I have an issue with sleep/wake. I used your instructions to go into standby on lid close. Now the notebook does a hard reboot when I reopen the lid...

Did I d something wrong with my dsdt? Can you have a look at my screen lid section, please?

I've uploaded my dsdt here:
http://www.mediafire.com/?dgku4j4ibwz

Go to your System Preferences, then to Energy Saver and enable "Automatically Restart After Power Loss" checkbox. If you don't have it there, run the following in your terminal:

The system doesnt reboot then but the display gets stuck and may show the shutdown screen.

Sleep is a tricky thing under OSX. A lot of things need to work correctly there. Do you have vanilla SpeedStep working 100%? Do you have proper video drivers? What kind of audio drivers do you use? Which kexts you have in your /E/E?

I have the original Apple Intel GMA 950 drivers working with EFI String (otherwise resolution is stuck at 1280x1024).

In my /E/E folder I have:
- AppleACPIBatteryManager v. 1.0.15
- AppleACPIPS2Nub v. 1.0.0d1
- ApplePS2Controller v. 1.1.0
- FakeSMC v. 2.1
- LegacyX60T
- TabletEnabler
- VoodooHDA v. 0.2.56
- VoodooSDHC v. 1.1d1
- VoodooTSync v. 1.1

I noticed that I don't have sound right now, but it was working yesterday.
One thing worth mentioning is that I did not add SSDT info for C-states into my DSDT, because I'm sure that C-states do already work according to my whining CPU when I unplug the AC cord.

I also do have a L7500 Core 2 Duo in my X60t, so I'm not sure if that's part of the problem because I use the LegacyX60t kext and common SMBIOS settings for this - but the vast majority of the X60's used Core Duo L2xxx series CPUs.

I've uploaded my current complete Extra folder here:
http://www.mediafire.com/?22yr5nfmmfm

I think you should modify your LegacyX60t.Kext : the SMC part is related to your CPU (as far as I understand Silencer's guide). I chose MacBookPro1,1 because of MY cpu. Maybe you should use a MacBookPro3,1 thermal profile ... and to do so you just have to follow the "Vanilla SpeedStep" part of Silencer's guide (and modify your SMBIOS.plist ...)

OK, I'll give it a try later. However I think it should be a 2.x MacBook Pro, because we have the Intel 945 "Napa" chip in our X60 which is also limited to 3GB RAM (Your link is Santa Rosa / Intel 965 which is what the X61 has).

I think this 2,2 model should be OK:
http://www.everymac.com/systems/apple/m ... specs.html

Ok, with C-states you need to check IORegistryExplorer, if you have CSTInfo set in ACPI_SMC_PlatformPlugin (check my guide). If it is set, then you're golden. Otherwise - sorry, C-states are not functional for you at the moment.

VoodoHDA is known to break sleep on 60-series ThinkPads. Try AppleAzaliaAudio.kext instead, or remove VoodooHDA and Azalia altogether and enabled HDEF device in DSDT for testing.

Well I have another take on it. I highly doubt that your code does effectively enable the CPU to enter C4 state. Sleep/Wake now works perfectly for me, but C4 does not. I have not checked with IORegistryExplorer but I'm very sure about this for two reasons: First, my CPU makes a clearly audible high pitched nois when it's idling in C4 (also known as "CPU whine"). Secondly, I took some time and made my own DSDT with the original SSDT that I dissassembled for my BIOS. This version effectively enters C4, I can tell by the CPU whine and the difference in estimated battery runtime (~1:15h vs 1:40h). However, unfortunately my version doesn't wake from sleep (btw, I uninstalled the audio kext before testing, so it won't affect sleep).

I compared the original code for the _CST method with your hacked methods. While the original does have various routines for stepping between the C-States (always one up / or one down), I think it may be just an incident that deeper sleep states may ever be entered by the simple hack.

So I'm no expert on this matter, but without proof (battery runtime and CPU whine) I doubt that a CPU may ever enter C4 on C2D / CD CPUs with the hack. Even when IORegistryExplorer is showing the capability, that does in no way mean that the CPU ever enters this state.

If you check RMClock for Windows, you will notice, that for C2D C4-state can only be enabled when replacing C3. Effectively C2D CPUs can have only 3 C-states defined and available at the same time.

If you check SSDT code, you will notice, that all the various sets of C-states are always no more than three(1; 1-2; 1-2-3; 1-2-4). I've also noticed high pitched sound when CPU enters C4 state, so I'm using 1-2-3 C-states, that's more than enough for me temperature-wise. Essentially, if one enables at lease C1-state, that is already good enough.

You most probably do have C-states enabled (judging by the noise you described), but it is best to double check this in IORegistryExplorer.

Regarding Sleep problem. C-states do not affect Sleep at all. But other patches in DSDT do: HPET and RTC fix, plus a bunch of minor IRQ fixes. Theoretically SATA fix also might help. For me the problem with Sleep was with VoodooHDA (check this thread from the beginning).

That makes sense indeed. Do you know which package enables which set of C-States?



I included all of them and I removed the audio driver completely for testing purpose.

Well, here's the tricky part. You NEED to have HDEF device in your DSDT and allow AppleHDA.kext to load in order to enable waking from Sleep. So if you remove HDEF and VoodooHDA altogether, AppleHDA won't load and you will have the same problem.

Nice. Ok, now I have Sleep/Wake working, C-States working as well. The last thing missing now is T-state frequency, it always stays at 1,5GHz (1498 MHz) according to CPU-X. The problem might be that I'm using a Macbook Air profile for my legacy kext, but this one had an Intel 965 chipset and is differnet to the thermal profile settings of the Intel 945 based Macs.

There never was a Mac with Intel 945/ICH7 + L7xxx CPU, so there is no correct info in the ACPI kext. I'm trying to add P-State info to the DSDT now to overcome this.

T-states are for throttling, this is a way for the system to pull handbreak on your CPU. They almost don't affect temperature or power consumption, but will affect your performance.

CPU-X will always show fixed frequency value, it is not reliable way to check your CPU status. You can use either MSR Tools (32-bit only) or VoodooMonitor.

I'm just about to install 10.6, trying to download the kext's, but your "/Extra/ folder" has reached its download limit.
Could you please upload it to somewhere again?

Thanks in advance.

---

and dsdt too.

Are you asking me or silencer? If me, whats your system configuration?

I'm asking Silencer. My configuration is almost identical to his except my cpu is a T7200 and video card is V5250 (V5250:71d4, V5200:71c4).

If you happen to have downloaded his kext's before and can upload them, that'll be great too.
Thanks.

I still plan to update the post with some new info and new files, can't find enough time though. But I've recently helped a guy with V5250 and 15.4" screen. Here are the files for him (with his EDID already embedded in com.apple.Boot.plist): http://www.mediafire.com/?yztwzmqzxhq

If you have different screen, you need to work a bit to get your own EDID. I will also update the post probably in the end of the week.

Thanks. This is great.
I'm going to try it now.

I'm also looking forward to the update!

---
I'm guessing all I need to do is
1) Insert my own EDID
2) Customize DSDT
3) See how the system will work on the advanced mini dock

---

Kernel Panics from ATIRadeonX1000.kext : |
can be by passed if I open MacOS/ATIRadeonX1000 in hex editor and replace every DE 71 with D4 71, but still no QE/CI.

1. Well, you need gfxutil for that and the original plist to generate EFI string from it. I will include the info on how to create EFI strings when I update the guide.
2. You can use the one I provided.
3. I have mini-dock, and it works fine. I disabled all legacy ports there, and I'm not using external monitor, so can't say much for that part.



No, you don't need any binary patches for original drivers. you need to create proper EFI strings and inject them into com.apple.Boot.plist. You can extract your EDID (per bozilla's guide) and post here, I'll upload the EFI string for you.

Snow Leopard 10.6.3 is live!

I will update in the evening and will post the results here. Some preliminary info can be found here. Basically, if you have vanilla speed step and no disablers, you should be fine.

I updated yesterday ... Everything is OK for me (I use VoodooHDA ... AppleHDA seems to be modified ... keep the 10.6.2 version somewhere)

Boot and restart are a little bit longer (about 10 seconds) ...

hmm, this is weird. I was assuming the dock might be doing something causing the kernel panic.

I have done the following to install 10.6:
1. Install bootloader (10.5, then replaced with AsereBLN) & 10.6 from a 10.5, put the kext's (provided) in E/E, and boot.
Got Kernel Panic from IOATAFamily.

2. Use DSDT (the one you provided) OR use the patched IOATAFamily I found on Insanelymac.
then Kernel Panic from AppleTyMCEDriver (It sometimes KP, sometimes doesn't : |, while it does I renamed it with .noload and the problem will be solved)

3. Then the ATIRadeonX1000.kext will kernel panic. (and the AppleTyMCEDriver IS loaded. I'm not sure why it doesn't cause KP's any more)
I'm now stuck here.

A few things I've tried to solve the problem.
1. I extracted the device property in the com.apple.boot.plist you uploaded, replace the EDID with mine. (gfxutil -i hex -o xml, change EDID, and the otherway around)
does nothing, still kernel panic.

2. Find another v5250 ATYInject that's written a little differently. make it into gfx string.
does nothing, still kernel panic.

3. Binary patch.
kernel panic go away, but no QE&CI. Screen stuck at native solution (1600x1200).

4. I tried to find another pair of ATINDRV and ATYInject (changed to my edid), placed them in S/L/E.
now the system shuts down/goes to sleep when I enter the system. : |

I use pfix to do rebuild the cache every time I change kext.

Maybe I did something wrong I wasn't aware of...
Your help is much appreciated : )

My EDID,


---
to probably save sometime between replies, here's some more information.

I just changed /S/L/E back to vanilla, replaced my whole /E/E with the one you zipped.
Pfix'ed, and, still, kernel panic from ATIRadeonX1000.kext.

This is my com.apple.boot.plist is here: http://pastebin.com/5htfeueh

Could you please possibly also upload your ATIRadeonX1000.kext (and maybe ATYInject) too?

Update to 10.6.3 went fine with completely vanilla /S/L/E. After reboot I had to run full pfix and then reboot again to enable audio (VoodooHDA) and legacy kexts. Other than that - not problems.

Which bootloader did you install initially? And what do you mean by "install from 10.5"? Did you make a 100% clean install or did you upgrade from 10.5?

You need to install Chameleon RC4 using readme provided with it, and that repeat the procedure for AsereBLN booter. Or you can install AsereBLN over any booter, but with full install:



Hm, so if you use my DSDT, does the problem with IOATAFamily.kext go away?

I've did a quick check and AppleTyMCEDriver is apparently related to SMBIOS and machine identification. Did you use com.apple.Boot.plist and SMBIOS.plist provided? What is your reported machine identification?

if you use chameleon, or other bootloader with smbios replacement support, better to make identify of your machine to be some imac or something else, MacPro4,1 identify may result in kernel panic, or you need to remove AppleTyMCEDriver.kext with this script, durring installation


Ok, this is the weird one. I've never had kernel panics with ATIRadeonX1000.kext. I mean ever.


Ok, this was a smart thing to do. Although when you extract original plist, you need to also change the device ID in two places. I have the following:
ATY,DeviceID = C4710000
device-id = 29 125

You need:
ATY,DeviceID = D4710000
device-id = 29 140


Try these files. Place them to /S/L/E (save the originals beforehand) and then run full pfix. They can only work in i386.



Well, you have already did a smart thing by trying your EDID in EFI string, and it didn't work, so no use for me trying. The one I posted is actually a bit wrong, so you better trying to insert into com.apple.Boot.plist info from other ATI injects, not the one I have. Check the package for V5250 I've uploaded and linked above. It has ATIinject which tries to use X1700 and not X1600 as me. So it might work better for you. They also do set video memory size, which I don't set in my EFI string.

It also looks that V5250 is actually different from V5200 and requires more work to make it work in OSX. Try searching InsanelyMac.com for more info: http://www.google.com/search?hl=en&q=V5 ... elymac.com

I've also checked your EFI string in your com.apple.Boot.plist, and it looks fine to me.

I installed PC-EFI 10.5, and 10.6 on a clean partition (an empty external hard disk).
It will boot into single user mode (-s), and I installed AsereBLN in single user mode following your directions.

The problem with IOATAFamily did go away with your DSDT.
And I realized the AppleTyMCEDriver probably stopped KP because I placed your smbios in E/E.

I'll try the new kext's in a while and report back :）

Ok, one bad finding with 10.6.3 - CPU temperature jumped back to 60-65 degrees as before vanilla speed step. IORegistryExplorer reports that CSTInfo is set, so trying to figure out what is wrong.

It looks like the fix for bad temperature was an easy one. I've checked in MSR Tools, and speed step was not working - always max frequency (2.33GHz).

Then I've checked the difference between my legacy kext (block ACPI_SMC_PlatformPlugin/IOPlatformThermalProfile) and /S/L/E/IOPlatformPluginFamily.kext/Contents/PlugIns/ACPI_SMC_PlatformPlugin.kext/Contents/Resources/MacBookPro2_1.plist which is the one I use as a reference. And the difference was in CtrlLoopArray/0/StepDataDict/SP1 block. It is usually the same across all models, but it was changed in 10.6.3. After update and restart temperature has fallen back to normal.

I forgot one more thing. One of the options that might work almost without any work - RadeonHD.kext. You need to delete/disable ATINDRV.kext and remove the device-properties block in com.apple.Boot.plist. Then you can place RadeonHD.kext to /E/E or /S/L/E and run full pfix caches regeneration.

If it doesn't work OOB for you, insert your EDID into RadeonHD.kext/Contents/Info.plist, and it will work. There are minor problems with mouse cursor, but they are minor, and everything else works like a charm.

i updated to 10.6.3 using the software updater. now os x won't boot and hangs after loading systemversion.plist.

i've reinstalled aserebln using a terminal from the install dvd and tried to fix disk settings with the disk utility. but no luck.