Slow wakeup problem

[rgviva]

Problem: my computer (Dell T5500) running 10.9 manages to go into sleep mode but when i press the power button to wake it up, it is "stuck" for 4-5 minutes and then finally it wakes up properly. Weird!

Saw similar reports here: https://github.com/tkrotoff/ASUS-P5E3Premium-Hackintosh/issues/1

Setup details:

Dell T5500, Dual Xeon X5560 - lspci attached

LPC/RTC kexts loaded (RTC CMOS Reset problem fixed)

NullCPU* kext removed

Chameleon-2.2svn-r2266 c-states p-states enabled

SMBios set to MacPro5,1

DSDT/Info attached

11/10/13 9:52:09.000 AM kernel[0]: Wake reason: VBTN (User)

11/10/13 9:56:19.000 AM kernel[0]: RTC: PowerByCalendarDate setting ignored

11/10/13 9:56:19.000 AM kernel[0]: Previous Sleep Cause: 5

11/10/13 9:56:19.000 AM kernel[0]: RTC: PowerByCalendarDate setting ignored

11/10/13 9:56:19.000 AM kernel[0]: USBF: 108.623 AppleUSBEHCI::CheckSleepCapability - controller will be unloaded across sleep

11/10/13 9:56:19.000 AM kernel[0]: USBF: 108.623 AppleUSBEHCI::CheckSleepCapability - controller will be unloaded across sleep

11/10/13 9:56:19.000 AM kernel[0]: The USB device HubDevice (Port 2 of Hub at 0x5d000000) may have caused a wake by being disconnected

11/10/13 9:56:19.174 AM WindowServer[92]: CGXDisplayDidWakeNotification [108779685774]: posting kCGSDisplayDidWake

11/10/13 9:56:19.174 AM WindowServer[92]: handle_will_sleep_auth_and_shield_windows: Reordering authw 0x7fa74a41c840(2004) (lock state: 3)

11/10/13 9:56:19.174 AM WindowServer[92]: handle_will_sleep_auth_and_shield_windows: err 0x0

11/10/13 9:56:19.000 AM kernel[0]: Auth result for: c8:60:00:e7:43:38 MAC AUTH succeeded

11/10/13 9:56:19.000 AM kernel[0]: AirPort: Link Up on en0

11/10/13 9:56:19.000 AM kernel[0]: en0: BSSID changed to c8:60:00:e7:43:38

11/10/13 9:56:19.919 AM configd[19]: LINKLOCAL en0: parent has no IP

Note ~4 minutes between Wake reason and RTC messages...

Any ideas?

T5500.zip

lspci.txt.zip

[pernacabeluda]

Did you try AppleACPIPlatform 1.3.6?

http://www.mediafire.com/download/l46x8zez97lta7y/AppleACPIPlatform.kext-1.3.6.zip

You may also need to patch AppleRTC

http://Olarila.com/forum/viewtopic.php?f=30&t=3416#p27223

Are you using this patch?

http://www.insanelymac.com/forum/topic/289798-applertc-patch-cmos-reset-after-sleepwake/?p=1966350

[rgviva]

Thanks for the ACPI Platform kext i will try it later today. I feel a bit uncomfortable installing 10.6.x kext on 10.9 but if it works then i'm good I also want to understand the problem if possible. I might attach a serial cable and do remote debug to see where the CPU is spending time.

The AppleRTC patch im using is the last port of the second link, i actually created this patch

perl -pi -e 's|\x41\x89\xd7\x41\x89\xf4|\xe9\xb8\x00\x00\x00\x90|;' /System/Library/Extensions/AppleRTC.kext/Contents/MacOS/AppleRTC

In AppleRTC there is a cmos write routine, so i short circuit it to not touch cmos at all (e9 b8 is jmp to end of routine). As far as i could see in IDA, cmos writes are used to update the system clock and to wake up the system after some time period (alarm). I don't care about those features to the patch is good for me. Haven't had any CMOS errors after applying it.

[rgviva]

AppleACPIPlatform kext v1.3.6 does not solve the problem

[pernacabeluda]

Try this

DSDT.aml.zip

[rgviva]

same