On Wed, 25 Jan 2023 15:22:00 -0300 Marcelo Tosatti <mtosatti@xxxxxxxxxx> > > Remote draining reduces interruptions whether CPU > is marked as isolated or not: > > - Allows isolated CPUs from benefiting of pcp caching. > - Removes the interruption to non isolated CPUs. See for example Why ask refill to take a pill because drain got a cough? > > https://lkml.org/lkml/2022/6/13/2769 > > "Minchan Kim tested this independently and reported; > > My workload is not NOHZ CPUs but run apps under heavy memory > pressure so they goes to direct reclaim and be stuck on > drain_all_pages until work on workqueue run. What sense are you trying to make by getting CPUs isolated and equipped with tight memory? > > unit: nanosecond > max(dur) avg(dur) count(dur) > 166713013 487511.77786438033 1283 > > From traces, system encountered the drain_all_pages 1283 times and > worst case was 166ms and avg was 487us. > > The other problem was alloc_contig_range in CMA. The PCP draining > takes several hundred millisecond sometimes though there is no > memory pressure or a few of pages to be migrated out but CPU were > fully booked. > > Your patch perfectly removed those wasted time."
