On Wed, 2020-05-27 at 08:22 +0200, Jan Beulich wrote: > On 26.05.2020 23:18, Dario Faggioli wrote: > > > > It looks like we need a way to rebalance the runqueues, which > > should be > > doable... But despite having spent a couple of days trying to come > > up > > with something decent, that I could include in v2 of this series, I > > couldn't get it to work sensibly. > > CPU on-/offlining may not need considering here at all imo. I think > it > would be quite reasonable as a first step to have a static model > where > from system topology (and perhaps a command line option) one can > predict which runqueue a particular CPU will end up in, no matter > when > it gets brought online. > Right. IAC, just FYI, after talking to Juergen --who suggested a nice solution to overcome the problem where I was stuck-- I have now a patch that successfully implement dynamic online rebalancing of runqueues. I'm polishing up the comments and changelog and will send it ASAP, as I'd really like for this series to go in... :-) Thanks and Regards -- Dario Faggioli, Ph.D http://about.me/dario.faggioli Virtualization Software Engineer SUSE Labs, SUSE https://www.suse.com/ ------------------------------------------------------------------- <> (Raistlin Majere)