GeneratePStates on E5-2600

[celstark]

I've got a dual E5-2600 based machine, generally up and going very nicely. The issue so far is that there is no TurboBoost (no enhanced P-states) under OS X (there in Linux).

This AM, I grabbed the current SVN Chimera, which I see supports the new processors (Thanks!). I compiled it up and copied sym/boot to /boot (is that all that's needed to upgrade Chimera?) and I've got GenerateCStates and GeneratePStates = Yes in org.Chameleon.Boot.plist.

Still, no dice. Now, I do need to run NullCPUPowerManagement. If I pass in -allowAppleCPUPM on boot, I get to the grey screen for a sec and then reboot hard. I'm still not entirely clear on how GeneratePStates operates with NullCPU and such.

I'm oh-so-close to having this 16-core machine really humming. Any guidance here would be greatly appreciated.

 

[MacMan]

You didn't need to compile Chimera, I always post an installer for the latest version in the Downloads section.

NullCPUPM disables Apple power management so you won't get TurboBoost. The only other thing I can think of is for you to try the flAKed SpeedStepper and then remove NullCPU.

 

[celstark]

Thanks for the advice. I'd already tried the Perl script (AICPMPatch) patch but tried this as well to no avail. The boot goes by quickly, but on video I've captured
SSDT with CPU C-States generated successfully
P-States: in 12, max 26
Number of P-states generated: 15
SSDT with CPU P-states generated successfully
RSDT: Added 2 SSDT table(s)
FADT: Restart Fix applied!
Added 2 SSDT tables(s) to XSDT
Starting Darwin x86_64
Boot Args: ....
...
AppleACPICPU: ProcessorId= (lists them all with Local nubers as enabled)
...
IOAPIC: Version 0x20 Vectors 64:87
IOAPIC: Version 0x20 Vectors 88:111
IOAPIC: Version 0x20 Vectors 112:135
ACPI: System State [S0 S4 S5] (S0)
PFM64 (44 cpu) 0xff10000000, 0xf000000
RTC: Only single RAM bank 128 bytes)
[ PCI configuration begin ]
FakeSMC: opensource SMC device emulator by netkas (C) 2009
FakeSMC: pl

we never get the rest of that. With NullCPUPM in there we have a few PCI configuration changed bits (3 of them). So, we're finding some P-states at least.