From: Balbir Singh <balbir@linux.vnet.ibm.com> To: Minchan Kim <minchan.kim@gmail.com> Cc: Andrew Morton <akpm@linux-foundation.org>, Rik van Riel <riel@redhat.com>, KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>, linux-mm <linux-mm@kvack.org>, LKML <linux-kernel@vger.kernel.org>, KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Subject: Re: [PATCH v4 3/7] move memcg reclaimable page into tail of inactive list Date: Mon, 6 Dec 2010 09:04:55 +0530 [thread overview] Message-ID: <20101206033455.GA3158@balbir.in.ibm.com> (raw) In-Reply-To: <a11d438e09af9808ac0cb0aba3e74c8a8deb4076.1291568905.git.minchan.kim@gmail.com> * MinChan Kim <minchan.kim@gmail.com> [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 <balbir@linux.vnet.ibm.com> > Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> > Cc: Rik van Riel <riel@redhat.com> > Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> > Signed-off-by: Minchan Kim <minchan.kim@gmail.com> > --- > include/linux/memcontrol.h | 6 ++++++ > mm/memcontrol.c | 27 +++++++++++++++++++++++++++ > mm/swap.c | 3 ++- > 3 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, > gfp_t gfp_mask); > extern void mem_cgroup_add_lru_list(struct page *page, enum lru_list lru); > extern void mem_cgroup_del_lru_list(struct page *page, enum lru_list lru); > +extern void mem_cgroup_rotate_reclaimable_page(struct page *page); > extern void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru); > extern void mem_cgroup_del_lru(struct page *page); > extern void mem_cgroup_move_lists(struct page *page, > @@ -207,6 +208,11 @@ static inline void mem_cgroup_del_lru_list(struct page *page, int lru) > return ; > } > > +static inline inline void mem_cgroup_rotate_reclaimable_page(struct page *page) > +{ > + return ; > +} > + > static inline void mem_cgroup_rotate_lru_list(struct page *page, int lru) > { > 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) > mem_cgroup_del_lru_list(page, page_lru(page)); > } > > +/* > + * Writeback is about to end against a page which has been marked for immediate > + * reclaim. If it still appears to be reclaimable, move it to the tail of the > + * inactive list. > + */ > +void mem_cgroup_rotate_reclaimable_page(struct page *page) > +{ > + struct mem_cgroup_per_zone *mz; > + struct page_cgroup *pc; > + enum lru_list lru = page_lru_base_type(page); > + > + if (mem_cgroup_disabled()) > + return; > + > + pc = lookup_page_cgroup(page); > + /* > + * Used bit is set without atomic ops but after smp_wmb(). > + * For making pc->mem_cgroup visible, insert smp_rmb() here. > + */ > + smp_rmb(); > + /* unused or root page is not rotated. */ > + if (!PageCgroupUsed(pc) || mem_cgroup_is_root(pc->mem_cgroup)) > + return; > + mz = page_cgroup_zoneinfo(pc); > + list_move_tail(&pc->lru, &mz->lists[lru]); > +} > + > void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru) > { > 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) > } > if (PageLRU(page) && !PageActive(page) && > !PageUnevictable(page)) { > - int lru = page_lru_base_type(page); > + enum lru_list lru = page_lru_base_type(page); > list_move_tail(&page->lru, &zone->lru[lru].list); > + mem_cgroup_rotate_reclaimable_page(page); > pgmoved++; > } > } 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. Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com> -- Three Cheers, Balbir
WARNING: multiple messages have this Message-ID (diff)
From: Balbir Singh <balbir@linux.vnet.ibm.com> To: Minchan Kim <minchan.kim@gmail.com> Cc: Andrew Morton <akpm@linux-foundation.org>, Rik van Riel <riel@redhat.com>, KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>, linux-mm <linux-mm@kvack.org>, LKML <linux-kernel@vger.kernel.org>, KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Subject: Re: [PATCH v4 3/7] move memcg reclaimable page into tail of inactive list Date: Mon, 6 Dec 2010 09:04:55 +0530 [thread overview] Message-ID: <20101206033455.GA3158@balbir.in.ibm.com> (raw) In-Reply-To: <a11d438e09af9808ac0cb0aba3e74c8a8deb4076.1291568905.git.minchan.kim@gmail.com> * MinChan Kim <minchan.kim@gmail.com> [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 <balbir@linux.vnet.ibm.com> > Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> > Cc: Rik van Riel <riel@redhat.com> > Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> > Signed-off-by: Minchan Kim <minchan.kim@gmail.com> > --- > include/linux/memcontrol.h | 6 ++++++ > mm/memcontrol.c | 27 +++++++++++++++++++++++++++ > mm/swap.c | 3 ++- > 3 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, > gfp_t gfp_mask); > extern void mem_cgroup_add_lru_list(struct page *page, enum lru_list lru); > extern void mem_cgroup_del_lru_list(struct page *page, enum lru_list lru); > +extern void mem_cgroup_rotate_reclaimable_page(struct page *page); > extern void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru); > extern void mem_cgroup_del_lru(struct page *page); > extern void mem_cgroup_move_lists(struct page *page, > @@ -207,6 +208,11 @@ static inline void mem_cgroup_del_lru_list(struct page *page, int lru) > return ; > } > > +static inline inline void mem_cgroup_rotate_reclaimable_page(struct page *page) > +{ > + return ; > +} > + > static inline void mem_cgroup_rotate_lru_list(struct page *page, int lru) > { > 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) > mem_cgroup_del_lru_list(page, page_lru(page)); > } > > +/* > + * Writeback is about to end against a page which has been marked for immediate > + * reclaim. If it still appears to be reclaimable, move it to the tail of the > + * inactive list. > + */ > +void mem_cgroup_rotate_reclaimable_page(struct page *page) > +{ > + struct mem_cgroup_per_zone *mz; > + struct page_cgroup *pc; > + enum lru_list lru = page_lru_base_type(page); > + > + if (mem_cgroup_disabled()) > + return; > + > + pc = lookup_page_cgroup(page); > + /* > + * Used bit is set without atomic ops but after smp_wmb(). > + * For making pc->mem_cgroup visible, insert smp_rmb() here. > + */ > + smp_rmb(); > + /* unused or root page is not rotated. */ > + if (!PageCgroupUsed(pc) || mem_cgroup_is_root(pc->mem_cgroup)) > + return; > + mz = page_cgroup_zoneinfo(pc); > + list_move_tail(&pc->lru, &mz->lists[lru]); > +} > + > void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru) > { > 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) > } > if (PageLRU(page) && !PageActive(page) && > !PageUnevictable(page)) { > - int lru = page_lru_base_type(page); > + enum lru_list lru = page_lru_base_type(page); > list_move_tail(&page->lru, &zone->lru[lru].list); > + mem_cgroup_rotate_reclaimable_page(page); > pgmoved++; > } > } 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. Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com> -- Three Cheers, Balbir -- 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: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2010-12-06 3:52 UTC|newest] Thread overview: 109+ messages / expand[flat|nested] mbox.gz Atom feed top 2010-12-05 17:29 [PATCH v4 0/7] f/madivse(DONTNEED) support Minchan Kim 2010-12-05 17:29 ` Minchan Kim 2010-12-05 17:29 ` [PATCH v4 1/7] Fix checkpatch's report in swap.c Minchan Kim 2010-12-05 17:29 ` Minchan Kim 2010-12-06 1:47 ` Rik van Riel 2010-12-06 1:47 ` Rik van Riel 2010-12-07 14:37 ` Johannes Weiner 2010-12-07 14:37 ` Johannes Weiner 2010-12-05 17:29 ` [PATCH v4 2/7] deactivate invalidated pages Minchan Kim 2010-12-05 17:29 ` Minchan Kim 2010-12-06 14:53 ` Mel Gorman 2010-12-06 14:53 ` Mel Gorman 2010-12-07 14:49 ` Johannes Weiner 2010-12-07 14:49 ` Johannes Weiner 2010-12-07 15:07 ` Minchan Kim 2010-12-07 15:07 ` Minchan Kim 2010-12-07 15:19 ` Johannes Weiner 2010-12-07 15:19 ` Johannes Weiner 2010-12-07 15:26 ` Minchan Kim 2010-12-07 15:26 ` Minchan Kim 2010-12-07 15:56 ` Johannes Weiner 2010-12-07 15:56 ` Johannes Weiner 2010-12-07 22:51 ` Minchan Kim 2010-12-07 22:51 ` Minchan Kim 2010-12-08 0:56 ` KAMEZAWA Hiroyuki 2010-12-08 0:56 ` KAMEZAWA Hiroyuki 2010-12-08 1:43 ` Minchan Kim 2010-12-08 1:43 ` Minchan Kim 2010-12-08 1:56 ` KAMEZAWA Hiroyuki 2010-12-08 1:56 ` KAMEZAWA Hiroyuki 2010-12-08 2:15 ` Minchan Kim 2010-12-08 2:15 ` Minchan Kim 2010-12-08 6:56 ` Balbir Singh 2010-12-08 6:56 ` Balbir Singh 2010-12-09 0:19 ` Minchan Kim 2010-12-09 0:19 ` Minchan Kim 2010-12-05 17:29 ` [PATCH v4 3/7] move memcg reclaimable page into tail of inactive list Minchan Kim 2010-12-05 17:29 ` Minchan Kim 2010-12-06 0:04 ` KAMEZAWA Hiroyuki 2010-12-06 0:04 ` KAMEZAWA Hiroyuki 2010-12-06 3:04 ` Rik van Riel 2010-12-06 3:04 ` Rik van Riel 2010-12-07 0:17 ` Minchan Kim 2010-12-07 0:17 ` Minchan Kim 2010-12-06 3:34 ` Balbir Singh [this message] 2010-12-06 3:34 ` Balbir Singh 2010-12-07 0:20 ` Minchan Kim 2010-12-07 0:20 ` Minchan Kim 2010-12-07 14:52 ` Johannes Weiner 2010-12-07 14:52 ` Johannes Weiner 2010-12-08 8:08 ` KOSAKI Motohiro 2010-12-08 8:08 ` KOSAKI Motohiro 2010-12-05 17:29 ` [PATCH v4 4/7] Reclaim invalidated page ASAP Minchan Kim 2010-12-05 17:29 ` Minchan Kim 2010-12-07 15:05 ` Johannes Weiner 2010-12-07 15:05 ` Johannes Weiner 2010-12-07 15:21 ` Minchan Kim 2010-12-07 15:21 ` Minchan Kim 2010-12-08 8:04 ` KOSAKI Motohiro 2010-12-08 8:04 ` KOSAKI Motohiro 2010-12-08 8:16 ` Minchan Kim 2010-12-08 8:16 ` Minchan Kim 2010-12-08 13:01 ` Ben Gamari 2010-12-08 13:01 ` Ben Gamari 2010-12-08 23:10 ` Minchan Kim 2010-12-08 23:10 ` Minchan Kim 2010-12-13 15:31 ` Minchan Kim 2010-12-13 15:31 ` Minchan Kim 2010-12-13 20:06 ` Ben Gamari 2010-12-13 20:06 ` Ben Gamari 2010-12-14 2:36 ` Minchan Kim 2010-12-14 2:36 ` Minchan Kim 2011-07-25 3:08 ` Ben Gamari 2011-07-25 3:08 ` Ben Gamari 2011-07-25 3:47 ` Minchan Kim 2011-07-25 3:47 ` Minchan Kim 2010-12-14 2:07 ` KAMEZAWA Hiroyuki 2010-12-14 2:07 ` KAMEZAWA Hiroyuki 2010-12-14 2:34 ` Minchan Kim 2010-12-14 2:34 ` Minchan Kim 2010-12-05 17:29 ` [PATCH v4 5/7] add profile information for invalidated page reclaim Minchan Kim 2010-12-05 17:29 ` Minchan Kim 2010-12-06 3:24 ` Rik van Riel 2010-12-06 3:24 ` Rik van Riel 2010-12-08 8:02 ` KOSAKI Motohiro 2010-12-08 8:02 ` KOSAKI Motohiro 2010-12-08 8:13 ` Minchan Kim 2010-12-08 8:13 ` Minchan Kim 2010-12-08 8:36 ` KOSAKI Motohiro 2010-12-08 8:36 ` KOSAKI Motohiro 2010-12-05 17:29 ` [PATCH v4 6/7] Remove zap_details NULL dependency Minchan Kim 2010-12-05 17:29 ` Minchan Kim 2010-12-06 3:25 ` Rik van Riel 2010-12-06 3:25 ` Rik van Riel 2010-12-07 4:26 ` Hugh Dickins 2010-12-07 4:26 ` Hugh Dickins 2010-12-07 5:30 ` Minchan Kim 2010-12-07 5:30 ` Minchan Kim 2010-12-05 17:29 ` [PATCH v4 7/7] Prevent activation of page in madvise_dontneed Minchan Kim 2010-12-05 17:29 ` Minchan Kim 2010-12-07 4:48 ` Hugh Dickins 2010-12-07 4:48 ` Hugh Dickins 2010-12-07 5:44 ` Minchan Kim 2010-12-08 7:26 ` Hugh Dickins 2010-12-08 7:26 ` Hugh Dickins 2010-12-08 7:55 ` Minchan Kim 2010-12-08 7:55 ` Minchan Kim [not found] ` <AANLkTim71krrCcmhTTCZTzxeUDkvOdBTOkeYQu6EXt32@mail.gmail.com> 2010-12-07 1:52 ` [PATCH v4 0/7] f/madivse(DONTNEED) support Ben Gamari 2010-12-07 2:16 ` Minchan Kim
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20101206033455.GA3158@balbir.in.ibm.com \ --to=balbir@linux.vnet.ibm.com \ --cc=akpm@linux-foundation.org \ --cc=kamezawa.hiroyu@jp.fujitsu.com \ --cc=kosaki.motohiro@jp.fujitsu.com \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=minchan.kim@gmail.com \ --cc=riel@redhat.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.