nsemenov

Are you using DSDT at all? Without it AICPUPM is liable to panic. BTW HPET is enabled in BIOS and is patched in DSDT. In case without dsdt SleepEnabler.kext NullCPUPM is needed to get it working. There is no "correct" platform-id that would work OOB for me and some other folks. In theory 5430 that has BIOS hardcoded 64MB of stollen memory one must select from 2,3,8,9. In practice id=4 worked for me but it has only one pipe and supposedly can only work for screen mirroring (no screen extention) so I have not pursued this path. So my question remains what id have you used in dsdt or otherwise. So far is not clear what connection strings are working for you and how (mirror, extended desktop)

Is AppleIntelCPUPowerManagement loading as it should? The behavior you describe is very similar to when NullCPUPowerManagement is used instead but without SleepEnabler.kext. Read this thread for further info http://www.insanelymac.com/forum/topic/280733-sleep-issue-solved/ Where exactly are changes, what id? I do not see any that I would expect.

Please, try more recent dsdt (I just uploaded). I think I redid some patches in middle march and sleep definitely works with it. For an audio to work copy AppleHDADisabler.kext together with VoodooHDA.kext if you do not want to bother deleting AppleHDA.kext esp. after each system update.

My fixed Capri and dstd make use of platform-id 9. String is not needed because the dsdt patch sets platform id. 4G are reporting 384MB Video, after upgrade to 8GB - 512MB. This is pretty consistent throughout platforms. 64MB of stollen memory is fixed in BIOS and cannot be changed. I have a problem with HDMI so far that is why your results with id 3 gives me a hint. Still there is a chance I got faulty hardware for in windows hdmi works only during driver install but does not after reboot.

No way 01660003 would work on my 5430 without modification to Capri kext. (01660004 did work) 05 03 00 00 02 00 00 00 30 00 is in 01660003 ig-profile-id 05 03 00 00 02 00 00 00 30 02 - in 01660004 in unmodified file This is very interesting. Could you post your AppleIntelFramebufferCapri.kext?

With my dsdt.aml sleep should be working. As for ssdt it was generated with the script from HP Probook installer (tonycrap.org) for i3-3120 (4 states)

That EFI string you effectively set platform-id to 01660004 without video DSDT patch. That id matches single port configuration with connection string @0, 05 03 00 00 02 00 00 00 03 02 00 00 in Capri kext. So it works, but do not expect HDMI to work.

Take kexts from here. https://www.dropbox.com/sh/ess8prezh1ptco6/c7V_00VCVx AppleIntelFramebufferCapri.kext is tuned for profile-id 01 66 00 09

To let you all know. After update to 10.8.3 the following changes were needed. 1) In AppleInterFramebufferCapri.kext change connector string for AAPL,ig-platform-id: 01660003 (if that's what u use) from @0, 05 03 00 00 02 00 00 00 03 00 00 00 // LVDS @1, 02 05 00 00 00 08 00 00 06 00 00 00 // DisplayPort @2, 03 04 00 00 00 04 00 00 81 00 00 00 // DisplayPort @3, 04 06 00 00 00 04 00 00 81 00 00 00 // DisplayPort to @0, 05 03 00 00 02 00 00 00 03 02 00 00 // LVDS @1, 02 05 00 00 00 08 00 00 06 00 00 00 // DisplayPort @2, 03 04 00 00 00 04 00 00 81 00 00 00 // DisplayPort @3, 04 06 00 00 00 04 00 00 81 00 00 00 // DisplayPort Otherwise screen won't lit. Alternatively use patched one from 10.8.2 Some other kexts were overwritten So I had to return the patched ones. AppleACPIPlatform.kext (for battery) IOBluetoothFamily.kext IOAudioFamily.kext (needed by VoodooHDA?) IONetworkingFamily.kext no longer needs patching as it includes BCM5722D.kext with pci144e,1681 it seems As for WiFi card (I use Atheros AR5B95) I patched IO80211Family.kext again by adding its id into match list of info.plist inside the AirPortAtheros40.kext (WiFi) Let me know how it goes in your case. BTW. Still trying to enable HDMI. I am almost sure that it is the 4th connector string i.e. @3, 04 06 00 00 00 04 00 00 81 00 00 00 // DisplayPort It senses hotplug and it worked on my HP Probook but some numbers have yet to be adjusted.

For correct speed stepping try using SSDT.aml I have added for i3-3210 to /E. Since CPU might differ I'd rather not include it in dsdt. I also have added and corrected some patches to dsdt.aml - HPET, LPCB and other. Would be interesting to compare with yours clover-fixed. BTW I have problems with sleep when booting with Clover - sleep is not entered fully and no wake up. How do you judge that no usb3 is present, CalDigitxxx aren't loading? https://www.dropbox.com/sh/ess8prezh1ptco6/c7V_00VCVx

If possible, I would like to hear how you were able to achieve this. I am only able to detect HDMI hot-plugging (LCDV display reconnects) but no ext display detected. It is the only feature I need to make work to call it a day. Your corrupt display may indicate that HDMI is connected to a pipe that is on the same PLL with the pipe hooked to LCDV. Two of the three IvyB's pipes are using common PLL by design.

Working: HD4000 graphics (ig-platform-id = 01660003) Native CPUPM, AGPM, SS Audio Ethernet Bluetooth Wi-Fi (Centrino replaced with AR5B95) SD Card Reader Battery Issues: Have not checked HDMI after latest change in AIFbCapri but hopeful. After sleep USB Flash disconnects and needs to be physically replugged. After eject SD Card hot-plug works with 1-2 min delay. OK on the first insert. I needed your DSDT to check what patches I might have missed. https://www.dropbox.com/sh/ess8prezh1ptco6/c7V_00VCVx

How did it go? Is there any way you can share dsdt.aml?