Radeon RX 6950 XT

[Edhawk]

Remove the unused Drivers and Tools.

Helps others when trying to assist you.

 

[van7one]

Hello, I just replaced a 6600 with an AMD Reference 6950XT. The 6600 was working perfectly in my 7.1 hack. The 6950XT not so much. B650 Board paired with a 7950x3d.Running Sonoma 14.2 and Opencore 0.9.7. The location info I got from windows is ACPI(_SB_)#ACPI(PCI0)#ACPI(GPP0)#ACPI(SWUS)#ACPI(SWDS)#ACPI(VGA_). I plugged the values in to create my SSDT, but no dice. Any suggestions ?

 

[trs96]

I just replaced a 6600 with an AMD Reference 6950XT. The 6600 was working perfectly in my 7.1 hack. The 6950XT not so much.

AMD cards that end in 50 need a spoofed ID to work. Spoof it to a 6900XT.
RX 6950XT has device-id 0x73A5 and it should be spoofed to 0x73BF.

 

[van7one]

AMD cards that end in 50 need a spoofed ID to work. Spoof it to a 6900XT.
RX 6950XT has device-id 0x73A5 and it should be spoofed to 0x73BF.

View attachment 577105

so the SSDT alone is not enough? Thanks.

 

[etorix]

The SSDT shall apply the spoof… if it targets the right object.
"VGA" is likely the GPU itself (which usually goes by "GFX0" on Intel builds).
Your SSDT defines a further object, "PBR0" at address 0, as a child of VGA, and having itself child GFX1, to which the spoof is applied. I do not know where the PCI address 0 comes from, and the ACPI string from Windows shows no unnamed bridge.

Try the attached version.

 

[Edhawk]

The Device Path for the RX 6950 XT shown in post #73 is from an AMD system, not an intel system.

@van7one where did you get the PBR0 entry in the spoof SSDT? It is not shown in the Device Path you provided.

ACPI(_SB_)#ACPI(PCI0)#ACPI(GPP0)#ACPI(SWUS)#ACPI(SWDS)#ACPI(VGA_).

Becomes this when the fluff is removed.

(_SB/PCI0.GPP0.SWUS.SWDS.VGA_)

There is no PBR0 in the device path.

Neither is there a GFX1 entry.

PBR0 entries highlighted


Using the wrong device path/Device location/IOReg Name is a sure way to have the SSDT fail.

 

[van7one]

The Device Path for the RX 6950 XT shown in post #73 is from an AMD system, not an intel system.

@van7one where did you get the PBR0 entry in the spoof SSDT? It is not shown in the Device Path you provided.

ACPI(_SB_)#ACPI(PCI0)#ACPI(GPP0)#ACPI(SWUS)#ACPI(SWDS)#ACPI(VGA_).

Becomes this when the fluff is removed.

(_SB/PCI0.GPP0.SWUS.SWDS.VGA_)

There is no PBR0 in the device path.

Neither is there a GFX1 entry.

View attachment 577116 PBR0 entries highlighted


Using the wrong device path/Device location/IOReg Name is a sure way to have the SSDT fail.

By watching a video that did not remove PBR0 when they replaced their other values. Foolishly I thought it was required. I did catch that when I tried the SSDT provided by etorix.

 

[Edhawk]

Add an entry to the DeviceProperties for the graphics card, entry named 'model' > String > RX 6950 XT 16GB and that will sort this cosmetic issue.

 

[van7one]

The SSDT shall apply the spoof… if it targets the right object.
"VGA" is likely the GPU itself (which usually goes by "GFX0" on Intel builds).
Your SSDT defines a further object, "PBR0" at address 0, as a child of VGA, and having itself child GFX1, to which the spoof is applied. I do not know where the PCI address 0 comes from, and the ACPI string from Windows shows no unnamed bridge.

Try the attached version.

Thank you, you nailed it.

 

[Edhawk]

You need to be using a recent version of RestrictEvents.kext plus the following boot argument (or NVRAM entries) in your config.plist. This will fix the PCI and Memory tab issues evident in your setup.

revpatch=auto

This boot argument will fix the PCI and Memory tab issues evident in your MacPro7,1 setup. I use the same boot argument in my AMD systems.

This boot argument is taken from the details below.

GitHub - acidanthera/RestrictEvents

Contribute to acidanthera/RestrictEvents development by creating an account on GitHub.

github.com