Daniel Smoczyk wrote: > Hello all, > > I have some question about powertop 1.11 - is this possible that > reporting of C-state is working different on .26 then .28 kernel? > > I have very similar problem like : > http://bugzilla.kernel.org/show_bug.cgi?id=12391 - on my debian > sid (thinkpad t41 - pentium M 1,6) > > my cpu in idle on .26 kernel is running ~100% C3 (reported by > powertop), but when switched into .28 it is ~100% in C2. In my opinion > processor is running in C2 (not a bug in powertop) because fan is on > more often then on .26.
Hi. I get almost the same behaviour in my laptop (pentium M 2.0). I built a custom 2.6.28 kernel with the two patches I regularly apply (TuxonIce and linuxPHC*), and the CPU gets stuck in C2 for 40-50% of the time. After suspend/resume it doesn't even reach C3. To check that the patches weren't to blame I also installed the Debian 2.6.28 kernel, and I observed the same behaviour. My last working kernel is 2.6.26, but I don't know about 2.6.27 since I skipped it. Please note that this doesn't _seem_ to be associated to some process banging the CPU constantly--actually, if anything, I get less wakeups/s with the latest kernels. Regards. Jose. (*) Undervolting of CPU: http://www.linux-phc.org/ > what is source of C-state statistics in powertop? if i get a prove > i'll reopen the bug on kernel.org > > Regards, > Daniel > > _______________________________________________ > Power mailing list > Power@bughost.org > http://www.bughost.org/mailman/listinfo/power > _______________________________________________ Power mailing list Power@bughost.org http://www.bughost.org/mailman/listinfo/power
