Messages in this thread | | | Date | Thu, 26 Nov 2015 10:33:59 +0530 | Subject | Re: [PATCH] mm/vmstat: retrieve more accurate vmstat value | From | vinayak menon <> |
| |
On Thu, Nov 26, 2015 at 7:26 AM, Joonsoo Kim <iamjoonsoo.kim@lge.com> wrote: > On Wed, Nov 25, 2015 at 01:00:22PM +0100, Michal Hocko wrote: >> On Tue 24-11-15 15:22:03, Joonsoo Kim wrote: >> > When I tested compaction in low memory condition, I found that >> > my benchmark is stuck in congestion_wait() at shrink_inactive_list(). >> > This stuck last for 1 sec and after then it can escape. More investigation >> > shows that it is due to stale vmstat value. vmstat is updated every 1 sec >> > so it is stuck for 1 sec. >> >> Wouldn't it be sufficient to use zone_page_state_snapshot in >> too_many_isolated? > This was done by this patch I believe, http://lkml.iu.edu/hypermail/linux/kernel/1501.2/00001.html, though the original issue (wait of more than 1 sec) was fixed by the vmstat changes.
| |