Contribute
Register

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

There might be something important I need to address in the Catalina Mini-Guide so let me ask some very specific questions:
  1. With the config-Internal-GPU.plist file in the Catalina Mini-Guide, the Catalina installer was not able to start up fully? It got hung up shortly after gIOScreenLockState?
  2. With the config.plist created by UniBeast and subsequently modified by following the various spoilers in Post #1, the Catalina installer was successfully able to start?
If the answers to both questions are YES, then can I ask you to try the following experiment:
  • Open the config.plist you created from the Catalina Mini-Guide(not from Mojave) and just make one change:
    • Clover Configurator --> Graphics --> Inject Intel --> check this ON.
  • Then save the config.plist into CLOVER folder of Catalina USB install disk (in the EFI partition).
    • Boot from the USB install disk.
    • Does the Catalina installer start up fully?
    • Don't worry, the installer will not make any changes to your installed system.
    • If the installer starts up, just go to Apple menu and click Restart.

YES to 1. and 2.
But no, it does not change. Installer does not start up. Hangs with gIOScreenlockState 3


I got Catalina up and running with the config and the kexts from the initial guide combined with a Catalina USB Installer as described above but need to tweak a lot now. iGpu gives me headaches. When i change the fake id to 0x0, i get a black screen. If do the Framebuffer patching I get a kernel panic... jezz...
 
I've attached my Clover folder to my original post, and, as requested, a screen shot of the drivers folder. I'm on a PC at the office, but I have access to a Mac if I need to get anything off the boot stick which I brought with me today :)

Again, much thanks!

Edit: I just noticed that second ._osxaptiofix2drv file. I didn't see that when I had the stick loaded in my Mac at home. Weird.
Yes, let's remove that extraneous file and try again...

UPDATE: Just downloaded the CLOVER folder you posted and that extraneous file is not present. So it might be something that Windows temporarily created?? In macOS Finder we can press CMD-Shift-. to toggle on/off the display of hidden files. We can also type ls -al in Terminal after cd to the drivers64UEFI directory. In both cases, there is no extra file.

Some suggestions:
  • If the extraneous file is indeed present, remove it.
  • If this alone does not fix the problem, try performing a CMOS Reset by shorting 2 designated pins on motherboard with a flathead screwdriver. Please follow the procedure in Designare Owner's Manual.
    • BIOS parameters will be reset to factory defaults, so re-configure them.
 
Last edited:
Yes, let's remove that extraneous file and try again...

UPDATE: Just downloaded the CLOVER folder you posted and that extraneous file is not present. So it might be something that Windows temporarily created?? In macOS Finder we can press CMD-Shift-. to toggle on/off the display of hidden files. We can also type ls -al in Terminal after cd to the drivers64UEFI directory. In both cases, there is no extra file.
Will try again and report back later on. Thanks again!
 
I got my system up and running using the Mini guide and everything seems to work (running Radeon VII in headless mode for QuickSync) except for good sleep. The computer keeps waking up on its own during the night, then going to sleep and waking up again and so on. Wake on LAN, Proximitywake and PowerNap is disabled, darkwake=0. If I look for wake reasons in logs, I get a lot of "Wake reason: RTC (Alarm)" during the night. Testing it during the day, doesn't work, since the system sleeps just fine and can't wait several hours to see when it starts the wake-sleep-wake cycle.

I didn't install RC scripts as per mini guide, but maybe I should have?
Yes, this is the multi-day sleep problem again. Is it necessary to keep the system asleep for long periods? I'm not aware of a specific solution, but there is an RTC Patch that you may try -- no guarantees...
  • Find* [HEX] = A00A9353 54415301
  • Replace* [HEX] = A00A910A FF0BFFFF
  • Comment = Fix 300-series RTC Bug
This is a single entry in Clover Configurator --> ACPI section. The result would look like this:
Screen Shot 2019-12-09 at 8.47.15 AM.png
 
Ok, I'll try this tonight and report back. I don't need the system asleep for long periods, but I do need it to sleep throughout the night since it's in the same room where I sleep and the constant waking up wakes me up too.

Yes, this is the multi-day sleep problem again. Is it necessary to keep the system asleep for long periods? I'm not aware of a specific solution, but there is an RTC Patch that you may try -- no guarantees...
  • Find* [HEX] = A00A9353 54415301
  • Replace* [HEX] = A00A910A FF0BFFFF
  • Comment = Fix 300-series RTC Bug
This is a single entry in Clover Configurator --> ACPI section. The result would look like this:
View attachment 439568
 
I have some trailer that definitely is protected.

In this documentation for Shiki, a downloadable trailer is linked. I believe it's FairPlay 1.0. Depending on your config you can't play this video. I noticed in the past that if you play remotely, it gets blocked out unlike other videos (which lag bad because VNC is not meant for video, but they play).

I did find some good results though.

I tested with iGPU on and off with an iMac19,1 SMBIOS and WEG 1.3.5.

TEST 1 = Protected (FairPlay 1.0?) movie trailer file in QuickTime from Shiki manual
TEST 2 = Amazon Prime Trailer (Safari)
TEST 3 = Amazon Prime Trailer (Firefox)
TEST 4 = Apple TV+ 7 day trial TV show (not movie)

From this, it would be great to get feedback on Netflix and Apple TV+ movies. The Apple TV+ exclusive shows were free to watch with the trial but not the movies so I didn't test those.

iGPU Shiki Config Test 1 Test 2 Test 3 Test 4
Enabled shikigva=16 Pass Pass Pass Pass
Enabled shikigva=80 Fail Fail Pass Fail
Enabled shikigva=144 Fail Pass Pass Pass
Enabled shikigva=32 shiki-id=Mac-7BA5B2D9E42DDD94 Pass Fail Pass Pass
Enabled shikigva=160 shiki-id=Mac-7BA5B2D9E42DDD94 Fail Fail Pass Pass
Enabled None Pass Pass Pass Fail
Disabled shikigva=16 Pass Fail Pass Pass
Disabled shikigva=80 Pass Fail Pass Fail
Disabled shikigva=144 Pass Fail Pass Pass
Disabled shikigva=32 shiki-id=Mac-7BA5B2D9E42DDD94 Pass Fail Pass Pass
Disabled shikigva=160 shiki-id=Mac-7BA5B2D9E42DDD94 Pass Fail Pass Pass
Disabled None Pass Fail Pass Pass


The only case where everything I tested worked was headless iGPU enabled with shikigva=16 and no shiki-id spoof. Nothing got Safari to play the Amazon trailer with iGPU off.

Some commits newer than the latest released WEG as well as this chart could use some further testing.

Apparently hardware and software decoding have been separated out because software decoding patches were breaking hardware decoding.

According to that chart, the best case scenario would to be to have iGPU enabled, be on an SMBIOS that has an iGPU (such as iMac19,1), and to use shikigva=80.

In my case, going from 16 to 80 just broke things. I compiled the latest WEG and will see later today if that fixes anything.

In addition, because from what I hear, Amazon Prime movies may use different protection for trailers as opposed to movies, movies should be tested. I also got a 30 day trial to Netflix to see if that works.
 
Some commits newer than the latest released WEG as well as this chart could use some further testing.

Apparently hardware and software decoding have been separated out because software decoding patches were breaking hardware decoding.

According to that chart, the best case scenario would to be to have iGPU enabled, be on an SMBIOS that has an iGPU (such as iMac19,1), and to use shikigva=80.

In my case, going from 16 to 80 just broke things. I compiled the latest WEG and will see later today if that fixes anything.

In addition, because from what I hear, Amazon Prime movies may use different protection for trailers as opposed to movies, movies should be tested. I also got a 30 day trial to Netflix to see if that works.
The DRM Chart at the acidanthera GitHub site is very helpful! I never saw that until now. It does state that IM/MM (iMac and Mac mini) SMBIOS require iGPU to be headless (platform ID 0x3E980003).
 
The DRM Chart at the acidanthera GitHub site is very helpful! I never saw that until now. It does state that IM/MM (iMac and Mac mini) SMBIOS require iGPU to be headless (platform ID 0x3E980003).

I currently do. I found in my last testing that not having an iGPU made Amazon Prime trailers fail in all tested cases.

I have had my iGPU headless ever since I got my RX 580. From research it sounds like RX 580 has issues especially on Catalina and results vary across makes of RX 580s.

Since there are so many makes of RX 580, and they likely matter, I note that I am specifically using a Sapphire Pulse RX 580 8GB 11265-05-20G. I currently have it in slot 2 now with nothing in slot 1. Just moved it when I moved the BCM4360 around because I kept losing bluetooth functionality after waking from sleep.
 
Yes, let's remove that extraneous file and try again...

UPDATE: Just downloaded the CLOVER folder you posted and that extraneous file is not present. So it might be something that Windows temporarily created?? In macOS Finder we can press CMD-Shift-. to toggle on/off the display of hidden files. We can also type ls -al in Terminal after cd to the drivers64UEFI directory. In both cases, there is no extra file.

Some suggestions:
  • If the extraneous file is indeed present, remove it.
  • If this alone does not fix the problem, try performing a CMOS Reset by shorting 2 designated pins on motherboard with a flathead screwdriver. Please follow the procedure in Designare Owner's Manual.
    • BIOS parameters will be reset to factory defaults, so re-configure them.
Ok so i shorted the cmos. Wiped clean. Reset all the parameters in bios. Still no good. Stuck at end
Randomseed.
I tried removing the video card and using the onboard (was able to boot and see everything, but no better)

Should i try removing one of the memory sticks. —edit: tried this. Didnt have any effect.
 
Last edited:
@CODYQX4 could you share the latest lilu and weg you compiled? I do like to test them but I cant compile them myself. AFAIK they both have been updated a few days ago.

thank you
 
Back
Top