From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail144.messagelabs.com (mail144.messagelabs.com [216.82.254.51]) by kanga.kvack.org (Postfix) with ESMTP id B06528D0039 for ; Thu, 3 Feb 2011 09:27:13 -0500 (EST) From: Johannes Weiner Subject: memcg: save 20% of per-page memcg memory overhead Date: Thu, 3 Feb 2011 15:26:01 +0100 Message-Id: <1296743166-9412-1-git-send-email-hannes@cmpxchg.org> Sender: owner-linux-mm@kvack.org List-ID: To: Andrew Morton Cc: KAMEZAWA Hiroyuki , Daisuke Nishimura , Balbir Singh , linux-mm@kvack.org, linux-kernel@vger.kernel.org This patch series removes the direct page pointer from struct page_cgroup, which saves 20% of per-page memcg memory overhead (Fedora and Ubuntu enable memcg per default, openSUSE apparently too). The node id or section number is encoded in the remaining free bits of pc->flags which allows calculating the corresponding page without the extra pointer. I ran, what I think is, a worst-case microbenchmark that just cats a large sparse file to /dev/null, because it means that walking the LRU list on behalf of per-cgroup reclaim and looking up pages from page_cgroups is happening constantly and at a high rate. But it made no measurable difference. A profile reported a 0.11% share of the new lookup_cgroup_page() function in this benchmark. Hannes -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. 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: email@kvack.org