Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757258Ab1EMLih (ORCPT ); Fri, 13 May 2011 07:38:37 -0400 Received: from fgwmail6.fujitsu.co.jp ([192.51.44.36]:47914 "EHLO fgwmail6.fujitsu.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752835Ab1EMLif (ORCPT ); Fri, 13 May 2011 07:38:35 -0400 X-SecurityPolicyCheck-FJ: OK by FujitsuOutboundMailChecker v1.3.1 Message-ID: <4DCD189D.2000207@jp.fujitsu.com> Date: Fri, 13 May 2011 20:40:13 +0900 From: KOSAKI Motohiro User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; ja; rv:1.9.2.17) Gecko/20110414 Lightning/1.0b2 Thunderbird/3.1.10 MIME-Version: 1.0 To: kosaki.motohiro@jp.fujitsu.com CC: linux-mm@kvack.org, linux-kernel@vger.kernel.org, akpm@linux-foundation.org, kamezawa.hiroyu@jp.fujitsu.com, minchan.kim@gmail.com, riel@redhat.com Subject: [PATCH 1/3] vmscan,memcg: memcg aware swap token References: <4DCD1824.1060801@jp.fujitsu.com> In-Reply-To: <4DCD1824.1060801@jp.fujitsu.com> Content-Type: text/plain; charset=ISO-2022-JP Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 7263 Lines: 259 Currently, memcg reclaim can disable swap token even if the swap token mm doesn't belong in its memory cgroup. It's slightly risky. If an admin creates very small mem-cgroup and silly guy runs contentious heavy memory pressure workload, every tasks are going to lose swap token and then system may become unresponsive. That's bad. This patch adds 'memcg' parameter into disable_swap_token(). and if the parameter doesn't match swap token, VM doesn't disable it. Signed-off-by: KOSAKI Motohiro --- include/linux/memcontrol.h | 6 +++ include/linux/swap.h | 8 +--- mm/memcontrol.c | 16 ++++----- mm/thrash.c | 73 ++++++++++++++++++++++++++++++++----------- mm/vmscan.c | 4 +- 5 files changed, 71 insertions(+), 36 deletions(-) diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h index 6a0cffd..df572af 100644 --- a/include/linux/memcontrol.h +++ b/include/linux/memcontrol.h @@ -84,6 +84,7 @@ int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem); extern struct mem_cgroup *try_get_mem_cgroup_from_page(struct page *page); extern struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p); +extern struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm); static inline int mm_match_cgroup(const struct mm_struct *mm, const struct mem_cgroup *cgroup) @@ -244,6 +245,11 @@ static inline struct mem_cgroup *try_get_mem_cgroup_from_page(struct page *page) return NULL; } +static inline struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm) +{ + return NULL; +} + static inline int mm_match_cgroup(struct mm_struct *mm, struct mem_cgroup *mem) { return 1; diff --git a/include/linux/swap.h b/include/linux/swap.h index 384eb5f..e705646 100644 --- a/include/linux/swap.h +++ b/include/linux/swap.h @@ -358,6 +358,7 @@ struct backing_dev_info; extern struct mm_struct *swap_token_mm; extern void grab_swap_token(struct mm_struct *); extern void __put_swap_token(struct mm_struct *); +extern void disable_swap_token(struct mem_cgroup *memcg); static inline int has_swap_token(struct mm_struct *mm) { @@ -370,11 +371,6 @@ static inline void put_swap_token(struct mm_struct *mm) __put_swap_token(mm); } -static inline void disable_swap_token(void) -{ - put_swap_token(swap_token_mm); -} - #ifdef CONFIG_CGROUP_MEM_RES_CTLR extern void mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout); @@ -500,7 +496,7 @@ static inline int has_swap_token(struct mm_struct *mm) return 0; } -static inline void disable_swap_token(void) +static inline void disable_swap_token(struct mem_cgroup *memcg) { } diff --git a/mm/memcontrol.c b/mm/memcontrol.c index c2776f1..1a78b3e 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -735,7 +735,7 @@ struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p) struct mem_cgroup, css); } -static struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm) +struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm) { struct mem_cgroup *mem = NULL; @@ -5194,18 +5194,16 @@ static void mem_cgroup_move_task(struct cgroup_subsys *ss, struct cgroup *old_cont, struct task_struct *p) { - struct mm_struct *mm; + struct mm_struct *mm = get_task_mm(p); - if (!mc.to) - /* no need to move charge */ - return; - - mm = get_task_mm(p); if (mm) { - mem_cgroup_move_charge(mm); + if (mc.to) + mem_cgroup_move_charge(mm); + put_swap_token(mm); mmput(mm); } - mem_cgroup_clear_mc(); + if (mc.to) + mem_cgroup_clear_mc(); } #else /* !CONFIG_MMU */ static int mem_cgroup_can_attach(struct cgroup_subsys *ss, diff --git a/mm/thrash.c b/mm/thrash.c index 2372d4e..32c07fd 100644 --- a/mm/thrash.c +++ b/mm/thrash.c @@ -21,14 +21,17 @@ #include #include #include +#include static DEFINE_SPINLOCK(swap_token_lock); struct mm_struct *swap_token_mm; +struct mem_cgroup *swap_token_memcg; static unsigned int global_faults; void grab_swap_token(struct mm_struct *mm) { int current_interval; + struct mem_cgroup *memcg; global_faults++; @@ -38,40 +41,72 @@ void grab_swap_token(struct mm_struct *mm) return; /* First come first served */ - if (swap_token_mm == NULL) { - mm->token_priority = mm->token_priority + 2; - swap_token_mm = mm; + if (!swap_token_mm) + goto replace_token; + + if (mm == swap_token_mm) { + mm->token_priority += 2; goto out; } - if (mm != swap_token_mm) { - if (current_interval < mm->last_interval) - mm->token_priority++; - else { - if (likely(mm->token_priority > 0)) - mm->token_priority--; - } - /* Check if we deserve the token */ - if (mm->token_priority > swap_token_mm->token_priority) { - mm->token_priority += 2; - swap_token_mm = mm; - } - } else { - /* Token holder came in again! */ - mm->token_priority += 2; + if (current_interval < mm->last_interval) + mm->token_priority++; + else { + if (likely(mm->token_priority > 0)) + mm->token_priority--; } + /* Check if we deserve the token */ + if (mm->token_priority > swap_token_mm->token_priority) + goto replace_token; + out: mm->faultstamp = global_faults; mm->last_interval = current_interval; spin_unlock(&swap_token_lock); + return; + +replace_token: + mm->token_priority += 2; + memcg = try_get_mem_cgroup_from_mm(mm); + if (memcg) + css_put(mem_cgroup_css(memcg)); + swap_token_mm = mm; + swap_token_memcg = memcg; + goto out; } /* Called on process exit. */ void __put_swap_token(struct mm_struct *mm) { spin_lock(&swap_token_lock); - if (likely(mm == swap_token_mm)) + if (likely(mm == swap_token_mm)) { swap_token_mm = NULL; + swap_token_memcg = NULL; + } spin_unlock(&swap_token_lock); } + +static bool match_memcg(struct mem_cgroup *a, struct mem_cgroup *b) +{ + if (!a) + return true; + if (!b) + return true; + if (a == b) + return true; + return false; +} + +void disable_swap_token(struct mem_cgroup *memcg) +{ + /* memcg reclaim don't disable unrelated mm token. */ + if (match_memcg(memcg, swap_token_memcg)) { + spin_lock(&swap_token_lock); + if (match_memcg(memcg, swap_token_memcg)) { + swap_token_mm = NULL; + swap_token_memcg = NULL; + } + spin_unlock(&swap_token_lock); + } +} diff --git a/mm/vmscan.c b/mm/vmscan.c index b3a569f..19e179b 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -2044,7 +2044,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist, for (priority = DEF_PRIORITY; priority >= 0; priority--) { sc->nr_scanned = 0; if (!priority) - disable_swap_token(); + disable_swap_token(sc->mem_cgroup); shrink_zones(priority, zonelist, sc); /* * Don't shrink slabs when reclaiming memory from @@ -2353,7 +2353,7 @@ loop_again: /* The swap token gets in the way of swapout... */ if (!priority) - disable_swap_token(); + disable_swap_token(NULL); all_zones_ok = 1; balanced = 0; -- 1.7.3.1 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/