Linux 3.8 Writeback机制源码分析

writeback相关数据结构

与writeback相关的数据结构主要有：

1. backing_dev_info，该数据结构描述了backing_dev的所有信息，通常块设备的request queue中会包含backing_dev对象。
2. bdi_writeback，该数据结构封装了writeback的内核线程以及需要操作的inode队列。
3. wb_writeback_work，该数据结构封装了writeback的工作任务。

它们的结构体分别如下：

struct backing_dev_info {
	struct list_head bdi_list;
	unsigned long ra_pages;	/* max readahead in PAGE_CACHE_SIZE units */
	unsigned long state;	/* Always use atomic bitops on this */
	unsigned int capabilities; /* Device capabilities */
	congested_fn *congested_fn; /* Function pointer if device is md/dm */
	void *congested_data;	/* Pointer to aux data for congested func */

	char *name;

	struct percpu_counter bdi_stat[NR_BDI_STAT_ITEMS];

	unsigned long bw_time_stamp;	/* last time write bw is updated */
	unsigned long dirtied_stamp;
	unsigned long written_stamp;	/* pages written at bw_time_stamp */
	unsigned long write_bandwidth;	/* the estimated write bandwidth */
	unsigned long avg_write_bandwidth; /* further smoothed write bw */

	/*
	 * The base dirty throttle rate, re-calculated on every 200ms.
	 * All the bdi tasks' dirty rate will be curbed under it.
	 * @dirty_ratelimit tracks the estimated @balanced_dirty_ratelimit
	 * in small steps and is much more smooth/stable than the latter.
	 */
	unsigned long dirty_ratelimit;
	unsigned long balanced_dirty_ratelimit;

	struct fprop_local_percpu completions;
	int dirty_exceeded;

	unsigned int min_ratio;
	unsigned int max_ratio, max_prop_frac;

	struct bdi_writeback wb;  /* default writeback info for this bdi */
	spinlock_t wb_lock;	  /* protects work_list */

	struct list_head work_list;

	struct device *dev;

	struct timer_list laptop_mode_wb_timer;

#ifdef CONFIG_DEBUG_FS
	struct dentry *debug_dir;
	struct dentry *debug_stats;
#endif
};

struct bdi_writeback {
	struct backing_dev_info *bdi;	/* our parent bdi */
	unsigned int nr;


	unsigned long last_old_flush;	/* last old data flush */
	unsigned long last_active;	/* last time bdi thread was active */


	struct task_struct *task;	/* writeback thread */
	struct timer_list wakeup_timer; /* used for delayed bdi thread wakeup */
	struct list_head b_dirty;	/* dirty inodes */
	struct list_head b_io;		/* parked for writeback */
	struct list_head b_more_io;	/* parked for more writeback */
	spinlock_t list_lock;		/* protects the b_* lists */
};

/*
 * Passed into wb_writeback(), essentially a subset of writeback_control
 */
struct wb_writeback_work {
	long nr_pages;
	struct super_block *sb;
	unsigned long *older_than_this;
	enum writeback_sync_modes sync_mode;
	unsigned int tagged_writepages:1;
	unsigned int for_kupdate:1;
	unsigned int range_cyclic:1;
	unsigned int for_background:1;
	enum wb_reason reason;		/* why was writeback initiated? */

	struct list_head list;		/* pending work list */
	struct completion *done;	/* set if the caller waits */
};

1. BDI数据结构是对块设备的一个描述。bdi对象在块设备添加的时候需要注册到系统的bdi队列中。对于ext3而言，在mount的时候需要将底层块设备的bdi对象联系到ext3 root_inode中。在bdi数据结构中有一条work_list，该队列维护了writeback内核线程需要处理的任务。如果该队列上没有work可以处理，那么writeback内核线程将会睡眠等待。
2. writeback对象封装了内核线程task以及需要处理的inode队列。当page cache/buffer cache需要刷新radix tree上的inode时，可以将该inode挂载到writeback对象的b_dirty队列上，然后唤醒writeback线程。在处理过程中，inode会被移到b_io队列上进行处理。多条链表的方式可以降低多线程之间的资源共享。
3. wb_writeback_work数据结构是对writeback任务的封装，不同的任务可以采用不同的刷新策略。writeback线程的处理对象就是writeback_work。如果writeback_work队列为空，那么内核线程就可以睡眠了。

writeback主要函数分析

writeback机制的主要函数包括如下两个方面：

1. 管理bdi对象并且fork相应的writeback内核线程处理cache数据的刷新工作。
2. writeback内核线程处理函数，实现dirty page的刷新操作

writeback线程管理

Linux中有一个内核守护线程，该线程用来管理系统bdi队列，并且负责为block device创建writeback thread。当bdi中有dirty page并且还没有为bdi分配内核线程的时候，bdi_forker_thread程序会为其分配线程资源；当一个writeback线程长时间（默认为5min）处于空闲状态时，bdi_forker_thread程序会释放该线程资源。

static int bdi_forker_thread(void *ptr)
{
	struct bdi_writeback *me = ptr;

	current->flags |= PF_SWAPWRITE;
	set_freezable();

	/*
	 * Our parent may run at a different priority, just set us to normal
	 */
	set_user_nice(current, 0);

	for (;;) {
		struct task_struct *task = NULL;
		struct backing_dev_info *bdi;
		enum {
			NO_ACTION,   /* Nothing to do */
			FORK_THREAD, /* Fork bdi thread */
			KILL_THREAD, /* Kill inactive bdi thread */
		} action = NO_ACTION;

		/*
		 * Temporary measure, we want to make sure we don't see
		 * dirty data on the default backing_dev_info
		 */
		if (wb_has_dirty_io(me) || !list_empty(&me->bdi->work_list)) {
			del_timer(&me->wakeup_timer);
			wb_do_writeback(me, 0);
		}

		spin_lock_bh(&bdi_lock);
		/*
		 * In the following loop we are going to check whether we have
		 * some work to do without any synchronization with tasks
		 * waking us up to do work for them. Set the task state here
		 * so that we don't miss wakeups after verifying conditions.
		 */
		set_current_state(TASK_INTERRUPTIBLE);

		list_for_each_entry(bdi, &bdi_list, bdi_list) {
			bool have_dirty_io;

			if (!bdi_cap_writeback_dirty(bdi) ||
			     bdi_cap_flush_forker(bdi))
				continue;

			WARN(!test_bit(BDI_registered, &bdi->state),
			     "bdi %p/%s is not registered!\n", bdi, bdi->name);

			have_dirty_io = !list_empty(&bdi->work_list) ||
					wb_has_dirty_io(&bdi->wb);

			/*
			 * If the bdi has work to do, but the thread does not
			 * exist - create it.
			 */
			if (!bdi->wb.task && have_dirty_io) {
				/*
				 * Set the pending bit - if someone will try to
				 * unregister this bdi - it'll wait on this bit.
				 */
				set_bit(BDI_pending, &bdi->state);
				action = FORK_THREAD;
				break;
			}

			spin_lock(&bdi->wb_lock);

			/*
			 * If there is no work to do and the bdi thread was
			 * inactive long enough - kill it. The wb_lock is taken
			 * to make sure no-one adds more work to this bdi and
			 * wakes the bdi thread up.
			 */
			if (bdi->wb.task && !have_dirty_io &&
			    time_after(jiffies, bdi->wb.last_active +
						bdi_longest_inactive())) {
				task = bdi->wb.task;
				bdi->wb.task = NULL;
				spin_unlock(&bdi->wb_lock);
				set_bit(BDI_pending, &bdi->state);
				action = KILL_THREAD;
				break;
			}
			spin_unlock(&bdi->wb_lock);
		}
		spin_unlock_bh(&bdi_lock);

		/* Keep working if default bdi still has things to do */
		if (!list_empty(&me->bdi->work_list))
			__set_current_state(TASK_RUNNING);

		switch (action) {
		case FORK_THREAD:
			__set_current_state(TASK_RUNNING);
			task = kthread_create(bdi_writeback_thread, &bdi->wb,
					      "flush-%s", dev_name(bdi->dev));
			if (IS_ERR(task)) {
				/*
				 * If thread creation fails, force writeout of
				 * the bdi from the thread. Hopefully 1024 is
				 * large enough for efficient IO.
				 */
				writeback_inodes_wb(&bdi->wb, 1024,
						    WB_REASON_FORKER_THREAD);
			} else {
				/*
				 * The spinlock makes sure we do not lose
				 * wake-ups when racing with 'bdi_queue_work()'.
				 * And as soon as the bdi thread is visible, we
				 * can start it.
				 */
				spin_lock_bh(&bdi->wb_lock);
				bdi->wb.task = task;
				spin_unlock_bh(&bdi->wb_lock);
				wake_up_process(task);
			}
			bdi_clear_pending(bdi);
			break;

		case KILL_THREAD:
			__set_current_state(TASK_RUNNING);
			kthread_stop(task);
			bdi_clear_pending(bdi);
			break;

		case NO_ACTION:
			if (!wb_has_dirty_io(me) || !dirty_writeback_interval)
				/*
				 * There are no dirty data. The only thing we
				 * should now care about is checking for
				 * inactive bdi threads and killing them. Thus,
				 * let's sleep for longer time, save energy and
				 * be friendly for battery-driven devices.
				 */
				schedule_timeout(bdi_longest_inactive());
			else
				schedule_timeout(msecs_to_jiffies(dirty_writeback_interval * 10));
			try_to_freeze();
			break;
		}
	}

	return 0;
}

Writeback工作线程

writeback线程是bdi_forker_thread 创建的，该线程的任务就是处理等待的数据回刷任务。线程处理函数为bdi_writeback_thread，该函数的实现如下：

/*
 * Handle writeback of dirty data for the device backed by this bdi. Also
 * wakes up periodically and does kupdated style flushing.
 */
int bdi_writeback_thread(void *data)
{
	struct bdi_writeback *wb = data;
	struct backing_dev_info *bdi = wb->bdi;
	long pages_written;

	current->flags |= PF_SWAPWRITE;
	set_freezable();
	wb->last_active = jiffies;

	/*
	 * Our parent may run at a different priority, just set us to normal
	 */
	set_user_nice(current, 0);

	trace_writeback_thread_start(bdi);

	while (!kthread_freezable_should_stop(NULL)) {
		/*
		 * Remove own delayed wake-up timer, since we are already awake
		 * and we'll take care of the periodic write-back.
		 */
		del_timer(&wb->wakeup_timer);

		pages_written = wb_do_writeback(wb, 0);

		trace_writeback_pages_written(pages_written);

		if (pages_written)
			wb->last_active = jiffies;

		set_current_state(TASK_INTERRUPTIBLE);
		if (!list_empty(&bdi->work_list) || kthread_should_stop()) {
			__set_current_state(TASK_RUNNING);
			continue;
		}

		if (wb_has_dirty_io(wb) && dirty_writeback_interval)
			schedule_timeout(msecs_to_jiffies(dirty_writeback_interval * 10));
		else {
			/*
			 * We have nothing to do, so can go sleep without any
			 * timeout and save power. When a work is queued or
			 * something is made dirty - we will be woken up.
			 */
			schedule();
		}
	}

	/* Flush any work that raced with us exiting */
	if (!list_empty(&bdi->work_list))
		wb_do_writeback(wb, 1);

	trace_writeback_thread_stop(bdi);
	return 0;
}

bdi_writeback_thread函数主要是调用wb_do_writeback()函数。

/*
 * Retrieve work items and do the writeback they describe
 */
long wb_do_writeback(struct bdi_writeback *wb, int force_wait)
{
	struct backing_dev_info *bdi = wb->bdi;
	struct wb_writeback_work *work;
	long wrote = 0;

	set_bit(BDI_writeback_running, &wb->bdi->state);
	while ((work = get_next_work_item(bdi)) != NULL) {
		/*
		 * Override sync mode, in case we must wait for completion
		 * because this thread is exiting now.
		 */
		if (force_wait)
			work->sync_mode = WB_SYNC_ALL;

		trace_writeback_exec(bdi, work);

		wrote += wb_writeback(wb, work);

		/*
		 * Notify the caller of completion if this is a synchronous
		 * work item, otherwise just free it.
		 */
		if (work->done)
			complete(work->done);
		else
			kfree(work);
	}

	/*
	 * Check for periodic writeback, kupdated() style
	 */
	wrote += wb_check_old_data_flush(wb);
	wrote += wb_check_background_flush(wb);
	clear_bit(BDI_writeback_running, &wb->bdi->state);

	return wrote;
}

wb_check_old_data_flush函数的主要功能是周期性的检查脏页并写回，它默认写回30s之前写入的脏页，每隔5s扫描一次。

static long wb_check_old_data_flush(struct bdi_writeback *wb)
{
	unsigned long expired;
	long nr_pages;

	/*
	 * When set to zero, disable periodic writeback
	 */
	if (!dirty_writeback_interval)
		return 0;

	expired = wb->last_old_flush +
			msecs_to_jiffies(dirty_writeback_interval * 10);
	if (time_before(jiffies, expired))
		return 0;

	wb->last_old_flush = jiffies;
	nr_pages = get_nr_dirty_pages();

	if (nr_pages) {
		struct wb_writeback_work work = {
			.nr_pages	= nr_pages,
			.sync_mode	= WB_SYNC_NONE,
			.for_kupdate	= 1,
			.range_cyclic	= 1,
			.reason		= WB_REASON_PERIODIC,
		};

		return wb_writeback(wb, &work);
	}

	return 0;
}

wb_check_background_flush的功能是在脏页达到一定比例时写回所有的脏页，直到脏页的比例达到阀值以下。

static long wb_check_background_flush(struct bdi_writeback *wb)
{
	if (over_bground_thresh(wb->bdi)) {

		struct wb_writeback_work work = {
			.nr_pages	= LONG_MAX,
			.sync_mode	= WB_SYNC_NONE,
			.for_background	= 1,
			.range_cyclic	= 1,
			.reason		= WB_REASON_BACKGROUND,
		};

		return wb_writeback(wb, &work);
	}

	return 0;
}

wb_check_background_flush和wb_check_old_data_flush的函数只是设置wb_writeback_work的各项参数，然后执行wb_writeback函数，该函数是Writeback机制中真正执行写回的函数。Writeback机制中的写回磁盘操作都是通过wb_writeback函数实现的，wb_writeback调用与文件系统有关的write函数，执行协会磁盘的操作。

/*
 * Explicit flushing or periodic writeback of "old" data.
 *
 * Define "old": the first time one of an inode's pages is dirtied, we mark the
 * dirtying-time in the inode's address_space.  So this periodic writeback code
 * just walks the superblock inode list, writing back any inodes which are
 * older than a specific point in time.
 *
 * Try to run once per dirty_writeback_interval.  But if a writeback event
 * takes longer than a dirty_writeback_interval interval, then leave a
 * one-second gap.
 *
 * older_than_this takes precedence over nr_to_write.  So we'll only write back
 * all dirty pages if they are all attached to "old" mappings.
 */
static long wb_writeback(struct bdi_writeback *wb,
			 struct wb_writeback_work *work)
{
	unsigned long wb_start = jiffies;
	long nr_pages = work->nr_pages;
	unsigned long oldest_jif;
	struct inode *inode;
	long progress;

	oldest_jif = jiffies;
	work->older_than_this = &oldest_jif;

	spin_lock(&wb->list_lock);
	for (;;) {
		/*
		 * Stop writeback when nr_pages has been consumed
		 */
		if (work->nr_pages <= 0)
			break;

		/*
		 * Background writeout and kupdate-style writeback may
		 * run forever. Stop them if there is other work to do
		 * so that e.g. sync can proceed. They'll be restarted
		 * after the other works are all done.
		 */
		if ((work->for_background || work->for_kupdate) &&
		    !list_empty(&wb->bdi->work_list))
			break;

		/*
		 * For background writeout, stop when we are below the
		 * background dirty threshold
		 */
		if (work->for_background && !over_bground_thresh(wb->bdi))
			break;

		/*
		 * Kupdate and background works are special and we want to
		 * include all inodes that need writing. Livelock avoidance is
		 * handled by these works yielding to any other work so we are
		 * safe.
		 */
		if (work->for_kupdate) {
			oldest_jif = jiffies -
				msecs_to_jiffies(dirty_expire_interval * 10);
		} else if (work->for_background)
			oldest_jif = jiffies;

		trace_writeback_start(wb->bdi, work);
		if (list_empty(&wb->b_io))
			queue_io(wb, work);
		if (work->sb)
			progress = writeback_sb_inodes(work->sb, wb, work);
		else
			progress = __writeback_inodes_wb(wb, work);
		trace_writeback_written(wb->bdi, work);

		wb_update_bandwidth(wb, wb_start);

		/*
		 * Did we write something? Try for more
		 *
		 * Dirty inodes are moved to b_io for writeback in batches.
		 * The completion of the current batch does not necessarily
		 * mean the overall work is done. So we keep looping as long
		 * as made some progress on cleaning pages or inodes.
		 */
		if (progress)
			continue;
		/*
		 * No more inodes for IO, bail
		 */
		if (list_empty(&wb->b_more_io))
			break;
		/*
		 * Nothing written. Wait for some inode to
		 * become available for writeback. Otherwise
		 * we'll just busyloop.
		 */
		if (!list_empty(&wb->b_more_io))  {
			trace_writeback_wait(wb->bdi, work);
			inode = wb_inode(wb->b_more_io.prev);
			spin_lock(&inode->i_lock);
			spin_unlock(&wb->list_lock);
			/* This function drops i_lock... */
			inode_sleep_on_writeback(inode);
			spin_lock(&wb->list_lock);
		}
	}
	spin_unlock(&wb->list_lock);

	return nr_pages - work->nr_pages;
}

总结

writeback机制是比较简单的，其核心是通过一个常驻内核线程为每个BDI对象分配writeback线程，实现对cache中dirty page的数据回刷。

参考：http://www.linuxidc.com/Linux/2013-01/77576.htm