Z490 & Z590 - Will Z590 ever have macOS Support ?

[vipermachine]

Today I desoldered and copied the rom (winbond 25Q80) from the Thunderbolt 4 controller from the Z590i Vision D.
I also wrote to Gigabyte about a hot-plug.

Wow that's good news @ori69

 

[dehjomz]

Today I desoldered and copied the ROM (Winbond 25Q80) from the Thunderbolt 4 controller from the Z590i Vision D.
I also wrote to Gigabyte about a hot-plug.

I wrote to them too about hot plug on the Z590 Vision D. Hopefully they are able to update the BIOS, and to do so soon. I hope it's not a controller problem as they may never be able to fix it, or maybe they can with a firmware update. With two people complaining, hopefully they realize it's a big problem with their implementation.

They also need to update to the latest Rocket Lake microcode...especially on Z490. The 0x34 microcode is bad. 13% worse multi-core performance compared to running in a Gigabyte board with the newer 0x3a microcode, and uses more power to run worse. Even the 0x3a microcode isn't the latest one available.

Asus is running the latest microcode... and their Maple Ridge Thunderbolt hotplug seemingly works. I may have to ditch Gigabyte...

 

[CaseySJ]

Today I desoldered and copied the ROM (Winbond 25Q80) from the Thunderbolt 4 controller from the Z590i Vision D.
I also wrote to Gigabyte about a hot-plug.

That was a very gutsy move! But now that you've done the deed, feel free to try the two attached modified firmware files. There is some guesswork involved in these mods so we should temper expectations. Some observations:

• The firmware chip is an 8MB Flash ROM versus 1MB for Titan Ridge
• NVM 26
• Active region at offset 0x4000 (typical)

 

[CaseySJ]

@vipermachine @ori69

If one of you can post your ACPI table set, I can modify the SSDT-TbtOnPCH to see whether we can activate Thunderbolt Bus with just an SSDT.

This is done with MaciASL as follows:

 

[vipermachine]

I wrote to them too about hot plug on the Z590 Vision D. Hopefully they are able to update the BIOS, and to do so soon. I hope it's not a controller problem as they may never be able to fix it, or maybe they can with a firmware update. With two people complaining, hopefully they realize it's a big problem with their implementation.

They also need to update to the latest Rocket Lake microcode...especially on Z490. The 0x34 microcode is bad. 13% worse multi-core performance compared to running in a Gigabyte board with the newer 0x3a microcode, and uses more power to run worse. Even the 0x3a microcode isn't the latest one available.

Asus is running the latest microcode... and their Maple Ridge Thunderbolt hotplug seemingly works. I may have to ditch Gigabyte...

There is a new BIOS available F5a

 

[vipermachine]

@vipermachine @ori69

If one of you can post your ACPI table set, I can modify the SSDT-TbtOnPCH to see whether we can activate Thunderbolt Bus with just an SSDT.

This is done with MaciASL as follows:
View attachment 513783

Here you go!

 

[dehjomz]

@vipermachine @ori69

If one of you can post your ACPI table set, I can modify the SSDT-TbtOnPCH to see whether we can activate Thunderbolt Bus with just an SSDT.

This is done with MaciASL as follows:
View attachment 513783

Hi Casey, I know you've spoken on this before but can you point me to some material on the challenges with Titan Ridge (and possibly Maple Ridge) with activating thunderbolt bus via SSDT?

There is a new BIOS available F5a

Yes, 5a on the z590 Vision D works okay with the 0x3a microcode, but microcode 0x34 is the latest available for the z490 Vision D.

 

[CaseySJ]

Hi Casey, I know you've spoken on this before but can you point me to some material on the challenges with Titan Ridge (and possibly Maple Ridge) with activating thunderbolt bus via SSDT?
...

Actually we have found through weeks of sheer experimentation (with great help from @Elias64Fr) that the Titan Ridge controller resets itself back to ICM mode (Internal Connection Manager) when we use SSDT to flip its state to OS Managed mode. The Alpine Ridge controller does not do this. I am therefore interested to find out how Maple Ridge responds when a SSDT flips its state to OS Managed mode.

 

[CaseySJ]

Here you go!

Okay, this is truly mad science, but attached is an initial attempt to activate Maple Ridge Thunderbolt Bus with just an SSDT. Some notes:

• Disable any previous Thunderbolt SSDT
• However, keep SSDT-DTPG.aml enabled
• Maple Ridge does not seem to use a typical line-level interrupt to signal a hot-plug event
  • There is no Thunderbolt-specific _GPE._Lxx method
  • However, this capability might have moved to the USB event handler AL6F in SSDT-6, which is coupled to line-level event _L6F
• The attached SSDT makes no attempt to handle hot plug events; it only attempts to put the Maple Ridge controller into OS Managed Mode

After installing the SSDT, simply reboot and post screenshot of RP05 section of IORegistryExplorer.

Probability of success for this 1st attempt: Close to 0%

 

[vipermachine]

Okay, this is truly mad science, but attached is an initial attempt to activate Maple Ridge Thunderbolt Bus with just an SSDT. Some notes:

• Disable any previous Thunderbolt SSDT
• However, keep SSDT-DTPG.aml enabled
• Maple Ridge does not seem to use a typical line-level interrupt to signal a hot-plug event
  • There is no Thunderbolt-specific _GPE._Lxx method
  • However, this capability might have moved to the USB event handler AL6F in SSDT-6, which is coupled to line-level event _L6F
• The attached SSDT makes no attempt to handle hot plug events; it only attempts to put the Maple Ridge controller into OS Managed Mode

After installing the SSDT, simply reboot and post screenshot of RP05 section of IORegistryExplorer.

Probability of success for this 1st attempt: Close to 0%

Here you go @CaseySJ