn't prior to this endeavor. I attached before & after screenshots.
 

Attachments

  • Screen Shot 2021-10-20 at 3.38.23 PM.png
    Screen Shot 2021-10-20 at 3.38.23 PM.png
    384.6 KB · Views: 79
  • Screen Shot 2021-10-28 at 4.12.21 PM.png
    Screen Shot 2021-10-28 at 4.12.21 PM.png
    438.9 KB · Views: 77
@Gigamaxx, @tedyun, I already tried that after we talked about it last night... didn't work.

The 5500's device-id is listed for "ATY,DeviceID" on GFX0 while "device-id" is "<"@s">." Bizarre, but that confirms that I'm on the right path.

I found the debug logs for Lilu & WEG and I think I may have found the issue on the WEG log:
"WhateverGreen unfair: @ (DBG) disabling unfair gva due to missing boot argument"
This is bizarre considering that "...weg: @ (DBG) agdpmod using config pikera" appears in the log long before that error occurs. Is there another boot-arg I'm missing for WEG? Note: there's no errors in the Lilu debug log.

Another interesting item: System Information lists that my 6900 XT supports metal when it didn't prior to this endeavor. I attached before & after screenshots.
unfairvga is a WEG boot arg. Try that.
 
According to MacRumors, macOS 12.1 beta1 has been released today to developers, and the update enables support for graphics cards that integrate the AMD Radeon 6600XT GPU.


I wonder if 6700 XT support is there as well?

More macOS support for big Navi is a good thing.
 
Back
Top