trx40 [Discussion] - TRX40 Bare Metal - Vanilla Patches

[meina222]

@iGPU - I tried to take it really conservative on the SSDT's

 

I so far only applied these

 

DTPG

PLUG

SBRG

SBUS (after checking my SMBus id in hackintool)

EC USBX

XHC rename (your rename one)

 

It boots and seems good. Could you pls check my IO registry? Did the rename work? I will try AX200 tmw - I'd like to understand how and why it works

 

I also did a test of my USB ports 1 by 1 and am a bit puzzled by the result. Will share those tmw. Would like to trim USB ports to a set that works w/out excess.

 

Btw, I inject my I210's in config.plist device properties and they show as built-in. Would I still need the SSDT? If so what does it do? Thanks for the insights!

 

iMac Pro.zip

Edited August 13, 2020 by meina222

[fabiosun]

19 minutes ago, iGPU said:

fabiosun,

 

another round for you to test USB-C activity. now there is new SSDT and a kext file to help inject USB properties. The kext file only injects USB for the TB card, no other devices.

 

(I used a similar type of kext file to increase USB speed on an ASRock Z390 ITX build.)

 

GB-TR-TB3.zip 5.21 kB · 1 download

 

kext seems not to load

Last login: Thu Aug 13 08:35:01 on console
fabios-iMac-Pro:~ fabio$ kextstat | grep -v com.apple
Index Refs Address            Size       Wired      Name (Version) UUID <Linked Against>
   36    2 0xffffff7f83318000 0x2a000    0x2a000    as.vit9696.Lilu (1.4.6) E1988EED-9336-3A82-A361-9F31FDE22EDE <7 5 4 3 2 1>
   37    0 0xffffff7f8335b000 0x145000   0x145000   as.vit9696.AppleALC (1.5.1) F3D6D4E7-6EEF-37E9-B814-027E16746B95 <36 12 7 5 4 3 2 1>
   50    0 0xffffff7f83342000 0x19000    0x19000    as.vit9696.VirtualSMC (1.1.5) D8B947B4-805E-357E-AD37-F96AB13AB87B <36 11 7 5 4 3 2 1>
   64    0 0xffffff7f834a9000 0x1a000    0x1a000    com.smalltree.driver.SmallTreeIntel82576 (1.0.6) F6C5E33A-97AA-37A0-8794-64FAA19FC153 <15 12 5 4 3 1>
  101    0 0xffffff7f817e9000 0x3000     0x3000     com.nvidia.NVDAStartupWeb (10.3.3) 4F3B6943-C73B-378C-9EA7-D36569B4B6D3 <12 4 3>
fabios-iMac-Pro:~ fabio$ 

 

[iGPU]

On 8/12/2020 at 9:06 PM, meina222 said:

I believe this the AX200.

 

I notice it is under the same root XHC0 node with the bluetooth device I want enabled (2 of them in this picture as I also plugged a BCM dongle BT to test) - the other 2 BTs are Apple compatible and should not require a driver. But this guy in the middle is the AX200.

 

I wish I could share better resolution - I paste from Apple cmd + shift + 3 tool and comes in blurry.

 

Edit - I hope to able to delete this port using this tutorial tomorrow

 

https://www.tonymacx86.com/threads/the-new-beginners-guide-to-usb-port-configuration.286553/

 

 

OK, based on this, here is a different approach to remove the AX200 by removing PRT5.

 

Use this SSDT (attached below) instead of the NoAX200 SSDT:  do not use NoAX200 nor the XHCx-Rename SSDTs.

 

Let me know how this gets on for you.

 

 

***

 

 

 

On 8/12/2020 at 11:38 PM, fabiosun said:

 

kext seems not to load

Last login: Thu Aug 13 08:35:01 on console
fabios-iMac-Pro:~ fabio$ kextstat | grep -v com.apple
Index Refs Address            Size       Wired      Name (Version) UUID <Linked Against>
   36    2 0xffffff7f83318000 0x2a000    0x2a000    as.vit9696.Lilu (1.4.6) E1988EED-9336-3A82-A361-9F31FDE22EDE <7 5 4 3 2 1>
   37    0 0xffffff7f8335b000 0x145000   0x145000   as.vit9696.AppleALC (1.5.1) F3D6D4E7-6EEF-37E9-B814-027E16746B95 <36 12 7 5 4 3 2 1>
   50    0 0xffffff7f83342000 0x19000    0x19000    as.vit9696.VirtualSMC (1.1.5) D8B947B4-805E-357E-AD37-F96AB13AB87B <36 11 7 5 4 3 2 1>
   64    0 0xffffff7f834a9000 0x1a000    0x1a000    com.smalltree.driver.SmallTreeIntel82576 (1.0.6) F6C5E33A-97AA-37A0-8794-64FAA19FC153 <15 12 5 4 3 1>
  101    0 0xffffff7f817e9000 0x3000     0x3000     com.nvidia.NVDAStartupWeb (10.3.3) 4F3B6943-C73B-378C-9EA7-D36569B4B6D3 <12 4 3>
fabios-iMac-Pro:~ fabio$ 

 

 

I don't think this type of kext shows up. Just like RadeonBoost nor AGPMInjector show up. I think if kext only contains an info.plist they do not show up. 

 

Simply look for TB USB function.

 

 

 

Edited August 23, 2020 by iGPU

[fabiosun]

ok, tested but no usb is seen if connected

I am using this:

 

and also an external disk (not thunderbolt)

fabio’s iMac Pro ssdt and kext by IGPU.zip

Edited August 13, 2020 by fabiosun
added IOREG

[iGPU]

31 minutes ago, meina222 said:

@iGPU - I tried to take it really conservative on the SSDT's

 

I so far only applied these

 

DTPG

PLUG

SBRG

SBUS (after checking my SMBus id in hackintool)

EC USBX

XHC rename (your rename one)

 

It boots and seems good. Could you pls check my IO registry? Did the rename work? I will try AX200 tmw - I'd like to understand how and why it works

 

I also did a test of my USB ports 1 by 1 and am a bit puzzled by the result. Will share those tmw. Would like to trim USB ports to a set that works w/out excess.

 

Btw, I inject my I210's in config.plist device properties and they show as built-in. Would I still need the SSDT? If so what does it do? Thanks for the insights!

 

iMac Pro.zip 6.16 MB · 1 download

 

Some worked; ETH0 didn't. DevProp injects info for PCI section. The SSDT simply renames devices; nothing spectacular.

 

 

[Ploddles]

10 hours ago, meina222 said:

@Ploddles

 

Try this EFI working on the TRX40 Designare which is very similar to the TRX40 Master. Just make sure to fill in your serial and ethernet ROM in the nvram section and try again

 

 

 

 

EFI.zip 6.89 MB · 1 download

Thanks, I’ll try it tonight. 

[iGPU]

I found 10.15.5 beta was stable under Proxmox, but I had crashes once I upgraded to 10.15.6. There were reports that 10.15.6 was unstable under VMware, so now an apparent fix:

 

 

Edited August 13, 2020 by iGPU

[meina222]

7 hours ago, iGPU said:

 

Some worked; ETH0 didn't. DevProp injects info for PCI section. The SSDT simply renames devices; nothing spectacular.

 

 

 

I didn't apply the ETH yet. I only added the rename XHC to try as I figured it'd be easier to check what works and what not one by one. I will try new AX200 later today when I get off work.Thank you! 

Edited August 13, 2020 by meina222

[meina222]

I noticed a strange behavior - when I try to shut down  Mac OS. on the bare metal (which reboots the PC) and then turn it off from the button, the next time I power it up, the PC and MacOS take much longer to start. Between the time the OC log is done, and I see the login screen it takes close to 10 sec. Once in MacOS I don't see anything unusual yet performance wise. The way to recover from this is to turn off the power button of the PSU to force un-powering the motherboard - then next reboot startup is very fast again (as I expect it to be).

 

I wonder if there are some USB power delivery issues causing this. I noticed that a few times in Proxmox where a reboot of the VM + hot PC reboot will leave some USB ports passed thru to MacOS with various errors and Linux takes longer to boot to diagnose these. Before it happened maybe 10-20% of the time per VM start/stop but now I see slowness every boot unless I de-power.

[iGPU]

18 minutes ago, meina222 said:

I noticed a strange behavior - when I try to shut down  Mac OS. on the bare metal (which reboots the PC) and then turn it off from the button, the next time I power it up, the PC and MacOS take much longer to start. Between the time the OC log is done, and I see the login screen it takes close to 10 sec. Once in MacOS I don't see anything unusual yet performance wise. The way to recover from this is to turn off the power button of the PSU to force un-powering the motherboard - then next reboot startup is very fast again (as I expect it to be).

 

I wonder if there are some USB power delivery issues causing this. I noticed that a few times in Proxmox where a reboot of the VM + hot PC reboot will leave some USB ports passed thru to MacOS with various errors and Linux takes longer to boot to diagnose these. Before it happened maybe 10-20% of the time per VM start/stop but now I see slowness every boot unless I de-power.

 

I'm seeing the exact same behavior. I think it briefly improved after re-setting NVRAM.

 

I have another set of SSDT and ACPI patches that work on the Intel side to see if they work here. One specifically addresses USB.

[Ploddles]

Getting further than yesterday thanks to @meina222 but still no joy. Stops at

 

[Configuring PCI]

 

when trying to install Big Sur, and just stays there. Waited up to 10 minutes but it doesn’t progress any further. 
 

Will try with Catalina tomorrow but I don’t hold much hope of it progressing beyond this point, but who knows. 
 

[tsongz]

Was able to install Catalina 10.15.6 (iMacPro1,1) bare metal with a new NVMe drive, and the stock EFI folder. One minor issue that some might run into if you're not writing over Proxmox is that after the first stage of the installer that goes back into the "Install OSX" reboot, if your config isn't mapped to your other storage devices, the installer will fail. In my case, I ended up just unplugging them.

 

What works now in Bare Metal:

 

WiFi from the PCIe Adapter (BRCM4360 - which confirms it was a Big Sur issue)

RAM Modules set and registered properly.

 

 

What doesn't work now:

 

Ironically the Bluetooth, as I also have AX200 showing up, despite disabling it in BIOS.

 

 

Minor benchmarking shows a slight dip in performance, but I'm willing to take that performance dip in exchange for system stability.

[tsongz]

8 minutes ago, Ploddles said:

Getting further than yesterday thanks to @meina222 but still no joy. Stops at

 

[Configuring PCI]

 

 

I ran into this issue once during the process. It's generally caused when you're mid-way through the install (first reboot), and then there's changes in your devices, whether that's from USB positions, moving disks around, devices around, etc. So it would be best to restart the install process from scratch on the disk, so it gets remapped during that first stage of the installer.

[meina222]

@iGPU

 

Your latest SSDT worked to disable the AX200! You got this!!

 

I have BT now working and enabled. I also finished the Catalina 10.15.6 suppl. update without an issue.

 

I was able to track down my delayed boots to one of my AIC NMVE's failing to initialize. I use the in the VM zfs pool - 4 NVMEs in the Aorus PCI riser. Not sure if this is MacOS's fault. I haven't applied your NVME SSDT yet. As switching off PSU power switch recovers it, it's not a faulty NVME but rather some power mismanagement.

 

Attached is latest IOReg explorer file and screenshot of Hackintool confirming disappearance of AX200.

 

 

iMac Pro 2.zip

Edited August 13, 2020 by meina222

[tsongz]

@meina222 running into the same boot/reboot issues, but it's slightly different. 

 

For some reason, whenever MacOS boots, or restarts, it seemingly disables the motherboard CMOS. What has happened like 10 times, is I do either restart/shutdown (from MacOS), and when it goes into the power cycle:

 

a) shutdown is basically a restart. Not sure why - and this is 50/50 whether b happens

b) restart causes blank screen (no BIOS POST, which can be seen from the lack of Dr CMOS lights). Fix is to unplug all, clear CMOS, and reboot. Nothing really gets affected on the MacOS volume itself. As it goes into MacOS (as I'm typing this) - I just noticed that DrCMOS is disabled. 

 

It may be worth noting that I have no SSDs or HDDs attached, just two NVME drives, one with BS (from Proxmox) and one with Catalina (metal), and there is a note during boot log that no cache is created because it's detecting NVME drives.

 

Interestingly, CPU performance benchmarks are noticeably lower, while after adding my Vega FE patches, Metal performance is significantly higher. Not only that, when measuring device bandwidth with Cling, it's gone up a lot as well. OpenCL is hit or miss. Results are also consistent with GB5 Results. The metal API on iMacPro for Vega drivers seems to be much better than for MacPro. 

 

 

 

Highlighted is results from PVE Big Sur (MacPro7,1 - same mods, - I'm certain that's my benchmark bc  nobody else has a Vega FE on these charts lol)

 

Previous for comparison, was only 8 GB/s for both Host to Device and Device to Host, and 148 GB/s for Device to Device. About 50% increase in bandwidth.

 

 

[iGPU]

FYI,

 

I tend to compile the latest OC commits. Currently, it is v061 12 Aug. However, this version and the commit from 11 Aug do not boot into Catalina on the TRX40. I have not taken the time to investigate as I'm working Shutdown and TB issues.

 

However, to get the computer to boot, I use v061 9 Aug, which seems very stable, and must perform a reset NVRAM from OC menu before booting. Otherwise, the computer only enters into Recovery. I've even tried this with Debug version. Perhaps some other settings in Quirks are necessary.

 

Anyhow, just a heads up. BTW, what the EFI that I'd uploaded earlier in this thread is v061 9 Aug, so this is good to go.

[meina222]

@iGPU- thanks for the nvram tip - I'll try it. I don't experience CMOS reset but I experience slowness and funkiness of NVME on reboot. I'll try NVRAM.

 

Q on SDDT - I noticed in MaciASL that you added port addresses closely matching iMacPro's SSDT (I presume) - the one where you disabled the AX200. With this, is there a point of me trying to generate a kext with Hackintool and remove whatever doesn't work to bring it to a smaller set? Or should I just leave it as. Most of my ports seem to work - but I've heard that if you narrow your set and disable limit of 15 removal your system would behave more stable and with perhaps better power management on devices - myth or truth?

 

Finally able to get some sane explanation on this:

 

https://aplus.rs/2020/usb-mapping-why/

https://aplus.rs/2020/usb-mapping-how/

 

Will study these more. I think something that may be of use for me in this guide is that I have 2 XHC0 controllers so it maybe beneficial to rename one of them.

Edited August 14, 2020 by meina222

[iGPU]

3 hours ago, meina222 said:

@iGPU

 

Your latest SSDT worked to disable the AX200! You got this!!

 

I have BT now working and enabled. I also finished the Catalina 10.15.6 suppl. update without an issue.

 

I was able to track down my delayed boots to one of my AIC NMVE's failing to initialize. I use the in the VM zfs pool - 4 NVMEs in the Aorus PCI riser. Not sure if this is MacOS's fault. I haven't applied your NVME SSDT yet. As switching off PSU power switch recovers it, it's not a faulty NVME but rather some power mismanagement.

 

Attached is latest IOReg explorer file and screenshot of Hackintool confirming disappearance of AX200.

 

 

iMac Pro 2.zip 9.06 MB · 0 downloads

 

I happy this worked!

 

This is technique I'd used on the X570 build. Basically, the SSDT simply re-defines all of the USB ports as is done in DSDT file, but leaving out PRT5. PRT5 then disappears.

 

I also re-named the 2 USB devices XHC and XHCI, but did not try to re-name PRTx or HS0x.

 

If you study the SSDT you'll understand how it works. If you ever wanted to re-activate PRT5, just enter it into the list, being careful of the port numbers.

[iGPU]

On 8/13/2020 at 8:05 PM, meina222 said:

@iGPU- thanks for the nvram tip - I'll try it. I don't experience CMOS reset but I experience slowness and funkiness of NVME on reboot. I'll try NVRAM.

 

Q on SDDT - I noticed in MaciASL that you added port addresses closely matching iMacPro's SSDT (I presume) - the one where you disabled the AX200. With this, is there a point of me trying to generate a kext with Hackintool and remove whatever doesn't work to bring it to a smaller set? Or should I just leave it as. Most of my ports seem to work - but I've heard that if you narrow your set and disable limit of 15 removal your system would behave more stable and with perhaps better power management on devices - myth or truth?

 

Finally able to get some sane explanation on this:

 

https://aplus.rs/2020/usb-mapping-why/

https://aplus.rs/2020/usb-mapping-how/

 

Will study these more. I think something that may be of use for me in this guide is that I have 2 XHC0 controllers so it maybe beneficial to rename one of them.

 

I'd given you other re-names, but they overlapped with the AX200 eliminator SSDT. So I re-did the batch for you. I think you can safely use all of them. Note that there is a v2 of the EC-USBX SSDT file.

 

 

Edited August 23, 2020 by iGPU

[iGPU]

I sourced an old GB-Titan Ridge AIC (previously flashed). The TB part works great, but not the USB. Very odd.

 

I know that the card is good as it does have proper TB/USB when used in other machines, including an AMD X570.

 

Shown below is what's seen when a TB HD is connected to one port; ideal result:

 

The other port has a USB-C drive connected, but the drive does not mount:

 

Without a TB SSDT, there is only 1 device shown under the 15ec USB section. This is not correct. There should be 4 items listed, not 1:

 

 

Edited August 14, 2020 by iGPU
added IORE without SSDT

[fabiosun]

DSB2 in a working x299 platform shows also hs01 and 02

 

ot 

is there a cheaper thunderbolt 3 device to buy to do more precise testing?

[meina222]

37 minutes ago, iGPU said:

 

I'd given you other re-names, but they overlapped with the AX200 eliminator SSDT. So I re-did the batch for you. I think you can safely use all of them. Note that there is a v2 of the EC-USBX SSDT file.

 

SSDT-TRX40-GB-Designare.zip 14.44 kB · 2 downloads

 

It works - thanks! I see that all controllers now have unique names in HT.

 

 

 

[tsongz]

Was able to fix my Bluetooth issues by using the aml file provided by @iGPU, and changing the removed HS0x to the one that was associated with my Asrock Creator, which was HS06. Did a nvram clean and booted into MacOS and no more AX200.

 

 

Now the only thing that's left is figuring out why MacOS won't shut down/reboot properly. I feel like it has something to do with RTC settings/flags, but haven't validated that yet. Figuring out that last step, this is a full fledged bare metal TRXintosh! Everyone here is awesome for being so helpful and responsive, and major kudos and credits to everyone involved.

 

Edit: Adding my IORE for a Asrock Creator data point

TS-iMacPro-AsrockCreator.zip

Edited August 14, 2020 by tsongz
Adding IO

[iGPU]

11 minutes ago, tsongz said:

Now the only thing that's left is figuring out why MacOS won't shut down/reboot properly. I feel like it has something to do with RTC settings/flags, but haven't validated that yet. Figuring out that last step, this is a full fledged bare metal TRXintosh! Everyone here is awesome for being so helpful and responsive, and major kudos and credits to everyone involved.

 

 

 

I've tried a couple of SSDT/patches and so far no success in preventing the shutdown reboot loop.

 

I had this issue with a Z390 build. I think I fixed it by settings in BIOS and boot loader, but that was on Intel

[tsongz]

Have confirmed that virtualization is enabled (I.e. Docker) even though the CPU flags aren't reported. Based on reports, they look at the BIOS rather than what's reported by the OS. I wasn't able to get it working in Proxmox.

 

Followed this guide, although with a few modifications.

 

Spoiler

Installing Docker on Metal

 

Prereqs: 

- Brew and xCode

 

1. Get Docker for Mac - Desktop, drag it into Applications, but don't open it. Or if you do, it'll fail, but that's fine.

2. brew cask install virtualbox

3. brew install docker-machine

4. Run to link Docker App with the functions

 

ln -s "/Applications/Docker.app/Contents//Resources/bin/docker-compose" /usr/local/bin/docker-compose
ln -s "/Applications/Docker.app/Contents//Resources/bin/docker-credential-desktop" /usr/local/bin/docker-credential-desktop
ln -s "/Applications/Docker.app/Contents//Resources/bin/docker-credential-osxkeychain" /usr/local/bin/docker-credential-osxkeychain

5. Create a default Docker Machine

 

docker-machine create --virtualbox-no-vtx-check --driver virtualbox default
docker-machine env default

6. brew install docker (for CLI docker commands)

 

And that should be it