From: KAMEZAWA Hiroyuki Date: Thu, 7 Feb 2008 08:14:24 +0000 (-0800) Subject: memory cgroup enhancements: add status accounting function for memory cgroup X-Git-Tag: v2.6.25-rc1~448 X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=d52aa412d43827033a8e2ce4415ef6e8f8d53635;p=linux-2.6 memory cgroup enhancements: add status accounting function for memory cgroup Add statistics account infrastructure for memory controller. All account information is stored per-cpu and caller will not have to take lock or use atomic ops. This will be used by memory.stat file later. CACHE includes swapcache now. I'd like to divide it to PAGECACHE and SWAPCACHE later. This patch adds 3 functions for accounting. * __mem_cgroup_stat_add() ... for usual routine. * __mem_cgroup_stat_add_safe ... for calling under irq_disabled section. * mem_cgroup_read_stat() ... for reading stat value. * renamed PAGECACHE to CACHE (because it may include swapcache *now*) [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: fix smp_processor_id-in-preemptible] [akpm@linux-foundation.org: uninline things] [akpm@linux-foundation.org: remove dead code] Signed-off-by: KAMEZAWA Hiroyuki Signed-off-by: YAMAMOTO Takashi Cc: Balbir Singh Cc: Pavel Emelianov Cc: Paul Menage Cc: Peter Zijlstra Cc: "Eric W. Biederman" Cc: Nick Piggin Cc: Kirill Korotaev Cc: Herbert Poetzl Cc: David Rientjes Cc: Vaidyanathan Srinivasan Cc: Kirill Korotaev Cc: Nick Piggin Cc: Paul Menage Cc: Pavel Emelianov Cc: Peter Zijlstra Cc: Vaidyanathan Srinivasan Cc: YAMAMOTO Takashi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 31c4f0cefd..5f3ad9c37b 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include @@ -34,6 +35,47 @@ struct cgroup_subsys mem_cgroup_subsys; static const int MEM_CGROUP_RECLAIM_RETRIES = 5; +/* + * Statistics for memory cgroup. + */ +enum mem_cgroup_stat_index { + /* + * For MEM_CONTAINER_TYPE_ALL, usage = pagecache + rss. + */ + MEM_CGROUP_STAT_CACHE, /* # of pages charged as cache */ + MEM_CGROUP_STAT_RSS, /* # of pages charged as rss */ + + MEM_CGROUP_STAT_NSTATS, +}; + +struct mem_cgroup_stat_cpu { + s64 count[MEM_CGROUP_STAT_NSTATS]; +} ____cacheline_aligned_in_smp; + +struct mem_cgroup_stat { + struct mem_cgroup_stat_cpu cpustat[NR_CPUS]; +}; + +/* + * For accounting under irq disable, no need for increment preempt count. + */ +static void __mem_cgroup_stat_add_safe(struct mem_cgroup_stat *stat, + enum mem_cgroup_stat_index idx, int val) +{ + int cpu = smp_processor_id(); + stat->cpustat[cpu].count[idx] += val; +} + +static s64 mem_cgroup_read_stat(struct mem_cgroup_stat *stat, + enum mem_cgroup_stat_index idx) +{ + int cpu; + s64 ret = 0; + for_each_possible_cpu(cpu) + ret += stat->cpustat[cpu].count[idx]; + return ret; +} + /* * The memory controller data structure. The memory controller controls both * page cache and RSS per cgroup. We would eventually like to provide @@ -63,6 +105,10 @@ struct mem_cgroup { */ spinlock_t lru_lock; unsigned long control_type; /* control RSS or RSS+Pagecache */ + /* + * statistics. + */ + struct mem_cgroup_stat stat; }; /* @@ -101,6 +147,24 @@ enum charge_type { MEM_CGROUP_CHARGE_TYPE_MAPPED, }; +/* + * Always modified under lru lock. Then, not necessary to preempt_disable() + */ +static void mem_cgroup_charge_statistics(struct mem_cgroup *mem, int flags, + bool charge) +{ + int val = (charge)? 1 : -1; + struct mem_cgroup_stat *stat = &mem->stat; + VM_BUG_ON(!irqs_disabled()); + + if (flags & PAGE_CGROUP_FLAG_CACHE) + __mem_cgroup_stat_add_safe(stat, + MEM_CGROUP_STAT_CACHE, val); + else + __mem_cgroup_stat_add_safe(stat, MEM_CGROUP_STAT_RSS, val); + +} + static struct mem_cgroup init_mem_cgroup; static inline @@ -175,8 +239,8 @@ static void __always_inline unlock_page_cgroup(struct page *page) * This can fail if the page has been tied to a page_cgroup. * If success, returns 0. */ -static inline int -page_cgroup_assign_new_page_cgroup(struct page *page, struct page_cgroup *pc) +static int page_cgroup_assign_new_page_cgroup(struct page *page, + struct page_cgroup *pc) { int ret = 0; @@ -198,8 +262,8 @@ page_cgroup_assign_new_page_cgroup(struct page *page, struct page_cgroup *pc) * clear_page_cgroup(page, pc) == pc */ -static inline struct page_cgroup * -clear_page_cgroup(struct page *page, struct page_cgroup *pc) +static struct page_cgroup *clear_page_cgroup(struct page *page, + struct page_cgroup *pc) { struct page_cgroup *ret; /* lock and clear */ @@ -211,7 +275,6 @@ clear_page_cgroup(struct page *page, struct page_cgroup *pc) return ret; } - static void __mem_cgroup_move_lists(struct page_cgroup *pc, bool active) { if (active) { @@ -426,6 +489,8 @@ retry: } spin_lock_irqsave(&mem->lru_lock, flags); + /* Update statistics vector */ + mem_cgroup_charge_statistics(mem, pc->flags, true); list_add(&pc->lru, &mem->active_list); spin_unlock_irqrestore(&mem->lru_lock, flags); @@ -496,6 +561,7 @@ void mem_cgroup_uncharge(struct page_cgroup *pc) res_counter_uncharge(&mem->res, PAGE_SIZE); spin_lock_irqsave(&mem->lru_lock, flags); list_del_init(&pc->lru); + mem_cgroup_charge_statistics(mem, pc->flags, false); spin_unlock_irqrestore(&mem->lru_lock, flags); kfree(pc); } @@ -572,6 +638,7 @@ retry: css_put(&mem->css); res_counter_uncharge(&mem->res, PAGE_SIZE); list_del_init(&pc->lru); + mem_cgroup_charge_statistics(mem, pc->flags, false); kfree(pc); } else /* being uncharged ? ...do relax */ break;