Green Screen of Death | Radeon RX 5700 XT - OpenCore 0.6.7 - Big Sur

[tmeuze]

Hi all,

 

My 0.6.7 OC build is working almost perfectly, except I am getting graphical glitches and regular "green screen of death" issues that make my computer completely unresponsive. The glitches and short occurrences of the green screen happen mostly when switching Mission Control spaces, but eventually and increasingly frequently, an unrecoverable GSoD will happen out of the blue.  This will occur on both of my monitors (1x DP, 1x HDMI). Unplugging and plugging them back in produces the same result. USB devices such as my wireless mouse will disconnect. The only resolution is to forcefully power cycle my machine.

 

This used to happen maybe once a week, and I used to accept it as a side-effect of hackintoshing. After increasing frequency and frustration (at least 3x per day now), I upgraded from OpenCore 0.6.3 -> 0.6.7, with a brand-new config.plist, built manually word-for-word from the guide, hoping it would fix the issue. I have also tried switching to @AudioGod's EFI here, but the issue persists.

 

Anyone here experienced this or have found a fix? You would think this is a GPU hardware issue (and it still may be), but I cannot reproduce it in Windows, despite stress tests and long gameplay. Nothing interesting in the system log, and no kernel panic. Boot args are:

keepsyms=1 agdpmod=pikera ncpi=0x2000

(I added ncpi=0x2000 on recommendation from a Reddit thread about a similar issue)

 

I appreciate any help or insight with this.

[AudioGod]

Check for this issue

https://spin.atomicobject.com/2018/08/24/macbook-pro-external-monitor-display-problem/

If it’s not that then try this

https://github.com/Pavo-IM/AGPMInjector

And if your going to use the boot argument ncpi=0x2000 then make sure above 4g is disabled in the bios. Never run both.

Edited March 25, 2021 by AudioGod

[tmeuze]

17 hours ago, AudioGod said:

Check for this issue

https://spin.atomicobject.com/2018/08/24/macbook-pro-external-monitor-display-problem/

If it’s not that then try this

https://github.com/Pavo-IM/AGPMInjector

And if your going to use the boot argument ncpi=0x2000 then make sure above 4g is disabled in the bios. Never run both.

 

Thank you so much for your response. I don't believe the EDID issue applies to me. I'd never heard of AGPMInjector; I am booting with this now, and will update this thread with the results.

 

Thanks, again.

[AudioGod]

You checked to see if you had the EDID fault or you don’t think?

it’s a universal problem that happens to real Macs and Hacks so check for it if you haven’t.

Edited March 26, 2021 by AudioGod

[tmeuze]

Unfortunately, the GSoD occurred again after applying the AGPMInjector kext. Below are logs up until the crash and forced reboot.

 

Any other ideas on what could be the cause, or have I just lost the silicon lottery here? It just doesn't make sense for it not to happen in another OS.

 

Thank you,

 

system.log

Spoiler

Mar 26 13:40:09 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.04000000-0700-0000-0000-000000000000[1381]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:40:09 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.07000000-0200-0000-0000-000000000000[1392]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:40:09 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.09000000-0200-0000-0000-000000000000[1394]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:40:09 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.0D000000-0700-0000-0000-000000000000[1490]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:40:15 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.06000000-0600-0000-0000-000000000000[1386]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:42:00 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.06000000-0700-0000-0000-000000000000[1528]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:42:24 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.0E000000-0000-0000-0000-000000000000[1550]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:42:27 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.09000000-0300-0000-0000-000000000000[1589]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:42:27 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.03000000-0200-0000-0000-000000000000[1489]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:42:27 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.10000000-0000-0000-0000-000000000000[1488]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:42:27 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.0D000000-0000-0000-0000-000000000000[1529]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:43:02 Equinox Spotify Helper[1699]: Libnotify: notify_register_coalesced_registration failed with code 9 on line 2835
Mar 26 13:43:12 --- last message repeated 16 times ---
Mar 26 13:43:12 Equinox iCloudNotificationAgent[1711]: objc[1711]: Class INActivity is implemented in both /System/Library/Frameworks/Intents.framework/Versions/A/Intents and /System/Library/PrivateFrameworks/iCloudNotification.framework/iCloudNotificationAgent. One of the two will be used. Which one is undefined.
Mar 26 13:43:30 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.0E000000-0100-0000-0000-000000000000[1686]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:43:30 Equinox com.apple.xpc.launchd[1]: Coalition Cache Hit: app<application.com.lighting.huesync.watchdog.7477244.7477250(502)> [843]
Mar 26 13:43:32 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.04000000-0000-0000-0000-000000000000[1626]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:43:50 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.07000000-0300-0000-0000-000000000000[1627]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:45:10 Equinox syslogd[62]: ASL Sender Statistics
Mar 26 13:45:10 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.0C000000-0600-0000-0000-000000000000[1812]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:45:24 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.06000000-0000-0000-0000-000000000000[1817]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:45:38 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.02000000-0000-0000-0000-000000000000[1815]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:45:38 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.10000000-0200-0000-0000-000000000000[1807]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:45:38 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.07000000-0400-0000-0000-000000000000[1809]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:45:43 Equinox com.apple.xpc.launchd[1] (com.apple.mdworker.shared.0E000000-0300-0000-0000-000000000000[1808]): Service exited due to SIGKILL | sent by mds[93]
Mar 26 13:46:59 localhost bootlog[0]: BOOT_TIME 1616780819 0

 

displaypolicyd.log (if useful)

Spoiler

2021-03-26 13:35:13.005411-0400: PolicyEngine started: 7800020000000000, Time: Fri Mar 26 13:35:13 2021
2021-03-26 13:35:13.010953-0400: NOTICE: kAGDCCallbackCapability: 0x1f [ Validate FastPulse SyncLinkTrain ModeSet ScanoutResCfg]
2021-03-26 13:35:13.157894-0400: NOTICE: Vendor portmap 0x1e mst 0xe ddc 0x1e aux 0xe
2021-03-26 13:35:13.158028-0400: NOTICE: Features policyd 0x20000000 kernel 0x0
2021-03-26 13:35:13.158789-0400: NOTICE: accepted No_Media_Mode enter, Fri Mar 26 13:35:12 2021
2021-03-26 13:35:13.158806-0400: NOTICE: accepted No_Media_Mode exit, Fri Mar 26 13:35:12 2021
2021-03-26 13:35:13.158837-0400: NOTICE: doChange: addr 1.0 [7]
2021-03-26 13:35:14.518668-0400: NOTICE: Populated addr 1.0 410c:93e UUID 4f710000 order 0:0 tx 2:2 dsp 1 WL 0 MS 0 edid 256 undef 0 DSC 0
2021-03-26 13:35:14.518730-0400: NOTICE: doChange: addr 2.0 [7]
2021-03-26 13:35:14.518769-0400: NOTICE: doChange: addr 3.0 [7]
2021-03-26 13:35:14.518804-0400: NOTICE: doChange: addr 4.0 [7]
2021-03-26 13:35:14.518936-0400: NOTICE: Populated addr 4.0 4e14:99d UUID 1010101 order 0:0 tx 1:1 dsp 2 WL 0 MS 0 edid 256 undef 0 DSC 0
2021-03-26 13:35:14.518991-0400: NOTICE: doChange: addr 1.0 [7]
2021-03-26 13:35:14.526636-0400: NOTICE: Populated addr 1.0 410c:93e UUID 4f710000 order 0:0 tx 2:2 dsp 1 WL 0 MS 0 edid 256 undef 0 DSC 0
2021-03-26 13:35:14.526650-0400: NOTICE: Device addr 1.0 is same ...
2021-03-26 13:35:14.526705-0400: NOTICE: doChange: addr 2.0 [7]
2021-03-26 13:35:14.526738-0400: NOTICE: doChange: addr 3.0 [7]
2021-03-26 13:35:14.526770-0400: NOTICE: doChange: addr 4.0 [7]
2021-03-26 13:35:14.529259-0400: NOTICE: Populated addr 4.0 4e14:99d UUID 1010101 order 0:0 tx 1:1 dsp 2 WL 0 MS 0 edid 256 undef 0 DSC 0
2021-03-26 13:35:14.529272-0400: NOTICE: Device addr 4.0 is same ...
2021-03-26 13:35:14.531538-0400: NOTICE: Do online device addr 1.0 processCount 1
2021-03-26 13:35:14.532330-0400: NOTICE: Sync LinkTrain Port: 1: SST, 4 lanes @ HBR2 FEC disabled
2021-03-26 13:35:14.543164-0400: NOTICE: FB AGDC state FB[0] - ON
2021-03-26 13:35:14.543184-0400: NOTICE: inject FB[0] change
2021-03-26 13:35:14.543202-0400: NOTICE: Do online device addr 1.0 id 410c:93e UUID 4f710000 order 0:0 edid 256 undef 0 - completed
2021-03-26 13:35:14.578150-0400: NOTICE: kIOFBNotifyOnlineChange 0x0 fb[ 0 ] -> online, Fri Mar 26 13:35:13 2021
2021-03-26 13:35:14.578169-0400: NOTICE: Update display map to fb[ 0 ] - online
2021-03-26 13:35:15.141656-0400: NOTICE: Do online device addr 4.0 processCount 1
2021-03-26 13:35:15.141984-0400: NOTICE: FB AGDC state FB[1] - ON
2021-03-26 13:35:15.142003-0400: NOTICE: inject FB[1] change
2021-03-26 13:35:15.142022-0400: NOTICE: Do online device addr 4.0 id 4e14:99d UUID 1010101 order 0:0 edid 256 undef 0 - completed
2021-03-26 13:35:15.172473-0400: NOTICE: kIOFBNotifyOnlineChange 0x1 fb[ 1 ] -> online, Fri Mar 26 13:35:14 2021
2021-03-26 13:35:15.172492-0400: NOTICE: Update display map to fb[ 1 ] - online
2021-03-26 13:35:15.613369-0400: LinkUpdate: 1.0 event: Insert [0]
2021-03-26 13:35:15.613403-0400: NOTICE: Addr: 1.0 event: Insert [0] source: 0
2021-03-26 13:35:15.613414-0400: NOTICE: doChange: addr 1.0 [4]
2021-03-26 13:35:15.614281-0400: ERROR: Port 1 result 0 bitRate 20:20 laneCount 4:4 status_0_1 0 status_2_3 0  error 0
2021-03-26 13:35:15.614290-0400: NOTICE: Port non-scanned FB[ 0 ] link is bad ...
2021-03-26 13:35:15.618984-0400: NOTICE: Populated addr 1.0 410c:93e UUID 4f710000 order 0:0 tx 2:2 dsp 1 WL 0 MS 0 edid 256 undef 0 DSC 0
2021-03-26 13:35:15.619002-0400: NOTICE: Device addr 1.0 is different ...
2021-03-26 13:35:15.619010-0400: NOTICE: Do offline device addr 1.0 flags 0x0
2021-03-26 13:35:15.619019-0400: NOTICE: FB AGDC state FB[0] - OFF
2021-03-26 13:35:15.619082-0400: NOTICE: inject FB[0] change
2021-03-26 13:35:15.923350-0400: NOTICE: kIOFBNotifyOnlineChange 0x3 fb[ 0 ] -> offline, Fri Mar 26 13:35:15 2021
2021-03-26 13:35:15.923369-0400: NOTICE: Update display map to fb[ 0 ] - offline
2021-03-26 13:35:15.923377-0400: NOTICE: Release fb[ 0 ] resources
2021-03-26 13:35:15.923400-0400: NOTICE: Do LT off on port 1
2021-03-26 13:35:15.923445-0400: NOTICE: Port[1] - DP link down
2021-03-26 13:35:15.923589-0400: NOTICE: Do online device addr 1.0 processCount 1
2021-03-26 13:35:15.924802-0400: NOTICE: Sync LinkTrain Port: 1: SST, 4 lanes @ HBR2 FEC disabled
2021-03-26 13:35:15.936105-0400: NOTICE: FB AGDC state FB[0] - ON
2021-03-26 13:35:15.936261-0400: NOTICE: inject FB[0] change
2021-03-26 13:35:15.936280-0400: NOTICE: Do online device addr 1.0 id 410c:93e UUID 4f710000 order 0:0 edid 256 undef 0 - completed
2021-03-26 13:35:16.005248-0400: NOTICE: kIOFBNotifyOnlineChange 0x2 fb[ 0 ] -> online, Fri Mar 26 13:35:15 2021
2021-03-26 13:35:16.005274-0400: NOTICE: Update display map to fb[ 0 ] - online
2021-03-26 13:35:17.005590-0400: LinkUpdate: 1.0 event: Insert [0]
2021-03-26 13:35:17.005626-0400: NOTICE: Addr: 1.0 event: Insert [0] source: 0
2021-03-26 13:35:17.005637-0400: NOTICE: doChange: addr 1.0 [4]
2021-03-26 13:35:17.006503-0400: ERROR: Port 1 result 0 bitRate 20:20 laneCount 4:4 status_0_1 0 status_2_3 0  error 0
2021-03-26 13:35:17.006512-0400: NOTICE: Port non-scanned FB[ 0 ] link is bad ...
2021-03-26 13:35:17.011061-0400: NOTICE: Populated addr 1.0 410c:93e UUID 4f710000 order 0:0 tx 2:2 dsp 1 WL 0 MS 0 edid 256 undef 0 DSC 0
2021-03-26 13:35:17.011079-0400: NOTICE: Device addr 1.0 is different ...
2021-03-26 13:35:17.011087-0400: NOTICE: Do offline device addr 1.0 flags 0x0
2021-03-26 13:35:17.011096-0400: NOTICE: FB AGDC state FB[0] - OFF
2021-03-26 13:35:17.011155-0400: NOTICE: inject FB[0] change
2021-03-26 13:35:17.013078-0400: NOTICE: kIOFBNotifyOnlineChange 0x3 fb[ 0 ] -> offline, Fri Mar 26 13:35:16 2021
2021-03-26 13:35:17.013096-0400: NOTICE: Update display map to fb[ 0 ] - offline
2021-03-26 13:35:17.013103-0400: NOTICE: Release fb[ 0 ] resources
2021-03-26 13:35:17.013138-0400: NOTICE: Do LT off on port 1
2021-03-26 13:35:17.013190-0400: NOTICE: Port[1] - DP link down
2021-03-26 13:35:17.263644-0400: NOTICE: Do online device addr 1.0 processCount 1
2021-03-26 13:35:17.264425-0400: NOTICE: Sync LinkTrain Port: 1: SST, 4 lanes @ HBR2 FEC disabled
2021-03-26 13:35:17.275195-0400: NOTICE: FB AGDC state FB[0] - ON
2021-03-26 13:35:17.275217-0400: NOTICE: inject FB[0] change
2021-03-26 13:35:17.275239-0400: NOTICE: Do online device addr 1.0 id 410c:93e UUID 4f710000 order 0:0 edid 256 undef 0 - completed
2021-03-26 13:35:17.320010-0400: NOTICE: kIOFBNotifyOnlineChange 0x2 fb[ 0 ] -> online, Fri Mar 26 13:35:16 2021
2021-03-26 13:35:17.320031-0400: NOTICE: Update display map to fb[ 0 ] - online
2021-03-26 13:47:07.901074-0400: PolicyEngine started: 7800020000000000, Time: Fri Mar 26 13:47:07 2021

 

EDIT:

 

3 minutes ago, AudioGod said:

You checked to see if you had the EDID fault or you don’t think?

it’s a universal problem that happens to real Macs and Hacks so check for it if you haven’t.

 

I will review this now. I don't have any of the symptoms they mention, but I do need to investigate this regardless.

 

Thanks,

Edited March 26, 2021 by tmeuze

[tmeuze]

Alright, overrides are in-place. Confirmed that both displays are in forced-RGB mode.

 

I will update here with the results.

 

Thank you,

[AudioGod]

Clear your nvram if you haven’t already bud

[tmeuze]

Hi @AudioGod, thanks again for your help.

 

New symptom now - instead of green screens, I now just get completely blank/black screens. What I've tried so far includes:

• EDID Patching
• AGPMInjector kext
• Reseating GPU power cables
• Clearing NVRAM

I will say, it seems to be less frequent than before, so I've got that going for me. Any other possible causes or ideas on this?

 

Thank you,

[thefat32]

Have you disabled your serial port in your motherboard? I know it sounds weird, but I have black screens on boot in Big Sur with an RX 5600XT until I disabled it. In my motherboard settings it is under SuperIO Configurations.

Edited April 1, 2021 by thefat32

[zzhou]

Hi, I am using gigabyte z390 master + gigabyte 5700xt Gpu, and the green screen problem has bothered me for almost half a year. I tried almost all the methods on the Internet but to no avail

 

But I found that after flash the bios of Sapphire 5700xt, this question never appeared again

 

https://www.techpowerup.com/vgabios/216970/sapphire-rx5700xt-8192-191101

 

I hope this answer can help people solve 5700xt green screens.

[Djlild7hina]

I bought a W5500 last month and have been running into the same green screen with the only way to fix is a reboot.  It's inconsistent when it happens (sometimes once a day sometimes multiple times in a row) but a minor workaround I'm using is keeping a video open in the background.  So far it hasn't crashed in the past 4 days *crosses fingers*

[yapan4]

47 minutes ago, Djlild7hina said:

I bought a W5500 last month and have been running into the same green screen with the only way to fix is a reboot.  It's inconsistent when it happens (sometimes once a day sometimes multiple times in a row) but a minor workaround I'm using is keeping a video open in the background.  So far it hasn't crashed in the past 4 days *crosses fingers*

Hi. @Djlild7hina

How much did you pay for this video card, where available? Sorry for offtop

 

Edited April 13, 2021 by yapan4

[Slice]

1 hour ago, Djlild7hina said:

I bought a W5500 last month and have been running into the same green screen with the only way to fix is a reboot.  It's inconsistent when it happens (sometimes once a day sometimes multiple times in a row) but a minor workaround I'm using is keeping a video open in the background.  So far it hasn't crashed in the past 4 days *crosses fingers*

Such workaround means you use wrong Graphics Power Management (AGPM). It depends on bootloader, WhateverGreen and SMBIOS model.

[Djlild7hina]

1 hour ago, yapan4 said:

Hi. @Djlild7hina

How much did you pay for this video card, where available? Sorry for offtop

 

 

Got lucky and found a new one on ebay for $350

 

57 minutes ago, Slice said:

Such workaround means you use wrong Graphics Power Management (AGPM). It depends on bootloader, WhateverGreen and SMBIOS model.

 

Hm I figured that might've been the case but not sure how to go about it.

Info:

OpenCore 0.6.8

SMBIOS: MacPro7,1

WhateverGreen - enabled w/boot arg agdpmod=pikera

 

I've also tried disabling WhateverGreen and generated a AGPMInjector kext using Pavo's app.  Green screen happens with and without it enabled.  Also tried SMBIOS iMacPro1,1

Edited April 13, 2021 by Djlild7hina

[Slice]

I can make advices for Clover only.