Jump to content

[80% Solved] Iris Xe iGPU on Tiger Lake successfully loaded ICLLP Frambuffer and VRAM also recognizes 1536MB! + However, some issues.


shl628
740 posts in this topic

Recommended Posts

3 hours ago, jalavoui said:

well old nblue as this guy name

 

image.png.4f32a47bb58d2900b117dc5558668054.png

 

and new nblue

 

image.png.fbb4f31699d73efddf584f968e082184.png

 

and ofc it is crashing

 

renaming didnt solve the issue. same kp

gonna try debug it

 

wonder what changed from old xcodeproj to new nred proj

 

hmm i doubt this kp address is correct

 

image.png.6e9c432a42b8620677fc2f85d8904aba.png

 

code is fresh from nred

image.png.917981f3b1954083b54f6312957a630c.png

 

this is latest source - can't make it load without lilu crashes

 

NootedBlue-master.zip 366.16 kB · 2 downloads

 

any1 with older source plz upload it -  i got no backup of old code

this is the one from yesterday jala. Hope it helps.

NootedBlue-master.zip

Link to comment
Share on other sites

tks guys. anyway i will eventually need to refactor the code - there's just 2 much junk

 

k so i've added the source back but this time with lilu.kext and sdk in the zip (it's with zip files of sle_internal and hookcase pg 24)

 

it's fine to build it like that or update lilu+sdk from github

 

i comment out code that could cause some unnecessary crashes doing framebuffer loading (but i think some displays will need the disabled patches)

 

default loading is icl system framebuffer with minimum patches 

 

here's a test release. just add it to opencore and boot os x

 

NootedBlue.kext.zip

 

this is the linux setup at earlier stage

do we need this also on os x ?

i don't get the BAR 7 and 9 setup

 

image.png.eb0631e49e94195c2c9be898af3b8f7b.png

 

got a funny boot hang bug

i boot from a usb pen

if i let the pen plugged after os x boot 

os x detects the usb pen as disk errors and just hangs the system

soa fter boot from the usb pen i remove it and os x decides to boot

tks os x

 

Edited by jalavoui
Link to comment
Share on other sites

I do believe that Apple was in the process of the development of the driver for the Intel DG1 of the laptop dgpu and/or the pcie gpu card.

 

They may have written enough code to do their initial testing, however only the "Asus Vivobook Flip TP470EZ 14" Touch Laptop Core i7-1165G7 16GB 512GB XᵉMAX W11" could tell. https://www.ebay.com/itm/375300423795?var=&widget_ver=artemis&media=COP . Its kind of hard to tell though since the DG1 that was shipped was on a ARM device, but they could have been testing the hardware internally.

 

It would be good if we had a clone of the partitions from the DTK, then attempt to start testing using the Asus Vivobook Flip. Thats a good starting point, to test the DTK's response to hardware similar to it, that way we can verify how far teh devlopers went in making the code.

 

This is Dortianas Guide  to using up to two laptops via serial to do debugging --->https://dortania.github.io/OpenCore-Install-Guide/troubleshooting/kernel-debugging.html#kernel-debug-kits-optional

Link to comment
Share on other sites

8 hours ago, jalavoui said:

 

this is the linux setup at earlier stage

do we need this also on os x ?

i don't get the BAR 7 and 9 setup

 

 

 

BAR 7 looks like main VRAM memory setup for the iGPU based on the range.

 

BAR 9 should be for either overflow, advanced memory buffer or command streams.

Link to comment
Share on other sites

masta latest bin works for you?

 

i notice tgl as encoded some settings that make sense for dgpu. therefore the display paths, plane and scaler are setup from those settings and

they will not work for igpu unless patched. it's kinda icl logic and the way the code forces frame1 tobe enabled for our displays

 

and ofc the dsbmemory will have tobe replaced as we do not have a DG1 card

Edited by jalavoui
Link to comment
Share on other sites

you seem to have a vey stable system. idk if you wanna share some of your configs on asus thread. i do wanna see ben system start to work

 

btw maybe u wanna play with this part of nblue. its disabled by default. i think the bytes from ventura  should be easy to find in sonoma

 

image.thumb.png.110b6c6962c62c58336027add1304eed.png

Edited by jalavoui
Link to comment
Share on other sites

I was going to upload the fully working version on my github, but i hadnt gotten permission to use the nootedblue from you.

 

Also I think thats a bad idea, so many non devs will flock here and pollute the thread asking for working driver....

 

Im uploading my current configuration with the nblue version of my config.plist here. ( i havent edited and added my corrected EDID to the nblue plist a iMac display so it wont show up as internal display in hackintool)

 

Config.plist

 

Showing my pci config, so you know where im coming from...

 

pcidevices.plist

 

The only difference would be replacing whatevergreen with nootedblue (latest) and removing the device properties for the igpu.

 

Did you update the Master for nblue based on this new kext I saw it was edited 8 hrs ago?

 

Post the link to the asus thread? or was it posted here? I can go through the entire thread but just wanted to be sure before i started a rat race.

 

 

Edited by Mastachief
Link to comment
Share on other sites

yes  i reverted base of nred. the bin is same release.

 

i mean your config can help others on asus intel thread not here

 

btw i'm curious if you can get tgl frame logs

Edited by jalavoui
  • Like 1
Link to comment
Share on other sites

15 hours ago, Mastachief said:

I was going to upload the fully working version on my github, but i hadnt gotten permission to use the nootedblue from you.

 

Also I think thats a bad idea, so many non devs will flock here and pollute the thread asking for working driver....

 

 

 

I think that's the right approach tbh, even if I would personally like to try and adapt your latest binaries to my working alder+raptor lake setups. There is no use in getting spammed by 'wen eta' requests on this channel, especially when such good progress is being made. Happy to remain a spectator until there's something more readily available to share...

Link to comment
Share on other sites

No, I dont Jala

 

im going to put back the kexts in le and fix permissions and try again, i also changed the settings for the info.plist in nblue to load on tglle,

 

however it boots to a black screen, it looks similar to when icl is booting it just goes black before the window manager, no backlight, but im on Sonoma so i figure it needs the register bytes for Sonoma.

Edited by Mastachief
Link to comment
Share on other sites

guys cmon just push the source to github if you want. all good

 

i just manged to make tgl frame debug version boot to a different stage. updating sources soon

 

btw the trick for tgl is to hack the connectors a bit

 

k so tested on tgl debug but also updated production patches

if any1 can figure this we might have a working tgl framebuffer. this patches just avoid crashing so need to figure out what is wrong

 

image.png.64d69ca554d20a1659a65219aa7095e5.png

 

current release will boot to a a black display with a working cursor and will produce logs

 

Edited by jalavoui
Link to comment
Share on other sites

no masta this is framebuffer only testing

 

btw i need to update nblue again cause of missing production symbol. tested without -disablegfxfrimware

image.thumb.png.8ae06e6a3910ae7297cd756c4d013a0f.png

 

but it think u guys should try debug version as it produces lots of logs.

 

about github if u wanna upload just post link here so others can follow

 

Edited by jalavoui
Link to comment
Share on other sites

i was looking at my nred and found out this. this allows injecting values on aGPm. if any1 wanna play with  it just add this to nblue info.plist

thinking a bit about this its possible that agpm just helps hang the driver - we will see in the future. setting this to a invalid mac as the "feature" to partially disable agpm

 

image.png.03bdf67a5ad73113f614ee014fcab1c9.png

 

tested tgl production code. it does same as debug with extra warn for low stolen memory.

to swap debug/prodution change patches calls in nblue

nblue src code at pg 24

 

Edited by jalavoui
  • Like 1
Link to comment
Share on other sites

21 hours ago, Mastachief said:

Yes Jala, the latest Bin works.

 

Need any logs etc?

 

Screenshot2024-11-04at4_56_06PM.png.0f59f361b430373a2edca94c1ed4bd5c.png

 

Screenshot2024-11-04at4_49_39PM.thumb.png.4a5721a72eea05377eb494e2be6a5dd1.png

 

Screenshot2024-11-04at4_49_26PM.png.315e00b1f51bc28d8fcd4bc60b87f3b4.png

 

Screenshot2024-11-04at4_49_12PM.thumb.png.5950e001116cb02f9f01b696b17f3c4a.png

 

Screenshot2024-11-04at4_48_49PM.png.16a1595bb78a45de12bb6c43e1ea5786.png

 

Which framebuffer are you loading, using nblue? @Mastachief 

ICL with Graphicsschedulerselect= 0 ? I thinked gen7ICL... you meant ICLLP.. ok ... at me it stall and gpu goes in overheat..

Edited by ASUS Vivobook
Link to comment
Share on other sites

Yes, I am loading ICL framebuffer.

I am using the latest nblue master.

 

x.log

 

See the attached bootlog

 

whats your info.plist? also did you make edits that are not in the master?

Edited by Mastachief
  • Like 1
Link to comment
Share on other sites

masta what am i missing here?

 

this is from your log

 

2024-11-05 13:27:01.547870-0500  localhost kernel[0]: (IOPCIFamily) [ PCI configuration begin ]
2024-11-05 13:27:01.547871-0500  localhost kernel[0]: (IOPCIFamily) IOPCIConfigurator::configure kIOPCIEnumerationWaitTime is 900
2024-11-05 13:27:01.547887-0500  localhost kernel[0]: kPEDisableScreen -1
2024-11-05 13:27:01.549221-0500  localhost kernel[0]: (AppleACPIEC) ACPI: no ECDT
2024-11-05 13:27:01.559425-0500  localhost kernel[0]: initialize_screen: b=4000000000, w=00000780, h=00000438, r=00001E00, d=00000001
2024-11-05 13:27:01.559522-0500  localhost kernel[0]: kPEEnableScreen 1
2024-11-05 13:27:01.559523-0500  localhost kernel[0]: (IOPCIFamily) console relocated to 0x4000000000
2024-11-05 13:27:01.566502-0500  localhost kernel[0]: (IOPCIFamily) [ PCI configuration end, bridges 7, devices 20 ]

and mine

 

2024-11-01 15:11:46.558790+0000  localhost kernel[0]: (IOPCIFamily) [ PCI configuration begin ]
2024-11-01 15:11:46.558792+0000  localhost kernel[0]: kPEDisableScreen -1
2024-11-01 15:11:46.565599+0000  localhost kernel[0]: (IOPCIFamily) head.start     0x0
2024-11-01 15:11:46.565600+0000  localhost kernel[0]: (IOPCIFamily) head.size      0xcf8
2024-11-01 15:11:46.565601+0000  localhost kernel[0]: (IOPCIFamily) head.end       0xcf8
2024-11-01 15:11:46.565602+0000  localhost kernel[0]: (IOPCIFamily) head.alignment 0x1
2024-11-01 15:11:46.565602+0000  localhost kernel[0]: (IOPCIFamily) allocs:
2024-11-01 15:11:46.565603+0000  localhost kernel[0]: (IOPCIFamily) [end]
2024-11-01 15:11:46.565603+0000  localhost kernel[0]: (IOPCIFamily) ------
2024-11-01 15:11:46.565604+0000  localhost kernel[0]: (IOPCIFamily) head.start     0xd00
2024-11-01 15:11:46.565604+0000  localhost kernel[0]: (IOPCIFamily) head.size      0xf300
2024-11-01 15:11:46.565605+0000  localhost kernel[0]: (IOPCIFamily) head.end       0x10000
2024-11-01 15:11:46.565605+0000  localhost kernel[0]: (IOPCIFamily) head.alignment 0x1
2024-11-01 15:11:46.565606+0000  localhost kernel[0]: (IOPCIFamily) allocs:
2024-11-01 15:11:46.565606+0000  localhost kernel[0]: (IOPCIFamily) [0].start     0x164e
2024-11-01 15:11:46.565607+0000  localhost kernel[0]: (IOPCIFamily) [0].size      0x2
2024-11-01 15:11:46.565607+0000  localhost kernel[0]: (IOPCIFamily) [0].end       0x1650
2024-11-01 15:11:46.565608+0000  localhost kernel[0]: (IOPCIFamily) [0].alignment 0x1
2024-11-01 15:11:46.565608+0000  localhost kernel[0]: (IOPCIFamily) [1].start     0x1854
2024-11-01 15:11:46.565609+0000  localhost kernel[0]: (IOPCIFamily) [1].size      0x4
2024-11-01 15:11:46.565609+0000  localhost kernel[0]: (IOPCIFamily) [1].end       0x1858
2024-11-01 15:11:46.565610+0000  localhost kernel[0]: (IOPCIFamily) [1].alignment 0x1
2024-11-01 15:11:46.565610+0000  localhost kernel[0]: (IOPCIFamily) [2].start     0x3000
2024-11-01 15:11:46.565611+0000  localhost kernel[0]: (IOPCIFamily) [2].size      0x1000
2024-11-01 15:11:46.565611+0000  localhost kernel[0]: (IOPCIFamily) [2].end       0x4000
2024-11-01 15:11:46.565612+0000  localhost kernel[0]: (IOPCIFamily) [2].alignment 0x1000
2024-11-01 15:11:46.565612+0000  localhost kernel[0]: (IOPCIFamily) [3].start     0x4000
2024-11-01 15:11:46.565613+0000  localhost kernel[0]: (IOPCIFamily) [3].size      0x40
2024-11-01 15:11:46.565613+0000  localhost kernel[0]: (IOPCIFamily) [3].end       0x4040
2024-11-01 15:11:46.565614+0000  localhost kernel[0]: (IOPCIFamily) [3].alignment 0x40
2024-11-01 15:11:46.565615+0000  localhost kernel[0]: (IOPCIFamily) [4].start     0xefa0
2024-11-01 15:11:46.565615+0000  localhost kernel[0]: (IOPCIFamily) [4].size      0x20
2024-11-01 15:11:46.565616+0000  localhost kernel[0]: (IOPCIFamily) [4].end       0xefc0
2024-11-01 15:11:46.565616+0000  localhost kernel[0]: (IOPCIFamily) [4].alignment 0x20
2024-11-01 15:11:46.565617+0000  localhost kernel[0]: (IOPCIFamily) [5].start     0xfd60
2024-11-01 15:11:46.565617+0000  localhost kernel[0]: (IOPCIFamily) [5].size      0x4
2024-11-01 15:11:46.565618+0000  localhost kernel[0]: (IOPCIFamily) [5].end       0xfd64
2024-11-01 15:11:46.565618+0000  localhost kernel[0]: (IOPCIFamily) [5].alignment 0x1
2024-11-01 15:11:46.565619+0000  localhost kernel[0]: (IOPCIFamily) [end]
2024-11-01 15:11:46.565619+0000  localhost kernel[0]: (IOPCIFamily) ------
2024-11-01 15:11:46.565620+0000  localhost kernel[0]: (IOPCIFamily) ------------------------------------
2024-11-01 15:11:46.565620+0000  localhost kernel[0]: (IOPCIFamily) head.start     0x0
2024-11-01 15:11:46.565621+0000  localhost kernel[0]: (IOPCIFamily) head.size      0xcf8
2024-11-01 15:11:46.565622+0000  localhost kernel[0]: (IOPCIFamily) head.end       0xcf8
2024-11-01 15:11:46.565622+0000  localhost kernel[0]: (IOPCIFamily) head.alignment 0x1
2024-11-01 15:11:46.565623+0000  localhost kernel[0]: (IOPCIFamily) allocs:
2024-11-01 15:11:46.565623+0000  localhost kernel[0]: (IOPCIFamily) [end]
2024-11-01 15:11:46.565623+0000  localhost kernel[0]: (IOPCIFamily) ------
2024-11-01 15:11:46.565624+0000  localhost kernel[0]: (IOPCIFamily) head.start     0xd00
2024-11-01 15:11:46.565624+0000  localhost kernel[0]: (IOPCIFamily) head.size      0xf300
2024-11-01 15:11:46.565625+0000  localhost kernel[0]: (IOPCIFamily) head.end       0x10000
2024-11-01 15:11:46.565626+0000  localhost kernel[0]: (IOPCIFamily) head.alignment 0x1
2024-11-01 15:11:46.565626+0000  localhost kernel[0]: (IOPCIFamily) allocs:
2024-11-01 15:11:46.565627+0000  localhost kernel[0]: (IOPCIFamily) [0].start     0x164e
2024-11-01 15:11:46.565627+0000  localhost kernel[0]: (IOPCIFamily) [0].size      0x2
2024-11-01 15:11:46.565628+0000  localhost kernel[0]: (IOPCIFamily) [0].end       0x1650
2024-11-01 15:11:46.565629+0000  localhost kernel[0]: (IOPCIFamily) [0].alignment 0x1
2024-11-01 15:11:46.565629+0000  localhost kernel[0]: (IOPCIFamily) [1].start     0x1854
2024-11-01 15:11:46.565630+0000  localhost kernel[0]: (IOPCIFamily) [1].size      0x4
2024-11-01 15:11:46.565630+0000  localhost kernel[0]: (IOPCIFamily) [1].end       0x1858
2024-11-01 15:11:46.565631+0000  localhost kernel[0]: (IOPCIFamily) [1].alignment 0x1
2024-11-01 15:11:46.565631+0000  localhost kernel[0]: (IOPCIFamily) [2].start     0x3000
2024-11-01 15:11:46.565632+0000  localhost kernel[0]: (IOPCIFamily) [2].size      0x1000
2024-11-01 15:11:46.565632+0000  localhost kernel[0]: (IOPCIFamily) [2].end       0x4000
2024-11-01 15:11:46.565633+0000  localhost kernel[0]: (IOPCIFamily) [2].alignment 0x1000
2024-11-01 15:11:46.565634+0000  localhost kernel[0]: (IOPCIFamily) [3].start     0x4000
2024-11-01 15:11:46.565634+0000  localhost kernel[0]: (IOPCIFamily) [3].size      0x40
2024-11-01 15:11:46.565635+0000  localhost kernel[0]: (IOPCIFamily) [3].end       0x4040
2024-11-01 15:11:46.565635+0000  localhost kernel[0]: (IOPCIFamily) [3].alignment 0x40
2024-11-01 15:11:46.565636+0000  localhost kernel[0]: (IOPCIFamily) [4].start     0xefa0
2024-11-01 15:11:46.565636+0000  localhost kernel[0]: (IOPCIFamily) [4].size      0x20
2024-11-01 15:11:46.565637+0000  localhost kernel[0]: (IOPCIFamily) [4].end       0xefc0
2024-11-01 15:11:46.565637+0000  localhost kernel[0]: (IOPCIFamily) [4].alignment 0x20
2024-11-01 15:11:46.565638+0000  localhost kernel[0]: (IOPCIFamily) [5].start     0xfd60
2024-11-01 15:11:46.565639+0000  localhost kernel[0]: (IOPCIFamily) [5].size      0x4
2024-11-01 15:11:46.565639+0000  localhost kernel[0]: (IOPCIFamily) [5].end       0xfd64
2024-11-01 15:11:46.565640+0000  localhost kernel[0]: (IOPCIFamily) [5].alignment 0x1
2024-11-01 15:11:46.565640+0000  localhost kernel[0]: (IOPCIFamily) [end]
2024-11-01 15:11:46.565641+0000  localhost kernel[0]: (IOPCIFamily) ------
2024-11-01 15:11:46.565641+0000  localhost kernel[0]: (IOPCIFamily) ------------------------------------
2024-11-01 15:11:46.566500+0000  localhost kernel[0]: (AppleACPIEC) ACPI: no ECDT
2024-11-01 15:11:46.568506+0000  localhost kernel[0]: initialize_screen: b=4000000000, w=00000780, h=00000438, r=00001E00, d=00000001
2024-11-01 15:11:46.568600+0000  localhost kernel[0]: kPEEnableScreen 1
2024-11-01 15:11:46.568601+0000  localhost kernel[0]: (IOPCIFamily) console relocated to 0x4000000000
2024-11-01 15:11:46.568603+0000  localhost kernel[0]: (IOPCIFamily) [ PCI configuration end, bridges 2, devices 18 ]

no idea why this

Link to comment
Share on other sites

I get a boot now, my default boot is on GitHub, I only added the IGLogLevel=8 and -liludbg and the following settings.

 

Settings in nblue info.plist, it correctly identifies the tigerlake igpu without us having to patch under another device id, this is using ICL Debug in LE with Graphiscschedulerselect=5, plist and logs attached.

 

Info.plist

 

x.log

 

lilu.log

 

My ID is 9A49

 

image.png.d751252d0409bbe7e2fc2d470c10e4d8.png

 

image.png.8c82d188488418b1c5008bec9c961e4e.png

 

 

image.thumb.png.26c764bbf317ab75f1c4e435e8555cf6.png

 

 

image.thumb.png.d0e7ec77d5eec441473f9a5794684260.png

 

image.thumb.png.8089c39c650d000f1e56630607a52559.png

 

 

 

image.png.da7dbc5f4444fb83ed988aa520e5c6cc.png

 

image.png.8d6ec2225eb67fda44c1b03430affdd8.png

 

 

 

 

Edited by Mastachief
  • Like 1
Link to comment
Share on other sites

masta the icl frame debug should work fine for you. try install in /L/E

 

the tgl graph you installed is inactive in nblue (need path change in code to /L/E and com.apple to com.xxxxx rename inside nblue code). anyway default opt is /S/L/E

 

current active nblue patches are for tgl frame debug (can be changed to production inside nblue code)

 

gonna check your acpi patches in github maybe i figure out my pci crash at system.log

 

i didnt change nblue info.plist (i only move current card ids to the frame i wanna test)

Edited by jalavoui
Link to comment
Share on other sites

so i get bored after reading all this posts so i decided to install this on my system

 

image.thumb.png.4dde90e57ca819fa87ed818dca83483e.png

 

is this a fake ?

 

x.log.zip

 

better put this here tobe safe

-added sleep code disable

 

 

 

 

 

 

 

 

ty all for the support

Edited by jalavoui
Link to comment
Share on other sites

I get a black screen when it completes booting, the backlight is on, no cursor,

 

I can see that its loading the tgl frame based on the on-screen verbose log while booting, however the log isnt saved to disk for some reason, using my updated EFI on my USB stick.

 

I know you said default opt is /S/L/E

 

image.png.aed0310ab43ba86dfcaf3b6b636aa3b0.png

 

but framebuffer is set to L/E in latest Nblue for TGL, or is it that all the tgl sle files go to sle and then use the tgl frame meant for le? 🧐

 

image.png.47fe3f08042c77d661ca02419b479208.png

Edited by Mastachief
Link to comment
Share on other sites

masta better keep /S/L/E logic for testing tgl graph kext cause of dependencies for now

 

this are the binarys of the AppleIntelTGLGraphicsFramebuffer.kext debug for testing

 

install to /L/E and fix permissions

 

type in terminal

sudo -i

rm /Library/KernelCollections/AuxiliaryKernelExtensions.kc

chmod -R 755 /Library/Extensions/*

chown -R root:wheel /Library/Extensions/*

 

install nblue to OC/kexts

 

remove -v if you get no boot hang

image.png.52b3b25c526846a9d22dc1a5bac423aa.png

 

 

 

 

Edited by jalavoui
Link to comment
Share on other sites

I get igpu hang,

I installed hookcase and tglframe in le, and used the latest nblue in my USB OC.

I made sure tgl kexts were not in sle anymore, did a kernel cache rebuild for both boot cache and system/library/extensions cache

Is it because I'm on Sonoma? maybe I have to reinstall Ventura.....maybe i have bugged sle as you said jala 😕.

Screenshot2024-11-06at12_25_32PM.png.9f7d25c22853e2c45431c5e1fba3d285.png

 

Screenshot2024-11-06at12_27_49PM.png.c3a7b1afe226138463844fe845289437.png


356b4e49a7ee57d2195ca688504074ff.jpg

Kernel cache rebuild, I tried taking out the USB a few seconds after I choose the boot volume in opencore picker

unsupported PCH, which is normal

4726967c42c8ec31d54443fa666fe03d.jpg

Lilu patch messages, which look okay.

37d88606a84128c88b3cce20b40229f3.jpg

Edited by Mastachief
Link to comment
Share on other sites

×
×
  • Create New...