10.10 for 24, no pressure.

[CRoland]

I think I remember reading at some point that cpufreq-aperf could report the turbo frequency, but I'm not sure if that's the case now.

[Touche]

Have 10.10 copy problems been resolved?

http://www.tomshardware.com/reviews/...x,2774-11.html

[CRoland]

Originally Posted by Touche

Have 10.10 copy problems been resolved?

http://www.tomshardware.com/reviews/...x,2774-11.html

That doesn't look like a reliable test. Such small test files fit in caches. For copy tests the total size should at least be significantly larger than RAM - 8GB in this case.

[toughluck]

CRoland, are you referring to the following? http://www.spinics.net/lists/cpufreq/msg01586.html
The problem is, on system, the p-states are detected correctly, so the above bug doesn't apply, but the system never uses the turbo state. Note that the guy in the report above has never seen cpufreq apply 3.6 GHz to the CPU. Frankly, the best way now seems to be to overclock the CPU and have powernow take care of just idling the CPU, but that's hardly what I'd call an elegant solution.

[toughluck]

Yay, turbo works.

After compiling 2.6.36 again (I was ultimately swayed by 36's long list of fixes), taking pains to configure it correctly and enable aperf, I have ran cpufreq-aperf and I have seen it report up to 3.3 GHz. So yes, it works, and works like a charm (in applications, too).

Now, on to resolving the IOMMU issue.

[PorscheRacer14]

Damn it! So I have to compile to get it to work... Do I compile to get turbo working and potentially mess up an otherwise perfectly running install?

[toughluck]

Try running cpufreq-aperf from root (or sudo cpufreq-aperf). The tool is still not working if I run it from a user account.

[PorscheRacer14]

OK, I'll try it out, thanks!