From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759807AbZENMfZ (ORCPT ); Thu, 14 May 2009 08:35:25 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1755362AbZENMfJ (ORCPT ); Thu, 14 May 2009 08:35:09 -0400 Received: from yw-out-2324.google.com ([74.125.46.28]:22012 "EHLO yw-out-2324.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755250AbZENMfG convert rfc822-to-8bit (ORCPT ); Thu, 14 May 2009 08:35:06 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; b=RUTReB5uNBULKghbTch9z2qtObkUUb5JxYdNZTPctXW9tUUXXMUBYbbwCSog6eo6Bw VdDdBB8NQeBQ6Zsd31WxMH7J9tyXNgtIAFVOIzj+shNWiBOo2PQzQ8yGrH3fmE+em7QL rb8BTqnQ/C1XkNsVLjvASZhN4poauh0jeqLTo= MIME-Version: 1.0 In-Reply-To: <20090514210839.9B90.A69D9226@jp.fujitsu.com> References: <20090514204033.9B87.A69D9226@jp.fujitsu.com> <28c262360905140505h2db7ac3bp5ca10fcf2b4301bb@mail.gmail.com> <20090514210839.9B90.A69D9226@jp.fujitsu.com> Date: Thu, 14 May 2009 21:35:07 +0900 Message-ID: <28c262360905140535w4bfdbbccp55f395daa18fbf0e@mail.gmail.com> Subject: Re: [PATCH] mmtom: Prevent shrinking of active anon lru list in case of no swap space V2 From: Minchan Kim To: KOSAKI Motohiro Cc: Andrew Morton , LKML , linux-mm , Johannes Weiner , Rik van Riel Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, May 14, 2009 at 9:11 PM, KOSAKI Motohiro wrote: >> On Thu, May 14, 2009 at 8:44 PM, KOSAKI Motohiro >> wrote: >> >> > >> >> > Changelog since V2 >> >> > ?o Add new function - can_reclaim_anon : it tests anon_list can be reclaim >> >> > >> >> > Changelog since V1 >> >> > ?o Use nr_swap_pages <= 0 in shrink_active_list to prevent scanning ?of active anon list. >> >> > >> >> > Now shrink_active_list is called several places. >> >> > But if we don't have a swap space, we can't reclaim anon pages. >> >> > So, we don't need deactivating anon pages in anon lru list. >> >> > >> >> > Signed-off-by: Minchan Kim >> >> > Cc: KOSAKI Motohiro >> >> > Cc: Johannes Weiner >> >> > Cc: Rik van Riel >> >> >> >> looks good to me. thanks :) >> > >> > Grr, my fault. >> > >> > >> > >> >> ?static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan, >> >> ? ? ? struct zone *zone, struct scan_control *sc, int priority) >> >> ?{ >> >> @@ -1399,7 +1412,7 @@ static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan, >> >> ? ? ? ? ? ? ? return 0; >> >> ? ? ? } >> >> >> >> - ? ? if (lru == LRU_ACTIVE_ANON && inactive_anon_is_low(zone, sc)) { >> >> + ? ? if (lru == LRU_ACTIVE_ANON && can_reclaim_anon(zone, sc)) { >> >> ? ? ? ? ? ? ? shrink_active_list(nr_to_scan, zone, sc, priority, file); >> >> ? ? ? ? ? ? ? return 0; >> > >> > you shouldn't do that. if nr_swap_pages==0, get_scan_ratio return anon=0%. >> > then, this branch is unnecessary. >> > >> >> But, I think at last it can be happen following as. >> >> 1515         * Even if we did not try to evict anon pages at all, we want to >> 1516         * rebalance the anon lru active/inactive ratio. >> 1517         */ >> 1518        if (inactive_anon_is_low(zone, sc)) >> 1519                shrink_active_list(SWAP_CLUSTER_MAX, zone, sc, priority, 0); > > I pointed to shrink_list(), but you replayed shrink_zone(). > I only talked about shrink_list(). > In shrink_zone, we call get_scan_ratio. it prevent scanning anon list. but, in shrink_all_zones can't prevent it. Also, I think shrink_list is not hot patch. So check of one condition adding is trivial If I don't understand your point, please explain detaily -- Kinds regards, Minchan Kim