Messages in this thread | | | Date | Mon, 08 Oct 2012 19:33:36 +0800 | From | Charles Wang <> | Subject | Re: Seems like "sched: Add missing call to calc_load_exit_idle()" should be reverted in 3.5 branch |
| |
On 10/06/2012 01:23 AM, Peter Zijlstra wrote: > On Fri, 2012-10-05 at 10:10 -0700, Jonathan Nieder wrote: >> Peter Zijlstra wrote: >>> On Thu, 2012-10-04 at 15:27 -0700, Greg Kroah-Hartman wrote: >> >>>> I'm puzzled as well. Any ideas if I should do anything here or not? >>> >>> So I think the current v3.5.5 code is fine. >> >> Now I'm puzzled. You wrote: >> >> | However, since we don't restart the tick, we won't be sampling load muck >> | and calling calc_load_exit_idle() from there is bound to confuse state. >> >> Doesn't that mean 900404e5d201 "sched: Add missing call to >> calc_load_exit_idle()" which is part of 3.5.5 was problematic? Or >> did I just miscount the number of "not"s? > > > Argh, yeah, so now I've managed to confuse everyone I'm afraid. > > You are right, v3.5.5 has one calc_load_exit_idle() too many, the one in > tick_nohz_update_jiffies() needs to go. > > Sorry.. I got entirely confused figuring out wth happened with 3.6. > High loadavg reported with v3.6, and I just checked the upstream code, which puzzled many people. Sorry for that~
| |