WhatEverGreen Support Topic

vit9696 · July 18, 2019

@Sherlocks, the only idea I might have about this is Xcode beta. Try building with Xcode 10.

The only functional change that may affect SKL+ IGPUs is modeset change. Try finding your CPU (e.g. SKL) and change

forceCompleteModeset = true;

to

forceCompleteModeset = false;

e.g. here https://github.com/acidanthera/WhateverGreen/blob/master/WhateverGreen/kern_igfx.cpp#L97 (it is duplicated per CPU).

Andres ZeroCross · July 18, 2019

1 hour ago, vandroiy2012 said:

@vit9696 implemented this patches in WhateverGreen.kext. Changes already in master. https://github.com/acidanthera/WhateverGreen/commit/bcd3f2ef73266f4a320ccfd0686bef2a969b2bde

Here is prebuild binary WhateverGreen-1.3.1-DEBUG.zip

Anyone welcome to test it.

I just compiled latest commit,, Still working good,

15 minutes ago, Sherlocks said:

can you upload latest lilu + WEG?
I built lilu + WEG with xcode 11 beta + catalina beta3. no luck all.

나의 SM-N960N 의 Tapatalk에서 보냄

Would you test this,, i just compiled it with XCode 10.2.1

Archive.zip

Sherlocks · July 18, 2019

[mention=980913]Sherlocks[/mention], the only idea I might have about this is Xcode beta. Try building with Xcode 10.
The only functional change that may affect SKL+ IGPUs is modeset change. Try finding your CPU (e.g. SKL) and change
forceCompleteModeset = true;
to
forceCompleteModeset = false;
e.g. here https://github.com/acidanthera/WhateverGreen/blob/master/WhateverGreen/kern_igfx.cpp#L97 (it is duplicated per CPU).

i'm using Version 11.0 beta 2 (11M337n) with catalina beta4

downloaded source from github, and built lilu debug version, then put it in WEG source folder, and built WEG. then test.

strangely WEG was not load when booting. is this xcode issue?

can you build weg(skylake cpu with forceCompleteModeset = false option)?

I just compiled latest commit,, Still working good,

Would you test this,, i just compiled it with XCode 10.2.1
Archive.zip

no success.

EDIT1.

i'm downloading xcode 11 beta4.

EDIT2

i can't build lilu with all plugins kext on xcode 11 beta4.

Edited July 18, 2019 by Sherlocks

Sherlocks · July 19, 2019

@vit9696

here is v1.3.1 log on skylake

2019-07-19 22:21:41.361365+0900 localhost kernel[0]: (kernel) WhateverGreen: igfx @ (DBG) hwregs multimonitor lookup failure, trying second
2019-07-19 22:21:41.361385+0900 localhost kernel[0]: (kernel) WhateverGreen: igfx @ (DBG) routed hwregs multimonitor
2019-07-19 22:21:41.361414+0900 localhost kernel[0]: (kernel) WhateverGreen: igfx @ (DBG) platform is snb 0 and list 0xFFFFFF7F82A4C850
2019-07-19 22:21:41.363497+0900 localhost kernel[0]: (kernel) WhateverGreen: igfx @ (DBG) found GuC accel config at 0xFFFFFF7F834F6BA5
2019-07-19 22:21:45.843051+0900 localhost kernel[0]: (kernel) WhateverGreen: weg @ (DBG) failed to obtain display mode
2019-07-19 22:21:45.848295+0900 localhost kernel[0]: (kernel) WhateverGreen: weg @ (DBG) failed to obtain display mode
2019-07-19 22:21:45.853671+0900 localhost kernel[0]: (kernel) WhateverGreen: weg @ (DBG) failed to obtain display mode
2019-07-19 22:21:45.951283+0900 localhost kernel[0]: (kernel) WhateverGreen: weg @ (DBG) fb info 1: -2147479552:0 7680:0:32
2019-07-19 22:21:45.956956+0900 localhost kernel[0]: (kernel) WhateverGreen: weg @ (DBG) fb info 2: 3:8 --------RRRRRRRRGGGGGGGGBBBBBBBB 0:1920:1080
2019-07-19 22:21:45.964962+0900 localhost kernel[0]: (kernel) WhateverGreen: weg @ (DBG) attempting to copy...

Edited July 19, 2019 by Sherlocks

uglyJoe · July 20, 2019

@vit9696

Thank you for fixing the purple lines at boot on AMD cards!

TheBloke · July 20, 2019

48 minutes ago, uglyJoe said:

@vit9696

Thank you for fixing the purple lines at boot on AMD cards!

Yes, thank you from me too.

And for me it was a much more important fix: since 10.14.5, I usually could not get a picture on any monitor from boot. I use an AMD 7970Ghz (R9 280X) with 5 monitors, booting on a Legacy Boot system. At most I only get a picture on two monitors from boot, and then I need to sleep & wake to get a picture on all of them. But since 10.14.5, I usually got a black screen on all monitors. This made it very hard to do the sleep - especially because, for reasons I don't understand, sometimes my keyboard also does not work immediately after boot! So I can't just do <Tab> then <Space> to activate the Sleep icon on the login screen. I usually had to unplug/replug a couple of monitors after boot, which would cause macOS to eventually show a picture on the main screen, so I could click Sleep.

Anyway, since installing the 1.3.1 build uploaded earlier, I am back to how it was with 10.14.1 - no purple lines, and I always get a picture on two monitors after boot. So I can see the login screen, and I can click Sleep. Then after wake, I get picture on all monitors as usual.

Thanks for all your ongoing work to keep my old system still running Mojave successfully!

Sherlocks · July 20, 2019

i really wonder WEG Build.

my process is here

1. download lastest lilu source in github.

2. then open it on xcode 11 beta4.

3. press build button l> icon. then go to the lilu-master/DerivedData/Lilu/Build/Products/Debug/Lilu-1.3.8-DEBUG.zip

4. extract this file. copy Lilu-1.3.8.kext to WhateverGreen-master/root folder.

5. open it on xcode11 beta4. then press build button l> icon. go to the WhateverGreen-master/DerivedData/WhateverGreen/Build/Products/Debug/WhateverGreen.kext

6. copy it into clover's other folder.

7. test.

why i can never load whatever green kext? my process wrong? i never experienced this issue before.

seems force modeset causes problem. i need to check this issue. and want to report it.

after build lilu and plugins on xcode 11 beta4, when use it, i can't never lilu kext with plugins.

if i use this script(https://github.com/corpnewt/Lilu-and-Friends), i can use built kexts by this script.

strangely i can't use kexts built by xcode.

also latest whatevergreen never boot on my system. always cause force reboot.

Edited July 20, 2019 by Sherlocks

vandroiy2012 · July 20, 2019

Your problem is in Xcode 11.

There is no compatibility with this Xcode version for now.

Sherlocks · July 20, 2019

7 minutes ago, vandroiy2012 said:

Your problem is in Xcode 11.

There is no compatibility with this Xcode version for now.

seems alright. now if use https://github.com/corpnewt/Lilu-and-Friends, i can use kexts.

but i want to know problem that i reported yesterday.

latest 1.3.1 whatevergreen have problem for me. v1.3.0 is everything great.

can you custom build and upload WEG?

https://github.com/acidanthera/WhateverGreen/blob/1512ff9186e58b4de48725dbed80d75fbdb37859/WhateverGreen/kern_igfx.cpp#L97

forceCompleteModeset.enable = true;

to

//forceCompleteModeset.enable = true;

seems this line causes various problem for me.

now i don't have any way for build and test.

Edited July 20, 2019 by Sherlocks

jsl2000 · July 21, 2019

On 7/18/2019 at 9:49 PM, vandroiy2012 said:

@vit9696 implemented this patches in WhateverGreen.kext. Changes already in master. https://github.com/acidanthera/WhateverGreen/commit/bcd3f2ef73266f4a320ccfd0686bef2a969b2bde

Here is prebuild binary WhateverGreen-1.3.1-DEBUG.zip

Anyone welcome to test it.

Thanks for this WEG 1.3.1 which can remove previous purple glitches during booting before login desktop with AMD RX 570 & R9 380X at 10.14.6 & 10.15 beta.

Only minimal trace of one multi-colored line remains from DP display port which was not so obviously now.

Completely normal from HDMI display port.

Edited July 22, 2019 by jsl2000

MacNB · July 22, 2019

On 7/18/2019 at 11:37 AM, vladie said:
for those who have problems with ATIController::TestVRAM crashes during boot, i've created a patch for clover for solving this problem. second patch removes magenta-colored lines which appears at screen during video driver startup.
	    <dict>
		<key>Comment</key>
		<string>ATIController TestVRAM boot failure patch</string>
		<key>Disabled</key>
		<false/>
		<key>Find</key>
		<data>
		uapVqlU7DJA=
		</data>
		<key>InfoPlistPatch</key>
		<false/>
		<key>Name</key>
		<string>AMDSupport</string>
		<key>Replace</key>
		<data>
		uapVqlU5yZA=
		</data>
	    </dict>
	    <dict>
		<key>Comment</key>
		<string>ATIController TestVRAM magenta-colored lines patch</string>
		<key>Disabled</key>
		<false/>
		<key>Find</key>
		<data>
		xwSxqlWqVQ==
		</data>
		<key>InfoPlistPatch</key>
		<false/>
		<key>Name</key>
		<string>AMDSupport</string>
		<key>Replace</key>
		<data>
		kJCQkJCQkA==
		</data>
	    </dict>
		
Insert this is KextsToPatch section.

PS: patch is beta version, tested only with 10.14.5 + Vega64, but must be compatible with any AMD card.

TestVRAM fills memory with 0xaa55aa55 bytes (it's magenta pixels you see at screen), and it's tries to test too much of address space, writing this bytes to framebuffer memory. If you have verbose mode enabled (as i have), there are possibility that framebuffer will scroll during this test, and TestVRAM will write 0xaa55aa55, but read 0x00000000 (because magenta-colored pixels are scrolled up).

So, does the second patch DISABLE the TestVRAM completely or limit the the test to the correct address space ?

Regi Yassin · July 23, 2019

On 7/18/2019 at 5:37 PM, vladie said:
for those who have problems with ATIController::TestVRAM crashes during boot, i've created a patch for clover for solving this problem. second patch removes magenta-colored lines which appears at screen during video driver startup.
	    <dict>
		<key>Comment</key>
		<string>ATIController TestVRAM boot failure patch</string>
		<key>Disabled</key>
		<false/>
		<key>Find</key>
		<data>
		uapVqlU7DJA=
		</data>
		<key>InfoPlistPatch</key>
		<false/>
		<key>Name</key>
		<string>AMDSupport</string>
		<key>Replace</key>
		<data>
		uapVqlU5yZA=
		</data>
	    </dict>
	    <dict>
		<key>Comment</key>
		<string>ATIController TestVRAM magenta-colored lines patch</string>
		<key>Disabled</key>
		<false/>
		<key>Find</key>
		<data>
		xwSxqlWqVQ==
		</data>
		<key>InfoPlistPatch</key>
		<false/>
		<key>Name</key>
		<string>AMDSupport</string>
		<key>Replace</key>
		<data>
		kJCQkJCQkA==
		</data>
	    </dict>
		
Insert this is KextsToPatch section.

PS: patch is beta version, tested only with 10.14.5 + Vega64, but must be compatible with any AMD card.

TestVRAM fills memory with 0xaa55aa55 bytes (it's magenta pixels you see at screen), and it's tries to test too much of address space, writing this bytes to framebuffer memory. If you have verbose mode enabled (as i have), there are possibility that framebuffer will scroll during this test, and TestVRAM will write 0xaa55aa55, but read 0x00000000 (because magenta-colored pixels are scrolled up).

this is no longer work with 10.14.6 Mojave, can you update the patch?

but its working with WEG, with this commit

Edited July 23, 2019 by Regi Yassin

hohoho · July 24, 2019

I would like to report that Whatevergreen is not working properly in my hack using Catalina beta (19A512f). My hack will successfully boot and then the Apple logo will just flash very fast and then it goes into a black screen with only the backlight turned on. I can't see the login screen, it's just black with the backlight turned on. But my hack is working perfectly because i can use a remote desktop app and i can see that MacOS is loaded properly and i can use it, but the only problem is that the screen is just black with the backlight turned on. Nothing is showing in my hack's screen. -wegoff fixed the problem. thanks.

[update] the latest whatevergreen with commit (Fix regression introduced in previous commit) fixed my black screen at startup problem. so no need for -wegoff anymore. thanks.

Edited July 31, 2019 by hohoho
problem fixed

Enrize · July 25, 2019

Thanks for fixing purple lines! However there is one last small glitch that's still remains. I use MSI Aero ITX RX560.

jsl2000 · July 25, 2019

3 hours ago, Enrize said:

Thanks for fixing purple lines! However there is one last small glitch that's still remains. I use MSI Aero ITX RX560.

It may disappear after repeated booting or connecting to HDMI port only.

That's what I have observed in my Z97 hackintosh.

Andres ZeroCross · July 26, 2019

@vit9696 with current latest Commit WhateverGreen + SSDT-PNLF.aml from Whatevergreen, i got problem ini HP HP Pavilion 14-BF002TX (KabyLake / IHD 620),, black screen when boot. Revert back to version 1.3.0 from Release version or delete SSDT-PNLF method are the solution. I am with macOS Mojave 10.14.6. Just for report.

Edited July 26, 2019 by Andres ZeroCross

brousseau6933 · July 26, 2019

On 7/18/2019 at 6:37 AM, vladie said:
for those who have problems with ATIController::TestVRAM crashes during boot, i've created a patch for clover for solving this problem. second patch removes magenta-colored lines which appears at screen during video driver startup.
	    <dict>
		<key>Comment</key>
		<string>ATIController TestVRAM boot failure patch</string>
		<key>Disabled</key>
		<false/>
		<key>Find</key>
		<data>
		uapVqlU7DJA=
		</data>
		<key>InfoPlistPatch</key>
		<false/>
		<key>Name</key>
		<string>AMDSupport</string>
		<key>Replace</key>
		<data>
		uapVqlU5yZA=
		</data>
	    </dict>
	    <dict>
		<key>Comment</key>
		<string>ATIController TestVRAM magenta-colored lines patch</string>
		<key>Disabled</key>
		<false/>
		<key>Find</key>
		<data>
		xwSxqlWqVQ==
		</data>
		<key>InfoPlistPatch</key>
		<false/>
		<key>Name</key>
		<string>AMDSupport</string>
		<key>Replace</key>
		<data>
		kJCQkJCQkA==
		</data>
	    </dict>
		
Insert this is KextsToPatch section.

PS: patch is beta version, tested only with 10.14.5 + Vega64, but must be compatible with any AMD card.

TestVRAM fills memory with 0xaa55aa55 bytes (it's magenta pixels you see at screen), and it's tries to test too much of address space, writing this bytes to framebuffer memory. If you have verbose mode enabled (as i have), there are possibility that framebuffer will scroll during this test, and TestVRAM will write 0xaa55aa55, but read 0x00000000 (because magenta-colored pixels are scrolled up).

Thanks for the patch Vladie, working well in 10.14.6!

telepati · July 26, 2019

which black screen patch have inside of the WEG?

braindefender · July 27, 2019

Hi, cannot get HiDPI (4k+) resolution to work in macOS Catalina 10.15 Beta.
Only 2560x1440 and lower resolutions are accessible.
Seems like enable-hdmi20 nor -cdfon boot arguments don't work.
Acceleration is present. I think I need patch CoreDisplay, but patch for Mojave doesn't work.
I have Intel 8600K (UHD 630 with 0x3E9B0007 platform-id) by the way. Any workaround?

Edited July 29, 2019 by braindefender

braindefender · July 29, 2019

On 7/27/2019 at 4:16 PM, braindefender said:

Hi, cannot get HiDPI (4k+) resolution to work in macOS Catalina 10.15 Beta.
Only 2560x1440 and lower resolutions are accessible.
Seems like enable-hdmi20 nor -cdfon boot arguments don't work.
Acceleration is present. I think I need patch CoreDisplay, but patch for Mojave doesn't work.
I have Intel 8600K (UHD 630 with 0x3E9B0007 platform-id) by the way. Any workaround?

Managed to get 4K with acceleration working by fake platform-id with Kaby Lake 0x59120000, but 4k is very laggish and slow at 200% interface size.
No lags at all while using 100%.

brousseau6933 · July 30, 2019

On 7/26/2019 at 4:58 PM, telepati said:

which black screen patch have inside of the WEG?

https://github.com/acidanthera/WhateverGreen/commit/bcd3f2ef73266f4a320ccfd0686bef2a969b2bde

telepati · July 30, 2019

7 hours ago, brousseau6933 said:

https://github.com/acidanthera/WhateverGreen/commit/bcd3f2ef73266f4a320ccfd0686bef2a969b2bde

This is not black screen patch.

tech8086 · July 31, 2019

On 7/20/2019 at 7:44 PM, jsl2000 said:

Thanks for this WEG 1.3.1 .....

Only minimal trace of one multi-colored line remains from DP display port which was not so obviously now.

Same behavior here. Purple lines now gone, multi-colored line still displays.

brousseau6933 · August 1, 2019

On 7/30/2019 at 5:18 AM, telepati said:

This is not black screen patch.

Sorry, misread...

brousseau6933 · August 1, 2019

On 7/22/2019 at 11:09 PM, Regi Yassin said:

this is no longer work with 10.14.6 Mojave, can you update the patch?

but its working with WEG, with this commit

Like I said before Regi, working fine for me without Whatevergreen on 10.14.6. Z97, i7 4790, RX580, iMac 14,2 here...

WhatEverGreen Support Topic

Recommended Posts

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites