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

[zgonda]

Hello Mr CaseySJ,
So I installed a 2.0 Titan Ridge GC card on my Gigabyte Z390 M Gaming without flashing it, but with an adequate SSDT (I guess !!!)
My LG Ultrafine 5K (Apple 1st generation) monitor is well recognized and functional, but my Samsung X5 NVMe SSD external drive (recognized in NVMe express system info and gone up on the desktop), freezes after a few seconds. This happens whether it is alone plugged into one of the Thunderbolt ports or with the LG Ultrafine also plugged in ...

An idea ?
Thank you in advance !!

PS : This Samsung X5 NVMe SSD works fine on my Gigabyte Z390 Designare with non-flashed Thunderbolt firmware...

Attachments are here :

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

@etc Can you tell me how you extracted the firmware for the Asus ThunderboltEX-3 TR ? Using CH341a programmer

www.tonymacx86.com

 

[CaseySJ]

Hello Mr CaseySJ,
So I installed a 2.0 Titan Ridge GC card on my Gigabyte Z390 M Gaming without flashing it, but with an adequate SSDT (I guess !!!)
My LG Ultrafine 5K (Apple 1st generation) monitor is well recognized and functional, but my Samsung X5 NVMe SSD external drive (recognized in NVMe express system info and gone up on the desktop), freezes after a few seconds. This happens whether it is alone plugged into one of the Thunderbolt ports or with the LG Ultrafine also plugged in ...

An idea ?
Thank you in advance !!

PS : This Samsung X5 NVMe SSD works fine on my Gigabyte Z390 Designare with non-flashed Thunderbolt firmware...

Attachments are here :

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

@etc Can you tell me how you extracted the firmware for the Asus ThunderboltEX-3 TR ? Using CH341a programmer

www.tonymacx86.com

Hello @zgonda,

Unfortunately there is not much we can do about this. We can only help you create a Thunderbolt hot plug SSDT, but after that it is up to macOS, the BIOS on your motherboard, and the firmware on the PCIe card. However, some suggestions:

• Does the Samsung X5 NVMe SSD work under Windows or Linux?
• Does the Samsung X5 NVMe SSD work if you connect it to a Thunderbolt 3 Dock and connect the Dock to your Z390 M Gaming system?

 

[zgonda]

Hello @zgonda,

Unfortunately there is not much we can do about this. We can only help you create a Thunderbolt hot plug SSDT, but after that it is up to macOS, the BIOS on your motherboard, and the firmware on the PCIe card. However, some suggestions:

• Does the Samsung X5 NVMe SSD work under Windows or Linux?
• Does the Samsung X5 NVMe SSD work if you connect it to a Thunderbolt 3 Dock and connect the Dock to your Z390 M Gaming system?

Thank you for your interest (as always) in my question. I don't have a Thunderbolt 3 Dock so I can't test.
On Windows, yes, the X5 works.
Thanks again

 

[CaseySJ]

Thank you for your interest (as always) in my question. I don't have a Thunderbolt 3 Dock so I can't test.
On Windows, yes, the X5 works.
Thanks again

Another option — if your motherboard has Thunderbolt header — is to connect GC-Titan Ridge to that header. Then check if BIOS contains a “Thunderbolt Boot” option. If so, try “Boot Once” or any of the other options.

 

[CaseySJ]

Here you go.

Interesting...

There is no parsing error for SSDT-HackinDROM.aml, and in fact the table is loaded successfully, as is SSDT-DTPG.aml. There are some ACPI parsing errors probably related to a rename of _OSI to XOSI, but those errors also occur when using the standard Thunderbolt SSDT.

The main difference between the standard SSDT and HackinDROM's version is that the standard one does not invoke DTGP method. It handles the return package by itself.

Perhaps this issue is limited to older motherboards/chipsets because we haven't seen this problem on Z370 and newer.

 

[CaseySJ]

You can find attached ZIP file with this firmware.
shasum= b302e2ab90ebfa64a04600276705675f0df57783

Methodology:
1. I only use Hex Fiend tool with UEFITool to result check

2. Replace all volume with GUID : AFDD39F1-19D7-4501-A730-CE5A27E1154B from F9i by F9j one (from 0xA70000 0xD70000 to 0xE3FFFF). This volume include new Microcodes.

3. At the end of volume 61C0F511-A691-4F54-974F-B9A42172CE53 and also the end of file, I change 32bits data from F9i by F9j one at 0xFFFFC0 address. This data is the Fixed Data address of raw file B52282EE-9B66-44B9-B1CF-7E5040F787C1 (moved because of micro codes size is bigger than previous one)

4. Now, to be flashable with Q-Flash, I have done following modifications on freeform with GUID 56E14F88-234B-4C34-B204-299670447247 (found on Win-Raid website forum):

• at address 0xFCDCEF, one byte replaced from 0x01 to 0x02 (0x63 offset address on freeform body)
• 32 bits word replaced by 0x00000000
  • address 0xFCDDE8 (0x15C offset address)
  • address 0xFCDDF8 (0x16C offset address)
  • address 0xFCDE08 (0x17C offset address)
  • address 0xFCDE18 (0x18C offset address)
  • address 0xFCDE28 (0x19C offset address)
  • address 0xFCDE38 (0x1AC offset address)
  • address 0xFCDE48 (0x1BC offset address)

All previous 32bits addresses are Firmware file adresses. Green offset address are related address if you want to check with UEFItool or to make modification on another Gigabyte firmware.

About my other customised firmware with included DSDT and SSDT, there are no benefits except to be independant on Boot loader patching with Clover or Opencore (all Mac devices and methods like MCHC and DTGP are correctly configured ) ... In fact, this a part of testing... Previously tried including many Mac DXE driver to have macOS Bootpicker (Real Macs Bootpicker) without success .. all of them a correctly loaded but no bootpicker graphically displayed.

Just updated to BIOS F9i-Elias64Fr Some observations:

• The update process went through quite quickly.
• No problems encountered.
• If the monitor supports HDR mode (High Dynamic Range), macOS will enable HDR mode again by default, so it's necessary to open System Preferences --> Displays and uncheck HDR.

 

[beelzebozo]

Interesting...

There is no parsing error for SSDT-HackinDROM.aml, and in fact the table is loaded successfully, as is SSDT-DTPG.aml. There are some ACPI parsing errors probably related to a rename of _OSI to XOSI, but those errors also occur when using the standard Thunderbolt SSDT.

The main difference between the standard SSDT and HackinDROM's version is that the standard one does not invoke DTGP method. It handles the return package by itself.

Perhaps this issue is limited to older motherboards/chipsets because we haven't seen this problem on Z370 and newer.

Figured it had to something to do with chipset generation.

Oh well, I guess I'm kind of lucky I have a SSDT that works at least . Thanks for looking.

 

[xtreem.p]

@CaseySJ Do you recommend to use BIOS F9i-Elias64Fr ... since im still on F9G ?

 

[CaseySJ]

@CaseySJ Do you recommend to use BIOS F9i-Elias64Fr ... since im still on F9G ?

The latest official BIOS for the Z390 Designare is F9j, but we don't recommend it because it breaks USB-C on Thunderbolt ports. But it does contain new Intel CPU micro code that patches some security vulnerabilities. @Elias64Fr has done us a huge favor by moving the new CPU micro code into BIOS F9i.

So now we have the best of both worlds:

• USB-C ports still work
• We get the latest CPU security patches

The decision to use the modified F9i is left to each person. As always, we make no guarantees and take no responsibility for any issues that may or may not arise. If issues do occur, one can always flash the original F9i.

 

[brousseau6933]

@Elias64Fr : I tried the Windows method with UBU (Universal BIOS Updater) but Q-Flash refused the updated BIOS. I’m glad that you succeeded and will try the modded BIOS. Kudos and thanks a lot, wonderful work!

Plus, no need to redo the BIOS settings, old BIOS settings F9i backup (on an USB key) should work as it’s the same BIOS version.