Contribute
Register

[Guide] Intel Framebuffer patching using WhateverGreen

Hi. Sorry for noob question and very bad English bc I'm Japanese.
Thanks for nice work.
Im trying Mojave on HD 4600.
It works nice Transparency and animation without using Whaterver Green.kext
using this config.
0x0a260006 >it-platform-id
0x04128086 >fake intel gfx id
and recognizes HD 4600 correctly on system information.
but freezes on safari video play and App Store load.
now I found this and I'm trying.
I tried using same platform id.
App Store works.also safari don't freeze on video play.
But transparency and animation doesn't work.System UI is very laggy.Shows display 4MB on about Mac.
I think it needs anything similiar fake id.
Can you please help? Thank you!
 
i tested all. all is fine but
correct hdmi patch is
<key>Key</key>
<string>framebuffer-con1-type</string>
<key>Value</key>
<integer>2048</integer>


<integer>0x00000800</integer> is wrong format in plistedit pro, Clover configurator.
Clover configurator automatically change string format if input 0x00000800 with number format.
so in CC, you need to put 2048 with number format in framebuffer-con1-type part
 
i tested all. all is fine but
correct hdmi patch is
<key>Key</key>
<string>framebuffer-con1-type</string>
<key>Value</key>
<integer>2048</integer>


<integer>0x00000800</integer> is wrong format in plistedit pro, Clover configurator.
Clover configurator automatically change string format if input 0x00000800 with number format.
so in CC, you need to put 2048 with number format in framebuffer-con1-type part
Correct, i did mention the same on post #19 but im still not sure if it works if we change it to 2048 or not ? (If this has been fixed or not)
 
Last edited:
right. i reported this problem to vit9696
 
Correct, i did mention the same on post #19 but im still not sure if it works if we change it to 2048 or not ? (If this has been fixed or not)

I know, but "-1" has a negative sign



Or maybe the wrong type, perhaps it should be <data> not <integer> .

try 2048 with number type. it's working
clover read hex values as string type. but whatevergreen is not same like clover
 
Not works on Xiaomi Notebook Pro (7mb 620HD) but on my MSI (5600HD) works perfectly although with Lag when starting the system
 
I know, but "-1" has a negative sign

I meant to say integers can be signed (but they can also be unsigned). When you use -1 it will convert to 0xFFFFFFFF (Data <FF FF FF FF>) which will then be cast to an 8 bit integer 0xFF which is the same as 255 or -1. -1 is used to DISABLE the port. It's just an example I use because I need to disable connector 3 otherwise my machine does not sleep correctly. Don't just cut and paste the examples into your config.plist and expect your graphics to work.

right. i reported this problem to vit9696
There is no problem to report.

i tested all. all is fine but
correct hdmi patch is
<key>Key</key>
<string>framebuffer-con1-type</string>
<key>Value</key>
<integer>2048</integer>

The example I gave works fine. If you enter 0x800 in Xcode's plist editor it will convert to 2048 but if you enter it as <integer>0x800</integer> in a text editor Clover will still read it correctly. The values are converted internally to Data eg. <00 00 00 80>. I contacted the author of CC and he is updating it to work the same as a plist editor ie. Entering it as 0x800 will convert to 2048. It may be easier for you to just enter it as decimal. It doesn't make a difference internally.
 
Last edited:
If you want to check the values you've entered in Clover's config.plist are correct check them in IORegistryExplorer. All values get converted to Data I just used examples that are easier to read. If you really want to be sure you can use Data format directly. Some people prefer it.
 

Attachments

  • FBPatch02.png
    FBPatch02.png
    648.6 KB · Views: 735
Last edited:
All values get converted to Data I just used examples that are easier to read. If you really want to be sure you can use Data format directly. Some people prefer it.

Ah, thanks for the clarification about the conversion . I prefer Data format directly since I prefer to inject those values via SSDT hot patch. For instance using rehabman's SSDT-IGPU.dsl. I just want to get familiar with the new method.
 

Attachments

  • Screen Shot 2018-07-25 at 1.40.44 PM.png
    Screen Shot 2018-07-25 at 1.40.44 PM.png
    507.6 KB · Views: 473
Back
Top