Ryzen 7000 Testing

[ExtremeXT]

@CaseySJ The 2 byte patch doesn't work for an MSI user, but the 4 byte one which you tested works fine, I guess we'll merge that one. Waiting for one more MSI tester.

EDIT: Another user got it working... You should try it too.

 

[CaseySJ]

@CaseySJ The 2 byte patch doesn't work for an MSI user, but the 4 byte one which you tested works fine, I guess we'll merge that one. Waiting for one more MSI tester.

EDIT: Another user got it working... You should try it too.

Sounds good. Just got back from lunch.

 

[CaseySJ]

For anyone wondering:

• Method 3 kernel patch allows 8-bit extended tag fields (8b ETFs) to be enabled, and only disables 10-bit ETFs (10b).
• Boot Argument pci=0x8000000 disables both 8-bit and 10-bit ETFs.

 

[CaseySJ]

@CaseySJ The 2 byte patch doesn't work for an MSI user, but the 4 byte one which you tested works fine, I guess we'll merge that one. Waiting for one more MSI tester.

EDIT: Another user got it working... You should try it too.

It works!!

 

[Shaneee]

Feel free to make pull request on the AMD_Vanilla GitHub with any final versions of patches/boot args used and I'll merge it. Also speak to @alyxferrari about adding it to the FAQ also if required.

Again awesome work guys

 

[CaseySJ]

Note to Self:

• Thunderbolt implementation on this AM5 board is quirky
• If Thunderbolt is disabled in BIOS, remember to disable the associated Thunderbolt SSDT, otherwise WiFi or any card installed in spare PCI slot will not work
• If Thunderbolt is enabled in BIOS, remember to enable the associated Thunderbolt SSDT, otherwise Ventura will not boot
• Sleep, Wake, Reboot, and Shutdown work when Thunderbolt is disabled in BIOS, but do not currently work when Thunderbolt is enabled (I'll look into this later)

 

[PoMpIs]

@CaseySJ The 2 byte patch doesn't work for an MSI user, but the 4 byte one which you tested works fine, I guess we'll merge that one. Waiting for one more MSI tester.

EDIT: Another user got it working... You should try it too.

So this is the minimalist patch that works?

 

[Ejh4286]

I have used the patch and the boot arg to boot Ventura. I installed Ventura on an M.2 NVME drive using an intel iMac, then copying my EFI on to the EFI partition. My Asus Strip X670E-F with the Rysen 7950 boots into Ventura, I log in, then after about two minutes, it freezes. I repeated the process using a 2.5 inch SATA SSD and it is rock stable. I an unsure if this is related to the PCI issue or not. I was running Big Sur on a M.2 NVME drive and it was also very stable. Any advice out there?

 

[CaseySJ]

So this is the minimalist patch that works?

View attachment 8271

Not exactly... The actual minimal Method 3 patch has been added to the bottom of this post:

Ryzen 7000 Testing

WOW!

forum.amd-osx.com

 

[CaseySJ]

I have used the patch and the boot arg to boot Ventura. I installed Ventura on an M.2 NVME drive using an intel iMac, then copying my EFI on to the EFI partition. My Asus Strip X670E-F with the Rysen 7950 boots into Ventura, I log in, then after about two minutes, it freezes. I repeated the process using a 2.5 inch SATA SSD and it is rock stable. I an unsure if this is related to the PCI issue or not. I was running Big Sur on a M.2 NVME drive and it was also very stable. Any advice out there?

Please note that we should use either the boot argument or the kernel patches, but not both. The latest kernel patch is Method 3 (see my post directly above).

 

[Galve2000]

If you’re using Monterey then yes, just try adding the boot argument e1000=0 and remove or disable all device properties for i225-V.

I'm using Monterey. Do I need OC 0.85?

 

[CaseySJ]

I'm using Monterey. Do I need OC 0.85?

Not necessary because the boot argument is independent of OpenCore.

 

[dmls]

Here's an MSI motherboard running Ventura from a tester, confirmed that it's working here too.

I'm glad to see the exciting work. Does it mean that the problem on MSI's motherboard on am4 has been solved?

 

[DarkSilentSC]

Tried parallel approach of adding device-id to DeviceProperties with my Aquantia but still gives me KP. uwu

 

[ExtremeXT]

@CaseySJ Should I send the PR with the 4 byte patch to the repo?

 

[mariettosun]

@ExtremeXT why a patches is advisable against the use of a bootarg 1revenger found in the kext code?
I mean is not more solid a bootarg if kext you patch change with an os upgrade?
Only to understand well
Thank you

 

[ExtremeXT]

@ExtremeXT why a patches is advisable against the use of a bootarg 1revenger found in the kext code?
I mean is not more solid a bootarg if kext you patch change with an os upgrade?
Only to understand well
Thank you

New kext patch should work for newer macOS versions. It's preferred because Zen 4 and MSI only have problems with the 10 bit tags, but the boot arg also disables the 8 bit tags which don't have to be disabled.

 

[mariettosun]

In the faq and also in amd vanilla GitHub repo should be left a track also of that boot arg in my opinion

 

[CaseySJ]

@CaseySJ Should I send the PR with the 4 byte patch to the repo?

Yes please do.

 

[CaseySJ]

@ExtremeXT why a patches is advisable against the use of a bootarg 1revenger found in the kext code?
I mean is not more solid a bootarg if kext you patch change with an os upgrade?
Only to understand well
Thank you

Are you referring to boot argument pci=0x8000000? We discovered it here:

Ryzen 7000 Testing

Yes, I would be grateful if you could send the EFI. I have a 7950x with an Asus Strix X670E-F. Why not make your own? If you encounter any problems, we will help you here.

forum.amd-osx.com