Gigabyte B550 Vision D (Thunderbolt 3) + AMD Ryzen 7 3700X + AMD RX 5600 XT

[cccip]

Thanks a lot folks.
1. Vision D-P differences (from Vision D) quite simple - one of blue USB3.2 ports from the back plate moved to mobo front to become internal USB-C front panel header and 2xIntel i211 1Gb controllers replaced with one 2.5Gb Intel i225 + one 2.5Gb Realtek RTL8125. Rest is identical.

2. Casey, I'm using Intel AX210 WiFi/BT module. Mobo been shipped with AX201 (sleep/wake issues were the same), did a slight upgrade.
3. In case we are out of ideas on what else might be causing it - suggest we leave it alone, not critical at all

 

[CaseySJ]

Thanks a lot folks.
1. Vision D-P differences (from Vision D) quite simple - one of blue USB3.2 ports from the back plate moved to mobo front to become internal USB-C front panel header ...

Oh, this means we need to modify the USB port map. Let me check this later today.

View attachment 568302
2. Casey, I'm using Intel AX210 WiFi/BT module. Mobo been shipped with AX201 (sleep/wake issues were the same), did a slight upgrade.
3. In case we are out of ideas on what else might be causing it - suggest we leave it alone, not critical at all

If using Intel WiFi/BT we should also use IntelBTPatcher.kext — If I recall correctly, this is needed in Ventura and newer.

If Intel Bluetooth is working in Monterey then adding IntelBTPatcher.kext is still a good idea for future upgrade.

 

[CaseySJ]

@cccip,

Please connect a USB-C device to the front panel USB-C port, check if it connects, then run IORegistryExplorer and post the IOReg file (File -> Save As...).

Also let me know what device you connected to front panel USB-C port.

I'm trying to modify the USB port map for B550 Vision D-P.

 

[cccip]

Hi Casey,
Used USB-C to lightning cable to connect "iPhone" (exact device name).
Here is the port map & IOReg export:

As for IntelBTPatcher.kext - "All of the features are already merged into IntelBluetoothFirmware project" which is been used in my setup.

 

[CaseySJ]

Hi Casey,
Used USB-C to lightning cable to connect "iPhone" (exact device name).
Here is the port map & IOReg export:

This shows that the USB port map is a bit more different than I expected. To map the ports correctly the following needs to be done:

1. iPhone is a USB 2.0 device; we can see that it connects to POT5 (front panel USB-C port). Now connect a USB 3.0 device to the front panel USB-C port such as a USB 3.x flash drive and check whether it attaches to POT1 or elsewhere. We can use IORegistryExplorer for this and the steps below.
2. Connect a USB 3.x device to the GREEN port in diagram below and let us know which port it connects to.
3. Connect a USB 3.x device to the RED port in diagram below and let us know which port it connects to.
4. Connect a USB 3.x device to the BLUE port in diagram below and let us know which port it connects to.

As for IntelBTPatcher.kext - "All of the features are already merged into IntelBluetoothFirmware project" which is been used in my setup.

This is a common misunderstanding due to the wording they used. IntelBTPatcher.kext is a separate kext. It has been added to the IntelBluetoothFirmware project, which is different from IntelBluetoothFirmware.kext.

If you download the ZIP file from their GitHub page, you'll find these three individual files inside:

But your config.plist does not include IntelBTPatcher.kext:

 

[cccip]

At last it makes some sense Casey, thanks.

1. Don't have a USB-C devices, except M1 MBP - plugged it to front USB-C and it is landing to XHC2/PTXH POT5 (i know it is weir to use M1 MacBook as USB3 flash disk, but better than nothing):

2. GREEN port USB3 disk lands at XHC/XHC0 PRT6

3. RED port USB3 disk lands at XHC PRT5 only:

4. BLUE port USB3 disk lands at XHC2 POT2 only:

5. Loading IntelBTPatcher did the trick and properly enabled BT portion of AX210 adapter.

 

[CaseySJ]

@cccip,

Although we haven't verified all of the USB ports, from the sample we did verify it seems that Gigabyte simply moved one port from the rear to the front and remapped a second port.

Let's try the attached modified USB port map files:

• First make a backup copy of the entire EFI folder. You may copy that folder to Documents or elsewhere.
• Then replace the following files with the attached versions:
  • EFI/OC/ACPI folder
    • Replace SSDT-B550-PTXH-TO-XHC2.aml with attached version
  • EFI/OC/Kexts folder
    • Replace USBPorts-XHC-XHC2.kext with attached version
• Reboot
• Check sleep/wake (keep expectations low!)

 

[cccip]

No joy with them...

2023-06-28 00:49:58 +0500 Sleep                   Entering Sleep state due to 'Software Sleep pid=186':TCPKeepAlive=active Using AC (Charge:0%) 19 secs
2023-06-28 00:50:17 +0500 DarkWake                DarkWake from Normal Sleep [CDN] : due to PT21 PT22 PT23/ Using AC (Charge:0%) 25 secs
2023-06-28 00:50:42 +0500 Wake                    DarkWake to FullWake from Normal Sleep [CDNVA] : due to HID Activity Using AC (Charge:0%)

Perhaps there is a way to disable PT21/PT22/PT23 entirely in ACPI?

 

[CaseySJ]

No joy with them...

2023-06-28 00:49:58 +0500 Sleep                   Entering Sleep state due to 'Software Sleep pid=186':TCPKeepAlive=active Using AC (Charge:0%) 19 secs
2023-06-28 00:50:17 +0500 DarkWake                DarkWake from Normal Sleep [CDN] : due to PT21 PT22 PT23/ Using AC (Charge:0%) 25 secs
2023-06-28 00:50:42 +0500 Wake                    DarkWake to FullWake from Normal Sleep [CDNVA] : due to HID Activity Using AC (Charge:0%)

Perhaps there is a way to disable PT21/PT22/PT23 entirely in ACPI?

These log messages are notoriously cryptic. We don't have to disable these in Vision D. Because the log line containing PT21, PT22 and PT23 says DarkWake from Normal Sleep, that is actually okay. The log entry we need to watch out for is the one containing the word FullWake.

If you disable all three of these and reboot, does it sleep and wake properly?

• ACPI:
  • SSDT-B550-PTXH-TO-XHC2.aml
  • SSDT-B550-XHC0-TO-XHC.aml
• Kernel:
  • USBPorts-XHC-XHC2.kext

 

[cccip]

These log messages are notoriously cryptic. We don't have to disable these in Vision D. Because the log line containing PT21, PT22 and PT23 says DarkWake from Normal Sleep, that is actually okay. The log entry we need to watch out for is the one containing the word FullWake.

If you disable all three of these and reboot, does it sleep and wake properly?

• ACPI:
  • SSDT-B550-PTXH-TO-XHC2.aml
  • SSDT-B550-XHC0-TO-XHC.aml
• Kernel:
  • USBPorts-XHC-XHC2.kext

Hi Casey,

Disabling all 3 does not make any difference, still complaining due to PT21 PT22 PT23