Contribute
Register

[help]black screen when uhd630 run with internal screen

Status
Not open for further replies.
For some reason I was using 0x3EA50009 so I've moved to 0x3E9B0000.

Do you think 0x3e9b0000 is a good default ig-platform-id if I was to add a plist to my guide plist collection for CoffeeLake native?
 
Do you think 0x3e9b0000 is a good default ig-platform-id if I was to add a plist to my guide plist collection for CoffeeLake native?

If you look at my guide you can see a list of recommended frames (from AppleLife forums).
 
If you look at my guide you can see a list of recommended frames (from AppleLife forums).

0x3ea50009 and 0x3e9b0000 seem very similar, but 0x3e9b0000 probably matches the device-id better (so non-connector data may match better).

I'm preparing a big update to the guide plists for them to use WhateverGreen style patching, Devices/Properties for ig-platform-id inject, device-id inject, etc. Will be updating the guide itself too (eg. WhateverGreen.kext instead of IntelGraphicsFixup.kext/CoreDisplayFixup.kext). Will add a UHD630 plist while I'm at it.

I suppose your suggestion is for 0x3e9b0000?

I have no CoffeeLake laptop to test with (would like to buy one when I find one worth buying...), but will be messing with my CoffeeLake NUC some more real soon...
 
Last edited:
I suppose your suggestion is for 0x3a9b0000?

0x3EA50009 and 0x3E9B0000 are very similar but I do think 0x3E9B0000 it better suited to UHD630 which matches the device-id of 0x3E9B.

As you mention there are a number of non-connector data differences between the two (which is easy to see using Intel FB-Patcher). For example 0x3E9B0000 has Camelia disabled. Also it has FBAvoidFastLinkTraining and FBAlternatePWMIncrement2 framebuffer flags set while 0x3EA50009 does not. 0x3EA50009 also has two FBUnknownFlag's that 0x3E9B0000 doesn't.

Unfortunately there is no easy way to patch the framebuffer flags using WhateverGreen (although in my earlier versions I did allow it we ultimately decided to remove it). From experience the flags don't seem to make much difference anyway.

I have no CoffeeLake laptop to test with, but will be messing with my CoffeeLake NUC some more real soon...

Nice. I'm a big fan of the small form factor PC's like the Brix and NUC's. Should make a great little hackintosh.
 
Last edited:
0x3EA50009 and 0x3E9B0000 are very similar but I do think 0x3E9B0000 it better suited to UHD360 which matches the device-id of 0x3E9B.

I'll use 0x3e9b0000 in the (newly added) config_UHD630.plist.

Unfortunately there is no easy way to patch the framebuffer flags using WhateverGreen (although in my earlier versions I did allow it we ultimately decided to remove it).

I think: framebuffer-patchX-find/framebuffer-patchX-replace.

But no reason to do that if you can just switch to the "more appropriate" ig-platform-id and use as-is (except for patching connectors/memory configuration, as needed).
 
Perhaps it's a BIOS setting you changed? Did you set "Fast Boot" to false by any chance? Changing that setting gave me issues.

Maybe it's some setting in your config.plist causing the issue? Did you give mine a try? You will notice my config.plist is very minimal. No junk so less chance of problems.

Hi,
Sorry i've been very very busy past weeks. And honestly i am a bit giving up on this issue by lack of time. I am already on Mojave Beta 10 and i don't imagine big changes coming with final release. Anyway i'll make some more trials on friday i think.
I indeed put back Fast Boot to True in BIOS, as you suggested, and get no more boot problem now (appart from the black screen). Thanks. But i cannot reproduce the "good start" without black screen i had sometimes. Maybe because of Fast Boot? Or maybe because i am using an external screen thru TB3?
Anyway for your question, when my boot was fine, i did get full hardware acceleration and informations seemed fine in System Infos. And as you suggested, using kextstat showed Whatevergreen actually loaded. So not the problem here i think.
But now i need to find a way to reproduce it and it's hard because it was very random..
 
Hi,
Sorry i've been very very busy past weeks. And honestly i am a bit giving up on this issue by lack of time. I am already on Mojave Beta 10 and i don't imagine big changes coming with final release.

Good to hear from you :) Yeah it's getting to the point where I don't think we can do much about. At least once the screen comes back on it's a working system so it's not the worst issue.

I indeed put back Fast Boot to True in BIOS, as you suggested, and get no more boot problem now (appart from the black screen). Thanks. But i cannot reproduce the "good start" without black screen i had sometimes. Maybe because of Fast Boot? Or maybe because i am using an external screen thru TB3?

I actually started getting the behavior you describe where the screen would come on right away and it appeared to boot perfectly. Seems completely random but I haven't seen it happen again in a while. I also use an external screen through TB3 but if I try and boot with it connected the internal screen will be very dark. So I need to boot up with the internal screen and then switch over to the external screen. I use a KVM to share the screen with my desktop so it's not really a problem.

BTW A moment ago I compiled Clover with modifications based on my investigation into the CFL framebuffer code and how AppleIntelFramebufferController::hwSetBacklight is different from previous versions in that it sets an extra register. But alas it didn't appear to make a difference.

Did you check out my post here? Turns out the IntelMaxValue needs to be shifted left 16 bits to be correct. So now when you use that you should see the terminal text on boot (until of course the 3 minute black screen starts).

RehabMan did suggest that I modify AppleIntelInfo.kext to get the output of critical IGPU registers before and after the black screen and compare them. It might help us locate the issue.
 
I'll use 0x3e9b0000 in the (newly added) config_UHD630.plist.



I think: framebuffer-patchX-find/framebuffer-patchX-replace.

But no reason to do that if you can just switch to the "more appropriate" ig-platform-id and use as-is (except for patching connectors/memory configuration, as needed).

Using latest self compiled debug versions of WhateverGreen and Lilu, getting KP with AppleIntelCFLGraphicsFrameBuffer.kext right after it says "WhateverGreen: igfx @ (DBG) Patching framebufferid 0x3EA50009 sucessful."

Happens with or without Devices/Properties. Attaching the kp photos.

More information -

MODEL - MSI GV 8RE.

OS - 10.13.6 (special build for CoffeeLake systems with AppleIntelCFLGraphicsFrameBuffer.kext support).

CPU - i7(8750H)

GFX - gtx1060(disabled), UHD630(cant get to work, as getting the KP).

SCREEN - 1920x1080 wide view Display(possibly eDP?), cant find much info about it.

Let me know if any other info is required. I can boot with 0x12345678 ig-platform-id set from Clover/config.plist/Graphics but not from Devices/Properties as it uses WG and it KPs.

This is all in order to get UHD630 acceleration working.

Also, is your UHD630 config.plist done?

I'd love to try it. :)
 

Attachments

  • 2018-09-20 23.38.34.jpg
    2018-09-20 23.38.34.jpg
    173.3 KB · Views: 98
  • 2018-09-20 23.38.46.jpg
    2018-09-20 23.38.46.jpg
    176.9 KB · Views: 103
Using latest self compiled debug versions of WhateverGreen and Lilu, getting KP with AppleIntelCFLGraphicsFrameBuffer.kext right after it says "WhateverGreen: igfx @ (DBG) Patching framebufferid 0x3EA50009 sucessful."

Happens with or without Devices/Properties. Attaching the kp photos.

More information -

MODEL - MSI GV 8RE.

OS - 10.13.6 (special build for CoffeeLake systems with AppleIntelCFLGraphicsFrameBuffer.kext support).

CPU - i7(8750H)

GFX - gtx1060(disabled), UHD630(cant get to work, as getting the KP).

SCREEN - 1920x1080 wide view Display(possibly eDP?), cant find much info about it.

Let me know if any other info is required. I can boot with 0x12345678 ig-platform-id set from Clover/config.plist/Graphics but not from Devices/Properties as it uses WG and it KPs.

This is all in order to get UHD630 acceleration working.

No "Problem Reporting" files attached.
Read FAQ, "Problem Reporting" again. Carefully. Attach all requested files/output.
https://www.tonymacx86.com/threads/faq-read-first-laptop-frequent-questions.164990/

Also, is your UHD630 config.plist done?

I'd love to try it. :)

Yes. It is in the beta branch.
 
No "Problem Reporting" files attached.
Read FAQ, "Problem Reporting" again. Carefully. Attach all requested files/output.
https://www.tonymacx86.com/threads/faq-read-first-laptop-frequent-questions.164990/



Yes. It is in the beta branch.
My laptop's hw is almost the same as @headkaze 's the screen seems different however).

FB-Patcher shows LVDS connected at 0 and as external, shouldn't it be internal for acceleration to work?(pic attached too).

Attached problem reporting files.
 

Attachments

  • debug_3948.zip
    572.8 KB · Views: 68
  • 2018-09-20 23.58.57.jpg
    2018-09-20 23.58.57.jpg
    51.6 KB · Views: 127
Status
Not open for further replies.
Back
Top