Contribute
Register

Skylake HD530 - Sleeps but won't wake

Status
Not open for further replies.
Just took out my Nvidia GPU yesterday and worked for hours trying to get machine to wake running Intel 630 graphics. Finally found this thread. Is it pretty much going to be impossible to get sleep wake with HD 630 until High Sierra (Or perhaps 10.12.6) because of this problem? I was going to return the GTX 1060 today because of the sleep glitch it has, but after an all nighter working on integrated graphics, it seems I was better off holding onto the GTX 1060 and hope Nvidia fixes the glitch. At least with the GPU I have sleep wake working.
 
Last edited:
No kexts installed. Works out of the box, including sleep/wake.
 
@bombardier10 how i can modify the setting at the 3^ step ? Set darkwake=10 . (Post 191)Sorry for the noob question but it's New of hackintosh . Thank you for support !
 
Still no solution? I have switched to a new GA z170N WIFI M-ITX and I have the same problem that it won't wake up! if anyone knows what works let me know! I have the integrate 530 graphics and nothing else. removed all kexts related to the intel gfx and still no.
 
Still no solution but I can give some additional info. Even my MBP with Intel Iris Pro when using an external monitor sometimes doesn't wake up the exact same way as my hack. Also a while back Windows users had the same problem until Intel fixed the drivers. Linux had the same problem too some developers got together and fixed the problem. All I can think is that these fixes have not made it to Apple's drivers yet, or they made a partial fix just enough to make MBPs a bit more reliable but the fix has no effect on hackintosh.

Last time I looked into it, I discovered it's caused by a bug in the dynamic display frequency calculation. The first time an external display is initialised it correctly calculates its frequency, however when a display is hot-plugged it fails to calculate and crashes, which you can recreate by unplugging and plugging in your monitor. This kind of bug usually comes about when the developer has not reused the same code for initialisation as for change of state, and it's likely only the init code was tested. On the Mac you'll see "invalid display frequency" output to the console or something similar. The reason why it doesn't happen on MBP internal displays is I believe the frequency settings are known and don't need to be calculated. On Linux they fixed the display frequency calculation and all is good for them now. I don't have it handy but I began to reverse engineer the method in AppleIntelSKLGraphicsFramebuffer.kext in attempt to find the bug, it was tricky though because it was quite different from the Linux driver code I was using as reference.
 
Well, I read that someone said they got it to work by readjusting screen resolution. The display "resets". Maybe there is a work around that involves causing the display to rethink its connection as it wakes.
 
@bombardier10 what was it you were saying about increasing resolution of screen sharing? I too have an issue where screensharing looks like crap
 
josephholder said:
Well, I read that someone said they got it to work by readjusting screen resolution. The display "resets". Maybe there is a work around that involves causing the display to rethink its connection as it wakes.
Click to expand...

Yes I remember reading that too. Someone had configured command+1 to change resolutions in QuickRes and did it while the screen was black and it made the MBP work again. And if a resolution change causes a complete initialisation rather than a more dynamic hot-plug then it fits with the theory. Probably would be quite simple to make a daemon which listens for wakeup and adjusts the screen res. I might give that a try actually.
 
Status
Not open for further replies.
Copyright © 2010 - 2024 tonymacx86 LLC
Back
Top