cpu scaling

[sedonix]

Why did I try this off-brand kernel(2.6.34-1.dmz.3-liquorix-amd64) Why? I was wondering why my TB drives were having a cup of coffee before the data showed up! I want my 2.8 GHZ speed at ALL times. I don't want to hear any "greenie" stuff about saving energy. I know how to diddle with the freqs after the fact. So, is this fixable or do i need to stay with a Debian kernel which comes sprinting out of the blocks at 2.8GHZ and not crawling out at 1GHZ.

[iotaka]

You can try to remove cpufrequtils and sysfsutils. I prefer have cpu scaling anyway ...

[sedonix]

Thanks iotaka! cpufrequtils did it and now I can see why. I didn't have
sysfsutils. I'm running at 2.813 GHZ forever. It does make a big difference when
you're dealing with getting data out of terabyte drives. I originally thought my
drives had died it took so long.

[damentz]

sedonix, don't worry about cpu scaling on these kernels. I applied a patch for the ondemand governor which accounts for IO when choosing the frequency to run at.

All vanilla kernels 2.6.34 and less should have the bug you're experiencing.

[sedonix]

What, me worry? I'm too old to worry. I want nothing to do with"scaling". I have hundreds of movies on my drives to watch and I can't wait around while some "greenie" software decides on how fast I get to them. I learned a long time ago how to set the cpu speed. It takes at least 10 seconds to change it in a console. Time wasted!

[DeepDayze]

Damentz, have you also applied the patch to the new 2.6.35 kernel as well?

[sedonix]

2.6.35 seems to be behaving as advertised.........so far.