blk_queue_bounce_limit(q, BLK_BOUNCE_HIGH);
blk_queue_activity_fn(q, NULL, NULL);
-
- INIT_LIST_HEAD(&q->drain_list);
}
EXPORT_SYMBOL(blk_queue_make_request);
rq->special = NULL;
rq->data_len = 0;
rq->data = NULL;
+ rq->nr_phys_segments = 0;
rq->sense = NULL;
rq->end_io = NULL;
rq->end_io_data = NULL;
struct request *rq = flush_rq->end_io_data;
request_queue_t *q = rq->q;
+ elv_completed_request(q, flush_rq);
+
rq->flags |= REQ_BAR_PREFLUSH;
if (!flush_rq->errors)
struct request *rq = flush_rq->end_io_data;
request_queue_t *q = rq->q;
+ elv_completed_request(q, flush_rq);
+
rq->flags |= REQ_BAR_POSTFLUSH;
q->end_flush_fn(q, flush_rq);
if (!list_empty(&rq->queuelist))
blkdev_dequeue_request(rq);
- elv_deactivate_request(q, rq);
-
flush_rq->end_io_data = rq;
flush_rq->end_io = blk_pre_flush_end_io;
/**
* blk_queue_find_tag - find a request by its tag and queue
- *
* @q: The request queue for the device
* @tag: The tag of the request
*
static char *rq_flags[] = {
"REQ_RW",
"REQ_FAILFAST",
+ "REQ_SORTED",
"REQ_SOFTBARRIER",
"REQ_HARDBARRIER",
"REQ_CMD",
"REQ_STARTED",
"REQ_DONTPREP",
"REQ_QUEUED",
+ "REQ_ELVPRIV",
"REQ_PC",
"REQ_BLOCK_PC",
"REQ_SENSE",
rl->count[READ] = rl->count[WRITE] = 0;
rl->starved[READ] = rl->starved[WRITE] = 0;
+ rl->elvpriv = 0;
init_waitqueue_head(&rl->wait[READ]);
init_waitqueue_head(&rl->wait[WRITE]);
- init_waitqueue_head(&rl->drain);
rl->rq_pool = mempool_create_node(BLKDEV_MIN_RQ, mempool_alloc_slab,
mempool_free_slab, request_cachep, q->node);
static int __make_request(request_queue_t *, struct bio *);
-request_queue_t *blk_alloc_queue(int gfp_mask)
+request_queue_t *blk_alloc_queue(gfp_t gfp_mask)
{
return blk_alloc_queue_node(gfp_mask, -1);
}
EXPORT_SYMBOL(blk_alloc_queue);
-request_queue_t *blk_alloc_queue_node(int gfp_mask, int node_id)
+request_queue_t *blk_alloc_queue_node(gfp_t gfp_mask, int node_id)
{
request_queue_t *q;
static inline void blk_free_request(request_queue_t *q, struct request *rq)
{
- elv_put_request(q, rq);
+ if (rq->flags & REQ_ELVPRIV)
+ elv_put_request(q, rq);
mempool_free(rq, q->rq.rq_pool);
}
static inline struct request *
-blk_alloc_request(request_queue_t *q, int rw, struct bio *bio, int gfp_mask)
+blk_alloc_request(request_queue_t *q, int rw, struct bio *bio,
+ int priv, gfp_t gfp_mask)
{
struct request *rq = mempool_alloc(q->rq.rq_pool, gfp_mask);
*/
rq->flags = rw;
- if (!elv_set_request(q, rq, bio, gfp_mask))
- return rq;
+ if (priv) {
+ if (unlikely(elv_set_request(q, rq, bio, gfp_mask))) {
+ mempool_free(rq, q->rq.rq_pool);
+ return NULL;
+ }
+ rq->flags |= REQ_ELVPRIV;
+ }
- mempool_free(rq, q->rq.rq_pool);
- return NULL;
+ return rq;
}
/*
* A request has just been released. Account for it, update the full and
* congestion status, wake up any waiters. Called under q->queue_lock.
*/
-static void freed_request(request_queue_t *q, int rw)
+static void freed_request(request_queue_t *q, int rw, int priv)
{
struct request_list *rl = &q->rq;
rl->count[rw]--;
+ if (priv)
+ rl->elvpriv--;
__freed_request(q, rw);
if (unlikely(rl->starved[rw ^ 1]))
__freed_request(q, rw ^ 1);
-
- if (!rl->count[READ] && !rl->count[WRITE]) {
- smp_mb();
- if (unlikely(waitqueue_active(&rl->drain)))
- wake_up(&rl->drain);
- }
}
#define blkdev_free_rq(list) list_entry((list)->next, struct request, queuelist)
* Returns !NULL on success, with queue_lock *not held*.
*/
static struct request *get_request(request_queue_t *q, int rw, struct bio *bio,
- int gfp_mask)
+ gfp_t gfp_mask)
{
struct request *rq = NULL;
struct request_list *rl = &q->rq;
struct io_context *ioc = current_io_context(GFP_ATOMIC);
-
- if (unlikely(test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags)))
- goto out;
+ int priv;
if (rl->count[rw]+1 >= q->nr_requests) {
/*
rl->starved[rw] = 0;
if (rl->count[rw] >= queue_congestion_on_threshold(q))
set_queue_congested(q, rw);
+
+ priv = !test_bit(QUEUE_FLAG_ELVSWITCH, &q->queue_flags);
+ if (priv)
+ rl->elvpriv++;
+
spin_unlock_irq(q->queue_lock);
- rq = blk_alloc_request(q, rw, bio, gfp_mask);
+ rq = blk_alloc_request(q, rw, bio, priv, gfp_mask);
if (!rq) {
/*
* Allocation failed presumably due to memory. Undo anything
* wait queue, but this is pretty rare.
*/
spin_lock_irq(q->queue_lock);
- freed_request(q, rw);
+ freed_request(q, rw, priv);
/*
* in the very unlikely event that allocation failed and no
return rq;
}
-struct request *blk_get_request(request_queue_t *q, int rw, int gfp_mask)
+struct request *blk_get_request(request_queue_t *q, int rw, gfp_t gfp_mask)
{
struct request *rq;
/**
* blk_rq_map_user - map user data to a request, for REQ_BLOCK_PC usage
* @q: request queue where request should be inserted
- * @rw: READ or WRITE data
+ * @rq: request structure to fill
* @ubuf: the user buffer
* @len: length of user data
*
* original bio must be passed back in to blk_rq_unmap_user() for proper
* unmapping.
*/
-struct request *blk_rq_map_user(request_queue_t *q, int rw, void __user *ubuf,
- unsigned int len)
+int blk_rq_map_user(request_queue_t *q, struct request *rq, void __user *ubuf,
+ unsigned int len)
{
unsigned long uaddr;
- struct request *rq;
struct bio *bio;
+ int reading;
if (len > (q->max_sectors << 9))
- return ERR_PTR(-EINVAL);
- if ((!len && ubuf) || (len && !ubuf))
- return ERR_PTR(-EINVAL);
+ return -EINVAL;
+ if (!len || !ubuf)
+ return -EINVAL;
- rq = blk_get_request(q, rw, __GFP_WAIT);
- if (!rq)
- return ERR_PTR(-ENOMEM);
+ reading = rq_data_dir(rq) == READ;
/*
* if alignment requirement is satisfied, map in user pages for
*/
uaddr = (unsigned long) ubuf;
if (!(uaddr & queue_dma_alignment(q)) && !(len & queue_dma_alignment(q)))
- bio = bio_map_user(q, NULL, uaddr, len, rw == READ);
+ bio = bio_map_user(q, NULL, uaddr, len, reading);
else
- bio = bio_copy_user(q, uaddr, len, rw == READ);
+ bio = bio_copy_user(q, uaddr, len, reading);
if (!IS_ERR(bio)) {
rq->bio = rq->biotail = bio;
rq->buffer = rq->data = NULL;
rq->data_len = len;
- return rq;
+ return 0;
}
/*
* bio is the err-ptr
*/
- blk_put_request(rq);
- return (struct request *) bio;
+ return PTR_ERR(bio);
}
EXPORT_SYMBOL(blk_rq_map_user);
+/**
+ * blk_rq_map_user_iov - map user data to a request, for REQ_BLOCK_PC usage
+ * @q: request queue where request should be inserted
+ * @rq: request to map data to
+ * @iov: pointer to the iovec
+ * @iov_count: number of elements in the iovec
+ *
+ * Description:
+ * Data will be mapped directly for zero copy io, if possible. Otherwise
+ * a kernel bounce buffer is used.
+ *
+ * A matching blk_rq_unmap_user() must be issued at the end of io, while
+ * still in process context.
+ *
+ * Note: The mapped bio may need to be bounced through blk_queue_bounce()
+ * before being submitted to the device, as pages mapped may be out of
+ * reach. It's the callers responsibility to make sure this happens. The
+ * original bio must be passed back in to blk_rq_unmap_user() for proper
+ * unmapping.
+ */
+int blk_rq_map_user_iov(request_queue_t *q, struct request *rq,
+ struct sg_iovec *iov, int iov_count)
+{
+ struct bio *bio;
+
+ if (!iov || iov_count <= 0)
+ return -EINVAL;
+
+ /* we don't allow misaligned data like bio_map_user() does. If the
+ * user is using sg, they're expected to know the alignment constraints
+ * and respect them accordingly */
+ bio = bio_map_user_iov(q, NULL, iov, iov_count, rq_data_dir(rq)== READ);
+ if (IS_ERR(bio))
+ return PTR_ERR(bio);
+
+ rq->bio = rq->biotail = bio;
+ blk_rq_bio_prep(q, rq, bio);
+ rq->buffer = rq->data = NULL;
+ rq->data_len = bio->bi_size;
+ return 0;
+}
+
+EXPORT_SYMBOL(blk_rq_map_user_iov);
+
/**
* blk_rq_unmap_user - unmap a request with user data
- * @rq: request to be unmapped
- * @bio: bio for the request
+ * @bio: bio to be unmapped
* @ulen: length of user buffer
*
* Description:
- * Unmap a request previously mapped by blk_rq_map_user().
+ * Unmap a bio previously mapped by blk_rq_map_user().
*/
-int blk_rq_unmap_user(struct request *rq, struct bio *bio, unsigned int ulen)
+int blk_rq_unmap_user(struct bio *bio, unsigned int ulen)
{
int ret = 0;
ret = bio_uncopy_user(bio);
}
- blk_put_request(rq);
- return ret;
+ return 0;
}
EXPORT_SYMBOL(blk_rq_unmap_user);
+/**
+ * blk_rq_map_kern - map kernel data to a request, for REQ_BLOCK_PC usage
+ * @q: request queue where request should be inserted
+ * @rq: request to fill
+ * @kbuf: the kernel buffer
+ * @len: length of user data
+ * @gfp_mask: memory allocation flags
+ */
+int blk_rq_map_kern(request_queue_t *q, struct request *rq, void *kbuf,
+ unsigned int len, gfp_t gfp_mask)
+{
+ struct bio *bio;
+
+ if (len > (q->max_sectors << 9))
+ return -EINVAL;
+ if (!len || !kbuf)
+ return -EINVAL;
+
+ bio = bio_map_kern(q, kbuf, len, gfp_mask);
+ if (IS_ERR(bio))
+ return PTR_ERR(bio);
+
+ if (rq_data_dir(rq) == WRITE)
+ bio->bi_rw |= (1 << BIO_RW);
+
+ rq->bio = rq->biotail = bio;
+ blk_rq_bio_prep(q, rq, bio);
+
+ rq->buffer = rq->data = NULL;
+ rq->data_len = len;
+ return 0;
+}
+
+EXPORT_SYMBOL(blk_rq_map_kern);
+
+/**
+ * blk_execute_rq_nowait - insert a request into queue for execution
+ * @q: queue to insert the request in
+ * @bd_disk: matching gendisk
+ * @rq: request to insert
+ * @at_head: insert request at head or tail of queue
+ * @done: I/O completion handler
+ *
+ * Description:
+ * Insert a fully prepared request at the back of the io scheduler queue
+ * for execution. Don't wait for completion.
+ */
+void blk_execute_rq_nowait(request_queue_t *q, struct gendisk *bd_disk,
+ struct request *rq, int at_head,
+ void (*done)(struct request *))
+{
+ int where = at_head ? ELEVATOR_INSERT_FRONT : ELEVATOR_INSERT_BACK;
+
+ rq->rq_disk = bd_disk;
+ rq->flags |= REQ_NOMERGE;
+ rq->end_io = done;
+ elv_add_request(q, rq, where, 1);
+ generic_unplug_device(q);
+}
+
/**
* blk_execute_rq - insert a request into queue for execution
* @q: queue to insert the request in
* @bd_disk: matching gendisk
* @rq: request to insert
+ * @at_head: insert request at head or tail of queue
*
* Description:
* Insert a fully prepared request at the back of the io scheduler queue
- * for execution.
+ * for execution and wait for completion.
*/
int blk_execute_rq(request_queue_t *q, struct gendisk *bd_disk,
- struct request *rq)
+ struct request *rq, int at_head)
{
DECLARE_COMPLETION(wait);
char sense[SCSI_SENSE_BUFFERSIZE];
int err = 0;
- rq->rq_disk = bd_disk;
-
/*
* we need an extra reference to the request, so we can look at
* it after io completion
rq->sense_len = 0;
}
- rq->flags |= REQ_NOMERGE;
rq->waiting = &wait;
- rq->end_io = blk_end_sync_rq;
- elv_add_request(q, rq, ELEVATOR_INSERT_BACK, 1);
- generic_unplug_device(q);
+ blk_execute_rq_nowait(q, bd_disk, rq, at_head, blk_end_sync_rq);
wait_for_completion(&wait);
rq->waiting = NULL;
if (!blk_fs_request(rq) || !rq->rq_disk)
return;
- if (rw == READ) {
- __disk_stat_add(rq->rq_disk, read_sectors, nr_sectors);
- if (!new_io)
- __disk_stat_inc(rq->rq_disk, read_merges);
- } else if (rw == WRITE) {
- __disk_stat_add(rq->rq_disk, write_sectors, nr_sectors);
- if (!new_io)
- __disk_stat_inc(rq->rq_disk, write_merges);
- }
- if (new_io) {
+ if (!new_io) {
+ __disk_stat_inc(rq->rq_disk, merges[rw]);
+ } else {
disk_round_stats(rq->rq_disk);
rq->rq_disk->in_flight++;
}
{
unsigned long now = jiffies;
- __disk_stat_add(disk, time_in_queue,
- disk->in_flight * (now - disk->stamp));
- disk->stamp = now;
+ if (now == disk->stamp)
+ return;
- if (disk->in_flight)
- __disk_stat_add(disk, io_ticks, (now - disk->stamp_idle));
- disk->stamp_idle = now;
+ if (disk->in_flight) {
+ __disk_stat_add(disk, time_in_queue,
+ disk->in_flight * (now - disk->stamp));
+ __disk_stat_add(disk, io_ticks, (now - disk->stamp));
+ }
+ disk->stamp = now;
}
/*
if (unlikely(--req->ref_count))
return;
+ elv_completed_request(q, req);
+
req->rq_status = RQ_INACTIVE;
req->rl = NULL;
*/
if (rl) {
int rw = rq_data_dir(req);
-
- elv_completed_request(q, req);
+ int priv = req->flags & REQ_ELVPRIV;
BUG_ON(!list_empty(&req->queuelist));
blk_free_request(q, req);
- freed_request(q, rw);
+ freed_request(q, rw, priv);
}
}
void blk_put_request(struct request *req)
{
+ unsigned long flags;
+ request_queue_t *q = req->q;
+
/*
- * if req->rl isn't set, this request didnt originate from the
- * block layer, so it's safe to just disregard it
+ * Gee, IDE calls in w/ NULL q. Fix IDE and remove the
+ * following if (q) test.
*/
- if (req->rl) {
- unsigned long flags;
- request_queue_t *q = req->q;
-
+ if (q) {
spin_lock_irqsave(q->queue_lock, flags);
__blk_put_request(q, req);
spin_unlock_irqrestore(q->queue_lock, flags);
if (bdev != bdev->bd_contains) {
struct hd_struct *p = bdev->bd_part;
+ const int rw = bio_data_dir(bio);
+
+ p->sectors[rw] += bio_sectors(bio);
+ p->ios[rw]++;
- switch (bio_data_dir(bio)) {
- case READ:
- p->read_sectors += bio_sectors(bio);
- p->reads++;
- break;
- case WRITE:
- p->write_sectors += bio_sectors(bio);
- p->writes++;
- break;
- }
bio->bi_sector += p->start_sect;
bio->bi_bdev = bdev->bd_contains;
}
}
-void blk_finish_queue_drain(request_queue_t *q)
-{
- struct request_list *rl = &q->rq;
- struct request *rq;
- int requeued = 0;
-
- spin_lock_irq(q->queue_lock);
- clear_bit(QUEUE_FLAG_DRAIN, &q->queue_flags);
-
- while (!list_empty(&q->drain_list)) {
- rq = list_entry_rq(q->drain_list.next);
-
- list_del_init(&rq->queuelist);
- elv_requeue_request(q, rq);
- requeued++;
- }
-
- if (requeued)
- q->request_fn(q);
-
- spin_unlock_irq(q->queue_lock);
-
- wake_up(&rl->wait[0]);
- wake_up(&rl->wait[1]);
- wake_up(&rl->drain);
-}
-
-static int wait_drain(request_queue_t *q, struct request_list *rl, int dispatch)
-{
- int wait = rl->count[READ] + rl->count[WRITE];
-
- if (dispatch)
- wait += !list_empty(&q->queue_head);
-
- return wait;
-}
-
-/*
- * We rely on the fact that only requests allocated through blk_alloc_request()
- * have io scheduler private data structures associated with them. Any other
- * type of request (allocated on stack or through kmalloc()) should not go
- * to the io scheduler core, but be attached to the queue head instead.
- */
-void blk_wait_queue_drained(request_queue_t *q, int wait_dispatch)
-{
- struct request_list *rl = &q->rq;
- DEFINE_WAIT(wait);
-
- spin_lock_irq(q->queue_lock);
- set_bit(QUEUE_FLAG_DRAIN, &q->queue_flags);
-
- while (wait_drain(q, rl, wait_dispatch)) {
- prepare_to_wait(&rl->drain, &wait, TASK_UNINTERRUPTIBLE);
-
- if (wait_drain(q, rl, wait_dispatch)) {
- __generic_unplug_device(q);
- spin_unlock_irq(q->queue_lock);
- io_schedule();
- spin_lock_irq(q->queue_lock);
- }
-
- finish_wait(&rl->drain, &wait);
- }
-
- spin_unlock_irq(q->queue_lock);
-}
-
-/*
- * block waiting for the io scheduler being started again.
- */
-static inline void block_wait_queue_running(request_queue_t *q)
-{
- DEFINE_WAIT(wait);
-
- while (unlikely(test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags))) {
- struct request_list *rl = &q->rq;
-
- prepare_to_wait_exclusive(&rl->drain, &wait,
- TASK_UNINTERRUPTIBLE);
-
- /*
- * re-check the condition. avoids using prepare_to_wait()
- * in the fast path (queue is running)
- */
- if (test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags))
- io_schedule();
-
- finish_wait(&rl->drain, &wait);
- }
-}
-
static void handle_bad_sector(struct bio *bio)
{
char b[BDEVNAME_SIZE];
if (unlikely(test_bit(QUEUE_FLAG_DEAD, &q->queue_flags)))
goto end_io;
- block_wait_queue_running(q);
-
/*
* If this device has partitions, remap block n
* of partition p to block n+start(p) of the disk.
(unsigned long long)req->sector);
}
+ if (blk_fs_request(req) && req->rq_disk) {
+ const int rw = rq_data_dir(req);
+
+ __disk_stat_add(req->rq_disk, sectors[rw], nr_bytes >> 9);
+ }
+
total_bytes = bio_nbytes = 0;
while ((bio = req->bio) != NULL) {
int nbytes;
if (disk && blk_fs_request(req)) {
unsigned long duration = jiffies - req->start_time;
- switch (rq_data_dir(req)) {
- case WRITE:
- __disk_stat_inc(disk, writes);
- __disk_stat_add(disk, write_ticks, duration);
- break;
- case READ:
- __disk_stat_inc(disk, reads);
- __disk_stat_add(disk, read_ticks, duration);
- break;
- }
+ const int rw = rq_data_dir(req);
+
+ __disk_stat_inc(disk, ios[rw]);
+ __disk_stat_add(disk, ticks[rw], duration);
disk_round_stats(disk);
disk->in_flight--;
}
* but since the current task itself holds a reference, the context can be
* used in general code, so long as it stays within `current` context.
*/
-struct io_context *current_io_context(int gfp_flags)
+struct io_context *current_io_context(gfp_t gfp_flags)
{
struct task_struct *tsk = current;
struct io_context *ret;
*
* This is always called in the context of the task which submitted the I/O.
*/
-struct io_context *get_io_context(int gfp_flags)
+struct io_context *get_io_context(gfp_t gfp_flags)
{
struct io_context *ret;
ret = current_io_context(gfp_flags);
projects / linux-2.6 / blobdiff