DSDT - Vanilla Speedstep - Generic Scope (_PR)

[mitch_de]

Anyone has an iMac9,1 smbios (i read that this works fro Cstate) or at least the main data i put in the smbios ?

 

PS: I see (kextstat) that com.apple.driver.AppleLPC (1.3.1) is loaded (unpatched).

[FKA]

I kept wondering why some people here had SSDT tables in their ACPI table, while others, including me had not. I noticed that the RSDT table on my Mac was different than the one on the hack so I searched and found this:

 

"When OSPM locates the structure, it looks at the physical address for the Root System Description Table or the Extended System Description Table. The Root System Description Table starts with the signature “RSDT”, while the Extended System Description Table starts with the signature “XSDT”. These tables contain one or more physical pointers to other system description tables that provide various information about the system. As shown in Figure 5-1, there is always a physical address in the Root System Description Table for the Fixed ACPI Description table (FADT).

 

When OSPM follows a physical pointer to another table, it examines each table for a known signature. Based on the signature, OSPM can then interpret the implementation-specific data within the description table.

 

The purpose of the FADT is to define various static system information related to configuration and power management. The Fixed ACPI Description Table starts with the “FACP” signature. The FADT describes the implementation and configuration details of the ACPI hardware registers on the platform."

 

This is the RSDT of a MacPro3,1:

/*
* Intel ACPI Component Architecture
* AML Disassembler version 20080926
*
* Disassembly of RSDT.dat, Sat Sep 19 13:56:53 2009
*
* ACPI Data Table [XSDT]
*
* Format: [HexOffset DecimalOffset ByteLength]  FieldName : FieldValue
*/

[000h 000  4]                    Signature : "XSDT"    /* Extended System Description Table */
[004h 004  4]                 Table Length : 000000EC
[008h 008  1]                     Revision : 01
[009h 009  1]                     Checksum : 47
[00Ah 010  6]                       Oem ID : "APPLE "
[010h 016  8]                 Oem Table ID : "Apple00 "
[018h 024  4]                 Oem Revision : 0000006C
[01Ch 028  4]              Asl Compiler ID : "    "
[020h 032  4]        Asl Compiler Revision : 01000013

[024h 036  8]       ACPI Table Address   0 : 000000007F740000
[02Ch 044  8]       ACPI Table Address   1 : 000000007F742000
[034h 052  8]       ACPI Table Address   2 : 000000007F73F000
[03Ch 060  8]       ACPI Table Address   3 : 000000007F73D000
[044h 068  8]       ACPI Table Address   4 : 000000007F73C000
[04Ch 076  8]       ACPI Table Address   5 : 000000007F736000
[054h 084  8]       ACPI Table Address   6 : 000000007F735000
[05Ch 092  8]       ACPI Table Address   7 : 000000007F734000
[064h 100  8]       ACPI Table Address   8 : 000000007F733000
[06Ch 108  8]       ACPI Table Address   9 : 000000007F732000
[074h 116  8]       ACPI Table Address  10 : 000000007F731000
[07Ch 124  8]       ACPI Table Address  11 : 000000007F730000
[084h 132  8]       ACPI Table Address  12 : 000000007F72F000
[08Ch 140  8]       ACPI Table Address  13 : 000000007F72E000
[094h 148  8]       ACPI Table Address  14 : 000000007F72D000
[09Ch 156  8]       ACPI Table Address  15 : 000000007F72C000
[0A4h 164  8]       ACPI Table Address  16 : 000000007F72B000
[0ACh 172  8]       ACPI Table Address  17 : 000000007F72A000
[0B4h 180  8]       ACPI Table Address  18 : 000000007F729000
[0BCh 188  8]       ACPI Table Address  19 : 000000007F728000
[0C4h 196  8]       ACPI Table Address  20 : 000000007F727000
[0CCh 204  8]       ACPI Table Address  21 : 000000007F726000
[0D4h 212  8]       ACPI Table Address  22 : 000000007F722000
[0DCh 220  8]       ACPI Table Address  23 : 000000007F71D000
[0E4h 228  8]       ACPI Table Address  24 : 000000007F71A000

Raw Table Data

 0000: 58 53 44 54 EC 00 00 00 01 47 41 50 50 4C 45 20  XSDT.....GAPPLE 
 0010: 41 70 70 6C 65 30 30 20 6C 00 00 00 20 20 20 20  Apple00 l...    
 0020: 13 00 00 01 00 00 74 7F 00 00 00 00 00 20 74 7F  ......t...... t.
 0030: 00 00 00 00 00 F0 73 7F 00 00 00 00 00 D0 73 7F  ......s.......s.
 0040: 00 00 00 00 00 C0 73 7F 00 00 00 00 00 60 73 7F  ......s......`s.
 0050: 00 00 00 00 00 50 73 7F 00 00 00 00 00 40 73 7F  .....Ps......@s.
 0060: 00 00 00 00 00 30 73 7F 00 00 00 00 00 20 73 7F  .....0s...... s.
 0070: 00 00 00 00 00 10 73 7F 00 00 00 00 00 00 73 7F  ......s.......s.
 0080: 00 00 00 00 00 F0 72 7F 00 00 00 00 00 E0 72 7F  ......r.......r.
 0090: 00 00 00 00 00 D0 72 7F 00 00 00 00 00 C0 72 7F  ......r.......r.
 00A0: 00 00 00 00 00 B0 72 7F 00 00 00 00 00 A0 72 7F  ......r.......r.
 00B0: 00 00 00 00 00 90 72 7F 00 00 00 00 00 80 72 7F  ......r.......r.
 00C0: 00 00 00 00 00 70 72 7F 00 00 00 00 00 60 72 7F  .....pr......`r.
 00D0: 00 00 00 00 00 20 72 7F 00 00 00 00 00 D0 71 7F  ..... r.......q.
 00E0: 00 00 00 00 00 A0 71 7F 00 00 00 00              ......q.....

 

And this of the hack:

/*
* Intel ACPI Component Architecture
* AML Disassembler version 20080926
*
* Disassembly of RSDT.dat, Sat Sep 19 13:45:06 2009
*
* ACPI Data Table [XSDT]
*
* Format: [HexOffset DecimalOffset ByteLength]  FieldName : FieldValue
*/

[000h 000  4]                    Signature : "XSDT"    /* Extended System Description Table */
[004h 004  4]                 Table Length : 00000054
[008h 008  1]                     Revision : 01
[009h 009  1]                     Checksum : A9
[00Ah 010  6]                       Oem ID : "A_M_I_"
[010h 016  8]                 Oem Table ID : "OEMXSDT "
[018h 024  4]                 Oem Revision : 12000816
[01Ch 028  4]              Asl Compiler ID : "MSFT"
[020h 032  4]        Asl Compiler Revision : 00000097

[024h 036  8]       ACPI Table Address   0 : 0000000000DEB000
[02Ch 044  8]       ACPI Table Address   1 : 00000000CFF80390
[034h 052  8]       ACPI Table Address   2 : 00000000CFF80400
[03Ch 060  8]       ACPI Table Address   3 : 00000000CFF8E040
[044h 068  8]       ACPI Table Address   4 : 00000000CFF88F80
[04Ch 076  8]       ACPI Table Address   5 : 00000000CFF88FC0

Raw Table Data

 0000: 58 53 44 54 54 00 00 00 01 A9 41 5F 4D 5F 49 5F  XSDTT.....A_M_I_
 0010: 4F 45 4D 58 53 44 54 20 16 08 00 12 4D 53 46 54  OEMXSDT ....MSFT
 0020: 97 00 00 00 00 B0 DE 00 00 00 00 00 90 03 F8 CF  ................
 0030: 00 00 00 00 00 04 F8 CF 00 00 00 00 40 E0 F8 CF  ............@...
 0040: 00 00 00 00 80 8F F8 CF 00 00 00 00 C0 8F F8 CF  ................
 0050: 00 00 00 00                                      ....

This explains the missing tables. Nothing more. And you might be smart/brave (stupid?) enough to flash a modified BIOS ROM. I however really hope that we can find a low-profile solution for everyone.

 

So you'd need RSDT override function in the bootloader !

 

I thought I read somewhere once that Chameleon can load SSDT.dat..? Maybe I'm confusing things.

 

If CHameleon and or EFI v10 can/could load SSDT.dat - how much would it take to override RSDT - if this is even a solution !

 

No way I'm prepared to flash BIOS with this!

[Master Chief]

So you'd need RSDT override function in the bootloader!

So it seems.

 

If CHameleon and or EFI v10 can/could load SSDT.dat - how much would it take to override RSDT - if this is even a solution !

It will for people with SSDT tables in their BIOS.

 

And here's an interesting link: http://www.insanelymac.com/forum/index.php?showtopic=145792

That loads my four tables (/SSDT-n.aml) and successfully it says, but nothing shows up in the ACPI table?!?!

 

No way I'm prepared to flash BIOS with this!

It can be a bag of hurt yes. And I know because I tried it yesterday.

[Squirrelf]

This is my dsdt with cst, we have almost the same board (ep45-ud3r)

I've put all the "SSDT stuff" in the beginning of the dsdt

 

do you have any problems with c-ctates?

you don't get any stuttering?

are your temps low (as low as having a disabler), are you loading appleLPC?

 

OK, I'll give CStates a go.

Do you mind posting your version of CStates. Is CState values specific to your hardware though? I didn't get a CST dump from Everest.

...

Any ideas on the Geekbench score?

 

try a bench again, but this time temporarily remove cpui again...

 

Use WinstonAce dsdt

he has the same board as me, just a newer CPU and has his own cst dump from linux/everest

Since he did get dump and I didn't on the same board, it seems to me that it is "CPU specific" (I think it is more specific to cpu generation rather than exact cpu)

I have a Q6600 and has either a 8xxx or a 9xxx

I don't have C2/C4 option in my bios because of my cpu...

 

I would say it looks like your 4 c-states are loaded 0 , 1, 2, 3

 

here's my screen grab. note im using c-states from mac-pro3,1 that only has c0, 1 and 2

NB they may be listed in IOREG and loaded but they dont function!

 

 

I don't think the 1, 2, 3, 4 are c-ctates, they seems to be p-states instead ( don't you have only 3 p-steps in you dsdt, if you have more then it is very strange...)

My Q6600 won't even support C2/C4, only supports C1E

as of now I have speedstep fix in dsdt (4 steps), which is not working right now because I have disabler (to get low temps)

and absolutely no c-steps fix added to dsdt

but I do have the 1, 2, 3, 4

ACPI_SMC_PlatformPlugin:

[Beerkex'd]

Re: Apple LPC loading, are you referring to the Device LCP fix mentioned earlier? If so, I don't have any reference to that in my DSDT so I assume its a new addition? And does it go anywhere specific?

Yes, you do have it in your DSDT, it just has another name. On Asus boards it's named SBRG, Gigabyte boards uses another name, etc etc.

If you rename it to LPCB you must rename all instances.

 

run LSPCI tools to see the device ID and address of your LPC device (look where it says ISA Bridge), then use this data to search for it in the DSDT.

How can i see if Cstates are found , where do they listed in the IOREGISTRY ?

Not sure if there's more but you should see at least "CSTInfo" under ACPI_SMC_PlatformPlugin.

( don't you have only 3 p-steps in you dsdt, if you have more then it is very strange...)

I have 8 P-states.

http://www.insanelymac.com/forum/topic/181631-dsdt-vanilla-speedstep-generic-scope-pr/page__st__220#entry1272462

Anyone has an iMac9,1 smbios (i read that this works fro Cstate) or at least the main data i put in the smbios ?

Here's mine. I don't think C-states are working 100% since I don't have working shutdown every time. Sometimes the CPU fan keeps running. Also my overall temps are a few degrees higher than before, even though the PStates are working. I'm pretty sure both issues are caused by something not being quite right with my CStates in the DSDT.

 

	<key>SMbiosvendor</key>
<string>Apple Inc.</string>
<key>SMbiosversion</key>
<string>IM91.88Z.008D.B08.0904271717</string>
<key>SMbiosdate</key>
<string>9/21/2009</string>
<key>SMmanufacter</key>
<string>Apple Inc.</string>
<key>SMproductname</key>
<string>iMac9,1</string>
<key>SMsystemversion</key>
<string>1.0</string>
<key>SMserial</key>
<string>xxxxxxxxxxx</string> (find/make your own!)
<key>SMfamily</key>
<string>Mac</string>
<key>SMboardmanufacter</key>
<string>Apple Inc.</string>
<key>SMboardproduct</key>
<string>Mac-F2218FA9</string>

 

About your other question,

 

PStates = CPU throttling according to load

TStates = CPU throttling according to temperature

CStates = Power Saving - see this article http://www.hardwares...com/article/611

 

Also see this mega-excellent post about how to make CStates work:

http://www.insanelym...60#entry1143210

[FKA]

It can be a bag of hurt yes. And I know because I tried it yesterday.

 

gob bless you ... thanks for being the guinea pig!

 

I don't think the 1, 2, 3, 4 are c-ctates, they seems to be p-states instead ( don't you have only 3 p-steps in you dsdt, if you have more then it is very strange...)

 

you are right!

[THe KiNG]

Mostly a style nit, like fixing IASL compiler warnings

 

 

I do not have this problem with Leopard / ChameleonSM (v1.1.12) where it returns:

 

hw.busfrequency = 1332000000

hw.cpufrequency = 2500000000

hw.tbfrequency = 1000000000

hw.tbfrequency: 1000000000

hw.cpufrequency_max: 2500000000

hw.cpufrequency_min: 2500000000

hw.cpufrequency: 2500000000

hw.busfrequency_max: 1332000000

hw.busfrequency_min: 1332000000

hw.busfrequency: 1332000000

 

Let's do some testing here... Done. I checked com.apple.Boot.plist and there I have this:

 

<key>SMmaximalclock</key>

<string>2500</string>

 

And without this snippet – because ChameleonSM won't work with Snow Leopard – the frequencies are wrong/set to 3800000000 (using an unknown processor requires SMBIOS patching).

I have no idea what you say there...

Chameleon and SMBIOS patching works just fine with RC3 version on all my hardware in snow leopard...

 

Back to this topic, a very simple way to dump all table from windows I described HERE

Presence of all table in OS X ioreg is not required, if they are set by default OK(most are wrong) and enabled in BIOS it will work w/o touching them.

Also be aware that smc_platform_plugin has some limitations on P-States nr., if they are less or more then original model(used) it will print an error(fixable with dummy smc_platform.kext)

 

Also for those who can't figure out how to get the value for .5 multi they can use voodoopower.kext and dump the log as described HERE

 

The result will be something like this(cutoff):

"PStates"=
({"DID"=0x0,"VID"=0x22,"FID"=0xc},
{"DID"=0x0,"VID"=0x21,"FID"=0xb},
{"DID"=0x0,"VID"=0x20,"FID"=0xa},
{"DID"=0x0,"VID"=0x1f,"FID"=0x9},
{"DID"=0x0,"VID"=0x1e,"FID"=0x8},
{"DID"=0x0,"VID"=0x1d,"FID"=0x7},
{"DID"=0x0,"VID"=0x1c,"FID"=0x6},
{"DID"=0x0,"VID"=0x1b,"FID"=0x8a},
{"DID"=0x0,"VID"=0x1a,"FID"=0x88},
{"DID"=0x0,"VID"=0x19,"FID"=0x86}),

That is from my T8300 CPU, FID+VID will get you what is needed for .5 multi:

            Name (_PSS, Package (0x0A)
           {
               Package (0x06)
               {
                   0x0960, 
                   0x88B8, 
                   0x0A, 
                   0x0A, 
                   0x0C22, 
                   0x0C22
               }, 

               Package (0x06)
               {
                   0x0898, 
                   0x71EE, 
                   0x0A, 
                   0x0A, 
                   0x0B21, 
                   0x0B21
               }, 

               Package (0x06)
               {
                   0x07D0, 
                   0x6D60, 
                   0x0A, 
                   0x0A, 
                   0x0A20, 
                   0x0A20
               }, 

               Package (0x06)
               {
                   0x0708, 
                   0x5BCC, 
                   0x0A, 
                   0x0A, 
                   0x091F, 
                   0x091F
               }, 

               Package (0x06)
               {
                   0x0640, 
                   0x50BA, 
                   0x0A, 
                   0x0A, 
                   0x081E, 
                   0x081E
               }, 

               Package (0x06)
               {
                   0x0578, 
                   0x459A, 
                   0x0A, 
                   0x0A, 
                   0x071D, 
                   0x071D
               }, 

               Package (0x06)
               {
                   0x04B0, 
                   0x3A98, 
                   0x0A, 
                   0x0A, 
                   0x061C, 
                   0x061C
               }, 

               Package (0x06)
               {
                   0x03E8, 
                   0x37DC, 
                   0x0A, 
                   0x0A, 
                   0x8A1B, 
                   0x8A1B
               }, 

               Package (0x06)
               {
                   0x0320, 
                   0x36B0, 
                   0x0A, 
                   0x0A, 
                   0x881A, 
                   0x881A
               }, 

               Package (0x06)
               {
                   0x0258, 
                   0x32C8, 
                   0x0A, 
                   0x0A, 
                   0x8619, 
                   0x8619
               }
           })

[keeza]

Having looked through quite a few cst tables from various sources the only value in the table that seems to vary is latency and of course the memory address the table is loaded to.

 

Im adding my Cstates to my SSDT now, does that mean that I need to change the CST memory addresses of my FFixedHW as shown in the attached text file (shown in red)?

CST.rtf

[Master Chief]

Thanks i will change that.

Here my IOREGISTRY output (before i changed that CFGD part) , PeformanceStates are loaded/used if i use superhais speedstep.

Or what did you meaned with "do you see the tables" ?

 

I was referring to: Root -> MacPro3,1 -> AppleACPIPlatformExpert -> ACPI Tables

 

Pic shows situation with no AppleINTELCPU and no speedstep loaded+active. CPU runs cool, even without speedstep loaded (MSR shows that it runs always at max speed). less than 40 Celsius idle, 55 Celsius CPU load OC from 2.66 to 3.0 GHZ.

Pic? As in picture? What picture are you talking about? I don't see one, except for the one in your post.

 

How can i see if Cstates are found , where do they listed in the IOREGISTRY ?

There should not be any _CST related errors at startup, and you see a CSTinfo property with IORegistryExplorer - you might want to read the whole thread

 

 

I have no idea what you say there...

Chameleon and SMBIOS patching works just fine with RC3 version on all my hardware in snow leopard...

What I said was pretty clear as in I use ChameleonSM v1.1.12 on Leopard, and that it won't set/select the correct CPU speed. At least not without that snippet in com.apple.Boot.plist. And it also won't work with Snow Leopard.

 

Presence of all table in OS X ioreg is not required...

If that was true, then why won't C-States work on my Mac without them in the ACPI table?

 

Also be aware that smc_platform_plugin has some limitations on P-States nr., if they are less or more then original model(used) it will print an error(fixable with dummy smc_platform.kext)

Interesting. I didn't knew that. Can you please elaborate how this works? Thanks!

 

Also for those who can't figure out how to get the value for .5 multi they can use voodoopower.kext and dump the log as described HERE

 

The result will be something like this...

I wonder if this method will list 0x4720 as maximum for my CPU because 0x8720 did end in a forced reboot (after the KP). See also Superhai's post #156

[FKA]

Since he did get dump and I didn't on the same board, it seems to me that it is "CPU specific" (I think it is more specific to cpu generation rather than exact cpu)

I have a Q6600 and has either a 8xxx or a 9xxx

I don't have C2/C4 option in my bios because of my cpu...

 

Again your quite right (And I should really spend more time reading before I jump in with both feet!)

 

My Intel Q9000 series only supports C1E which is obviously why I have no C0 ,1, 2 etc support in BIOS.

 

Question though - C1E comprises various CPU halt and sleep states which my processor doesn't seem to be entering (unless the whole system is put to sleep.) as CPUi shows the CPU is drawing a minimum of around 1.1mV (lowest p-state.) when left idle.

 

As mitch_de has said the iMac uses C2D that has C1E feature.

 

Does anyboy know if Apple make use of C1E features with the iMac?

 

I think to borrow from iMac is my next step. Thanks mich_de and Eliade for the tip.

 

D.

[THe KiNG]

What I said was pretty clear as in I use ChameleonSM v1.1.12 on Leopard, and that it won't set/select the correct CPU speed. At least not without that snippet in com.apple.Boot.plist. And it also won'\t work with Snow Leopard.

I still don't get why you use such old Chameleon on leopard, why not use latest RC3?

For me freq is reported OK w/o Smbios.plist, that Chameleon version is not designed for snow leopard...

 

If that was true, then why won't C-States work on my Mac without them in the ACPI table?

That assume you have them in ACPI, if not or add them in BIOS or use _CST override.

 

Interesting. I didn't knew that. Can you please elaborate how this works? Thanks!

How to override SMC_Platform_Plugin.kext I described HERE

If you look on original you will find on some models "<key>num-states</key>":

<key>MacBookPro4,1</key>
<array>
<dict>
<key>num-states</key>
<integer>9</integer>
<key>program</key>
<string>SP4</string>
</dict>
</array>

 

I have to change that to:

<key>MacBookPro4,1</key>

<string>SP4</string>

so I can use 10 P-States or less then 9, you can play also with "program" stuff...

 

I wonder if this method will list 0x4720 as maximum for my CPU because 0x8720 did end in a forced reboot (after the KP).

You have one way to find out, try...

 

Good luck!

[FKA]

Back to this topic, a very simple way to dump all table from windows I described HERE

Presence of all table in OS X ioreg is not required, if they are set by default OK(most are wrong) and enabled in BIOS it will work w/o touching them.

Also be aware that smc_platform_plugin has some limitations on P-States nr., if they are less or more then original model(used) it will print an error(fixable with dummy smc_platform.kext)

 

Also for those who can't figure out how to get the value for .5 multi they can use voodoopower.kext and dump the log as described HERE

 

How to override SMC_Platform_Plugin.kext I described HERE

 

Some really interesting stuff!

Many thanks for this THe KiNG!

 

D.

[Superhai]

I wonder if this method will list 0x4720 as maximum for my CPU because 0x8720 did end in a forced reboot (after the KP). See also Superhai's post #156

0x8720 would be odd (and very high voltage for such a low frequency). You can have 0x88 (800MHz on 200MHz FSB) or 0x86 (600MHz on 200MHz FSB) if your CPU support SLFM.

0x47 is multiplier 7.5.

[barnum]

/EDIT

 

I've patched the LPC device ID in my DSDT using zhell's method. Vanilla AppleLPC loads.

 

Hi,

can you give me the method to patch the LPC device ID in DSDT ?

 

Thanks,

barnum

[FKA]

Hi,

can you give me the method to patch the LPC device ID in DSDT ?

 

Thanks,

barnum

 

see zhell's thread http://www.insanelymac.com/forum/index.php?showtopic=168014

[Master Chief]

0x8720 would be odd (and very high voltage for such a low frequency). You can have 0x88 (800MHz on 200MHz FSB) or 0x86 (600MHz on 200MHz FSB) if your CPU support SLFM.

0x47 is multiplier 7.5.

Just what I thought. Thanks.

 

Now a very important note: I somehow smoked a perfectly good functioning Intel Q9300 running @2.5Ghz so everyone here is warned!!!

 

n.b. I have a picture of the BIOS startup screen showing only two cores (I had to boot into Leopard on the same hack first, simply because of the stutter (that's how bad it is). Anyway, enjoy the attached picture while I take a shower and rethink what I did.

 

I was happy to have two of them – thanks to the supplier for making the error – including another P5K PRO motherboard, but I will not continue experimenting. And now I have to fix this craptastic stutter, but I have no idea why, especially since I am using my Leopard DSDT?!? Any pointers as to where I should be looking first?

[FKA]

Just what I thought. Thanks.

 

Now a very important note: I somehow smoked a perfectly good functioning Intel Q9300 running @2.5Ghz so everyone here is warned!!!

 

ouch! do you have any idea what you did wrong? overvolted with the aformention VID ? or the corrupt BIOS you flashed ?

 

D

[Master Chief]

ouch! do you have any idea what you did wrong? overvolted with the aformention VID ? or the corrupt BIOS you flashed ?

 

D

Not really. And that BIOS issue I was having was really my own fault: I used the DVD that came with the motherboard to flash the original BIOS. Which of course was way too old and thus did not support my CPU and thus I had to install two other CPU's to get it going again. That's how old it was.

 

Anyway I was using the latest BIOS (1303) and everything was working just fine. And for some time already. Until I got this KP and had to do a reboot. Which of course failed for only 2 cores. At first I didn't even noticed that it said something about the CPU, but then I did another reboot and noticed the 2 in my boot screen. I checked the BIOS settings and "Max CPUID Value Limit" was disabled so that was not it.

 

My P-State changes are very limited/relaxed so that can't be it (planning to attach them later). I did try The Kings 0x8720 multiplier and got a KP, but that was it. At least that's what I thought. I am however not blaming him nor anyone else but myself for getting into this trouble.

 

Now back to Snow Leopard and fixing the stutter.

[THe KiNG]

...I did try The Kings 0x8720 multiplier and got a KP, but that was it. At least that's what I thought. I am however not blaming him nor anyone else but myself for getting into this trouble.

.....

There is no 0x8720 in what I posted, and the example I posted was from my mobile T8300 that is far from your cpu...

And I advised to use voodoopower.kext debug to get your own p-states/multi, not mine.

Sorry to hear that you get in trouble...

[Master Chief]

There is no 0x8720 in what I posted, and the example I posted was from my mobile T8300 that is far from your cpu...

And I advised to use voodoopower.kext debug to get your own p-states/multi, not mine.

Sorry to hear that you get in trouble...

I know. No worries. And I changed my post (note the bold part) to make clear what I was doing. And it isn't even clear that this caused this specific hardware failure.

 

Guys, I need to reapply all updates to this Snow Leopard installation, but the stress is killing me... so I can't even remember how to do this. What was it again that I have to remove to trigger software update again? HELP!

 

Edit: I restored a backup with Time Machine but still have that stutter problem. Time for a fresh installation, and it appears to be working like before.

[yeehaa]

I know. No worries. And I changed my post (note the bold part) to make clear what I was doing. And it isn't even clear that this caused this specific hardware failure.

 

Guys, I need to reapply all updates to this Snow Leopard installation, but the stress is killing me... so I can't even remember how to do this. What was it again that I have to remove to trigger software update again? HELP!

 

Edit: I restored a backup with Time Machine but still have that stutter problem.

 

Thank you!

 

you are not talking about the Cstate stutter, are you?

i also had a KP yesterday and had to recover the same way from TM. some edit to the DSDT or geekbench or some other {censored}. couldn't figure out what! snow leopard indeed is a beast! (and i forget to restart and check after every single modification i make)

and for my setup i figured out that MacPro3,1 was the safest configuration to use. had to sit and stare at that SMC_platform plist for quite a long time. now the random KPs have stopped (i hope)

[Master Chief]

you are not talking about the Cstate stutter, are you?

Nope. I wiped my Snow Leopard HD and did a fresh install – including the usual updates – and everything appears to be working again.

 

i also had a KP yesterday and had to recover the same way from TM. some edit to the DSDT or geekbench or some other {censored}. couldn't figure out what! snow leopard indeed is a beast! (and i forget to restart and check after every single modification i make)

and for my setup i figured out that MacPro3,1 was the safest configuration to use. had to sit and stare at that SMC_platform plist for quite a long time. now the random KPs have stopped (i hope)

My hack was running without KP's until yesterday. I also had it set to MacPro3,1 but I won't change anything until after Time Machine has finished.

[Squirrelf]

cpui now work without disabler

all I needed to do was to have the mkext alone, I don't know why I kept Extensions along with it for a while, I guess I didn't see any problem (but cpui KP during boot was one of the problem), until I decided to clean the setup a little.

and I finally get nice values in my pstate screen, before th voltage was at max for all steps

 

 

My final setup:

voodoopowermini + disabler is perfect for me

(I don't know if the ssdt in dsdt still do anything, but there still there)

I get full speedstep (speed and voltage are lowered) and low temperature

(without disabler temperature would be in the upper 40s)

[keeza]

Beerkex'd,

 

Does LSPCI happen to be a Linux utility?

I cant seem to locate it for OSX.

 

Care to provide a link?

[Master Chief]

Beerkex'd,

 

Does LSPCI happen to be a Linux utility?

I cant seem to locate it for OSX.

 

Care to provide a link?

 

Not the beer person but that should be OSX86 Tools by p c w i z c o m p u t e r . c o m That to me would have been a helpful link for keeza, but I can't link to it and thus had to mangle it a little.

 

I checked the forum rules and all I can find is this: "The advertising, comparison, or off-topic discussion of sites similar to InsanelyMac will not be allowed. The staff will decide what is and is not "advertising" based on the content and spirit of the post. Helpful links are encouraged!".

 

I'm confused!