#include <linux/task_io_accounting_ops.h>
#include <linux/blktrace_api.h>
#include <linux/fault-inject.h>
-#include <trace/block.h>
+
+#define CREATE_TRACE_POINTS
+#include <trace/events/block.h>
#include "blk.h"
-DEFINE_TRACE(block_plug);
-DEFINE_TRACE(block_unplug_io);
-DEFINE_TRACE(block_unplug_timer);
-DEFINE_TRACE(block_getrq);
-DEFINE_TRACE(block_sleeprq);
-DEFINE_TRACE(block_rq_requeue);
-DEFINE_TRACE(block_bio_backmerge);
-DEFINE_TRACE(block_bio_frontmerge);
-DEFINE_TRACE(block_bio_queue);
-DEFINE_TRACE(block_rq_complete);
-DEFINE_TRACE(block_remap); /* Also used in drivers/md/dm.c */
EXPORT_TRACEPOINT_SYMBOL_GPL(block_remap);
+EXPORT_TRACEPOINT_SYMBOL_GPL(block_bio_complete);
static int __make_request(struct request_queue *q, struct bio *bio);
int rw = rq_data_dir(rq);
int cpu;
- if (!blk_fs_request(rq) || !rq->rq_disk)
+ if (!blk_fs_request(rq) || !blk_do_io_stat(rq))
return;
cpu = part_stat_lock();
INIT_HLIST_NODE(&rq->hash);
RB_CLEAR_NODE(&rq->rb_node);
rq->cmd = rq->__cmd;
+ rq->cmd_len = BLK_MAX_CDB;
rq->tag = -1;
rq->ref_count = 1;
}
{
if (unlikely(blk_queue_stopped(q)))
return;
-
- if (!blk_remove_plug(q))
+ if (!blk_remove_plug(q) && !blk_queue_nonrot(q))
return;
q->request_fn(q);
{
del_timer_sync(&q->unplug_timer);
del_timer_sync(&q->timeout);
- kblockd_flush_work(&q->unplug_work);
+ cancel_work_sync(&q->unplug_work);
}
EXPORT_SYMBOL(blk_sync_queue);
{
struct request_list *rl = &q->rq;
- rl->count[READ] = rl->count[WRITE] = 0;
- rl->starved[READ] = rl->starved[WRITE] = 0;
+ rl->count[BLK_RW_SYNC] = rl->count[BLK_RW_ASYNC] = 0;
+ rl->starved[BLK_RW_SYNC] = rl->starved[BLK_RW_ASYNC] = 0;
rl->elvpriv = 0;
- init_waitqueue_head(&rl->wait[READ]);
- init_waitqueue_head(&rl->wait[WRITE]);
+ init_waitqueue_head(&rl->wait[BLK_RW_SYNC]);
+ init_waitqueue_head(&rl->wait[BLK_RW_ASYNC]);
rl->rq_pool = mempool_create_node(BLKDEV_MIN_RQ, mempool_alloc_slab,
mempool_free_slab, request_cachep, q->node);
q->request_fn = rfn;
q->prep_rq_fn = NULL;
q->unplug_fn = generic_unplug_device;
- q->queue_flags = (1 << QUEUE_FLAG_CLUSTER |
- 1 << QUEUE_FLAG_STACKABLE);
+ q->queue_flags = QUEUE_FLAG_DEFAULT;
q->queue_lock = lock;
- blk_queue_segment_boundary(q, BLK_SEG_BOUNDARY_MASK);
-
+ /*
+ * This also sets hw/phys segments, boundary and size
+ */
blk_queue_make_request(q, __make_request);
- blk_queue_max_segment_size(q, MAX_SEGMENT_SIZE);
-
- blk_queue_max_hw_segments(q, MAX_HW_SEGMENTS);
- blk_queue_max_phys_segments(q, MAX_PHYS_SEGMENTS);
q->sg_reserved_size = INT_MAX;
}
static struct request *
-blk_alloc_request(struct request_queue *q, int rw, int priv, gfp_t gfp_mask)
+blk_alloc_request(struct request_queue *q, int flags, int priv, gfp_t gfp_mask)
{
struct request *rq = mempool_alloc(q->rq.rq_pool, gfp_mask);
blk_rq_init(q, rq);
- rq->cmd_flags = rw | REQ_ALLOCED;
+ rq->cmd_flags = flags | REQ_ALLOCED;
if (priv) {
if (unlikely(elv_set_request(q, rq, gfp_mask))) {
ioc->last_waited = jiffies;
}
-static void __freed_request(struct request_queue *q, int rw)
+static void __freed_request(struct request_queue *q, int sync)
{
struct request_list *rl = &q->rq;
- if (rl->count[rw] < queue_congestion_off_threshold(q))
- blk_clear_queue_congested(q, rw);
+ if (rl->count[sync] < queue_congestion_off_threshold(q))
+ blk_clear_queue_congested(q, sync);
- if (rl->count[rw] + 1 <= q->nr_requests) {
- if (waitqueue_active(&rl->wait[rw]))
- wake_up(&rl->wait[rw]);
+ if (rl->count[sync] + 1 <= q->nr_requests) {
+ if (waitqueue_active(&rl->wait[sync]))
+ wake_up(&rl->wait[sync]);
- blk_clear_queue_full(q, rw);
+ blk_clear_queue_full(q, sync);
}
}
* 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(struct request_queue *q, int rw, int priv)
+static void freed_request(struct request_queue *q, int sync, int priv)
{
struct request_list *rl = &q->rq;
- rl->count[rw]--;
+ rl->count[sync]--;
if (priv)
rl->elvpriv--;
- __freed_request(q, rw);
+ __freed_request(q, sync);
- if (unlikely(rl->starved[rw ^ 1]))
- __freed_request(q, rw ^ 1);
+ if (unlikely(rl->starved[sync ^ 1]))
+ __freed_request(q, sync ^ 1);
}
-#define blkdev_free_rq(list) list_entry((list)->next, struct request, queuelist)
/*
* Get a free request, queue_lock must be held.
* Returns NULL on failure, with queue_lock held.
struct request *rq = NULL;
struct request_list *rl = &q->rq;
struct io_context *ioc = NULL;
- const int rw = rw_flags & 0x01;
+ const bool is_sync = rw_is_sync(rw_flags) != 0;
int may_queue, priv;
may_queue = elv_may_queue(q, rw_flags);
if (may_queue == ELV_MQUEUE_NO)
goto rq_starved;
- if (rl->count[rw]+1 >= queue_congestion_on_threshold(q)) {
- if (rl->count[rw]+1 >= q->nr_requests) {
+ if (rl->count[is_sync]+1 >= queue_congestion_on_threshold(q)) {
+ if (rl->count[is_sync]+1 >= q->nr_requests) {
ioc = current_io_context(GFP_ATOMIC, q->node);
/*
* The queue will fill after this allocation, so set
* This process will be allowed to complete a batch of
* requests, others will be blocked.
*/
- if (!blk_queue_full(q, rw)) {
+ if (!blk_queue_full(q, is_sync)) {
ioc_set_batching(q, ioc);
- blk_set_queue_full(q, rw);
+ blk_set_queue_full(q, is_sync);
} else {
if (may_queue != ELV_MQUEUE_MUST
&& !ioc_batching(q, ioc)) {
}
}
}
- blk_set_queue_congested(q, rw);
+ blk_set_queue_congested(q, is_sync);
}
/*
* limit of requests, otherwise we could have thousands of requests
* allocated with any setting of ->nr_requests
*/
- if (rl->count[rw] >= (3 * q->nr_requests / 2))
+ if (rl->count[is_sync] >= (3 * q->nr_requests / 2))
goto out;
- rl->count[rw]++;
- rl->starved[rw] = 0;
+ rl->count[is_sync]++;
+ rl->starved[is_sync] = 0;
priv = !test_bit(QUEUE_FLAG_ELVSWITCH, &q->queue_flags);
if (priv)
rl->elvpriv++;
+ if (blk_queue_io_stat(q))
+ rw_flags |= REQ_IO_STAT;
spin_unlock_irq(q->queue_lock);
rq = blk_alloc_request(q, rw_flags, priv, gfp_mask);
* wait queue, but this is pretty rare.
*/
spin_lock_irq(q->queue_lock);
- freed_request(q, rw, priv);
+ freed_request(q, is_sync, priv);
/*
* in the very unlikely event that allocation failed and no
* rq mempool into READ and WRITE
*/
rq_starved:
- if (unlikely(rl->count[rw] == 0))
- rl->starved[rw] = 1;
+ if (unlikely(rl->count[is_sync] == 0))
+ rl->starved[is_sync] = 1;
goto out;
}
if (ioc_batching(q, ioc))
ioc->nr_batch_requests--;
- trace_block_getrq(q, bio, rw);
+ trace_block_getrq(q, bio, rw_flags & 1);
out:
return rq;
}
static struct request *get_request_wait(struct request_queue *q, int rw_flags,
struct bio *bio)
{
- const int rw = rw_flags & 0x01;
+ const bool is_sync = rw_is_sync(rw_flags) != 0;
struct request *rq;
rq = get_request(q, rw_flags, bio, GFP_NOIO);
struct io_context *ioc;
struct request_list *rl = &q->rq;
- prepare_to_wait_exclusive(&rl->wait[rw], &wait,
+ prepare_to_wait_exclusive(&rl->wait[is_sync], &wait,
TASK_UNINTERRUPTIBLE);
- trace_block_sleeprq(q, bio, rw);
+ trace_block_sleeprq(q, bio, rw_flags & 1);
__generic_unplug_device(q);
spin_unlock_irq(q->queue_lock);
ioc_set_batching(q, ioc);
spin_lock_irq(q->queue_lock);
- finish_wait(&rl->wait[rw], &wait);
+ finish_wait(&rl->wait[is_sync], &wait);
rq = get_request(q, rw_flags, bio, GFP_NOIO);
};
elv_completed_request(q, req);
+ /* this is a bio leak */
+ WARN_ON(req->bio != NULL);
+
/*
* Request may not have originated from ll_rw_blk. if not,
* it didn't come out of our reserved rq pools
*/
if (req->cmd_flags & REQ_ALLOCED) {
- int rw = rq_data_dir(req);
+ int is_sync = rq_is_sync(req) != 0;
int priv = req->cmd_flags & REQ_ELVPRIV;
BUG_ON(!list_empty(&req->queuelist));
BUG_ON(!hlist_unhashed(&req->hash));
blk_free_request(q, req);
- freed_request(q, rw, priv);
+ freed_request(q, is_sync, priv);
}
}
EXPORT_SYMBOL_GPL(__blk_put_request);
req->cmd_flags |= REQ_RW_SYNC;
if (bio_rw_meta(bio))
req->cmd_flags |= REQ_RW_META;
+ if (bio_noidle(bio))
+ req->cmd_flags |= REQ_NOIDLE;
req->errors = 0;
req->hard_sector = req->sector = bio->bi_sector;
blk_rq_bio_prep(req->q, req, bio);
}
+/*
+ * Only disabling plugging for non-rotational devices if it does tagging
+ * as well, otherwise we do need the proper merging
+ */
+static inline bool queue_should_plug(struct request_queue *q)
+{
+ return !(blk_queue_nonrot(q) && blk_queue_tagged(q));
+}
+
static int __make_request(struct request_queue *q, struct bio *bio)
{
struct request *req;
- int el_ret, nr_sectors, barrier, discard, err;
+ int el_ret, nr_sectors;
const unsigned short prio = bio_prio(bio);
const int sync = bio_sync(bio);
+ const int unplug = bio_unplug(bio);
int rw_flags;
nr_sectors = bio_sectors(bio);
*/
blk_queue_bounce(q, &bio);
- barrier = bio_barrier(bio);
- if (unlikely(barrier) && bio_has_data(bio) &&
- (q->next_ordered == QUEUE_ORDERED_NONE)) {
- err = -EOPNOTSUPP;
- goto end_io;
- }
-
- discard = bio_discard(bio);
- if (unlikely(discard) && !q->prepare_discard_fn) {
- err = -EOPNOTSUPP;
- goto end_io;
- }
-
spin_lock_irq(q->queue_lock);
- if (unlikely(barrier) || elv_queue_empty(q))
+ if (unlikely(bio_barrier(bio)) || elv_queue_empty(q))
goto get_rq;
el_ret = elv_merge(q, &req, bio);
if (test_bit(QUEUE_FLAG_SAME_COMP, &q->queue_flags) ||
bio_flagged(bio, BIO_CPU_AFFINE))
req->cpu = blk_cpu_to_group(smp_processor_id());
- if (elv_queue_empty(q))
+ if (queue_should_plug(q) && elv_queue_empty(q))
blk_plug_device(q);
add_request(q, req);
out:
- if (sync)
+ if (unplug || !queue_should_plug(q))
__generic_unplug_device(q);
spin_unlock_irq(q->queue_lock);
return 0;
-
-end_io:
- bio_endio(bio, err);
- return 0;
}
/*
bio->bi_bdev = bdev->bd_contains;
trace_block_remap(bdev_get_queue(bio->bi_bdev), bio,
- bdev->bd_dev, bio->bi_sector,
+ bdev->bd_dev,
bio->bi_sector - p->start_sect);
}
}
char b[BDEVNAME_SIZE];
q = bdev_get_queue(bio->bi_bdev);
- if (!q) {
+ if (unlikely(!q)) {
printk(KERN_ERR
"generic_make_request: Trying to access "
"nonexistent block-device %s (%Lu)\n",
bdevname(bio->bi_bdev, b),
(long long) bio->bi_sector);
-end_io:
- bio_endio(bio, err);
- break;
+ goto end_io;
}
if (unlikely(nr_sectors > q->max_hw_sectors)) {
goto end_io;
if (old_sector != -1)
- trace_block_remap(q, bio, old_dev, bio->bi_sector,
- old_sector);
+ trace_block_remap(q, bio, old_dev, old_sector);
trace_block_bio_queue(q, bio);
if (bio_check_eod(bio, nr_sectors))
goto end_io;
- if ((bio_empty_barrier(bio) && !q->prepare_flush_fn) ||
- (bio_discard(bio) && !q->prepare_discard_fn)) {
+
+ if (bio_discard(bio) && !q->prepare_discard_fn) {
+ err = -EOPNOTSUPP;
+ goto end_io;
+ }
+ if (bio_barrier(bio) && bio_has_data(bio) &&
+ (q->next_ordered == QUEUE_ORDERED_NONE)) {
err = -EOPNOTSUPP;
goto end_io;
}
ret = q->make_request_fn(q, bio);
} while (ret);
+
+ return;
+
+end_io:
+ bio_endio(bio, err);
}
/*
}
EXPORT_SYMBOL(blkdev_dequeue_request);
+static void blk_account_io_completion(struct request *req, unsigned int bytes)
+{
+ if (!blk_do_io_stat(req))
+ return;
+
+ if (blk_fs_request(req)) {
+ const int rw = rq_data_dir(req);
+ struct hd_struct *part;
+ int cpu;
+
+ cpu = part_stat_lock();
+ part = disk_map_sector_rcu(req->rq_disk, req->sector);
+ part_stat_add(cpu, part, sectors[rw], bytes >> 9);
+ part_stat_unlock();
+ }
+}
+
+static void blk_account_io_done(struct request *req)
+{
+ if (!blk_do_io_stat(req))
+ return;
+
+ /*
+ * Account IO completion. bar_rq isn't accounted as a normal
+ * IO on queueing nor completion. Accounting the containing
+ * request is enough.
+ */
+ if (blk_fs_request(req) && req != &req->q->bar_rq) {
+ unsigned long duration = jiffies - req->start_time;
+ const int rw = rq_data_dir(req);
+ struct hd_struct *part;
+ int cpu;
+
+ cpu = part_stat_lock();
+ part = disk_map_sector_rcu(req->rq_disk, req->sector);
+
+ part_stat_inc(cpu, part, ios[rw]);
+ part_stat_add(cpu, part, ticks[rw], duration);
+ part_round_stats(cpu, part);
+ part_dec_in_flight(part);
+
+ part_stat_unlock();
+ }
+}
+
/**
* __end_that_request_first - end I/O on a request
* @req: the request being processed
trace_block_rq_complete(req->q, req);
/*
- * for a REQ_TYPE_BLOCK_PC request, we want to carry any eventual
- * sense key with us all the way through
+ * For fs requests, rq is just carrier of independent bio's
+ * and each partial completion should be handled separately.
+ * Reset per-request error on each partial completion.
+ *
+ * TODO: tj: This is too subtle. It would be better to let
+ * low level drivers do what they see fit.
*/
- if (!blk_pc_request(req))
+ if (blk_fs_request(req))
req->errors = 0;
if (error && (blk_fs_request(req) && !(req->cmd_flags & REQ_QUIET))) {
(unsigned long long)req->sector);
}
- if (blk_fs_request(req) && req->rq_disk) {
- const int rw = rq_data_dir(req);
- struct hd_struct *part;
- int cpu;
-
- cpu = part_stat_lock();
- part = disk_map_sector_rcu(req->rq_disk, req->sector);
- part_stat_add(cpu, part, sectors[rw], nr_bytes >> 9);
- part_stat_unlock();
- }
+ blk_account_io_completion(req, nr_bytes);
total_bytes = bio_nbytes = 0;
while ((bio = req->bio) != NULL) {
int nbytes;
- /*
- * For an empty barrier request, the low level driver must
- * store a potential error location in ->sector. We pass
- * that back up in ->bi_sector.
- */
- if (blk_empty_barrier(req))
- bio->bi_sector = req->sector;
-
if (nr_bytes >= bio->bi_size) {
req->bio = bio->bi_next;
nbytes = bio->bi_size;
} else {
int idx = bio->bi_idx + next_idx;
- if (unlikely(bio->bi_idx >= bio->bi_vcnt)) {
+ if (unlikely(idx >= bio->bi_vcnt)) {
blk_dump_rq_flags(req, "__end_that");
printk(KERN_ERR "%s: bio idx %d >= vcnt %d\n",
- __func__, bio->bi_idx, bio->bi_vcnt);
+ __func__, idx, bio->bi_vcnt);
break;
}
*/
static void end_that_request_last(struct request *req, int error)
{
- struct gendisk *disk = req->rq_disk;
-
if (blk_rq_tagged(req))
blk_queue_end_tag(req->q, req);
blk_delete_timer(req);
- /*
- * Account IO completion. bar_rq isn't accounted as a normal
- * IO on queueing nor completion. Accounting the containing
- * request is enough.
- */
- if (disk && blk_fs_request(req) && req != &req->q->bar_rq) {
- unsigned long duration = jiffies - req->start_time;
- const int rw = rq_data_dir(req);
- struct hd_struct *part;
- int cpu;
-
- cpu = part_stat_lock();
- part = disk_map_sector_rcu(disk, req->sector);
-
- part_stat_inc(cpu, part, ios[rw]);
- part_stat_add(cpu, part, ticks[rw], duration);
- part_round_stats(cpu, part);
- part_dec_in_flight(part);
-
- part_stat_unlock();
- }
+ blk_account_io_done(req);
if (req->end_io)
req->end_io(req, error);
}
EXPORT_SYMBOL(kblockd_schedule_work);
-void kblockd_flush_work(struct work_struct *work)
-{
- cancel_work_sync(work);
-}
-EXPORT_SYMBOL(kblockd_flush_work);
-
int __init blk_dev_init(void)
{
kblockd_workqueue = create_workqueue("kblockd");
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff