struct page *page,
bool anon, int nr_pages)
{
- preempt_disable();
-
/*
* Here, RSS means 'mapped anon' and anon's SwapCache. Shmem/tmpfs is
* counted as CACHE even if it's on ANON LRU.
}
__this_cpu_add(memcg->stat->nr_page_events, nr_pages);
-
- preempt_enable();
}
unsigned long
return mem_cgroup_from_css(task_css(p, memory_cgrp_id));
}
-struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm)
+static struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm)
{
struct mem_cgroup *memcg = NULL;
- if (!mm)
- return NULL;
- /*
- * Because we have no locks, mm->owner's may be being moved to other
- * cgroup. We use css_tryget() here even if this looks
- * pessimistic (rather than adding locks here).
- */
rcu_read_lock();
do {
memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
if (unlikely(!memcg))
- break;
+ memcg = root_mem_cgroup;
} while (!css_tryget(&memcg->css));
rcu_read_unlock();
return memcg;
p = find_lock_task_mm(task);
if (p) {
- curr = try_get_mem_cgroup_from_mm(p->mm);
+ curr = get_mem_cgroup_from_mm(p->mm);
task_unlock(p);
} else {
/*
css_get(&curr->css);
rcu_read_unlock();
}
- if (!curr)
- return false;
/*
* We should check use_hierarchy of "memcg" not "curr". Because checking
* use_hierarchy of "curr" here make this function true if hierarchy is
}
-/* See __mem_cgroup_try_charge() for details */
+/* See mem_cgroup_try_charge() for details */
enum {
CHARGE_OK, /* success */
CHARGE_RETRY, /* need to retry but retry is not bad */
return CHARGE_NOMEM;
}
-/*
- * __mem_cgroup_try_charge() does
- * 1. detect memcg to be charged against from passed *mm and *ptr,
- * 2. update res_counter
- * 3. call memory reclaim if necessary.
- *
- * In some special case, if the task is fatal, fatal_signal_pending() or
- * has TIF_MEMDIE, this function returns -EINTR while writing root_mem_cgroup
- * to *ptr. There are two reasons for this. 1: fatal threads should quit as soon
- * as possible without any hazards. 2: all pages should have a valid
- * pc->mem_cgroup. If mm is NULL and the caller doesn't pass a valid memcg
- * pointer, that is treated as a charge to root_mem_cgroup.
- *
- * So __mem_cgroup_try_charge() will return
- * 0 ... on success, filling *ptr with a valid memcg pointer.
- * -ENOMEM ... charge failure because of resource limits.
- * -EINTR ... if thread is fatal. *ptr is filled with root_mem_cgroup.
+/**
+ * mem_cgroup_try_charge - try charging a memcg
+ * @memcg: memcg to charge
+ * @nr_pages: number of pages to charge
+ * @oom: trigger OOM if reclaim fails
*
- * Unlike the exported interface, an "oom" parameter is added. if oom==true,
- * the oom-killer can be invoked.
+ * Returns 0 if @memcg was charged successfully, -EINTR if the charge
+ * was bypassed to root_mem_cgroup, and -ENOMEM if the charge failed.
*/
-static int __mem_cgroup_try_charge(struct mm_struct *mm,
- gfp_t gfp_mask,
- unsigned int nr_pages,
- struct mem_cgroup **ptr,
- bool oom)
+static int mem_cgroup_try_charge(struct mem_cgroup *memcg,
+ gfp_t gfp_mask,
+ unsigned int nr_pages,
+ bool oom)
{
unsigned int batch = max(CHARGE_BATCH, nr_pages);
int nr_oom_retries = MEM_CGROUP_RECLAIM_RETRIES;
- struct mem_cgroup *memcg = NULL;
int ret;
+ if (mem_cgroup_is_root(memcg))
+ goto done;
/*
- * Unlike gloval-vm's OOM-kill, we're not in memory shortage
- * in system level. So, allow to go ahead dying process in addition to
- * MEMDIE process.
+ * Unlike in global OOM situations, memcg is not in a physical
+ * memory shortage. Allow dying and OOM-killed tasks to
+ * bypass the last charges so that they can exit quickly and
+ * free their memory.
*/
- if (unlikely(test_thread_flag(TIF_MEMDIE)
- || fatal_signal_pending(current)))
+ if (unlikely(test_thread_flag(TIF_MEMDIE) ||
+ fatal_signal_pending(current)))
goto bypass;
if (unlikely(task_in_memcg_oom(current)))
if (gfp_mask & __GFP_NOFAIL)
oom = false;
-
- /*
- * We always charge the cgroup the mm_struct belongs to.
- * The mm_struct's mem_cgroup changes on task migration if the
- * thread group leader migrates. It's possible that mm is not
- * set, if so charge the root memcg (happens for pagecache usage).
- */
- if (!*ptr && !mm)
- *ptr = root_mem_cgroup;
again:
- if (*ptr) { /* css should be a valid one */
- memcg = *ptr;
- if (mem_cgroup_is_root(memcg))
- goto done;
- if (consume_stock(memcg, nr_pages))
- goto done;
- css_get(&memcg->css);
- } else {
- struct task_struct *p;
-
- rcu_read_lock();
- p = rcu_dereference(mm->owner);
- /*
- * Because we don't have task_lock(), "p" can exit.
- * In that case, "memcg" can point to root or p can be NULL with
- * race with swapoff. Then, we have small risk of mis-accouning.
- * But such kind of mis-account by race always happens because
- * we don't have cgroup_mutex(). It's overkill and we allo that
- * small race, here.
- * (*) swapoff at el will charge against mm-struct not against
- * task-struct. So, mm->owner can be NULL.
- */
- memcg = mem_cgroup_from_task(p);
- if (!memcg)
- memcg = root_mem_cgroup;
- if (mem_cgroup_is_root(memcg)) {
- rcu_read_unlock();
- goto done;
- }
- if (consume_stock(memcg, nr_pages)) {
- /*
- * It seems dagerous to access memcg without css_get().
- * But considering how consume_stok works, it's not
- * necessary. If consume_stock success, some charges
- * from this memcg are cached on this cpu. So, we
- * don't need to call css_get()/css_tryget() before
- * calling consume_stock().
- */
- rcu_read_unlock();
- goto done;
- }
- /* after here, we may be blocked. we need to get refcnt */
- if (!css_tryget(&memcg->css)) {
- rcu_read_unlock();
- goto again;
- }
- rcu_read_unlock();
- }
+ if (consume_stock(memcg, nr_pages))
+ goto done;
do {
bool invoke_oom = oom && !nr_oom_retries;
/* If killed, bypass charge */
- if (fatal_signal_pending(current)) {
- css_put(&memcg->css);
+ if (fatal_signal_pending(current))
goto bypass;
- }
ret = mem_cgroup_do_charge(memcg, gfp_mask, batch,
nr_pages, invoke_oom);
break;
case CHARGE_RETRY: /* not in OOM situation but retry */
batch = nr_pages;
- css_put(&memcg->css);
- memcg = NULL;
goto again;
case CHARGE_WOULDBLOCK: /* !__GFP_WAIT */
- css_put(&memcg->css);
goto nomem;
case CHARGE_NOMEM: /* OOM routine works */
- if (!oom || invoke_oom) {
- css_put(&memcg->css);
+ if (!oom || invoke_oom)
goto nomem;
- }
nr_oom_retries--;
break;
}
if (batch > nr_pages)
refill_stock(memcg, batch - nr_pages);
- css_put(&memcg->css);
done:
- *ptr = memcg;
return 0;
nomem:
- if (!(gfp_mask & __GFP_NOFAIL)) {
- *ptr = NULL;
+ if (!(gfp_mask & __GFP_NOFAIL))
return -ENOMEM;
- }
bypass:
- *ptr = root_mem_cgroup;
return -EINTR;
}
+/**
+ * mem_cgroup_try_charge_mm - try charging a mm
+ * @mm: mm_struct to charge
+ * @nr_pages: number of pages to charge
+ * @oom: trigger OOM if reclaim fails
+ *
+ * Returns the charged mem_cgroup associated with the given mm_struct or
+ * NULL the charge failed.
+ */
+static struct mem_cgroup *mem_cgroup_try_charge_mm(struct mm_struct *mm,
+ gfp_t gfp_mask,
+ unsigned int nr_pages,
+ bool oom)
+
+{
+ struct mem_cgroup *memcg;
+ int ret;
+
+ memcg = get_mem_cgroup_from_mm(mm);
+ ret = mem_cgroup_try_charge(memcg, gfp_mask, nr_pages, oom);
+ css_put(&memcg->css);
+ if (ret == -EINTR)
+ memcg = root_mem_cgroup;
+ else if (ret)
+ memcg = NULL;
+
+ return memcg;
+}
+
/*
* Somemtimes we have to undo a charge we got by try_charge().
* This function is for that and do uncharge, put css's refcnt.
static int memcg_charge_kmem(struct mem_cgroup *memcg, gfp_t gfp, u64 size)
{
struct res_counter *fail_res;
- struct mem_cgroup *_memcg;
int ret = 0;
ret = res_counter_charge(&memcg->kmem, size, &fail_res);
if (ret)
return ret;
- _memcg = memcg;
- ret = __mem_cgroup_try_charge(NULL, gfp, size >> PAGE_SHIFT,
- &_memcg, oom_gfp_allowed(gfp));
-
+ ret = mem_cgroup_try_charge(memcg, gfp, size >> PAGE_SHIFT,
+ oom_gfp_allowed(gfp));
if (ret == -EINTR) {
/*
- * __mem_cgroup_try_charge() chosed to bypass to root due to
+ * mem_cgroup_try_charge() chosed to bypass to root due to
* OOM kill or fatal signal. Since our only options are to
* either fail the allocation or charge it to this cgroup, do
* it as a temporary condition. But we can't fail. From a
*
* This condition will only trigger if the task entered
* memcg_charge_kmem in a sane state, but was OOM-killed during
- * __mem_cgroup_try_charge() above. Tasks that were already
+ * mem_cgroup_try_charge() above. Tasks that were already
* dying when the allocation triggers should have been already
* directed to the root cgroup in memcontrol.h
*/
return 0;
}
+char *memcg_create_cache_name(struct mem_cgroup *memcg,
+ struct kmem_cache *root_cache)
+{
+ static char *buf = NULL;
+
+ /*
+ * We need a mutex here to protect the shared buffer. Since this is
+ * expected to be called only on cache creation, we can employ the
+ * slab_mutex for that purpose.
+ */
+ lockdep_assert_held(&slab_mutex);
+
+ if (!buf) {
+ buf = kmalloc(NAME_MAX + 1, GFP_KERNEL);
+ if (!buf)
+ return NULL;
+ }
+
+ cgroup_name(memcg->css.cgroup, buf, NAME_MAX + 1);
+ return kasprintf(GFP_KERNEL, "%s(%d:%s)", root_cache->name,
+ memcg_cache_id(memcg), buf);
+}
+
int memcg_alloc_cache_params(struct mem_cgroup *memcg, struct kmem_cache *s,
struct kmem_cache *root_cache)
{
s->memcg_params->root_cache = root_cache;
INIT_WORK(&s->memcg_params->destroy,
kmem_cache_destroy_work_func);
+ css_get(&memcg->css);
} else
s->memcg_params->is_root_cache = true;
void memcg_free_cache_params(struct kmem_cache *s)
{
+ if (!s->memcg_params)
+ return;
+ if (!s->memcg_params->is_root_cache)
+ css_put(&s->memcg_params->memcg->css);
kfree(s->memcg_params);
}
memcg = s->memcg_params->memcg;
id = memcg_cache_id(memcg);
- css_get(&memcg->css);
-
-
/*
* Since readers won't lock (see cache_from_memcg_idx()), we need a
* barrier here to ensure nobody will see the kmem_cache partially
* after removing it from the memcg_slab_caches list, otherwise we can
* fail to convert memcg_params_to_cache() while traversing the list.
*/
- VM_BUG_ON(!root->memcg_params->memcg_caches[id]);
+ VM_BUG_ON(root->memcg_params->memcg_caches[id] != s);
root->memcg_params->memcg_caches[id] = NULL;
-
- css_put(&memcg->css);
}
/*
schedule_work(&cachep->memcg_params->destroy);
}
-static struct kmem_cache *memcg_create_kmem_cache(struct mem_cgroup *memcg,
- struct kmem_cache *s)
-{
- struct kmem_cache *new = NULL;
- static char *tmp_path = NULL, *tmp_name = NULL;
- static DEFINE_MUTEX(mutex); /* protects tmp_name */
-
- BUG_ON(!memcg_can_account_kmem(memcg));
-
- mutex_lock(&mutex);
- /*
- * kmem_cache_create_memcg duplicates the given name and
- * cgroup_name for this name requires RCU context.
- * This static temporary buffer is used to prevent from
- * pointless shortliving allocation.
- */
- if (!tmp_path || !tmp_name) {
- if (!tmp_path)
- tmp_path = kmalloc(PATH_MAX, GFP_KERNEL);
- if (!tmp_name)
- tmp_name = kmalloc(NAME_MAX + 1, GFP_KERNEL);
- if (!tmp_path || !tmp_name)
- goto out;
- }
-
- cgroup_name(memcg->css.cgroup, tmp_name, NAME_MAX + 1);
- snprintf(tmp_path, PATH_MAX, "%s(%d:%s)", s->name,
- memcg_cache_id(memcg), tmp_name);
-
- new = kmem_cache_create_memcg(memcg, tmp_path, s->object_size, s->align,
- (s->flags & ~SLAB_PANIC), s->ctor, s);
- if (new)
- new->allocflags |= __GFP_KMEMCG;
- else
- new = s;
-out:
- mutex_unlock(&mutex);
- return new;
-}
-
-void kmem_cache_destroy_memcg_children(struct kmem_cache *s)
+int __kmem_cache_destroy_memcg_children(struct kmem_cache *s)
{
struct kmem_cache *c;
- int i;
-
- if (!s->memcg_params)
- return;
- if (!s->memcg_params->is_root_cache)
- return;
+ int i, failed = 0;
/*
* If the cache is being destroyed, we trust that there is no one else
c->memcg_params->dead = false;
cancel_work_sync(&c->memcg_params->destroy);
kmem_cache_destroy(c);
+
+ if (cache_from_memcg_idx(s, i))
+ failed++;
}
mutex_unlock(&activate_kmem_mutex);
+ return failed;
}
-struct create_work {
- struct mem_cgroup *memcg;
- struct kmem_cache *cachep;
- struct work_struct work;
-};
-
static void mem_cgroup_destroy_all_caches(struct mem_cgroup *memcg)
{
struct kmem_cache *cachep;
mutex_unlock(&memcg->slab_caches_mutex);
}
+struct create_work {
+ struct mem_cgroup *memcg;
+ struct kmem_cache *cachep;
+ struct work_struct work;
+};
+
static void memcg_create_cache_work_func(struct work_struct *w)
{
- struct create_work *cw;
+ struct create_work *cw = container_of(w, struct create_work, work);
+ struct mem_cgroup *memcg = cw->memcg;
+ struct kmem_cache *cachep = cw->cachep;
- cw = container_of(w, struct create_work, work);
- memcg_create_kmem_cache(cw->memcg, cw->cachep);
- css_put(&cw->memcg->css);
+ kmem_cache_create_memcg(memcg, cachep);
+ css_put(&memcg->css);
kfree(cw);
}
if (!current->mm || current->memcg_kmem_skip_account)
return true;
- memcg = try_get_mem_cgroup_from_mm(current->mm);
-
- /*
- * very rare case described in mem_cgroup_from_task. Unfortunately there
- * isn't much we can do without complicating this too much, and it would
- * be gfp-dependent anyway. Just let it go
- */
- if (unlikely(!memcg))
- return true;
+ memcg = get_mem_cgroup_from_mm(current->mm);
if (!memcg_can_account_kmem(memcg)) {
css_put(&memcg->css);
}
#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
-static inline
-void mem_cgroup_move_account_page_stat(struct mem_cgroup *from,
- struct mem_cgroup *to,
- unsigned int nr_pages,
- enum mem_cgroup_stat_index idx)
-{
- /* Update stat data for mem_cgroup */
- preempt_disable();
- __this_cpu_sub(from->stat->count[idx], nr_pages);
- __this_cpu_add(to->stat->count[idx], nr_pages);
- preempt_enable();
-}
-
/**
* mem_cgroup_move_account - move account of the page
* @page: the page
move_lock_mem_cgroup(from, &flags);
- if (!anon && page_mapped(page))
- mem_cgroup_move_account_page_stat(from, to, nr_pages,
- MEM_CGROUP_STAT_FILE_MAPPED);
+ if (!anon && page_mapped(page)) {
+ __this_cpu_sub(from->stat->count[MEM_CGROUP_STAT_FILE_MAPPED],
+ nr_pages);
+ __this_cpu_add(to->stat->count[MEM_CGROUP_STAT_FILE_MAPPED],
+ nr_pages);
+ }
- if (PageWriteback(page))
- mem_cgroup_move_account_page_stat(from, to, nr_pages,
- MEM_CGROUP_STAT_WRITEBACK);
+ if (PageWriteback(page)) {
+ __this_cpu_sub(from->stat->count[MEM_CGROUP_STAT_WRITEBACK],
+ nr_pages);
+ __this_cpu_add(to->stat->count[MEM_CGROUP_STAT_WRITEBACK],
+ nr_pages);
+ }
mem_cgroup_charge_statistics(from, page, anon, -nr_pages);
return ret;
}
-/*
- * Charge the memory controller for page usage.
- * Return
- * 0 if the charge was successful
- * < 0 if the cgroup is over its limit
- */
-static int mem_cgroup_charge_common(struct page *page, struct mm_struct *mm,
- gfp_t gfp_mask, enum charge_type ctype)
+int mem_cgroup_charge_anon(struct page *page,
+ struct mm_struct *mm, gfp_t gfp_mask)
{
- struct mem_cgroup *memcg = NULL;
unsigned int nr_pages = 1;
+ struct mem_cgroup *memcg;
bool oom = true;
- int ret;
+
+ if (mem_cgroup_disabled())
+ return 0;
+
+ VM_BUG_ON_PAGE(page_mapped(page), page);
+ VM_BUG_ON_PAGE(page->mapping && !PageAnon(page), page);
+ VM_BUG_ON(!mm);
if (PageTransHuge(page)) {
nr_pages <<= compound_order(page);
oom = false;
}
- ret = __mem_cgroup_try_charge(mm, gfp_mask, nr_pages, &memcg, oom);
- if (ret == -ENOMEM)
- return ret;
- __mem_cgroup_commit_charge(memcg, page, nr_pages, ctype, false);
+ memcg = mem_cgroup_try_charge_mm(mm, gfp_mask, nr_pages, oom);
+ if (!memcg)
+ return -ENOMEM;
+ __mem_cgroup_commit_charge(memcg, page, nr_pages,
+ MEM_CGROUP_CHARGE_TYPE_ANON, false);
return 0;
}
-int mem_cgroup_newpage_charge(struct page *page,
- struct mm_struct *mm, gfp_t gfp_mask)
-{
- if (mem_cgroup_disabled())
- return 0;
- VM_BUG_ON_PAGE(page_mapped(page), page);
- VM_BUG_ON_PAGE(page->mapping && !PageAnon(page), page);
- VM_BUG_ON(!mm);
- return mem_cgroup_charge_common(page, mm, gfp_mask,
- MEM_CGROUP_CHARGE_TYPE_ANON);
-}
-
/*
* While swap-in, try_charge -> commit or cancel, the page is locked.
* And when try_charge() successfully returns, one refcnt to memcg without
gfp_t mask,
struct mem_cgroup **memcgp)
{
- struct mem_cgroup *memcg;
+ struct mem_cgroup *memcg = NULL;
struct page_cgroup *pc;
int ret;
* in turn serializes uncharging.
*/
if (PageCgroupUsed(pc))
- return 0;
- if (!do_swap_account)
- goto charge_cur_mm;
- memcg = try_get_mem_cgroup_from_page(page);
+ goto out;
+ if (do_swap_account)
+ memcg = try_get_mem_cgroup_from_page(page);
if (!memcg)
- goto charge_cur_mm;
- *memcgp = memcg;
- ret = __mem_cgroup_try_charge(NULL, mask, 1, memcgp, true);
+ memcg = get_mem_cgroup_from_mm(mm);
+ ret = mem_cgroup_try_charge(memcg, mask, 1, true);
css_put(&memcg->css);
if (ret == -EINTR)
- ret = 0;
- return ret;
-charge_cur_mm:
- ret = __mem_cgroup_try_charge(mm, mask, 1, memcgp, true);
- if (ret == -EINTR)
- ret = 0;
- return ret;
+ memcg = root_mem_cgroup;
+ else if (ret)
+ return ret;
+out:
+ *memcgp = memcg;
+ return 0;
}
int mem_cgroup_try_charge_swapin(struct mm_struct *mm, struct page *page,
gfp_t gfp_mask, struct mem_cgroup **memcgp)
{
- *memcgp = NULL;
- if (mem_cgroup_disabled())
+ if (mem_cgroup_disabled()) {
+ *memcgp = NULL;
return 0;
+ }
/*
* A racing thread's fault, or swapoff, may have already
* updated the pte, and even removed page from swap cache: in
* there's also a KSM case which does need to charge the page.
*/
if (!PageSwapCache(page)) {
- int ret;
+ struct mem_cgroup *memcg;
- ret = __mem_cgroup_try_charge(mm, gfp_mask, 1, memcgp, true);
- if (ret == -EINTR)
- ret = 0;
- return ret;
+ memcg = mem_cgroup_try_charge_mm(mm, gfp_mask, 1, true);
+ if (!memcg)
+ return -ENOMEM;
+ *memcgp = memcg;
+ return 0;
}
return __mem_cgroup_try_charge_swapin(mm, page, gfp_mask, memcgp);
}
MEM_CGROUP_CHARGE_TYPE_ANON);
}
-int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm,
+int mem_cgroup_charge_file(struct page *page, struct mm_struct *mm,
gfp_t gfp_mask)
{
- struct mem_cgroup *memcg = NULL;
enum charge_type type = MEM_CGROUP_CHARGE_TYPE_CACHE;
+ struct mem_cgroup *memcg;
int ret;
if (mem_cgroup_disabled())
if (PageCompound(page))
return 0;
- if (!PageSwapCache(page))
- ret = mem_cgroup_charge_common(page, mm, gfp_mask, type);
- else { /* page is swapcache/shmem */
+ if (PageSwapCache(page)) { /* shmem */
ret = __mem_cgroup_try_charge_swapin(mm, page,
gfp_mask, &memcg);
- if (!ret)
- __mem_cgroup_commit_charge_swapin(page, memcg, type);
+ if (ret)
+ return ret;
+ __mem_cgroup_commit_charge_swapin(page, memcg, type);
+ return 0;
}
- return ret;
+
+ /*
+ * Page cache insertions can happen without an actual mm
+ * context, e.g. during disk probing on boot.
+ */
+ if (unlikely(!mm))
+ memcg = root_mem_cgroup;
+ else {
+ memcg = mem_cgroup_try_charge_mm(mm, gfp_mask, 1, true);
+ if (!memcg)
+ return -ENOMEM;
+ }
+ __mem_cgroup_commit_charge(memcg, page, 1, type, false);
+ return 0;
}
static void mem_cgroup_do_uncharge(struct mem_cgroup *memcg,
batch_count = PRECHARGE_COUNT_AT_ONCE;
cond_resched();
}
- ret = __mem_cgroup_try_charge(NULL,
- GFP_KERNEL, 1, &memcg, false);
+ ret = mem_cgroup_try_charge(memcg, GFP_KERNEL, 1, false);
if (ret)
/* mem_cgroup_clear_mc() will do uncharge later */
return ret;
git.samba.org / sfrench / cifs-2.6.git / blobdiff