AMD Ryzen R9 5900HX+6600m Install Monterey kernel Panic

[daliansky]

hey, @ExtremeXT, @Visual, I just got a minisforum hx90g, when I install monterey 12.6, I get a kernel panic message, I tried a few patches made by @Visual, but the kernel panic message will Recurrence; if it is convenient, can you provide directions or ideas to solve the problem; thank you very much.

 

[Middleman]

huh, that's strange

No not strange, it’s just that the config.plist has been cleaned up more extensively so works better now for RX6600M setups.

@daliansky - I would really be interested to find out what the Geekbench 5 scores for this is.

 

[Jo-Toku]

It’s the way it was implemented. From my experience some AMD systems needed a separate SSDT-EC and SSDT-USBX, and also added a SSDT-PLUG, SSDT-SBUS-MCHC & SSDT-XOSI. Also GenericUSBXHCI was not enabled and we used @daliansky’s existing USBPorts map.
We added those plus some new kernel patches from @Visual and removed the device properties tree completely. FeatureUnlock kext was also added.

To be honest I am very surprised how it all turned out. I was half expecting to KP but it didn’t. This means the Minisforum HX90g is a worthy contender as a Mac Mini alternative with upgrade potential. The GPU performance alone should beat an M1 Mac. Kudos to everyone for this!

The 6600m does beat the m1 base. Compared to the top end macbook pro or base studio on Geekbench/ Metal the 32core is 66k while the 6600m on mac is 61k

 

[daliansky]

并不奇怪，只是 config.plist 已被更广泛地清理，因此现在更适合 RX6600M 设置。

@daliansky - 我真的很想知道 Geekbench 5 的得分是多少。

今天可能更新不了，明天附上geek bench 5的测试结果今天可能更新不了，明天附上geek bench 5的测试结果

No not strange, it’s just that the config.plist has been cleaned up more extensively so works better now for RX6600M setups.

@daliansky - I would really be interested to find out what the Geekbench 5 scores for this is.

May not be able to update today, tomorrow I will attach the test results of geek bench 5

 

[Jo-Toku]

May not be able to update today, tomorrow I will attach the test results of geek bench 5

I posted my Geekbench of the 6600m a few months ago. Metal was 61k. It will be interesting what score yours get

 

[ExtremeXT]

@daliansky Glad you got it working! Enjoy your NUC!

 

[WalidB]

Not a laptop, it's a mini PC just like Intel NUC, using 5900HX+6600M. According to the screenshots, it has completely the same USB comtrollers as Legion 5…… That's confusing

Hi I bought the same and should get it today. did you manage installing Monterey on it?

 

[Middleman]

Hi I bought the same and should get it today. did you manage installing Monterey on it?

It is not the same system. Yours is a HX90 with only the 5900HX, this is a HX90g with 5900HX PLUS RX6600M.

 

[daliansky]

Geek Bench5 Scores

 

[Jo-Toku]

Geek Bench5 ScoresView attachment 7696View attachment 7697View attachment 7698

Excellent score!

It seem Monterey fixed the problem laptop opus have. The dedicated GPU performance is at hand performance. The 6600m seem to be running at full potential. I am curious if my 5700m would run at full potential under Monterey. Currently running half per under Big Sur. RadeonBoost don't seem to work

 

[alexjaixd]

Does anyone know if the 6900hx version of HX90G, the HX99G will also work the same way as 5900hx version? Would be pretty interested in getting the 6900hx version if it also works the same way. Both of them also uses the 6600M.

 

[daliansky]

Does anyone know if the 6900hx version of HX90G, the HX99G will also work the same way as 5900hx version? Would be pretty interested in getting the 6900hx version if it also works the same way. Both of them also uses the 6600M.

Yes, I'll get a prototype of the HX99G shortly, and I'll also try installing ventura, and if it goes well, I'll update the usage on this thread.

 

[daliansky]

hey,@ExtremeXT,
The brand new HX99G has been obtained, and dmesg and DSDT are attached for reference.

 

[ExtremeXT]

hey,@ExtremeXT,
The brand new HX99G has been obtained, and dmesg and DSDT are attached for reference.

Did macOS boot on it just fine without problems?

 

[daliansky]

它在OP中，报告.zip。

Did macOS boot on it just fine without problems?

Re-made ssdt-usb-reset, occasionally can enter the installation interface; but can not enter the installed ventura system, it feels like the graphics card is not properly driven.

 

[ExtremeXT]

Re-made ssdt-usb-reset, occasionally can enter the installation interface; but can not enter the installed ventura system, it feels like the graphics card is not properly driven.

Did you re-do the USB map, or used the one from the other NUC? Could you also send a HWiNFO64 hardware report?

 

[daliansky]

Did you re-do the USB map, or used the one from the other NUC? Could you also send a HWiNFO64 hardware report?

 

[daliansky]

I noticed that the two USB4 displays are connected to the core display. If the IGPU is disabled in the BIOS, the USB4 has no display output. That is to say, under macOS, there may only be two HDMIs that support display output, and the other two USB4s only support input.

 

[CaseySJ]

Please try the attached config.plist that includes the latest kernel patch to circumvent a problem with 10-bit extended tag fields. From the Linux dmesg we can see that extended tags are enabled:

[ 0.328998] pci 0000:00:01.3: enabling Extended Tags
[ 0.329051] pci 0000:00:01.3: PME# supported from D0 D3hot D3cold
[ 0.329129] pci 0000:00:02.0: [1022:14b7] type 00 class 0x060000
[ 0.329170] pci 0000:00:02.1: [1022:14ba] type 01 class 0x060400
[ 0.329255] pci 0000:00:02.1: PME# supported from D0 D3hot D3cold
[ 0.329330] pci 0000:00:02.2: [1022:14ba] type 01 class 0x060400
[ 0.329363] pci 0000:00:02.2: enabling Extended Tags
[ 0.329415] pci 0000:00:02.2: PME# supported from D0 D3hot D3cold
[ 0.329495] pci 0000:00:02.4: [1022:14ba] type 01 class 0x060400
[ 0.329577] pci 0000:00:02.4: PME# supported from D0 D3hot D3cold
[ 0.329655] pci 0000:00:03.0: [1022:14b7] type 00 class 0x060000
[ 0.329694] pci 0000:00:03.1: [1022:14cd] type 01 class 0x060400
[ 0.329729] pci 0000:00:03.1: enabling Extended Tags
[ 0.329783] pci 0000:00:03.1: PME# supported from D0 D3hot D3cold
[ 0.330474] pci 0000:00:04.0: [1022:14b7] type 00 class 0x060000
[ 0.330518] pci 0000:00:04.1: [1022:14cd] type 01 class 0x060400
[ 0.330553] pci 0000:00:04.1: enabling Extended Tags
[ 0.330606] pci 0000:00:04.1: PME# supported from D0 D3hot D3cold
[ 0.331285] pci 0000:00:08.0: [1022:14b7] type 00 class 0x060000
[ 0.331325] pci 0000:00:08.1: [1022:14b9] type 01 class 0x060400
[ 0.331346] pci 0000:00:08.1: enabling Extended Tags
[ 0.331378] pci 0000:00:08.1: PME# supported from D0 D3hot D3cold
[ 0.331439] pci 0000:00:08.3: [1022:14b9] type 01 class 0x060400
[ 0.331460] pci 0000:00:08.3: enabling Extended Tags

I've also attached a newer version of SSDT-I225V.aml.

 

[daliansky]

Please try the attached config.plist that includes the latest kernel patch to circumvent a problem with 10-bit extended tag fields. From the Linux dmesg we can see that extended tags are enabled:

[ 0.328998] pci 0000:00:01.3: enabling Extended Tags
[ 0.329051] pci 0000:00:01.3: PME# supported from D0 D3hot D3cold
[ 0.329129] pci 0000:00:02.0: [1022:14b7] type 00 class 0x060000
[ 0.329170] pci 0000:00:02.1: [1022:14ba] type 01 class 0x060400
[ 0.329255] pci 0000:00:02.1: PME# supported from D0 D3hot D3cold
[ 0.329330] pci 0000:00:02.2: [1022:14ba] type 01 class 0x060400
[ 0.329363] pci 0000:00:02.2: enabling Extended Tags
[ 0.329415] pci 0000:00:02.2: PME# supported from D0 D3hot D3cold
[ 0.329495] pci 0000:00:02.4: [1022:14ba] type 01 class 0x060400
[ 0.329577] pci 0000:00:02.4: PME# supported from D0 D3hot D3cold
[ 0.329655] pci 0000:00:03.0: [1022:14b7] type 00 class 0x060000
[ 0.329694] pci 0000:00:03.1: [1022:14cd] type 01 class 0x060400
[ 0.329729] pci 0000:00:03.1: enabling Extended Tags
[ 0.329783] pci 0000:00:03.1: PME# supported from D0 D3hot D3cold
[ 0.330474] pci 0000:00:04.0: [1022:14b7] type 00 class 0x060000
[ 0.330518] pci 0000:00:04.1: [1022:14cd] type 01 class 0x060400
[ 0.330553] pci 0000:00:04.1: enabling Extended Tags
[ 0.330606] pci 0000:00:04.1: PME# supported from D0 D3hot D3cold
[ 0.331285] pci 0000:00:08.0: [1022:14b7] type 00 class 0x060000
[ 0.331325] pci 0000:00:08.1: [1022:14b9] type 01 class 0x060400
[ 0.331346] pci 0000:00:08.1: enabling Extended Tags
[ 0.331378] pci 0000:00:08.1: PME# supported from D0 D3hot D3cold
[ 0.331439] pci 0000:00:08.3: [1022:14b9] type 01 class 0x060400
[ 0.331460] pci 0000:00:08.3: enabling Extended Tags

I've also attached a newer version of SSDT-I225V.aml.

After adding CaseySJ - probeBusGated Disable 10 bit tags - 12.0/13.0 patch, booting is fast
Unfortunately, when the boot is about to complete and enter the desktop, the screen is black, the USB keyboard is unresponsive, and the machine is unresponsive. I recorded the guide video, and the final picture is like this. . .

 

[CaseySJ]

After adding CaseySJ - probeBusGated Disable 10 bit tags - 12.0/13.0 patch, booting is fast
Unfortunately, when the boot is about to complete and enter the desktop, the screen is black, the USB keyboard is unresponsive, and the machine is unresponsive. I recorded the guide video, and the final picture is like this. . .

@daliansky,

I wasn’t able to open the ACPI tables that were posted earlier (DSDT, SSDT, FACP, etc). If you set the following parameters in OpenCore config.plist, those files will be generated in a folder called SysReport in EFI partition:

• Misc -> Debug -> Target = 67
• Misc -> Debug -> SysReport = True

After making these changes we just need to reboot and get to the OpenCore boot picker. Now the SysReport should be ready.

The two parameters above can then be changed back to their previous settings.