Here is my page_cgroup diet series v3. Since v2, "remove PCG_CACHE" is alread merged. This series changes page-stat-accounting per memcg from: if (change page's state) mem_cgroup_update_page_state() to: mem_cgroup_begin_update_page_state() if (change page's state) mem_cgroup_update_page_state() mem_cgroup_end_update_page_state() (see patch 4 for details.) This allows us not to duplicate page struct's information in page_cgroup's flag field. Because above sequence adds 2 extra calls to hot-path, performance will be problem. Patch 6 is a fix for performance, and I don't see performance regression in my small test. (see patch 6 for details.) Thanks, -Kame -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>