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

[AlexD]

Btw, I was in Apple Store yesterday and dumped SSDTs and IOReg from the newest Macbook Pro 2018, which uses the same Titan Ridge JHL7540 chip right on PCH Root Port 05). There's a very interesting table named TBTONPCH (which I'm sure is exactly what we need for our Titan Ridge chip which is also operates on PCH).

There's a lot going on inside, and I can see the methods that operate with switching between USB/TB modes, initialization methods, proper notifications and many many more functions that we're missing in our SSDTs.

The very interesting fact is that if in our case we have 15EA (Titan Ridge) both on UPSB and DSB0 devices, while Macbook Pro has 1578 (Alpine Ridge) on UPSB and 15EA (Titan Ridge) on DSB0. That seem to be a very strange decision. Maybe the guys at Apple know something and placed an older controller in front, to provide a better compatibility with older devices? Maybe there's an answer to my problem with Thunderbolt display? Lot of questions, though.

NHI is the same 15EB in both cases, and TB XHC is 15EC is the same both in Macbook Pro 2018 and Z390 Designare.
Though, Root Port 05 has different IDs: our is A33C and Macbook Pro's is 9DBC.

I believe if someone with a strong knowledge of ACPI/SSDT and interest in Thunderbolt technology will be able to help the community with a proper SSDT for Titan Ridge JHL7540 on PCH, that would mean finally a complete and full Thunderbolt support for all of us.

 

[xanderevo]

So basically are you saying that everything that isn't TB3 but lower won't work on this Motherboard also with an adapter? I am a little lost.

 

[CaseySJ]

Btw, I was in Apple Store yesterday and dumped SSDTs and IOReg from the newest Macbook Pro 2018, which uses the same Titan Ridge JHL7540 chip right on PCH Root Port 05). There's a very interesting table named TBTONPCH (which I'm sure is exactly what we need for our Titan Ridge chip which is also operates on PCH).

There's a lot going on inside, and I can see the methods that operate with switching between USB/TB modes, initialization methods, proper notifications and many many more functions that we're missing in our SSDTs.

The very interesting fact is that if in our case we have 15EA (Titan Ridge) both on UPSB and DSB0 devices, while Macbook Pro has 1578 (Alpine Ridge) on UPSB and 15EA (Titan Ridge) on DSB0. That seem to be a very strange decision. Maybe the guys at Apple know something and placed an older controller in front, to provide a better compatibility with older devices? Maybe there's an answer to my problem with Thunderbolt display? Lot of questions, though.

NHI is the same 15EB in both cases, and TB XHC is 15EC is the same both in Macbook Pro 2018 and Z390 Designare.
Though, Root Port 05 has different IDs: our is A33C and Macbook Pro's is 9DBC.

I believe if someone with a strong knowledge of ACPI/SSDT and interest in Thunderbolt technology will be able to help the community with a proper SSDT for Titan Ridge JHL7540 on PCH, that would mean finally a complete and full Thunderbolt support for all of us.

Awesome report...glad to see you could use Apple Labs (oops, I mean Apple Store) to get this comparative information. If you captured the entire IOReg dump, perhaps you could post it in the SSDT thread:

https://www.tonymacx86.com/threads/in-progress-ssdt-for-thunderbolt-3-hotplug.248784/

 

[CaseySJ]

So basically are you saying that everything that isn't TB3 but lower won't work on this Motherboard also with an adapter? I am a little lost.

It seems only complex devices such as TB2 display monitors may not work with an adapter. Audio devices such as the UAD Octo Satellite work fine with adapter.

 

[AlexD]

If you captured the entire IOReg dump, perhaps you could post it in the SSDT thread:

https://www.tonymacx86.com/threads/in-progress-ssdt-for-thunderbolt-3-hotplug.248784/

Of course, I'll do. Hopefully, it may help someone to figure things out.

I'm also wondering if it will be possible to connect and use LG Ultrafine 5k 27MD5KB-B display. Seems like it is a Thunderbolt Display as well (but may contain DisplayPort vis USB-C fallback, which may allow it to connect without messing with TB). Pretty expensive, though.

Btw, really love the Planar IX2790, which was probably already mentioned by @CaseySJ here. Dark, "iMac Pro" exterior with an absolutely gorgeous glass surface.

 

[xanderevo]

A modern Mac is not a Mac without Thunderbolt 3

I'd say a modern Mac is not a Mac without the startup Chime,
credits goes to "Carillon" from nicoSWD

Carillon 1.2

If you want to customize the sound just put whatever you like into /System/Library/Sounds and rename it to Chime.aiff

 

[CaseySJ]

I'd say a modern Mac is not a Mac without the startup Chime,
credits goes to "Carillon" from nicoSWD

Carillon 1.2

If you want to customize the sound just put whatever you like into /System/Library/Sounds and rename it to Chime.aiff

Ha, love it! But the party pooper in me seems to enjoy the sound of silence.

 

[CaseySJ]

Of course, I'll do. Hopefully, it may help someone to figure things out.

I'm also wondering if it will be possible to connect and use LG Ultrafine 5k 27MD5KB-B display. Seems like it is a Thunderbolt Display as well (but may contain DisplayPort vis USB-C fallback, which may allow it to connect without messing with TB). Pretty expensive, though.

Btw, really love the Planar IX2790, which was probably already mentioned by @CaseySJ here. Dark, "iMac Pro" exterior with an absolutely gorgeous glass surface.

I've been trying to better understand some of the thorny details of SSDTs and IOReg. If my understanding is correct, then:

• Some of the more complex nodes in the IOReg tree are inserted at run-time by kernel extensions/drivers.
• These kexts/drivers need to see a minimum number of "anchor" devices in the SSDT to which they attach additional nodes and sub-nodes.
• So we would need to draw a distinction between extra devices and sub-devices in boot-time SSDT versus the additional nodes inserted at run-time by the O/S.
• Can the O/S insert new devices to the SSDT at run-time, or is the SSDT frozen before the O/S starts? My guess is that the O/S may be able to modify the SSDT at run-time.
• So if we can capture the unmodified SSDT from the Macbook Pro, we can compare that against our SSDT, identify the additional devices, paths, and methods, and duplicate them.

Regarding the Planar IX2790, I too am tempted by the promise of a 5K monitor driven by a single DisplayPort 1.4 cable from the RX 580, hence bypassing Thunderbolt 3 entirely! But with 2019 already upon us, I am willing to wait for a rumored new Apple Display to accompany the modular Mac Pro. With the Hackintosh having diverted money away from Apple, I feel I should buy "something" from Apple to compensate!

 

[AlexD]

@CaseySJ Fully agree with all said above. There's definitely a subtle balance that needs to be respected. Because if we add too much stuff manually, it may potentially prevent the system from properly inject all the rest where it all should be.

I've seen a lot of various TB SSDTs, especially in another big thread on another forum. Some of them are very minimal, and some of them are way more complicated. Some guys tried to fully adapt an iMac Pro Thunderbolt SSDT, for example. Still half way there. Some devices connect perfectly, some of them not.

But none of them achieved complete Thunderbolt Local Node implementation which will mean, that Thunderbolt is fully recognized and supported by the macOS.

My belief is that if this particular table TbtOnPch is persist and hard coded for Macbook Pro 2018, it means that most of the magic stuff is done right here, in software. And properly adapting it may allow us to get a complete Thunderbolt support on our systems.

Anyway, I'm also attaching the MBP 2018 IOJones dump, as well as DSDT.aml, TBT-ON-PCH.aml, XHCI.aml files here.

 

[CaseySJ]

@CaseySJ Fully agree with all said above. There's definitely a subtle balance that needs to be respected. Because if we add too much stuff manually, it may potentially prevent the system from properly inject all the rest where it all should be.

I've seen a lot of various TB SSDTs, especially in another big thread on another forum. Some of them are very minimal, and some of them are way more complicated. Some guys tried to fully adapt an iMac Pro Thunderbolt SSDT, for example. Still half way there. Some devices connect perfectly, some of them not.

But none of them achieved complete Thunderbolt Local Node implementation which will mean, that Thunderbolt is fully recognized and supported by the macOS.

My belief is that if this particular table TbtOnPch is persist and hard coded for Macbook Pro 2018, it means that most of the magic stuff is done right here, in software. And properly adapting it may allow us to get a complete Thunderbolt support on our systems.

Anyway, I'm also attaching the MBP 2018 IOJones dump, as well as DSDT.aml, TBT-ON-PCH.aml, XHCI.aml files here.

This gives me an interesting homework assignment for tonight. I think baby steps are the way forward. If we cram too much into our SSDT at once then we no longer have a controlled experiment in which a particular cause is associated with a particular effect. The Thunderbolt-on-PCH node (TbtOnPch) does seem like a good place to start.