Guide To Install Mavericks with Clover Bootloader

[custom-mac]

Note: You do need a relatively recent version of Clover. I tested with r2372.

Did you mean CC 2352? I was able to switch to Kpm following your instructions. I also added it to AddArguments in custom Entry for Mac OS X, so I do not need to type it on boot.

 

[RehabMan]

Did you mean CC 2352? I was able to switch to Kpm following your instructions.

KernelPm only works in recent versions of Clover. I'm not sure of the exact version when it started working. When I wrote "Clover r2372," I meant "Clover r2372." You can set the option all you want in your config.plist, but if it isn't implemented correctly in the version of Clover you're using, your laptop will simply reboot on attempting to boot Mavericks/ML. It is not a big deal (until you set -xcpm in config.plist), just boot without -xcpm.

 

[custom-mac]

I could confirm - it is working on Clover r2352 (available to general public). Thanks for patch and other clarifications. I think kpm added additional state 18 and 24 to my otherwise very limited set of states on i3.

 

[RehabMan]

I could confirm - it is working on Clover r2352 (available to general public). Thanks for patch and other clarifications. I think kpm added additional state 18 and 24 to my otherwise very limited set of states on i3.

I build clover from source, which is probably the discrepancy/confusion in versions...

Note: Using xcpm on Ivy hardware should be considered experimental. One would think there is a reason Apple did not enable it for their own Ivy hardware and instead left it disabled behind an undocumented kernel flag.

 

[nguyenmac]

Nice, I also get some more states in between after switching to XCPM and a strange state :/

Here is my log with DPCI Manager PStates:

AICPM
P States: 12, 22, 26, 28, 29, 30, 31
XCPM
P States: 12, 19, 22, 26, 28, 29, 31, 87

 

[rob1980]

whilst getting those extra p states, is there a noticeable difference in performance?

Has anyone ran some benchmark software to test?

Or am I missing the point of what this could / does achieve?

thanks

 

[pwd]

whilst getting those extra p states, is there a noticeable difference in performance?

Has anyone ran some benchmark software to test?

Or am I missing the point of what this could / does achieve?

thanks

i think there could be some performance improvements but the most important think in speed stepping is the power consumption. More pstates => less power consumed by the cpu when doing task's.

 

[rob1980]

i think there could be some performance improvements but the most important think in speed stepping is the power consumption. More pstates => less power consumed by the cpu when doing task's.

cooler cpu also then?

 

[rob1980]

Ive not tried this yet, but I seem to have quite a few p states already (i5-3360m)

[FONT=Menlo]kernel[0]: AICPUPMI: CPU P-States [ 12 13 17 18 19 22 28 29 30 31 32 33 34 35 ][/FONT]

 

[nguyenmac]

Ive not tried this yet, but I seem to have quite a few p states already (i5-3360m)

[FONT=Menlo]kernel[0]: AICPUPMI: CPU P-States [ 12 13 17 18 19 22 28 29 30 31 32 33 34 35 ][/FONT]

Very nice. It's just for testing, so no confirmation yet.

Btw, do your sound work now?