Linux-3.14.12内存管理笔记【内存溢出保护机制（OOM）】

   Linux系统内存管理中存在着一个称之为OOM killer（Out-Of-Memory killer）的机制，该机制主要用于内存监控，监控进程的内存使用量，当系统的内存耗尽时，其将根据算法选择性地kill了部分进程。本文分析的内存溢出保护机制，也就是OOM killer机制了。

回到伙伴管理算法中涉及的一函数__alloc_pages_nodemask()，其里面调用的__alloc_pages_slowpath()并未展开深入，而内存溢出保护机制则在此函数中。

先行查看一下__alloc_pages_slowpath()的实现：

【file:/ mm/page_alloc.h】
static inline struct page *
__alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
    struct zonelist *zonelist, enum zone_type high_zoneidx,
    nodemask_t *nodemask, struct zone *preferred_zone,
    int migratetype)
{
    const gfp_t wait = gfp_mask & __GFP_WAIT;
    struct page *page = NULL;
    int alloc_flags;
    unsigned long pages_reclaimed = 0;
    unsigned long did_some_progress;
    bool sync_migration = false;
    bool deferred_compaction = false;
    bool contended_compaction = false;

    /*
     * In the slowpath, we sanity check order to avoid ever trying to
     * reclaim >= MAX_ORDER areas which will never succeed. Callers may
     * be using allocators in order of preference for an area that is
     * too large.
     */
    if (order >= MAX_ORDER) {
        WARN_ON_ONCE(!(gfp_mask & __GFP_NOWARN));
        return NULL;
    }

    /*
     * GFP_THISNODE (meaning __GFP_THISNODE, __GFP_NORETRY and
     * __GFP_NOWARN set) should not cause reclaim since the subsystem
     * (f.e. slab) using GFP_THISNODE may choose to trigger reclaim
     * using a larger set of nodes after it has established that the
     * allowed per node queues are empty and that nodes are
     * over allocated.
     */
    if (IS_ENABLED(CONFIG_NUMA) &&
        (gfp_mask & GFP_THISNODE) == GFP_THISNODE)
        goto nopage;

restart:
    if (!(gfp_mask & __GFP_NO_KSWAPD))
        wake_all_kswapds(order, zonelist, high_zoneidx, preferred_zone);

    /*
     * OK, we're below the kswapd watermark and have kicked background
     * reclaim. Now things get more complex, so set up alloc_flags according
     * to how we want to proceed.
     */
    alloc_flags = gfp_to_alloc_flags(gfp_mask);

    /*
     * Find the true preferred zone if the allocation is unconstrained by
     * cpusets.
     */
    if (!(alloc_flags & ALLOC_CPUSET) && !nodemask)
        first_zones_zonelist(zonelist, high_zoneidx, NULL,
                    &preferred_zone);

rebalance:
    /* This is the last chance, in general, before the goto nopage. */
    page = get_page_from_freelist(gfp_mask, nodemask, order, zonelist,
            high_zoneidx, alloc_flags & ~ALLOC_NO_WATERMARKS,
            preferred_zone, migratetype);
    if (page)
        goto got_pg;

    /* Allocate without watermarks if the context allows */
    if (alloc_flags & ALLOC_NO_WATERMARKS) {
        /*
         * Ignore mempolicies if ALLOC_NO_WATERMARKS on the grounds
         * the allocation is high priority and these type of
         * allocations are system rather than user orientated
         */
        zonelist = node_zonelist(numa_node_id(), gfp_mask);

        page = __alloc_pages_high_priority(gfp_mask, order,
                zonelist, high_zoneidx, nodemask,
                preferred_zone, migratetype);
        if (page) {
            goto got_pg;
        }
    }

    /* Atomic allocations - we can't balance anything */
    if (!wait) {
        /*
         * All existing users of the deprecated __GFP_NOFAIL are
         * blockable, so warn of any new users that actually allow this
         * type of allocation to fail.
         */
        WARN_ON_ONCE(gfp_mask & __GFP_NOFAIL);
        goto nopage;
    }

    /* Avoid recursion of direct reclaim */
    if (current->flags & PF_MEMALLOC)
        goto nopage;

    /* Avoid allocations with no watermarks from looping endlessly */
    if (test_thread_flag(TIF_MEMDIE) && !(gfp_mask & __GFP_NOFAIL))
        goto nopage;

    /*
     * Try direct compaction. The first pass is asynchronous. Subsequent
     * attempts after direct reclaim are synchronous
     */
    page = __alloc_pages_direct_compact(gfp_mask, order,
                    zonelist, high_zoneidx,
                    nodemask,
                    alloc_flags, preferred_zone,
                    migratetype, sync_migration,
                    &contended_compaction,
                    &deferred_compaction,
                    &did_some_progress);
    if (page)
        goto got_pg;
    sync_migration = true;

    /*
     * If compaction is deferred for high-order allocations, it is because
     * sync compaction recently failed. In this is the case and the caller
     * requested a movable allocation that does not heavily disrupt the
     * system then fail the allocation instead of entering direct reclaim.
     */
    if ((deferred_compaction || contended_compaction) &&
                        (gfp_mask & __GFP_NO_KSWAPD))
        goto nopage;

    /* Try direct reclaim and then allocating */
    page = __alloc_pages_direct_reclaim(gfp_mask, order,
                    zonelist, high_zoneidx,
                    nodemask,
                    alloc_flags, preferred_zone,
                    migratetype, &did_some_progress);
    if (page)
        goto got_pg;

    /*
     * If we failed to make any progress reclaiming, then we are
     * running out of options and have to consider going OOM
     */
    if (!did_some_progress) {
        if (oom_gfp_allowed(gfp_mask)) {
            if (oom_killer_disabled)
                goto nopage;
            /* Coredumps can quickly deplete all memory reserves */
            if ((current->flags & PF_DUMPCORE) &&
                !(gfp_mask & __GFP_NOFAIL))
                goto nopage;
            page = __alloc_pages_may_oom(gfp_mask, order,
                    zonelist, high_zoneidx,
                    nodemask, preferred_zone,
                    migratetype);
            if (page)
                goto got_pg;

            if (!(gfp_mask & __GFP_NOFAIL)) {
                /*
                 * The oom killer is not called for high-order
                 * allocations that may fail, so if no progress
                 * is being made, there are no other options and
                 * retrying is unlikely to help.
                 */
                if (order > PAGE_ALLOC_COSTLY_ORDER)
                    goto nopage;
                /*
                 * The oom killer is not called for lowmem
                 * allocations to prevent needlessly killing
                 * innocent tasks.
                 */
                if (high_zoneidx < ZONE_NORMAL)
                    goto nopage;
            }

            goto restart;
        }
    }

    /* Check if we should retry the allocation */
    pages_reclaimed += did_some_progress;
    if (should_alloc_retry(gfp_mask, order, did_some_progress,
                        pages_reclaimed)) {
        /* Wait for some write requests to complete then retry */
        wait_iff_congested(preferred_zone, BLK_RW_ASYNC, HZ/50);
        goto rebalance;
    } else {
        /*
         * High-order allocations do not necessarily loop after
         * direct reclaim and reclaim/compaction depends on compaction
         * being called after reclaim so call directly if necessary
         */
        page = __alloc_pages_direct_compact(gfp_mask, order,
                    zonelist, high_zoneidx,
                    nodemask,
                    alloc_flags, preferred_zone,
                    migratetype, sync_migration,
                    &contended_compaction,
                    &deferred_compaction,
                    &did_some_progress);
        if (page)
            goto got_pg;
    }

nopage:
    warn_alloc_failed(gfp_mask, order, NULL);
    return page;
got_pg:
    if (kmemcheck_enabled)
        kmemcheck_pagealloc_alloc(page, order, gfp_mask);

    return page;
}

该函数首先判断调用者是否禁止唤醒kswapd线程，若不做禁止则唤醒线程进行内存回收工作，然后通过gfp_to_alloc_flags()对内存分配标识进行调整，而后再次调用get_page_from_freelist()尝试分配，如果分配到则退出。否则继续尝试内存分配，继续尝试分配则先行判断是否设置了ALLOC_NO_WATERMARKS标识，如果设置了，则将忽略watermark，调用__alloc_pages_high_priority()进行分配。

__alloc_pages_high_priority()函数实现：

【file:/ mm/page_alloc.h】
/*
 * This is called in the allocator slow-path if the allocation request is of
 * sufficient urgency to ignore watermarks and take other desperate measures
 */
static inline struct page *
__alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order,
    struct zonelist *zonelist, enum zone_type high_zoneidx,
    nodemask_t *nodemask, struct zone *preferred_zone,
    int migratetype)
{
    struct page *page;

    do {
        page = get_page_from_freelist(gfp_mask, nodemask, order,
            zonelist, high_zoneidx, ALLOC_NO_WATERMARKS,
            preferred_zone, migratetype);

        if (!page && gfp_mask & __GFP_NOFAIL)
            wait_iff_congested(preferred_zone, BLK_RW_ASYNC, HZ/50);
    } while (!page && (gfp_mask & __GFP_NOFAIL));

    return page;
}

可以看到该函数根据分配标识__GFP_NOFAIL不断地调用get_page_from_freelist()循环尝试去获得内存。

接着回到__alloc_pages_slowpath()中，其从__alloc_pages_high_priority()退出后继而判断是否设置了__GFP_WAIT标识，如果设置则表示内存分配运行休眠，否则直接以分配内存失败而退出。接着将会调用__alloc_pages_direct_compact()和__alloc_pages_direct_reclaim()尝试回收内存并尝试分配。基于上面的多种尝试内存分配仍然失败的情况，将会调用__alloc_pages_may_oom()触发OOM killer机制。OOM killer将进程kill后会重新再次尝试内存分配，最后则是分配失败或分配成功的收尾处理。

__alloc_pages_slowpath()暂且分析至此，回到本文重点函数__alloc_pages_may_oom()中进一步进行分析。

【file:/ mm/page_alloc.h】
static inline struct page *
__alloc_pages_may_oom(gfp_t gfp_mask, unsigned int order,
    struct zonelist *zonelist, enum zone_type high_zoneidx,
    nodemask_t *nodemask, struct zone *preferred_zone,
    int migratetype)
{
    struct page *page;

    /* Acquire the OOM killer lock for the zones in zonelist */
    if (!try_set_zonelist_oom(zonelist, gfp_mask)) {
        schedule_timeout_uninterruptible(1);
        return NULL;
    }

    /*
     * Go through the zonelist yet one more time, keep very high watermark
     * here, this is only to catch a parallel oom killing, we must fail if
     * we're still under heavy pressure.
     */
    page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, nodemask,
        order, zonelist, high_zoneidx,
        ALLOC_WMARK_HIGH|ALLOC_CPUSET,
        preferred_zone, migratetype);
    if (page)
        goto out;

    if (!(gfp_mask & __GFP_NOFAIL)) {
        /* The OOM killer will not help higher order allocs */
        if (order > PAGE_ALLOC_COSTLY_ORDER)
            goto out;
        /* The OOM killer does not needlessly kill tasks for lowmem */
        if (high_zoneidx < ZONE_NORMAL)
            goto out;
        /*
         * GFP_THISNODE contains __GFP_NORETRY and we never hit this.
         * Sanity check for bare calls of __GFP_THISNODE, not real OOM.
         * The caller should handle page allocation failure by itself if
         * it specifies __GFP_THISNODE.
         * Note: Hugepage uses it but will hit PAGE_ALLOC_COSTLY_ORDER.
         */
        if (gfp_mask & __GFP_THISNODE)
            goto out;
    }
    /* Exhausted what can be done so it's blamo time */
    out_of_memory(zonelist, gfp_mask, order, nodemask, false);

out:
    clear_zonelist_oom(zonelist, gfp_mask);
    return page;
}

该函数首先通过try_set_zonelist_oom()判断OOM killer是否已经在其他核进行killing操作，如果没有的情况下将会在try_set_zonelist_oom()内部进行锁操作，确保只有一个核执行killing的操作。继而调用get_page_from_freelist()在高watermark的情况下尝试再次获取内存，不过这里注定会失败。接着就是调用到了关键函数out_of_memory()。最后函数退出时将会调用clear_zonelist_oom()清除掉try_set_zonelist_oom()里面的锁操作。

着重分析一下out_of_memory()：

【file:/ mm/oom_kill.c】
/**
 * out_of_memory - kill the "best" process when we run out of memory
 * @zonelist: zonelist pointer
 * @gfp_mask: memory allocation flags
 * @order: amount of memory being requested as a power of 2
 * @nodemask: nodemask passed to page allocator
 * @force_kill: true if a task must be killed, even if others are exiting
 *
 * If we run out of memory, we have the choice between either
 * killing a random task (bad), letting the system crash (worse)
 * OR try to be smart about which process to kill. Note that we
 * don't have to be perfect here, we just have to be good.
 */
void out_of_memory(struct zonelist *zonelist, gfp_t gfp_mask,
        int order, nodemask_t *nodemask, bool force_kill)
{
    const nodemask_t *mpol_mask;
    struct task_struct *p;
    unsigned long totalpages;
    unsigned long freed = 0;
    unsigned int uninitialized_var(points);
    enum oom_constraint constraint = CONSTRAINT_NONE;
    int killed = 0;

    blocking_notifier_call_chain(&oom_notify_list, 0, &freed);
    if (freed > 0)
        /* Got some memory back in the last second. */
        return;

    /*
     * If current has a pending SIGKILL or is exiting, then automatically
     * select it. The goal is to allow it to allocate so that it may
     * quickly exit and free its memory.
     */
    if (fatal_signal_pending(current) || current->flags & PF_EXITING) {
        set_thread_flag(TIF_MEMDIE);
        return;
    }

    /*
     * Check if there were limitations on the allocation (only relevant for
     * NUMA) that may require different handling.
     */
    constraint = constrained_alloc(zonelist, gfp_mask, nodemask,
                        &totalpages);
    mpol_mask = (constraint == CONSTRAINT_MEMORY_POLICY) ? nodemask : NULL;
    check_panic_on_oom(constraint, gfp_mask, order, mpol_mask);

    if (sysctl_oom_kill_allocating_task && current->mm &&
        !oom_unkillable_task(current, NULL, nodemask) &&
        current->signal->oom_score_adj != OOM_SCORE_ADJ_MIN) {
        get_task_struct(current);
        oom_kill_process(current, gfp_mask, order, 0, totalpages, NULL,
                 nodemask,
                 "Out of memory (oom_kill_allocating_task)");
        goto out;
    }

    p = select_bad_process(&points, totalpages, mpol_mask, force_kill);
    /* Found nothing?!?! Either we hang forever, or we panic. */
    if (!p) {
        dump_header(NULL, gfp_mask, order, NULL, mpol_mask);
        panic("Out of memory and no killable processes...\n");
    }
    if (p != (void *)-1UL) {
        oom_kill_process(p, gfp_mask, order, points, totalpages, NULL,
                 nodemask, "Out of memory");
        killed = 1;
    }
out:
    /*
     * Give the killed threads a good chance of exiting before trying to
     * allocate memory again.
     */
    if (killed)
        schedule_timeout_killable(1);
}

该函数首先调用blocking_notifier_call_chain()进行OOM的内核通知链回调处理；接着的if (fatal_signal_pending(current) || current->flags & PF_EXITING)判断则是用于检查是否有SIGKILL信号挂起或者正在信号处理中，如果有则退出；再接着通过constrained_alloc()检查内存分配限制以及check_panic_on_oom()检查是否报linux内核panic；继而判断sysctl_oom_kill_allocating_task变量及进程检查，如果符合条件判断，则将当前分配的内存kill掉；否则最后，将通过select_bad_process()选出最佳的进程，进而调用oom_kill_process()对其进行kill操作。

最后分析一下select_bad_process()和oom_kill_process()，其中select_bad_process()的实现：

【file:/ mm/oom_kill.c】
/*
 * Simple selection loop. We chose the process with the highest
 * number of 'points'. Returns -1 on scan abort.
 *
 * (not docbooked, we don't want this one cluttering up the manual)
 */
static struct task_struct *select_bad_process(unsigned int *ppoints,
        unsigned long totalpages, const nodemask_t *nodemask,
        bool force_kill)
{
    struct task_struct *g, *p;
    struct task_struct *chosen = NULL;
    unsigned long chosen_points = 0;

    rcu_read_lock();
    for_each_process_thread(g, p) {
        unsigned int points;

        switch (oom_scan_process_thread(p, totalpages, nodemask,
                        force_kill)) {
        case OOM_SCAN_SELECT:
            chosen = p;
            chosen_points = ULONG_MAX;
            /* fall through */
        case OOM_SCAN_CONTINUE:
            continue;
        case OOM_SCAN_ABORT:
            rcu_read_unlock();
            return (struct task_struct *)(-1UL);
        case OOM_SCAN_OK:
            break;
        };
        points = oom_badness(p, NULL, nodemask, totalpages);
        if (!points || points < chosen_points)
            continue;
        /* Prefer thread group leaders for display purposes */
        if (points == chosen_points && thread_group_leader(chosen))
            continue;

        chosen = p;
        chosen_points = points;
    }
    if (chosen)
        get_task_struct(chosen);
    rcu_read_unlock();

    *ppoints = chosen_points * 1000 / totalpages;
    return chosen;
}

    此函数通过for_each_process_thread()宏遍历所有进程，进而借用oom_scan_process_thread()获得进程扫描类型然后通过switch-case作特殊化处理，例如存在某进程退出中则中断扫描、某进程占用内存过多且被标识为优先kill掉则优选等特殊处理。而正常情况则会通过oom_badness()计算出进程的分值，然后根据最高分值将进程控制块返回回去。

    顺便研究一下oom_badness()的实现:

【file:/ mm/oom_kill.c】
/**
 * oom_badness - heuristic function to determine which candidate task to kill
 * @p: task struct of which task we should calculate
 * @totalpages: total present RAM allowed for page allocation
 *
 * The heuristic for determining which task to kill is made to be as simple and
 * predictable as possible. The goal is to return the highest value for the
 * task consuming the most memory to avoid subsequent oom failures.
 */
unsigned long oom_badness(struct task_struct *p, struct mem_cgroup *memcg,
              const nodemask_t *nodemask, unsigned long totalpages)
{
    long points;
    long adj;

    if (oom_unkillable_task(p, memcg, nodemask))
        return 0;

    p = find_lock_task_mm(p);
    if (!p)
        return 0;

    adj = (long)p->signal->oom_score_adj;
    if (adj == OOM_SCORE_ADJ_MIN) {
        task_unlock(p);
        return 0;
    }

    /*
     * The baseline for the badness score is the proportion of RAM that each
     * task's rss, pagetable and swap space use.
     */
    points = get_mm_rss(p->mm) + atomic_long_read(&p->mm->nr_ptes) +
         get_mm_counter(p->mm, MM_SWAPENTS);
    task_unlock(p);

    /*
     * Root processes get 3% bonus, just like the __vm_enough_memory()
     * implementation used by LSMs.
     */
    if (has_capability_noaudit(p, CAP_SYS_ADMIN))
        points -= (points * 3) / 100;

    /* Normalize to oom_score_adj units */
    adj *= totalpages / 1000;
    points += adj;

    /*
     * Never return 0 for an eligible task regardless of the root bonus and
     * oom_score_adj (oom_score_adj can't be OOM_SCORE_ADJ_MIN here).
     */
    return points > 0 ? points : 1;
}

    计算进程分值的函数中，首先排除了不可OOM kill的进程以及oom_score_adj值为OOM_SCORE_ADJ_MIN（即-1000）的进程，其中oom_score_adj取值范围是-1000到1000；接着就是计算进程的RSS、页表以及SWAP空间的使用量占RAM的比重，如果该进程是超级进程，则去除3%的权重；最后将oom_score_adj和points归一后，但凡小于0值的都返回1，其他的则返回原值。由此可知，分值越低的则越不会被kill，而且该值可以通过修改oom_score_adj进行调整。

    最后分析一下找到了最“bad”的进程后，其享受的“待遇”oom_kill_process()：

【file:/ mm/oom_kill.c】
/*
 * Must be called while holding a reference to p, which will be released upon
 * returning.
 */
void oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order,
              unsigned int points, unsigned long totalpages,
              struct mem_cgroup *memcg, nodemask_t *nodemask,
              const char *message)
{
    struct task_struct *victim = p;
    struct task_struct *child;
    struct task_struct *t;
    struct mm_struct *mm;
    unsigned int victim_points = 0;
    static DEFINE_RATELIMIT_STATE(oom_rs, DEFAULT_RATELIMIT_INTERVAL,
                          DEFAULT_RATELIMIT_BURST);

    /*
     * If the task is already exiting, don't alarm the sysadmin or kill
     * its children or threads, just set TIF_MEMDIE so it can die quickly
     */
    if (p->flags & PF_EXITING) {
        set_tsk_thread_flag(p, TIF_MEMDIE);
        put_task_struct(p);
        return;
    }

    if (__ratelimit(&oom_rs))
        dump_header(p, gfp_mask, order, memcg, nodemask);

    task_lock(p);
    pr_err("%s: Kill process %d (%s) score %d or sacrifice child\n",
        message, task_pid_nr(p), p->comm, points);
    task_unlock(p);

    /*
     * If any of p's children has a different mm and is eligible for kill,
     * the one with the highest oom_badness() score is sacrificed for its
     * parent. This attempts to lose the minimal amount of work done while
     * still freeing memory.
     */
    read_lock(&tasklist_lock);
    for_each_thread(p, t) {
        list_for_each_entry(child, &t->children, sibling) {
            unsigned int child_points;

            if (child->mm == p->mm)
                continue;
            /*
             * oom_badness() returns 0 if the thread is unkillable
             */
            child_points = oom_badness(child, memcg, nodemask,
                                totalpages);
            if (child_points > victim_points) {
                put_task_struct(victim);
                victim = child;
                victim_points = child_points;
                get_task_struct(victim);
            }
        }
    }
    read_unlock(&tasklist_lock);

    p = find_lock_task_mm(victim);
    if (!p) {
        put_task_struct(victim);
        return;
    } else if (victim != p) {
        get_task_struct(p);
        put_task_struct(victim);
        victim = p;
    }

    /* mm cannot safely be dereferenced after task_unlock(victim) */
    mm = victim->mm;
    pr_err("Killed process %d (%s) total-vm:%lukB, anon-rss:%lukB, file-rss:%lukB\n",
        task_pid_nr(victim), victim->comm, K(victim->mm->total_vm),
        K(get_mm_counter(victim->mm, MM_ANONPAGES)),
        K(get_mm_counter(victim->mm, MM_FILEPAGES)));
    task_unlock(victim);

    /*
     * Kill all user processes sharing victim->mm in other thread groups, if
     * any. They don't get access to memory reserves, though, to avoid
     * depletion of all memory. This prevents mm->mmap_sem livelock when an
     * oom killed thread cannot exit because it requires the semaphore and
     * its contended by another thread trying to allocate memory itself.
     * That thread will now get access to memory reserves since it has a
     * pending fatal signal.
     */
    rcu_read_lock();
    for_each_process(p)
        if (p->mm == mm && !same_thread_group(p, victim) &&
            !(p->flags & PF_KTHREAD)) {
            if (p->signal->oom_score_adj == OOM_SCORE_ADJ_MIN)
                continue;

            task_lock(p); /* Protect ->comm from prctl() */
            pr_err("Kill process %d (%s) sharing same memory\n",
                task_pid_nr(p), p->comm);
            task_unlock(p);
            do_send_sig_info(SIGKILL, SEND_SIG_FORCED, p, true);
        }
    rcu_read_unlock();

    set_tsk_thread_flag(victim, TIF_MEMDIE);
    do_send_sig_info(SIGKILL, SEND_SIG_FORCED, victim, true);
    put_task_struct(victim);
}

    该函数将会判断当前被kill的进程情况，如果该进程处于退出状态，则设置TIF_MEMDIE标志，不做kill操作；接着会通过list_for_each_entry()遍历该进程的子进程信息，如果某个子进程拥有不同的mm且合适被kill掉，将会优先考虑将该子进程替代父进程kill掉，这样可以避免kill掉父进程带来的接管子进程的工作开销；再往下通过find_lock_task_mm()找到持有mm锁的进程，如果进程处于退出状态，则return，否则继续处理，若此时的进程与传入的不是同一个时则更新victim；继而接着通过for_each_process()查找与当前被kill进程使用到了同样的共享内存的进程进行一起kill掉，kill之前将对应的进程添加标识TIF_MEMDIE，而kill的动作则是通过发送SICKILL信号给对应进程，由被kill进程从内核态返回用户态时进行处理。

    至此，OOM kill处理分析完毕。