Contribute
Register

Gigabyte Z490 Vision D (Thunderbolt 3) + i5-10400 + AMD RX 580

Hi @CaseySJ, @tdcam, and @emistral,

I just found out another thread discussing about this topic: https://www.tonymacx86.com/threads/screen-goes-black-for-2-seconds.287053/

It indeed seems to also affect the DP connection, even though for @CaseySJ and me it's not the case. I will leave a short message there as well later on, describing my issue.

I don't have access to my hackintosh right now, but I will give a try ASAP playing with the HDMI version support in my monitor OSD. I know that there is something like this in the menu. I saw it! :think: ...
something like: HDMI 2.0 On - HDMI 2.0 Off and also for DP: DisplayPort 1.1 - DisplayPort 1.2

Speaking about my personal case (so HDMI issues), I fear that it's not possible to switch to HDMI 2.0 off while still supporting my ultra wide resolution: 3440 x 1440... but I will let you know anyway ASAP!
 
Hi @CaseySJ, @tdcam, and @emistral,

I just found out another thread discussing about this topic: https://www.tonymacx86.com/threads/screen-goes-black-for-2-seconds.287053/

It indeed seems to also affect the DP connection, even though for @CaseySJ and me it's not the case. I will leave a short message there as well later on, describing my issue.

I don't have access to my hackintosh right now, but I will give a try ASAP playing with the HDMI version support in my monitor OSD. I know that there is something like this in the menu. I saw it! :think: ...
something like: HDMI 2.0 On - HDMI 2.0 Off and also for DP: DisplayPort 1.1 - DisplayPort 1.2

Speaking about my personal case (so HDMI issues), I fear that it's not possible to switch to HDMI 2.0 off while still supporting my ultra wide resolution: 3440 x 1440... but I will let you know anyway ASAP!
My ASRock X570 Taichi Ryzentosh with RX 580 GPU is also connected to the same monitor. Will see if HDMI blank-out occurs with RX 580. My Z490 Vision D and Z390 Test Bench are connected to an AOC AGON monitor, both via HDMI and both with RX 580 -- with no blanking problem. So this might affect RX 5000 series only.
 
Hi @CaseySJ, @tdcam, and @emistral,

I just found out another thread discussing about this topic: https://www.tonymacx86.com/threads/screen-goes-black-for-2-seconds.287053/

It indeed seems to also affect the DP connection, even though for @CaseySJ and me it's not the case. I will leave a short message there as well later on, describing my issue.

I don't have access to my hackintosh right now, but I will give a try ASAP playing with the HDMI version support in my monitor OSD. I know that there is something like this in the menu. I saw it! :think: ...
something like: HDMI 2.0 On - HDMI 2.0 Off and also for DP: DisplayPort 1.1 - DisplayPort 1.2

Speaking about my personal case (so HDMI issues), I fear that it's not possible to switch to HDMI 2.0 off while still supporting my ultra wide resolution: 3440 x 1440... but I will let you know anyway ASAP!
Thanks I can try a different cable but I need 1.4 . I do think it's a 5700 issue as it does the same black screen for me when logging into windows
 
Ladies and gents,

I wanted to use my iGPU to run macOS instead of my Radeon 560, which I wanted to put into another system.
I enabled the Comet Lake iGPU patches for the framebuffer and device ID per Opencore's dortania installation guide, and found out that the iGPU output works via thunderbolt (displayport). However, whenever booting macOS with the iGPU, the system was unstable, and would crash whenever I opened Chrome/Edge, or whenever a macOS system update was being applied. That's the reason why I used the 560 in the first place... for stability.


The dortania guide doesn't talk about this..... But what I found out is for stability using the iGPU in macOS, we have to go into the Gigabyte Vision D BIOS, and set the Initial Graphics Output to the IGPU (and not the PCIe slot 1). We also have to set DVMT-Preallocated to 96 MB for macOS to fully load "Intel UHD 630 1536 MB." If we use 64 MB or lower, the iGPU graphics driver doesn't fully load, and there are artifacts in Chrome.

Once we set these properly, the IGPU works well, the system can go to sleep and awaken no problem. And best of all, no crashing. We can also disable the pcie gpu by setting a flag in opencore's boot-args: -wegnoegpu. I did this to disable my 6800XT until Apple releases a driver, and sleep/wake works well.

Hopefully this helps anyone that needs to use the iGPU instead of a dedicated GPU, especially if you were like me and experienced instability.
 
Last edited:
@andyxm715 I can confirm that deleting Audiodxe driver and disabling the PlayChime in the config.plist significantly reduces the boot time. Now between displaying the Gigabyte logo and the OC picker, it almost instantaneous.
Thanks for the tip.
 
Hi @CaseySJ, @tdcam, and @emistral,

I just found out another thread discussing about this topic: https://www.tonymacx86.com/threads/screen-goes-black-for-2-seconds.287053/

It indeed seems to also affect the DP connection, even though for @CaseySJ and me it's not the case. I will leave a short message there as well later on, describing my issue.

I don't have access to my hackintosh right now, but I will give a try ASAP playing with the HDMI version support in my monitor OSD. I know that there is something like this in the menu. I saw it! :think: ...
something like: HDMI 2.0 On - HDMI 2.0 Off and also for DP: DisplayPort 1.1 - DisplayPort 1.2

Speaking about my personal case (so HDMI issues), I fear that it's not possible to switch to HDMI 2.0 off while still supporting my ultra wide resolution: 3440 x 1440... but I will let you know anyway ASAP!
Although I have the possibility in my monitor to set up DP to 1.1 or 1.2, I can't use DP1.1 because the max resolution is 4k @ 30hz. I mean I could but given that when I am at a 4k 60hz resolution it happens only once in a while, it is not bothering me yet.
 
Ladies and gents,

I wanted to use my iGPU to run macOS instead of my Radeon 560, which I wanted to put into another system.
I enabled the Comet Lake iGPU patches for the framebuffer and device ID per Opencore's dortania installation guide, and found out that the iGPU output works via thunderbolt (displayport). However, whenever booting macOS with the iGPU, the system was unstable, and would crash whenever I opened Chrome/Edge, or whenever a macOS system update was being applied. That's the reason why I used the 560 in the first place... for stability.


The dortania guide doesn't talk about this..... But what I found out is for stability using the iGPU in macOS, we have to go into the Gigabyte Vision D BIOS, and set the Initial Graphics Output to the IGPU (and not the PCIe slot 1). We also have to set DVMT-Preallocated to 96 MB for macOS to fully load "Intel UHD 630 1536 MB." If we use 64 MB or lower, the iGPU graphics driver doesn't fully load, and there are artifacts in Chrome.

Once we set these properly, the IGPU works well, the system can go to sleep and awaken no problem. And best of all, no crashing. We can also disable the pcie gpu by setting a flag in opencore's boot-args: -wegnoegpu. I did this to disable my 6800XT until Apple releases a driver, and sleep/wake works well.

Hopefully this helps anyone that needs to use the iGPU instead of a dedicated GPU, especially if you were like me and experienced instability.
@CaseySJ hi there, has any of you guys played with disabling a GPU in slot 1 or 2 via the config.plist? I was looking on the net for a solution other than the SSDT or weg when I found a link to the OC installation guide where there was a section on disabling a GPU. On the page (https://dortania.github.io/OpenCore-Install-Guide/extras/spoof.html#windows-gpu-selection) there are 3 sections to disable the GPU; one via the -wegnoegpu, one via the config.plist, one via the SSDT and it is stated that it is supported in 0.6.7.
I don't remember seeing that page in previous version of OC. Maybe it is something new in 0.6.7.
Anyway I am interested in the config.plist solution and have tried it but not sure if it works because under system report -> graphic displays, I still see my 5600 XT and also a "display" entry which is listed as being in slot 2 which is where the 6800 XT is.
On the page, there are 3 parameters under the PCI root path of the PEG1 (which is the 6800 XT).
Name with a "data" type and a value of 23646973706C6179. Any idea what this value is and if it is a generic value or specific to PEG0 or PEG1 (the example on the page is for PEG0).
In IORegistryExplorer, under PEG1/GFX1, the Name is also of data type but with a value of <"display">. It could be well be that this is a conversion to a string of the above value.
IOName is a string type with a value of #display in the config.plist and in IO registry explorer it is of string type but with a value of "display" (without the quotes)
Finally class-code in the config.plist is of "data" type with a value of FFFFFFFF but in IORegistryexplorer it is of data type but with a value of <00 00 03 00>.

Note that when comparing GFX0 and GFX1 in IOregistryexplorer, IOName and class-code entries have exactly the same values. They only differ by Name which is ATY_GPU for GFX0 and <"display"> for GFX1.
My CPIroot is PciRoot(0x0)/Pci(0x1,0x1)/Pci(0x0,0x0)/Pci(0x0,0x0)/Pci(0x0,0x0) and matches the pciroot path given by gfxutil for GFX1 and hackintool
Any suggestion if what I am doing is right or does not sound right?
Thanks in advance.
 
Last edited:
@CaseySJ hi there, has any of you guys played with disabling a GPU in slot 1 or 2 via the config.plist? I was looking on the net for a solution other than the SSDT or weg when I found a link to the OC installation guide where there was a section on disabling a GPU. On the page (https://dortania.github.io/OpenCore-Install-Guide/extras/spoof.html#windows-gpu-selection) there are 3 sections to disable the GPU; one via the -wegnoegpu, one via the config.plist, one via the SSDT and it is stated that it is supported in 0.6.7.
The config.plist method is identical to the SSDT method. Both are used to inject the same device properties. In the Dortania guide they inject these properties:
  • name = #display
  • IOName = #display
  • class-code = 0xFFFFFFFF

I don't remember seeing that page in previous version of OC. Maybe it is something new in 0.6.7.
Anyway I am interested in the config.plist solution and have tried it but not sure if it works because under system report -> graphic displays, I still see my 5600 XT and also a "display" entry which is listed as being in slot 2 which is where the 6800 XT is.
On the page, there are 3 parameters under the PCI root path of the PEG1 (which is the 6800 XT).
Name with a "data" type and a value of 23646973706C6179. Any idea what this value is and if it is a generic value or specific to PEG0 or PEG1 (the example on the page is for PEG0).
"23646973706C6179" is just the ASCII code equivalent of "#display". A string can be specified either as (a) type String followed by the letters and symbols, or (b) type Data followed by the ASCII codes of each character of the string.

In IORegistryExplorer, under PEG1/GFX1, the Name is also of data type but with a value of <"display">. It could be well be that this is a conversion to a string of the above value.
IOName is a string type with a value of #display in the config.plist and in IO registry explorer it is of string type but with a value of "display" (without the quotes)
Finally class-code in the config.plist is of "data" type with a value of FFFFFFFF but in IORegistryexplorer it is of data type but with a value of <00 00 03 00>.

Note that when comparing GFX0 and GFX1 in IOregistryexplorer, IOName and class-code entries have exactly the same values. They only differ by Name which is ATY_GPU for GFX0 and <"display"> for GFX1.
My CPIroot is PciRoot(0x0)/Pci(0x1,0x1)/Pci(0x0,0x0)/Pci(0x0,0x0)/Pci(0x0,0x0) and matches the pciroot path given by gfxutil for GFX1 and hackintool
Any suggestion if what I am doing is right or does not sound right?
Thanks in advance.
Have you tried using the Dortania SSDT? If so, does it work?
 
@CaseySJ,

I want to ask you about Thunderbolt. What is better, Thunderbolt on the MB or PCIe Card. Which one is also good or the same because I want to buy the Vision G which does not have the Thunderbolt. The Z590 maybe has that VisonLink, but I don't want to risk it. Because in the future, I know Apple is going to release a cheaper display as the XDR, and, maybe, I will buy it if the price is good. Knowing Apple, they might release it without DP. So I want to be covered. I like the VisIon G. But,, if the PCIe Card TB is difficult to get it to work, then I'll have to buy a MB with Thunderbolt.

i appreciate your advice. So what do you think?
Much thanks in advance for you time for answering this.

 
@CaseySJ,

I want to ask you about Thunderbolt. What is better, Thunderbolt on the MB or PCIe Card. Which one is also good or the same because I want to buy the Vision G which does not have the Thunderbolt. The Z590 maybe has that VisonLink, but I don't want to risk it. Because in the future, I know Apple is going to release a cheaper display as the XDR, and, maybe, I will buy it if the price is good. Knowing Apple, they might release it without DP. So I want to be covered. I like the VisIon G. But,, if the PCIe Card TB is difficult to get it to work, then I'll have to buy a MB with Thunderbolt.

i appreciate your advice. So what do you think?
Much thanks in advance for you time for answering this.

This is very easy to answer:
  • If you plan to flash the Thunderbolt controller, you will get significantly more reliable Thunderbolt performance with an on-board Thunderbolt controller.
  • If you don't plan to flash the controller, an add-in PCIe card will work just fine.
 
  • Like
Reactions: jd6
Back
Top