From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail143.messagelabs.com (mail143.messagelabs.com [216.82.254.35]) by kanga.kvack.org (Postfix) with ESMTP id BF4576B0023 for ; Mon, 9 May 2011 04:36:28 -0400 (EDT) Received: from m1.gw.fujitsu.co.jp (unknown [10.0.50.71]) by fgwmail5.fujitsu.co.jp (Postfix) with ESMTP id D95B83EE0C3 for ; Mon, 9 May 2011 17:36:24 +0900 (JST) Received: from smail (m1 [127.0.0.1]) by outgoing.m1.gw.fujitsu.co.jp (Postfix) with ESMTP id BB67945DE64 for ; Mon, 9 May 2011 17:36:24 +0900 (JST) Received: from s1.gw.fujitsu.co.jp (s1.gw.fujitsu.co.jp [10.0.50.91]) by m1.gw.fujitsu.co.jp (Postfix) with ESMTP id A0A4645DE58 for ; Mon, 9 May 2011 17:36:24 +0900 (JST) Received: from s1.gw.fujitsu.co.jp (localhost.localdomain [127.0.0.1]) by s1.gw.fujitsu.co.jp (Postfix) with ESMTP id 8E0CFEF800B for ; Mon, 9 May 2011 17:36:24 +0900 (JST) Received: from m105.s.css.fujitsu.com (m105.s.css.fujitsu.com [10.240.81.145]) by s1.gw.fujitsu.co.jp (Postfix) with ESMTP id 5876FEF8001 for ; Mon, 9 May 2011 17:36:24 +0900 (JST) From: KOSAKI Motohiro Subject: Re: [PATCH 2/2] Allocate memory cgroup structures in local nodes v4 In-Reply-To: <1304716637-19556-2-git-send-email-andi@firstfloor.org> References: <1304716637-19556-1-git-send-email-andi@firstfloor.org> <1304716637-19556-2-git-send-email-andi@firstfloor.org> Message-Id: <20110509173806.1678.A69D9226@jp.fujitsu.com> MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Date: Mon, 9 May 2011 17:36:23 +0900 (JST) Sender: owner-linux-mm@kvack.org List-ID: To: Andi Kleen Cc: kosaki.motohiro@jp.fujitsu.com, akpm@linux-foundation.org, linux-mm@kvack.org, Andi Kleen , rientjes@google.com, Michal Hocko , Dave Hansen , Balbir Singh , Johannes Weiner > From: Andi Kleen > > dde79e005a769 added a regression that the memory cgroup data structures > all end up in node 0 because the first attempt at allocating them > would not pass in a node hint. Since the initialization runs on CPU #0 > it would all end up node 0. This is a problem on large memory systems, > where node 0 would lose a lot of memory. > > Change the alloc_pages_exact to alloc_pages_exact_node. This will > still fall back to other nodes if not enough memory is available. > > [RED-PEN: right now it would fall back first before trying > vmalloc_node. Probably not the best strategy ... But I left it like > that for now.] > > v4: Remove debugging code. > v3: Really call the correct function now. Thanks for everyone who commented. > Reported-by: Doug Nelson > Cc: rientjes@google.com > CC: Michal Hocko > Cc: Dave Hansen > Cc: Balbir Singh > Cc: Johannes Weiner > Signed-off-by: Andi Kleen > --- > mm/page_cgroup.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/mm/page_cgroup.c b/mm/page_cgroup.c > index 9905501..2daadc3 100644 > --- a/mm/page_cgroup.c > +++ b/mm/page_cgroup.c > @@ -134,7 +134,7 @@ static void *__init_refok alloc_page_cgroup(size_t size, int nid) > { > void *addr = NULL; > > - addr = alloc_pages_exact(size, GFP_KERNEL | __GFP_NOWARN); > + addr = alloc_pages_exact_nid(nid, size, GFP_KERNEL | __GFP_NOWARN); > if (addr) > return addr; I guess every developers dislike dis quirk name. but I have no idea another naming. Reviewed-by: KOSAKI Motohiro -- 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