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 SMTP id DD0286B0087 for ; Mon, 6 Dec 2010 19:21:01 -0500 (EST) Received: by iwn5 with SMTP id 5so353817iwn.14 for ; Mon, 06 Dec 2010 16:21:00 -0800 (PST) MIME-Version: 1.0 In-Reply-To: <20101206033455.GA3158@balbir.in.ibm.com> References: <20101206033455.GA3158@balbir.in.ibm.com> Date: Tue, 7 Dec 2010 09:20:59 +0900 Message-ID: Subject: Re: [PATCH v4 3/7] move memcg reclaimable page into tail of inactive list From: Minchan Kim Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Sender: owner-linux-mm@kvack.org To: balbir@linux.vnet.ibm.com Cc: Andrew Morton , Rik van Riel , KOSAKI Motohiro , linux-mm , LKML , KAMEZAWA Hiroyuki List-ID: Hi Balbir, On Mon, Dec 6, 2010 at 12:34 PM, Balbir Singh w= rote: > * MinChan Kim [2010-12-06 02:29:11]: > >> Golbal page reclaim moves reclaimalbe pages into inactive list > > Some typos here and Rik already pointed out some other changes. > >> to reclaim asap. This patch apply the rule in memcg. >> It can help to prevent unnecessary working page eviction of memcg. >> >> Cc: Balbir Singh >> Cc: KAMEZAWA Hiroyuki >> Cc: Rik van Riel >> Cc: KOSAKI Motohiro >> Signed-off-by: Minchan Kim >> --- >> =A0include/linux/memcontrol.h | =A0 =A06 ++++++ >> =A0mm/memcontrol.c =A0 =A0 =A0 =A0 =A0 =A0| =A0 27 +++++++++++++++++++++= ++++++ >> =A0mm/swap.c =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0 =A03 ++- >> =A03 files changed, 35 insertions(+), 1 deletions(-) >> >> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h >> index 067115c..8317f5c 100644 >> --- a/include/linux/memcontrol.h >> +++ b/include/linux/memcontrol.h >> @@ -62,6 +62,7 @@ extern int mem_cgroup_cache_charge(struct page *page, = struct mm_struct *mm, >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 = =A0 gfp_t gfp_mask); >> =A0extern void mem_cgroup_add_lru_list(struct page *page, enum lru_list = lru); >> =A0extern void mem_cgroup_del_lru_list(struct page *page, enum lru_list = lru); >> +extern void mem_cgroup_rotate_reclaimable_page(struct page *page); >> =A0extern void mem_cgroup_rotate_lru_list(struct page *page, enum lru_li= st lru); >> =A0extern void mem_cgroup_del_lru(struct page *page); >> =A0extern void mem_cgroup_move_lists(struct page *page, >> @@ -207,6 +208,11 @@ static inline void mem_cgroup_del_lru_list(struct p= age *page, int lru) >> =A0 =A0 =A0 return ; >> =A0} >> >> +static inline inline void mem_cgroup_rotate_reclaimable_page(struct pag= e *page) >> +{ >> + =A0 =A0 return ; >> +} >> + >> =A0static inline void mem_cgroup_rotate_lru_list(struct page *page, int = lru) >> =A0{ >> =A0 =A0 =A0 return ; >> diff --git a/mm/memcontrol.c b/mm/memcontrol.c >> index 729beb7..f9435be 100644 >> --- a/mm/memcontrol.c >> +++ b/mm/memcontrol.c >> @@ -829,6 +829,33 @@ void mem_cgroup_del_lru(struct page *page) >> =A0 =A0 =A0 mem_cgroup_del_lru_list(page, page_lru(page)); >> =A0} >> >> +/* >> + * Writeback is about to end against a page which has been marked for i= mmediate >> + * reclaim. =A0If it still appears to be reclaimable, move it to the ta= il of the >> + * inactive list. >> + */ >> +void mem_cgroup_rotate_reclaimable_page(struct page *page) >> +{ >> + =A0 =A0 struct mem_cgroup_per_zone *mz; >> + =A0 =A0 struct page_cgroup *pc; >> + =A0 =A0 enum lru_list lru =3D page_lru_base_type(page); >> + >> + =A0 =A0 if (mem_cgroup_disabled()) >> + =A0 =A0 =A0 =A0 =A0 =A0 return; >> + >> + =A0 =A0 pc =3D lookup_page_cgroup(page); >> + =A0 =A0 /* >> + =A0 =A0 =A0* Used bit is set without atomic ops but after smp_wmb(). >> + =A0 =A0 =A0* For making pc->mem_cgroup visible, insert smp_rmb() here. >> + =A0 =A0 =A0*/ >> + =A0 =A0 smp_rmb(); >> + =A0 =A0 /* unused or root page is not rotated. */ >> + =A0 =A0 if (!PageCgroupUsed(pc) || mem_cgroup_is_root(pc->mem_cgroup)) >> + =A0 =A0 =A0 =A0 =A0 =A0 return; >> + =A0 =A0 mz =3D page_cgroup_zoneinfo(pc); >> + =A0 =A0 list_move_tail(&pc->lru, &mz->lists[lru]); >> +} >> + >> =A0void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru) >> =A0{ >> =A0 =A0 =A0 struct mem_cgroup_per_zone *mz; >> diff --git a/mm/swap.c b/mm/swap.c >> index 1f36f6f..0fe98e7 100644 >> --- a/mm/swap.c >> +++ b/mm/swap.c >> @@ -122,8 +122,9 @@ static void pagevec_move_tail(struct pagevec *pvec) >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 } >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 if (PageLRU(page) && !PageActive(page) && >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 = =A0 !PageUnevictable(page)) { >> - =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 int lru =3D page_lru_base_type= (page); >> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 enum lru_list lru =3D page_lru= _base_type(page); >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 list_move_tail(&page->lru, &= zone->lru[lru].list); >> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 mem_cgroup_rotate_reclaimable_= page(page); >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 pgmoved++; >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 } >> =A0 =A0 =A0 } > > Looks good, do you have any numbers, workloads that benefit? I agree > that keeping both global and memcg reclaim in sync is a good idea. This patch series for Ben's fadvise problem in rsync. As I fix the global reclaim, I found this patch could help memcg, too. If Ben is busy, I will measure the benefit. > > Acked-by: Balbir Singh Thanks, Balbir. > > > -- > =A0 =A0 =A0 =A0Three Cheers, > =A0 =A0 =A0 =A0Balbir > --=20 Kind regards, Minchan Kim -- 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 policy in Canada: sign http://dissolvethecrtc.ca/ Don't email: email@kvack.org