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

[CaseySJ]

I do it for the community and to break one of the last hackintosh wall

• I don't use GPIO3 Force Power because I don't have it on BIOS (only last Maximus firmware). Asus had remove it ! You should disable it for one time to see what happen. Better is to test with and without it.
• Right, but on first time, we need to have magic Thunderbolt Local node tree.
• Suggest to do all tests with NO devices connected on two thunderbolt ports >> one COLD BOOT, extracting ACPIDebug log and IoReg screenshot, >> one HOT BOOT , extracting ACPIDebug log and IoReg screenshot.
• "PCIHotplugCapable" is Zero (like Macs), probably for drivers ?!

On Z270 Maximus, UpstreamBridge device-id is 0x15D3 which is different for Macs TitanRidge and AlpineRidge (0x1578) .. Drivers seems to work well with it ... for the moment.

UPDATE: Geekbench 5.1.0 OpenCL performance with OMEN Accelerator EGPU case / Sapphire RX580 PULSE have 39077 score .. which is good ?! .. or not

UPDATE2: Screen shot of the magic button for safely extract EGPU:

View attachment 445901​

Which method or device of special MUX are you refer on SSDT ?

Just finished testing the two new SSDTs. There is good news and bad news!

Good News:

• Both SSDT-TbtOnPch-Designare_Draft05.aml (I renamed it to Draft05) and SSDT-TbtOnPch-Designare_Simplified_Draft01.aml cause the full set of Thunderbolt devices to appear in IOReg! This solves the problem of missing child nodes.

Bad news:

• The system locks up (freezes completely) after about 2-3 minutes with both SSDTs if any Thunderbolt device is connected.
  • If no Thunderbolt device is connected, then the system works reliably with both SSDTs.
• Hot plug does not work whether GPIO3 Force Power is enabled or disabled.
• No Thunderbolt Local Node or Thunderbolt Bus.

Spoiler: ACPIDebug Log for Draft05 with TB3 Device Connected

kernel: (ACPIDebug) ACPIDebug: Version 0.1.4 starting on OS X Darwin 19.2.
kernel: (ACPIDebug) ACPIDebug: 0x40039854
kernel: (ACPIDebug) ACPIDebug: 0xe00e4000
kernel: (ACPIDebug) ACPIDebug: 0xe0800000
kernel: (ACPIDebug) ACPIDebug: 0xe0900000
kernel: (ACPIDebug) ACPIDebug: 0xe0908000
kernel: (ACPIDebug) ACPIDebug: 0xe0910000
kernel: (ACPIDebug) ACPIDebug: 0xe0920000
kernel: (ACPIDebug) ACPIDebug: 0xe0a00000
kernel: (ACPIDebug) ACPIDebug: "_INI Root Port"
kernel: (ACPIDebug) ACPIDebug: "ICMBoot"
kernel: (ACPIDebug) ACPIDebug: "Tbt TWIN set to One"
kernel: (ACPIDebug) ACPIDebug: "Tbt Force Power with FPG0"
kernel: (ACPIDebug) ACPIDebug: "Up Stream VID/DID =0000000015EA8086"
kernel: (ACPIDebug) ACPIDebug: "Delay to appear = 187 us"
kernel: (ACPIDebug) ACPIDebug: "ICMS"
kernel: (ACPIDebug) ACPIDebug: "ICME_0="
kernel: (ACPIDebug) ACPIDebug: 0xffffffff
kernel: (ACPIDebug) ACPIDebug: "Configure NHI called"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI RP OK!"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI UPSB OK!"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI Dp 0 done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI Dp 3 done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI Dp 4 done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI Dp 5 done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI DPs done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI0 all done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI End"
kernel: (ACPIDebug) ACPIDebug: "ICME_1="
kernel: (ACPIDebug) ACPIDebug: 0x800001a1
kernel: (ACPIDebug) ACPIDebug: "ICME_2="
kernel: (ACPIDebug) ACPIDebug: 0x800001a1
kernel: (ACPIDebug) ACPIDebug: "ICME_3="
kernel: (ACPIDebug) ACPIDebug: 0x800001a7
kernel: (ACPIDebug) ACPIDebug: 0x800003f5
kernel: (ACPIDebug) ACPIDebug: "ICME_4="
kernel: (ACPIDebug) ACPIDebug: 0x800001a7
kernel: (ACPIDebug) ACPIDebug: "ISTA="
kernel: (ACPIDebug) ACPIDebug: 0x800003f5
kernel: (ACPIDebug) ACPIDebug: "Configuration delay = 1004 ms"
kernel: (ACPIDebug) ACPIDebug: "_L61"
kernel: (ACPIDebug) ACPIDebug: "_PS0 RP05"
kernel: (ACPIDebug) ACPIDebug: "PCEU RP05"
kernel: (ACPIDebug) ACPIDebug: "PCEU RP05- Put upstream bridge back into D0 "
kernel: (ACPIDebug) ACPIDebug: "_PS0 UPSB"
kernel: (ACPIDebug) ACPIDebug: "PCED UPSB"
kernel: (ACPIDebug) ACPIDebug: "PCED - enable GPIO"
kernel: (ACPIDebug) ACPIDebug: "UGIO - PCI wants on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - NHI wants on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - XHCI wants on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - TBT forced on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - USB forced on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - TBT GPIO should be on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - USB GPIO should be on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - Make sure TBT & USBC is on"
kernel: (ACPIDebug) ACPIDebug: "PCED UPSB- restored flag, THUNDERBOLT_PCI_LINK_MGMT_DEVICE.PRSR"
kernel: (ACPIDebug) ACPIDebug: "RTPC NHI0 called with arg: 0000000000000000"
kernel: (ACPIDebug) ACPIDebug: "_PS3 NHI0, Nothing to do!"
kernel: (ACPIDebug) ACPIDebug: "_PS3 DSB0"
kernel: (ACPIDebug) ACPIDebug: "PCDA DSB0"
kernel: (ACPIDebug) ACPIDebug: "POFF DSB0"
kernel: (ACPIDebug) ACPIDebug: "PCDA DSB0- Put upstream bridge into D3"
kernel: (ACPIDebug) ACPIDebug: "PCDA DSB0- Set link disable on upstream bridge"
kernel: (ACPIDebug) ACPIDebug: "PCDA DSB0- Wait for link to drop..."
kernel: (ACPIDebug) ACPIDebug: "PCDA DSB0 - NHI0 not present"
kernel: (ACPIDebug) ACPIDebug: "PCDA DSB0- disable GPIO"
kernel: (ACPIDebug) ACPIDebug: "UGIO - PCI wants on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - NHI wants off"
kernel: (ACPIDebug) ACPIDebug: "UGIO - XHCI wants on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - TBT allows off"
kernel: (ACPIDebug) ACPIDebug: "UGIO - USB forced on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - TBT GPIO should be off"
kernel: (ACPIDebug) ACPIDebug: "UGIO - USB GPIO should be on"
kernel: (ACPIDebug) ACPIDebug: "UGIO - Make sure TBT & USBC is on"

Spoiler: ACPIDebug Log for Simplified SSDT with TB3 device connected

kernel: (ACPIDebug) ACPIDebug: Version 0.1.4 starting on OS X Darwin 19.2.
kernel: (ACPIDebug) ACPIDebug: 0x40039854
kernel: (ACPIDebug) ACPIDebug: 0xe00e4000
kernel: (ACPIDebug) ACPIDebug: 0xe0800000
kernel: (ACPIDebug) ACPIDebug: 0xe0900000
kernel: (ACPIDebug) ACPIDebug: 0xe0908000
kernel: (ACPIDebug) ACPIDebug: 0xe0910000
kernel: (ACPIDebug) ACPIDebug: 0xe0920000
kernel: (ACPIDebug) ACPIDebug: 0xe0a00000
kernel: (ACPIDebug) ACPIDebug: "_INI Root Port"
kernel: (ACPIDebug) ACPIDebug: "ICMBoot"
kernel: (ACPIDebug) ACPIDebug: "Tbt TWIN set to One"
kernel: (ACPIDebug) ACPIDebug: "Tbt Force Power with FPG0"
kernel: (ACPIDebug) ACPIDebug: "Up Stream VID/DID =0000000015EA8086"
kernel: (ACPIDebug) ACPIDebug: "Delay to appear = 190 us"
kernel: (ACPIDebug) ACPIDebug: "ICMS"
kernel: (ACPIDebug) ACPIDebug: "ICME_0="
kernel: (ACPIDebug) ACPIDebug: 0xffffffff
kernel: (ACPIDebug) ACPIDebug: "Configure NHI called"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI RP OK!"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI UPSB OK!"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI Dp 0 done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI Dp 3 done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI Dp 4 done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI Dp 5 done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI DPs done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI0 all done"
kernel: (ACPIDebug) ACPIDebug: "Configure NHI End"
kernel: (ACPIDebug) ACPIDebug: "ICME_1="
kernel: (ACPIDebug) ACPIDebug: 0x800001a1
kernel: (ACPIDebug) ACPIDebug: "ICME_2="
kernel: (ACPIDebug) ACPIDebug: 0x800001a1
kernel: (ACPIDebug) ACPIDebug: "ICME_3="
kernel: (ACPIDebug) ACPIDebug: 0x800001a7
kernel: (ACPIDebug) ACPIDebug: 0x800003f5
kernel: (ACPIDebug) ACPIDebug: "Seem to be ICM Firmware boot or NVM Auth done!"
kernel: (ACPIDebug) ACPIDebug: "ICME_4="
kernel: (ACPIDebug) ACPIDebug: 0x800001a7
kernel: (ACPIDebug) ACPIDebug: "ISTA="
kernel: (ACPIDebug) ACPIDebug: 0x800003f5
kernel: (ACPIDebug) ACPIDebug: "Configuration delay = 1004 ms"
kernel: (ACPIDebug) ACPIDebug: "_L61"

... then 7 seconds later ...

kernel: (ACPIDebug) ACPIDebug: "RTPC NHI0 called with arg: 0000000000000000"

Thing(s) to Note:

• An _L61 event is taking place. Do we need to provide any special handling for _L61?

 

[CaseySJ]

 * PMC device has nothing to do to LPC bus, but is added to its scope for
* faster initialisation. If we add it to PCI0, where it normally exists,
* it will start in the end of PCI configuration, which is too late for
* NVRAM support.

@CaseySJ So that's probably why we failed in our attempts.

I read that paragraph and came to the same conclusion. Now if "someone" has time to try it...

 

[CaseySJ]

My system occasionally freezes (normally when I have one or more CAD programs open). I was thinking of changing the graphics card from a MSI RX 5700 XT to something else as a start? I couldn't even follow the miniguide with this board so I think it's suspect. I was thinking going to a Vega 64 or Radeon VII. Any suggestions? Thanks !

@tdcam @Anim8me2 @skroutz999 and anyone else with sudden lockups:

This sudden lockup problem seems to be growing. It would be helpful to see what your systems have in common to determine whether there's any pattern. It's important that all of these questions be answered:

• What SMBIOS name are you using? iMacPro1,1 or iMac19,1?
• What is the make/model of your GPU?
• What are your config.plist boot arguments?
• What version of Lilu and WhateverGreen are you using?
• What version of macOS are you running?
• How much RAM is in your system and what is the configuration of the DIMM modules (e.g. 8GB modules x 4)?
• When did you first notice the lockups?

 

[manfriday]

Thanks -- this is very promising! The main difference between their version and ours is that they're adding device PMCR to \_SB.PCI0.LPCB, whereas we were adding it to \_SB.PCI0.

The attached version follows the acidanthera convention and moves PMCR to \_SB.PCI0.LPCB. Anyone with OpenCore experience is welcome to try, but it requires removal of PMCR from the DSDT.aml that accompanies the OpenCore mini-guide.

EDIT:

I take it back.

The new SSDT DOES work without having to be in a DSDT.
I had just dropped it into the ACPI folder ala clover.
I forgot I needed to update my config.plist with the SSDT.

sweet!

 

[tdcam]

@tdcam @Anim8me2 @skroutz999 and anyone else with sudden lockups:

This sudden lockup problem seems to be growing. It would be helpful to see what your systems have in common to determine whether there's any pattern. It's important that all of these questions be answered:

• What SMBIOS iMacPro1,1
• What is the make/model of your GPU? msi 5700 xt mech
• What are your config.plist boot arguments? keepsyms=1 debug=0x100 dart=0 slide=0 shikigva=32 shiki-id-mac-7ba5b2d9e42ddd94 agdpmod=pikera
• What version of Lilu and WhateverGreen are you using? The newest just updated 1.3.6 /1.4.1
• What version of macOS are you running? 10.15.2
• How much RAM is (8GB modules x 2)
• When did you first notice the lockups? The last week

Edit noticed on Newegg this card has lots of crashes on Windows. https://www.newegg.com/msi-radeon-rx-5700-xt-rx-5700-xt-mech-oc/p/N82E16814137465

 

[mango1122]

I read that paragraph and came to the same conclusion. Now if "someone" has time to try it...

Works with z390 ASRock ITX. OpenCore 0.5.4.
Followed the following steps

• Copy the SSDT-PMC.aml to EFI/OC/ACPI
• Modify config.plist to add SSDT-PMC.aml ACPI -> Add
• Booter -> Quirks -> DisableVariableWrite -> NO
• NVRAM -> LegacyEnable -> NO
• Delete the the logouthook command
• Delete nvram.plist (if it exists in the root of the EFI folder)

 

[CaseySJ]

Works with z390 ASRock ITX. OpenCore 0.5.4.
Followed the following steps

• Copy the SSDT-PMC.aml to EFI/OC/ACPI
• Modify config.plist to add SSDT-PMC.aml ACPI -> Add
• Booter -> Quirks -> DisableVariableWrite -> NO
• NVRAM -> LegacyEnable -> NO
• Delete the the logouthook command
• Delete nvram.plist (if it exists in the root of the EFI folder)

This might work in Clover as well!

 

[CODYQX4]

I read that paragraph and came to the same conclusion. Now if "someone" has time to try it...

I shutdown, turned off PSU, booted, set nvram test, restarted and it worked. Works in Clover and previous testers say it works in OC.

The properties you set though don't work, so it doesn't show in PCI.

 

[jleahy2]

This sudden lockup problem seems to be growing. It would be helpful to see what your systems have in common to determine whether there's any pattern. It's important that all of these questions be answered:

• What SMBIOS name are you using? iMacPro1,1 or iMac19,1?
• What is the make/model of your GPU?
• What are your config.plist boot arguments?
• What version of Lilu and WhateverGreen are you using?
• What version of macOS are you running?
• How much RAM is in your system and what is the configuration of the DIMM modules (e.g. 8GB modules x 4)?
• When did you first notice the lockups?

I'm sad to say I experienced another one tonight, although it wasn't a complete lockup. Interestingly enough, the system became unresponsive, but the mouse still worked and started 'painting' blue lines across the screen wherever it went.

Here's the answer to your questions:

• What SMBIOS name are you using? iMacPro1,1 or iMac19,1? iMac19,1
• What is the make/model of your GPU? Sapphire Radeon Nitro+ Rx 5700 Xt
• What are your config.plist boot arguments? keepsyms=1, dart=0, debug=0x100, slide=0, agdpmod=pikera
• What version of Lilu and WhateverGreen are you using? Latest: L=1.4.1 WG=1.3.6
• What version of macOS are you running? 10.15.2
• How much RAM is in your system and what is the configuration of the DIMM modules (e.g. 8GB modules x 4)? 32gb (4x8gb DIMMs)
• When did you first notice the lockups? Saturday (first day of new build)

I shutdown, turned off PSU, booted, set nvram test, restarted and it worked. Works in Clover and previous testers say it works in OC.

The properties you set though don't work, so it doesn't show in PCI.

I tried this with my Clover setup. Removed EmuVariableUefi.efi and nvram.plist, and installed the .aml file, rebooted, and nvram looks good on my end too.

 

[willl]

Hi thanks for the guide and all the work put into this. I just built and am posting from my first hackintosh thanks to this. I'm having a problem though with my GPU. I tried to find answers in the thread first, but haven't been able to yet.

I have

• gigabyte designare z390 (firmare f8)
• i7-9700K
• Sapphire Radeon RX 5700 XT 8 GB PULSE
• 2x32GB corsair vegeance lpx
• 1tb samsung 970 evo plus (came already on newest firmware, which I found out when trying to update it)

I was able to follow the fresh Catalina mini guide (with 15-dec-2019 update) to install 10.15.2 but I had to follow the "if you only have the igpu" route to get it to work. When I tried to use config-AMD-GPU.plist as the base, the black apple progress bar screen would get to around half way and hang there forever (I tried once leaving it overnight). However, when I used the config-Internal-GPU.plist as the base, everything worked fine.

I'll also note, that I haven't been able to see any of the BIOS splash screens connecting my monitor to the sapphire card, neither HDMI, nor any of the display ports. All of the working install has been the motherboard HDMI port. I do have "initial display output " set to "PCIe 1 slot" in the bios, and power going to the card. The sapphire logo lights up if that's worth anything.

I'm downloading windows now to install on another ssd to see if the card will work there at all, but I'm not sure what to try on the Mac side in the meantime, and would appreciate any help.