《Linux驱动：块设备的读写流程（ ll_rw_block 接口分析）》

一，前言

在内核空间构建了块设备驱动程序，创建了块设备的设备节点，那么用户空间的APP如果通过该设备节点去读写对应的块设备呢？接下来便逐步分析这个过程。分析用户程序是如何一步步读写到块设备的数据的。

二， 应用层读写到块设备数据框架图

通用块层提供了一个接口ll_rw_block( )用来对逻辑块进行读写操作。应用层对块设备的读写操作都会调用到ll_rw_block( )函数来执行。
void ll_rw_block(int rw, int nr, struct buffer_head *bhs[])；
rw：进行读操作还是写操作，一般可选的值为 READ、WRITE 或者 READA 等；
nr：bhs数组元素个数；
bhs：要进行读写操作的数据块数组；

三，ll_rw_block接口分析

函数原型：
void ll_rw_block(int rw, int nr, struct buffer_head *bhs[])；
在分析ll_rw_block接口前，先看下buffer_head结构体

3.1 buffer_head结构体

struct buffer_head {
	unsigned long b_state;		/* buffer state bitmap (see above) */
	struct buffer_head *b_this_page;/* circular list of page's buffers */
	struct page *b_page;		/* the page this bh is mapped to */

	sector_t b_blocknr;		/* start block number */  // 数据块号
	size_t b_size;			/* size of mapping */     // 数据块大小
	char *b_data;			/* pointer to data within the page */

	struct block_device *b_bdev;                        // 数据块所属设备
	bh_end_io_t *b_end_io;		/* I/O completion */
 	void *b_private;		/* reserved for b_end_io */
	struct list_head b_assoc_buffers; /* associated with another mapping */
	struct address_space *b_assoc_map;	/* mapping this buffer is
						   associated with */
	atomic_t b_count;		/* users using this buffer_head */
};

3.2 ll_rw_block接口实现

void ll_rw_block(int rw, int nr, struct buffer_head *bhs[])
{
	int i;

	// 遍历数据块数组，取出每一个数据块
	for (i = 0; i < nr; i++) {
		struct buffer_head *bh = bhs[i];

		if (rw == SWRITE)
			lock_buffer(bh);
		else if (test_set_buffer_locked(bh))
			continue;

		if (rw == WRITE || rw == SWRITE) {
			if (test_clear_buffer_dirty(bh)) {
				bh->b_end_io = end_buffer_write_sync;
				get_bh(bh);
				submit_bh(WRITE, bh);  // 提交写数据块
				continue;
			}
		} else {
			if (!buffer_uptodate(bh)) {
				bh->b_end_io = end_buffer_read_sync;
				get_bh(bh);
				submit_bh(rw, bh);	// 提交读数据块
				continue;
			}
		}
		unlock_buffer(bh);
	}
}

3.3 构建bio、获取设备i/o请求队列

submit_bh(WRITE, bh)/submit_bh(rw, bh) ->
    bio = bio_alloc(GFP_NOIO, 1);
	bio->bi_sector = bh->b_blocknr * (bh->b_size >> 9);
	bio->bi_bdev = bh->b_bdev;
	bio->bi_io_vec[0].bv_page = bh->b_page;
	bio->bi_io_vec[0].bv_len = bh->b_size;
	bio->bi_io_vec[0].bv_offset = bh_offset(bh);

	bio->bi_vcnt = 1;
	bio->bi_idx = 0;
	bio->bi_size = bh->b_size;

	bio->bi_end_io = end_bio_bh_io_sync;
	bio->bi_private = bh;

	bio_get(bio);
	// 提交bio
	submit_bio(rw, bio) ->
    	generic_make_request(bio) -> 
            __generic_make_request(bio) ->
                q = bdev_get_queue(bio->bi_bdev) -> //获取块设备的i/o请求队列
                    ret = q->make_request_fn(q, bio) -> // // 调用队列的"构造请求函数"处理bio

3.4 队列的"构造请求函数"由来

队列的"构造请求函数"有两个来源，一是，使用默认的函数__make_request；二是，在块设备驱动程序中设置。

3.4.1 默认的__make_request接口

__make_request被设置为队列的"构造请求函数"的过程

blk_init_queue ->
    blk_init_queue_node ->
        blk_queue_make_request(q, __make_request);
        	q->make_request_fn = mfn;

__make_request接口实现

static int __make_request(request_queue_t *q, struct bio *bio)
{
	struct request *req;
	int el_ret, nr_sectors, barrier, err;
	const unsigned short prio = bio_prio(bio);
	const int sync = bio_sync(bio);
	int rw_flags;

	nr_sectors = bio_sectors(bio);  // 要读写的扇区个数

	/*
	 * low level driver can indicate that it wants pages above a
	 * certain limit bounced to low memory (ie for highmem, or even
	 * ISA dma in theory)
	 */
	blk_queue_bounce(q, &bio);

	barrier = bio_barrier(bio);
	if (unlikely(barrier) && (q->next_ordered == QUEUE_ORDERED_NONE)) {
		err = -EOPNOTSUPP;
		goto end_io;
	}

	spin_lock_irq(q->queue_lock);

	if (unlikely(barrier) || elv_queue_empty(q))
		goto get_rq;

	// 首先通过调用 elv_merge 方法尝试将当前I/O请求与其他正在排队的I/O请求进行合并，
	// 如果当前I/O请求与正在排队的I/O请求相邻，那么就可以合并为一个I/O请求，从而减少对设备I/O请求的次数。
	el_ret = elv_merge(q, &req, bio);
	switch (el_ret) {
		case ELEVATOR_BACK_MERGE:  // 与其他I/O请求合并成功
			BUG_ON(!rq_mergeable(req));

			if (!ll_back_merge_fn(q, req, bio))
				break;

			blk_add_trace_bio(q, bio, BLK_TA_BACKMERGE);

			req->biotail->bi_next = bio;
			req->biotail = bio;
			req->nr_sectors = req->hard_nr_sectors += nr_sectors;
			req->ioprio = ioprio_best(req->ioprio, prio);
			drive_stat_acct(req, nr_sectors, 0);
			if (!attempt_back_merge(q, req))
				elv_merged_request(q, req, el_ret);
			goto out;

		case ELEVATOR_FRONT_MERGE:  // 与其他I/O请求合并成功
			BUG_ON(!rq_mergeable(req));

			if (!ll_front_merge_fn(q, req, bio))
				break;

			blk_add_trace_bio(q, bio, BLK_TA_FRONTMERGE);

			bio->bi_next = req->bio;
			req->bio = bio;

			/*
			 * may not be valid. if the low level driver said
			 * it didn't need a bounce buffer then it better
			 * not touch req->buffer either...
			 */
			req->buffer = bio_data(bio);
			req->current_nr_sectors = bio_cur_sectors(bio);
			req->hard_cur_sectors = req->current_nr_sectors;
			req->sector = req->hard_sector = bio->bi_sector;
			req->nr_sectors = req->hard_nr_sectors += nr_sectors;
			req->ioprio = ioprio_best(req->ioprio, prio);
			drive_stat_acct(req, nr_sectors, 0);
			if (!attempt_front_merge(q, req))
				elv_merged_request(q, req, el_ret);
			goto out;

		/* ELV_NO_MERGE: elevator says don't/can't merge. */
		default:
			;
	}

// 无法和其他I/O请求合并，则构建一个i/o请求。
get_rq:
	/*
	 * This sync check and mask will be re-done in init_request_from_bio(),
	 * but we need to set it earlier to expose the sync flag to the
	 * rq allocator and io schedulers.
	 */
	rw_flags = bio_data_dir(bio);
	if (sync)
		rw_flags |= REQ_RW_SYNC;

	/*
	 * Grab a free request. This is might sleep but can not fail.
	 * Returns with the queue unlocked.
	 */
	req = get_request_wait(q, rw_flags, bio);

	/*
	 * After dropping the lock and possibly sleeping here, our request
	 * may now be mergeable after it had proven unmergeable (above).
	 * We don't worry about that case for efficiency. It won't happen
	 * often, and the elevators are able to handle it.
	 */

	// 根据bio构建一个i/o请求
	init_request_from_bio(req, bio);

	spin_lock_irq(q->queue_lock);
	if (elv_queue_empty(q))
		blk_plug_device(q);

	// 向队列提交i/o请求
	add_request(q, req);
out:
    // 同步处理队列的i/o请求
	if (sync)
		__generic_unplug_device(q);

	spin_unlock_irq(q->queue_lock);
	return 0;

end_io:
	bio_endio(bio, nr_sectors << 9, err);
	return 0;
}

__generic_unplug_device(q) ->
    q->request_fn(q); // 调用队列的处理函数，由块设备驱动程序提供 ramblock_queue = blk_init_queue(do_ramblock_request, &ramblock_lock);

四，总结

用户程序通过块设备驱动节点，读写块设备数据的流程

• 块设备驱动程序初始化时调用blk_init_queue接口初始一个i/o请求队列，并为该队列提供一个队列"请求处理函数"；
• 用户程序调用read()/write()，会转换成通用块层ll_rw_block接口的调用；
• ll_rw_block接口通过调用submit_bh接口提交读写数据块；
• submit_bh接口构建并设置bio，然后通过submit_bio接口提交bio；
• 调用块设备队列的"构造请求函数"处理bio，即__make_request接口处理bio；
• __make_request接口首先尝试和其他i/o请求合并bio以减少对设备I/O请求的次数；
• 如果合并不成功，则构建一个新的i/o请求；
• 所有bio处理（和其他i/o请求合并或构建一个新的i/o请求）完成或者用户强制sync时，会处理所有的i/o请求，调用__generic_unplug_device接口，调用队列的"请求处理函数"处理所有请求（读写数据）。