ll_rw_blk.c

Linux内核源代码 为压缩文件 是<<Linux内核>>一书中的源代码

/*
 *  linux/drivers/block/ll_rw_blk.c
 *
 * Copyright (C) 1991, 1992 Linus Torvalds
 * Copyright (C) 1994,      Karl Keyte: Added support for disk statistics
 * Elevator latency, (C) 2000  Andrea Arcangeli <[e-mail]> SuSE
 * Queue request tables / lock, selectable elevator, Jens Axboe <[e-mail]>
 * kernel-doc documentation started by NeilBrown <[e-mail]> -  July2000
 */

/*
 * This handles all read/write requests to block devices
 */
#include <linux/sched.h>
#include <linux/kernel.h>
#include <linux/kernel_stat.h>
#include <linux/errno.h>
#include <linux/string.h>
#include <linux/config.h>
#include <linux/locks.h>
#include <linux/mm.h>
#include <linux/init.h>
#include <linux/smp_lock.h>

#include <asm/system.h>
#include <asm/io.h>
#include <linux/blk.h>
#include <linux/highmem.h>
#include <linux/raid/md.h>

#include <linux/module.h>

/*
 * MAC Floppy IWM hooks
 */

#ifdef CONFIG_MAC_FLOPPY_IWM
extern int mac_floppy_init(void);
#endif

extern int lvm_init(void);

/*
 * For the allocated request tables
 */
static kmem_cache_t *request_cachep;

/*
 * The "disk" task queue is used to start the actual requests
 * after a plug
 */
DECLARE_TASK_QUEUE(tq_disk);

/*
 * Protect the request list against multiple users..
 *
 * With this spinlock the Linux block IO subsystem is 100% SMP threaded
 * from the IRQ event side, and almost 100% SMP threaded from the syscall
 * side (we still have protect against block device array operations, and
 * the do_request() side is casually still unsafe. The kernel lock protects
 * this part currently.).
 *
 * there is a fair chance that things will work just OK if these functions
 * are called with no global kernel lock held ...
 */
spinlock_t io_request_lock = SPIN_LOCK_UNLOCKED;

/* This specifies how many sectors to read ahead on the disk. */

int read_ahead[MAX_BLKDEV];

/* blk_dev_struct is:
 *	*request_fn
 *	*current_request
 */
struct blk_dev_struct blk_dev[MAX_BLKDEV]; /* initialized by blk_dev_init() */

/*
 * blk_size contains the size of all block-devices in units of 1024 byte
 * sectors:
 *
 * blk_size[MAJOR][MINOR]
 *
 * if (!blk_size[MAJOR]) then no minor size checking is done.
 */
int * blk_size[MAX_BLKDEV];

/*
 * blksize_size contains the size of all block-devices:
 *
 * blksize_size[MAJOR][MINOR]
 *
 * if (!blksize_size[MAJOR]) then 1024 bytes is assumed.
 */
int * blksize_size[MAX_BLKDEV];

/*
 * hardsect_size contains the size of the hardware sector of a device.
 *
 * hardsect_size[MAJOR][MINOR]
 *
 * if (!hardsect_size[MAJOR])
 *		then 512 bytes is assumed.
 * else
 *		sector_size is hardsect_size[MAJOR][MINOR]
 * This is currently set by some scsi devices and read by the msdos fs driver.
 * Other uses may appear later.
 */
int * hardsect_size[MAX_BLKDEV];

/*
 * The following tunes the read-ahead algorithm in mm/filemap.c
 */
int * max_readahead[MAX_BLKDEV];

/*
 * Max number of sectors per request
 */
int * max_sectors[MAX_BLKDEV];

static inline int get_max_sectors(kdev_t dev)
{
	if (!max_sectors[MAJOR(dev)])
		return MAX_SECTORS;
	return max_sectors[MAJOR(dev)][MINOR(dev)];
}

static inline request_queue_t *__blk_get_queue(kdev_t dev)
{
	struct blk_dev_struct *bdev = blk_dev + MAJOR(dev);

	if (bdev->queue)
		return bdev->queue(dev);
	else
		return &blk_dev[MAJOR(dev)].request_queue;
}

/*
 * NOTE: the device-specific queue() functions
 * have to be atomic!
 */
request_queue_t *blk_get_queue(kdev_t dev)
{
	request_queue_t *ret;
	unsigned long flags;

	spin_lock_irqsave(&io_request_lock,flags);
	ret = __blk_get_queue(dev);
	spin_unlock_irqrestore(&io_request_lock,flags);

	return ret;
}

static int __blk_cleanup_queue(struct list_head *head)
{
	struct list_head *entry;
	struct request *rq;
	int i = 0;

	if (list_empty(head))
		return 0;

	entry = head->next;
	do {
		rq = list_entry(entry, struct request, table);
		entry = entry->next;
		list_del(&rq->table);
		kmem_cache_free(request_cachep, rq);
		i++;
	} while (!list_empty(head));

	return i;
}

/**
 * blk_cleanup_queue: - release a &request_queue_t when it is no longer needed
 * @q:    the request queue to be released
 *
 * Description:
 *     blk_cleanup_queue is the pair to blk_init_queue().  It should
 *     be called when a request queue is being released; typically
 *     when a block device is being de-registered.  Currently, its
 *     primary task it to free all the &struct request structures that
 *     were allocated to the queue.
 * Caveat: 
 *     Hopefully the low level driver will have finished any
 *     outstanding requests first...
 **/
void blk_cleanup_queue(request_queue_t * q)
{
	int count = QUEUE_NR_REQUESTS;

	count -= __blk_cleanup_queue(&q->request_freelist[READ]);
	count -= __blk_cleanup_queue(&q->request_freelist[WRITE]);

	if (count)
		printk("blk_cleanup_queue: leaked requests (%d)\n", count);

	memset(q, 0, sizeof(*q));
}

/**
 * blk_queue_headactive - indicate whether head of request queue may be active
 * @q:       The queue which this applies to.
 * @active:  A flag indication where the head of the queue is active.
 *
 * Description:
 *    The driver for a block device may choose to leave the currently active
 *    request on the request queue, removing it only when it has completed.
 *    The queue handling routines assume this by default for safety reasons
 *    and will not involve the head of the request queue in any merging or
 *    reordering of requests when the queue is unplugged (and thus may be
 *    working on this particular request).
 *
 *    If a driver removes requests from the queue before processing them, then
 *    it may indicate that it does so, there by allowing the head of the queue
 *    to be involved in merging and reordering.  This is done be calling
 *    blk_queue_headactive() with an @active flag of %0.
 *
 *    If a driver processes several requests at once, it must remove them (or
 *    at least all but one of them) from the request queue.
 *
 *    When a queue is plugged (see blk_queue_pluggable()) the head will be
 *    assumed to be inactive.
 **/
 
void blk_queue_headactive(request_queue_t * q, int active)
{
	q->head_active = active;
}

/**
 * blk_queue_pluggable - define a plugging function for a request queue
 * @q:   the request queue to which the function will apply
 * @plug: the function to be called to plug a queue
 *
 * Description:
 *   A request queue will be "plugged" if a request is added to it
 *   while it is empty.  This allows a number of requests to be added
 *   before any are processed, thus providing an opportunity for these
 *   requests to be merged or re-ordered.
 *   The default plugging function (generic_plug_device()) sets the
 *   "plugged" flag for the queue and adds a task to the $tq_disk task
 *   queue to unplug the queue and call the request function at a
 *   later time.
 *
 *   A device driver may provide an alternate plugging function by
 *   passing it to blk_queue_pluggable().  This function should set
 *   the "plugged" flag if it want calls to the request_function to be
 *   blocked, and should place a task on $tq_disk which will unplug
 *   the queue.  Alternately it can simply do nothing and there-by
 *   disable plugging of the device.
 **/

void blk_queue_pluggable (request_queue_t * q, plug_device_fn *plug)
{
	q->plug_device_fn = plug;
}


/**
 * blk_queue_make_request - define an alternate make_request function for a device
 * @q:  the request queue for the device to be affected
 * @mfn: the alternate make_request function
 *
 * Description:
 *    The normal way for &struct buffer_heads to be passed to a device
 *    driver is for them to be collected into requests on a request
 *    queue, and then to allow the device driver to select requests
 *    off that queue when it is ready.  This works well for many block
 *    devices. However some block devices (typically virtual devices
 *    such as md or lvm) do not benefit from the processing on the
 *    request queue, and are served best by having the requests passed
 *    directly to them.  This can be achieved by providing a function
 *    to blk_queue_make_request().
 *
 * Caveat:
 *    The driver that does this *must* be able to deal appropriately
 *    with buffers in "highmemory", either by calling bh_kmap() to get
 *    a kernel mapping, to by calling create_bounce() to create a
 *    buffer in normal memory.
 **/

void blk_queue_make_request(request_queue_t * q, make_request_fn * mfn)
{
	q->make_request_fn = mfn;
}

static inline int ll_new_segment(request_queue_t *q, struct request *req, int max_segments)
{
	if (req->nr_segments < max_segments) {
		req->nr_segments++;
		q->elevator.nr_segments++;
		return 1;
	}
	return 0;
}

static int ll_back_merge_fn(request_queue_t *q, struct request *req, 
			    struct buffer_head *bh, int max_segments)
{
	if (req->bhtail->b_data + req->bhtail->b_size == bh->b_data)
		return 1;
	return ll_new_segment(q, req, max_segments);
}

static int ll_front_merge_fn(request_queue_t *q, struct request *req, 
			     struct buffer_head *bh, int max_segments)
{
	if (bh->b_data + bh->b_size == req->bh->b_data)
		return 1;
	return ll_new_segment(q, req, max_segments);
}

static int ll_merge_requests_fn(request_queue_t *q, struct request *req,
				struct request *next, int max_segments)
{
	int total_segments = req->nr_segments + next->nr_segments;
	int same_segment;

	same_segment = 0;
	if (req->bhtail->b_data + req->bhtail->b_size == next->bh->b_data) {
		total_segments--;
		same_segment = 1;
	}
    
	if (total_segments > max_segments)
		return 0;

	q->elevator.nr_segments -= same_segment;
	req->nr_segments = total_segments;
	return 1;
}

/*
 * "plug" the device if there are no outstanding requests: this will
 * force the transfer to start only after we have put all the requests
 * on the list.
 *
 * This is called with interrupts off and no requests on the queue.
 * (and with the request spinlock acquired)
 */
static void generic_plug_device(request_queue_t *q, kdev_t dev)
{
	/*
	 * no need to replug device
	 */
	if (!list_empty(&q->queue_head) || q->plugged)
		return;

	q->plugged = 1;
	queue_task(&q->plug_tq, &tq_disk);
}

/*
 * remove the plug and let it rip..
 */
static inline void __generic_unplug_device(request_queue_t *q)
{
	if (q->plugged) {
		q->plugged = 0;
		if (!list_empty(&q->queue_head))
			q->request_fn(q);
	}
}

static void generic_unplug_device(void *data)
{
	request_queue_t *q = (request_queue_t *) data;
	unsigned long flags;

	spin_lock_irqsave(&io_request_lock, flags);
	__generic_unplug_device(q);
	spin_unlock_irqrestore(&io_request_lock, flags);
}

static void blk_init_free_list(request_queue_t *q)
{
	struct request *rq;
	int i;

	/*
	 * Divide requests in half between read and write. This used to
	 * be a 2/3 advantage for reads, but now reads can steal from
	 * the write free list.
	 */
	for (i = 0; i < QUEUE_NR_REQUESTS; i++) {
		rq = kmem_cache_alloc(request_cachep, SLAB_KERNEL);
		rq->rq_status = RQ_INACTIVE;
		list_add(&rq->table, &q->request_freelist[i & 1]);
	}

	init_waitqueue_head(&q->wait_for_request);
	spin_lock_init(&q->request_lock);
}

static int __make_request(request_queue_t * q, int rw, struct buffer_head * bh);

/**
 * blk_init_queue  - prepare a request queue for use with a block device
 * @q:    The &request_queue_t to be initialised
 * @rfn:  The function to be called to process requests that have been
 *        placed on the queue.
 *
 * Description:
 *    If a block device wishes to use the standard request handling procedures,
 *    which sorts requests and coalesces adjacent requests, then it must
 *    call blk_init_queue().  The function @rfn will be called when there
 *    are requests on the queue that need to be processed.  If the device
 *    supports plugging, then @rfn may not be called immediately when requests
 *    are available on the queue, but may be called at some time later instead.
 *    Plugged queues are generally unplugged when a buffer belonging to one
 *    of the requests on the queue is needed, or due to memory pressure.
 *
 *    @rfn is not required, or even expected, to remove all requests off the
 *    queue, but only as many as it can handle at a time.  If it does leave
 *    requests on the queue, it is responsible for arranging that the requests
 *    get dealt with eventually.
 *
 *    A global spin lock $io_request_lock must be held while manipulating the
 *    requests on the request queue.
 *
 *    The request on the head of the queue is by default assumed to be
 *    potentially active, and it is not considered for re-ordering or merging
 *    whenever the given queue is unplugged. This behaviour can be changed with
 *    blk_queue_headactive().
 *
 * Note:
 *    blk_init_queue() must be paired with a blk_cleanup-queue() call
 *    when the block device is deactivated (such as at module unload).