On Tue 14-05-13 17:13:07, Sha Zhengju wrote: > On Mon, May 13, 2013 at 9:38 PM, Michal Hocko <mhocko@xxxxxxx> wrote: > > > > On Mon 13-05-13 15:12:51, Michal Hocko wrote: > > [...] > > > I am sorry but I do not think this is the right approach. IMO we should > > > focus on mem_cgroup_begin_update_page_stat and make it really recursive > > > safe - ideally without any additional overhead (which sounds like a real > > > challenge) > > > > Or maybe we should just not over complicate this and simply consider > > recursivness when it starts being an issue. It is not a problem for > > rmap accounting anymore and dirty pages accounting seems to be safe as > > well and pages under writeback accounting was OK even previously. > > It doesn't make much sense to block dirty pages accounting by a > > non-existing problem. > > > > Yes, the dirty/writeback accounting seems okay now. I sent this patch > out to see if I can do something to simplify the locks but this > approach seems to have its own drawbacks. Since you and Kame are NAK > to this, in the order of importance I'll put the patch aside and > continue the work of dirty page accounting. :) Thanks for your work! > Thanks for the teaching! -- Michal Hocko SUSE Labs -- To unsubscribe from this list: send the line "unsubscribe cgroups" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
