Search results

You have to find the EC address for the fan/s and how it reports the fan rpms from 0-100% or 100-0%, big or little endian, or 2 byte or more. There are a lot of configs with addresses here that hopefully include the one you're after. If not you need rweverything or similar to analyze the EC...

Yeah so it looks like all your ACPI errors are from the USB table I mentioned before, Drop the table "xh_cfht4" and load this SSDT instead, it's the same table but with a conditional that should make it work.

Yeah so what I thought, you have a bunch of ACPI errors. To solve the USB one, I find there's a bug in the MacOS interpreter and it ignores the whole table if it links to the same USB port twice, so to solve this you need to drop the SSDT - xh_cfhd4 table. Then replace it with a copy of the same...

Using Hackintool, go to Logs tab and under System and generate a boot log and post here. You will need to manually map your display ports. I have a 5491 and it used a couple of weird edits to get it perfect.

Sounds about right

Try this, just changed your ZWAK patch, there was an error in it. It was looking for 5F57414B01 and the match in your DSDT is 5F57414B09 Also removed my patch. everything else is as it was.

This is the patch and how it works, I thought that the external \_SB.PCI0.PEG0.PEGP.INIO under _WAK was the method waking the DGPU but it seems that is not the case, so will have to dig deeper. The patch replaced that call with a ADBG nonsense call that will be ignored. End result: So...

Try this, if it works we'll have to make an SSDT, since this patch will also affect windows.

In my Zbook G5 there is an acpi method called RDSS that is triggered by the embedded controller on wake that turned the DGPU on. In my case I had to replace the method and inject a modified version that didn't wake the DGPU. Maybe there is something similar in your ACPI tables. Dump all your...

You need to configure the bios so that DVMT is 64MB or higher. Some bioses give the option, some you have to find and set the variable using ru or modified efi shell.

DVMT > 64MB?

My broken screen dead battery mbp / thunderbolt display "macbook-pro-imac" with macpro1 cpu heatsink for silent fanless operation. The Ducktape just adds to the beauty.

This may work https://github.com/acidanthera/HibernationFixup

Just had another idea, have you tried disabling digital audio with flag "no-hda-gfx" in igpu device properties? It is possible the audio codec is causing the KP on wake.

Just for completeness maybe try imac 18,1 SMBIOS. It is kabylake, but uses IGPU only...

From reading about this in linux forums, it seems to be "fixed" by disabling deep sleep for igpu using kernel flag intel_idle.max_cstate=1. Some people claim their problem went away after a bios update, so I'm starting to think this is an intel driver/chip bug, and the OEM's just deactivated...

Only other idea that comes to mind is to go back to regular WEG and try faking device-id for Kabylake. Skylake and KB are very similar though so the wake issue may still be there. So keep the same info you have under AAPL,ig-platform-id but add a new entry under it for "device-id" KBL: 0x5912...

Try with the WEG kext above, and lilu 1.5, and remove the extra boot flags, also remove the PTS-WAK stuff above. Reverting back to as clean as possible and only this one patch on WEG. Use lilu debug version and try to get a full boot log (even if it crashes)

Also you can try this version of WEG, (you may need to downgrade to LILU 1.5) , I compiled it adding modForceWakeWorkaround.enabled = true; for Skylake iGPU.

you did both renames, _WAK to ZWAK and _PTS to ZPTS?