Posted 01 August 2012 - 06:28 AM
With my AMD, you're right about the problem being graphics-related (thank you both a lot!), and i discovered using the connected server method that my graphics wasn't waking. Any suggestion of fixes?
With Atom, i tried a more conservative approach, like suggested by james (thanks!) with various different sleepenablers (using voodoo power mini, so i could get rid of nullpowermanagement), to no avail. The auto sleep doesn't work, pressing the power switch calls a dialog box with four options, sleep among them, but chosing this option only "sleep" the screen and pause iTunes (sleep with quotes, because screen freezes at the last image exihibited), as does the option at the apple menu. Using the hotkey (fn+f2) gives me an outright shutdown, and a header size error when booting after this shutdown (but the system boots fine besides the error). The patched RTC with conti's method (myhack) solved the CMOS reset.
So now i'm prepared to follow eep's advice and go catch my patch. That's all!
Once again, thank you for the help.
Posted 01 August 2012 - 08:41 AM
You may need a kext or 2.
Just remove them both and rely on VoodooPowerMini, use chameleon to generate your c+p states for you (set in your org.chameleon.Boot.plist)
Use DSDT Editor to extract your DSDT then patch the relevant sections.
You can download patches to apply to your DSDT, just use google.
If you get stuck just post in the DSDT sub-forum and I'm sure someone will help you if you ask nicely.
Glad you got it sorted but use the perl script next time. You will need to re-patch after most major updates, i.e. 10.7.5, 6 or any subsequent update.
Posted 01 August 2012 - 02:46 PM
As for the Atom sleep issues, i totally overlooked the c and p states options. Worth a try! I'd already removed AppleIntelCPUPowerManagement.
Posted 01 August 2012 - 10:48 PM
DarkWake=0 doesn't seem to be mentioned anywhere in the Chameleon documentation but it works on my system whereas DarkWake=No doesn't.
Posted 03 August 2012 - 03:17 PM
As for Atom, could you please provide a link to the sleepenabler that you use to get sleep working in your netbook? I tried quite a lot of them - i'm on 10.7.4 - and nothing worked. I tried the c and p states, but had no perceptible change. Thank you.
Posted 03 August 2012 - 07:17 PM
I can give you the sleep enabler I use but I use it on 10.6.8. My atom is an n270 so is only 32 bit and won't reliable run lion. It's actually my girlfriends netbook so don't want it crashing on her all the time because she'll only moan...more!
Posted 04 August 2012 - 01:41 AM
Thank you for the answer, James. I'm afraid the dsdt patching patch is the only left. But there's one last question before starting to patch: I did remove NullCPUPowerManagement indeed, as well as AppleIntelPowerManagement. But i left alone the kext AppleIntelPowerManagementClient, since it appears to do no harm. Could it have something to do with the lack of sleeping capability in my Atom netbook?
Posted 04 August 2012 - 10:12 AM
It's easy enough to restore using an install disk/USB if things go wrong.
Posted 04 August 2012 - 02:11 PM
Posted 22 September 2012 - 11:21 PM
After a month trying to patch my dsdt so sleep works in my Atom netbook, still no joy...
I successfully dealt with those errors of the last post of mine, which used to prevent me from effectively compile my patches. Oldnapalm point me the solution for it: i had to add DTGP patch, so i did it and finally compiled my first patched dsdt. The problem is that the patch that supposedly could grant my computer sleep/speedstep capabilities simply showed no result. Better saying, gave me a kernel panic (ACPI, non-supported CPU) when i removed nullCPUpowermanagement.kext and sleepenabler.kext. from S/L/E/. I tried other related patches to no avail.
In the meantime, i tried some "conservative" fixes for sleep/speedstep: VodooPowerMini (no result), a lot of sleep enablers (including the newest, sleepenableruniversal.kext), using the sleep enablers with the couple NullCPU, AppleCPU, with Voodoo alone, with voodoo+AppleCPU+NullCPU. Tried an heterodox fix i found in the Clover development thread (altering some info in the IOfamily kext plugin for the model of Mac my smBIOS assigned to my netbook). Any combination of those tries and fixes either didn't change a iota or messed with my system altogether.
I should've thrown the towel, but i'm not the one giving up here, so i return to dsdt, but still no positive results. Can someone take a look at my "virgin" dsdt, so to advise me on possible solutions for my problem? I'm not lazy and i'm not wanting nobody to do my job, but any light on the issue would help me a lot, because i'm simply stuck with this problem. Tried the ICH7 fix with and without the "conservative" methods above, but made no advances.
I uploaded my unedited dsdt.aml. Thank you for any bit of advice.
dsdt.aml.zip 16.22K 1 downloads
Posted 23 September 2012 - 06:14 PM
Posted 23 September 2012 - 08:41 PM
You're asking me about any of my dsdt attempts to use native power management? No, i didn't, since i was already in Mountain Lion when started to patch my dsdt. In fact, i'm using an APCI rollback (from 10.8.0) since i updated the system to 10.8.2, but this rollback, i think, is needed for all hackintoshes, right?
However, you may be asking about my "conservative" attempts, using legacy kexts, when i was running Lion. No, i didn't roll back. I tried all those legacy kexts in Mountain Lion, too, to no avail. Should i give a shot to the combination of legacy kexts/Snow Leo ACPI rollback with Mountain Lion? If it's the case, what's the combination of kexts do you advise me?
Should i try it? Could it be compatible with ML anyhow?
Thank you very much indeed, eeep!
Posted 23 September 2012 - 10:22 PM
P.S.: i have already an ACPI rollback in my Atom 10.8.2 hackbook: that rollback was necessary because of a change in the kernel that made it have incompatibilities with fakeSMC (more precisely, osmBIOS.kext). Will this older kext be a good replacement to the one already have, with the plus of enabling native power management?
Thank you, eeep! Much appreciated, really!
Posted 23 September 2012 - 10:36 PM
Posted 23 September 2012 - 11:12 PM
In fact, the one i'm using is the one uploaded by the user unixservice in the topic "10.8.2 Final (C54)". With that ACPI kext, my system is already running very smoothly, except for the annoying sleep issues and another one related to restart/reboot. That's why i made so many questions: i must confess that i'm afraid to mess my whole system just by trying to fix sleep. But i think i'll just give it a try.
Posted 24 September 2012 - 04:42 AM
Posted 24 September 2012 - 04:53 AM
In fact i think i'm very well prepared for this possibility, and it was thanks to it that i managed to make so much experiences and troubleshooting until this day: i have a full installation of Mountain Lion 10.8 in a 16GB USB stick. Slow as hell, takes an eternity to start up, but has, besides a working 10.8 running with the first 12.0 patched by meklort, also a comprehensive set of troubleshooting utilities that i collected during my (short) time of hackintoshing. So it would not be the end if it does not boot but, if you forgive the lousy pun, it would be a pain in the OS.
I'll give it a try ASAP, eeep! Thank you very much for the advice and the upload.
Posted 24 September 2012 - 04:39 PM
The sleep issues are still unresolved, but the older APCI kext solved my shutdown/restart issues, and appears to have shortened the boot time. Now, back to the sleep fixes!
Thank you once more!
Posted 24 September 2012 - 04:46 PM
Take a look and see if anything might be preventing sleep, in terminal type
pmset -g assertionsalso right after attempting to sleep, check the console log. Keep in mind, as long as NullCPUPowerManagement.kext is installed, sleep will never work, so you'll still need to be able to boot without it first. Don't know atom's very well, but maybe SSDT tables in /Extra could help
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users