Ryzen 7000 Testing

[fabiosun]

and this are my IOREG
Big Sur with all BIOS option enabled
i225 works without any kext
WIFI/BT works vanilla without any kext
Thunderbolt works (hot plug no)
Audio USB works (with mapping)

Ventura 13.2 with all BIOS option enabled (with PCI bootargs or CaseySJ Kernel patch)
I225 does not work
WIFI/BT does not work
Thunderbolt does not work
Audio usb does not work (not checked HDMI)

with @Lorys89 had tested many pci bootargs and also others boot in Ventura but without improvement

attached ioreg for Ventura and BigSur with logs for both

edit:
Monterey with all BIOS option enabled (with CaseySJ Kernel patch)
i225 works
WIFI/BT works vanilla without any kext
Thunderbolt works (hot plug no)
Audio USB works (with mapping)

 

[pbs888]

My X670E Gene exhibits the same behavior, namely:

• Internal WiFi disabled
  • Allows Thunderbolt to be enabled
  • Allows i225-V to work
  • Allows Broadcom WiFi/BT in PCIe slot to work
• Internal Thunderbolt disabled
  • Allows Intel WiFi/BT to work
  • Allows i225-V to work
  • Allows Broadcom WiFi/BT in PCIe slot to work

There appears to be a conflict in macOS between internal WiFi and internal Thunderbolt. We cannot enable both at the same time.

I’ll check whether this is true in Big Sur and Monterey, or only in Ventura.

Exactly same on my Asus X670E Extreme. I've reseted bios to stock, disabled the usual stuff as per open core config and no ethernet / no fenvi wifi/BT.

 

[leesurone]

some thoughts...

My system is:
ASUS Rog X67E Hero
7950x
128 Gb DDR5 5200 Mhz (@3600)
AMD 6900 XT
a couple of internal NVME and an external thunderbolt case with an NVME drive.
In my bios I had thunderbol enabled and only wifi and BT disabled as I was using the card below

View attachment 9618

with this card mounted in the last slot I had all the functionality without any problems (Thunderbolt without Hot Plug, airdrop, wifi and so on)
I then decided to make this board compatible with OSX:

View attachment 9619

I enabled wifi and bluetooth from bios and verified in windows that I had mounted the wifi card well instead of the stock one
all working

i start in OSX and this is where the problems start (i am using Ventura but today i want to try Big Sur)

Basically with everything on you lose many functions (i225 ethernet, wifi, BT, USB, thunderbolt)
to get everything but wifi working again, you have to disable wifi in the bios
On the other hand, if I disable thunderbolt and enable everything else, everything works except thunderbolt of course

It seems like there is no physical space for all our devices (said approximately and without technical pretensions)

what are your experiences in this ?

With the PCIE card everything was perfect (obviously disabling WIFI AND BT from bios)

I have the same issue

 

[leesurone]

and this are my IOREG
Big Sur with all BIOS option enabled
i225 works without any kext
WIFI/BT works vanilla without any kext
Thunderbolt works (hot plug no)
Audio USB works (with mapping)

Ventura 13.2 with all BIOS option enabled (with PCI bootargs or CaseySJ Kernel patch)
I225 does not work
WIFI/BT does not work
Thunderbolt does not work
Audio usb does not work (not checked HDMI)

with @Lorys89 had tested many pci bootargs and also others boot in Ventura but without improvement

attached ioreg for Ventura and BigSur with logs for both

edit:
Monterey with all BIOS option enabled (with CaseySJ Kernel patch)
i225 works
WIFI/BT works vanilla without any kext
Thunderbolt works (hot plug no)
Audio USB works (with mapping)

Well that's interesting that everything works in Monterey.

 

[fabiosun]

I have the same issue

ALL we have
I did further testing
Big SUR and Monterey works perfectly with all active in BIOS
from Ventura 13,01 we loose some device (I think USB and others) and to have system booting we have to disable WIFI or thunderbolt
So maybe patch which allows us to boot properly needs some adjustment
I discovered the lately because in my further tests I have always BT and WIFI disabled because I had a fenvi on PCIE

 

[Galve2000]

It seems like there is no physical space for all our devices (said approximately and without technical pretensions)

There appears to be a conflict in macOS between internal WiFi and internal Thunderbolt. We cannot enable both at the same time.

from Monterey 13,01 we loose some device (I think USB and others) and to have system booting we have to disable WIFI or thunderbolt
So maybe patch which allows us to boot properly needs some adjustment

All very informative -- thank you!

for me, macOS is not macOS if I cannot use Airdfrop, handoff, and the original "continuity" so as much as I'm sure the onboard wifi and BT are amazing, they shall remain disabled in favor of the Fenvi T919.

I do wish these Asus boards had one more PCI slot.

 

[fabiosun]

All very informative -- thank you!

for me, macOS is not macOS if I cannot use Airdfrop, handoff, and the original "continuity" so as much as I'm sure the onboard wifi and BT are amazing, they shall remain disabled in favor of the Fenvi T919.

I do wish these Asus boards had one more PCI slot.

There are also other problem for users like us with also as media SATA controller with 2+4 sata ports:



on second nvme (4.0) I have Big Sur installed via USB
I can update to Monterey (with all bios option enabled (TB,WIFI,and so on)
then I can update to Ventura 13.01 disabling WIFI on bios (TB, i225 , audio work perfectly using Casey patch or booting)

only for the sake of testing I have installed Monterey on my ASMedia controller
I have there a BD Pioneer burner and a Samsung SSD
installation went fine
but, from there I can't update to Ventura 13.01
On monterey installed on NVME I can update till 13.2 with usual attention to disable wifi

Something is changed in macOS or the patches are not fully working in latest OS

 

[leesurone]

ALL we have
I did further testing
Big SUR and Monterey works perfectly with all active in BIOS
from Ventura 13,01 we loose some device (I think USB and others) and to have system booting we have to disable WIFI or thunderbolt
So maybe patch which allows us to boot properly needs some adjustment
I discovered the lately because in my further tests I have always BT and WIFI disabled because I had a fenvi on PCIE

Hi @mariettosun, so everything works fine disabling the onboard Wifi/ Bluetooth module and using a Fenvi adapter instead in Ventura? Thanks in advance

 

[fabiosun]

so everything works fine disabling the onboard Wifi/ Bluetooth module and using a Fenvi adapter instead in Ventura?

Yes (you have probably to map your usb for audio and BT if you have ab ASUS hero with USB Realtek audio)

 

[Galve2000]

Yes (you have probably to map your usb for audio and BT if you have ab ASUS hero with USB Realtek audio)

I did not include any audio boot args using your EFI as a jumping off point and it works well -- at least using the rear "lime green" output. i have not tied HDMI audio from my 6650XT but will do so in the next few days / weeks when I move this system from my test bench to my LG OLED.

 

[CaseySJ]

Investigation Ideas:

Option 1:

• We can enable full PCI logs with boot argument pci=0x3
• We'll also need DebugEnhancer.kext with boot arguments msgbuf=3145728 io=0xff -dbgenhdbg -dbgenhiolog
• With Thunderbolt and internal WiFi both enabled in BIOS we can then boot Ventura
• After logging in we should very quickly:
  • Open Terminal
  • Capture these log files

  • sudo dmesg > ~/Documents/dmesg1.txt
    log show --last boot > ~/Documents/bootlog1.txt

• Then compress and post the two log files from Documents folder:
  • dmesg1.txt
  • bootlog1.txt
• These logs will show us complete details of PCI bus discovery and enumeration
  • In other words, it will provide details on every device that macOS finds on the PCI bus before it attaches any drivers

Option 2:

• We can download source code for IOPCIFamily for both Monterey and Ventura
• We can use a diff tool to compare files
• We can also use GitHub to look at commits made in Ventura version of IOPCIFamily
  • This is easy enough to do by clicking here
• It's possible that a change was made to AppleACPIPlatform.kext, in which case looking at IOPCIFamily will not help
  • AppleACPIPlatform source code is not available in the Apple Open Source repository

 

[leesurone]

Yes (you have probably to map your usb for audio and BT if you have ab ASUS hero with USB Realtek audio)

I decided to take a shot at enabling Thunderbolt (again) this weekend after reading your earlier post that showed it working. Needless to say I struck out and even though everything works as expected I started questioning the efficacy of my EFI build. I've got the same board as Casey but for whatever reason his masterpiece would not boot for me. Reading your post today and then see everyone chime in that they had the same issues was awesome. Just wanted to say thanks to everyone contributing here and for the new information on Thunderbolt conflicting with WFI/ Bluetooth. I've got a Fenvi sitting on a shelf but until I get bored I'm leaving my system as. It works perfectly and I really don't need Thunderbolt on this build. Knowing I could enable though is worth a lot.

 

[fabiosun]

Same Bios Settings
macOS Ventura


macOS Monterey



macOS Big Sur

 

[fabiosun]

Well that's interesting that everything works in Monterey.

also in BigSur

 

[fabiosun]

further thoughts on the topic (problems in Ventura)

Although some do not consider clover a viable bootloader today (no controversy, but it is a fact that it is considered obsolete and outdated)

for me it has always been a source of proof and confirmation of some theories

The tremendous work done by Casey has allowed the AM5 platform to work well, still not perfectly in all OSes but usable professionally

For my part I would never have noticed the problem we have identified (partially) now if I had not replaced the internal wifi card in my motherboard with an OSX compatible one

back to CLover, we know that the bootloader developers have not yet transposed the fundamental changes to be able to use audio seamlessly on this platform.

the modification made in opencore brings problems though, especially in some areas

If you go and read some old posts that I can't find now about the various benchmarks and how to use fake ways to get higher performance for example on GeekBench (downloadfritz also talks about it somewhere in another geological era ) you will see that modifying data hub and TSC gives higher results at the expense of audio desynch and other oddities
closed this digression, using clover or opencore before the patch for AMD audio and an old GPU benchmark (Cinebench R15) with my 6900 XT I get results of about 160 fps. With Opencore incorporating the appropriate modification to remove the audio issues a stratospheric result of about 300 fps
Now many would be happy..but here is a problem!
How does it manifest itself in other fields? I don't know!

back to the macOS Ventura problem
Using old opencore and the bootarg that was discovered before CaseySJ's final patch was written all OSes start even if you have everything enabled in the bios (tb, wifi i225 an so on). clearly ventura will not have all these peripherals available, while older OSX will

Using new opencore and the bootarg pci=0x8 etc etc you have an instant reboot on ventura, while all ok on older OSes
Using new opencore and Casey's patch you will start quietly with the usual problems in Ventura

all this to say that in my opinion those who know about these things should investigate the problem of TSC screwing up the results (in some benchmarks and not in all tests) and those who are interested should help with logs and with their own experiences those who are working hard to perfect this platform (I'm referring to Casey and also to ExtremeXT and Lorys89 who have been working hard over time)

sorry for the length of these thoughts!

 

[fabiosun]

Investigation Ideas:

Option 1:

these are the bootargs used:
pci=0x3 msgbuf=3145728 io=0xff -dbgenhdbg -dbgenhiolog revpatch=cpuname e1000=0 agdpmod=pikera

with these enabling Thunderbolt in Ventura system has an instant reboot
without enabling thunderbolt these are logs system creates

with tb enabled in bios and same EFI and bootargs Big Sur and Monterey boot fine and these are related logs

 

[CaseySJ]

It's a very busy week at the office so I may need to wait until weekend for more thorough investigation. My first task will be to obtain PCI debug logs from Ventura. This should be possible perhaps by removing io=0xff from boot arguments. I'll also examine standard Ventura boot log to see if there are any AppleACPIPlatform log messages that provide a clue. Standard boot log is obtained simply as follows:

log show --last boot | head -1500 > ~/Documents/bootlog.txt

But comparing PCI debug logs between Monterey and Ventura may be more effective than any other option at this time. So it's imperative that we generate PCI debug log from Ventura.

 

[CaseySJ]

@mariettosun,

Do you have a flashed GC-Titan Ridge add-in-card?

If so, a simple experiment to try would be:

• Disable on-board Thunderbolt in BIOS
• Enable on-board WiFi
• Install GC-Titan Ridge and bridge top and middle pins of vertical 5-pin J1 header to power on the card
• Modify Thunderbolt SSDT with PCI path of GC-Titan Ridge
• Boot Ventura and see if GC-Titan Ridge works (including hot plug)

 

[Galve2000]

with these enabling Thunderbolt in Ventura system has an instant reboot

I am getting random instant reboots on powering up in Monterey as well. I did not know that it was TB related. I have not yet tested TB bc I am temporarily using the only TB peripheral I own, a TB NVME enclosure, as a TimeMachine drive on my X570/5950X build.

my work-around is to boot from the USB stick containing the Install Monterey , for some reason that will get me to the desktop of my already built system. I can then reboot without the USB stick into the desktop as normal.

unfortunately this process is long and tedious what with having to

1. try to boot in to the desktop as normal.
2, rebooting into the BIOS hopefully having enough time to insert USB stick so that BIOS can recognize it.
3. Boot into MacOS drive from USB OP Boot Picker
4. Reboot into MacOS without the USB.

the EFI on the USB stick is identical to the EFI on my MacOS drive save for a RadeonSensor.kext -- I will add this tonight to the USB stick EFI so that they can be exactly identical. I do not think the lack of this kext is what is causing a reboot. I do get an error that the kext is missing when RadeonSensor launches on booting into the desktop.

 

[fabiosun]

I am getting random instant reboots on powering up in Monterey as well

I suggest to use a minimal set o f kext and check your mmio
If all is fine there try to enable only the second one

@CaseySJ i have not tested Titan ridge card I will try soon but I think it will work