Jump to content

WhatEverGreen Support Topic


MattsCreative
1,501 posts in this topic

Recommended Posts

On 2/1/2018 at 2:57 PM, xtddd said:

@Hackmac101 i now use 1.1.5 whatevergreen and no problems...

test it but no success

After repeated testing with different versions of WhateverGreen.kext I found that WEG version 1.1.5, 1.1.6 or 1.1.7 got blank screen from AMD RX 560 if Intel HD 4600 enabled at BIOS.

Only version 1.1.3 or 1.1.2 got normal display from both AMD RX 560 and Intel HD 4600 at 10.13.4 & 10.13.5 beta1(10.13.5 beta 2 & 3 will broke it again !).

Link to comment
Share on other sites

45 minutes ago, Hackmac101 said:

I had the same problem, I mentioned it in my previous reply. I fixed it by change connectorless ig-platform-id.

I can not fully understand what you mean in your previous post.

Can you explain more in details how to solve this problem ?

If possible step by step please .

Link to comment
Share on other sites

5 hours ago, jsl2000 said:

I can not fully understand what you mean in your previous post.

Can you explain more in details how to solve this problem ?

 If possible step by step please .

Yeah, my previous replies are only the problem, not the solution.

So the solution is to use a connectorless ig-platform-id, or just disable the iGPU in the BIOS.

If you don't know what is a connectorless ig-platform-id, follow the settings of the screenshot:

1085418917_CloverConfigurator_2018-05-0517-58-22@2x.thumb.png.8c47bb2bc730196d03ddfdf77b6b97f9.png

Leave ig-platform-id blank

download the latest IntelGraphicsFixup, put it to the same folder you have with Lilu.

IntelGraphicsFixup will implement a connectorless ig-platform-id automatically if a dGPU is present.

Link to comment
Share on other sites

4 hours ago, Hackmac101 said:

Yeah, my previous replies are only the problem, not the solution.

So the solution is to use a connectorless ig-platform-id, or just disable the iGPU in the BIOS.

If you don't know what is a connectorless ig-platform-id, follow the settings of the screenshot:

1085418917_CloverConfigurator_2018-05-0517-58-22@2x.thumb.png.8c47bb2bc730196d03ddfdf77b6b97f9.png

Leave ig-platform-id blank

download the latest IntelGraphicsFixup, put it to the same folder you have with Lilu.

IntelGraphicsFixup will implement a connectorless ig-platform-id automatically if a dGPU is present.

Thanks for your prompt reply and help.

Unfortunately either connectorless ig-platform-id by config.plist or add IntelGraphicsFixup.kext can not fix 10.13.5 beta2 or beta 3 yet.

Only disable iGPU at BIOS can login desktop from AMD RX 560 at 10.13.5 beta2 or beta3 in my hackintosh.

WEG 1.1.2 ~ 1.1.4 worked at 10.13.4 & 10.13.5 beta1.

Link to comment
Share on other sites

On 5/2/2018 at 1:12 AM, MMido said:

I would most definitely say this is a firmware bug or major driver bug, likely the later as the voltage configuration should be consistent, which it's definitely not. I don't know whether .775 (or lower for that matter) is hard-coded into the firmware or not, but it's definitely not a stable voltage for when the card is active and ready to jump to >500MHz (~25W), coupled with the fact that the core itself will gate certain parts, allowing for a drastic inrush under sudden demand...

 

Also note that at the default profile, the core will run at low, basic clocks below 200MHz, which is fine really, however in this particular state it'll constantly jump between .775 and .825, instead of just remaining at .825...

 

As of now, however, I've somewhat verified that .825V (state 1) is stable with a 4k+1200p dual display setup, so, for now, it seems that they only need to correct state 0 (the full-auto state) to have a minimum vcore of .825V. The potential cause of the hangs is not enough voltage to the display controller/SoC when 4k+ display setups are used, or just in general low loads.

 

I had a similar black screen fault,  I fixed it with locking State 1 for the ram (500mhz HBM). in Hight Serria 10.3.4 and Windows 10 x64,

With this fix the system is solid, but once I allow the switching it crashes with the strange hang, that you had.

Another difference between DP and HDMI is DP can be 4k@60Hz in my case but HDMI can only be 4k@30 Hz, that could also be a reason. The faster refresh rate forces more switching of the ram controller and then it crashes.

 

Good Luck if that fixed your Problem.

 

 

I don't fully understand what you are saying, but from what I gather, you are saying there is a problem when 4k monitors need to switch between 60Hz and 30Hz.

 

I am going to try use a single monitor which should be in 60Hz and see if that fixes the problem.

Link to comment
Share on other sites

Confirming the OS X 10.13.4 with Vega 64 still locks up on a single monitor.

 

Need to increase the power limit on the Vega 64. Is there a way to do this or can the new whatevergreen can fix this?

Edited by Raima
clarification
Link to comment
Share on other sites

On 5/19/2018 at 12:07 PM, Raima said:

Confirming the OS X 10.13.4 with Vega 64 still locks up on a single monitor.

 

Need to increase the power limit on the Vega 64. Is there a way to do this or can the new whatevergreen can fix this?

1

 

 

You can try and use this Kext originally done by@wyhtc  @okrasit.

 

Put it in 

EFI/CLOVER/kexts/Other 

 

 

and it works. without the system locks up because of the first state in these settings are more than 800mv so this should work. 

Core: 1442MHz/1075mV ,Memory: 1045MHz/1075mV,Temp target: 57,Idle fan: 700rpm.

 

 

RadeonVegaLowFansSpeed.kext.zip

Link to comment
Share on other sites

On 5/23/2018 at 3:02 AM, MMido said:

 

 

You can try and use this Kext originally done by@wyhtc  @okrasit.

 

Put it in 


EFI/CLOVER/kexts/Other 

 

 

and it works. without the system locks up because of the first state in these settings are more than 800mv so this should work. 

Core: 1442MHz/1075mV ,Memory: 1045MHz/1075mV,Temp target: 57,Idle fan: 700rpm.

 

 

RadeonVegaLowFansSpeed.kext.zip

 

Thanks, I'll give this a try. I made a bit of progress since my last post, but still came across problems.

 

My vega 64 has an after market air cooler on it with 2 x 120mm fans. After seeing a video on a RX480 with water-cooling, a youtuber noticed how much lower the power draw was. This gave me the idea to pull the power for the fans from the GPU and plug it directly into the motherboard.

 

At first I did some tests where I loaded up the GPU and nothing failed. I was ecstatic. This was the same day as an RX580 arrived in the mail as my backup plan. Come the weekend, I experienced the problem again. At first the system rebooted when I was out of the room. I didn't think too much of it as the system normally locks up. Then later that night, the system did lock up. It took much longer to do so, but it did. I decided to then give up and pop in the RX580.

 

My feeling were mixed. The RX580 was not that much slower than the Vega 64. Bruce X 5k was exporting in about 14.5 seconds with the Vega 64, while the RX580 was doing it in 16 seconds. Cinebench 15 the Vega 64 was getting about 153 fps, while the RX5808 was getting 139 fps. Geekbence 4.2.0 produced much bigger differences. Vega 64 scored 194995 while the RX 580 scored 144894.

 

Right now I've got a game in the background loading the GPU. I have the kext installed.

 

Hopefully it will run stable. The activity monitor is showing about 98% load on the Vega 64. Will report back later.

Edited by Raima
Link to comment
Share on other sites

On 5/23/2018 at 3:02 AM, MMido said:

 

 

You can try and use this Kext originally done by@wyhtc  @okrasit.

 

Put it in 


EFI/CLOVER/kexts/Other 

 

 

and it works. without the system locks up because of the first state in these settings are more than 800mv so this should work. 

Core: 1442MHz/1075mV ,Memory: 1045MHz/1075mV,Temp target: 57,Idle fan: 700rpm.

 

 

RadeonVegaLowFansSpeed.kext.zip

 

It's been 2 days of testing. No crashes so far. Looking great. You may have fixed my problem :)

Edited by Raima
Link to comment
Share on other sites

  • 3 weeks later...

Let me warn you about the state of WhateverGreen and Mojave Beta compatibility. In brief, certain functionality (including connector prioritsation) is partially incompatible and needs some work. For those in need of WhateverGreen I advise to wait for an update and stay on a current stable version of the operating system, which works great.

 

One of the known issues affecting both High Sierra and Mojave is multimonitor compatibility, and the next update will try to address it in a better way. The current solution (which disables AGDC) seems to sometimes cause graphical glitches when waking from sleep. A more proper and reliable way seems to be using a more standard AGDP disabling code with mine or Piker-Alpha's patch commonly found in NvidiaGraphicsFixup.

 

In the coming updates, for simplicity, maintainability, and unmeasurable :) boot speed improvements certain Lilu plugins dedicated for GPU patching will be merged in one, and appear as a single module.

  • Like 10
Link to comment
Share on other sites

  • 2 weeks later...

Hello, there! I'm here because I'm completely hopeless. I've been trying for a week to properly setup 2 monitors, one at DisplayPort and another at DVI with supplied VGA adapter.

On Sierra, this was extremely easy using Verde.kext but now on High Sierra (10.13.5) I'm struggling with a lot of configurations found in all the past pages and around the Internet.

With WhateverGreen.kext, DisplayPort works as expected (including sound) and so HDMI, but the DVI port sends no signal after boot (or it glitches the same image displayed over DP distorted and painted green).

 

I have this card https://www.techpowerup.com/gpudb/b653/sapphire-hd-7750-low-profile. I have already flashed a patched BIOS to enable GOP and disabled CSM.

 

I already tried to patch manually the connectors into frame buffer, using every release of Lilu and WhateverGreen, the -raddvi flag, disabling or enabling IGPU, booting into safe mode and much more. No luck.

As I said, I'm completely hopeless. If someone can help me, I'm attaching all files I think that could affect and I'll be absurdly glad.

iMac-de-Matheus:Downloads matheus$ ./redsock_bios_decoder < uefi.rom 
231P03H6.S25:
E231 VERDE PRO 1G GDDR5 64MX32                                                  
Subsystem Vendor ID: 174b
       Subsystem ID: e231
Object Header Structure Size: 246
Connector Object Table Offset: 3e
Router Object Table Offset: 0
Encoder Object Table Offset: b5
Display Path Table Offset: 12
Connector Object Id [19] which is [DISPLAY_PORT]
	encoder obj id [0x21] which is [INTERNAL_UNIPHY2 (osx txmit 0x22 [duallink 0x2] enc 0x5)] linkb: true
Connector Object Id [12] which is [HDMI_TYPE_A]
	encoder obj id [0x21] which is [INTERNAL_UNIPHY2 (osx txmit 0x12 [duallink 0x2] enc 0x4)] linkb: false
Connector Object Id [2] which is [DVI_I]
	encoder obj id [0x1e] which is [INTERNAL_UNIPHY (osx txmit 0x10 [duallink 0x0] enc 0x0)] linkb: false
Connector Object Id [2] which is [DVI_I]
	encoder obj id [0x15] which is [INTERNAL_KLDSCP_DAC1 (osx txmit 0x00 enc 0x10?)] linkb: false
iMac-de-Matheus:Downloads matheus$ ./radeon_bios_decode < uefi.rom 
ATOM BIOS Rom: 
	SubsystemVendorID: 0x174b SubsystemID: 0xe231
	IOBaseAddress: 0x0000
	Filename: 231P03H6.S25
	BIOS Bootup Message: E231 VERDE PRO 1G GDDR5 64MX32                                                  
PCI ID: 1002:683f
Connector at index 0
	Type [@offset 44902]: DisplayPort (10)
	Encoder [@offset 44906]: INTERNAL_UNIPHY2 (0x21)
	i2cid [@offset 44986]: 0x91, OSX senseid: 0x2
	HotPlugID: 4
Connector at index 1
	Type [@offset 44912]: HDMI-A (11)
	Encoder [@offset 44916]: INTERNAL_UNIPHY2 (0x21)
	i2cid [@offset 45013]: 0x94, OSX senseid: 0x5
	HotPlugID: 1
Connector at index 2
	Type [@offset 44922]: DVI-I (2)
	Encoder [@offset 44926]: INTERNAL_UNIPHY (0x1e)
	i2cid [@offset 45050]: 0x93, OSX senseid: 0x4
	HotPlugID: 5
Connector at index 3
	Type [@offset 44932]: DVI-I (2)
	Encoder [@offset 44936]: INTERNAL_KLDSCP_DAC1 (0x15)
	i2cid [@offset 45050]: 0x93, OSX senseid: 0x4
	HotPlugID: 5

 

DSDT.dsl

uefi.rom

config.plist

log_show

Link to comment
Share on other sites

2 hours ago, mths0x5f said:

Hello, there! I'm here because I'm completely hopeless. I've been trying for a week to properly setup 2 monitors, one at DisplayPort and another at DVI with supplied VGA adapter.

On Sierra, this was extremely easy using Verde.kext but now on High Sierra (10.13.5) I'm struggling with a lot of configurations found in all the past pages and around the Internet.

With WhateverGreen.kext, DisplayPort works as expected (including sound) and so HDMI, but the DVI port sends no signal after boot (or it glitches the same image displayed over DP distorted and painted green).

 

I have this card https://www.techpowerup.com/gpudb/b653/sapphire-hd-7750-low-profile. I have already flashed a patched BIOS to enable GOP and disabled CSM.

 

I already tried to patch manually the connectors into frame buffer, using every release of Lilu and WhateverGreen, the -raddvi flag, disabling or enabling IGPU, booting into safe mode and much more. No luck.

As I said, I'm completely hopeless. If someone can help me, I'm attaching all files I think that could affect and I'll be absurdly glad.


iMac-de-Matheus:Downloads matheus$ ./redsock_bios_decoder < uefi.rom 
231P03H6.S25:
E231 VERDE PRO 1G GDDR5 64MX32                                                  
Subsystem Vendor ID: 174b
       Subsystem ID: e231
Object Header Structure Size: 246
Connector Object Table Offset: 3e
Router Object Table Offset: 0
Encoder Object Table Offset: b5
Display Path Table Offset: 12
Connector Object Id [19] which is [DISPLAY_PORT]
	encoder obj id [0x21] which is [INTERNAL_UNIPHY2 (osx txmit 0x22 [duallink 0x2] enc 0x5)] linkb: true
Connector Object Id [12] which is [HDMI_TYPE_A]
	encoder obj id [0x21] which is [INTERNAL_UNIPHY2 (osx txmit 0x12 [duallink 0x2] enc 0x4)] linkb: false
Connector Object Id [2] which is [DVI_I]
	encoder obj id [0x1e] which is [INTERNAL_UNIPHY (osx txmit 0x10 [duallink 0x0] enc 0x0)] linkb: false
Connector Object Id [2] which is [DVI_I]
	encoder obj id [0x15] which is [INTERNAL_KLDSCP_DAC1 (osx txmit 0x00 enc 0x10?)] linkb: false

iMac-de-Matheus:Downloads matheus$ ./radeon_bios_decode < uefi.rom 
ATOM BIOS Rom: 
	SubsystemVendorID: 0x174b SubsystemID: 0xe231
	IOBaseAddress: 0x0000
	Filename: 231P03H6.S25
	BIOS Bootup Message: E231 VERDE PRO 1G GDDR5 64MX32                                                  
PCI ID: 1002:683f
Connector at index 0
	Type [@offset 44902]: DisplayPort (10)
	Encoder [@offset 44906]: INTERNAL_UNIPHY2 (0x21)
	i2cid [@offset 44986]: 0x91, OSX senseid: 0x2
	HotPlugID: 4
Connector at index 1
	Type [@offset 44912]: HDMI-A (11)
	Encoder [@offset 44916]: INTERNAL_UNIPHY2 (0x21)
	i2cid [@offset 45013]: 0x94, OSX senseid: 0x5
	HotPlugID: 1
Connector at index 2
	Type [@offset 44922]: DVI-I (2)
	Encoder [@offset 44926]: INTERNAL_UNIPHY (0x1e)
	i2cid [@offset 45050]: 0x93, OSX senseid: 0x4
	HotPlugID: 5
Connector at index 3
	Type [@offset 44932]: DVI-I (2)
	Encoder [@offset 44936]: INTERNAL_KLDSCP_DAC1 (0x15)
	i2cid [@offset 45050]: 0x93, OSX senseid: 0x4
	HotPlugID: 5

 

DSDT.dsl

uefi.rom

config.plist

log_show

 

I have figured that if I include this in my SSDT patch, i.e., if I reenable AGDC, I get the Apple logo on system boot. I'm starting to regain hope, patching each frame buffer config manually. But I still like to get help.

"CFG,CFG_USE_AGDC",
0x01,

 

 

Link to comment
Share on other sites

Can someone tell me if they have gotten WhateverGreen working with a Radeon 7770 card?  I have both Lilu and WhateverGreen in my other kext folder and the radpg=15  boot argument, but I keep getting a black screen after the apple logo boot bar goes about half way when trying to install High Sierra.  I'm using the DVI port but I got a black screen when I tried the HDMI port too (some people say HDMI always works).  Is there something I'm missing that might be special to a the 7xxx cards?  My motherboard is a X58 if that matters.

Link to comment
Share on other sites

Hi,

 

kern_weg.cpp


// Graphics policy patches do harm on Apple MacBooks, see:

// https://github.com/acidanthera/bugtracker/issues/260

// if (info->firmwareVendor == DeviceInfo::FirmwareVendor::Apple)

// return false;

 

For people who like me heavy tweak bios and replace "AMI" with "Apple" in DMI. (just for fun ;) )

You'll need to comment lines above and build, if you don't do that you'll get a black screen.

 

Fred

 

Edited by FredWst
Link to comment
Share on other sites

I use a KextsToPatch entry to get HDMI/DP-Audio working on my Vega56:

Name:    com.apple.driver.AppleHDAController
Find:    021030AA
Replace: 0210F8AA
Comment: Vega56 HDMI-Audio

Just find the Ven-/Dev-ID for your HDMI-Audio-Device and byte-flip the numbers

  • Like 1
Link to comment
Share on other sites

16 hours ago, al6042 said:

I use a KextsToPatch entry to get HDMI/DP-Audio working on my Vega56:


Name:    com.apple.driver.AppleHDAController
Find:    021030AA
Replace: 0210F8AA
Comment: Vega56 HDMI-Audio

Just find the Ven-/Dev-ID for your HDMI-Audio-Device and byte-flip the numbers

Did you add the hda-gfx property via DSDT or SSDT? I can't get my SSDT to work no matter what I try to name as the Devices for the pci-bridges.

Link to comment
Share on other sites

@vit9696 Just wanted to say, you are the man whatever you did with the latest comment to Lilu and Whatevergreen fixed the naming issues with my Vega 64 and added DP/HDMI audio perfectly thank you very much.

0S9gpXw.png

85kHgmg.png

S4jjgMv.png

Cgn0Swu.png

  • Like 1
Link to comment
Share on other sites

21 hours ago, al6042 said:

I use a KextsToPatch entry to get HDMI/DP-Audio working on my Vega56:


Name:    com.apple.driver.AppleHDAController
Find:    021030AA
Replace: 0210F8AA
Comment: Vega56 HDMI-Audio

Just find the Ven-/Dev-ID for your HDMI-Audio-Device and byte-flip the numbers

My Vega 56 has HDMI/DP-Audio working OOB, no need to apply any patch. Just WhateverGreen installed.

 

  • Like 2
Link to comment
Share on other sites

×
×
  • Create New...