Tahoe - Intel UDH 630 - dual monitor problems

[deeveedee]

@fermento Attached are new config.plists (based on config-test5-4.plist) as follows:

• config-test5-9.plist: framebuffer-conX-flags = 0x0147, add igfxfw=2
• config-test5-10.plist: framebuffer-conX-flags = 0x01E7, add igfxfw=2
• config-test6-1.plist: disable Kernel > Add > AdvancedMap.kext, Add boot-arg igfxonlnfbs=0x00, add igfxfw=2
• config-test6-2.plist: disable Kernel > Add > AdvancedMap.kext, Add boot-arg igfxonlnfbs=0x01, add igfxfw=2
• config-test6-3.plist: disable Kernel > Add > AdvancedMap.kext, Add boot-arg igfxonlnfbs=0x02, add igfxfw=2

 

EDiT: information about this hack when using the attached config.plists:

• con0: Gigabyte M27Q Display (DP -> DP)

• con1: Dell E248WFP Display (HDMI -> DVI-D)

• framebuffer-portcount = 2

config-test5-9.plist.zip config-test5-10.plist.zip config-test6-1.plist.zip config-test6-2.plist.zip config-test6-3.plist.zip

Edited January 13 by deeveedee
Fixed typos

[fermento]

@deeveedee

 

Tested the configs

 

• config-test5-9 -> not able to reach the login
• all the rest -> flickering issues

[deeveedee]

@fermento We're now venturing into undocumented, highly experimental territory.  This is why most people give up when attempting framebuffer patching.  This next test is going to require more work and more patience for you.  I spent some time creating the table below that defines the framebuffer-flags that we haven't tested.  Since you have already tested the candidate framebuffer ig-platform-ids for Coffee Lake / Comet Lake, you have already tested many framebuffer-flags combinations (maybe without realizing that you did).  My table below shows the framebuffer-flags bits that we did not test (because they were the same and did not change for all framebuffer ig-platform-ids).

 

Framebuffer-Flags to be tested

Spoiler

1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 0 0 1 0 AAPL,ig-platform-ID: 0x3e9b0007, framebuffer-Flags 0x00801302
1 0 0 0 0 0 1 1 0 0 0 0 1 0 1 1 0 0 0 0 1 0 1 0 ID: 0x3EA50009, Flags: 0x00830B0A
1 0 0 0 0 0 1 1 0 0 0 1 0 0 1 1 0 0 0 0 1 0 1 0 ID: 0x3E920009, Flags: 0x0083130A
1 0 0 0 0 0 1 1 0 0 0 1 0 0 1 1 0 0 0 0 1 0 1 0 ID: 0x3E9B0009, Flags: 0x0083130A
0 0 0 0 0 0 1 1 0 0 0 0 1 0 1 1 0 0 0 0 1 0 1 1 ID: 0x3EA50000, Flags: 0x00030B0B
0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 1 0 1 1 ID: 0x3E920000, Flags: 0x0000130B
0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 1 0 1 1 ID: 0x3E000000, Flags: 0x0000130B
0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 1 0 1 1 ID: 0x3E9B0000, Flags: 0x0000130B
1 1 1 0 0 0 1 1 0 0 0 0 1 0 1 1 0 0 0 0 1 0 1 0 ID: 0x3EA50004, Flags: 0x00E30B0A
1 1 1 0 0 0 1 1 0 0 0 0 1 0 1 1 0 0 0 0 1 0 1 0 ID: 0x3EA50005, Flags: 0x00E30B0A
1 1 1 0 0 0 1 1 0 0 0 0 1 0 1 1 0 0 0 0 1 0 1 0 ID: 0x3EA60005, Flags: 0x00E30B0A
1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 0 1 1 0 ID: 0x3e9b0007 with FBFramebufferCompression, Flags 0x00801306
1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 1 0 0 1 0 ID: 0x3e9b0007 with FBDynamicFBCEnable, Flags 0x00801312
1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 1 0 0 0 1 0 ID: 0x3e9b0007 with FBUseVideoTurbo, Flags 0x00801322
1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 1 0 0 0 0 1 0 ID: 0x3e9b0007 with FBForcePowerAlwaysConnected, Flags 0x00801342
1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 1 0 0 0 0 0 1 0 ID: 0x3e9b0007 with FBDisableHighBitrateMode2, Flags 0x00801382
1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 without FBBoostPixelFrequencyLimit, Flags 0x00801202
1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 without FBLimit4KSourceSize, Flags 0x00801102
1 0 0 0 0 0 0 0 0 0 0 1 0 1 1 1 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 with FBAlternatePWMIncrement1, Flags 0x00801702
1 0 0 0 0 0 0 0 0 0 1 1 0 0 1 1 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 with FBUnknownFlag_2000, Flags 0x00803302
1 0 0 0 0 0 0 0 0 1 0 1 0 0 1 1 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 with FBAllowConnectorRecover, Flags 0x00805302
1 0 0 0 0 0 0 0 1 0 0 1 0 0 1 1 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 with FBUnknownFlag_8000, Flags 0x00809302
1 0 0 0 0 1 0 0 0 0 0 1 0 0 1 1 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 with FBDisableGFMPPFM, Flags 0x00841302
1 0 0 0 1 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 with FBUnknownFlag_80000, Flags 0x00881302
1 0 0 1 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 0 0 1 0 ID: 0x3e9b0007 with FBUnknownFlag_100000, Flags 0x00901302
| | | | | | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | | | | | +- FBAvoidFastLinkTraining
| | | | | | | | | | | | | | | | | | | | | | +--- FBFramebufferCommonMemory
| | | | | | | | | | | | | | | | | | | | | +----- FBFramebufferCompression
| | | | | | | | | | | | | | | | | | | | +------- FBEnableSliceFeatures
| | | | | | | | | | | | | | | | | | | +--------- FBDynamicFBCEnable
| | | | | | | | | | | | | | | | | | +----------- FBUseVideoTurbo
| | | | | | | | | | | | | | | | | +------------- FBForcePowerAlwaysConnected
| | | | | | | | | | | | | | | | +--------------- FBDisableHighBitrateMode2
| | | | | | | | | | | | | | | +----------------- FBBoostPixelFrequencyLimit
| | | | | | | | | | | | | | +------------------- FBLimit4KSourceSize
| | | | | | | | | | | | | +--------------------- FBAlternatePWMIncrement1
| | | | | | | | | | | | +----------------------- FBAlternatePWMIncrement2
| | | | | | | | | | | +------------------------- FBDisableFeatureIPS
| | | | | | | | | | +--------------------------- FBUnknownFlag_2000
| | | | | | | | | +----------------------------- FBAllowConnectorRecover
| | | | | | | | +------------------------------- FBUnknownFlag_8000
| | | | | | | +--------------------------------- FBUnknownFlag_10000
| | | | | | +----------------------------------- FBUnknownFlag_20000
| | | | | +------------------------------------- FBDisableGFMPPFM
| | | | +--------------------------------------- FBUnknownFlag_80000
| | | +----------------------------------------- FBUnknownFlag_100000
| | +------------------------------------------- FBEnableDynamicCDCLK
| +--------------------------------------------- FBUnknownFlag_400000
+----------------------------------------------- FBSupport5KSourceSize

framebuffer flags to test
FBFramebufferCompression: 1
FBDynamicFBCEnable: 1
FBUseVideoTurbo: 1
FBForcePowerAlwaysConnected: 1
FBDisableHighBitrateMode2: 1
FBBoostPixelFrequencyLimit: 0
FBLimit4KSourceSize: 0
FBAlternatePWMIncrement1: 1
FBUnknownFlag_2000: 1
FBAllowConnectorRecover: 1
FBUnknownFlag_8000: 1
FBDisableGFMPPFM: 1
FBUnknownFlag_80000: 1
FBUnknownFlag_100000: 1

 

 

 The attached config.plists based on config-test5-4.plist give you a starting point for your new tests as follows:

• config-test7-0.plist: define framebuffer-flags to be unchanged from the default framebuffer-flags for ig-platform-id 0x3e9b0007.  Test with this config.plist first.  If it fails for any reason, report the failure and do not test any further.
• config-test7-1.plist: set framebuffer-flags bit FBFramebufferCompression
• config-test7-2.plist: set framebuffer-flags bit FBDynamicFBCEnable

After you test the config.plists that I created for you, if the flickering problem is not fixed, you will need to test the additional framebuffer-flags bits:

Spoiler

• FBUseVideoTurbo: 1
• FBForcePowerAlwaysConnected: 1
• FBDisableHighBitrateMode2: 1
• FBBoostPixelFrequencyLimit: 0
• FBLimit4KSourceSize: 0
• FBAlternatePWMIncrement1: 1
• FBUnknownFlag_2000: 1
• FBAllowConnectorRecover: 1
• FBUnknownFlag_8000: 1
• FBDisableGFMPPFM: 1
• FBUnknownFlag_80000: 1
• FBUnknownFlag_100000: 1

 

You will test these additional framebuffer-flags bits by editing the config.plist DeviceProperty framebuffer-flags.  I have inserted comments into the config.plist DeviceProperties for each framebuffer-flags bit to be tested, so you can simply copy and paste to perform each test.  For example, to test the FBUseVideoTurbo framebuffer-flags bit, copy the data value from #framebuffer-flags-FBUseVideoTurbo and paste it into the data value for framebuffer-flags.  Then test the resulting new config.plist.

 

Let me know if you have any questions.  Unfortunately, I still can't guarantee any success from these tests.  This is the nature of framebuffer patching experimentation.  When you report your test results, indicate any behavioral differences that you observe for each framebuffer-flags bit.  You don't need to post any screenshots unless you think they are necessary to explain a behavior.  Good luck.

config-test7-0.plist.zip config-test7-1.plist.zip config-test7-2.plist.zip

Edited January 16 by deeveedee
Removed stray text from bad copy/paste & removed unnecessary text (no change to framebuffer-flags)

[deeveedee]

@fermento My HDMI->DVI-D adapter arrived ahead of schedule.  Unfortunately, it doesn't work at all in Windows or macOS (just lots of display flickering and static).  Since it doesn't work in Windows, I'm not going to attempt to fix it in macOS.  Sorry my own experimentation won't help, but I'm happy to continue offering suggestions for your testing.

Edited January 15 by deeveedee
Changed after testing adapter

[fermento]

Hi! I'm back, I was not able to continue before. I just tested the latest 3 configs.

 

config-test7-0 -> flickering

config-test7-1 -> flickering

config-test7-2 -> both displays black, not able to access the system

 

I will try the additional frame buffer-flags bits during the week and give feedback.

[deeveedee]

I received a new HDMI->DVI-D adapter (Tripp Lite 8-inch HDMI-M to DVI-D Cable Adapter (M/F), 8-in. (P132-08N) 8" HDDVI Black).  It works with my HackBookPro16,2.  My sample size is small (only tested two adapters), but this suggests for me a few things:

• HDMI->DVI-D can work just fine with macOS (currently testing with Tahoe 26.3)
• HDMI->DVI-D adapter brand does matter.  One of my purchased adapters did not work at all.
• With the Tripp Lite adapter, HDMI->DVI-D works when booting with display connected at boot and when hot-plugged after booting macOS.

 

EDIT: external display connected via HDMI->DVI-D sleeps and wakes without issues

Edited February 6 by deeveedee

[deeveedee]

I think I might know why we did not observe any difference in HDMI port behavior when we added lspcon DeviceProperties.  According to this, 

 

Starting from version 1.3.0, WEG now provides driver support for the onboard LSPCON by automatically configuring the adapter to run in PCON mode on new HDMI connections, and hence solves the black screen issue on some platforms.

 

 

[deeveedee]

@verdazil You may be right.  I have a laptop that has an HDMI 2.0 port with UHD630 and I don't need to configure LSPCON in order for the HDMI 2.0 port to work perfectly.  I am going to test this laptop with WhateverGreen.kext DEGUG build so that I can observe the debug log output.  I'll report my findings which I suspect will show that PCON mode has been enabled automatically.

[deeveedee]

13 hours ago, verdazil said:

@deeveedee, I understand it a little differently.

 

I think you are correct.  Without DeviceProperty "enable-lspcon-support" I don't see any debug log entries to indicate that WEG is attempting to detect LSPCON support.  Maybe this comment below in the WEG documentation simply means that when LSPCON  support is enabled, WEG assumes PCON mode unless specifically configured for LS mode.

 

Starting from version 1.3.0, WEG now provides driver support for the onboard LSPCON by automatically configuring the adapter to run in PCON mode on new HDMI connections, and hence solves the black screen issue on some platforms.

 

 

After reviewing the debug log on my laptop, I did confirm that WEG does not detect support for LSPCON, which is why adding LSPCON configuration to my laptop's DeviceProperties does not make any difference.

 

 

[deeveedee]

@fermento I have been experimenting with framebuffer-flags on my hack with HDMI->DVI (it's a laptop with mobile framebuffer, which is different from yours, but it still allows me to experiment).  I am not finding any benefit from changing framebuffer-flags on my laptop.  If you haven't found any of the framebuffer-flags here that improve the "flickering" problem, I think that the next step would be to examine your WhateverGreen DEBUG log.

[deeveedee]

@fermento Long time, no talk.  I finally had some time to document my successful HDMI patching on my laptop.  The patching strategy implements some of the techniques discussed in your thread here.  If you're still struggling with your HDMI->DVI, it may help to look at this thread since my patch works with HDMI->DVI adapter.

 

EDIT: Note that I have multiple DVI displays.  I am finding that graphics quality on the DVI displays varies from one to the next.  On one of my DVI displays, the image when connected via HDMI->DVI from my laptop is "snowy" and fades in and out.  With another DVI display, the display quality is perfect.  So the DVI display does matter (although I don't have enough of a sample to tell you why one display is good and another is not).  

 

The HDMI->DVI adapter that I am using is Tripp-Lite Model P132-08N.

Edited May 7 by deeveedee