Leaderboard

Grazie @AlexAMD funge davvero bene Grazie ad @carlo_67 per i suoi preziosi consigli e la pazienza dimostrata unico neo che mi sono accorto perlomeno nel mio caso funziona wifi ma non la Ethenet pazienza chi troppo vuole nulla stringe 😂😂😂

In this thread we can talk about any subject related to AMD 3970x Proxmox VE guide published here:

scusa ma prima che kext usavi per la ethernet?

my microcode is 1.0003 patch c i see some gigabyte /asus mb have 1.0003 patch b i do not know if this difference could create problem for our task in my case I have flashed back bios to test i haven’t see any difference doing this

@iGPUlink posted it is useful for windows to update drivers in it? it not seems bios related Or maybe I miss something?

they say it could help also now I am trying this args: pcie_no_flr=1022:149c,1022:1487 does not work

I opened a thread on Proxmox support forum https://forum.proxmox.com/threads/msi-trx40-pro-10g-amd-3970x.69426/ in there some other useful link: https://forum.level1techs.com/t/attention-flr-kernel-patch-fixes-usb-audio-passthrough-issues-on-agesa-1-0-0-4b/151877 https://git.proxmox.com/?p=pve-kernel.git;a=summary https://bugzilla.kernel.org/show_bug.cgi?id=206543

@iGPU have you tested that Kext in a real applications? I mean in app like DaVinci Resolve or others? I my opinion and in our tests it improves Geekbench benchmarks but it has no other real application benefit in using osx with AMD gpu card. try ie CAndle test benchmark in DaVinci and you will see it is useful or not

about (my) internal audio problem: 23:00.4 Audio device [0403]: Advanced Micro Devices, Inc. [AMD] Starship/Matisse HD Audio Controller [1022:1487] passed as hostpciX: 23:00.4 and dpcimanager app see it as below: you can also see here a weird audio device: 08086293E (intel related??) if I do not blacklist blacklist snd_hda_intel when I boot on proxmox I see a message snd_hda_intel codec not found. this message seems related to starship/matisse HD Audio controller (famous 23:00.4) So or Proxmox Ve associated internal ALC audio to intel codec..or we should think how to eliminate that weird audio device you can see in pictures above If you have some ideas to share it is welcomed as usual 🙂

-dcata vm assistant option here is working maybe problem is country related?

Today I have tested my system only with host as cpu and two patches for kernel system is stable as before and this is the end for that part of my config i would like to know from users with a working vm (also am4 chipset) if audio is working well..I mean internal audio. Gpu audio is perfect. I think audio problem is also related to our common problem with some usb controller incongruences i have also tested an usb audio and the weird thing it is detected as audio usb also in those not passed USB ports... let me know 🙏

thank you @iGPU. It is above my Knowledge this..so for now I give up 🙂 For now tb is not mandatory for me..but if you find news or you would like to share your finding I will happy to read it and to post in a future TB section of main guide @Driftwood It seems most AMD GPU are affected from a bug very annoying It is called reset bug and it hangs OSX if you restart it I think people are trying to find a solution for now it seems to be not possible to avoid it I have no direct experience on it

thank you for this file--it is named like for Titan Ridge but inside I see some referral to Asus card Have you some link to guide or similar stuff to follow? thank you

I see here in my outpu some group I can pass easily..so maybe problems are different.

this could be our problem with internal audio in my opinion.. but.. I have set a windows VM, with same Proxmox configuration...in windows audio is perfect (in dp and from internal speaker from USB Realtek USB 2.0 audio controller) So..for this I am not so positive to solve this in OSX 😉

This I mean when I said you made my day 🙂 A positive and helpful approach from other users to find a solution. Uefi problem: I will put in guide's OP, I didn't think was an important step because in modern system Uefi are always advised . In my case system boots fine also in CSM and 4g (on/off) but it could be different for others Usb slowness for high capacity drive I do not have this behaviour so we have to investigate also with other user Thunderbolt part: I have an old Asus Thunderbolt Ex 3 and I didn't find any clue it could work without a thunderbolt header. So here I can't say any interesting things. vfio-pci: I did a fresh install from scratch on a clean ssd drive I have noticed a weird thing. All I need to have system booting fine in Proxmox / OSX is to complete module file in /etc/modprobe.d path to have modules loaded from kernel No blacklist.conf, no vfio.conf, no other things found on traditional way to set Proxmox ( I have asked in private to some users to test this, a x570 user confirmed these but I can't be sure for different configurations) Internal audio stuff and USB controller (related?) problems: Above screenshot is the best I can achieve It seems perfect but it isn't so..realtek USB2.0 Audio (internal audio on motherboard) is bad, scratchy and it looses synch during play audio or video viewing DP audio from Nvidia is perfect. I am investigating with @tomnic (similar problem for him on a X570 rig) and with @Gengik84with his skills to produce solutions for weird situations 🙂 ) In my try I have noticed the same things you write above: An impossibility to pass two USB controllers in may case for USB: 04:00.3 23:00.3 And in my rig these "unpassed" stuff produces 4 USB 3.x rear ports non working I have no tested internal USB 2.0 ports for now ASmedia USB 3.x ports work well (connected to my Thermaltake X9 case) I have passed fine 23.00.4 (matisse audio controller) but in my opinion it conflicts with a device passed in automatic way from VM basic parameters (I think, not sure, investigating on this) I have discovered an important thing in my rig yesterday with all things I can pass on, with audio output connected on green jack of rear motherboard backplate, booting from scratch produce this: in this situation only DP audio is working perfectly if I disconnect green jack from back, reboot linux and VM I see: Also in this situation I have a perfect DP audio from Nvidia and a bad and scratchy audio from internal matisse audio controller. in following box you can see my passed parts: 00:14.0 SMBus [0c05]: Advanced Micro Devices, Inc. [AMD] FCH SMBus Controller [1022:790b] (rev 61) Subsystem: Micro-Star International Co., Ltd. [MSI] FCH SMBus Controller [1462:7c60] Kernel driver in use: vfio-pci Kernel modules: i2c_piix4, sp5100_tco 01:00.0 Non-Volatile memory controller [0108]: Samsung Electronics Co Ltd NVMe SSD Controller SM961/PM961 [144d:a804] Subsystem: Samsung Electronics Co Ltd NVMe SSD Controller SM961/PM961 [144d:a801] Kernel driver in use: vfio-pci 02:00.0 Non-Volatile memory controller [0108]: Phison Electronics Corporation E12 NVMe Controller [1987:5012] (rev 01) Subsystem: Phison Electronics Corporation E12 NVMe Controller [1987:5012] Kernel driver in use: vfio-pci 21:00.0 VGA compatible controller [0300]: NVIDIA Corporation GP102 [TITAN Xp] [10de:1b02] (rev a1) Subsystem: NVIDIA Corporation GP102 [TITAN Xp] [10de:123e] Kernel driver in use: vfio-pci Kernel modules: nvidiafb, nouveau 21:00.1 Audio device [0403]: NVIDIA Corporation GP102 HDMI Audio Controller [10de:10ef] (rev a1) Subsystem: NVIDIA Corporation GP102 HDMI Audio Controller [10de:123e] Kernel driver in use: vfio-pci Kernel modules: snd_hda_intel 23:00.4 Audio device [0403]: Advanced Micro Devices, Inc. [AMD] Starship/Matisse HD Audio Controller [1022:1487] Subsystem: Micro-Star International Co., Ltd. [MSI] Starship/Matisse HD Audio Controller [1462:cb60] Kernel driver in use: vfio-pci Kernel modules: snd_hda_intel 44:00.0 Ethernet controller [0200]: Intel Corporation I211 Gigabit Network Connection [8086:1539] (rev 03) Subsystem: Micro-Star International Co., Ltd. [MSI] I211 Gigabit Network Connection [1462:7c60] Kernel driver in use: vfio-pci Kernel modules: igb 45:00.0 Ethernet controller [0200]: Intel Corporation I211 Gigabit Network Connection [8086:1539] (rev 03) Subsystem: Micro-Star International Co., Ltd. [MSI] I211 Gigabit Network Connection [1462:7c60] Kernel driver in use: vfio-pci Kernel modules: igb 46:00.0 Non-Essential Instrumentation [1300]: Advanced Micro Devices, Inc. [AMD] Starship/Matisse Reserved SPP [1022:1485] Subsystem: Advanced Micro Devices, Inc. [AMD] Starship/Matisse Reserved SPP [1022:1485] Kernel driver in use: vfio-pci 46:00.1 USB controller [0c03]: Advanced Micro Devices, Inc. [AMD] Matisse USB 3.0 Host Controller [1022:149c] Subsystem: Advanced Micro Devices, Inc. [AMD] Matisse USB 3.0 Host Controller [1022:1486] Kernel driver in use: vfio-pci 46:00.3 USB controller [0c03]: Advanced Micro Devices, Inc. [AMD] Matisse USB 3.0 Host Controller [1022:149c] Subsystem: Advanced Micro Devices, Inc. [AMD] Matisse USB 3.0 Host Controller [1022:148c] Kernel driver in use: vfio-pci 47:00.0 SATA controller [0106]: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] [1022:7901] (rev 51) Subsystem: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] [1022:7901] Kernel driver in use: vfio-pci Kernel modules: ahci 48:00.0 SATA controller [0106]: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] [1022:7901] (rev 51) Subsystem: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] [1022:7901] Kernel driver in use: vfio-pci Kernel modules: ahci 49:00.0 Ethernet controller [0200]: Aquantia Corp. AQC107 NBase-T/IEEE 802.3bz Ethernet Controller [AQtion] [1d6a:07b1] (rev 02) Subsystem: Micro-Star International Co., Ltd. [MSI] AQC107 NBase-T/IEEE 802.3bz Ethernet Controller [AQtion] [1462:b912] Kernel driver in use: vfio-pci Kernel modules: atlantic in this box you can see my PCI Device in VM conf: hostpci0: 21:00,pcie=1,x-vga=1 hostpci1: 02:00.0 #nvme controller hostpci10: 00:14.0 #smbus controller hostpci11: 49:00 #Aquantia ethernet controller hostpci12: 43:00 #Asmedia USB controller hostpci2: 44:00 #I211 ethernet controller hostpci3: 46:00.1 #USB controller hostpci4: 46:00.3 #USB controller hostpci5: 47:00 #Sata controller hostpci6: 48:00 #Satacontroller hostpci7: 01:00.0 #Nvme controller hostpci8: 45:00 #I211 ethernet controller hostpci9: 23:00.4 #matisse audio controller Actual /etc/modules: # /etc/modules: kernel modules to load at boot time. # # This file contains the names of kernel modules that should be loaded # at boot time, one per line. Lines beginning with "#" are ignored. vfio vfio_iommu_type1 vfio_pci vfio_virqfd actual /etc/modprobe.d/blacklist.conf #blacklist nouveau #blacklist radeon #blacklist snd_hda_codec_hdmi #blacklist snd_hda_intel #blacklist snd_hda_codec #blacklist snd_hda_core #blacklist xhci_hcd #blacklist i2c_piix4 #blacklist sp5100_tco all commentedl (I try stuff..not useful in my testing..maybe for others it could be?) vfio.conf #options vfio-pci ids=1022:148c #options vfio-pci ids=1462:cb60 #options vfio-pci ids=1022:1487 all commented (obviously tested but no useful result for my rig)

@Driftwood from terminal VM_assistant -dcata here it is working maybe an Apple server problem from your country?

I @iGPU you made my day! 🙂 very happy for your message (this is all I have wished from day 0 from other user! I explain later 🙂

thanks @Rocket88..attached a simple screenshot:

@AllubzVso many question 😛 boot order: here is well explained (I hope) how to (I will revert also OVMF patched file in an original one..but for you could be a problem..in my case it is working original one) higher Cinebench : CPU: Penryn delete all args (all I mean) in this way benchmark are higher..sytem has weird behaviour..maybe I need to insert +invtsc boot arg I DO NOT advise to do in this way..I am trying for testing also experimental patches.. Correct name in cinebench only with OpenCore for now in args host as cpu then CPU: Penryn in vm config it is mandatory to use two "AMD Vanilla patches" by algrey..Search on the net because I think it is not a good way for us (patches slow our GPU in my testing) It is beautiful to see Cinebench, MacCPUid and other app recognise our cpu without faking anything. For other your problems you can test some deb files patched by @Pavo you can find in his GitHub (read his indication for this). link here This files could help to separate in a better way IOMMU and its shared peripherals..it solves some problems on X570 based rigs.

a test for AMD Gpu users. If you know what you are doing and you are booting your OSX from a passed disk try to comment EFI disk in your VM config In my system it boots fine the same except for Nvidia driver not loaded correctly, I think it is for me a Nvram problem.. I am curious to see what happens with OSX supported GPU to comment VM line you can put this # at line's start

hi I see 4 potential problems: 1) args: -device isa-applesmc,osk="ourhardworkbythesewordsguardedpleasedontsteal(c)Apple$ this is not complete 2) net0: vmxnet3=92:C8:08:B1:2D:BE,bridge=vmbr0 net1: e1000=FE:8F:05:47:A8:1E,bridge=vmbr0 net2: rtl8139=46:D4:B9:63:16:44,bridge=vmbr0 net3: virtio=36:6C:02:76:47:2E,bridge=vmbr0 3) sockets: 3 4) you miss installer iso in your VM (HS,Catalina, Mojave) I see only clover iso also a potential error is to leave media cdrom on ide0 line you should follow the guide and put there cache=unsafe for point 3) put 1 and use all your 48 cores in cpu I do not think this is a 3960x problem...I have 64 core and 1 socket from day 0 and yes, if you pass your VGA you have to put none in display part

Hi post your VM config and shell output of: lspci -nn

😂😂😂😂

@Renato however: then type: edit config.plist and configure as you want it is not necessary for the goal of this guide..but here you can see "How to" 🙂

@Renato hi 😉 if you followed the guide or downloaded a proper clover iso searching in the net..if you examine it you will see it has a very basic config.plist SMBIOS part is blank and you have to fill with proper data according to your system To boot well is not mandatory to have some Apple services configured well it is necessary to have a working chance!

ehmm have you formatted it in disk util? then..why scsi? I never tested scsi..but I think it is not your problem (also never checked backup, some time SSD emulation but it is not relevant @Renatofor your disk miss! see also If you see it here :

I do not understand well this however I have installed with clover, then I have tried Opencore and it is now working better (it solves bad graphic I some condition) Your problem is a linux problem..you can't see your mouse I think so OSX is not yet involved in this problem

ciao IGPU when osx boots..you have the same problem you have in a regular hackintosh However not tested Intel BT/wifi stuff I have had x299 gigabyte designare mb before this one and I have had many problems with its integrated thunderbolt on my previous rig With sage 10G ant titan ridge all was working fine (hot plug and so on) In this one I have not tested any TB adapter for now. My MSI has not a thunderbolt header and I would have to test with titan Ridge and the cable trick

you can edit via shell not simple fs0:/ or as your EFI it is named in your system.. then terminal command

You can also try to pass completely your unique usb controller hostpci3: 00:14 in your conf file but i have no others idea

I see @Renato i think in this way is fine so if you cant see a mouse attached to it the only way for now is to connect your mouse on a onboard usb port I cant see your hardware in your signature but i think some spare usb port could be there on your motherboard

if you can post also cinebencz 15 🙂 if I understand well in Proxmox is better for you than previous patched method?

have you passed this one? if you do not pass it, in Web interface/hardware/usb part what do you see?

@Pavo cpu is fine, you can see comparing in real way with no vm benchmark using traditional patches gpu benchmarks are missing for now have you some problem there? @Renato for inserting pictures you can drag and drop when you are writing your message

original order of passed sata controller changement to original position saving and committing as it should be it seems fine... after a restart:

thank you for uploading pictures..but you have also to delete all your qemu drive (if you have in VM conf) and see if after a cold boot your boot sequence is fine, in my case it starts always with network entering in bios and select sata passed disk it boots fine

to test experimental kernel patches I can't say anything because I do not know your lspci output.. but if it is ok for you.. 🙂 add: by the way it seems fine to me, maybe you miss some cpu features args and I do not see internal motherboard audio

here in the guide the indication requested.. However I will try to explain better @Renato

Hi Renato, this guide is in continue evolution.. Passed most of things and that link is very useful.. I am trying to clear all other things I did and put here in a clear way 🙂 These are some of passed drive (passed 2 sata controllers) and following Asmedia and USB 3.1 and 3.2 USB controllers passed completely : I discover also a big OVMF bug..but I am investigating it 😉 ah sorry HS.iso is basesystem.dmg converted in iso and uploaded on your host via web interface clover.iso is the same for clover.

Ciao Renato You have to be logged inside this forum to download any file you want

On my test ssd (with Catalina 10.15.4...System starts without any "QEMU" virtual drive...

@Imanol only fp16c is not taken..bu I have not investigated for now with these features I can install and boot fine with all OSX version

@Pavoput yours and check well linux part of passthrough (iommu and so on)

then take a look here stepping : 0 microcode : 0x8301025 cpu MHz : 3510.071 cache size : 512 KB physical id : 0 siblings : 64 core id : 5 cpu cores : 32 apicid : 11 initial apicid : 11 fpu : yes fpu_exception : yes cpuid level : 16 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate sme ssbd mba sev ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif umip rdpid overflow_recov succor smca bugs : sysret_ss_attrs spectre_v1 spectre_v2 spec_store_bypass bogomips : 7400.02 TLB size : 3072 4K pages clflush size : 64 cache_alignment : 64 address sizes : 43 bits physical, 48 bits virtual power management: ts ttp tm hwpstate cpb eff_freq_ro [13] [14] start to pass in your vm boot args this one on one starting from tested ones execute cat /proc/cpuinfo in linux shell.

Camera Raw in action 🙂

Happy for this i am not a linux expert so i have tried to stay as simple possible i suggest to start with a clover iso after you have properly configured Proxmox environment in this way i have had success to configure with an only EFI all main OSX system 🙂

Benchmark VM: Windows: Benchmark Windows (NO VM): Benchmark VM: OSX macOSHigh Sierra 10.13.6 (17G12034) Benchmark (NO VM): OSX macOSHigh Sierra 10.13.6 (17G12034) NOT AVAILABLE see here: