NootedGreen.kext is on air! It's a long long road to complete..

[Stezza88]

15 minutes ago, jalavoui said:

careful using com.xxxxx.driver.AppleIntelTGLGraphics that is old patch i've done. the original com.apple.driver.AppleIntelTGLGraphics is signed and maybe make os x happy.

 

if AppleIntelTGLGraphics loads in /L/E and the metal plugin gets load maybe it can work otherwise only path is /S/L/E

 

gonna try explain better why u might need revert ngreen

this patch in ngreen lacks a var fix

 

done on nblue

 

this is just a sample idk how long ai wil take to see this and other stuff

 

the info.plist of ngreen lacks all the properties that the kexts need. e.g

in ngreen

 

in nblue and nred logic u use the base kext properties, change as need  and just add pci ids in iopciprimarymatch if u want osx to load at boot

in /L/E u change iopciprimarymatch to 0x12345678 so it doesnt auto load

 

 

But you always told me to use only one IOPCIPrimaryMatch and it's conceptually wrong....

[jalavoui]

ah i see hmm sec ive removed the ids in current mblue

so this is original frambe that is instaled to /L/E and the pcimatch does nothing

 

if u copy this block to ngreen and change iopciprimarymatcg to your card id ngeen codo will load it

extra bonus: u can change also featurecontrol properties

 

same logic for com.apple.driver.AppleIntelTGLGraphics but u can try with the unsigned version u got in ngreen if works in /L/E

Edited April 23 by jalavoui

[Stezza88]

We are gonna breakthrough Mac OS or not? we are not here to go fishing.. ghghghgh

Take the good things from NootedGreen and import them! I'm gonna continue with ngreen fork of your nblue..

Edited April 23 by Stezza88

[Stezza88]

23 minutes ago, jalavoui said:

ah i see hmm sec ive removed the ids in current mblue

so this is original frambe that is instaled to /L/E and the pcimatch does nothing

 

if u copy this block to ngreen and change iopciprimarymatcg to your card id ngeen codo will load it

extra bonus: u can change also featurecontrol properties

 

same logic for com.apple.driver.AppleIntelTGLGraphics but u can try with the unsigned version u got in ngreen if works in /L/E

 

Man framebuffer and graphics must load simultaneously not separated way ... this is why ngreen born...

 

LE

\xxxx . fb

\xxxx . graphics 

 

OR

 

SLE

 apple. fb

 apple . graphics

Edited April 23 by Stezza88

[jalavoui]

i just remenbered why i made the com.xxxxx.driver.AppleIntelTGLGraphics dam im happy you are testing it - idea at that time was force load in /L/E

 

if u do this in ngreen it will load both kexts (replace com.apple.driver.AppleIntelTGLGraphics with com.xxxxx.driver.AppleIntelTGLGraphics)

 

 

i know cause this is nred logic and also cause i'm testing both this way

 

found ids on old release of nblue

 

0x9A608086 0x9A688086 0x9A708086 0x9A408086 0x9A498086 0x9A598086 0x9A788086 0x9AC08086 0x9AC98086 0x9AD98086 0x9AF88086 0x46808086 0x46828086 0x46888086 0x468A8086 0x468B8086 0x46908086 0x46928086 0x46938086 0x46A08086 0x46A18086 0x46A28086 0x46A38086 0x46A68086 0x46A88086 0x46AA8086 0x462A8086 0x46268086 0x46288086 0x46B08086 0x46B18086 0x46B28086 0x46B38086 0x46C08086 0x46C18086 0x46C28086 0x46C38086 0x46D08086 0x46D18086 0x46D28086 0x46D38086 0x46D48086 0xff058086 0x8A708086 0x8A718086 0x8A518086 0x8A5C8086 0x8A5D8086 0x8A528086 0x8A538086 0x8A5A8086 0x8A5B8086 0x4C808086 0x4C8A8086 0x4C8B8086 0x4C8C8086 0x4C908086 0x4C9A8086 0xA7808086 0xA7818086 0xA7828086 0xA7838086 0xA7888086 0xA7898086 0xA78A8086 0xA78B8086 0xA7218086 0xA7A18086 0xA7A98086 0xA7AC8086 0xA7AD8086 0xA7208086 0xA7A08086 0xA7A88086 0xA7AA8086 0xA7AB8086

but adding ids to both frame and acelerator will hang if u dont have them instaled

btw somewhere in code there is a functions that also setups some of this parameters

 

lol i posted wrong picture of my tests in loading another projet AppleIntelTGLEGLFramebuffer but it also works

https://github.com/pawan295/Appleinteltgldriver.kext

heres the nblue conf if any1 wanna test it - same logic add 0x12345667 to original file before install in /L/E

 

i just realized i updated tigerlake thread with old nblue source code

fixed you can get latest from there

my bad

Edited April 24 by jalavoui

[Stezza88]

4 hours ago, jalavoui said:

i just remenbered why i made the com.xxxxx.driver.AppleIntelTGLGraphics dam im happy you are testing it - idea at that time was force load in /L/E

 

if u do this in ngreen it will load both kexts (replace com.apple.driver.AppleIntelTGLGraphics with com.xxxxx.driver.AppleIntelTGLGraphics)

 

 

i know cause this is nred logic and also cause i'm testing both this way

 

found ids on old release of nblue

 

0x9A608086 0x9A688086 0x9A708086 0x9A408086 0x9A498086 0x9A598086 0x9A788086 0x9AC08086 0x9AC98086 0x9AD98086 0x9AF88086 0x46808086 0x46828086 0x46888086 0x468A8086 0x468B8086 0x46908086 0x46928086 0x46938086 0x46A08086 0x46A18086 0x46A28086 0x46A38086 0x46A68086 0x46A88086 0x46AA8086 0x462A8086 0x46268086 0x46288086 0x46B08086 0x46B18086 0x46B28086 0x46B38086 0x46C08086 0x46C18086 0x46C28086 0x46C38086 0x46D08086 0x46D18086 0x46D28086 0x46D38086 0x46D48086 0xff058086 0x8A708086 0x8A718086 0x8A518086 0x8A5C8086 0x8A5D8086 0x8A528086 0x8A538086 0x8A5A8086 0x8A5B8086 0x4C808086 0x4C8A8086 0x4C8B8086 0x4C8C8086 0x4C908086 0x4C9A8086 0xA7808086 0xA7818086 0xA7828086 0xA7838086 0xA7888086 0xA7898086 0xA78A8086 0xA78B8086 0xA7218086 0xA7A18086 0xA7A98086 0xA7AC8086 0xA7AD8086 0xA7208086 0xA7A08086 0xA7A88086 0xA7AA8086 0xA7AB8086

but adding ids to both frame and acelerator will hang if u dont have them instaled

btw somewhere in code there is a functions that also setups some of this parameters

 

lol i posted wrong picture of my tests in loading another projet AppleIntelTGLEGLFramebuffer but it also works

https://github.com/pawan295/Appleinteltgldriver.kext

heres the nblue conf if any1 wanna test it - same logic add 0x12345667 to original file before install in /L/E

 

i just realized i updated tigerlake thread with old nblue source code

fixed you can get latest from there

my bad

 

When i saw a "fragmented apple logo flash brief" my brain saw the light for a moment.. i still got that version maybe i will post at you so you can help me to achieve it another time

 

[Stezza88]

4 hours ago, jalavoui said:

i just remenbered why i made the com.xxxxx.driver.AppleIntelTGLGraphics dam im happy you are testing it - idea at that time was force load in /L/E

 

if u do this in ngreen it will load both kexts (replace com.apple.driver.AppleIntelTGLGraphics with com.xxxxx.driver.AppleIntelTGLGraphics)

 

 

i know cause this is nred logic and also cause i'm testing both this way

 

found ids on old release of nblue

 

0x9A608086 0x9A688086 0x9A708086 0x9A408086 0x9A498086 0x9A598086 0x9A788086 0x9AC08086 0x9AC98086 0x9AD98086 0x9AF88086 0x46808086 0x46828086 0x46888086 0x468A8086 0x468B8086 0x46908086 0x46928086 0x46938086 0x46A08086 0x46A18086 0x46A28086 0x46A38086 0x46A68086 0x46A88086 0x46AA8086 0x462A8086 0x46268086 0x46288086 0x46B08086 0x46B18086 0x46B28086 0x46B38086 0x46C08086 0x46C18086 0x46C28086 0x46C38086 0x46D08086 0x46D18086 0x46D28086 0x46D38086 0x46D48086 0xff058086 0x8A708086 0x8A718086 0x8A518086 0x8A5C8086 0x8A5D8086 0x8A528086 0x8A538086 0x8A5A8086 0x8A5B8086 0x4C808086 0x4C8A8086 0x4C8B8086 0x4C8C8086 0x4C908086 0x4C9A8086 0xA7808086 0xA7818086 0xA7828086 0xA7838086 0xA7888086 0xA7898086 0xA78A8086 0xA78B8086 0xA7218086 0xA7A18086 0xA7A98086 0xA7AC8086 0xA7AD8086 0xA7208086 0xA7A08086 0xA7A88086 0xA7AA8086 0xA7AB8086

but adding ids to both frame and acelerator will hang if u dont have them instaled

btw somewhere in code there is a functions that also setups some of this parameters

 

lol i posted wrong picture of my tests in loading another projet AppleIntelTGLEGLFramebuffer but it also works

https://github.com/pawan295/Appleinteltgldriver.kext

heres the nblue conf if any1 wanna test it - same logic add 0x12345667 to original file before install in /L/E

 

i just realized i updated tigerlake thread with old nblue source code

fixed you can get latest from there

my bad

 

Let's push to the final breakthrough. I'm stronger than ever.

[jkbuha]

19 hours ago, Mastachief said:

 

Sorry, I got the scanlines and scrambled screen issue fixed.

 

remove the boot-args that are blocking metal if you get the yellow tint in the background.

 

Use this version, it also has the blit mem fix from nblue as well to lower the window server usage was over 150% usage.

 

I had a version that booted and started the ioaccelerator and intelaccelarator, but its missing code.

 

NootedGreen.kext.zip 90.02 kB · 4 downloads

 

 

still scrambled screen with your updated kext - new logfiles attached

Lilu_1.7.3_24.6.txt x.log

[Mastachief]

the scrambled should be fixed after you put your ID in IOPCIPrimaryMatch, 0X9A498086 in both the tgl graphics kept and the tgl framebuffer kext

[jkbuha]

I got it to work by removing -allow3d and replacing -ngreenAllowMetal with -ngreenNoMetal

 

Screen loads fine, framebuffer loads and 1536MB is allocated to graphics.

[jlrycm]

44 minutes ago, jkbuha said:

I got it to work by removing -allow3d and replacing -ngreenAllowMetal with -ngreenNoMetal

 

Screen loads fine, framebuffer loads and 1536MB is allocated to graphics.

With acceleration?

[Stezza88]

57 minutes ago, jlrycm said:

With acceleration?

 

* it is possible soon or later
the problem to me is Metal API at the moment... but gpu seems working when patching with dyld
..but I wanna avoid dyld patches reaching a full MTL path so... crashes on crashes about MTL and CoreDisplay

* another problem is : only xxxx kexts are working in Library/Extensions?

Edited April 24 by Stezza88

[jkbuha]

4 minutes ago, Stezza88 said:

 

it is possible soon or later
the problem to me is Metal API at the moment... but gpu seems working when patching with dyld
..but I wanna avoid dyld patches to have a full MTL path so... crashes on crashes about MTL and CoreDisplay

 

For some reason, with your version of NootedGreen (including all updated repo commits), I don't get the framebuffer loading correctly on TigerLake - it stays at 8MB. Masta's patched NootedGreen works for me - ofc without acceleration, for now at least.

[Stezza88]

3 minutes ago, jkbuha said:

 

For some reason, with your version of NootedGreen (including all updated repo commits), I don't get the framebuffer loading correctly on TigerLake - it stays at 8MB. Masta's patched NootedGreen works for me - ofc without acceleration, for now at least.

 

to me NootedGreen is loading fb very well, gpu is working... but in black screen and only with dyld patches

......

 

i will fix every possibility but... I need to make my system works first of all..

Edited April 24 by Stezza88

[Stezza88]

On 4/24/2026 at 2:10 AM, jalavoui said:

i just remenbered why i made the com.xxxxx.driver.AppleIntelTGLGraphics dam im happy you are testing it - idea at that time was force load in /L/E

 

if u do this in ngreen it will load both kexts (replace com.apple.driver.AppleIntelTGLGraphics with com.xxxxx.driver.AppleIntelTGLGraphics)

 

 

i know cause this is nred logic and also cause i'm testing both this way

 

found ids on old release of nblue

 

0x9A608086 0x9A688086 0x9A708086 0x9A408086 0x9A498086 0x9A598086 0x9A788086 0x9AC08086 0x9AC98086 0x9AD98086 0x9AF88086 0x46808086 0x46828086 0x46888086 0x468A8086 0x468B8086 0x46908086 0x46928086 0x46938086 0x46A08086 0x46A18086 0x46A28086 0x46A38086 0x46A68086 0x46A88086 0x46AA8086 0x462A8086 0x46268086 0x46288086 0x46B08086 0x46B18086 0x46B28086 0x46B38086 0x46C08086 0x46C18086 0x46C28086 0x46C38086 0x46D08086 0x46D18086 0x46D28086 0x46D38086 0x46D48086 0xff058086 0x8A708086 0x8A718086 0x8A518086 0x8A5C8086 0x8A5D8086 0x8A528086 0x8A538086 0x8A5A8086 0x8A5B8086 0x4C808086 0x4C8A8086 0x4C8B8086 0x4C8C8086 0x4C908086 0x4C9A8086 0xA7808086 0xA7818086 0xA7828086 0xA7838086 0xA7888086 0xA7898086 0xA78A8086 0xA78B8086 0xA7218086 0xA7A18086 0xA7A98086 0xA7AC8086 0xA7AD8086 0xA7208086 0xA7A08086 0xA7A88086 0xA7AA8086 0xA7AB8086

but adding ids to both frame and acelerator will hang if u dont have them instaled

btw somewhere in code there is a functions that also setups some of this parameters

 

lol i posted wrong picture of my tests in loading another projet AppleIntelTGLEGLFramebuffer but it also works

https://github.com/pawan295/Appleinteltgldriver.kext

heres the nblue conf if any1 wanna test it - same logic add 0x12345667 to original file before install in /L/E

 

i just realized i updated tigerlake thread with old nblue source code

fixed you can get latest from there

my bad

 

* I'm just now passed to signed Apple drivers... from this moment I have loaded them in L/E

...i've always used your com.xxxxx driver but we will think about later...

Edited April 25 by Stezza88

[jkbuha]

16 minutes ago, Stezza88 said:

 

to me NootedGreen is loading fb very well, gpu is working... but in black screen and only with dyld patches

......

 

i will fix every possibility but... I need to make my system works first of all..

It's not an issue - I can build and compile both yours and his forked repo with 8 diffs but all your commits. If you need any additional logfiles or testing any code, lmk

[Stezza88]

.

Edited April 26 by Stezza88

[Stezza88]

* fixed first page and more clear readme
* added credits to various developer..

Edited April 25 by Stezza88

[Stezza88]

i was off road ... now I start with signed kexts...

[jkbuha]

On 4/25/2026 at 6:38 AM, Stezza88 said:

* fixed first page and more clear readme
* added credits to various developer..

Attaching updated files as per debugging request. This is using @Mastachief'a updated fork (1.0.4) so the display and fb load on my TGL 

Lilu_1.7.3_24.6.txt fb.log x.log

[Stezza88]

With signed I'm having a new series of kp.. wait for big fix for vanilla ngreen...

With new update they kill me my Claude opus... now only Claude sonnet is available...

Edited April 26 by Stezza88

[Stezza88]

LOL give me my money back!

 

Quote

Sorry, you have been rate-limited. Please wait 10 hours 45 minutes before trying again or consider switching to Auto.

 

[Stezza88]

I'm patching like a beast just wait

[Stezza88]

Actual state, still not committed, with signed kexts : start() returns, the scheduler exists, and encodeFailureStack[1] = 1 is still set. Then submitBlit keeps getting a null task 32 times and the ring never executes a single command.

We are so close.. fresh logs with more logging!

 

 

The culprit is the task producer, not start() and not the blit context getter. Log shows start() returned 1, the scheduler exists, but submitBlit still gets IGAccelTask * = NULL 32 times while the render ring stays dead at HEAD=TAIL=0x108, ACTHD=0, device state 0x1e. That means the failure is upstream in task creation, specifically the IGAccelTask::withOptions / createUserGPUTask path.

I patched that producer path in kern_gen11.cpp:798 and kern_gen11.cpp:1167, with the declarations/state in kern_gen11.hpp:1305 and kern_gen11.hpp:1409. On spoofed RPL only, if createUserGPUTask() returns null, NootedGreen now reuses Apple’s stable kernel task at accel+0x150, or the last successful task from IGAccelTask::withOptions(). Build succeeded and the edited files are error-free.

 

Edit : i had wrong ABI submitblit ... now new kp....  now resolved.. waiting 200 sec for new lilu logs..

Edit: submitblit is a big function... so need to be careful..

Lilu_1.7.2_23.6.txt

Edited April 26 by Stezza88

[jalavoui]

type in terminal after boot. use IGLogLevel=0xe

 

log show --style syslog --predicate 'processID == 0' --last 1h --info --debug > x.log

 

check if close to this

 

xx.log

 

try this also this - ngreen from github without junk

for framebufer only boot remove ids from IOPCIPrimaryMatch in Gen7TGL info.plist 

logic is gpu drivers + bundles in /S/L/E  for /L/E refactor info.plist and DYLDPatches

Arquivo.zipNootedGreen-main.zip

 

masta this is from github so i think old parameters u guys testing will still work. altough i removed some

btw i think u also have some nice patches so carefull with what u choose

 

 

Edited April 27 by jalavoui