John McCabe-Dansted wrote: > On Feb 17, 2008 2:33 AM, Arjan van de Ven <[EMAIL PROTECTED]> wrote: >> John McCabe-Dansted wrote: >>> Would you agree that ignore_nice_load should not be set by default? >> yes with one exception: there's a class of nice tasks such as [EMAIL >> PROTECTED] and stuff >> where people do want it ignored > > Still, sleeping the process when on battery would be even better? > >>> Also what up_threshold do you recommend? (Ubuntu uses 31% by default, >>> but AFAICT, race to idle would mean that a lower up_threshold would >>> actually save power) >> 31% is just a silly value that is so wrong on many levels; they should just >> leave >> the kernel default. > > Would you care to elaborate? > > In martin's benchmark (below) ondemand is 2% slower than performance. > For someone on AC, this may make them choose performance over > ondemand. > > http://martin.ankerl.com/2006/08/16/how-to-make-firefox-40-percent-faster/ >
note the "2006" in this. Todays ondemand is by far not the same as the 2008 > How does performance compare to ondemand with a low up_threshold? the up_threshold is actually ONLY there because of measurements errors in the sampling. It has NOTHING to do with how fast the kernel goes to full speed. That's a whole different tunable ;) (and yes some gnome people got that one very nastily wrong) > > It seems that ondemand with a low up_threshold could give > "performance" like speed while saving power over the long periods when > the desktop is unattended. no; for that you want to increase the sampling rate, that actually has makes the system go to full speed faster. Changing up_threshold is just giving the algorithm more cushing for sampling errors (btw the measurement errors are going away really soon, Venki is switching ondemand over to an exact measurement rather than sampling, at which point up_threshold can go away too) > > The only reference to up_threshold I found on this list was > http://www.mail-archive.com/[email protected]/msg00699.html > which mentions using an up_threshold of about 75% when not in > power-saving mode. > > So, there is little point in going much below 75%, even on AC power? yep Also, you're just making the second most common thinko that the gnome/desktop folks tend to make. "On AC" is not equivalent with "don't care about power saving". Sadly too many people make that leap... but just talk to a datacenter operator and he'll explain to you that "on AC" means "must save more power" :) > _______________________________________________ Power mailing list [email protected] http://www.bughost.org/mailman/listinfo/power
