Testers needed: CPU Power Management for SB and IB Xeon(s) or i7-39xx on X79 or C60x chipset

omni · January 28, 2014

Yes, but egos get in the way and next thing you know someone gets hurt.

And I should know, I was in the hackint0sh community from the very beginning so I've seen people burn and crash. At one point there was even a huge divide between two groups/forums, a lot of drama and what not.

shilohhh · January 28, 2014

Hey omni, I noticed you crossed off the Xeon E5-v2 invitation because they are getting native stepping w beta 13C44. I think the 4930K and such are in the same boat but I'm not positive.

I'd love to be of more help but I lack the knowledge. If its not a pain in your butt, I'd love to know more about your workflow, what tools you're using and what exactly you've done to get the results you've achieved thus far. Do you think you'll make a tool or script to auto patch future versions of AICPUPM? Please let me know if there is anything I can do to be of more use at this point.

omni · January 28, 2014

I don't have a IB v2 CPU so I made a guesstimate.

The assumption is that because IB-EP Xeon v2 and the IB-E i7-49x0 are the same model numbers (0x3e), AICPM already has the necessary code to detect that model and do power management.

Seeing that you have a SB-E, you are in the same boat with me (model 0x2d) and need a patched AICPM to get anything. I've been testing various platforms and it seems that using ACPI_SMC is better than X86Platform with regards to stepping to lower states. Although more testing is needed to see whether it's the platform or the ACPI_SMC vs X86Platform that does it.

What I don't like about ACPI_SMC is that it appears to be going through SMC to manage states and since we don't have a real SMC that could be a problem. More testing is needed and I think I'm going to badger netkas about SMC as well.

Also, I will be posting something new soon, just need to do more testing and add a bit more code to it.

jasjeet · January 28, 2014

Ive checked in CPUz and my 4960X is also Model 3E, but no native PM as far as i can tell. Im using HWmonitor to see the Multiplier as i cant get the AICPMI kext to work.

Heres my startup log using Mac Mini 6,1, it gives locked x36 Multi, though voltage steps to 0.8v.

Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=0 LocalApicId=0 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=2 LocalApicId=2 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=4 LocalApicId=4 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=6 LocalApicId=6 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=8 LocalApicId=8 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=10 LocalApicId=10 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=1 LocalApicId=1 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=3 LocalApicId=3 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=5 LocalApicId=5 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=7 LocalApicId=7 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=9 LocalApicId=9 Enabled
Jan 28 11:56:16 localhost kernel[0]: AppleACPICPU: ProcessorId=11 LocalApicId=11 Enabled
Jan 28 11:56:16 localhost kernel[0]: pci (build 20:00:24 Jan 16 2014), flags 0x61008, pfm64 (44 cpu) 0xfff80000000, 0x80000000
Jan 28 11:56:16 localhost kernel[0]: AppleIntelCPUPowerManagement: Turbo Ratios 112334
Jan 28 11:56:16 localhost kernel[0]: AppleIntelCPUPowerManagement: (built 19:46:50 Jan 16 2014) initialization complete
Jan 28 11:56:16 localhost kernel[0]: CPUSensors: [Fatal] found unsupported Intel processor, using default Tjmax
Jan 28 11:56:16 localhost kernel[0]: CPUSensors: CPU family 0x6, model 0x3e, stepping 0x4, cores 6, threads 12, TJmax 0
Jan 28 11:56:16 localhost kernel[0]: CPUSensors: base CPU multiplier is 36
Jan 28 11:56:16 localhost kernel[0]: CPUSensors: RAPL units power: 0x3 energy: 0x10 time: 0xa
Jan 28 11:56:16 localhost kernel[0]: mcache: 12 CPU(s), 64 bytes CPU cache line size
Jan 28 11:56:16 localhost kernel[0]: AppleIntelCPUPowerManagementClient: ready
Jan 28 11:56:17 Jasjeets-Mac-Pro kernel[0]: SMC::smcPublishLightshowVersion ERROR: smcGetLightshowVers failed (kSMCKeyNotFound)
Jan 28 11:56:17 Jasjeets-Mac-Pro kernel[0]: SMC::smcInitHelper ERROR: smcPublishLightshowVersion failed (kSMCKeyNotFound)
Jan 28 11:56:18 Jasjeets-Mac-Pro kernel[0]: WARNING: IOPlatformPluginUtil : getCPUIDInfo: this is an unknown CPU model 0x3e
Jan 28 11:56:18 Jasjeets-Mac-Pro kernel[0]: ACPI_SMC_PlatformPlugin::pushCPU_CSTData - _CST evaluation failed
Jan 28 11:56:20 Jasjeets-Mac-Pro kernel[0]: ACPI_SMC_PlatformPlugin::pushCPU_CSTData - _CST evaluation failed

Using 6.1 mac Pro SMBIOS, i get the same thing, locked x36 Multi.

Using Mac Pro 5,1 SMBIOS i get x36, x33, x12 Multi's.

All tests done with the CPU type 2561 in SMBIOS, but Mac Pro 5,1 1281.

SSDT from RamapageDev.

I dont understand this part of OP

"you are using either X86Platform or ACPI_SMC_Platform", can somebody elaborate?

stinga11 · January 28, 2014

Files:

~~Patched AppleIntelCPUPowerManagement v001: http://www.putlocker.com/file/773B0DD943BC0F94~~

~~Patched AppleIntelCPUPowerManagement v002 (SB+IB): http://www.putlocker.com/file/31C44ABB95501C61~~

Patched AppleIntelCPUPowerManagement v003 (SB+IB): http://www.putlocker.com/file/4F9CFD5146AA3569

Hey man, what patch you use on the appleintelpowermanagment?

magnifico · January 28, 2014

Hey man, what patch you use on the appleintelpowermanagment?

v003 of course

jasjeet · January 28, 2014

I generated my SSDT with C State info, then added RampageDevs edits to it.

I used original AICPM kext

I used SMBIOS Mac Pro 5.1

Here is the output from MSRDumper, is it working?

28/01/2014 17:41:36.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:36.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:37.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:37.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:37.000 kernel[0]: MSRDumper CoreMulti(39) 
28/01/2014 17:41:37.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:38.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:38.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:38.000 kernel[0]: MSRDumper CoreMulti(37) 
28/01/2014 17:41:38.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:39.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:39.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:39.000 kernel[0]: MSRDumper CoreMulti(38) 
28/01/2014 17:41:39.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:40.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:40.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:40.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:40.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:41.000 kernel[0]: MSRDumper CoreMulti(12)

magnifico · January 28, 2014

I generated my SSDT with C State info, then added RampageDevs edits to it.

I used original AICPM kext

I used SMBIOS Mac Pro 5.1

Here is the output from MSRDumper, is it working?

28/01/2014 17:41:36.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:36.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:37.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:37.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:37.000 kernel[0]: MSRDumper CoreMulti(39) 
28/01/2014 17:41:37.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:38.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:38.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:38.000 kernel[0]: MSRDumper CoreMulti(37) 
28/01/2014 17:41:38.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:39.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:39.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:39.000 kernel[0]: MSRDumper CoreMulti(38) 
28/01/2014 17:41:39.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:40.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:40.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:40.000 kernel[0]: MSRDumper CoreMulti(12) 
28/01/2014 17:41:40.000 kernel[0]: MSRDumper PStatesReached: 12 16 37 38 39 
28/01/2014 17:41:41.000 kernel[0]: MSRDumper CoreMulti(12)

without overclock?

jasjeet · January 28, 2014

Yes no overclock. My CPU is 4960X maximum turbo 4Ghz.

In IOReg i have 12 P States per CPU Core and 3 C States per CPU Core under X86PlatformPlugin.

stinga11 · January 28, 2014

Hey man, what patch you use on the appleintelpowermanagment?

v003 of course

hahaha Asked because I compare the original aicp with patched aicp v003 and see that have nine hex patch. Is only for curiosity and for the future version of mac os x

omni · January 28, 2014

@jasjeet

Please read my OP (Original Post) if you wish to participate in this test/thread.

I cannot, nor will, support other people's hacks, as I do not have the time to invest into following what or how they are doing whatever they are doing in detail, that's their business.

Just to refresh your memory, from my OP:

- we are NOT using MSRDumper but AppleIntelCPUPowerManagementInfo.

- most people use Pike's ssdtPRgen.sh script to make their SSDT, but I'm a proponent of writing own based on existing OEM SSDT. I am NOT encouraging nor condoning use of an "special" SSDT edits done by others because that skews the test base and puts everyone off the "main" page.

If you are doing any of those things, or others not mentioned in my OP, please don't ask questions about them here as you are steering this thread off-topic. Start your own thread and ask questions with it.

Thanks.

jasjeet · January 28, 2014

I cant get AICPMI kext to load it causes a KP, hence why i tried MSRDumper. Try to explain things so more people can help. For example is there a particular way to load this kext, maybe im making a an error here, installing to S/L/E gives KP on bootup and using kextload it either KP's or locks up the system. You have written 2 replies to me which have allowed me to progress 0%. I have been researching many things on your OP which could of been clarified on your first reply. There are probably very simple reasons im having a bit of trouble getting started here.

If that is too much trouble, then just say.

magnifico · January 28, 2014

omni...listen me please..i use clover boot..witout ssdt table in acpi /patched folder of clover...and i use mac pro 5.1 smbios...

It seems that everything is working?

kundica · January 28, 2014

I cant get AICPMI kext to load it causes a KP, hence why i tried MSRDumper. Try to explain things so more people can help. For example is there a particular way to load this kext, maybe im making a an error here, installing to S/L/E gives KP on bootup and using kextload it either KP's or locks up the system. You have written 2 replies to me which have allowed me to progress 0%. I have been researching many things on your OP which could of been clarified on your first reply. There are probably very simple reasons im having a bit of trouble getting started here.

If that is too much trouble, then just say.

You need to compile AICPUPMI from Pike's source if it KPs for you. It's fairly easy to do, you just need Xcode installed. Pike's source directory contains an xcode file so just open that then chose to build the kext. You might want to change the ~~value of igpu~~ to 0 (Fabiosun lists the setting you need to change above).

Regarding the other SSDT that you're using... it doesn't define stepping, just injects device ids, etc. For proper stepping you need to generate/build an ssdt with pikes script, making sure that plugin-type 1 is in it. I'm not 100% about generating and SSDT with pikes script for your 4960x cpu, but if it should look something like this for stock settings:

./ssdtPRGen.sh 'i7-4960K' 4000 130 1

I think you cpu is listed as K not X in pikes script.

Edited January 28, 2014 by kundica

jasjeet · January 28, 2014

Excellent guys, now i can get AICPMI working.

kundica · January 28, 2014

Kundica, could you elaborate plugin type =1?

in launching sh script? or inside code?

I have this but I don't know if it is useful

Method (_DSM, 4, NotSerialized)

{

Store ("Method C000._DSM Called", Debug)

If (LEqual (Arg2, Zero))

{

Return (Buffer (One)

{

0x03

})

}

Return (Package (0x02)

{

"plugin-type",

One

})

}

That's it. It should be generated in the SSDT when you run pikes script for ivy bridge cpu.

magnifico · January 28, 2014

- we are NOT using MSRDumper but AppleIntelCPUPowerManagementInfo.

omni can you attached also new ACPMI kext ?

this is a strange for me

sh-3.2# cd desktop
sh-3.2# sudo sh
sh-3.2# sudo chown -R root:wheel AppleIntelCPUPowerManagementInfo.kext
sh-3.2# sudo chmod -R 755 AppleIntelCPUPowerManagementInfo.kext
sh-3.2# sudo touch desktop
sh-3.2# sudo -k
sh-3.2# kextload AppleIntelCPUPowerManagementInfo.kext
sh-3.2# cat /var/log/system.log | grep "AICPUPMI:"
Jan 28 20:34:20 pro-di-carmine kernel[0]: AICPUPMI: CPU P-States [ 47 ]
Jan 28 20:34:21 pro-di-carmine kernel[0]: AICPUPMI: CPU P-States [ 12 47 ]

whyle IPG read another

reading, did you see my screen?

omni · January 28, 2014

@jasjeet,

My OP is very clear what to do so I won't quote myself from it. Please read it more closely about AICPMI. Once again, if you are having problems with it, that's an issue for Pike, not me, I just require you use it for this test.

@magnifico,

Please do dmesg instead of cat /var/log/system.log as not everything ends up in the system log. The above screenshots appear to indicate that you are reaching quite a number of P-states. Not sure about your scores, but 4,000+ for a single core appears that you are running Turbo Boost as well.

magnifico · January 28, 2014

omni i do

cat /var/log/system.log | grep "AICPUPMI:"

but i see only two state again

Jan 28 20:34:20 pro-di-carmine kernel[0]: AICPUPMI: CPU P-States [ 47 ]
Jan 28 20:34:21 pro-di-carmine kernel[0]: AICPUPMI: CPU P-States [ 12 47 ]

mhmmm :no:

yangshun1029 · January 28, 2014

Here's what I get in my kernel log:

kernel[0]: AppleIntelCPUPowerManagement: Turbo Ratios 1122
kernel[0]: AppleIntelCPUPowerManagement: (built 20:35:09 Oct 17 2013) initialization complete
kernel[0]: SMC: successfully initialized
kernel[0]: AppleIntelCPUPowerManagementClient: ready

kernel[0]: X86PlatformPlugin::init() - kIOPPFBootArgs boot-args set debugLevel 3, debugFlags 3
kernel[0]: X86PlatformPlugin::start - fPanicErrors 0
kernel[0]: X86PlatformPlugin::startDelayedPlatformDatabaseRead - Mac-F60DEB81FF30ACF6.plist
kernel[0]: X86PlatformPlugin::getCPUPStates - P 0: coreFreq 3800, power 130000, latency 10, busLatency 10, control 0x2600, status 0x2600
kernel[0]: X86PlatformPlugin::getCPUPStates - P 1: coreFreq 3700, power 130000, latency 10, busLatency 10, control 0x2500, status 0x2500
kernel[0]: X86PlatformPlugin::getCPUPStates - P 2: coreFreq 3600, power 130000, latency 10, busLatency 10, control 0x2400, status 0x2400
kernel[0]: X86PlatformPlugin::getCPUPStates - P 3: coreFreq 3500, power 124956, latency 10, busLatency 10, control 0x2300, status 0x2300
kernel[0]: X86PlatformPlugin::getCPUPStates - P 4: coreFreq 3400, power 120003, latency 10, busLatency 10, control 0x2200, status 0x2200
kernel[0]: X86PlatformPlugin::getCPUPStates - P 5: coreFreq 3300, power 115138, latency 10, busLatency 10, control 0x2100, status 0x2100
kernel[0]: X86PlatformPlugin::getCPUPStates - P 6: coreFreq 3200, power 110362, latency 10, busLatency 10, control 0x2000, status 0x2000
kernel[0]: X86PlatformPlugin::getCPUPStates - P 7: coreFreq 3100, power 105674, latency 10, busLatency 10, control 0x1f00, status 0x1f00
kernel[0]: X86PlatformPlugin::getCPUPStates - P 8: coreFreq 3000, power 101072, latency 10, busLatency 10, control 0x1e00, status 0x1e00
kernel[0]: X86PlatformPlugin::getCPUPStates - P 9: coreFreq 2900, power 96557, latency 10, busLatency 10, control 0x1d00, status 0x1d00
kernel[0]: X86PlatformPlugin::getCPUPStates - P10: coreFreq 2800, power 92127, latency 10, busLatency 10, control 0x1c00, status 0x1c00
kernel[0]: X86PlatformPlugin::getCPUPStates - P11: coreFreq 2700, power 87783, latency 10, busLatency 10, control 0x1b00, status 0x1b00
kernel[0]: X86PlatformPlugin::getCPUPStates - P12: coreFreq 2600, power 83522, latency 10, busLatency 10, control 0x1a00, status 0x1a00
kernel[0]: X86PlatformPlugin::getCPUPStates - P13: coreFreq 2500, power 79345, latency 10, busLatency 10, control 0x1900, status 0x1900
kernel[0]: X86PlatformPlugin::getCPUPStates - P14: coreFreq 2400, power 75251, latency 10, busLatency 10, control 0x1800, status 0x1800
kernel[0]: X86PlatformPlugin::getCPUPStates - P15: coreFreq 2300, power 71238, latency 10, busLatency 10, control 0x1700, status 0x1700
kernel[0]: X86PlatformPlugin::getCPUPStates - P16: coreFreq 2200, power 67308, latency 10, busLatency 10, control 0x1600, status 0x1600
kernel[0]: X86PlatformPlugin::getCPUPStates - P17: coreFreq 2100, power 63457, latency 10, busLatency 10, control 0x1500, status 0x1500
kernel[0]: X86PlatformPlugin::getCPUPStates - P18: coreFreq 2000, power 59687, latency 10, busLatency 10, control 0x1400, status 0x1400
kernel[0]: X86PlatformPlugin::getCPUPStates - P19: coreFreq 1900, power 55996, latency 10, busLatency 10, control 0x1300, status 0x1300
kernel[0]: X86PlatformPlugin::getCPUPStates - P20: coreFreq 1800, power 52384, latency 10, busLatency 10, control 0x1200, status 0x1200
kernel[0]: X86PlatformPlugin::getCPUPStates - P21: coreFreq 1700, power 48849, latency 10, busLatency 10, control 0x1100, status 0x1100
kernel[0]: X86PlatformPlugin::getCPUPStates - P22: coreFreq 1600, power 45392, latency 10, busLatency 10, control 0x1000, status 0x1000
kernel[0]: X86PlatformPlugin::getCPUPStates - P23: coreFreq 1500, power 42011, latency 10, busLatency 10, control 0xf00, status 0xf00
kernel[0]: X86PlatformPlugin::getCPUPStates - P24: coreFreq 1400, power 38706, latency 10, busLatency 10, control 0xe00, status 0xe00
kernel[0]: X86PlatformPlugin::getCPUPStates - P25: coreFreq 1300, power 35476, latency 10, busLatency 10, control 0xd00, status 0xd00
kernel[0]: X86PlatformPlugin::getCPUPStates - P26: coreFreq 1200, power 32320, latency 10, busLatency 10, control 0xc00, status 0xc00
kernel[0]: X86PlatformPlugin::publishACPIStates - Succeeded in getting max non-turbo PState 2
kernel[0]: X86PlatformPlugin::publishACPIStates - Succeeded in getting max efficient PState ratio difference 1
kernel[0]: X86PlatformPlugin::publishACPIStates - Found max efficient PState to be 25
kernel[0]: IOPPF: AppleIntelCPUPowerManagement mode
kernel[0]: X86PlatformPlugin::cgPublishCPUStates - Got all the CPU callbacks!
kernel[0]: X86PlatformPlugin::configResourceHandler - setting Thermal Profile from resource
kernel[0]: X86PlatformPlugin::cgConfigResourceHandler - waiting for commandWake on num proccessors to be 0: 0
kernel[0]: X86PlatformPlugin::setRingTable - AICPM failed to load ring table with status 0x0: Get=0, Load=0, Install=0
kernel[0]: X86PlatformPlugin::configResourceHandler - Failed to set ring table!

kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[0]: vid 0x0, voltage 7000, frequency 3800, multiplier 038
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[1]: vid 0x0, voltage 7000, frequency 3700, multiplier 037
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[2]: vid 0x0, voltage 7000, frequency 3600, multiplier 036
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[3]: vid 0x0, voltage 7000, frequency 3500, multiplier 035
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[4]: vid 0x0, voltage 7000, frequency 3400, multiplier 034
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[5]: vid 0x0, voltage 7000, frequency 3300, multiplier 033
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[6]: vid 0x0, voltage 7000, frequency 3200, multiplier 032
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[7]: vid 0x0, voltage 7000, frequency 3100, multiplier 031
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[8]: vid 0x0, voltage 7000, frequency 3000, multiplier 030
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[9]: vid 0x0, voltage 7000, frequency 2900, multiplier 029
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[10]: vid 0x0, voltage 7000, frequency 2800, multiplier 028
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[11]: vid 0x0, voltage 7000, frequency 2700, multiplier 027
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[12]: vid 0x0, voltage 7000, frequency 2600, multiplier 026
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[13]: vid 0x0, voltage 7000, frequency 2500, multiplier 025
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[14]: vid 0x0, voltage 7000, frequency 2400, multiplier 024
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[15]: vid 0x0, voltage 7000, frequency 2300, multiplier 023
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[16]: vid 0x0, voltage 7000, frequency 2200, multiplier 022
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[17]: vid 0x0, voltage 7000, frequency 2100, multiplier 021
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[18]: vid 0x0, voltage 7000, frequency 2000, multiplier 020
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[19]: vid 0x0, voltage 7000, frequency 1900, multiplier 019
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[20]: vid 0x0, voltage 7000, frequency 1800, multiplier 018
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[21]: vid 0x0, voltage 7000, frequency 1700, multiplier 017
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[22]: vid 0x0, voltage 7000, frequency 1600, multiplier 016
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[23]: vid 0x0, voltage 7000, frequency 1500, multiplier 015
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[24]: vid 0x0, voltage 7000, frequency 1400, multiplier 014
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[25]: vid 0x0, voltage 7000, frequency 1300, multiplier 013
kernel[0]: X86PlatformShim::setPStateEntryInfoAICPM - PState[26]: vid 0x0, voltage 7000, frequency 1200, multiplier 012

kernel[0]: X86PlatformPlugin::publishACPIStates - Succeeded in getting max non-turbo PState 2
kernel[0]: X86PlatformPlugin::publishACPIStates - Succeeded in getting max efficient PState ratio difference 1
kernel[0]: X86PlatformPlugin::publishACPIStates - Found max efficient PState to be 25
kernel[0]: X86PlatformShim::start - registering pLimitNotificationCallback for cpu plimits
kernel[0]: X86PlatformPlugin::setPLimitRange - CPU Limit Range [0, 26]
kernel[0]: X86PlatformPlugin::cgDispatchCallback - no initial cpu-plimit-notification plimit value sent
kernel[0]: X86PlatformShim::start - registering pLimitNotificationCallback for cpu idle plimitskernel[0]: X86PlatformPlugin::setPLimitRange - Force Idle Limit Range [0, 100]
kernel[0]: X86PlatformPlugin::cgDispatchCallback - no initial cpu-idle-plimit-notification plimit value sent

And use the AppleIntelCPUPowerManagementInfo.kext

14-1-29 14:50:40.000 kernel[0]: AICPUPMI: v3.3 Copyright © 2012-2014 Pike R. Alpha. All rights reserved
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: logMSRs............................: 1
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: logIGPU............................: 0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: logCStates.........................: 1
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: logIPGStyle........................: 1
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MWAIT C-States.....................: 135456
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_CORE_THREAD_COUNT......(0x35)  : 0x40008
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PLATFORM_INFO..........(0xCE)  : 0xC0070012400
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PMG_CST_CONFIG_CONTROL.(0xE2)  : 0x1E000407
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PMG_IO_CAPTURE_BASE....(0xE4)  : 0x10414
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: IA32_MPERF.................(0xE7)  : 0xEBA4EE5AC
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: IA32_APERF.................(0xE8)  : 0x9810C8C57
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_FLEX_RATIO.............(0x194) : 0xC0000
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_IA32_PERF_STATUS.......(0x198) : 0x273B00002400
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_IA32_PERF_CONTROL......(0x199) : 0x2400
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: IA32_CLOCK_MODULATION......(0x19A) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: IA32_THERM_STATUS..........(0x19C) : 0x88380000
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: IA32_MISC_ENABLES..........(0x1A0) : 0x850089
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_MISC_PWR_MGMT..........(0x1AA) : 0x400001
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_TURBO_RATIO_LIMIT......(0x1AD) : 0x2525252525252626
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: IA32_ENERGY_PERF_BIAS......(0x1B0) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_POWER_CTL..............(0x1FC) : 0x2504005B
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_RAPL_POWER_UNIT........(0x606) : 0xA1003
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKG_POWER_LIMIT........(0x610) : 0x68960005AFFFF
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKG_ENERGY_STATUS......(0x611) : 0x29977BB
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKGC3_IRTL.............(0x60a) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKGC6_IRTL.............(0x60b) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKGC7_IRTL.............(0x60c) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PP0_CURRENT_CONFIG.....(0x601) : 0x14149480000640
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PP0_POWER_LIMIT........(0x638) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PP0_ENERGY_STATUS......(0x639) : 0x1DEF427
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PP0_POLICY.............(0x63a) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKG_C2_RESIDENCY.......(0x60d) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKG_C3_RESIDENCY.......(0x3f8) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKG_C6_RESIDENCY.......(0x3f9) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: MSR_PKG_C7_RESIDENCY.......(0x3fa) : 0x0
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: CPU Low Frequency Mode.............: 1200 MHz
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: CPU Maximum non-Turbo Frequency....: 3600 MHz
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: CPU Maximum Turbo Frequency........: 3800 MHz
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: CPU P-States [ (12) 35 36 ]
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: CPU P-States [ 12 27 35 (36) ]
14-1-29 14:50:40.000 kernel[0]: AICPUPMI: CPU P-States [ (12) 25 27 35 36 ]
14-1-29 14:50:41.000 kernel[0]: AICPUPMI: CPU P-States [ (12) 25 27 34 35 36 ]
14-1-29 14:50:44.000 kernel[0]: AICPUPMI: CPU P-States [ 12 25 27 34 35 36 (37) ]
14-1-29 14:50:44.000 kernel[0]: AICPUPMI: CPU C3-Cores [ 2 3 6 7 ]
14-1-29 14:50:44.000 kernel[0]: AICPUPMI: CPU C6-Cores [ 2 3 5 6 ]
14-1-29 14:50:44.000 kernel[0]: AICPUPMI: CPU P-States [ 12 25 27 31 34 35 36 (37) ]
14-1-29 14:50:44.000 kernel[0]: AICPUPMI: CPU C3-Cores [ 2 3 4 6 7 ]
14-1-29 14:50:44.000 kernel[0]: AICPUPMI: CPU C6-Cores [ 2 3 4 5 6 7 ]
14-1-29 14:50:45.000 kernel[0]: AICPUPMI: CPU C3-Cores [ 0 2 3 4 5 6 7 ]
14-1-29 14:50:45.000 kernel[0]: AICPUPMI: CPU C6-Cores [ 1 2 3 4 5 6 7 ]
14-1-29 14:50:46.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (24) 25 27 31 34 35 36 37 ]
14-1-29 14:50:46.000 kernel[0]: AICPUPMI: CPU C3-Cores [ 0 1 2 3 4 5 6 7 ]
14-1-29 14:50:46.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (13) 24 25 27 31 34 35 36 37 ]
14-1-29 14:50:47.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (13) 22 24 25 27 31 34 35 36 37 ]
14-1-29 14:50:47.000 kernel[0]: AICPUPMI: CPU C6-Cores [ 0 1 2 3 4 5 6 7 ]
14-1-29 14:50:51.000 kernel[0]: AICPUPMI: CPU P-States [ 12 13 22 24 25 27 31 32 34 35 36 (37) ]
14-1-29 14:50:55.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (13) 22 24 25 26 27 31 32 34 35 36 37 ]
14-1-29 14:50:56.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (13) 20 22 24 25 26 27 31 32 34 35 36 37 ]
14-1-29 14:50:58.000 kernel[0]: AICPUPMI: CPU P-States [ (12) 13 20 22 24 25 26 27 31 32 33 34 35 36 37 ]
14-1-29 14:50:58.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (13) 20 22 24 25 26 27 29 31 32 33 34 35 36 37 ]
14-1-29 14:51:01.000 kernel[0]: AICPUPMI: CPU P-States [ (12) 13 20 22 23 24 25 26 27 29 31 32 33 34 35 36 37 ]
14-1-29 14:51:02.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (13) 20 21 22 23 24 25 26 27 29 31 32 33 34 35 36 37 ]
14-1-29 14:51:02.000 kernel[0]: AICPUPMI: CPU P-States [ 12 13 20 21 22 23 24 25 26 27 29 (30) 31 32 33 34 35 36 37 ]
14-1-29 14:51:03.000 kernel[0]: AICPUPMI: CPU P-States [ 12 13 19 20 21 22 23 24 25 26 27 29 30 31 32 33 34 35 36 (37) ]
14-1-29 14:51:03.000 kernel[0]: AICPUPMI: CPU P-States [ 12 13 19 20 21 22 23 24 25 26 27 29 30 31 32 33 34 35 36 37 (38) ]
14-1-29 14:51:05.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (13) 18 19 20 21 22 23 24 25 26 27 29 30 31 32 33 34 35 36 37 38 ]
14-1-29 14:51:06.000 kernel[0]: AICPUPMI: CPU P-States [ 12 (13) 17 18 19 20 21 22 23 24 25 26 27 29 30 31 32 33 34 35 36 37 38 ]
14-1-29 14:51:06.000 kernel[0]: AICPUPMI: CPU P-States [ (12) 13 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 ]
14-1-29 14:51:07.000 kernel[0]: AICPUPMI: CPU P-States [ (12) 13 14 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 ]
14-1-29 14:51:08.000 kernel[0]: AICPUPMI: CPU P-States [ (12) 13 14 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 ]
14-1-29 14:51:13.000 kernel[0]: AICPUPMI: CPU P-States [ 12 13 14 15 16 17 18 19 20 21 22 23 (24) 25 26 27 28 29 30 31 32 33 34 35 36 37 38 ]

There are any error need to fixed?

kundica · January 28, 2014

omni i do

cat /var/log/system.log | grep "AICPUPMI:"

but i see only two state again

Jan 28 20:34:20 pro-di-carmine kernel[0]: AICPUPMI: CPU P-States [ 47 ]
Jan 28 20:34:21 pro-di-carmine kernel[0]: AICPUPMI: CPU P-States [ 12 47 ]

mhmmm :no:

I believe that's all you'll get with 3930k when you OC. I only get 2 unless I run stock, in which case, I also get 32 and states for each turbo multiplier that's not 38. IPG works differently than AICPUPMI so don't expect them to yield the same exact results.

magnifico · January 28, 2014

I believe that's all you'll get with 3930k when you OC. I only get 2 unless I run stock, in which case, I also get 32 and states for each turbo multiplier that's not 38. IPG works differently than AICPUPMI so don't expect them to yield the same exact results.

so I have to trust only IPG ?

For you work perfect my hack?

Riley Freeman · January 28, 2014

so I have to trust only IPG ?

For you work perfect my hack?

I just did a fresh install of 10.9.2 with MacPro5,1 SMBIOS. Replaced the AICPUPM binary with omni's patched SB one and I'm getting the same behaviour here with my 3930k.

HWMonitor shows the CPU fluctuating between 12x, 32x and 38x but the Intel Power Gadget shows various speed states (and nothing close to 32x or 38x). Using my stock SSDT and Chameleon.

I'd be inclined to trust IPG as I assume it's reading the hardware directly. But perhaps someone can prove this is the case.

Also, HWMonitor's CPU voltage reading matches IPG's. So while the CPU states might be getting reported incorrectly, it looks like the CPU is speedstepping correctly.

magnifico · January 28, 2014

I just did a fresh install of 10.9.2 with MacPro5,1 SMBIOS. Replaced the AICPUPM binary with omni's patched SB one and I'm getting the same behaviour here with my 3930k.

HWMonitor shows the CPU fluctuating between 12x, 32x and 38x but the Intel Power Gadget shows various speed states (and nothing close to 32x or 38x). Using my stock SSDT and Chameleon.

I'd be inclined to trust IPG as I assume it's reading the hardware directly. But perhaps someone can prove this is the case.

Also, HWMonitor's CPU voltage reading matches IPG's. Così, mentre gli stati della CPU potrebbero essere sempre riportati in modo errato, sembra che la CPU sia speedstepping correttamente.

hi friend how are you?

so for you all work perfect ?

Friend i dont use ssdt table...i use only clover boot and no dsdt only table bios

Riley Freeman · January 28, 2014

hi friend how are you?

so for you all work perfect ?

Friend i dont use ssdt table...i use only clover boot and no dsdt only table bios

At the hardware level it seems to be. I disabled C1E as a test and it still stepped up and down.

By stock SSDT I mean I'm not using a SSDT override either. It's just using the one in my BIOS.

So basically it's working much the same as if we were still using NullCPUPM and C1E. But instead of loading a hack kext, we're patching an Apple one.

It's good progress so far but looking at omni's patched binary it's quite involved. So if Apple don't add support for SB-E chips later it might not be practical to have to keep patching AICPUPM.

Testers needed: CPU Power Management for SB and IB Xeon(s) or i7-39xx on X79 or C60x chipset

Recommended Posts

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites