Merge branch 'for-2.6.28' of git://git.kernel.dk/linux-2.6-block

* 'for-2.6.28' of git://git.kernel.dk/linux-2.6-block: (132 commits)
  doc/cdrom: Trvial documentation error, file not present
  block_dev: fix kernel-doc in new functions
  block: add some comments around the bio read-write flags
  block: mark bio_split_pool static
  block: Find bio sector offset given idx and offset
  block: gendisk integrity wrapper
  block: Switch blk_integrity_compare from bdev to gendisk
  block: Fix double put in blk_integrity_unregister
  block: Introduce integrity data ownership flag
  block: revert part of d7533ad0e132f92e75c1b2eb7c26387b25a583c1
  bio.h: Remove unused conditional code
  block: remove end_{queued|dequeued}_request()
  block: change elevator to use __blk_end_request()
  gdrom: change to use __blk_end_request()
  memstick: change to use __blk_end_request()
  virtio_blk: change to use __blk_end_request()
  blktrace: use BLKTRACE_BDEV_SIZE as the name size for setup structure
  block: add lld busy state exporting interface
  block: Fix blk_start_queueing() to not kick a stopped queue
  include blktrace_api.h in headers_install
  ...

24 files changed, 1894 insertions, 880 deletions

diff --git a/block/Makefile b/block/Makefile
index 208000b0750..bfe73049f93 100644
--- a/block/Makefile
+++ b/block/Makefile
@@ -4,8 +4,8 @@
 
 obj-$(CONFIG_BLOCK) := elevator.o blk-core.o blk-tag.o blk-sysfs.o \
 			blk-barrier.o blk-settings.o blk-ioc.o blk-map.o \
-			blk-exec.o blk-merge.o ioctl.o genhd.o scsi_ioctl.o \
-			cmd-filter.o
+			blk-exec.o blk-merge.o blk-softirq.o blk-timeout.o \
+			ioctl.o genhd.o scsi_ioctl.o cmd-filter.o
 
 obj-$(CONFIG_BLK_DEV_BSG)	+= bsg.o
 obj-$(CONFIG_IOSCHED_NOOP)	+= noop-iosched.o
diff --git a/block/as-iosched.c b/block/as-iosched.c
index cf4eb0eefbb..71f0abb219e 100644
--- a/block/as-iosched.c
+++ b/block/as-iosched.c
@@ -462,7 +462,7 @@ static void as_antic_stop(struct as_data *ad)
 			del_timer(&ad->antic_timer);
 		ad->antic_status = ANTIC_FINISHED;
 		/* see as_work_handler */
-		kblockd_schedule_work(&ad->antic_work);
+		kblockd_schedule_work(ad->q, &ad->antic_work);
 	}
 }
 
@@ -483,7 +483,7 @@ static void as_antic_timeout(unsigned long data)
 		aic = ad->io_context->aic;
 
 		ad->antic_status = ANTIC_FINISHED;
-		kblockd_schedule_work(&ad->antic_work);
+		kblockd_schedule_work(q, &ad->antic_work);
 
 		if (aic->ttime_samples == 0) {
 			/* process anticipated on has exited or timed out*/
@@ -745,6 +745,14 @@ static int as_can_break_anticipation(struct as_data *ad, struct request *rq)
  */
 static int as_can_anticipate(struct as_data *ad, struct request *rq)
 {
+#if 0 /* disable for now, we need to check tag level as well */
+	/*
+	 * SSD device without seek penalty, disable idling
+	 */
+	if (blk_queue_nonrot(ad->q)) axman
+		return 0;
+#endif
+
 	if (!ad->io_context)
 		/*
 		 * Last request submitted was a write
@@ -844,7 +852,7 @@ static void as_completed_request(struct request_queue *q, struct request *rq)
 	if (ad->changed_batch && ad->nr_dispatched == 1) {
 		ad->current_batch_expires = jiffies +
 					ad->batch_expire[ad->batch_data_dir];
-		kblockd_schedule_work(&ad->antic_work);
+		kblockd_schedule_work(q, &ad->antic_work);
 		ad->changed_batch = 0;
 
 		if (ad->batch_data_dir == REQ_SYNC)
diff --git a/block/blk-barrier.c b/block/blk-barrier.c
index a09ead19f9c..5c99ff8d2db 100644
--- a/block/blk-barrier.c
+++ b/block/blk-barrier.c
@@ -293,7 +293,7 @@ int blkdev_issue_flush(struct block_device *bdev, sector_t *error_sector)
 	bio->bi_end_io = bio_end_empty_barrier;
 	bio->bi_private = &wait;
 	bio->bi_bdev = bdev;
-	submit_bio(1 << BIO_RW_BARRIER, bio);
+	submit_bio(WRITE_BARRIER, bio);
 
 	wait_for_completion(&wait);
 
@@ -315,3 +315,73 @@ int blkdev_issue_flush(struct block_device *bdev, sector_t *error_sector)
 	return ret;
 }
 EXPORT_SYMBOL(blkdev_issue_flush);
+
+static void blkdev_discard_end_io(struct bio *bio, int err)
+{
+	if (err) {
+		if (err == -EOPNOTSUPP)
+			set_bit(BIO_EOPNOTSUPP, &bio->bi_flags);
+		clear_bit(BIO_UPTODATE, &bio->bi_flags);
+	}
+
+	bio_put(bio);
+}
+
+/**
+ * blkdev_issue_discard - queue a discard
+ * @bdev:	blockdev to issue discard for
+ * @sector:	start sector
+ * @nr_sects:	number of sectors to discard
+ * @gfp_mask:	memory allocation flags (for bio_alloc)
+ *
+ * Description:
+ *    Issue a discard request for the sectors in question. Does not wait.
+ */
+int blkdev_issue_discard(struct block_device *bdev,
+			 sector_t sector, sector_t nr_sects, gfp_t gfp_mask)
+{
+	struct request_queue *q;
+	struct bio *bio;
+	int ret = 0;
+
+	if (bdev->bd_disk == NULL)
+		return -ENXIO;
+
+	q = bdev_get_queue(bdev);
+	if (!q)
+		return -ENXIO;
+
+	if (!q->prepare_discard_fn)
+		return -EOPNOTSUPP;
+
+	while (nr_sects && !ret) {
+		bio = bio_alloc(gfp_mask, 0);
+		if (!bio)
+			return -ENOMEM;
+
+		bio->bi_end_io = blkdev_discard_end_io;
+		bio->bi_bdev = bdev;
+
+		bio->bi_sector = sector;
+
+		if (nr_sects > q->max_hw_sectors) {
+			bio->bi_size = q->max_hw_sectors << 9;
+			nr_sects -= q->max_hw_sectors;
+			sector += q->max_hw_sectors;
+		} else {
+			bio->bi_size = nr_sects << 9;
+			nr_sects = 0;
+		}
+		bio_get(bio);
+		submit_bio(DISCARD_BARRIER, bio);
+
+		/* Check if it failed immediately */
+		if (bio_flagged(bio, BIO_EOPNOTSUPP))
+			ret = -EOPNOTSUPP;
+		else if (!bio_flagged(bio, BIO_UPTODATE))
+			ret = -EIO;
+		bio_put(bio);
+	}
+	return ret;
+}
+EXPORT_SYMBOL(blkdev_issue_discard);
diff --git a/block/blk-core.c b/block/blk-core.c
index 2cba5ef97b2..2d053b58441 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -26,8 +26,6 @@
 #include <linux/swap.h>
 #include <linux/writeback.h>
 #include <linux/task_io_accounting_ops.h>
-#include <linux/interrupt.h>
-#include <linux/cpu.h>
 #include <linux/blktrace_api.h>
 #include <linux/fault-inject.h>
 
@@ -50,27 +48,26 @@ struct kmem_cache *blk_requestq_cachep;
  */
 static struct workqueue_struct *kblockd_workqueue;
 
-static DEFINE_PER_CPU(struct list_head, blk_cpu_done);
-
 static void drive_stat_acct(struct request *rq, int new_io)
 {
 	struct hd_struct *part;
 	int rw = rq_data_dir(rq);
+	int cpu;
 
 	if (!blk_fs_request(rq) || !rq->rq_disk)
 		return;
 
-	part = get_part(rq->rq_disk, rq->sector);
+	cpu = part_stat_lock();
+	part = disk_map_sector_rcu(rq->rq_disk, rq->sector);
+
 	if (!new_io)
-		__all_stat_inc(rq->rq_disk, part, merges[rw], rq->sector);
+		part_stat_inc(cpu, part, merges[rw]);
 	else {
-		disk_round_stats(rq->rq_disk);
-		rq->rq_disk->in_flight++;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight++;
-		}
+		part_round_stats(cpu, part);
+		part_inc_in_flight(part);
 	}
+
+	part_stat_unlock();
 }
 
 void blk_queue_congestion_threshold(struct request_queue *q)
@@ -113,7 +110,8 @@ void blk_rq_init(struct request_queue *q, struct request *rq)
 	memset(rq, 0, sizeof(*rq));
 
 	INIT_LIST_HEAD(&rq->queuelist);
-	INIT_LIST_HEAD(&rq->donelist);
+	INIT_LIST_HEAD(&rq->timeout_list);
+	rq->cpu = -1;
 	rq->q = q;
 	rq->sector = rq->hard_sector = (sector_t) -1;
 	INIT_HLIST_NODE(&rq->hash);
@@ -308,7 +306,7 @@ void blk_unplug_timeout(unsigned long data)
 	blk_add_trace_pdu_int(q, BLK_TA_UNPLUG_TIMER, NULL,
 				q->rq.count[READ] + q->rq.count[WRITE]);
 
-	kblockd_schedule_work(&q->unplug_work);
+	kblockd_schedule_work(q, &q->unplug_work);
 }
 
 void blk_unplug(struct request_queue *q)
@@ -325,6 +323,21 @@ void blk_unplug(struct request_queue *q)
 }
 EXPORT_SYMBOL(blk_unplug);
 
+static void blk_invoke_request_fn(struct request_queue *q)
+{
+	/*
+	 * one level of recursion is ok and is much faster than kicking
+	 * the unplug handling
+	 */
+	if (!queue_flag_test_and_set(QUEUE_FLAG_REENTER, q)) {
+		q->request_fn(q);
+		queue_flag_clear(QUEUE_FLAG_REENTER, q);
+	} else {
+		queue_flag_set(QUEUE_FLAG_PLUGGED, q);
+		kblockd_schedule_work(q, &q->unplug_work);
+	}
+}
+
 /**
  * blk_start_queue - restart a previously stopped queue
  * @q:    The &struct request_queue in question
@@ -339,18 +352,7 @@ void blk_start_queue(struct request_queue *q)
 	WARN_ON(!irqs_disabled());
 
 	queue_flag_clear(QUEUE_FLAG_STOPPED, q);
-
-	/*
-	 * one level of recursion is ok and is much faster than kicking
-	 * the unplug handling
-	 */
-	if (!queue_flag_test_and_set(QUEUE_FLAG_REENTER, q)) {
-		q->request_fn(q);
-		queue_flag_clear(QUEUE_FLAG_REENTER, q);
-	} else {
-		blk_plug_device(q);
-		kblockd_schedule_work(&q->unplug_work);
-	}
+	blk_invoke_request_fn(q);
 }
 EXPORT_SYMBOL(blk_start_queue);
 
@@ -408,15 +410,8 @@ void __blk_run_queue(struct request_queue *q)
 	 * Only recurse once to avoid overrunning the stack, let the unplug
 	 * handling reinvoke the handler shortly if we already got there.
 	 */
-	if (!elv_queue_empty(q)) {
-		if (!queue_flag_test_and_set(QUEUE_FLAG_REENTER, q)) {
-			q->request_fn(q);
-			queue_flag_clear(QUEUE_FLAG_REENTER, q);
-		} else {
-			blk_plug_device(q);
-			kblockd_schedule_work(&q->unplug_work);
-		}
-	}
+	if (!elv_queue_empty(q))
+		blk_invoke_request_fn(q);
 }
 EXPORT_SYMBOL(__blk_run_queue);
 
@@ -441,6 +436,14 @@ void blk_put_queue(struct request_queue *q)
 
 void blk_cleanup_queue(struct request_queue *q)
 {
+	/*
+	 * We know we have process context here, so we can be a little
+	 * cautious and ensure that pending block actions on this device
+	 * are done before moving on. Going into this function, we should
+	 * not have processes doing IO to this device.
+	 */
+	blk_sync_queue(q);
+
 	mutex_lock(&q->sysfs_lock);
 	queue_flag_set_unlocked(QUEUE_FLAG_DEAD, q);
 	mutex_unlock(&q->sysfs_lock);
@@ -496,6 +499,8 @@ struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id)
 	}
 
 	init_timer(&q->unplug_timer);
+	setup_timer(&q->timeout, blk_rq_timed_out_timer, (unsigned long) q);
+	INIT_LIST_HEAD(&q->timeout_list);
 
 	kobject_init(&q->kobj, &blk_queue_ktype);
 
@@ -531,7 +536,7 @@ EXPORT_SYMBOL(blk_alloc_queue_node);
  *    request queue; this lock will be taken also from interrupt context, so irq
  *    disabling is needed for it.
  *
- *    Function returns a pointer to the initialized request queue, or NULL if
+ *    Function returns a pointer to the initialized request queue, or %NULL if
  *    it didn't succeed.
  *
  * Note:
@@ -569,7 +574,8 @@ blk_init_queue_node(request_fn_proc *rfn, spinlock_t *lock, int node_id)
 	q->request_fn		= rfn;
 	q->prep_rq_fn		= NULL;
 	q->unplug_fn		= generic_unplug_device;
-	q->queue_flags		= (1 << QUEUE_FLAG_CLUSTER);
+	q->queue_flags		= (1 << QUEUE_FLAG_CLUSTER |
+				   1 << QUEUE_FLAG_STACKABLE);
 	q->queue_lock		= lock;
 
 	blk_queue_segment_boundary(q, 0xffffffff);
@@ -624,10 +630,6 @@ blk_alloc_request(struct request_queue *q, int rw, int priv, gfp_t gfp_mask)
 
 	blk_rq_init(q, rq);
 
-	/*
-	 * first three bits are identical in rq->cmd_flags and bio->bi_rw,
-	 * see bio.h and blkdev.h
-	 */
 	rq->cmd_flags = rw | REQ_ALLOCED;
 
 	if (priv) {
@@ -888,9 +890,11 @@ EXPORT_SYMBOL(blk_get_request);
  */
 void blk_start_queueing(struct request_queue *q)
 {
-	if (!blk_queue_plugged(q))
+	if (!blk_queue_plugged(q)) {
+		if (unlikely(blk_queue_stopped(q)))
+			return;
 		q->request_fn(q);
-	else
+	} else
 		__generic_unplug_device(q);
 }
 EXPORT_SYMBOL(blk_start_queueing);
@@ -907,6 +911,8 @@ EXPORT_SYMBOL(blk_start_queueing);
  */
 void blk_requeue_request(struct request_queue *q, struct request *rq)
 {
+	blk_delete_timer(rq);
+	blk_clear_rq_complete(rq);
 	blk_add_trace_rq(q, rq, BLK_TA_REQUEUE);
 
 	if (blk_rq_tagged(rq))
@@ -917,7 +923,7 @@ void blk_requeue_request(struct request_queue *q, struct request *rq)
 EXPORT_SYMBOL(blk_requeue_request);
 
 /**
- * blk_insert_request - insert a special request in to a request queue
+ * blk_insert_request - insert a special request into a request queue
  * @q:		request queue where request should be inserted
  * @rq:		request to be inserted
  * @at_head:	insert request at head or tail of queue
@@ -927,8 +933,8 @@ EXPORT_SYMBOL(blk_requeue_request);
  *    Many block devices need to execute commands asynchronously, so they don't
  *    block the whole kernel from preemption during request execution.  This is
  *    accomplished normally by inserting aritficial requests tagged as
- *    REQ_SPECIAL in to the corresponding request queue, and letting them be
- *    scheduled for actual execution by the request queue.
+ *    REQ_TYPE_SPECIAL in to the corresponding request queue, and letting them
+ *    be scheduled for actual execution by the request queue.
  *
  *    We have the option of inserting the head or the tail of the queue.
  *    Typically we use the tail for new ioctls and so forth.  We use the head
@@ -982,8 +988,22 @@ static inline void add_request(struct request_queue *q, struct request *req)
 	__elv_add_request(q, req, ELEVATOR_INSERT_SORT, 0);
 }
 
-/*
- * disk_round_stats()	- Round off the performance stats on a struct
+static void part_round_stats_single(int cpu, struct hd_struct *part,
+				    unsigned long now)
+{
+	if (now == part->stamp)
+		return;
+
+	if (part->in_flight) {
+		__part_stat_add(cpu, part, time_in_queue,
+				part->in_flight * (now - part->stamp));
+		__part_stat_add(cpu, part, io_ticks, (now - part->stamp));
+	}
+	part->stamp = now;
+}
+
+/**
+ * part_round_stats()	- Round off the performance stats on a struct
  * disk_stats.
  *
  * The average IO queue length and utilisation statistics are maintained
@@ -997,36 +1017,15 @@ static inline void add_request(struct request_queue *q, struct request *req)
  * /proc/diskstats.  This accounts immediately for all queue usage up to
  * the current jiffies and restarts the counters again.
  */
-void disk_round_stats(struct gendisk *disk)
+void part_round_stats(int cpu, struct hd_struct *part)
 {
 	unsigned long now = jiffies;
 
-	if (now == disk->stamp)
-		return;
-
-	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;
-}
-EXPORT_SYMBOL_GPL(disk_round_stats);
-
-void part_round_stats(struct hd_struct *part)
-{
-	unsigned long now = jiffies;
-
-	if (now == part->stamp)
-		return;
-
-	if (part->in_flight) {
-		__part_stat_add(part, time_in_queue,
-				part->in_flight * (now - part->stamp));
-		__part_stat_add(part, io_ticks, (now - part->stamp));
-	}
-	part->stamp = now;
+	if (part->partno)
+		part_round_stats_single(cpu, &part_to_disk(part)->part0, now);
+	part_round_stats_single(cpu, part, now);
 }
+EXPORT_SYMBOL_GPL(part_round_stats);
 
 /*
  * queue lock must be held
@@ -1070,6 +1069,7 @@ EXPORT_SYMBOL(blk_put_request);
 
 void init_request_from_bio(struct request *req, struct bio *bio)
 {
+	req->cpu = bio->bi_comp_cpu;
 	req->cmd_type = REQ_TYPE_FS;
 
 	/*
@@ -1081,7 +1081,12 @@ void init_request_from_bio(struct request *req, struct bio *bio)
 	/*
 	 * REQ_BARRIER implies no merging, but lets make it explicit
 	 */
-	if (unlikely(bio_barrier(bio)))
+	if (unlikely(bio_discard(bio))) {
+		req->cmd_flags |= REQ_DISCARD;
+		if (bio_barrier(bio))
+			req->cmd_flags |= REQ_SOFTBARRIER;
+		req->q->prepare_discard_fn(req->q, req);
+	} else if (unlikely(bio_barrier(bio)))
 		req->cmd_flags |= (REQ_HARDBARRIER | REQ_NOMERGE);
 
 	if (bio_sync(bio))
@@ -1099,7 +1104,7 @@ void init_request_from_bio(struct request *req, struct bio *bio)
 static int __make_request(struct request_queue *q, struct bio *bio)
 {
 	struct request *req;
-	int el_ret, nr_sectors, barrier, err;
+	int el_ret, nr_sectors, barrier, discard, err;
 	const unsigned short prio = bio_prio(bio);
 	const int sync = bio_sync(bio);
 	int rw_flags;
@@ -1114,7 +1119,14 @@ static int __make_request(struct request_queue *q, struct bio *bio)
 	blk_queue_bounce(q, &bio);
 
 	barrier = bio_barrier(bio);
-	if (unlikely(barrier) && (q->next_ordered == QUEUE_ORDERED_NONE)) {
+	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;
 	}
@@ -1138,6 +1150,8 @@ static int __make_request(struct request_queue *q, struct bio *bio)
 		req->biotail = bio;
 		req->nr_sectors = req->hard_nr_sectors += nr_sectors;
 		req->ioprio = ioprio_best(req->ioprio, prio);
+		if (!blk_rq_cpu_valid(req))
+			req->cpu = bio->bi_comp_cpu;
 		drive_stat_acct(req, 0);
 		if (!attempt_back_merge(q, req))
 			elv_merged_request(q, req, el_ret);
@@ -1165,6 +1179,8 @@ static int __make_request(struct request_queue *q, struct bio *bio)
 		req->sector = req->hard_sector = bio->bi_sector;
 		req->nr_sectors = req->hard_nr_sectors += nr_sectors;
 		req->ioprio = ioprio_best(req->ioprio, prio);
+		if (!blk_rq_cpu_valid(req))
+			req->cpu = bio->bi_comp_cpu;
 		drive_stat_acct(req, 0);
 		if (!attempt_front_merge(q, req))
 			elv_merged_request(q, req, el_ret);
@@ -1200,13 +1216,15 @@ get_rq:
 	init_request_from_bio(req, bio);
 
 	spin_lock_irq(q->queue_lock);
+	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))
 		blk_plug_device(q);
 	add_request(q, req);
 out:
 	if (sync)
 		__generic_unplug_device(q);
-
 	spin_unlock_irq(q->queue_lock);
 	return 0;
 
@@ -1260,8 +1278,9 @@ __setup("fail_make_request=", setup_fail_make_request);
 
 static int should_fail_request(struct bio *bio)
 {
-	if ((bio->bi_bdev->bd_disk->flags & GENHD_FL_FAIL) ||
-	    (bio->bi_bdev->bd_part && bio->bi_bdev->bd_part->make_it_fail))
+	struct hd_struct *part = bio->bi_bdev->bd_part;
+
+	if (part_to_disk(part)->part0.make_it_fail || part->make_it_fail)
 		return should_fail(&fail_make_request, bio->bi_size);
 
 	return 0;
@@ -1314,7 +1333,7 @@ static inline int bio_check_eod(struct bio *bio, unsigned int nr_sectors)
 }
 
 /**
- * generic_make_request: hand a buffer to its device driver for I/O
+ * generic_make_request - hand a buffer to its device driver for I/O
  * @bio:  The bio describing the location in memory and on the device.
  *
  * generic_make_request() is used to make I/O requests of block
@@ -1409,7 +1428,8 @@ end_io:
 
 		if (bio_check_eod(bio, nr_sectors))
 			goto end_io;
-		if (bio_empty_barrier(bio) && !q->prepare_flush_fn) {
+		if ((bio_empty_barrier(bio) && !q->prepare_flush_fn) ||
+		    (bio_discard(bio) && !q->prepare_discard_fn)) {
 			err = -EOPNOTSUPP;
 			goto end_io;
 		}
@@ -1471,13 +1491,13 @@ void generic_make_request(struct bio *bio)
 EXPORT_SYMBOL(generic_make_request);
 
 /**
- * submit_bio: submit a bio to the block device layer for I/O
+ * submit_bio - submit a bio to the block device layer for I/O
  * @rw: whether to %READ or %WRITE, or maybe to %READA (read ahead)
  * @bio: The &struct bio which describes the I/O
  *
  * submit_bio() is very similar in purpose to generic_make_request(), and
  * uses that function to do most of the work. Both are fairly rough
- * interfaces, @bio must be presetup and ready for I/O.
+ * interfaces; @bio must be presetup and ready for I/O.
  *
  */
 void submit_bio(int rw, struct bio *bio)
@@ -1490,11 +1510,7 @@ void submit_bio(int rw, struct bio *bio)
 	 * If it's a regular read/write or a barrier with data attached,
 	 * go through the normal accounting stuff before submission.
 	 */
-	if (!bio_empty_barrier(bio)) {
-
-		BIO_BUG_ON(!bio->bi_size);
-		BIO_BUG_ON(!bio->bi_io_vec);
-
+	if (bio_has_data(bio)) {
 		if (rw & WRITE) {
 			count_vm_events(PGPGOUT, count);
 		} else {
@@ -1517,9 +1533,90 @@ void submit_bio(int rw, struct bio *bio)
 EXPORT_SYMBOL(submit_bio);
 
 /**
+ * blk_rq_check_limits - Helper function to check a request for the queue limit
+ * @q:  the queue
+ * @rq: the request being checked
+ *
+ * Description:
+ *    @rq may have been made based on weaker limitations of upper-level queues
+ *    in request stacking drivers, and it may violate the limitation of @q.
+ *    Since the block layer and the underlying device driver trust @rq
+ *    after it is inserted to @q, it should be checked against @q before
+ *    the insertion using this generic function.
+ *
+ *    This function should also be useful for request stacking drivers
+ *    in some cases below, so export this fuction.
+ *    Request stacking drivers like request-based dm may change the queue
+ *    limits while requests are in the queue (e.g. dm's table swapping).
+ *    Such request stacking drivers should check those requests agaist
+ *    the new queue limits again when they dispatch those requests,
+ *    although such checkings are also done against the old queue limits
+ *    when submitting requests.
+ */
+int blk_rq_check_limits(struct request_queue *q, struct request *rq)
+{
+	if (rq->nr_sectors > q->max_sectors ||
+	    rq->data_len > q->max_hw_sectors << 9) {
+		printk(KERN_ERR "%s: over max size limit.\n", __func__);
+		return -EIO;
+	}
+
+	/*
+	 * queue's settings related to segment counting like q->bounce_pfn
+	 * may differ from that of other stacking queues.
+	 * Recalculate it to check the request correctly on this queue's
+	 * limitation.
+	 */
+	blk_recalc_rq_segments(rq);
+	if (rq->nr_phys_segments > q->max_phys_segments ||
+	    rq->nr_phys_segments > q->max_hw_segments) {
+		printk(KERN_ERR "%s: over max segments limit.\n", __func__);
+		return -EIO;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(blk_rq_check_limits);
+
+/**
+ * blk_insert_cloned_request - Helper for stacking drivers to submit a request
+ * @q:  the queue to submit the request
+ * @rq: the request being queued
+ */
+int blk_insert_cloned_request(struct request_queue *q, struct request *rq)
+{
+	unsigned long flags;
+
+	if (blk_rq_check_limits(q, rq))
+		return -EIO;
+
+#ifdef CONFIG_FAIL_MAKE_REQUEST
+	if (rq->rq_disk && rq->rq_disk->part0.make_it_fail &&
+	    should_fail(&fail_make_request, blk_rq_bytes(rq)))
+		return -EIO;
+#endif
+
+	spin_lock_irqsave(q->queue_lock, flags);
+
+	/*
+	 * Submitting request must be dequeued before calling this function
+	 * because it will be linked to another request_queue
+	 */
+	BUG_ON(blk_queued_rq(rq));
+
+	drive_stat_acct(rq, 1);
+	__elv_add_request(q, rq, ELEVATOR_INSERT_BACK, 0);
+
+	spin_unlock_irqrestore(q->queue_lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(blk_insert_cloned_request);
+
+/**
  * __end_that_request_first - end I/O on a request
  * @req:      the request being processed
- * @error:    0 for success, < 0 for error
+ * @error:    %0 for success, < %0 for error
  * @nr_bytes: number of bytes to complete
  *
  * Description:
@@ -1527,8 +1624,8 @@ EXPORT_SYMBOL(submit_bio);
  *     for the next range of segments (if any) in the cluster.
  *
  * Return:
- *     0 - we are done with this request, call end_that_request_last()
- *     1 - still buffers pending for this request
+ *     %0 - we are done with this request, call end_that_request_last()
+ *     %1 - still buffers pending for this request
  **/
 static int __end_that_request_first(struct request *req, int error,
 				    int nr_bytes)
@@ -1539,7 +1636,7 @@ static int __end_that_request_first(struct request *req, int error,
 	blk_add_trace_rq(req->q, req, BLK_TA_COMPLETE);
 
 	/*
-	 * for a REQ_BLOCK_PC request, we want to carry any eventual
+	 * for a REQ_TYPE_BLOCK_PC request, we want to carry any eventual
 	 * sense key with us all the way through
 	 */
 	if (!blk_pc_request(req))
@@ -1552,11 +1649,14 @@ static int __end_that_request_first(struct request *req, int error,
 	}
 
 	if (blk_fs_request(req) && req->rq_disk) {
-		struct hd_struct *part = get_part(req->rq_disk, req->sector);
 		const int rw = rq_data_dir(req);
+		struct hd_struct *part;
+		int cpu;
 
-		all_stat_add(req->rq_disk, part, sectors[rw],
-				nr_bytes >> 9, req->sector);
+		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();
 	}
 
 	total_bytes = bio_nbytes = 0;
@@ -1641,88 +1741,14 @@ static int __end_that_request_first(struct request *req, int error,
 }
 
 /*
- * splice the completion data to a local structure and hand off to
- * process_completion_queue() to complete the requests
- */
-static void blk_done_softirq(struct softirq_action *h)
-{
-	struct list_head *cpu_list, local_list;
-
-	local_irq_disable();
-	cpu_list = &__get_cpu_var(blk_cpu_done);
-	list_replace_init(cpu_list, &local_list);
-	local_irq_enable();
-
-	while (!list_empty(&local_list)) {
-		struct request *rq;
-
-		rq = list_entry(local_list.next, struct request, donelist);
-		list_del_init(&rq->donelist);
-		rq->q->softirq_done_fn(rq);
-	}
-}
-
-static int __cpuinit blk_cpu_notify(struct notifier_block *self,
-				    unsigned long action, void *hcpu)
-{
-	/*
-	 * If a CPU goes away, splice its entries to the current CPU
-	 * and trigger a run of the softirq
-	 */
-	if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) {
-		int cpu = (unsigned long) hcpu;
-
-		local_irq_disable();
-		list_splice_init(&per_cpu(blk_cpu_done, cpu),
-				 &__get_cpu_var(blk_cpu_done));
-		raise_softirq_irqoff(BLOCK_SOFTIRQ);
-		local_irq_enable();
-	}
-
-	return NOTIFY_OK;
-}
-
-
-static struct notifier_block blk_cpu_notifier __cpuinitdata = {
-	.notifier_call	= blk_cpu_notify,
-};
-
-/**
- * blk_complete_request - end I/O on a request
- * @req:      the request being processed
- *
- * Description:
- *     Ends all I/O on a request. It does not handle partial completions,
- *     unless the driver actually implements this in its completion callback
- *     through requeueing. The actual completion happens out-of-order,
- *     through a softirq handler. The user must have registered a completion
- *     callback through blk_queue_softirq_done().
- **/
-
-void blk_complete_request(struct request *req)
-{
-	struct list_head *cpu_list;
-	unsigned long flags;
-
-	BUG_ON(!req->q->softirq_done_fn);
-
-	local_irq_save(flags);
-
-	cpu_list = &__get_cpu_var(blk_cpu_done);
-	list_add_tail(&req->donelist, cpu_list);
-	raise_softirq_irqoff(BLOCK_SOFTIRQ);
-
-	local_irq_restore(flags);
-}
-EXPORT_SYMBOL(blk_complete_request);
-
-/*
  * queue lock must be held
  */
 static void end_that_request_last(struct request *req, int error)
 {
 	struct gendisk *disk = req->rq_disk;
 
+	blk_delete_timer(req);
+
 	if (blk_rq_tagged(req))
 		blk_queue_end_tag(req->q, req);
 
@@ -1740,16 +1766,18 @@ static void end_that_request_last(struct request *req, int error)
 	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 = get_part(disk, req->sector);
-
-		__all_stat_inc(disk, part, ios[rw], req->sector);
-		__all_stat_add(disk, part, ticks[rw], duration, req->sector);
-		disk_round_stats(disk);
-		disk->in_flight--;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight--;
-		}
+		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();
 	}
 
 	if (req->end_io)
@@ -1762,17 +1790,6 @@ static void end_that_request_last(struct request *req, int error)
 	}
 }
 
-static inline void __end_request(struct request *rq, int uptodate,
-				 unsigned int nr_bytes)
-{
-	int error = 0;
-
-	if (uptodate <= 0)
-		error = uptodate ? uptodate : -EIO;
-
-	__blk_end_request(rq, error, nr_bytes);
-}
-
 /**
  * blk_rq_bytes - Returns bytes left to complete in the entire request
  * @rq: the request being processed
@@ -1803,74 +1820,57 @@ unsigned int blk_rq_cur_bytes(struct request *rq)
 EXPORT_SYMBOL_GPL(blk_rq_cur_bytes);
 
 /**
- * end_queued_request - end all I/O on a queued request
- * @rq:		the request being processed
- * @uptodate:	error value or 0/1 uptodate flag
- *
- * Description:
- *     Ends all I/O on a request, and removes it from the block layer queues.
- *     Not suitable for normal IO completion, unless the driver still has
- *     the request attached to the block layer.
- *
- **/
-void end_queued_request(struct request *rq, int uptodate)
-{
-	__end_request(rq, uptodate, blk_rq_bytes(rq));
-}
-EXPORT_SYMBOL(end_queued_request);
-
-/**
- * end_dequeued_request - end all I/O on a dequeued request
- * @rq:		the request being processed
- * @uptodate:	error value or 0/1 uptodate flag
- *
- * Description:
- *     Ends all I/O on a request. The request must already have been
- *     dequeued using blkdev_dequeue_request(), as is normally the case
- *     for most drivers.
- *
- **/
-void end_dequeued_request(struct request *rq, int uptodate)
-{
-	__end_request(rq, uptodate, blk_rq_bytes(rq));
-}
-EXPORT_SYMBOL(end_dequeued_request);
-
-
-/**
  * end_request - end I/O on the current segment of the request
  * @req:	the request being processed
- * @uptodate:	error value or 0/1 uptodate flag
+ * @uptodate:	error value or %0/%1 uptodate flag
  *
  * Description:
  *     Ends I/O on the current segment of a request. If that is the only
  *     remaining segment, the request is also completed and freed.
  *
- *     This is a remnant of how older block drivers handled IO completions.
- *     Modern drivers typically end IO on the full request in one go, unless
+ *     This is a remnant of how older block drivers handled I/O completions.
+ *     Modern drivers typically end I/O on the full request in one go, unless
  *     they have a residual value to account for. For that case this function
  *     isn't really useful, unless the residual just happens to be the
  *     full current segment. In other words, don't use this function in new
- *     code. Either use end_request_completely(), or the
- *     end_that_request_chunk() (along with end_that_request_last()) for
- *     partial completions.
- *
+ *     code. Use blk_end_request() or __blk_end_request() to end a request.
  **/
 void end_request(struct request *req, int uptodate)
 {
-	__end_request(req, uptodate, req->hard_cur_sectors << 9);
+	int error = 0;
+
+	if (uptodate <= 0)
+		error = uptodate ? uptodate : -EIO;
+
+	__blk_end_request(req, error, req->hard_cur_sectors << 9);
 }
 EXPORT_SYMBOL(end_request);
 
+static int end_that_request_data(struct request *rq, int error,
+				 unsigned int nr_bytes, unsigned int bidi_bytes)
+{
+	if (rq->bio) {
+		if (__end_that_request_first(rq, error, nr_bytes))
+			return 1;
+
+		/* Bidi request must be completed as a whole */
+		if (blk_bidi_rq(rq) &&
+		    __end_that_request_first(rq->next_rq, error, bidi_bytes))
+			return 1;
+	}
+
+	return 0;
+}
+
 /**
  * blk_end_io - Generic end_io function to complete a request.
  * @rq:           the request being processed
- * @error:        0 for success, < 0 for error
+ * @error:        %0 for success, < %0 for error
  * @nr_bytes:     number of bytes to complete @rq
  * @bidi_bytes:   number of bytes to complete @rq->next_rq
  * @drv_callback: function called between completion of bios in the request
  *                and completion of the request.
- *                If the callback returns non 0, this helper returns without
+ *                If the callback returns non %0, this helper returns without
  *                completion of the request.
  *
  * Description:
@@ -1878,8 +1878,8 @@ EXPORT_SYMBOL(end_request);
  *     If @rq has leftover, sets it up for the next range of segments.
  *
  * Return:
- *     0 - we are done with this request
- *     1 - this request is not freed yet, it still has pending buffers.
+ *     %0 - we are done with this request
+ *     %1 - this request is not freed yet, it still has pending buffers.
  **/
 static int blk_end_io(struct request *rq, int error, unsigned int nr_bytes,
 		      unsigned int bidi_bytes,
@@ -1888,15 +1888,8 @@ static int blk_end_io(struct request *rq, int error, unsigned int nr_bytes,
 	struct request_queue *q = rq->q;
 	unsigned long flags = 0UL;
 
-	if (blk_fs_request(rq) || blk_pc_request(rq)) {
-		if (__end_that_request_first(rq, error, nr_bytes))
-			return 1;
-
-		/* Bidi request must be completed as a whole */
-		if (blk_bidi_rq(rq) &&
-		    __end_that_request_first(rq->next_rq, error, bidi_bytes))
-			return 1;
-	}
+	if (end_that_request_data(rq, error, nr_bytes, bidi_bytes))
+		return 1;
 
 	/* Special feature for tricky drivers */
 	if (drv_callback && drv_callback(rq))
@@ -1914,7 +1907,7 @@ static int blk_end_io(struct request *rq, int error, unsigned int nr_bytes,
 /**
  * blk_end_request - Helper function for drivers to complete the request.
  * @rq:       the request being processed
- * @error:    0 for success, < 0 for error
+ * @error:    %0 for success, < %0 for error
  * @nr_bytes: number of bytes to complete
  *
  * Description:
@@ -1922,8 +1915,8 @@ static int blk_end_io(struct request *rq, int error, unsigned int nr_bytes,
  *     If @rq has leftover, sets it up for the next range of segments.
  *
  * Return:
- *     0 - we are done with this request
- *     1 - still buffers pending for this request
+ *     %0 - we are done with this request
+ *     %1 - still buffers pending for this request
  **/
 int blk_end_request(struct request *rq, int error, unsigned int nr_bytes)
 {
@@ -1934,22 +1927,20 @@ EXPORT_SYMBOL_GPL(blk_end_request);
 /**
  * __blk_end_request - Helper function for drivers to complete the request.
  * @rq:       the request being processed
- * @error:    0 for success, < 0 for error
+ * @error:    %0 for success, < %0 for error
  * @nr_bytes: number of bytes to complete
  *
  * Description:
  *     Must be called with queue lock held unlike blk_end_request().
  *
  * Return:
- *     0 - we are done with this request
- *     1 - still buffers pending for this request
+ *     %0 - we are done with this request
+ *     %1 - still buffers pending for this request
  **/
 int __blk_end_request(struct request *rq, int error, unsigned int nr_bytes)
 {
-	if (blk_fs_request(rq) || blk_pc_request(rq)) {
-		if (__end_that_request_first(rq, error, nr_bytes))
-			return 1;
-	}
+	if (rq->bio && __end_that_request_first(rq, error, nr_bytes))
+		return 1;
 
 	add_disk_randomness(rq->rq_disk);
 
@@ -1962,7 +1953,7 @@ EXPORT_SYMBOL_GPL(__blk_end_request);
 /**
  * blk_end_bidi_request - Helper function for drivers to complete bidi request.
  * @rq:         the bidi request being processed
- * @error:      0 for success, < 0 for error
+ * @error:      %0 for success, < %0 for error
  * @nr_bytes:   number of bytes to complete @rq
  * @bidi_bytes: number of bytes to complete @rq->next_rq
  *
@@ -1970,8 +1961,8 @@ EXPORT_SYMBOL_GPL(__blk_end_request);
  *     Ends I/O on a number of bytes attached to @rq and @rq->next_rq.
  *
  * Return:
- *     0 - we are done with this request
- *     1 - still buffers pending for this request
+ *     %0 - we are done with this request
+ *     %1 - still buffers pending for this request
  **/
 int blk_end_bidi_request(struct request *rq, int error, unsigned int nr_bytes,
 			 unsigned int bidi_bytes)
@@ -1981,13 +1972,43 @@ int blk_end_bidi_request(struct request *rq, int error, unsigned int nr_bytes,
 EXPORT_SYMBOL_GPL(blk_end_bidi_request);
 
 /**
+ * blk_update_request - Special helper function for request stacking drivers
+ * @rq:           the request being processed
+ * @error:        %0 for success, < %0 for error
+ * @nr_bytes:     number of bytes to complete @rq
+ *
+ * Description:
+ *     Ends I/O on a number of bytes attached to @rq, but doesn't complete
+ *     the request structure even if @rq doesn't have leftover.
+ *     If @rq has leftover, sets it up for the next range of segments.
+ *
+ *     This special helper function is only for request stacking drivers
+ *     (e.g. request-based dm) so that they can handle partial completion.
+ *     Actual device drivers should use blk_end_request instead.
+ */
+void blk_update_request(struct request *rq, int error, unsigned int nr_bytes)
+{
+	if (!end_that_request_data(rq, error, nr_bytes, 0)) {
+		/*
+		 * These members are not updated in end_that_request_data()
+		 * when all bios are completed.
+		 * Update them so that the request stacking driver can find
+		 * how many bytes remain in the request later.
+		 */
+		rq->nr_sectors = rq->hard_nr_sectors = 0;
+		rq->current_nr_sectors = rq->hard_cur_sectors = 0;
+	}
+}
+EXPORT_SYMBOL_GPL(blk_update_request);
+
+/**
  * blk_end_request_callback - Special helper function for tricky drivers
  * @rq:           the request being processed
- * @error:        0 for success, < 0 for error
+ * @error:        %0 for success, < %0 for error
  * @nr_bytes:     number of bytes to complete
  * @drv_callback: function called between completion of bios in the request
  *                and completion of the request.
- *                If the callback returns non 0, this helper returns without
+ *                If the callback returns non %0, this helper returns without
  *                completion of the request.
  *
  * Description:
@@ -2000,10 +2021,10 @@ EXPORT_SYMBOL_GPL(blk_end_bidi_request);
  *     Don't use this interface in other places anymore.
  *
  * Return:
- *     0 - we are done with this request
- *     1 - this request is not freed yet.
- *         this request still has pending buffers or
- *         the driver doesn't want to finish this request yet.
+ *     %0 - we are done with this request
+ *     %1 - this request is not freed yet.
+ *          this request still has pending buffers or
+ *          the driver doesn't want to finish this request yet.
  **/
 int blk_end_request_callback(struct request *rq, int error,
 			     unsigned int nr_bytes,
@@ -2016,15 +2037,17 @@ EXPORT_SYMBOL_GPL(blk_end_request_callback);
 void blk_rq_bio_prep(struct request_queue *q, struct request *rq,
 		     struct bio *bio)
 {
-	/* first two bits are identical in rq->cmd_flags and bio->bi_rw */
+	/* Bit 0 (R/W) is identical in rq->cmd_flags and bio->bi_rw, and
+	   we want BIO_RW_AHEAD (bit 1) to imply REQ_FAILFAST (bit 1). */
 	rq->cmd_flags |= (bio->bi_rw & 3);
 
-	rq->nr_phys_segments = bio_phys_segments(q, bio);
-	rq->nr_hw_segments = bio_hw_segments(q, bio);
+	if (bio_has_data(bio)) {
+		rq->nr_phys_segments = bio_phys_segments(q, bio);
+		rq->buffer = bio_data(bio);
+	}
 	rq->current_nr_sectors = bio_cur_sectors(bio);
 	rq->hard_cur_sectors = rq->current_nr_sectors;
 	rq->hard_nr_sectors = rq->nr_sectors = bio_sectors(bio);
-	rq->buffer = bio_data(bio);
 	rq->data_len = bio->bi_size;
 
 	rq->bio = rq->biotail = bio;
@@ -2033,7 +2056,35 @@ void blk_rq_bio_prep(struct request_queue *q, struct request *rq,
 		rq->rq_disk = bio->bi_bdev->bd_disk;
 }
 
-int kblockd_schedule_work(struct work_struct *work)
+/**
+ * blk_lld_busy - Check if underlying low-level drivers of a device are busy
+ * @q : the queue of the device being checked
+ *
+ * Description:
+ *    Check if underlying low-level drivers of a device are busy.
+ *    If the drivers want to export their busy state, they must set own
+ *    exporting function using blk_queue_lld_busy() first.
+ *
+ *    Basically, this function is used only by request stacking drivers
+ *    to stop dispatching requests to underlying devices when underlying
+ *    devices are busy.  This behavior helps more I/O merging on the queue
+ *    of the request stacking driver and prevents I/O throughput regression
+ *    on burst I/O load.
+ *
+ * Return:
+ *    0 - Not busy (The request stacking driver should dispatch request)
+ *    1 - Busy (The request stacking driver should stop dispatching request)
+ */
+int blk_lld_busy(struct request_queue *q)
+{
+	if (q->lld_busy_fn)
+		return q->lld_busy_fn(q);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(blk_lld_busy);
+
+int kblockd_schedule_work(struct request_queue *q, struct work_struct *work)
 {
 	return queue_work(kblockd_workqueue, work);
 }
@@ -2047,8 +2098,6 @@ EXPORT_SYMBOL(kblockd_flush_work);
 
 int __init blk_dev_init(void)
 {
-	int i;
-
 	kblockd_workqueue = create_workqueue("kblockd");
 	if (!kblockd_workqueue)
 		panic("Failed to create kblockd\n");
@@ -2059,12 +2108,6 @@ int __init blk_dev_init(void)
 	blk_requestq_cachep = kmem_cache_create("blkdev_queue",
 			sizeof(struct request_queue), 0, SLAB_PANIC, NULL);
 
-	for_each_possible_cpu(i)
-		INIT_LIST_HEAD(&per_cpu(blk_cpu_done, i));
-
-	open_softirq(BLOCK_SOFTIRQ, blk_done_softirq);
-	register_hotcpu_notifier(&blk_cpu_notifier);
-
 	return 0;
 }
 
diff --git a/block/blk-exec.c b/block/blk-exec.c
index 9bceff7674f..6af716d1e54 100644
--- a/block/blk-exec.c
+++ b/block/blk-exec.c
@@ -16,7 +16,7 @@
 /**
  * blk_end_sync_rq - executes a completion event on a request
  * @rq: request to complete
- * @error: end io status of the request
+ * @error: end I/O status of the request
  */
 static void blk_end_sync_rq(struct request *rq, int error)
 {
@@ -41,7 +41,7 @@ static void blk_end_sync_rq(struct request *rq, int error)
  * @done:	I/O completion handler
  *
  * Description:
- *    Insert a fully prepared request at the back of the io scheduler queue
+ *    Insert a fully prepared request at the back of the I/O scheduler queue
  *    for execution.  Don't wait for completion.
  */
 void blk_execute_rq_nowait(struct request_queue *q, struct gendisk *bd_disk,
@@ -72,7 +72,7 @@ EXPORT_SYMBOL_GPL(blk_execute_rq_nowait);
  * @at_head:    insert request at head or tail of queue
  *
  * Description:
- *    Insert a fully prepared request at the back of the io scheduler queue
+ *    Insert a fully prepared request at the back of the I/O scheduler queue
  *    for execution and wait for completion.
  */
 int blk_execute_rq(struct request_queue *q, struct gendisk *bd_disk,
diff --git a/block/blk-integrity.c b/block/blk-integrity.c
index 3f1a8478cc3..61a8e2f8fdd 100644
--- a/block/blk-integrity.c
+++ b/block/blk-integrity.c
@@ -108,51 +108,51 @@ new_segment:
 EXPORT_SYMBOL(blk_rq_map_integrity_sg);
 
 /**
- * blk_integrity_compare - Compare integrity profile of two block devices
- * @b1:		Device to compare
- * @b2:		Device to compare
+ * blk_integrity_compare - Compare integrity profile of two disks
+ * @gd1:	Disk to compare
+ * @gd2:	Disk to compare
  *
  * Description: Meta-devices like DM and MD need to verify that all
  * sub-devices use the same integrity format before advertising to
  * upper layers that they can send/receive integrity metadata.  This
- * function can be used to check whether two block devices have
+ * function can be used to check whether two gendisk devices have
  * compatible integrity formats.
  */
-int blk_integrity_compare(struct block_device *bd1, struct block_device *bd2)
+int blk_integrity_compare(struct gendisk *gd1, struct gendisk *gd2)
 {
-	struct blk_integrity *b1 = bd1->bd_disk->integrity;
-	struct blk_integrity *b2 = bd2->bd_disk->integrity;
+	struct blk_integrity *b1 = gd1->integrity;
+	struct blk_integrity *b2 = gd2->integrity;
 
-	BUG_ON(bd1->bd_disk == NULL);
-	BUG_ON(bd2->bd_disk == NULL);
+	if (!b1 && !b2)
+		return 0;
 
 	if (!b1 || !b2)
-		return 0;
+		return -1;
 
 	if (b1->sector_size != b2->sector_size) {
 		printk(KERN_ERR "%s: %s/%s sector sz %u != %u\n", __func__,
-		       bd1->bd_disk->disk_name, bd2->bd_disk->disk_name,
+		       gd1->disk_name, gd2->disk_name,
 		       b1->sector_size, b2->sector_size);
 		return -1;
 	}
 
 	if (b1->tuple_size != b2->tuple_size) {
 		printk(KERN_ERR "%s: %s/%s tuple sz %u != %u\n", __func__,
-		       bd1->bd_disk->disk_name, bd2->bd_disk->disk_name,
+		       gd1->disk_name, gd2->disk_name,
 		       b1->tuple_size, b2->tuple_size);
 		return -1;
 	}
 
 	if (b1->tag_size && b2->tag_size && (b1->tag_size != b2->tag_size)) {
 		printk(KERN_ERR "%s: %s/%s tag sz %u != %u\n", __func__,
-		       bd1->bd_disk->disk_name, bd2->bd_disk->disk_name,
+		       gd1->disk_name, gd2->disk_name,
 		       b1->tag_size, b2->tag_size);
 		return -1;
 	}
 
 	if (strcmp(b1->name, b2->name)) {
 		printk(KERN_ERR "%s: %s/%s type %s != %s\n", __func__,
-		       bd1->bd_disk->disk_name, bd2->bd_disk->disk_name,
+		       gd1->disk_name, gd2->disk_name,
 		       b1->name, b2->name);
 		return -1;
 	}
@@ -331,7 +331,8 @@ int blk_integrity_register(struct gendisk *disk, struct blk_integrity *template)
 			return -1;
 
 		if (kobject_init_and_add(&bi->kobj, &integrity_ktype,
-					 &disk->dev.kobj, "%s", "integrity")) {
+					 &disk_to_dev(disk)->kobj,
+					 "%s", "integrity")) {
 			kmem_cache_free(integrity_cachep, bi);
 			return -1;
 		}
@@ -375,7 +376,7 @@ void blk_integrity_unregister(struct gendisk *disk)
 
 	kobject_uevent(&bi->kobj, KOBJ_REMOVE);
 	kobject_del(&bi->kobj);
-	kobject_put(&disk->dev.kobj);
 	kmem_cache_free(integrity_cachep, bi);
+	disk->integrity = NULL;
 }
 EXPORT_SYMBOL(blk_integrity_unregister);
diff --git a/block/blk-map.c b/block/blk-map.c
index af37e4ae62f..4849fa36161 100644
--- a/block/blk-map.c
+++ b/block/blk-map.c
@@ -41,10 +41,10 @@ static int __blk_rq_unmap_user(struct bio *bio)
 }
 
 static int __blk_rq_map_user(struct request_queue *q, struct request *rq,
-			     void __user *ubuf, unsigned int len)
+			     struct rq_map_data *map_data, void __user *ubuf,
+			     unsigned int len, int null_mapped, gfp_t gfp_mask)
 {
 	unsigned long uaddr;
-	unsigned int alignment;
 	struct bio *bio, *orig_bio;
 	int reading, ret;
 
@@ -55,15 +55,17 @@ static int __blk_rq_map_user(struct request_queue *q, struct request *rq,
 	 * direct dma. else, set up kernel bounce buffers
 	 */
 	uaddr = (unsigned long) ubuf;
-	alignment = queue_dma_alignment(q) | q->dma_pad_mask;
-	if (!(uaddr & alignment) && !(len & alignment))
-		bio = bio_map_user(q, NULL, uaddr, len, reading);
+	if (blk_rq_aligned(q, ubuf, len) && !map_data)
+		bio = bio_map_user(q, NULL, uaddr, len, reading, gfp_mask);
 	else
-		bio = bio_copy_user(q, uaddr, len, reading);
+		bio = bio_copy_user(q, map_data, uaddr, len, reading, gfp_mask);
 
 	if (IS_ERR(bio))
 		return PTR_ERR(bio);
 
+	if (null_mapped)
+		bio->bi_flags |= (1 << BIO_NULL_MAPPED);
+
 	orig_bio = bio;
 	blk_queue_bounce(q, &bio);
 
@@ -85,17 +87,19 @@ static int __blk_rq_map_user(struct request_queue *q, struct request *rq,
 }
 
 /**
- * blk_rq_map_user - map user data to a request, for REQ_BLOCK_PC usage
+ * blk_rq_map_user - map user data to a request, for REQ_TYPE_BLOCK_PC usage
  * @q:		request queue where request should be inserted
  * @rq:		request structure to fill
+ * @map_data:   pointer to the rq_map_data holding pages (if necessary)
  * @ubuf:	the user buffer
  * @len:	length of user data
+ * @gfp_mask:	memory allocation flags
  *
  * Description:
- *    Data will be mapped directly for zero copy io, if possible. Otherwise
+ *    Data will be mapped directly for zero copy I/O, if possible. Otherwise
  *    a kernel bounce buffer is used.
  *
- *    A matching blk_rq_unmap_user() must be issued at the end of io, while
+ *    A matching blk_rq_unmap_user() must be issued at the end of I/O, while
  *    still in process context.
  *
  *    Note: The mapped bio may need to be bounced through blk_queue_bounce()
@@ -105,16 +109,22 @@ static int __blk_rq_map_user(struct request_queue *q, struct request *rq,
  *    unmapping.
  */
 int blk_rq_map_user(struct request_queue *q, struct request *rq,
-		    void __user *ubuf, unsigned long len)
+		    struct rq_map_data *map_data, void __user *ubuf,
+		    unsigned long len, gfp_t gfp_mask)
 {
 	unsigned long bytes_read = 0;
 	struct bio *bio = NULL;
-	int ret;
+	int ret, null_mapped = 0;
 
 	if (len > (q->max_hw_sectors << 9))
 		return -EINVAL;
-	if (!len || !ubuf)
+	if (!len)
 		return -EINVAL;
+	if (!ubuf) {
+		if (!map_data || rq_data_dir(rq) != READ)
+			return -EINVAL;
+		null_mapped = 1;
+	}
 
 	while (bytes_read != len) {
 		unsigned long map_len, end, start;
@@ -132,7 +142,8 @@ int blk_rq_map_user(struct request_queue *q, struct request *rq,
 		if (end - start > BIO_MAX_PAGES)
 			map_len -= PAGE_SIZE;
 
-		ret = __blk_rq_map_user(q, rq, ubuf, map_len);
+		ret = __blk_rq_map_user(q, rq, map_data, ubuf, map_len,
+					null_mapped, gfp_mask);
 		if (ret < 0)
 			goto unmap_rq;
 		if (!bio)
@@ -154,18 +165,20 @@ unmap_rq:
 EXPORT_SYMBOL(blk_rq_map_user);
 
 /**
- * blk_rq_map_user_iov - map user data to a request, for REQ_BLOCK_PC usage
+ * blk_rq_map_user_iov - map user data to a request, for REQ_TYPE_BLOCK_PC usage
  * @q:		request queue where request should be inserted
  * @rq:		request to map data to
+ * @map_data:   pointer to the rq_map_data holding pages (if necessary)
  * @iov:	pointer to the iovec
  * @iov_count:	number of elements in the iovec
  * @len:	I/O byte count
+ * @gfp_mask:	memory allocation flags
  *
  * Description:
- *    Data will be mapped directly for zero copy io, if possible. Otherwise
+ *    Data will be mapped directly for zero copy I/O, if possible. Otherwise
  *    a kernel bounce buffer is used.
  *
- *    A matching blk_rq_unmap_user() must be issued at the end of io, while
+ *    A matching blk_rq_unmap_user() must be issued at the end of I/O, while
  *    still in process context.
  *
  *    Note: The mapped bio may need to be bounced through blk_queue_bounce()
@@ -175,7 +188,8 @@ EXPORT_SYMBOL(blk_rq_map_user);
  *    unmapping.
  */
 int blk_rq_map_user_iov(struct request_queue *q, struct request *rq,
-			struct sg_iovec *iov, int iov_count, unsigned int len)
+			struct rq_map_data *map_data, struct sg_iovec *iov,
+			int iov_count, unsigned int len, gfp_t gfp_mask)
 {
 	struct bio *bio;
 	int i, read = rq_data_dir(rq) == READ;
@@ -193,10 +207,11 @@ int blk_rq_map_user_iov(struct request_queue *q, struct request *rq,
 		}
 	}
 
-	if (unaligned || (q->dma_pad_mask & len))
-		bio = bio_copy_user_iov(q, iov, iov_count, read);
+	if (unaligned || (q->dma_pad_mask & len) || map_data)
+		bio = bio_copy_user_iov(q, map_data, iov, iov_count, read,
+					gfp_mask);
 	else
-		bio = bio_map_user_iov(q, NULL, iov, iov_count, read);
+		bio = bio_map_user_iov(q, NULL, iov, iov_count, read, gfp_mask);
 
 	if (IS_ERR(bio))
 		return PTR_ERR(bio);
@@ -216,6 +231,7 @@ int blk_rq_map_user_iov(struct request_queue *q, struct request *rq,
 	rq->buffer = rq->data = NULL;
 	return 0;
 }
+EXPORT_SYMBOL(blk_rq_map_user_iov);
 
 /**
  * blk_rq_unmap_user - unmap a request with user data
@@ -224,7 +240,7 @@ int blk_rq_map_user_iov(struct request_queue *q, struct request *rq,
  * Description:
  *    Unmap a rq previously mapped by blk_rq_map_user(). The caller must
  *    supply the original rq->bio from the blk_rq_map_user() return, since
- *    the io completion may have changed rq->bio.
+ *    the I/O completion may have changed rq->bio.
  */
 int blk_rq_unmap_user(struct bio *bio)
 {
@@ -250,7 +266,7 @@ int blk_rq_unmap_user(struct bio *bio)
 EXPORT_SYMBOL(blk_rq_unmap_user);
 
 /**
- * blk_rq_map_kern - map kernel data to a request, for REQ_BLOCK_PC usage
+ * blk_rq_map_kern - map kernel data to a request, for REQ_TYPE_BLOCK_PC usage
  * @q:		request queue where request should be inserted
  * @rq:		request to fill
  * @kbuf:	the kernel buffer
@@ -264,8 +280,6 @@ EXPORT_SYMBOL(blk_rq_unmap_user);
 int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf,
 		    unsigned int len, gfp_t gfp_mask)
 {
-	unsigned long kaddr;
-	unsigned int alignment;
 	int reading = rq_data_dir(rq) == READ;
 	int do_copy = 0;
 	struct bio *bio;
@@ -275,11 +289,7 @@ int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf,
 	if (!len || !kbuf)
 		return -EINVAL;
 
-	kaddr = (unsigned long)kbuf;
-	alignment = queue_dma_alignment(q) | q->dma_pad_mask;
-	do_copy = ((kaddr & alignment) || (len & alignment) ||
-		   object_is_on_stack(kbuf));
-
+	do_copy = !blk_rq_aligned(q, kbuf, len) || object_is_on_stack(kbuf);
 	if (do_copy)
 		bio = bio_copy_kern(q, kbuf, len, gfp_mask, reading);
 	else
diff --git a/block/blk-merge.c b/block/blk-merge.c
index 5efc9e7a68b..908d3e11ac5 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -11,7 +11,7 @@
 
 void blk_recalc_rq_sectors(struct request *rq, int nsect)
 {
-	if (blk_fs_request(rq)) {
+	if (blk_fs_request(rq) || blk_discard_rq(rq)) {
 		rq->hard_sector += nsect;
 		rq->hard_nr_sectors -= nsect;
 
@@ -41,12 +41,9 @@ void blk_recalc_rq_sectors(struct request *rq, int nsect)
 void blk_recalc_rq_segments(struct request *rq)
 {
 	int nr_phys_segs;
-	int nr_hw_segs;
 	unsigned int phys_size;
-	unsigned int hw_size;
 	struct bio_vec *bv, *bvprv = NULL;
 	int seg_size;
-	int hw_seg_size;
 	int cluster;
 	struct req_iterator iter;
 	int high, highprv = 1;
@@ -56,8 +53,8 @@ void blk_recalc_rq_segments(struct request *rq)
 		return;
 
 	cluster = test_bit(QUEUE_FLAG_CLUSTER, &q->queue_flags);
-	hw_seg_size = seg_size = 0;
-	phys_size = hw_size = nr_phys_segs = nr_hw_segs = 0;
+	seg_size = 0;
+	phys_size = nr_phys_segs = 0;
 	rq_for_each_segment(bv, rq, iter) {
 		/*
 		 * the trick here is making sure that a high page is never
@@ -66,7 +63,7 @@ void blk_recalc_rq_segments(struct request *rq)
 		 */
 		high = page_to_pfn(bv->bv_page) > q->bounce_pfn;
 		if (high || highprv)
-			goto new_hw_segment;
+			goto new_segment;
 		if (cluster) {
 			if (seg_size + bv->bv_len > q->max_segment_size)
 				goto new_segment;
@@ -74,40 +71,19 @@ void blk_recalc_rq_segments(struct request *rq)
 				goto new_segment;
 			if (!BIOVEC_SEG_BOUNDARY(q, bvprv, bv))
 				goto new_segment;
-			if (BIOVEC_VIRT_OVERSIZE(hw_seg_size + bv->bv_len))
-				goto new_hw_segment;
 
 			seg_size += bv->bv_len;
-			hw_seg_size += bv->bv_len;
 			bvprv = bv;
 			continue;
 		}
 new_segment:
-		if (BIOVEC_VIRT_MERGEABLE(bvprv, bv) &&
-		    !BIOVEC_VIRT_OVERSIZE(hw_seg_size + bv->bv_len))
-			hw_seg_size += bv->bv_len;
-		else {
-new_hw_segment:
-			if (nr_hw_segs == 1 &&
-			    hw_seg_size > rq->bio->bi_hw_front_size)
-				rq->bio->bi_hw_front_size = hw_seg_size;
-			hw_seg_size = BIOVEC_VIRT_START_SIZE(bv) + bv->bv_len;
-			nr_hw_segs++;
-		}
-
 		nr_phys_segs++;
 		bvprv = bv;
 		seg_size = bv->bv_len;
 		highprv = high;
 	}
 
-	if (nr_hw_segs == 1 &&
-	    hw_seg_size > rq->bio->bi_hw_front_size)
-		rq->bio->bi_hw_front_size = hw_seg_size;
-	if (hw_seg_size > rq->biotail->bi_hw_back_size)
-		rq->biotail->bi_hw_back_size = hw_seg_size;
 	rq->nr_phys_segments = nr_phys_segs;
-	rq->nr_hw_segments = nr_hw_segs;
 }
 
 void blk_recount_segments(struct request_queue *q, struct bio *bio)
@@ -120,7 +96,6 @@ void blk_recount_segments(struct request_queue *q, struct bio *bio)
 	blk_recalc_rq_segments(&rq);
 	bio->bi_next = nxt;
 	bio->bi_phys_segments = rq.nr_phys_segments;
-	bio->bi_hw_segments = rq.nr_hw_segments;
 	bio->bi_flags |= (1 << BIO_SEG_VALID);
 }
 EXPORT_SYMBOL(blk_recount_segments);
@@ -131,13 +106,17 @@ static int blk_phys_contig_segment(struct request_queue *q, struct bio *bio,
 	if (!test_bit(QUEUE_FLAG_CLUSTER, &q->queue_flags))
 		return 0;
 
-	if (!BIOVEC_PHYS_MERGEABLE(__BVEC_END(bio), __BVEC_START(nxt)))
-		return 0;
 	if (bio->bi_size + nxt->bi_size > q->max_segment_size)
 		return 0;
 
+	if (!bio_has_data(bio))
+		return 1;
+
+	if (!BIOVEC_PHYS_MERGEABLE(__BVEC_END(bio), __BVEC_START(nxt)))
+		return 0;
+
 	/*
-	 * bio and nxt are contigous in memory, check if the queue allows
+	 * bio and nxt are contiguous in memory; check if the queue allows
 	 * these two to be merged into one
 	 */
 	if (BIO_SEG_BOUNDARY(q, bio, nxt))
@@ -146,22 +125,6 @@ static int blk_phys_contig_segment(struct request_queue *q, struct bio *bio,
 	return 0;
 }
 
-static int blk_hw_contig_segment(struct request_queue *q, struct bio *bio,
-				 struct bio *nxt)
-{
-	if (!bio_flagged(bio, BIO_SEG_VALID))
-		blk_recount_segments(q, bio);
-	if (!bio_flagged(nxt, BIO_SEG_VALID))
-		blk_recount_segments(q, nxt);
-	if (!BIOVEC_VIRT_MERGEABLE(__BVEC_END(bio), __BVEC_START(nxt)) ||
-	    BIOVEC_VIRT_OVERSIZE(bio->bi_hw_back_size + nxt->bi_hw_front_size))
-		return 0;
-	if (bio->bi_hw_back_size + nxt->bi_hw_front_size > q->max_segment_size)
-		return 0;
-
-	return 1;
-}
-
 /*
  * map a request to scatterlist, return number of sg entries setup. Caller
  * must make sure sg can hold rq->nr_phys_segments entries
@@ -275,10 +238,9 @@ static inline int ll_new_hw_segment(struct request_queue *q,
 				    struct request *req,
 				    struct bio *bio)
 {
-	int nr_hw_segs = bio_hw_segments(q, bio);
 	int nr_phys_segs = bio_phys_segments(q, bio);
 
-	if (req->nr_hw_segments + nr_hw_segs > q->max_hw_segments
+	if (req->nr_phys_segments + nr_phys_segs > q->max_hw_segments
 	    || req->nr_phys_segments + nr_phys_segs > q->max_phys_segments) {
 		req->cmd_flags |= REQ_NOMERGE;
 		if (req == q->last_merge)
@@ -290,7 +252,6 @@ static inline int ll_new_hw_segment(struct request_queue *q,
 	 * This will form the start of a new hw segment.  Bump both
 	 * counters.
 	 */
-	req->nr_hw_segments += nr_hw_segs;
 	req->nr_phys_segments += nr_phys_segs;
 	return 1;
 }
@@ -299,7 +260,6 @@ int ll_back_merge_fn(struct request_queue *q, struct request *req,
 		     struct bio *bio)
 {
 	unsigned short max_sectors;
-	int len;
 
 	if (unlikely(blk_pc_request(req)))
 		max_sectors = q->max_hw_sectors;
@@ -316,19 +276,6 @@ int ll_back_merge_fn(struct request_queue *q, struct request *req,
 		blk_recount_segments(q, req->biotail);
 	if (!bio_flagged(bio, BIO_SEG_VALID))
 		blk_recount_segments(q, bio);
-	len = req->biotail->bi_hw_back_size + bio->bi_hw_front_size;
-	if (BIOVEC_VIRT_MERGEABLE(__BVEC_END(req->biotail), __BVEC_START(bio))
-	    && !BIOVEC_VIRT_OVERSIZE(len)) {
-		int mergeable =  ll_new_mergeable(q, req, bio);
-
-		if (mergeable) {
-			if (req->nr_hw_segments == 1)
-				req->bio->bi_hw_front_size = len;
-			if (bio->bi_hw_segments == 1)
-				bio->bi_hw_back_size = len;
-		}
-		return mergeable;
-	}
 
 	return ll_new_hw_segment(q, req, bio);
 }
@@ -337,7 +284,6 @@ int ll_front_merge_fn(struct request_queue *q, struct request *req,
 		      struct bio *bio)
 {
 	unsigned short max_sectors;
-	int len;
 
 	if (unlikely(blk_pc_request(req)))
 		max_sectors = q->max_hw_sectors;
@@ -351,23 +297,10 @@ int ll_front_merge_fn(struct request_queue *q, struct request *req,
 			q->last_merge = NULL;
 		return 0;
 	}
-	len = bio->bi_hw_back_size + req->bio->bi_hw_front_size;
 	if (!bio_flagged(bio, BIO_SEG_VALID))
 		blk_recount_segments(q, bio);
 	if (!bio_flagged(req->bio, BIO_SEG_VALID))
 		blk_recount_segments(q, req->bio);
-	if (BIOVEC_VIRT_MERGEABLE(__BVEC_END(bio), __BVEC_START(req->bio)) &&
-	    !BIOVEC_VIRT_OVERSIZE(len)) {
-		int mergeable =  ll_new_mergeable(q, req, bio);
-
-		if (mergeable) {
-			if (bio->bi_hw_segments == 1)
-				bio->bi_hw_front_size = len;
-			if (req->nr_hw_segments == 1)
-				req->biotail->bi_hw_back_size = len;
-		}
-		return mergeable;
-	}
 
 	return ll_new_hw_segment(q, req, bio);
 }
@@ -376,7 +309,6 @@ static int ll_merge_requests_fn(struct request_queue *q, struct request *req,
 				struct request *next)
 {
 	int total_phys_segments;
-	int total_hw_segments;
 
 	/*
 	 * First check if the either of the requests are re-queued
@@ -398,26 +330,11 @@ static int ll_merge_requests_fn(struct request_queue *q, struct request *req,
 	if (total_phys_segments > q->max_phys_segments)
 		return 0;
 
-	total_hw_segments = req->nr_hw_segments + next->nr_hw_segments;
-	if (blk_hw_contig_segment(q, req->biotail, next->bio)) {
-		int len = req->biotail->bi_hw_back_size +
-				next->bio->bi_hw_front_size;
-		/*
-		 * propagate the combined length to the end of the requests
-		 */
-		if (req->nr_hw_segments == 1)
-			req->bio->bi_hw_front_size = len;
-		if (next->nr_hw_segments == 1)
-			next->biotail->bi_hw_back_size = len;
-		total_hw_segments--;
-	}
-
-	if (total_hw_segments > q->max_hw_segments)
+	if (total_phys_segments > q->max_hw_segments)
 		return 0;
 
 	/* Merge is OK... */
 	req->nr_phys_segments = total_phys_segments;
-	req->nr_hw_segments = total_hw_segments;
 	return 1;
 }
 
@@ -470,17 +387,21 @@ static int attempt_merge(struct request_queue *q, struct request *req,
 	elv_merge_requests(q, req, next);
 
 	if (req->rq_disk) {
-		struct hd_struct *part
-			= get_part(req->rq_disk, req->sector);
-		disk_round_stats(req->rq_disk);
-		req->rq_disk->in_flight--;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight--;
-		}
+		struct hd_struct *part;
+		int cpu;
+
+		cpu = part_stat_lock();
+		part = disk_map_sector_rcu(req->rq_disk, req->sector);
+
+		part_round_stats(cpu, part);
+		part_dec_in_flight(part);
+
+		part_stat_unlock();
 	}
 
 	req->ioprio = ioprio_best(req->ioprio, next->ioprio);
+	if (blk_rq_cpu_valid(next))
+		req->cpu = next->cpu;
 
 	__blk_put_request(q, next);
 	return 1;
diff --git a/block/blk-settings.c b/block/blk-settings.c
index dfc77012843..b21dcdb6415 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -33,6 +33,23 @@ void blk_queue_prep_rq(struct request_queue *q, prep_rq_fn *pfn)
 EXPORT_SYMBOL(blk_queue_prep_rq);
 
 /**
+ * blk_queue_set_discard - set a discard_sectors function for queue
+ * @q:		queue
+ * @dfn:	prepare_discard function
+ *
+ * It's possible for a queue to register a discard callback which is used
+ * to transform a discard request into the appropriate type for the
+ * hardware. If none is registered, then discard requests are failed
+ * with %EOPNOTSUPP.
+ *
+ */
+void blk_queue_set_discard(struct request_queue *q, prepare_discard_fn *dfn)
+{
+	q->prepare_discard_fn = dfn;
+}
+EXPORT_SYMBOL(blk_queue_set_discard);
+
+/**
  * blk_queue_merge_bvec - set a merge_bvec function for queue
  * @q:		queue
  * @mbfn:	merge_bvec_fn
@@ -60,6 +77,24 @@ void blk_queue_softirq_done(struct request_queue *q, softirq_done_fn *fn)
 }
 EXPORT_SYMBOL(blk_queue_softirq_done);
 
+void blk_queue_rq_timeout(struct request_queue *q, unsigned int timeout)
+{
+	q->rq_timeout = timeout;
+}
+EXPORT_SYMBOL_GPL(blk_queue_rq_timeout);
+
+void blk_queue_rq_timed_out(struct request_queue *q, rq_timed_out_fn *fn)
+{
+	q->rq_timed_out_fn = fn;
+}
+EXPORT_SYMBOL_GPL(blk_queue_rq_timed_out);
+
+void blk_queue_lld_busy(struct request_queue *q, lld_busy_fn *fn)
+{
+	q->lld_busy_fn = fn;
+}
+EXPORT_SYMBOL_GPL(blk_queue_lld_busy);
+
 /**
  * blk_queue_make_request - define an alternate make_request function for a device
  * @q:  the request queue for the device to be affected
@@ -127,7 +162,7 @@ EXPORT_SYMBOL(blk_queue_make_request);
  *    Different hardware can have different requirements as to what pages
  *    it can do I/O directly to. A low level driver can call
  *    blk_queue_bounce_limit to have lower memory pages allocated as bounce
- *    buffers for doing I/O to pages residing above @page.
+ *    buffers for doing I/O to pages residing above @dma_addr.
  **/
 void blk_queue_bounce_limit(struct request_queue *q, u64 dma_addr)
 {
@@ -212,7 +247,7 @@ EXPORT_SYMBOL(blk_queue_max_phys_segments);
  * Description:
  *    Enables a low level driver to set an upper limit on the number of
  *    hw data segments in a request.  This would be the largest number of
- *    address/length pairs the host adapter can actually give as once
+ *    address/length pairs the host adapter can actually give at once
  *    to the device.
  **/
 void blk_queue_max_hw_segments(struct request_queue *q,
@@ -393,7 +428,7 @@ EXPORT_SYMBOL(blk_queue_segment_boundary);
  * @mask:  alignment mask
  *
  * description:
- *    set required memory and length aligment for direct dma transactions.
+ *    set required memory and length alignment for direct dma transactions.
  *    this is used when buiding direct io requests for the queue.
  *
  **/
@@ -409,7 +444,7 @@ EXPORT_SYMBOL(blk_queue_dma_alignment);
  * @mask:  alignment mask
  *
  * description:
- *    update required memory and length aligment for direct dma transactions.
+ *    update required memory and length alignment for direct dma transactions.
  *    If the requested alignment is larger than the current alignment, then
  *    the current queue alignment is updated to the new value, otherwise it
  *    is left alone.  The design of this is to allow multiple objects
diff --git a/block/blk-softirq.c b/block/blk-softirq.c
new file mode 100644
index 00000000000..e660d26ca65
--- /dev/null
+++ b/block/blk-softirq.c
@@ -0,0 +1,175 @@
+/*
+ * Functions related to softirq rq completions
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/bio.h>
+#include <linux/blkdev.h>
+#include <linux/interrupt.h>
+#include <linux/cpu.h>
+
+#include "blk.h"
+
+static DEFINE_PER_CPU(struct list_head, blk_cpu_done);
+
+/*
+ * Softirq action handler - move entries to local list and loop over them
+ * while passing them to the queue registered handler.
+ */
+static void blk_done_softirq(struct softirq_action *h)
+{
+	struct list_head *cpu_list, local_list;
+
+	local_irq_disable();
+	cpu_list = &__get_cpu_var(blk_cpu_done);
+	list_replace_init(cpu_list, &local_list);
+	local_irq_enable();
+
+	while (!list_empty(&local_list)) {
+		struct request *rq;
+
+		rq = list_entry(local_list.next, struct request, csd.list);
+		list_del_init(&rq->csd.list);
+		rq->q->softirq_done_fn(rq);
+	}
+}
+
+#if defined(CONFIG_SMP) && defined(CONFIG_USE_GENERIC_SMP_HELPERS)
+static void trigger_softirq(void *data)
+{
+	struct request *rq = data;
+	unsigned long flags;
+	struct list_head *list;
+
+	local_irq_save(flags);
+	list = &__get_cpu_var(blk_cpu_done);
+	list_add_tail(&rq->csd.list, list);
+
+	if (list->next == &rq->csd.list)
+		raise_softirq_irqoff(BLOCK_SOFTIRQ);
+
+	local_irq_restore(flags);
+}
+
+/*
+ * Setup and invoke a run of 'trigger_softirq' on the given cpu.
+ */
+static int raise_blk_irq(int cpu, struct request *rq)
+{
+	if (cpu_online(cpu)) {
+		struct call_single_data *data = &rq->csd;
+
+		data->func = trigger_softirq;
+		data->info = rq;
+		data->flags = 0;
+
+		__smp_call_function_single(cpu, data);
+		return 0;
+	}
+
+	return 1;
+}
+#else /* CONFIG_SMP && CONFIG_USE_GENERIC_SMP_HELPERS */
+static int raise_blk_irq(int cpu, struct request *rq)
+{
+	return 1;
+}
+#endif
+
+static int __cpuinit blk_cpu_notify(struct notifier_block *self,
+				    unsigned long action, void *hcpu)
+{
+	/*
+	 * If a CPU goes away, splice its entries to the current CPU
+	 * and trigger a run of the softirq
+	 */
+	if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) {
+		int cpu = (unsigned long) hcpu;
+
+		local_irq_disable();
+		list_splice_init(&per_cpu(blk_cpu_done, cpu),
+				 &__get_cpu_var(blk_cpu_done));
+		raise_softirq_irqoff(BLOCK_SOFTIRQ);
+		local_irq_enable();
+	}
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block __cpuinitdata blk_cpu_notifier = {
+	.notifier_call	= blk_cpu_notify,
+};
+
+void __blk_complete_request(struct request *req)
+{
+	struct request_queue *q = req->q;
+	unsigned long flags;
+	int ccpu, cpu, group_cpu;
+
+	BUG_ON(!q->softirq_done_fn);
+
+	local_irq_save(flags);
+	cpu = smp_processor_id();
+	group_cpu = blk_cpu_to_group(cpu);
+
+	/*
+	 * Select completion CPU
+	 */
+	if (test_bit(QUEUE_FLAG_SAME_COMP, &q->queue_flags) && req->cpu != -1)
+		ccpu = req->cpu;
+	else
+		ccpu = cpu;
+
+	if (ccpu == cpu || ccpu == group_cpu) {
+		struct list_head *list;
+do_local:
+		list = &__get_cpu_var(blk_cpu_done);
+		list_add_tail(&req->csd.list, list);
+
+		/*
+		 * if the list only contains our just added request,
+		 * signal a raise of the softirq. If there are already
+		 * entries there, someone already raised the irq but it
+		 * hasn't run yet.
+		 */
+		if (list->next == &req->csd.list)
+			raise_softirq_irqoff(BLOCK_SOFTIRQ);
+	} else if (raise_blk_irq(ccpu, req))
+		goto do_local;
+
+	local_irq_restore(flags);
+}
+
+/**
+ * blk_complete_request - end I/O on a request
+ * @req:      the request being processed
+ *
+ * Description:
+ *     Ends all I/O on a request. It does not handle partial completions,
+ *     unless the driver actually implements this in its completion callback
+ *     through requeueing. The actual completion happens out-of-order,
+ *     through a softirq handler. The user must have registered a completion
+ *     callback through blk_queue_softirq_done().
+ **/
+void blk_complete_request(struct request *req)
+{
+	if (unlikely(blk_should_fake_timeout(req->q)))
+		return;
+	if (!blk_mark_rq_complete(req))
+		__blk_complete_request(req);
+}
+EXPORT_SYMBOL(blk_complete_request);
+
+__init int blk_softirq_init(void)
+{
+	int i;
+
+	for_each_possible_cpu(i)
+		INIT_LIST_HEAD(&per_cpu(blk_cpu_done, i));
+
+	open_softirq(BLOCK_SOFTIRQ, blk_done_softirq);
+	register_hotcpu_notifier(&blk_cpu_notifier);
+	return 0;
+}
+subsys_initcall(blk_softirq_init);
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index 304ec73ab82..21e275d7eed 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -156,6 +156,30 @@ static ssize_t queue_nomerges_store(struct request_queue *q, const char *page,
 	return ret;
 }
 
+static ssize_t queue_rq_affinity_show(struct request_queue *q, char *page)
+{
+	unsigned int set = test_bit(QUEUE_FLAG_SAME_COMP, &q->queue_flags);
+
+	return queue_var_show(set != 0, page);
+}
+
+static ssize_t
+queue_rq_affinity_store(struct request_queue *q, const char *page, size_t count)
+{
+	ssize_t ret = -EINVAL;
+#if defined(CONFIG_USE_GENERIC_SMP_HELPERS)
+	unsigned long val;
+
+	ret = queue_var_store(&val, page, count);
+	spin_lock_irq(q->queue_lock);
+	if (val)
+		queue_flag_set(QUEUE_FLAG_SAME_COMP, q);
+	else
+		queue_flag_clear(QUEUE_FLAG_SAME_COMP,  q);
+	spin_unlock_irq(q->queue_lock);
+#endif
+	return ret;
+}
 
 static struct queue_sysfs_entry queue_requests_entry = {
 	.attr = {.name = "nr_requests", .mode = S_IRUGO | S_IWUSR },
@@ -197,6 +221,12 @@ static struct queue_sysfs_entry queue_nomerges_entry = {
 	.store = queue_nomerges_store,
 };
 
+static struct queue_sysfs_entry queue_rq_affinity_entry = {
+	.attr = {.name = "rq_affinity", .mode = S_IRUGO | S_IWUSR },
+	.show = queue_rq_affinity_show,
+	.store = queue_rq_affinity_store,
+};
+
 static struct attribute *default_attrs[] = {
 	&queue_requests_entry.attr,
 	&queue_ra_entry.attr,
@@ -205,6 +235,7 @@ static struct attribute *default_attrs[] = {
 	&queue_iosched_entry.attr,
 	&queue_hw_sector_size_entry.attr,
 	&queue_nomerges_entry.attr,
+	&queue_rq_affinity_entry.attr,
 	NULL,
 };
 
@@ -310,7 +341,7 @@ int blk_register_queue(struct gendisk *disk)
 	if (!q->request_fn)
 		return 0;
 
-	ret = kobject_add(&q->kobj, kobject_get(&disk->dev.kobj),
+	ret = kobject_add(&q->kobj, kobject_get(&disk_to_dev(disk)->kobj),
 			  "%s", "queue");
 	if (ret < 0)
 		return ret;
@@ -339,6 +370,6 @@ void blk_unregister_queue(struct gendisk *disk)
 
 		kobject_uevent(&q->kobj, KOBJ_REMOVE);
 		kobject_del(&q->kobj);
-		kobject_put(&disk->dev.kobj);
+		kobject_put(&disk_to_dev(disk)->kobj);
 	}
 }
diff --git a/block/blk-tag.c b/block/blk-tag.c
index ed5166fbc59..c0d419e84ce 100644
--- a/block/blk-tag.c
+++ b/block/blk-tag.c
@@ -29,7 +29,7 @@ EXPORT_SYMBOL(blk_queue_find_tag);
  * __blk_free_tags - release a given set of tag maintenance info
  * @bqt:	the tag map to free
  *
- * Tries to free the specified @bqt@.  Returns true if it was
+ * Tries to free the specified @bqt.  Returns true if it was
  * actually freed and false if there are still references using it
  */
 static int __blk_free_tags(struct blk_queue_tag *bqt)
@@ -78,7 +78,7 @@ void __blk_queue_free_tags(struct request_queue *q)
  * blk_free_tags - release a given set of tag maintenance info
  * @bqt:	the tag map to free
  *
- * For externally managed @bqt@ frees the map.  Callers of this
+ * For externally managed @bqt frees the map.  Callers of this
  * function must guarantee to have released all the queues that
  * might have been using this tag map.
  */
@@ -94,7 +94,7 @@ EXPORT_SYMBOL(blk_free_tags);
  * @q:  the request queue for the device
  *
  *  Notes:
- *	This is used to disabled tagged queuing to a device, yet leave
+ *	This is used to disable tagged queuing to a device, yet leave
  *	queue in function.
  **/
 void blk_queue_free_tags(struct request_queue *q)
@@ -271,7 +271,7 @@ EXPORT_SYMBOL(blk_queue_resize_tags);
  * @rq: the request that has completed
  *
  *  Description:
- *    Typically called when end_that_request_first() returns 0, meaning
+ *    Typically called when end_that_request_first() returns %0, meaning
  *    all transfers have been done for a request. It's important to call
  *    this function before end_that_request_last(), as that will put the
  *    request back on the free list thus corrupting the internal tag list.
@@ -337,6 +337,7 @@ EXPORT_SYMBOL(blk_queue_end_tag);
 int blk_queue_start_tag(struct request_queue *q, struct request *rq)
 {
 	struct blk_queue_tag *bqt = q->queue_tags;
+	unsigned max_depth, offset;
 	int tag;
 
 	if (unlikely((rq->cmd_flags & REQ_QUEUED))) {
@@ -350,10 +351,19 @@ int blk_queue_start_tag(struct request_queue *q, struct request *rq)
 	/*
 	 * Protect against shared tag maps, as we may not have exclusive
 	 * access to the tag map.
+	 *
+	 * We reserve a few tags just for sync IO, since we don't want
+	 * to starve sync IO on behalf of flooding async IO.
 	 */
+	max_depth = bqt->max_depth;
+	if (rq_is_sync(rq))
+		offset = 0;
+	else
+		offset = max_depth >> 2;
+
 	do {
-		tag = find_first_zero_bit(bqt->tag_map, bqt->max_depth);
-		if (tag >= bqt->max_depth)
+		tag = find_next_zero_bit(bqt->tag_map, max_depth, offset);
+		if (tag >= max_depth)
 			return 1;
 
 	} while (test_and_set_bit_lock(tag, bqt->tag_map));
diff --git a/block/blk-timeout.c b/block/blk-timeout.c
new file mode 100644
index 00000000000..972a63f848f
--- /dev/null
+++ b/block/blk-timeout.c
@@ -0,0 +1,238 @@
+/*
+ * Functions related to generic timeout handling of requests.
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/blkdev.h>
+#include <linux/fault-inject.h>
+
+#include "blk.h"
+
+#ifdef CONFIG_FAIL_IO_TIMEOUT
+
+static DECLARE_FAULT_ATTR(fail_io_timeout);
+
+static int __init setup_fail_io_timeout(char *str)
+{
+	return setup_fault_attr(&fail_io_timeout, str);
+}
+__setup("fail_io_timeout=", setup_fail_io_timeout);
+
+int blk_should_fake_timeout(struct request_queue *q)
+{
+	if (!test_bit(QUEUE_FLAG_FAIL_IO, &q->queue_flags))
+		return 0;
+
+	return should_fail(&fail_io_timeout, 1);
+}
+
+static int __init fail_io_timeout_debugfs(void)
+{
+	return init_fault_attr_dentries(&fail_io_timeout, "fail_io_timeout");
+}
+
+late_initcall(fail_io_timeout_debugfs);
+
+ssize_t part_timeout_show(struct device *dev, struct device_attribute *attr,
+			  char *buf)
+{
+	struct gendisk *disk = dev_to_disk(dev);
+	int set = test_bit(QUEUE_FLAG_FAIL_IO, &disk->queue->queue_flags);
+
+	return sprintf(buf, "%d\n", set != 0);
+}
+
+ssize_t part_timeout_store(struct device *dev, struct device_attribute *attr,
+			   const char *buf, size_t count)
+{
+	struct gendisk *disk = dev_to_disk(dev);
+	int val;
+
+	if (count) {
+		struct request_queue *q = disk->queue;
+		char *p = (char *) buf;
+
+		val = simple_strtoul(p, &p, 10);
+		spin_lock_irq(q->queue_lock);
+		if (val)
+			queue_flag_set(QUEUE_FLAG_FAIL_IO, q);
+		else
+			queue_flag_clear(QUEUE_FLAG_FAIL_IO, q);
+		spin_unlock_irq(q->queue_lock);
+	}
+
+	return count;
+}
+
+#endif /* CONFIG_FAIL_IO_TIMEOUT */
+
+/*
+ * blk_delete_timer - Delete/cancel timer for a given function.
+ * @req:	request that we are canceling timer for
+ *
+ */
+void blk_delete_timer(struct request *req)
+{
+	struct request_queue *q = req->q;
+
+	/*
+	 * Nothing to detach
+	 */
+	if (!q->rq_timed_out_fn || !req->deadline)
+		return;
+
+	list_del_init(&req->timeout_list);
+
+	if (list_empty(&q->timeout_list))
+		del_timer(&q->timeout);
+}
+
+static void blk_rq_timed_out(struct request *req)
+{
+	struct request_queue *q = req->q;
+	enum blk_eh_timer_return ret;
+
+	ret = q->rq_timed_out_fn(req);
+	switch (ret) {
+	case BLK_EH_HANDLED:
+		__blk_complete_request(req);
+		break;
+	case BLK_EH_RESET_TIMER:
+		blk_clear_rq_complete(req);
+		blk_add_timer(req);
+		break;
+	case BLK_EH_NOT_HANDLED:
+		/*
+		 * LLD handles this for now but in the future
+		 * we can send a request msg to abort the command
+		 * and we can move more of the generic scsi eh code to
+		 * the blk layer.
+		 */
+		break;
+	default:
+		printk(KERN_ERR "block: bad eh return: %d\n", ret);
+		break;
+	}
+}
+
+void blk_rq_timed_out_timer(unsigned long data)
+{
+	struct request_queue *q = (struct request_queue *) data;
+	unsigned long flags, uninitialized_var(next), next_set = 0;
+	struct request *rq, *tmp;
+
+	spin_lock_irqsave(q->queue_lock, flags);
+
+	list_for_each_entry_safe(rq, tmp, &q->timeout_list, timeout_list) {
+		if (time_after_eq(jiffies, rq->deadline)) {
+			list_del_init(&rq->timeout_list);
+
+			/*
+			 * Check if we raced with end io completion
+			 */
+			if (blk_mark_rq_complete(rq))
+				continue;
+			blk_rq_timed_out(rq);
+		}
+		if (!next_set) {
+			next = rq->deadline;
+			next_set = 1;
+		} else if (time_after(next, rq->deadline))
+			next = rq->deadline;
+	}
+
+	if (next_set && !list_empty(&q->timeout_list))
+		mod_timer(&q->timeout, round_jiffies(next));
+
+	spin_unlock_irqrestore(q->queue_lock, flags);
+}
+
+/**
+ * blk_abort_request -- Request request recovery for the specified command
+ * @req:	pointer to the request of interest
+ *
+ * This function requests that the block layer start recovery for the
+ * request by deleting the timer and calling the q's timeout function.
+ * LLDDs who implement their own error recovery MAY ignore the timeout
+ * event if they generated blk_abort_req. Must hold queue lock.
+ */
+void blk_abort_request(struct request *req)
+{
+	if (blk_mark_rq_complete(req))
+		return;
+	blk_delete_timer(req);
+	blk_rq_timed_out(req);
+}
+EXPORT_SYMBOL_GPL(blk_abort_request);
+
+/**
+ * blk_add_timer - Start timeout timer for a single request
+ * @req:	request that is about to start running.
+ *
+ * Notes:
+ *    Each request has its own timer, and as it is added to the queue, we
+ *    set up the timer. When the request completes, we cancel the timer.
+ */
+void blk_add_timer(struct request *req)
+{
+	struct request_queue *q = req->q;
+	unsigned long expiry;
+
+	if (!q->rq_timed_out_fn)
+		return;
+
+	BUG_ON(!list_empty(&req->timeout_list));
+	BUG_ON(test_bit(REQ_ATOM_COMPLETE, &req->atomic_flags));
+
+	if (req->timeout)
+		req->deadline = jiffies + req->timeout;
+	else {
+		req->deadline = jiffies + q->rq_timeout;
+		/*
+		 * Some LLDs, like scsi, peek at the timeout to prevent
+		 * a command from being retried forever.
+		 */
+		req->timeout = q->rq_timeout;
+	}
+	list_add_tail(&req->timeout_list, &q->timeout_list);
+
+	/*
+	 * If the timer isn't already pending or this timeout is earlier
+	 * than an existing one, modify the timer. Round to next nearest
+	 * second.
+	 */
+	expiry = round_jiffies(req->deadline);
+
+	/*
+	 * We use ->deadline == 0 to detect whether a timer was added or
+	 * not, so just increase to next jiffy for that specific case
+	 */
+	if (unlikely(!req->deadline))
+		req->deadline = 1;
+
+	if (!timer_pending(&q->timeout) ||
+	    time_before(expiry, q->timeout.expires))
+		mod_timer(&q->timeout, expiry);
+}
+
+/**
+ * blk_abort_queue -- Abort all request on given queue
+ * @queue:	pointer to queue
+ *
+ */
+void blk_abort_queue(struct request_queue *q)
+{
+	unsigned long flags;
+	struct request *rq, *tmp;
+
+	spin_lock_irqsave(q->queue_lock, flags);
+
+	elv_abort_queue(q);
+
+	list_for_each_entry_safe(rq, tmp, &q->timeout_list, timeout_list)
+		blk_abort_request(rq);
+
+	spin_unlock_irqrestore(q->queue_lock, flags);
+
+}
+EXPORT_SYMBOL_GPL(blk_abort_queue);
diff --git a/block/blk.h b/block/blk.h
index c79f30e1df5..e5c57976996 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -17,6 +17,42 @@ void __blk_queue_free_tags(struct request_queue *q);
 
 void blk_unplug_work(struct work_struct *work);
 void blk_unplug_timeout(unsigned long data);
+void blk_rq_timed_out_timer(unsigned long data);
+void blk_delete_timer(struct request *);
+void blk_add_timer(struct request *);
+
+/*
+ * Internal atomic flags for request handling
+ */
+enum rq_atomic_flags {
+	REQ_ATOM_COMPLETE = 0,
+};
+
+/*
+ * EH timer and IO completion will both attempt to 'grab' the request, make
+ * sure that only one of them suceeds
+ */
+static inline int blk_mark_rq_complete(struct request *rq)
+{
+	return test_and_set_bit(REQ_ATOM_COMPLETE, &rq->atomic_flags);
+}
+
+static inline void blk_clear_rq_complete(struct request *rq)
+{
+	clear_bit(REQ_ATOM_COMPLETE, &rq->atomic_flags);
+}
+
+#ifdef CONFIG_FAIL_IO_TIMEOUT
+int blk_should_fake_timeout(struct request_queue *);
+ssize_t part_timeout_show(struct device *, struct device_attribute *, char *);
+ssize_t part_timeout_store(struct device *, struct device_attribute *,
+				const char *, size_t);
+#else
+static inline int blk_should_fake_timeout(struct request_queue *q)
+{
+	return 0;
+}
+#endif
 
 struct io_context *current_io_context(gfp_t gfp_flags, int node);
 
@@ -59,4 +95,16 @@ static inline int queue_congestion_off_threshold(struct request_queue *q)
 
 #endif /* BLK_DEV_INTEGRITY */
 
+static inline int blk_cpu_to_group(int cpu)
+{
+#ifdef CONFIG_SCHED_MC
+	cpumask_t mask = cpu_coregroup_map(cpu);
+	return first_cpu(mask);
+#elif defined(CONFIG_SCHED_SMT)
+	return first_cpu(per_cpu(cpu_sibling_map, cpu));
+#else
+	return cpu;
+#endif
+}
+
 #endif
diff --git a/block/blktrace.c b/block/blktrace.c
index eb9651ccb24..85049a7e7a1 100644
--- a/block/blktrace.c
+++ b/block/blktrace.c
@@ -111,23 +111,9 @@ static int act_log_check(struct blk_trace *bt, u32 what, sector_t sector,
  */
 static u32 ddir_act[2] __read_mostly = { BLK_TC_ACT(BLK_TC_READ), BLK_TC_ACT(BLK_TC_WRITE) };
 
-/*
- * Bio action bits of interest
- */
-static u32 bio_act[9] __read_mostly = { 0, BLK_TC_ACT(BLK_TC_BARRIER), BLK_TC_ACT(BLK_TC_SYNC), 0, BLK_TC_ACT(BLK_TC_AHEAD), 0, 0, 0, BLK_TC_ACT(BLK_TC_META) };
-
-/*
- * More could be added as needed, taking care to increment the decrementer
- * to get correct indexing
- */
-#define trace_barrier_bit(rw)	\
-	(((rw) & (1 << BIO_RW_BARRIER)) >> (BIO_RW_BARRIER - 0))
-#define trace_sync_bit(rw)	\
-	(((rw) & (1 << BIO_RW_SYNC)) >> (BIO_RW_SYNC - 1))
-#define trace_ahead_bit(rw)	\
-	(((rw) & (1 << BIO_RW_AHEAD)) << (2 - BIO_RW_AHEAD))
-#define trace_meta_bit(rw)	\
-	(((rw) & (1 << BIO_RW_META)) >> (BIO_RW_META - 3))
+/* The ilog2() calls fall out because they're constant */
+#define MASK_TC_BIT(rw, __name) ( (rw & (1 << BIO_RW_ ## __name)) << \
+	  (ilog2(BLK_TC_ ## __name) + BLK_TC_SHIFT - BIO_RW_ ## __name) )
 
 /*
  * The worker for the various blk_add_trace*() types. Fills out a
@@ -147,10 +133,11 @@ void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
 		return;
 
 	what |= ddir_act[rw & WRITE];
-	what |= bio_act[trace_barrier_bit(rw)];
-	what |= bio_act[trace_sync_bit(rw)];
-	what |= bio_act[trace_ahead_bit(rw)];
-	what |= bio_act[trace_meta_bit(rw)];
+	what |= MASK_TC_BIT(rw, BARRIER);
+	what |= MASK_TC_BIT(rw, SYNC);
+	what |= MASK_TC_BIT(rw, AHEAD);
+	what |= MASK_TC_BIT(rw, META);
+	what |= MASK_TC_BIT(rw, DISCARD);
 
 	pid = tsk->pid;
 	if (unlikely(act_log_check(bt, what, sector, pid)))
@@ -382,7 +369,8 @@ int do_blk_trace_setup(struct request_queue *q, char *name, dev_t dev,
 	if (!buts->buf_size || !buts->buf_nr)
 		return -EINVAL;
 
-	strcpy(buts->name, name);
+	strncpy(buts->name, name, BLKTRACE_BDEV_SIZE);
+	buts->name[BLKTRACE_BDEV_SIZE - 1] = '\0';
 
 	/*
 	 * some device names have larger paths - convert the slashes
diff --git a/block/bsg.c b/block/bsg.c
index 0aae8d7ba99..56cb343c76d 100644
--- a/block/bsg.c
+++ b/block/bsg.c
@@ -283,7 +283,8 @@ bsg_map_hdr(struct bsg_device *bd, struct sg_io_v4 *hdr, int has_write_perm)
 		next_rq->cmd_type = rq->cmd_type;
 
 		dxferp = (void*)(unsigned long)hdr->din_xferp;
-		ret =  blk_rq_map_user(q, next_rq, dxferp, hdr->din_xfer_len);
+		ret =  blk_rq_map_user(q, next_rq, NULL, dxferp,
+				       hdr->din_xfer_len, GFP_KERNEL);
 		if (ret)
 			goto out;
 	}
@@ -298,7 +299,8 @@ bsg_map_hdr(struct bsg_device *bd, struct sg_io_v4 *hdr, int has_write_perm)
 		dxfer_len = 0;
 
 	if (dxfer_len) {
-		ret = blk_rq_map_user(q, rq, dxferp, dxfer_len);
+		ret = blk_rq_map_user(q, rq, NULL, dxferp, dxfer_len,
+				      GFP_KERNEL);
 		if (ret)
 			goto out;
 	}
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index 1e2aff812ee..6a062eebbd1 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -39,6 +39,7 @@ static int cfq_slice_idle = HZ / 125;
 #define CFQ_MIN_TT		(2)
 
 #define CFQ_SLICE_SCALE		(5)
+#define CFQ_HW_QUEUE_MIN	(5)
 
 #define RQ_CIC(rq)		\
 	((struct cfq_io_context *) (rq)->elevator_private)
@@ -86,7 +87,14 @@ struct cfq_data {
 
 	int rq_in_driver;
 	int sync_flight;
+
+	/*
+	 * queue-depth detection
+	 */
+	int rq_queued;
 	int hw_tag;
+	int hw_tag_samples;
+	int rq_in_driver_peak;
 
 	/*
 	 * idle window management
@@ -244,7 +252,7 @@ static inline void cfq_schedule_dispatch(struct cfq_data *cfqd)
 {
 	if (cfqd->busy_queues) {
 		cfq_log(cfqd, "schedule dispatch");
-		kblockd_schedule_work(&cfqd->unplug_work);
+		kblockd_schedule_work(cfqd->queue, &cfqd->unplug_work);
 	}
 }
 
@@ -654,15 +662,6 @@ static void cfq_activate_request(struct request_queue *q, struct request *rq)
 	cfq_log_cfqq(cfqd, RQ_CFQQ(rq), "activate rq, drv=%d",
 						cfqd->rq_in_driver);
 
-	/*
-	 * If the depth is larger 1, it really could be queueing. But lets
-	 * make the mark a little higher - idling could still be good for
-	 * low queueing, and a low queueing number could also just indicate
-	 * a SCSI mid layer like behaviour where limit+1 is often seen.
-	 */
-	if (!cfqd->hw_tag && cfqd->rq_in_driver > 4)
-		cfqd->hw_tag = 1;
-
 	cfqd->last_position = rq->hard_sector + rq->hard_nr_sectors;
 }
 
@@ -686,6 +685,7 @@ static void cfq_remove_request(struct request *rq)
 	list_del_init(&rq->queuelist);
 	cfq_del_rq_rb(rq);
 
+	cfqq->cfqd->rq_queued--;
 	if (rq_is_meta(rq)) {
 		WARN_ON(!cfqq->meta_pending);
 		cfqq->meta_pending--;
@@ -878,6 +878,14 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd)
 	struct cfq_io_context *cic;
 	unsigned long sl;
 
+	/*
+	 * SSD device without seek penalty, disable idling. But only do so
+	 * for devices that support queuing, otherwise we still have a problem
+	 * with sync vs async workloads.
+	 */
+	if (blk_queue_nonrot(cfqd->queue) && cfqd->hw_tag)
+		return;
+
 	WARN_ON(!RB_EMPTY_ROOT(&cfqq->sort_list));
 	WARN_ON(cfq_cfqq_slice_new(cfqq));
 
@@ -1833,6 +1841,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq,
 {
 	struct cfq_io_context *cic = RQ_CIC(rq);
 
+	cfqd->rq_queued++;
 	if (rq_is_meta(rq))
 		cfqq->meta_pending++;
 
@@ -1880,6 +1889,31 @@ static void cfq_insert_request(struct request_queue *q, struct request *rq)
 	cfq_rq_enqueued(cfqd, cfqq, rq);
 }
 
+/*
+ * Update hw_tag based on peak queue depth over 50 samples under
+ * sufficient load.
+ */
+static void cfq_update_hw_tag(struct cfq_data *cfqd)
+{
+	if (cfqd->rq_in_driver > cfqd->rq_in_driver_peak)
+		cfqd->rq_in_driver_peak = cfqd->rq_in_driver;
+
+	if (cfqd->rq_queued <= CFQ_HW_QUEUE_MIN &&
+	    cfqd->rq_in_driver <= CFQ_HW_QUEUE_MIN)
+		return;
+
+	if (cfqd->hw_tag_samples++ < 50)
+		return;
+
+	if (cfqd->rq_in_driver_peak >= CFQ_HW_QUEUE_MIN)
+		cfqd->hw_tag = 1;
+	else
+		cfqd->hw_tag = 0;
+
+	cfqd->hw_tag_samples = 0;
+	cfqd->rq_in_driver_peak = 0;
+}
+
 static void cfq_completed_request(struct request_queue *q, struct request *rq)
 {
 	struct cfq_queue *cfqq = RQ_CFQQ(rq);
@@ -1890,6 +1924,8 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq)
 	now = jiffies;
 	cfq_log_cfqq(cfqd, cfqq, "complete");
 
+	cfq_update_hw_tag(cfqd);
+
 	WARN_ON(!cfqd->rq_in_driver);
 	WARN_ON(!cfqq->dispatched);
 	cfqd->rq_in_driver--;
@@ -2200,6 +2236,7 @@ static void *cfq_init_queue(struct request_queue *q)
 	cfqd->cfq_slice[1] = cfq_slice_sync;
 	cfqd->cfq_slice_async_rq = cfq_slice_async_rq;
 	cfqd->cfq_slice_idle = cfq_slice_idle;
+	cfqd->hw_tag = 1;
 
 	return cfqd;
 }
diff --git a/block/cmd-filter.c b/block/cmd-filter.c
index 79c14996ac1..e669aed4c6b 100644
--- a/block/cmd-filter.c
+++ b/block/cmd-filter.c
@@ -211,14 +211,10 @@ int blk_register_filter(struct gendisk *disk)
 {
 	int ret;
 	struct blk_cmd_filter *filter = &disk->queue->cmd_filter;
-	struct kobject *parent = kobject_get(disk->holder_dir->parent);
 
-	if (!parent)
-		return -ENODEV;
-
-	ret = kobject_init_and_add(&filter->kobj, &rcf_ktype, parent,
+	ret = kobject_init_and_add(&filter->kobj, &rcf_ktype,
+				   &disk_to_dev(disk)->kobj,
 				   "%s", "cmd_filter");
-
 	if (ret < 0)
 		return ret;
 
@@ -231,7 +227,6 @@ void blk_unregister_filter(struct gendisk *disk)
 	struct blk_cmd_filter *filter = &disk->queue->cmd_filter;
 
 	kobject_put(&filter->kobj);
-	kobject_put(disk->holder_dir->parent);
 }
 EXPORT_SYMBOL(blk_unregister_filter);
 #endif
diff --git a/block/compat_ioctl.c b/block/compat_ioctl.c
index c23177e4623..1e559fba7bd 100644
--- a/block/compat_ioctl.c
+++ b/block/compat_ioctl.c
@@ -788,6 +788,7 @@ long compat_blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg)
 		return compat_hdio_getgeo(disk, bdev, compat_ptr(arg));
 	case BLKFLSBUF:
 	case BLKROSET:
+	case BLKDISCARD:
 	/*
 	 * the ones below are implemented in blkdev_locked_ioctl,
 	 * but we call blkdev_ioctl, which gets the lock for us
diff --git a/block/deadline-iosched.c b/block/deadline-iosched.c
index 342448c3d2d..fd311179f44 100644
--- a/block/deadline-iosched.c
+++ b/block/deadline-iosched.c
@@ -33,7 +33,7 @@ struct deadline_data {
 	 */
 	struct rb_root sort_list[2];	
 	struct list_head fifo_list[2];
-	
+
 	/*
 	 * next in sort order. read, write or both are NULL
 	 */
@@ -53,7 +53,11 @@ struct deadline_data {
 
 static void deadline_move_request(struct deadline_data *, struct request *);
 
-#define RQ_RB_ROOT(dd, rq)	(&(dd)->sort_list[rq_data_dir((rq))])
+static inline struct rb_root *
+deadline_rb_root(struct deadline_data *dd, struct request *rq)
+{
+	return &dd->sort_list[rq_data_dir(rq)];
+}
 
 /*
  * get the request after `rq' in sector-sorted order
@@ -72,15 +76,11 @@ deadline_latter_request(struct request *rq)
 static void
 deadline_add_rq_rb(struct deadline_data *dd, struct request *rq)
 {
-	struct rb_root *root = RQ_RB_ROOT(dd, rq);
+	struct rb_root *root = deadline_rb_root(dd, rq);
 	struct request *__alias;
 
-retry:
-	__alias = elv_rb_add(root, rq);
-	if (unlikely(__alias)) {
+	while (unlikely(__alias = elv_rb_add(root, rq)))
 		deadline_move_request(dd, __alias);
-		goto retry;
-	}
 }
 
 static inline void
@@ -91,7 +91,7 @@ deadline_del_rq_rb(struct deadline_data *dd, struct request *rq)
 	if (dd->next_rq[data_dir] == rq)
 		dd->next_rq[data_dir] = deadline_latter_request(rq);
 
-	elv_rb_del(RQ_RB_ROOT(dd, rq), rq);
+	elv_rb_del(deadline_rb_root(dd, rq), rq);
 }
 
 /*
@@ -106,7 +106,7 @@ deadline_add_request(struct request_queue *q, struct request *rq)
 	deadline_add_rq_rb(dd, rq);
 
 	/*
-	 * set expire time (only used for reads) and add to fifo list
+	 * set expire time and add to fifo list
 	 */
 	rq_set_fifo_time(rq, jiffies + dd->fifo_expire[data_dir]);
 	list_add_tail(&rq->queuelist, &dd->fifo_list[data_dir]);
@@ -162,7 +162,7 @@ static void deadline_merged_request(struct request_queue *q,
 	 * if the merge was a front merge, we need to reposition request
 	 */
 	if (type == ELEVATOR_FRONT_MERGE) {
-		elv_rb_del(RQ_RB_ROOT(dd, req), req);
+		elv_rb_del(deadline_rb_root(dd, req), req);
 		deadline_add_rq_rb(dd, req);
 	}
 }
@@ -212,7 +212,7 @@ deadline_move_request(struct deadline_data *dd, struct request *rq)
 	dd->next_rq[WRITE] = NULL;
 	dd->next_rq[data_dir] = deadline_latter_request(rq);
 
-	dd->last_sector = rq->sector + rq->nr_sectors;
+	dd->last_sector = rq_end_sector(rq);
 
 	/*
 	 * take it off the sort and fifo list, move
@@ -222,7 +222,7 @@ deadline_move_request(struct deadline_data *dd, struct request *rq)
 }
 
 /*
- * deadline_check_fifo returns 0 if there are no expired reads on the fifo,
+ * deadline_check_fifo returns 0 if there are no expired requests on the fifo,
  * 1 otherwise. Requires !list_empty(&dd->fifo_list[data_dir])
  */
 static inline int deadline_check_fifo(struct deadline_data *dd, int ddir)
@@ -258,17 +258,9 @@ static int deadline_dispatch_requests(struct request_queue *q, int force)
 	else
 		rq = dd->next_rq[READ];
 
-	if (rq) {
-		/* we have a "next request" */
-		
-		if (dd->last_sector != rq->sector)
-			/* end the batch on a non sequential request */
-			dd->batching += dd->fifo_batch;
-		
-		if (dd->batching < dd->fifo_batch)
-			/* we are still entitled to batch */
-			goto dispatch_request;
-	}
+	if (rq && dd->batching < dd->fifo_batch)
+		/* we have a next request are still entitled to batch */
+		goto dispatch_request;
 
 	/*
 	 * at this point we are not running a batch. select the appropriate
diff --git a/block/elevator.c b/block/elevator.c
index ed6f8f32d27..04518921db3 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -34,8 +34,9 @@
 #include <linux/delay.h>
 #include <linux/blktrace_api.h>
 #include <linux/hash.h>
+#include <linux/uaccess.h>
 
-#include <asm/uaccess.h>
+#include "blk.h"
 
 static DEFINE_SPINLOCK(elv_list_lock);
 static LIST_HEAD(elv_list);
@@ -75,6 +76,12 @@ int elv_rq_merge_ok(struct request *rq, struct bio *bio)
 		return 0;
 
 	/*
+	 * Don't merge file system requests and discard requests
+	 */
+	if (bio_discard(bio) != bio_discard(rq->bio))
+		return 0;
+
+	/*
 	 * different data direction or already started, don't merge
 	 */
 	if (bio_data_dir(bio) != rq_data_dir(rq))
@@ -438,6 +445,8 @@ void elv_dispatch_sort(struct request_queue *q, struct request *rq)
 	list_for_each_prev(entry, &q->queue_head) {
 		struct request *pos = list_entry_rq(entry);
 
+		if (blk_discard_rq(rq) != blk_discard_rq(pos))
+			break;
 		if (rq_data_dir(rq) != rq_data_dir(pos))
 			break;
 		if (pos->cmd_flags & stop_flags)
@@ -607,7 +616,7 @@ void elv_insert(struct request_queue *q, struct request *rq, int where)
 		break;
 
 	case ELEVATOR_INSERT_SORT:
-		BUG_ON(!blk_fs_request(rq));
+		BUG_ON(!blk_fs_request(rq) && !blk_discard_rq(rq));
 		rq->cmd_flags |= REQ_SORTED;
 		q->nr_sorted++;
 		if (rq_mergeable(rq)) {
@@ -692,7 +701,7 @@ void __elv_add_request(struct request_queue *q, struct request *rq, int where,
 		 * this request is scheduling boundary, update
 		 * end_sector
 		 */
-		if (blk_fs_request(rq)) {
+		if (blk_fs_request(rq) || blk_discard_rq(rq)) {
 			q->end_sector = rq_end_sector(rq);
 			q->boundary_rq = rq;
 		}
@@ -745,7 +754,7 @@ struct request *elv_next_request(struct request_queue *q)
 		 * not ever see it.
 		 */
 		if (blk_empty_barrier(rq)) {
-			end_queued_request(rq, 1);
+			__blk_end_request(rq, 0, blk_rq_bytes(rq));
 			continue;
 		}
 		if (!(rq->cmd_flags & REQ_STARTED)) {
@@ -764,6 +773,12 @@ struct request *elv_next_request(struct request_queue *q)
 			 */
 			rq->cmd_flags |= REQ_STARTED;
 			blk_add_trace_rq(q, rq, BLK_TA_ISSUE);
+
+			/*
+			 * We are now handing the request to the hardware,
+			 * add the timeout handler
+			 */
+			blk_add_timer(rq);
 		}
 
 		if (!q->boundary_rq || q->boundary_rq == rq) {
@@ -782,7 +797,6 @@ struct request *elv_next_request(struct request_queue *q)
 			 * device can handle
 			 */
 			rq->nr_phys_segments++;
-			rq->nr_hw_segments++;
 		}
 
 		if (!q->prep_rq_fn)
@@ -805,14 +819,13 @@ struct request *elv_next_request(struct request_queue *q)
 				 * so that we don't add it again
 				 */
 				--rq->nr_phys_segments;
-				--rq->nr_hw_segments;
 			}
 
 			rq = NULL;
 			break;
 		} else if (ret == BLKPREP_KILL) {
 			rq->cmd_flags |= REQ_QUIET;
-			end_queued_request(rq, 0);
+			__blk_end_request(rq, -EIO, blk_rq_bytes(rq));
 		} else {
 			printk(KERN_ERR "%s: bad return=%d\n", __func__, ret);
 			break;
@@ -901,6 +914,19 @@ int elv_may_queue(struct request_queue *q, int rw)
 	return ELV_MQUEUE_MAY;
 }
 
+void elv_abort_queue(struct request_queue *q)
+{
+	struct request *rq;
+
+	while (!list_empty(&q->queue_head)) {
+		rq = list_entry_rq(q->queue_head.next);
+		rq->cmd_flags |= REQ_QUIET;
+		blk_add_trace_rq(q, rq, BLK_TA_ABORT);
+		__blk_end_request(rq, -EIO, blk_rq_bytes(rq));
+	}
+}
+EXPORT_SYMBOL(elv_abort_queue);
+
 void elv_completed_request(struct request_queue *q, struct request *rq)
 {
 	elevator_t *e = q->elevator;
diff --git a/block/genhd.c b/block/genhd.c
index e0ce23ac2ec..4cd3433c99a 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -16,6 +16,7 @@
 #include <linux/kobj_map.h>
 #include <linux/buffer_head.h>
 #include <linux/mutex.h>
+#include <linux/idr.h>
 
 #include "blk.h"
 
@@ -24,8 +25,194 @@ static DEFINE_MUTEX(block_class_lock);
 struct kobject *block_depr;
 #endif
 
+/* for extended dynamic devt allocation, currently only one major is used */
+#define MAX_EXT_DEVT		(1 << MINORBITS)
+
+/* For extended devt allocation.  ext_devt_mutex prevents look up
+ * results from going away underneath its user.
+ */
+static DEFINE_MUTEX(ext_devt_mutex);
+static DEFINE_IDR(ext_devt_idr);
+
 static struct device_type disk_type;
 
+/**
+ * disk_get_part - get partition
+ * @disk: disk to look partition from
+ * @partno: partition number
+ *
+ * Look for partition @partno from @disk.  If found, increment
+ * reference count and return it.
+ *
+ * CONTEXT:
+ * Don't care.
+ *
+ * RETURNS:
+ * Pointer to the found partition on success, NULL if not found.
+ */
+struct hd_struct *disk_get_part(struct gendisk *disk, int partno)
+{
+	struct hd_struct *part = NULL;
+	struct disk_part_tbl *ptbl;
+
+	if (unlikely(partno < 0))
+		return NULL;
+
+	rcu_read_lock();
+
+	ptbl = rcu_dereference(disk->part_tbl);
+	if (likely(partno < ptbl->len)) {
+		part = rcu_dereference(ptbl->part[partno]);
+		if (part)
+			get_device(part_to_dev(part));
+	}
+
+	rcu_read_unlock();
+
+	return part;
+}
+EXPORT_SYMBOL_GPL(disk_get_part);
+
+/**
+ * disk_part_iter_init - initialize partition iterator
+ * @piter: iterator to initialize
+ * @disk: disk to iterate over
+ * @flags: DISK_PITER_* flags
+ *
+ * Initialize @piter so that it iterates over partitions of @disk.
+ *
+ * CONTEXT:
+ * Don't care.
+ */
+void disk_part_iter_init(struct disk_part_iter *piter, struct gendisk *disk,
+			  unsigned int flags)
+{
+	struct disk_part_tbl *ptbl;
+
+	rcu_read_lock();
+	ptbl = rcu_dereference(disk->part_tbl);
+
+	piter->disk = disk;
+	piter->part = NULL;
+
+	if (flags & DISK_PITER_REVERSE)
+		piter->idx = ptbl->len - 1;
+	else if (flags & DISK_PITER_INCL_PART0)
+		piter->idx = 0;
+	else
+		piter->idx = 1;
+
+	piter->flags = flags;
+
+	rcu_read_unlock();
+}
+EXPORT_SYMBOL_GPL(disk_part_iter_init);
+
+/**
+ * disk_part_iter_next - proceed iterator to the next partition and return it
+ * @piter: iterator of interest
+ *
+ * Proceed @piter to the next partition and return it.
+ *
+ * CONTEXT:
+ * Don't care.
+ */
+struct hd_struct *disk_part_iter_next(struct disk_part_iter *piter)
+{
+	struct disk_part_tbl *ptbl;
+	int inc, end;
+
+	/* put the last partition */
+	disk_put_part(piter->part);
+	piter->part = NULL;
+
+	/* get part_tbl */
+	rcu_read_lock();
+	ptbl = rcu_dereference(piter->disk->part_tbl);
+
+	/* determine iteration parameters */
+	if (piter->flags & DISK_PITER_REVERSE) {
+		inc = -1;
+		if (piter->flags & DISK_PITER_INCL_PART0)
+			end = -1;
+		else
+			end = 0;
+	} else {
+		inc = 1;
+		end = ptbl->len;
+	}
+
+	/* iterate to the next partition */
+	for (; piter->idx != end; piter->idx += inc) {
+		struct hd_struct *part;
+
+		part = rcu_dereference(ptbl->part[piter->idx]);
+		if (!part)
+			continue;
+		if (!(piter->flags & DISK_PITER_INCL_EMPTY) && !part->nr_sects)
+			continue;
+
+		get_device(part_to_dev(part));
+		piter->part = part;
+		piter->idx += inc;
+		break;
+	}
+
+	rcu_read_unlock();
+
+	return piter->part;
+}
+EXPORT_SYMBOL_GPL(disk_part_iter_next);
+
+/**
+ * disk_part_iter_exit - finish up partition iteration
+ * @piter: iter of interest
+ *
+ * Called when iteration is over.  Cleans up @piter.
+ *
+ * CONTEXT:
+ * Don't care.
+ */
+void disk_part_iter_exit(struct disk_part_iter *piter)
+{
+	disk_put_part(piter->part);
+	piter->part = NULL;
+}
+EXPORT_SYMBOL_GPL(disk_part_iter_exit);
+
+/**
+ * disk_map_sector_rcu - map sector to partition
+ * @disk: gendisk of interest
+ * @sector: sector to map
+ *
+ * Find out which partition @sector maps to on @disk.  This is
+ * primarily used for stats accounting.
+ *
+ * CONTEXT:
+ * RCU read locked.  The returned partition pointer is valid only
+ * while preemption is disabled.
+ *
+ * RETURNS:
+ * Found partition on success, part0 is returned if no partition matches
+ */
+struct hd_struct *disk_map_sector_rcu(struct gendisk *disk, sector_t sector)
+{
+	struct disk_part_tbl *ptbl;
+	int i;
+
+	ptbl = rcu_dereference(disk->part_tbl);
+
+	for (i = 1; i < ptbl->len; i++) {
+		struct hd_struct *part = rcu_dereference(ptbl->part[i]);
+
+		if (part && part->start_sect <= sector &&
+		    sector < part->start_sect + part->nr_sects)
+			return part;
+	}
+	return &disk->part0;
+}
+EXPORT_SYMBOL_GPL(disk_map_sector_rcu);
+
 /*
  * Can be deleted altogether. Later.
  *
@@ -43,14 +230,14 @@ static inline int major_to_index(int major)
 }
 
 #ifdef CONFIG_PROC_FS
-void blkdev_show(struct seq_file *f, off_t offset)
+void blkdev_show(struct seq_file *seqf, off_t offset)
 {
 	struct blk_major_name *dp;
 
 	if (offset < BLKDEV_MAJOR_HASH_SIZE) {
 		mutex_lock(&block_class_lock);
 		for (dp = major_names[offset]; dp; dp = dp->next)
-			seq_printf(f, "%3d %s\n", dp->major, dp->name);
+			seq_printf(seqf, "%3d %s\n", dp->major, dp->name);
 		mutex_unlock(&block_class_lock);
 	}
 }
@@ -136,6 +323,118 @@ EXPORT_SYMBOL(unregister_blkdev);
 
 static struct kobj_map *bdev_map;
 
+/**
+ * blk_mangle_minor - scatter minor numbers apart
+ * @minor: minor number to mangle
+ *
+ * Scatter consecutively allocated @minor number apart if MANGLE_DEVT
+ * is enabled.  Mangling twice gives the original value.
+ *
+ * RETURNS:
+ * Mangled value.
+ *
+ * CONTEXT:
+ * Don't care.
+ */
+static int blk_mangle_minor(int minor)
+{
+#ifdef CONFIG_DEBUG_BLOCK_EXT_DEVT
+	int i;
+
+	for (i = 0; i < MINORBITS / 2; i++) {
+		int low = minor & (1 << i);
+		int high = minor & (1 << (MINORBITS - 1 - i));
+		int distance = MINORBITS - 1 - 2 * i;
+
+		minor ^= low | high;	/* clear both bits */
+		low <<= distance;	/* swap the positions */
+		high >>= distance;
+		minor |= low | high;	/* and set */
+	}
+#endif
+	return minor;
+}
+
+/**
+ * blk_alloc_devt - allocate a dev_t for a partition
+ * @part: partition to allocate dev_t for
+ * @gfp_mask: memory allocation flag
+ * @devt: out parameter for resulting dev_t
+ *
+ * Allocate a dev_t for block device.
+ *
+ * RETURNS:
+ * 0 on success, allocated dev_t is returned in *@devt.  -errno on
+ * failure.
+ *
+ * CONTEXT:
+ * Might sleep.
+ */
+int blk_alloc_devt(struct hd_struct *part, dev_t *devt)
+{
+	struct gendisk *disk = part_to_disk(part);
+	int idx, rc;
+
+	/* in consecutive minor range? */
+	if (part->partno < disk->minors) {
+		*devt = MKDEV(disk->major, disk->first_minor + part->partno);
+		return 0;
+	}
+
+	/* allocate ext devt */
+	do {
+		if (!idr_pre_get(&ext_devt_idr, GFP_KERNEL))
+			return -ENOMEM;
+		rc = idr_get_new(&ext_devt_idr, part, &idx);
+	} while (rc == -EAGAIN);
+
+	if (rc)
+		return rc;
+
+	if (idx > MAX_EXT_DEVT) {
+		idr_remove(&ext_devt_idr, idx);
+		return -EBUSY;
+	}
+
+	*devt = MKDEV(BLOCK_EXT_MAJOR, blk_mangle_minor(idx));
+	return 0;
+}
+
+/**
+ * blk_free_devt - free a dev_t
+ * @devt: dev_t to free
+ *
+ * Free @devt which was allocated using blk_alloc_devt().
+ *
+ * CONTEXT:
+ * Might sleep.
+ */
+void blk_free_devt(dev_t devt)
+{
+	might_sleep();
+
+	if (devt == MKDEV(0, 0))
+		return;
+
+	if (MAJOR(devt) == BLOCK_EXT_MAJOR) {
+		mutex_lock(&ext_devt_mutex);
+		idr_remove(&ext_devt_idr, blk_mangle_minor(MINOR(devt)));
+		mutex_unlock(&ext_devt_mutex);
+	}
+}
+
+static char *bdevt_str(dev_t devt, char *buf)
+{
+	if (MAJOR(devt) <= 0xff && MINOR(devt) <= 0xff) {
+		char tbuf[BDEVT_SIZE];
+		snprintf(tbuf, BDEVT_SIZE, "%02x%02x", MAJOR(devt), MINOR(devt));
+		snprintf(buf, BDEVT_SIZE, "%-9s", tbuf);
+	} else
+		snprintf(buf, BDEVT_SIZE, "%03x:%05x", MAJOR(devt), MINOR(devt));
+
+	return buf;
+}
+
 /*
  * Register device numbers dev..(dev+range-1)
  * range must be nonzero
@@ -157,11 +456,11 @@ void blk_unregister_region(dev_t devt, unsigned long range)
 
 EXPORT_SYMBOL(blk_unregister_region);
 
-static struct kobject *exact_match(dev_t devt, int *part, void *data)
+static struct kobject *exact_match(dev_t devt, int *partno, void *data)
 {
 	struct gendisk *p = data;
 
-	return &p->dev.kobj;
+	return &disk_to_dev(p)->kobj;
 }
 
 static int exact_lock(dev_t devt, void *data)
@@ -179,21 +478,46 @@ static int exact_lock(dev_t devt, void *data)
  *
  * This function registers the partitioning information in @disk
  * with the kernel.
+ *
+ * FIXME: error handling
  */
 void add_disk(struct gendisk *disk)
 {
 	struct backing_dev_info *bdi;
+	dev_t devt;
 	int retval;
 
+	/* minors == 0 indicates to use ext devt from part0 and should
+	 * be accompanied with EXT_DEVT flag.  Make sure all
+	 * parameters make sense.
+	 */
+	WARN_ON(disk->minors && !(disk->major || disk->first_minor));
+	WARN_ON(!disk->minors && !(disk->flags & GENHD_FL_EXT_DEVT));
+
 	disk->flags |= GENHD_FL_UP;
-	blk_register_region(MKDEV(disk->major, disk->first_minor),
-			    disk->minors, NULL, exact_match, exact_lock, disk);
+
+	retval = blk_alloc_devt(&disk->part0, &devt);
+	if (retval) {
+		WARN_ON(1);
+		return;
+	}
+	disk_to_dev(disk)->devt = devt;
+
+	/* ->major and ->first_minor aren't supposed to be
+	 * dereferenced from here on, but set them just in case.
+	 */
+	disk->major = MAJOR(devt);
+	disk->first_minor = MINOR(devt);
+
+	blk_register_region(disk_devt(disk), disk->minors, NULL,
+			    exact_match, exact_lock, disk);
 	register_disk(disk);
 	blk_register_queue(disk);
 
 	bdi = &disk->queue->backing_dev_info;
-	bdi_register_dev(bdi, MKDEV(disk->major, disk->first_minor));
-	retval = sysfs_create_link(&disk->dev.kobj, &bdi->dev->kobj, "bdi");
+	bdi_register_dev(bdi, disk_devt(disk));
+	retval = sysfs_create_link(&disk_to_dev(disk)->kobj, &bdi->dev->kobj,
+				   "bdi");
 	WARN_ON(retval);
 }
 
@@ -202,78 +526,71 @@ EXPORT_SYMBOL(del_gendisk);	/* in partitions/check.c */
 
 void unlink_gendisk(struct gendisk *disk)
 {
-	sysfs_remove_link(&disk->dev.kobj, "bdi");
+	sysfs_remove_link(&disk_to_dev(disk)->kobj, "bdi");
 	bdi_unregister(&disk->queue->backing_dev_info);
 	blk_unregister_queue(disk);
-	blk_unregister_region(MKDEV(disk->major, disk->first_minor),
-			      disk->minors);
+	blk_unregister_region(disk_devt(disk), disk->minors);
 }
 
 /**
  * get_gendisk - get partitioning information for a given device
- * @dev: device to get partitioning information for
+ * @devt: device to get partitioning information for
+ * @part: returned partition index
  *
  * This function gets the structure containing partitioning
- * information for the given device @dev.
+ * information for the given device @devt.
  */
-struct gendisk *get_gendisk(dev_t devt, int *part)
+struct gendisk *get_gendisk(dev_t devt, int *partno)
 {
-	struct kobject *kobj = kobj_lookup(bdev_map, devt, part);
-	struct device *dev = kobj_to_dev(kobj);
+	struct gendisk *disk = NULL;
+
+	if (MAJOR(devt) != BLOCK_EXT_MAJOR) {
+		struct kobject *kobj;
+
+		kobj = kobj_lookup(bdev_map, devt, partno);
+		if (kobj)
+			disk = dev_to_disk(kobj_to_dev(kobj));
+	} else {
+		struct hd_struct *part;
 
-	return  kobj ? dev_to_disk(dev) : NULL;
+		mutex_lock(&ext_devt_mutex);
+		part = idr_find(&ext_devt_idr, blk_mangle_minor(MINOR(devt)));
+		if (part && get_disk(part_to_disk(part))) {
+			*partno = part->partno;
+			disk = part_to_disk(part);
+		}
+		mutex_unlock(&ext_devt_mutex);
+	}
+
+	return disk;
 }
 
-/*
- * print a partitions - intended for places where the root filesystem can't be
- * mounted and thus to give the victim some idea of what went wrong
+/**
+ * bdget_disk - do bdget() by gendisk and partition number
+ * @disk: gendisk of interest
+ * @partno: partition number
+ *
+ * Find partition @partno from @disk, do bdget() on it.
+ *
+ * CONTEXT:
+ * Don't care.
+ *
+ * RETURNS:
+ * Resulting block_device on success, NULL on failure.
  */
-static int printk_partition(struct device *dev, void *data)
+struct block_device *bdget_disk(struct gendisk *disk, int partno)
 {
-	struct gendisk *sgp;
-	char buf[BDEVNAME_SIZE];
-	int n;
-
-	if (dev->type != &disk_type)
-		goto exit;
+	struct hd_struct *part;
+	struct block_device *bdev = NULL;
 
-	sgp = dev_to_disk(dev);
-	/*
-	 * Don't show empty devices or things that have been surpressed
-	 */
-	if (get_capacity(sgp) == 0 ||
-	    (sgp->flags & GENHD_FL_SUPPRESS_PARTITION_INFO))
-		goto exit;
+	part = disk_get_part(disk, partno);
+	if (part)
+		bdev = bdget(part_devt(part));
+	disk_put_part(part);
 
-	/*
-	 * Note, unlike /proc/partitions, I am showing the numbers in
-	 * hex - the same format as the root= option takes.
-	 */
-	printk("%02x%02x %10llu %s",
-		sgp->major, sgp->first_minor,
-		(unsigned long long)get_capacity(sgp) >> 1,
-		disk_name(sgp, 0, buf));
-	if (sgp->driverfs_dev != NULL &&
-	    sgp->driverfs_dev->driver != NULL)
-		printk(" driver: %s\n",
-			sgp->driverfs_dev->driver->name);
-	else
-		printk(" (driver?)\n");
-
-	/* now show the partitions */
-	for (n = 0; n < sgp->minors - 1; ++n) {
-		if (sgp->part[n] == NULL)
-			goto exit;
-		if (sgp->part[n]->nr_sects == 0)
-			goto exit;
-		printk("  %02x%02x %10llu %s\n",
-			sgp->major, n + 1 + sgp->first_minor,
-			(unsigned long long)sgp->part[n]->nr_sects >> 1,
-			disk_name(sgp, n + 1, buf));
-	}
-exit:
-	return 0;
+	return bdev;
 }
+EXPORT_SYMBOL(bdget_disk);
 
 /*
  * print a full list of all partitions - intended for places where the root
@@ -282,120 +599,145 @@ exit:
  */
 void __init printk_all_partitions(void)
 {
-	mutex_lock(&block_class_lock);
-	class_for_each_device(&block_class, NULL, NULL, printk_partition);
-	mutex_unlock(&block_class_lock);
+	struct class_dev_iter iter;
+	struct device *dev;
+
+	class_dev_iter_init(&iter, &block_class, NULL, &disk_type);
+	while ((dev = class_dev_iter_next(&iter))) {
+		struct gendisk *disk = dev_to_disk(dev);
+		struct disk_part_iter piter;
+		struct hd_struct *part;
+		char name_buf[BDEVNAME_SIZE];
+		char devt_buf[BDEVT_SIZE];
+
+		/*
+		 * Don't show empty devices or things that have been
+		 * surpressed
+		 */
+		if (get_capacity(disk) == 0 ||
+		    (disk->flags & GENHD_FL_SUPPRESS_PARTITION_INFO))
+			continue;
+
+		/*
+		 * Note, unlike /proc/partitions, I am showing the
+		 * numbers in hex - the same format as the root=
+		 * option takes.
+		 */
+		disk_part_iter_init(&piter, disk, DISK_PITER_INCL_PART0);
+		while ((part = disk_part_iter_next(&piter))) {
+			bool is_part0 = part == &disk->part0;
+
+			printk("%s%s %10llu %s", is_part0 ? "" : "  ",
+			       bdevt_str(part_devt(part), devt_buf),
+			       (unsigned long long)part->nr_sects >> 1,
+			       disk_name(disk, part->partno, name_buf));
+			if (is_part0) {
+				if (disk->driverfs_dev != NULL &&
+				    disk->driverfs_dev->driver != NULL)
+					printk(" driver: %s\n",
+					      disk->driverfs_dev->driver->name);
+				else
+					printk(" (driver?)\n");
+			} else
+				printk("\n");
+		}
+		disk_part_iter_exit(&piter);
+	}
+	class_dev_iter_exit(&iter);
 }
 
 #ifdef CONFIG_PROC_FS
 /* iterator */
-static int find_start(struct device *dev, void *data)
+static void *disk_seqf_start(struct seq_file *seqf, loff_t *pos)
 {
-	loff_t *k = data;
+	loff_t skip = *pos;
+	struct class_dev_iter *iter;
+	struct device *dev;
 
-	if (dev->type != &disk_type)
-		return 0;
-	if (!*k)
-		return 1;
-	(*k)--;
-	return 0;
+	iter = kmalloc(sizeof(*iter), GFP_KERNEL);
+	if (!iter)
+		return ERR_PTR(-ENOMEM);
+
+	seqf->private = iter;
+	class_dev_iter_init(iter, &block_class, NULL, &disk_type);
+	do {
+		dev = class_dev_iter_next(iter);
+		if (!dev)
+			return NULL;
+	} while (skip--);
+
+	return dev_to_disk(dev);
 }
 
-static void *part_start(struct seq_file *part, loff_t *pos)
+static void *disk_seqf_next(struct seq_file *seqf, void *v, loff_t *pos)
 {
 	struct device *dev;
-	loff_t k = *pos;
-
-	if (!k)
-		part->private = (void *)1LU;	/* tell show to print header */
 
-	mutex_lock(&block_class_lock);
-	dev = class_find_device(&block_class, NULL, &k, find_start);
-	if (dev) {
-		put_device(dev);
+	(*pos)++;
+	dev = class_dev_iter_next(seqf->private);
+	if (dev)
 		return dev_to_disk(dev);
-	}
+
 	return NULL;
 }
 
-static int find_next(struct device *dev, void *data)
+static void disk_seqf_stop(struct seq_file *seqf, void *v)
 {
-	if (dev->type == &disk_type)
-		return 1;
-	return 0;
-}
+	struct class_dev_iter *iter = seqf->private;
 
-static void *part_next(struct seq_file *part, void *v, loff_t *pos)
-{
-	struct gendisk *gp = v;
-	struct device *dev;
-	++*pos;
-	dev = class_find_device(&block_class, &gp->dev, NULL, find_next);
-	if (dev) {
-		put_device(dev);
-		return dev_to_disk(dev);
+	/* stop is called even after start failed :-( */
+	if (iter) {
+		class_dev_iter_exit(iter);
+		kfree(iter);
 	}
-	return NULL;
 }
 
-static void part_stop(struct seq_file *part, void *v)
+static void *show_partition_start(struct seq_file *seqf, loff_t *pos)
 {
-	mutex_unlock(&block_class_lock);
+	static void *p;
+
+	p = disk_seqf_start(seqf, pos);
+	if (!IS_ERR(p) && p && !*pos)
+		seq_puts(seqf, "major minor  #blocks  name\n\n");
+	return p;
 }
 
-static int show_partition(struct seq_file *part, void *v)
+static int show_partition(struct seq_file *seqf, void *v)
 {
 	struct gendisk *sgp = v;
-	int n;
+	struct disk_part_iter piter;
+	struct hd_struct *part;
 	char buf[BDEVNAME_SIZE];
 
-	/*
-	 * Print header if start told us to do.  This is to preserve
-	 * the original behavior of not printing header if no
-	 * partition exists.  This hackery will be removed later with
-	 * class iteration clean up.
-	 */
-	if (part->private) {
-		seq_puts(part, "major minor  #blocks  name\n\n");
-		part->private = NULL;
-	}
-
 	/* Don't show non-partitionable removeable devices or empty devices */
-	if (!get_capacity(sgp) ||
-			(sgp->minors == 1 && (sgp->flags & GENHD_FL_REMOVABLE)))
+	if (!get_capacity(sgp) || (!disk_partitionable(sgp) &&
+				   (sgp->flags & GENHD_FL_REMOVABLE)))
 		return 0;
 	if (sgp->flags & GENHD_FL_SUPPRESS_PARTITION_INFO)
 		return 0;
 
 	/* show the full disk and all non-0 size partitions of it */
-	seq_printf(part, "%4d  %4d %10llu %s\n",
-		sgp->major, sgp->first_minor,
-		(unsigned long long)get_capacity(sgp) >> 1,
-		disk_name(sgp, 0, buf));
-	for (n = 0; n < sgp->minors - 1; n++) {
-		if (!sgp->part[n])
-			continue;
-		if (sgp->part[n]->nr_sects == 0)
-			continue;
-		seq_printf(part, "%4d  %4d %10llu %s\n",
-			sgp->major, n + 1 + sgp->first_minor,
-			(unsigned long long)sgp->part[n]->nr_sects >> 1 ,
-			disk_name(sgp, n + 1, buf));
-	}
+	disk_part_iter_init(&piter, sgp, DISK_PITER_INCL_PART0);
+	while ((part = disk_part_iter_next(&piter)))
+		seq_printf(seqf, "%4d  %7d %10llu %s\n",
+			   MAJOR(part_devt(part)), MINOR(part_devt(part)),
+			   (unsigned long long)part->nr_sects >> 1,
+			   disk_name(sgp, part->partno, buf));
+	disk_part_iter_exit(&piter);
 
 	return 0;
 }
 
 const struct seq_operations partitions_op = {
-	.start	= part_start,
-	.next	= part_next,
-	.stop	= part_stop,
+	.start	= show_partition_start,
+	.next	= disk_seqf_next,
+	.stop	= disk_seqf_stop,
 	.show	= show_partition
 };
 #endif
 
 
-static struct kobject *base_probe(dev_t devt, int *part, void *data)
+static struct kobject *base_probe(dev_t devt, int *partno, void *data)
 {
 	if (request_module("block-major-%d-%d", MAJOR(devt), MINOR(devt)) > 0)
 		/* Make old-style 2.4 aliases work */
@@ -431,29 +773,29 @@ static ssize_t disk_range_show(struct device *dev,
 	return sprintf(buf, "%d\n", disk->minors);
 }
 
-static ssize_t disk_removable_show(struct device *dev,
+static ssize_t disk_ext_range_show(struct device *dev,
 				   struct device_attribute *attr, char *buf)
 {
 	struct gendisk *disk = dev_to_disk(dev);
 
-	return sprintf(buf, "%d\n",
-		       (disk->flags & GENHD_FL_REMOVABLE ? 1 : 0));
+	return sprintf(buf, "%d\n", disk_max_parts(disk));
 }
 
-static ssize_t disk_ro_show(struct device *dev,
+static ssize_t disk_removable_show(struct device *dev,
 				   struct device_attribute *attr, char *buf)
 {
 	struct gendisk *disk = dev_to_disk(dev);
 
-	return sprintf(buf, "%d\n", disk->policy ? 1 : 0);
+	return sprintf(buf, "%d\n",
+		       (disk->flags & GENHD_FL_REMOVABLE ? 1 : 0));
 }
 
-static ssize_t disk_size_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
+static ssize_t disk_ro_show(struct device *dev,
+				   struct device_attribute *attr, char *buf)
 {
 	struct gendisk *disk = dev_to_disk(dev);
 
-	return sprintf(buf, "%llu\n", (unsigned long long)get_capacity(disk));
+	return sprintf(buf, "%d\n", get_disk_ro(disk) ? 1 : 0);
 }
 
 static ssize_t disk_capability_show(struct device *dev,
@@ -464,73 +806,26 @@ static ssize_t disk_capability_show(struct device *dev,
 	return sprintf(buf, "%x\n", disk->flags);
 }
 
-static ssize_t disk_stat_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
-{
-	struct gendisk *disk = dev_to_disk(dev);
-
-	preempt_disable();
-	disk_round_stats(disk);
-	preempt_enable();
-	return sprintf(buf,
-		"%8lu %8lu %8llu %8u "
-		"%8lu %8lu %8llu %8u "
-		"%8u %8u %8u"
-		"\n",
-		disk_stat_read(disk, ios[READ]),
-		disk_stat_read(disk, merges[READ]),
-		(unsigned long long)disk_stat_read(disk, sectors[READ]),
-		jiffies_to_msecs(disk_stat_read(disk, ticks[READ])),
-		disk_stat_read(disk, ios[WRITE]),
-		disk_stat_read(disk, merges[WRITE]),
-		(unsigned long long)disk_stat_read(disk, sectors[WRITE]),
-		jiffies_to_msecs(disk_stat_read(disk, ticks[WRITE])),
-		disk->in_flight,
-		jiffies_to_msecs(disk_stat_read(disk, io_ticks)),
-		jiffies_to_msecs(disk_stat_read(disk, time_in_queue)));
-}
-
-#ifdef CONFIG_FAIL_MAKE_REQUEST
-static ssize_t disk_fail_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
-{
-	struct gendisk *disk = dev_to_disk(dev);
-
-	return sprintf(buf, "%d\n", disk->flags & GENHD_FL_FAIL ? 1 : 0);
-}
-
-static ssize_t disk_fail_store(struct device *dev,
-			       struct device_attribute *attr,
-			       const char *buf, size_t count)
-{
-	struct gendisk *disk = dev_to_disk(dev);
-	int i;
-
-	if (count > 0 && sscanf(buf, "%d", &i) > 0) {
-		if (i == 0)
-			disk->flags &= ~GENHD_FL_FAIL;
-		else
-			disk->flags |= GENHD_FL_FAIL;
-	}
-
-	return count;
-}
-
-#endif
-
 static DEVICE_ATTR(range, S_IRUGO, disk_range_show, NULL);
+static DEVICE_ATTR(ext_range, S_IRUGO, disk_ext_range_show, NULL);
 static DEVICE_ATTR(removable, S_IRUGO, disk_removable_show, NULL);
 static DEVICE_ATTR(ro, S_IRUGO, disk_ro_show, NULL);
-static DEVICE_ATTR(size, S_IRUGO, disk_size_show, NULL);
+static DEVICE_ATTR(size, S_IRUGO, part_size_show, NULL);
 static DEVICE_ATTR(capability, S_IRUGO, disk_capability_show, NULL);
-static DEVICE_ATTR(stat, S_IRUGO, disk_stat_show, NULL);
+static DEVICE_ATTR(stat, S_IRUGO, part_stat_show, NULL);
 #ifdef CONFIG_FAIL_MAKE_REQUEST
 static struct device_attribute dev_attr_fail =
-	__ATTR(make-it-fail, S_IRUGO|S_IWUSR, disk_fail_show, disk_fail_store);
+	__ATTR(make-it-fail, S_IRUGO|S_IWUSR, part_fail_show, part_fail_store);
+#endif
+#ifdef CONFIG_FAIL_IO_TIMEOUT
+static struct device_attribute dev_attr_fail_timeout =
+	__ATTR(io-timeout-fail,  S_IRUGO|S_IWUSR, part_timeout_show,
+		part_timeout_store);
 #endif
 
 static struct attribute *disk_attrs[] = {
 	&dev_attr_range.attr,
+	&dev_attr_ext_range.attr,
 	&dev_attr_removable.attr,
 	&dev_attr_ro.attr,
 	&dev_attr_size.attr,
@@ -539,6 +834,9 @@ static struct attribute *disk_attrs[] = {
 #ifdef CONFIG_FAIL_MAKE_REQUEST
 	&dev_attr_fail.attr,
 #endif
+#ifdef CONFIG_FAIL_IO_TIMEOUT
+	&dev_attr_fail_timeout.attr,
+#endif
 	NULL
 };
 
@@ -551,13 +849,87 @@ static struct attribute_group *disk_attr_groups[] = {
 	NULL
 };
 
+static void disk_free_ptbl_rcu_cb(struct rcu_head *head)
+{
+	struct disk_part_tbl *ptbl =
+		container_of(head, struct disk_part_tbl, rcu_head);
+
+	kfree(ptbl);
+}
+
+/**
+ * disk_replace_part_tbl - replace disk->part_tbl in RCU-safe way
+ * @disk: disk to replace part_tbl for
+ * @new_ptbl: new part_tbl to install
+ *
+ * Replace disk->part_tbl with @new_ptbl in RCU-safe way.  The
+ * original ptbl is freed using RCU callback.
+ *
+ * LOCKING:
+ * Matching bd_mutx locked.
+ */
+static void disk_replace_part_tbl(struct gendisk *disk,
+				  struct disk_part_tbl *new_ptbl)
+{
+	struct disk_part_tbl *old_ptbl = disk->part_tbl;
+
+	rcu_assign_pointer(disk->part_tbl, new_ptbl);
+	if (old_ptbl)
+		call_rcu(&old_ptbl->rcu_head, disk_free_ptbl_rcu_cb);
+}
+
+/**
+ * disk_expand_part_tbl - expand disk->part_tbl
+ * @disk: disk to expand part_tbl for
+ * @partno: expand such that this partno can fit in
+ *
+ * Expand disk->part_tbl such that @partno can fit in.  disk->part_tbl
+ * uses RCU to allow unlocked dereferencing for stats and other stuff.
+ *
+ * LOCKING:
+ * Matching bd_mutex locked, might sleep.
+ *
+ * RETURNS:
+ * 0 on success, -errno on failure.
+ */
+int disk_expand_part_tbl(struct gendisk *disk, int partno)
+{
+	struct disk_part_tbl *old_ptbl = disk->part_tbl;
+	struct disk_part_tbl *new_ptbl;
+	int len = old_ptbl ? old_ptbl->len : 0;
+	int target = partno + 1;
+	size_t size;
+	int i;
+
+	/* disk_max_parts() is zero during initialization, ignore if so */
+	if (disk_max_parts(disk) && target > disk_max_parts(disk))
+		return -EINVAL;
+
+	if (target <= len)
+		return 0;
+
+	size = sizeof(*new_ptbl) + target * sizeof(new_ptbl->part[0]);
+	new_ptbl = kzalloc_node(size, GFP_KERNEL, disk->node_id);
+	if (!new_ptbl)
+		return -ENOMEM;
+
+	INIT_RCU_HEAD(&new_ptbl->rcu_head);
+	new_ptbl->len = target;
+
+	for (i = 0; i < len; i++)
+		rcu_assign_pointer(new_ptbl->part[i], old_ptbl->part[i]);
+
+	disk_replace_part_tbl(disk, new_ptbl);
+	return 0;
+}
+
 static void disk_release(struct device *dev)
 {
 	struct gendisk *disk = dev_to_disk(dev);
 
 	kfree(disk->random);
-	kfree(disk->part);
-	free_disk_stats(disk);
+	disk_replace_part_tbl(disk, NULL);
+	free_part_stats(&disk->part0);
 	kfree(disk);
 }
 struct class block_class = {
@@ -578,83 +950,31 @@ static struct device_type disk_type = {
  * The output looks suspiciously like /proc/partitions with a bunch of
  * extra fields.
  */
-
-static void *diskstats_start(struct seq_file *part, loff_t *pos)
-{
-	struct device *dev;
-	loff_t k = *pos;
-
-	mutex_lock(&block_class_lock);
-	dev = class_find_device(&block_class, NULL, &k, find_start);
-	if (dev) {
-		put_device(dev);
-		return dev_to_disk(dev);
-	}
-	return NULL;
-}
-
-static void *diskstats_next(struct seq_file *part, void *v, loff_t *pos)
-{
-	struct gendisk *gp = v;
-	struct device *dev;
-
-	++*pos;
-	dev = class_find_device(&block_class, &gp->dev, NULL, find_next);
-	if (dev) {
-		put_device(dev);
-		return dev_to_disk(dev);
-	}
-	return NULL;
-}
-
-static void diskstats_stop(struct seq_file *part, void *v)
-{
-	mutex_unlock(&block_class_lock);
-}
-
-static int diskstats_show(struct seq_file *s, void *v)
+static int diskstats_show(struct seq_file *seqf, void *v)
 {
 	struct gendisk *gp = v;
+	struct disk_part_iter piter;
+	struct hd_struct *hd;
 	char buf[BDEVNAME_SIZE];
-	int n = 0;
+	int cpu;
 
 	/*
-	if (&gp->dev.kobj.entry == block_class.devices.next)
-		seq_puts(s,	"major minor name"
+	if (&disk_to_dev(gp)->kobj.entry == block_class.devices.next)
+		seq_puts(seqf,	"major minor name"
 				"     rio rmerge rsect ruse wio wmerge "
 				"wsect wuse running use aveq"
 				"\n\n");
 	*/
  
-	preempt_disable();
-	disk_round_stats(gp);
-	preempt_enable();
-	seq_printf(s, "%4d %4d %s %lu %lu %llu %u %lu %lu %llu %u %u %u %u\n",
-		gp->major, n + gp->first_minor, disk_name(gp, n, buf),
-		disk_stat_read(gp, ios[0]), disk_stat_read(gp, merges[0]),
-		(unsigned long long)disk_stat_read(gp, sectors[0]),
-		jiffies_to_msecs(disk_stat_read(gp, ticks[0])),
-		disk_stat_read(gp, ios[1]), disk_stat_read(gp, merges[1]),
-		(unsigned long long)disk_stat_read(gp, sectors[1]),
-		jiffies_to_msecs(disk_stat_read(gp, ticks[1])),
-		gp->in_flight,
-		jiffies_to_msecs(disk_stat_read(gp, io_ticks)),
-		jiffies_to_msecs(disk_stat_read(gp, time_in_queue)));
-
-	/* now show all non-0 size partitions of it */
-	for (n = 0; n < gp->minors - 1; n++) {
-		struct hd_struct *hd = gp->part[n];
-
-		if (!hd || !hd->nr_sects)
-			continue;
-
-		preempt_disable();
-		part_round_stats(hd);
-		preempt_enable();
-		seq_printf(s, "%4d %4d %s %lu %lu %llu "
+	disk_part_iter_init(&piter, gp, DISK_PITER_INCL_PART0);
+	while ((hd = disk_part_iter_next(&piter))) {
+		cpu = part_stat_lock();
+		part_round_stats(cpu, hd);
+		part_stat_unlock();
+		seq_printf(seqf, "%4d %7d %s %lu %lu %llu "
 			   "%u %lu %lu %llu %u %u %u %u\n",
-			   gp->major, n + gp->first_minor + 1,
-			   disk_name(gp, n + 1, buf),
+			   MAJOR(part_devt(hd)), MINOR(part_devt(hd)),
+			   disk_name(gp, hd->partno, buf),
 			   part_stat_read(hd, ios[0]),
 			   part_stat_read(hd, merges[0]),
 			   (unsigned long long)part_stat_read(hd, sectors[0]),
@@ -668,14 +988,15 @@ static int diskstats_show(struct seq_file *s, void *v)
 			   jiffies_to_msecs(part_stat_read(hd, time_in_queue))
 			);
 	}
+	disk_part_iter_exit(&piter);
  
 	return 0;
 }
 
 const struct seq_operations diskstats_op = {
-	.start	= diskstats_start,
-	.next	= diskstats_next,
-	.stop	= diskstats_stop,
+	.start	= disk_seqf_start,
+	.next	= disk_seqf_next,
+	.stop	= disk_seqf_stop,
 	.show	= diskstats_show
 };
 #endif /* CONFIG_PROC_FS */
@@ -690,7 +1011,7 @@ static void media_change_notify_thread(struct work_struct *work)
 	 * set enviroment vars to indicate which event this is for
 	 * so that user space will know to go check the media status.
 	 */
-	kobject_uevent_env(&gd->dev.kobj, KOBJ_CHANGE, envp);
+	kobject_uevent_env(&disk_to_dev(gd)->kobj, KOBJ_CHANGE, envp);
 	put_device(gd->driverfs_dev);
 }
 
@@ -703,42 +1024,29 @@ void genhd_media_change_notify(struct gendisk *disk)
 EXPORT_SYMBOL_GPL(genhd_media_change_notify);
 #endif  /*  0  */
 
-struct find_block {
-	const char *name;
-	int part;
-};
-
-static int match_id(struct device *dev, void *data)
+dev_t blk_lookup_devt(const char *name, int partno)
 {
-	struct find_block *find = data;
+	dev_t devt = MKDEV(0, 0);
+	struct class_dev_iter iter;
+	struct device *dev;
 
-	if (dev->type != &disk_type)
-		return 0;
-	if (strcmp(dev->bus_id, find->name) == 0) {
+	class_dev_iter_init(&iter, &block_class, NULL, &disk_type);
+	while ((dev = class_dev_iter_next(&iter))) {
 		struct gendisk *disk = dev_to_disk(dev);
-		if (find->part < disk->minors)
-			return 1;
-	}
-	return 0;
-}
+		struct hd_struct *part;
 
-dev_t blk_lookup_devt(const char *name, int part)
-{
-	struct device *dev;
-	dev_t devt = MKDEV(0, 0);
-	struct find_block find;
+		if (strcmp(dev->bus_id, name))
+			continue;
 
-	mutex_lock(&block_class_lock);
-	find.name = name;
-	find.part = part;
-	dev = class_find_device(&block_class, NULL, &find, match_id);
-	if (dev) {
-		put_device(dev);
-		devt = MKDEV(MAJOR(dev->devt),
-			     MINOR(dev->devt) + part);
+		part = disk_get_part(disk, partno);
+		if (part) {
+			devt = part_devt(part);
+			disk_put_part(part);
+			break;
+		}
+		disk_put_part(part);
 	}
-	mutex_unlock(&block_class_lock);
-
+	class_dev_iter_exit(&iter);
 	return devt;
 }
 EXPORT_SYMBOL(blk_lookup_devt);
@@ -747,6 +1055,7 @@ struct gendisk *alloc_disk(int minors)
 {
 	return alloc_disk_node(minors, -1);
 }
+EXPORT_SYMBOL(alloc_disk);
 
 struct gendisk *alloc_disk_node(int minors, int node_id)
 {
@@ -755,32 +1064,28 @@ struct gendisk *alloc_disk_node(int minors, int node_id)
 	disk = kmalloc_node(sizeof(struct gendisk),
 				GFP_KERNEL | __GFP_ZERO, node_id);
 	if (disk) {
-		if (!init_disk_stats(disk)) {
+		if (!init_part_stats(&disk->part0)) {
 			kfree(disk);
 			return NULL;
 		}
-		if (minors > 1) {
-			int size = (minors - 1) * sizeof(struct hd_struct *);
-			disk->part = kmalloc_node(size,
-				GFP_KERNEL | __GFP_ZERO, node_id);
-			if (!disk->part) {
-				free_disk_stats(disk);
-				kfree(disk);
-				return NULL;
-			}
+		if (disk_expand_part_tbl(disk, 0)) {
+			free_part_stats(&disk->part0);
+			kfree(disk);
+			return NULL;
 		}
+		disk->part_tbl->part[0] = &disk->part0;
+
 		disk->minors = minors;
 		rand_initialize_disk(disk);
-		disk->dev.class = &block_class;
-		disk->dev.type = &disk_type;
-		device_initialize(&disk->dev);
+		disk_to_dev(disk)->class = &block_class;
+		disk_to_dev(disk)->type = &disk_type;
+		device_initialize(disk_to_dev(disk));
 		INIT_WORK(&disk->async_notify,
 			media_change_notify_thread);
+		disk->node_id = node_id;
 	}
 	return disk;
 }
-
-EXPORT_SYMBOL(alloc_disk);
 EXPORT_SYMBOL(alloc_disk_node);
 
 struct kobject *get_disk(struct gendisk *disk)
@@ -793,7 +1098,7 @@ struct kobject *get_disk(struct gendisk *disk)
 	owner = disk->fops->owner;
 	if (owner && !try_module_get(owner))
 		return NULL;
-	kobj = kobject_get(&disk->dev.kobj);
+	kobj = kobject_get(&disk_to_dev(disk)->kobj);
 	if (kobj == NULL) {
 		module_put(owner);
 		return NULL;
@@ -807,27 +1112,28 @@ EXPORT_SYMBOL(get_disk);
 void put_disk(struct gendisk *disk)
 {
 	if (disk)
-		kobject_put(&disk->dev.kobj);
+		kobject_put(&disk_to_dev(disk)->kobj);
 }
 
 EXPORT_SYMBOL(put_disk);
 
 void set_device_ro(struct block_device *bdev, int flag)
 {
-	if (bdev->bd_contains != bdev)
-		bdev->bd_part->policy = flag;
-	else
-		bdev->bd_disk->policy = flag;
+	bdev->bd_part->policy = flag;
 }
 
 EXPORT_SYMBOL(set_device_ro);
 
 void set_disk_ro(struct gendisk *disk, int flag)
 {
-	int i;
-	disk->policy = flag;
-	for (i = 0; i < disk->minors - 1; i++)
-		if (disk->part[i]) disk->part[i]->policy = flag;
+	struct disk_part_iter piter;
+	struct hd_struct *part;
+
+	disk_part_iter_init(&piter, disk,
+			    DISK_PITER_INCL_EMPTY | DISK_PITER_INCL_PART0);
+	while ((part = disk_part_iter_next(&piter)))
+		part->policy = flag;
+	disk_part_iter_exit(&piter);
 }
 
 EXPORT_SYMBOL(set_disk_ro);
@@ -836,18 +1142,15 @@ int bdev_read_only(struct block_device *bdev)
 {
 	if (!bdev)
 		return 0;
-	else if (bdev->bd_contains != bdev)
-		return bdev->bd_part->policy;
-	else
-		return bdev->bd_disk->policy;
+	return bdev->bd_part->policy;
 }
 
 EXPORT_SYMBOL(bdev_read_only);
 
-int invalidate_partition(struct gendisk *disk, int index)
+int invalidate_partition(struct gendisk *disk, int partno)
 {
 	int res = 0;
-	struct block_device *bdev = bdget_disk(disk, index);
+	struct block_device *bdev = bdget_disk(disk, partno);
 	if (bdev) {
 		fsync_bdev(bdev);
 		res = __invalidate_device(bdev);
diff --git a/block/ioctl.c b/block/ioctl.c
index 77185e5c026..38bee321e1f 100644
--- a/block/ioctl.c
+++ b/block/ioctl.c
@@ -12,11 +12,12 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user
 {
 	struct block_device *bdevp;
 	struct gendisk *disk;
+	struct hd_struct *part;
 	struct blkpg_ioctl_arg a;
 	struct blkpg_partition p;
+	struct disk_part_iter piter;
 	long long start, length;
-	int part;
-	int i;
+	int partno;
 	int err;
 
 	if (!capable(CAP_SYS_ADMIN))
@@ -28,8 +29,8 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user
 	disk = bdev->bd_disk;
 	if (bdev != bdev->bd_contains)
 		return -EINVAL;
-	part = p.pno;
-	if (part <= 0 || part >= disk->minors)
+	partno = p.pno;
+	if (partno <= 0)
 		return -EINVAL;
 	switch (a.op) {
 		case BLKPG_ADD_PARTITION:
@@ -43,36 +44,37 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user
 				    || pstart < 0 || plength < 0)
 					return -EINVAL;
 			}
-			/* partition number in use? */
+
 			mutex_lock(&bdev->bd_mutex);
-			if (disk->part[part - 1]) {
-				mutex_unlock(&bdev->bd_mutex);
-				return -EBUSY;
-			}
-			/* overlap? */
-			for (i = 0; i < disk->minors - 1; i++) {
-				struct hd_struct *s = disk->part[i];
 
-				if (!s)
-					continue;
-				if (!(start+length <= s->start_sect ||
-				      start >= s->start_sect + s->nr_sects)) {
+			/* overlap? */
+			disk_part_iter_init(&piter, disk,
+					    DISK_PITER_INCL_EMPTY);
+			while ((part = disk_part_iter_next(&piter))) {
+				if (!(start + length <= part->start_sect ||
+				      start >= part->start_sect + part->nr_sects)) {
+					disk_part_iter_exit(&piter);
 					mutex_unlock(&bdev->bd_mutex);
 					return -EBUSY;
 				}
 			}
+			disk_part_iter_exit(&piter);
+
 			/* all seems OK */
-			err = add_partition(disk, part, start, length, ADDPART_FLAG_NONE);
+			err = add_partition(disk, partno, start, length,
+					    ADDPART_FLAG_NONE);
 			mutex_unlock(&bdev->bd_mutex);
 			return err;
 		case BLKPG_DEL_PARTITION:
-			if (!disk->part[part-1])
-				return -ENXIO;
-			if (disk->part[part - 1]->nr_sects == 0)
+			part = disk_get_part(disk, partno);
+			if (!part)
 				return -ENXIO;
-			bdevp = bdget_disk(disk, part);
+
+			bdevp = bdget(part_devt(part));
+			disk_put_part(part);
 			if (!bdevp)
 				return -ENOMEM;
+
 			mutex_lock(&bdevp->bd_mutex);
 			if (bdevp->bd_openers) {
 				mutex_unlock(&bdevp->bd_mutex);
@@ -84,7 +86,7 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user
 			invalidate_bdev(bdevp);
 
 			mutex_lock_nested(&bdev->bd_mutex, 1);
-			delete_partition(disk, part);
+			delete_partition(disk, partno);
 			mutex_unlock(&bdev->bd_mutex);
 			mutex_unlock(&bdevp->bd_mutex);
 			bdput(bdevp);
@@ -100,7 +102,7 @@ static int blkdev_reread_part(struct block_device *bdev)
 	struct gendisk *disk = bdev->bd_disk;
 	int res;
 
-	if (disk->minors == 1 || bdev != bdev->bd_contains)
+	if (!disk_partitionable(disk) || bdev != bdev->bd_contains)
 		return -EINVAL;
 	if (!capable(CAP_SYS_ADMIN))
 		return -EACCES;
@@ -111,6 +113,69 @@ static int blkdev_reread_part(struct block_device *bdev)
 	return res;
 }
 
+static void blk_ioc_discard_endio(struct bio *bio, int err)
+{
+	if (err) {
+		if (err == -EOPNOTSUPP)
+			set_bit(BIO_EOPNOTSUPP, &bio->bi_flags);
+		clear_bit(BIO_UPTODATE, &bio->bi_flags);
+	}
+	complete(bio->bi_private);
+}
+
+static int blk_ioctl_discard(struct block_device *bdev, uint64_t start,
+			     uint64_t len)
+{
+	struct request_queue *q = bdev_get_queue(bdev);
+	int ret = 0;
+
+	if (start & 511)
+		return -EINVAL;
+	if (len & 511)
+		return -EINVAL;
+	start >>= 9;
+	len >>= 9;
+
+	if (start + len > (bdev->bd_inode->i_size >> 9))
+		return -EINVAL;
+
+	if (!q->prepare_discard_fn)
+		return -EOPNOTSUPP;
+
+	while (len && !ret) {
+		DECLARE_COMPLETION_ONSTACK(wait);
+		struct bio *bio;
+
+		bio = bio_alloc(GFP_KERNEL, 0);
+		if (!bio)
+			return -ENOMEM;
+
+		bio->bi_end_io = blk_ioc_discard_endio;
+		bio->bi_bdev = bdev;
+		bio->bi_private = &wait;
+		bio->bi_sector = start;
+
+		if (len > q->max_hw_sectors) {
+			bio->bi_size = q->max_hw_sectors << 9;
+			len -= q->max_hw_sectors;
+			start += q->max_hw_sectors;
+		} else {
+			bio->bi_size = len << 9;
+			len = 0;
+		}
+		submit_bio(DISCARD_NOBARRIER, bio);
+
+		wait_for_completion(&wait);
+
+		if (bio_flagged(bio, BIO_EOPNOTSUPP))
+			ret = -EOPNOTSUPP;
+		else if (!bio_flagged(bio, BIO_UPTODATE))
+			ret = -EIO;
+		bio_put(bio);
+	}
+	return ret;
+}
+
 static int put_ushort(unsigned long arg, unsigned short val)
 {
 	return put_user(val, (unsigned short __user *)arg);
@@ -258,6 +323,19 @@ int blkdev_ioctl(struct inode *inode, struct file *file, unsigned cmd,
 		set_device_ro(bdev, n);
 		unlock_kernel();
 		return 0;
+
+	case BLKDISCARD: {
+		uint64_t range[2];
+
+		if (!(file->f_mode & FMODE_WRITE))
+			return -EBADF;
+
+		if (copy_from_user(range, (void __user *)arg, sizeof(range)))
+			return -EFAULT;
+
+		return blk_ioctl_discard(bdev, range[0], range[1]);
+	}
+
 	case HDIO_GETGEO: {
 		struct hd_geometry geo;
 
diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c
index ec4b7f23462..c34272a348f 100644
--- a/block/scsi_ioctl.c
+++ b/block/scsi_ioctl.c
@@ -185,6 +185,7 @@ void blk_set_cmd_filter_defaults(struct blk_cmd_filter *filter)
 	__set_bit(GPCMD_PREVENT_ALLOW_MEDIUM_REMOVAL, filter->write_ok);
 	__set_bit(GPCMD_LOAD_UNLOAD, filter->write_ok);
 	__set_bit(GPCMD_SET_STREAMING, filter->write_ok);
+	__set_bit(GPCMD_SET_READ_AHEAD, filter->write_ok);
 }
 EXPORT_SYMBOL_GPL(blk_set_cmd_filter_defaults);
 
@@ -313,11 +314,12 @@ static int sg_io(struct file *file, struct request_queue *q,
 			goto out;
 		}
 
-		ret = blk_rq_map_user_iov(q, rq, iov, hdr->iovec_count,
-					  hdr->dxfer_len);
+		ret = blk_rq_map_user_iov(q, rq, NULL, iov, hdr->iovec_count,
+					  hdr->dxfer_len, GFP_KERNEL);
 		kfree(iov);
 	} else if (hdr->dxfer_len)
-		ret = blk_rq_map_user(q, rq, hdr->dxferp, hdr->dxfer_len);
+		ret = blk_rq_map_user(q, rq, NULL, hdr->dxferp, hdr->dxfer_len,
+				      GFP_KERNEL);
 
 	if (ret)
 		goto out;