diff options
Diffstat (limited to 'mm/memcontrol.c')
-rw-r--r-- | mm/memcontrol.c | 312 |
1 files changed, 195 insertions, 117 deletions
diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 287933005e11..e50a2db5b4ff 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -687,10 +687,119 @@ mem_cgroup_largest_soft_limit_node(struct mem_cgroup_tree_per_node *mctz) return mz; } -static unsigned long memcg_sum_events(struct mem_cgroup *memcg, - int event) +/** + * __mod_memcg_state - update cgroup memory statistics + * @memcg: the memory cgroup + * @idx: the stat item - can be enum memcg_stat_item or enum node_stat_item + * @val: delta to add to the counter, can be negative + */ +void __mod_memcg_state(struct mem_cgroup *memcg, int idx, int val) +{ + long x; + + if (mem_cgroup_disabled()) + return; + + x = val + __this_cpu_read(memcg->vmstats_percpu->stat[idx]); + if (unlikely(abs(x) > MEMCG_CHARGE_BATCH)) { + struct mem_cgroup *mi; + + atomic_long_add(x, &memcg->vmstats_local[idx]); + for (mi = memcg; mi; mi = parent_mem_cgroup(mi)) + atomic_long_add(x, &mi->vmstats[idx]); + x = 0; + } + __this_cpu_write(memcg->vmstats_percpu->stat[idx], x); +} + +static struct mem_cgroup_per_node * +parent_nodeinfo(struct mem_cgroup_per_node *pn, int nid) +{ + struct mem_cgroup *parent; + + parent = parent_mem_cgroup(pn->memcg); + if (!parent) + return NULL; + return mem_cgroup_nodeinfo(parent, nid); +} + +/** + * __mod_lruvec_state - update lruvec memory statistics + * @lruvec: the lruvec + * @idx: the stat item + * @val: delta to add to the counter, can be negative + * + * The lruvec is the intersection of the NUMA node and a cgroup. This + * function updates the all three counters that are affected by a + * change of state at this level: per-node, per-cgroup, per-lruvec. + */ +void __mod_lruvec_state(struct lruvec *lruvec, enum node_stat_item idx, + int val) { - return atomic_long_read(&memcg->events[event]); + pg_data_t *pgdat = lruvec_pgdat(lruvec); + struct mem_cgroup_per_node *pn; + struct mem_cgroup *memcg; + long x; + + /* Update node */ + __mod_node_page_state(pgdat, idx, val); + + if (mem_cgroup_disabled()) + return; + + pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec); + memcg = pn->memcg; + + /* Update memcg */ + __mod_memcg_state(memcg, idx, val); + + /* Update lruvec */ + x = val + __this_cpu_read(pn->lruvec_stat_cpu->count[idx]); + if (unlikely(abs(x) > MEMCG_CHARGE_BATCH)) { + struct mem_cgroup_per_node *pi; + + atomic_long_add(x, &pn->lruvec_stat_local[idx]); + for (pi = pn; pi; pi = parent_nodeinfo(pi, pgdat->node_id)) + atomic_long_add(x, &pi->lruvec_stat[idx]); + x = 0; + } + __this_cpu_write(pn->lruvec_stat_cpu->count[idx], x); +} + +/** + * __count_memcg_events - account VM events in a cgroup + * @memcg: the memory cgroup + * @idx: the event item + * @count: the number of events that occured + */ +void __count_memcg_events(struct mem_cgroup *memcg, enum vm_event_item idx, + unsigned long count) +{ + unsigned long x; + + if (mem_cgroup_disabled()) + return; + + x = count + __this_cpu_read(memcg->vmstats_percpu->events[idx]); + if (unlikely(x > MEMCG_CHARGE_BATCH)) { + struct mem_cgroup *mi; + + atomic_long_add(x, &memcg->vmevents_local[idx]); + for (mi = memcg; mi; mi = parent_mem_cgroup(mi)) + atomic_long_add(x, &mi->vmevents[idx]); + x = 0; + } + __this_cpu_write(memcg->vmstats_percpu->events[idx], x); +} + +static unsigned long memcg_events(struct mem_cgroup *memcg, int event) +{ + return atomic_long_read(&memcg->vmevents[event]); +} + +static unsigned long memcg_events_local(struct mem_cgroup *memcg, int event) +{ + return atomic_long_read(&memcg->vmevents_local[event]); } static void mem_cgroup_charge_statistics(struct mem_cgroup *memcg, @@ -722,7 +831,7 @@ static void mem_cgroup_charge_statistics(struct mem_cgroup *memcg, nr_pages = -nr_pages; /* for event */ } - __this_cpu_add(memcg->stat_cpu->nr_page_events, nr_pages); + __this_cpu_add(memcg->vmstats_percpu->nr_page_events, nr_pages); } static bool mem_cgroup_event_ratelimit(struct mem_cgroup *memcg, @@ -730,8 +839,8 @@ static bool mem_cgroup_event_ratelimit(struct mem_cgroup *memcg, { unsigned long val, next; - val = __this_cpu_read(memcg->stat_cpu->nr_page_events); - next = __this_cpu_read(memcg->stat_cpu->targets[target]); + val = __this_cpu_read(memcg->vmstats_percpu->nr_page_events); + next = __this_cpu_read(memcg->vmstats_percpu->targets[target]); /* from time_after() in jiffies.h */ if ((long)(next - val) < 0) { switch (target) { @@ -747,7 +856,7 @@ static bool mem_cgroup_event_ratelimit(struct mem_cgroup *memcg, default: break; } - __this_cpu_write(memcg->stat_cpu->targets[target], next); + __this_cpu_write(memcg->vmstats_percpu->targets[target], next); return true; } return false; @@ -1325,12 +1434,14 @@ void mem_cgroup_print_oom_meminfo(struct mem_cgroup *memcg) if (memcg1_stats[i] == MEMCG_SWAP && !do_swap_account) continue; pr_cont(" %s:%luKB", memcg1_stat_names[i], - K(memcg_page_state(iter, memcg1_stats[i]))); + K(memcg_page_state_local(iter, + memcg1_stats[i]))); } for (i = 0; i < NR_LRU_LISTS; i++) pr_cont(" %s:%luKB", mem_cgroup_lru_names[i], - K(memcg_page_state(iter, NR_LRU_BASE + i))); + K(memcg_page_state_local(iter, + NR_LRU_BASE + i))); pr_cont("\n"); } @@ -2076,7 +2187,7 @@ static void drain_all_stock(struct mem_cgroup *root_memcg) static int memcg_hotplug_cpu_dead(unsigned int cpu) { struct memcg_stock_pcp *stock; - struct mem_cgroup *memcg; + struct mem_cgroup *memcg, *mi; stock = &per_cpu(memcg_stock, cpu); drain_stock(stock); @@ -2088,9 +2199,12 @@ static int memcg_hotplug_cpu_dead(unsigned int cpu) int nid; long x; - x = this_cpu_xchg(memcg->stat_cpu->count[i], 0); - if (x) - atomic_long_add(x, &memcg->stat[i]); + x = this_cpu_xchg(memcg->vmstats_percpu->stat[i], 0); + if (x) { + atomic_long_add(x, &memcg->vmstats_local[i]); + for (mi = memcg; mi; mi = parent_mem_cgroup(mi)) + atomic_long_add(x, &memcg->vmstats[i]); + } if (i >= NR_VM_NODE_STAT_ITEMS) continue; @@ -2100,17 +2214,24 @@ static int memcg_hotplug_cpu_dead(unsigned int cpu) pn = mem_cgroup_nodeinfo(memcg, nid); x = this_cpu_xchg(pn->lruvec_stat_cpu->count[i], 0); - if (x) - atomic_long_add(x, &pn->lruvec_stat[i]); + if (x) { + atomic_long_add(x, &pn->lruvec_stat_local[i]); + do { + atomic_long_add(x, &pn->lruvec_stat[i]); + } while ((pn = parent_nodeinfo(pn, nid))); + } } } for (i = 0; i < NR_VM_EVENT_ITEMS; i++) { long x; - x = this_cpu_xchg(memcg->stat_cpu->events[i], 0); - if (x) - atomic_long_add(x, &memcg->events[i]); + x = this_cpu_xchg(memcg->vmstats_percpu->events[i], 0); + if (x) { + atomic_long_add(x, &memcg->vmevents_local[i]); + for (mi = memcg; mi; mi = parent_mem_cgroup(mi)) + atomic_long_add(x, &memcg->vmevents[i]); + } } } @@ -2940,50 +3061,15 @@ static int mem_cgroup_hierarchy_write(struct cgroup_subsys_state *css, return retval; } -struct accumulated_stats { - unsigned long stat[MEMCG_NR_STAT]; - unsigned long events[NR_VM_EVENT_ITEMS]; - unsigned long lru_pages[NR_LRU_LISTS]; - const unsigned int *stats_array; - const unsigned int *events_array; - int stats_size; - int events_size; -}; - -static void accumulate_memcg_tree(struct mem_cgroup *memcg, - struct accumulated_stats *acc) -{ - struct mem_cgroup *mi; - int i; - - for_each_mem_cgroup_tree(mi, memcg) { - for (i = 0; i < acc->stats_size; i++) - acc->stat[i] += memcg_page_state(mi, - acc->stats_array ? acc->stats_array[i] : i); - - for (i = 0; i < acc->events_size; i++) - acc->events[i] += memcg_sum_events(mi, - acc->events_array ? acc->events_array[i] : i); - - for (i = 0; i < NR_LRU_LISTS; i++) - acc->lru_pages[i] += memcg_page_state(mi, - NR_LRU_BASE + i); - } -} - static unsigned long mem_cgroup_usage(struct mem_cgroup *memcg, bool swap) { - unsigned long val = 0; + unsigned long val; if (mem_cgroup_is_root(memcg)) { - struct mem_cgroup *iter; - - for_each_mem_cgroup_tree(iter, memcg) { - val += memcg_page_state(iter, MEMCG_CACHE); - val += memcg_page_state(iter, MEMCG_RSS); - if (swap) - val += memcg_page_state(iter, MEMCG_SWAP); - } + val = memcg_page_state(memcg, MEMCG_CACHE) + + memcg_page_state(memcg, MEMCG_RSS); + if (swap) + val += memcg_page_state(memcg, MEMCG_SWAP); } else { if (!swap) val = page_counter_read(&memcg->memory); @@ -3324,7 +3410,7 @@ static unsigned long mem_cgroup_node_nr_lru_pages(struct mem_cgroup *memcg, for_each_lru(lru) { if (!(BIT(lru) & lru_mask)) continue; - nr += lruvec_page_state(lruvec, NR_LRU_BASE + lru); + nr += lruvec_page_state_local(lruvec, NR_LRU_BASE + lru); } return nr; } @@ -3338,7 +3424,7 @@ static unsigned long mem_cgroup_nr_lru_pages(struct mem_cgroup *memcg, for_each_lru(lru) { if (!(BIT(lru) & lru_mask)) continue; - nr += memcg_page_state(memcg, NR_LRU_BASE + lru); + nr += memcg_page_state_local(memcg, NR_LRU_BASE + lru); } return nr; } @@ -3414,7 +3500,6 @@ static int memcg_stat_show(struct seq_file *m, void *v) unsigned long memory, memsw; struct mem_cgroup *mi; unsigned int i; - struct accumulated_stats acc; BUILD_BUG_ON(ARRAY_SIZE(memcg1_stat_names) != ARRAY_SIZE(memcg1_stats)); BUILD_BUG_ON(ARRAY_SIZE(mem_cgroup_lru_names) != NR_LRU_LISTS); @@ -3423,17 +3508,17 @@ static int memcg_stat_show(struct seq_file *m, void *v) if (memcg1_stats[i] == MEMCG_SWAP && !do_memsw_account()) continue; seq_printf(m, "%s %lu\n", memcg1_stat_names[i], - memcg_page_state(memcg, memcg1_stats[i]) * + memcg_page_state_local(memcg, memcg1_stats[i]) * PAGE_SIZE); } for (i = 0; i < ARRAY_SIZE(memcg1_events); i++) seq_printf(m, "%s %lu\n", memcg1_event_names[i], - memcg_sum_events(memcg, memcg1_events[i])); + memcg_events_local(memcg, memcg1_events[i])); for (i = 0; i < NR_LRU_LISTS; i++) seq_printf(m, "%s %lu\n", mem_cgroup_lru_names[i], - memcg_page_state(memcg, NR_LRU_BASE + i) * + memcg_page_state_local(memcg, NR_LRU_BASE + i) * PAGE_SIZE); /* Hierarchical information */ @@ -3448,27 +3533,21 @@ static int memcg_stat_show(struct seq_file *m, void *v) seq_printf(m, "hierarchical_memsw_limit %llu\n", (u64)memsw * PAGE_SIZE); - memset(&acc, 0, sizeof(acc)); - acc.stats_size = ARRAY_SIZE(memcg1_stats); - acc.stats_array = memcg1_stats; - acc.events_size = ARRAY_SIZE(memcg1_events); - acc.events_array = memcg1_events; - accumulate_memcg_tree(memcg, &acc); - for (i = 0; i < ARRAY_SIZE(memcg1_stats); i++) { if (memcg1_stats[i] == MEMCG_SWAP && !do_memsw_account()) continue; seq_printf(m, "total_%s %llu\n", memcg1_stat_names[i], - (u64)acc.stat[i] * PAGE_SIZE); + (u64)memcg_page_state(memcg, i) * PAGE_SIZE); } for (i = 0; i < ARRAY_SIZE(memcg1_events); i++) seq_printf(m, "total_%s %llu\n", memcg1_event_names[i], - (u64)acc.events[i]); + (u64)memcg_events(memcg, i)); for (i = 0; i < NR_LRU_LISTS; i++) seq_printf(m, "total_%s %llu\n", mem_cgroup_lru_names[i], - (u64)acc.lru_pages[i] * PAGE_SIZE); + (u64)memcg_page_state(memcg, NR_LRU_BASE + i) * + PAGE_SIZE); #ifdef CONFIG_DEBUG_VM { @@ -3901,11 +3980,11 @@ struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb) */ static unsigned long memcg_exact_page_state(struct mem_cgroup *memcg, int idx) { - long x = atomic_long_read(&memcg->stat[idx]); + long x = atomic_long_read(&memcg->vmstats[idx]); int cpu; for_each_online_cpu(cpu) - x += per_cpu_ptr(memcg->stat_cpu, cpu)->count[idx]; + x += per_cpu_ptr(memcg->vmstats_percpu, cpu)->stat[idx]; if (x < 0) x = 0; return x; @@ -4445,7 +4524,7 @@ static void __mem_cgroup_free(struct mem_cgroup *memcg) for_each_node(node) free_mem_cgroup_per_node_info(memcg, node); - free_percpu(memcg->stat_cpu); + free_percpu(memcg->vmstats_percpu); kfree(memcg); } @@ -4474,8 +4553,8 @@ static struct mem_cgroup *mem_cgroup_alloc(void) if (memcg->id.id < 0) goto fail; - memcg->stat_cpu = alloc_percpu(struct mem_cgroup_stat_cpu); - if (!memcg->stat_cpu) + memcg->vmstats_percpu = alloc_percpu(struct memcg_vmstats_percpu); + if (!memcg->vmstats_percpu) goto fail; for_each_node(node) @@ -5561,7 +5640,6 @@ static int memory_events_show(struct seq_file *m, void *v) static int memory_stat_show(struct seq_file *m, void *v) { struct mem_cgroup *memcg = mem_cgroup_from_seq(m); - struct accumulated_stats acc; int i; /* @@ -5575,31 +5653,27 @@ static int memory_stat_show(struct seq_file *m, void *v) * Current memory state: */ - memset(&acc, 0, sizeof(acc)); - acc.stats_size = MEMCG_NR_STAT; - acc.events_size = NR_VM_EVENT_ITEMS; - accumulate_memcg_tree(memcg, &acc); - seq_printf(m, "anon %llu\n", - (u64)acc.stat[MEMCG_RSS] * PAGE_SIZE); + (u64)memcg_page_state(memcg, MEMCG_RSS) * PAGE_SIZE); seq_printf(m, "file %llu\n", - (u64)acc.stat[MEMCG_CACHE] * PAGE_SIZE); + (u64)memcg_page_state(memcg, MEMCG_CACHE) * PAGE_SIZE); seq_printf(m, "kernel_stack %llu\n", - (u64)acc.stat[MEMCG_KERNEL_STACK_KB] * 1024); + (u64)memcg_page_state(memcg, MEMCG_KERNEL_STACK_KB) * 1024); seq_printf(m, "slab %llu\n", - (u64)(acc.stat[NR_SLAB_RECLAIMABLE] + - acc.stat[NR_SLAB_UNRECLAIMABLE]) * PAGE_SIZE); + (u64)(memcg_page_state(memcg, NR_SLAB_RECLAIMABLE) + + memcg_page_state(memcg, NR_SLAB_UNRECLAIMABLE)) * + PAGE_SIZE); seq_printf(m, "sock %llu\n", - (u64)acc.stat[MEMCG_SOCK] * PAGE_SIZE); + (u64)memcg_page_state(memcg, MEMCG_SOCK) * PAGE_SIZE); seq_printf(m, "shmem %llu\n", - (u64)acc.stat[NR_SHMEM] * PAGE_SIZE); + (u64)memcg_page_state(memcg, NR_SHMEM) * PAGE_SIZE); seq_printf(m, "file_mapped %llu\n", - (u64)acc.stat[NR_FILE_MAPPED] * PAGE_SIZE); + (u64)memcg_page_state(memcg, NR_FILE_MAPPED) * PAGE_SIZE); seq_printf(m, "file_dirty %llu\n", - (u64)acc.stat[NR_FILE_DIRTY] * PAGE_SIZE); + (u64)memcg_page_state(memcg, NR_FILE_DIRTY) * PAGE_SIZE); seq_printf(m, "file_writeback %llu\n", - (u64)acc.stat[NR_WRITEBACK] * PAGE_SIZE); + (u64)memcg_page_state(memcg, NR_WRITEBACK) * PAGE_SIZE); /* * TODO: We should eventually replace our own MEMCG_RSS_HUGE counter @@ -5608,43 +5682,47 @@ static int memory_stat_show(struct seq_file *m, void *v) * where the page->mem_cgroup is set up and stable. */ seq_printf(m, "anon_thp %llu\n", - (u64)acc.stat[MEMCG_RSS_HUGE] * PAGE_SIZE); + (u64)memcg_page_state(memcg, MEMCG_RSS_HUGE) * PAGE_SIZE); for (i = 0; i < NR_LRU_LISTS; i++) seq_printf(m, "%s %llu\n", mem_cgroup_lru_names[i], - (u64)acc.lru_pages[i] * PAGE_SIZE); + (u64)memcg_page_state(memcg, NR_LRU_BASE + i) * + PAGE_SIZE); seq_printf(m, "slab_reclaimable %llu\n", - (u64)acc.stat[NR_SLAB_RECLAIMABLE] * PAGE_SIZE); + (u64)memcg_page_state(memcg, NR_SLAB_RECLAIMABLE) * + PAGE_SIZE); seq_printf(m, "slab_unreclaimable %llu\n", - (u64)acc.stat[NR_SLAB_UNRECLAIMABLE] * PAGE_SIZE); + (u64)memcg_page_state(memcg, NR_SLAB_UNRECLAIMABLE) * + PAGE_SIZE); /* Accumulated memory events */ - seq_printf(m, "pgfault %lu\n", acc.events[PGFAULT]); - seq_printf(m, "pgmajfault %lu\n", acc.events[PGMAJFAULT]); + seq_printf(m, "pgfault %lu\n", memcg_events(memcg, PGFAULT)); + seq_printf(m, "pgmajfault %lu\n", memcg_events(memcg, PGMAJFAULT)); seq_printf(m, "workingset_refault %lu\n", - acc.stat[WORKINGSET_REFAULT]); + memcg_page_state(memcg, WORKINGSET_REFAULT)); seq_printf(m, "workingset_activate %lu\n", - acc.stat[WORKINGSET_ACTIVATE]); + memcg_page_state(memcg, WORKINGSET_ACTIVATE)); seq_printf(m, "workingset_nodereclaim %lu\n", - acc.stat[WORKINGSET_NODERECLAIM]); - - seq_printf(m, "pgrefill %lu\n", acc.events[PGREFILL]); - seq_printf(m, "pgscan %lu\n", acc.events[PGSCAN_KSWAPD] + - acc.events[PGSCAN_DIRECT]); - seq_printf(m, "pgsteal %lu\n", acc.events[PGSTEAL_KSWAPD] + - acc.events[PGSTEAL_DIRECT]); - seq_printf(m, "pgactivate %lu\n", acc.events[PGACTIVATE]); - seq_printf(m, "pgdeactivate %lu\n", acc.events[PGDEACTIVATE]); - seq_printf(m, "pglazyfree %lu\n", acc.events[PGLAZYFREE]); - seq_printf(m, "pglazyfreed %lu\n", acc.events[PGLAZYFREED]); + memcg_page_state(memcg, WORKINGSET_NODERECLAIM)); + + seq_printf(m, "pgrefill %lu\n", memcg_events(memcg, PGREFILL)); + seq_printf(m, "pgscan %lu\n", memcg_events(memcg, PGSCAN_KSWAPD) + + memcg_events(memcg, PGSCAN_DIRECT)); + seq_printf(m, "pgsteal %lu\n", memcg_events(memcg, PGSTEAL_KSWAPD) + + memcg_events(memcg, PGSTEAL_DIRECT)); + seq_printf(m, "pgactivate %lu\n", memcg_events(memcg, PGACTIVATE)); + seq_printf(m, "pgdeactivate %lu\n", memcg_events(memcg, PGDEACTIVATE)); + seq_printf(m, "pglazyfree %lu\n", memcg_events(memcg, PGLAZYFREE)); + seq_printf(m, "pglazyfreed %lu\n", memcg_events(memcg, PGLAZYFREED)); #ifdef CONFIG_TRANSPARENT_HUGEPAGE - seq_printf(m, "thp_fault_alloc %lu\n", acc.events[THP_FAULT_ALLOC]); + seq_printf(m, "thp_fault_alloc %lu\n", + memcg_events(memcg, THP_FAULT_ALLOC)); seq_printf(m, "thp_collapse_alloc %lu\n", - acc.events[THP_COLLAPSE_ALLOC]); + memcg_events(memcg, THP_COLLAPSE_ALLOC)); #endif /* CONFIG_TRANSPARENT_HUGEPAGE */ return 0; @@ -6080,7 +6158,7 @@ static void uncharge_batch(const struct uncharge_gather *ug) __mod_memcg_state(ug->memcg, MEMCG_RSS_HUGE, -ug->nr_huge); __mod_memcg_state(ug->memcg, NR_SHMEM, -ug->nr_shmem); __count_memcg_events(ug->memcg, PGPGOUT, ug->pgpgout); - __this_cpu_add(ug->memcg->stat_cpu->nr_page_events, nr_pages); + __this_cpu_add(ug->memcg->vmstats_percpu->nr_page_events, nr_pages); memcg_check_events(ug->memcg, ug->dummy_page); local_irq_restore(flags); |