Skylake HD530 - Sleeps but won't wake

[spunog]

Hi, I've noticed something. Hackintosh goes to sleep fine and wakes up but no video signal restores back to tv from intel hdmi. Using skylake with 530 on a asus z170-A board. So, what I noticed was, that if I turn on the computer normally from the off state, and THEN turn on the TV, I get no signal, just like when waking up from sleep. If, however, I turn on the TV first and then the computer then the usual screen comes up, clover drive selection. I'm thinking maybe it's some bios settings, any theories? (bios settings used during set-up:
BIOS Features: Fast boot - Enabled PS2 Devices Support - Disabled CSM Support - Disabled Secure Boot - Disabled

Peripherals: XHCI Hand-Off - Enabled Super IO Configuration -> Serial Port - Disabled

Chipset: VT-D - Disabled IOAPIC 24-119 - Disabled)

 

[brittonberkan]

So, still no way to fix this?

Just got everything else to work on my Hackintosh, and am preparing a guide for this hardware setup. Would like to have this fixed before posting a guide though

 

[pgriss]

I have the same problem.

it's the same issue as AMD graphics cards- the machine sleeps fine but the screen doesn't wake up.

It may be important to realize that this problem with the HD530 is not limited to the screen not waking up. The OP accurately described the issue that I am also seeing, which is that the computer seems to power on when waking up but nothing works, not just the video signal. I can't ping the machine and keyboard input is not processed.

 

[pgriss]

if I turn on the computer normally from the off state, and THEN turn on the TV, I get no signal, just like when waking up from sleep. If, however, I turn on the TV first and then the computer then the usual screen comes up, clover drive selection.

I think this is a different issue. If my computer is running and I just turn off the monitor and turn it on again, the signal doesn't come back. The computer itself still functions though. I think this is the situation that you are describing, except that you start with the computer off.

The problem with sleep/wake is more severe in that it's not limited to the video signal, see my previous post.

 

[brittonberkan]

Can confirm that this is basically killing the machine, not just a problem with the display not waking up.

However, i think it might be entirely unrelated to sleep. The system is in the same broken state when I just unplug and replug the HDMI or power cable on the monitor, no sleep required. Or does that only happen on my end?

 

[ideabit]

Finally, I got it fixed (i5 6500). Following steps is not detailed enough, just for your information.

1) Set hibernate mode
$ sudo pmset -a hibernatemode 25

2) Make sure image file can be written.
$ ls -lh /var/vm/sleepimage
-rw------T 1 root wheel 8.0G Dec 1 23:35 /var/vm/sleepimage

3) Not sure if these configuration is required.
System Preferences >> Energy Saver >> Power Nap disable
Clover: Boot parameter >> darkwake=0

4) Download RehabMan-FakePCIID-2016-1115.zip from https://bitbucket.org/RehabMan/os-x-fake-pci-id/downloads
extracting zip file, then copy:
FakePCIID.kext
FakePCIID_Intel_HD_Graphics.kext
from Release to /Volumes/EFI/EFI/CLOVER/kexts/Other/

5) Modify /Volumes/EFI/EFI/CLOVER/config.plist according attached config-Intel-HD530.plist
NOTE: I also changed the value of ig-platform-id from 'AAAWGQ==\n' to 'AAASGQ==\n'

6) Restart.

Good luck.

 

[ideabit]

'AAAWGQ==' is the encoding of 0x191600, and
'AAASGQ==' is the encoding of 0x191200.
* it seems that the '\n' should not be included.

 

[thedman]

Can confirm that this is basically killing the machine, not just a problem with the display not waking up.

However, i think it might be entirely unrelated to sleep. The system is in the same broken state when I just unplug and replug the HDMI or power cable on the monitor, no sleep required. Or does that only happen on my end?

No no, you are not alone, I get this exactly. If the monitor disconnects in anyway I cannot use the system anymore.

 

[thedman]

I'm a bit scared to follow the work of idea bit above, has anyone else tried???

 

[RehabMan]

'AAAWGQ==' is the encoding of 0x191600, and
'AAASGQ==' is the encoding of 0x191200.
* it seems that the '\n' should not be included.

Both 0x191600 and 0x191200 are invalid ig-platform-id values.
But they actually decode to 0x19160000 and 0x19120000 (which are valid).
And yes, the \n is wrong.