[SUCCESS] Gigabyte Designare Z390 (Thunderbolt 3) + i7-9700K + AMD RX 580

[CaseySJ]

Well @CaseySJ Sleep is finally working!

Just to be clear that water cooling on the cpu will continue to work after this or not?

I found some interesting reading here in the forum, I have a h115i Corsair water cooler, that USB port (Internal HS11) was preventing sleep (finally found that after 1 year).

The Water cooling will continue to work, those are the steps I did:

-connect everything as normal
-adjust the settings in the corsair app on windows (or use the bios to configure the fan curves)
-disable the port HS11 (I also have leds on it but whatever)

Now I stumbled upon this app on GitHub, maybe I can make it work but needs some testing (same as liquidctl)

GitHub - audiohacked/OpenCorsairLink: Linux and Mac OS support for the CorsairLink Devices

Linux and Mac OS support for the CorsairLink Devices - GitHub - audiohacked/OpenCorsairLink: Linux and Mac OS support for the CorsairLink Devices

github.com

still working on those audio pops

liquidctl also supports the Corsair H115i, but of course this requires reactivating HS11, which in turn will reactivate insomnia.

 

[skwebby]

good luck! the guide in post #1 is excellent, as are all the other mini-guides Casey has created. my strong advice is READ THROUGH the guide in its entirety before starting the installation. then follow every step to the letter. it will work.

Great - I have read, reread, re-reread!

 

[DazzboDuck]

Great - I have read, reread, re-reread!

The only advice I would offer is to try and get your head around Open Core, things for me just feels more compatible than Clover, mainly little differences but it seems to run nicer.....good luck

 

[Elias64Fr]

@Elias64Fr,

Bonjour! Here's a summary of yesterday's action items and their outcomes:

Next Action Items:

• Replace ThunderboltConfig with the string you provided above.
  • As long as SSP1/SSP2 --> UsbCPortNumber matches the last 4 bytes in ThunderboltConfig, TNODE/TBUS will appear (but TNODE/TBUS depends also on detecting USB cable state change, etc.). I've tried successfully with both 01000200 and 03000400.
• Boot without ThunderboltDROM and observe outcome.
  • (deferred to later today)
• Remove Sleep() from DBG1 and DBG2. Reduce ACPIDebug.kext polling interval instead.
  • Polling reduced to 50ms from default of 100ms. Removing Sleep() from DBG1/DBG2 seemed to miss a few entries in ACPIDebug log, probably because 50ms polling interval is still too long. So for now I've put back the Sleep() statements.
• Uncomment the AVND logging lines from the end of ICMB and see if AVND values remain valid.
  • When AVND is logged in ICMB, something bad happens, causing subsequent NHI0/XHC2 PCED methods to see AVND = 0xFFFFFFFF.
  • Attached log shows results with and without AVND logging, with relevant lines in blue.
• Show XHC2 AVND from PCED.
  • I modified all PCED methods to display AVND, and those values look good.
  • Even XHC2 AVND finally looks correct every time!
• Fully disable ASPM in BIOS.
  • Not sure I did this correctly. I've disabled ASPM in BIOS, but pci-aspm-default values for UPSB, DSB0, and NHI0 were still non-zero yesterday evening.
  • All of the pci-aspm-default values are 0x0 when TNODE/TBUS is not present, but they are non-zero when TNODE/TBUS is present.
    • Update: They are sometimes 0x0 and sometimes non-zero. Trying to determine the conditions...
• Test DisplayPort pass-through via Platform ID 0x3E9B0007.
  • TNODE/TBUS appeared one time yesterday evening, but the system was booted with 0x3E980003.
  • Will try again.
• Add modified _L6F handler.
  • All tests yesterday evening were with this change.
  • Did not see any _L6F events in the log.

Salut @CaseySJ

OK for test results !

About polling time, I have changed default value to 10ms.

About ASPM, I currently work on CNHI methods modifications to have correct configuration (including ASPM) of differents registers of Root Port, UpStreamBridge, DownStreamBridge, XHC2 and NHI0.This include Link Control Register.
With that, we should have a configuration closest to rMac and less depending on BIOS setting (ASPM, Common clock and Cache Line Size at once, then more later .. with some brain wave .. more setting). Have tested a part of this and It work like a French charm

In order to verify your TR registers configuration, when you will back at home or start some test, could you send me LSPCI files from DarwinDumper app with a thunderbolt device connected ?

Merci d'avance !

 

[mazop89]

Hello @CaseySJ , i followed your guide for Configure Carbon Copy Cloner for One-Key Bootable Backup for catalina users. But I don't understand the "important note" inside the STEP 2. My file is inside my downloads, I unzipped it but I don't understand what I have to do with the terminal and "cd". Each time I try something, for example copying the path of the file and typing the line that you've wrote ("xattr -d com.apple.quarantine EFIClone-v4.sh"), the Terminal says that there's no file. So when I follow the next steps, at the end there's nothing inside the EFI. Could you just explain tis step please ?

 

[CaseySJ]

Hello @CaseySJ , i followed your guide for Configure Carbon Copy Cloner for One-Key Bootable Backup for catalina users. But I don't understand the "important note" inside the STEP 2. My file is inside my downloads, I unzipped it but I don't understand what I have to do with the terminal and "cd". Each time I try something, for example copying the path of the file and typing the line that you've wrote ("xattr -d com.apple.quarantine EFIClone-v4.sh"), the Terminal says that there's no file. So when I follow the next steps, at the end there's nothing inside the EFI. Could you just explain tis step please ?

Hello @mazop89,

If the file is in your Downloads folder, simply type this in Terminal:

cd ~/Downloads
xattr -d com.apple.quarantine EFIClone-v4.sh

Also remember that TEST_SWITCH is set to Y, which puts the script in test mode. Once you have examined EFIClone.log from a test run and verified that it looks okay (i.e. only the files from Source EFI are being copied to Destination EFI), then change TEST_SWITCH to N.

 

[CaseySJ]

Salut @CaseySJ

OK for test results !

About polling time, I have changed default value to 10ms.

About ASPM, I currently work on CNHI methods modifications to have correct configuration (including ASPM) of differents registers of Root Port, UpStreamBridge, DownStreamBridge, XHC2 and NHI0.This include Link Control Register.
With that, we should have a configuration closest to rMac and less depending on BIOS setting (ASPM, Common clock and Cache Line Size at once, then more later .. with some brain wave .. more setting). Have tested a part of this and It work like a French charm

Magnifique!

In order to verify your TR registers configuration, when you will back at home or start some test, could you send me LSPCI files from DarwinDumper app with a thunderbolt device connected ?

Certainly -- will send LSPCI files with Thunderbolt device connected.

I'll also post some BIOS screenshots to make sure there aren't other ASPM settings I might have overlooked.

I'm also thinking of calling RP05._PS0() before calling RP05.UPSB._PS0(). Just want to see if that makes any difference. Currently, the sequence is:

• _INI --> PINI --> ICMB --> UPSB._PS0()
• Then macOS calls RP05._PS0(). But this should ideally be done before UPSB._PS0().
• Then macOS does not call UPSB._PS0()
• Then macOS calls DSB0._PS0()
• Etc.

So I would like to check this:

• _INI --> PINI --> ICMB --> RP05._PS0() and then UPSB._PS0()
• Then I want to see what macOS does...
  • Will it still call RP05._PS0()? We'll see...

 

[CaseySJ]

Great - I have read, reread, re-reread!

We fully expect your next post to be a carbon copy of this post!

 

[zzmadd]

Hi guys,

I have an issue that's probably due to graphic card: RX 5700XT PowerColor.
If I drop in a RX 570 4GB all runs fine. With the RX 5700XT if I run the Unigine Valley benchmark I get a reboot straightaway.
I'm still using EmuVariable + OsxAptioFix2Drv-free2000
I have agdpmod=pikera to avoid framebuffer black screen.
I'm on 10.15.3
Is there a way to fix this issue?

thanks.

 

[CaseySJ]

Hi guys,

I have an issue that's probably due to graphic card: RX 5700XT PowerColor.
If I drop in a RX 570 4GB all runs fine. With the RX 5700XT if I run the Unigine Valley benchmark I get a reboot straightaway.
I'm still using EmuVariable + OsxAptioFix2Drv-free2000
I have agdpmod=pikera to avoid framebuffer black screen.
I'm on 10.15.3
Is there a way to fix this issue?

thanks.

Please do not quote the full build guide. It says so right there at the top...

Next, see the Warning for RX 5700 and RX 5700 XT Owners also at the top. Are you sure the reboot happens only when running Unigine Valley Benchmark, and nothing else causes a reboot?