Changed checkfs to auto reboot after correctable fsck fixes.

[people/pmueller/ipfire-2.x.git] / src / patches / suse-2.6.27.25 / patches.fixes / block-git-fixes

From: Jens Axboe <jens.axboe@oracle.com>
Subject: Block layer fixes for 2.6.28
Patch-Mainline: 2.6.28

This is a combined patchset with block layer fixes from 2.6.28.
Commit IDs:
97dee27d1c4d6041ff1cc8150db95fe3eab6be5a
00bbda44114e70fc9879731be3c888122b1de8b1
7452d2a2be657becb2f385d0e0864ba51f1ae694
075a108f7d4dd24b8b69e59edcdf1a0fd84e6541
7a1b6029bf9ff3d0636e318d2482031dc493df16
b3a5faf3cefbff4b69ca181767b882bbd6189aaf
8fe902de23b4f4012db91f538cafd864c63308e7
dfef13dad8d34d0a9e83adee3e8cd9f94cca465e
d2629dd70132f90f9d1bca07572197e9adea25b1
1f08a4484a223cb337e0466042005421cd55d22b
fcdc7361d2925596d69d0538d738c08c221a69c9
cd93bcfa9ca9b15051220614160131c53d7f33f0
d371ca6b8a21a617b8607d23f7202197ad40482a
910ee03b1e61d5cfb121dfb1ee7c127f18bdae01

Signed-off-by: Hannes Reinecke <hare@suse.de>

---
 Documentation/DocBook/kernel-api.tmpl    |    4 
 Documentation/block/deadline-iosched.txt |   14 +-
 block/Makefile                           |    4 
 block/blk-core.c                         |  166 +++++++------------------------
 block/blk-exec.c                         |    6 -
 block/blk-integrity.c                    |    4 
 block/blk-map.c                          |   16 +-
 block/blk-merge.c                        |  100 ------------------
 block/blk-settings.c                     |    8 -
 block/blk-softirq.c                      |  103 +++++++++++++++++++
 block/blk-tag.c                          |    8 -
 block/cfq-iosched.c                      |   47 +++++++-
 block/deadline-iosched.c                 |   40 ++-----
 block/elevator.c                         |    5 
 block/genhd.c                            |    5 
 drivers/block/ps3disk.c                  |    9 +
 drivers/block/virtio_blk.c               |    4 
 drivers/md/raid1.c                       |    4 
 drivers/md/raid10.c                      |    4 
 drivers/md/raid5.c                       |   66 +++++++++---
 fs/bio.c                                 |   16 --
 include/linux/bio.h                      |   33 ------
 include/linux/blkdev.h                   |   18 +--
 23 files changed, 310 insertions(+), 374 deletions(-)

--- 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>
 
@@ -50,8 +48,6 @@ 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;
@@ -531,7 +527,7 @@ 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:
@@ -913,7 +909,7 @@ void blk_requeue_request(struct request_
 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
@@ -923,8 +919,8 @@ 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
@@ -1322,7 +1318,7 @@ static inline int bio_check_eod(struct b
 }
 
 /**
- * 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
@@ -1480,13 +1476,13 @@ void generic_make_request(struct bio *bi
 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)
@@ -1524,7 +1520,7 @@ EXPORT_SYMBOL(submit_bio);
 /**
  * __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:
@@ -1532,8 +1528,8 @@ 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)
@@ -1544,7 +1540,7 @@ static int __end_that_request_first(stru
 	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))
@@ -1646,82 +1642,6 @@ static int __end_that_request_first(stru
 }
 
 /*
- * 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)
@@ -1810,11 +1730,11 @@ 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
+ * @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
+ *     Not suitable for normal I/O completion, unless the driver still has
  *     the request attached to the block layer.
  *
  **/
@@ -1827,7 +1747,7 @@ 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
+ * @uptodate:	error value or %0/%1 uptodate flag
  *
  * Description:
  *     Ends all I/O on a request. The request must already have been
@@ -1845,14 +1765,14 @@ 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
@@ -1870,12 +1790,12 @@ EXPORT_SYMBOL(end_request);
 /**
  * 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:
@@ -1883,8 +1803,8 @@ 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,
@@ -1893,7 +1813,7 @@ static int blk_end_io(struct request *rq
 	struct request_queue *q = rq->q;
 	unsigned long flags = 0UL;
 
-	if (bio_has_data(rq->bio) || blk_discard_rq(rq)) {
+	if (rq->bio) {
 		if (__end_that_request_first(rq, error, nr_bytes))
 			return 1;
 
@@ -1919,7 +1839,7 @@ static int blk_end_io(struct request *rq
 /**
  * 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:
@@ -1927,8 +1847,8 @@ static int blk_end_io(struct request *rq
  *     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)
 {
@@ -1939,20 +1859,19 @@ 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 ((bio_has_data(rq->bio) || blk_discard_rq(rq)) &&
-	    __end_that_request_first(rq, error, nr_bytes))
+	if (rq->bio && __end_that_request_first(rq, error, nr_bytes))
 		return 1;
 
 	add_disk_randomness(rq->rq_disk);
@@ -1966,7 +1885,7 @@ 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
  *
@@ -1974,8 +1893,8 @@ 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)
@@ -1987,11 +1906,11 @@ EXPORT_SYMBOL_GPL(blk_end_bidi_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:
@@ -2004,10 +1923,10 @@ 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,
@@ -2026,7 +1945,6 @@ void blk_rq_bio_prep(struct request_queu
 
 	if (bio_has_data(bio)) {
 		rq->nr_phys_segments = bio_phys_segments(q, bio);
-		rq->nr_hw_segments = bio_hw_segments(q, bio);
 		rq->buffer = bio_data(bio);
 	}
 	rq->current_nr_sectors = bio_cur_sectors(bio);
@@ -2054,8 +1972,6 @@ 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");
@@ -2066,12 +1982,6 @@ 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;
 }
 
--- 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 reque
  * @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,
--- a/block/blk-integrity.c
+++ b/block/blk-integrity.c
@@ -109,8 +109,8 @@ 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
+ * @bd1:	Device to compare
+ * @bd2:	Device to compare
  *
  * Description: Meta-devices like DM and MD need to verify that all
  * sub-devices use the same integrity format before advertising to
--- a/block/blk-map.c
+++ b/block/blk-map.c
@@ -85,17 +85,17 @@ static int __blk_rq_map_user(struct requ
 }
 
 /**
- * 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
  * @ubuf:	the user buffer
  * @len:	length of user data
  *
  * 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()
@@ -154,7 +154,7 @@ 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
  * @iov:	pointer to the iovec
@@ -162,10 +162,10 @@ EXPORT_SYMBOL(blk_rq_map_user);
  * @len:	I/O byte count
  *
  * 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()
@@ -224,7 +224,7 @@ int blk_rq_map_user_iov(struct request_q
  * 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 +250,7 @@ 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
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -41,12 +41,9 @@ void blk_recalc_rq_sectors(struct reques
 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 reque
 		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 reque
 		 */
 		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;
@@ -74,27 +71,12 @@ void blk_recalc_rq_segments(struct reque
 				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++;
-		}
-
 		if (nr_phys_segs == 1 && seg_size > rq->bio->bi_seg_front_size)
 			rq->bio->bi_seg_front_size = seg_size;
 
@@ -104,17 +86,11 @@ new_hw_segment:
 		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;
 	if (nr_phys_segs == 1 && seg_size > rq->bio->bi_seg_front_size)
 		rq->bio->bi_seg_front_size = seg_size;
 	if (seg_size > rq->biotail->bi_seg_back_size)
 		rq->biotail->bi_seg_back_size = 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)
@@ -127,7 +103,6 @@ void blk_recount_segments(struct request
 	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);
@@ -158,23 +133,6 @@ static int blk_phys_contig_segment(struc
 	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 (bio_has_data(bio) &&
-	    (!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
@@ -288,10 +246,9 @@ static inline int ll_new_hw_segment(stru
 				    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)
@@ -303,7 +260,6 @@ static inline int ll_new_hw_segment(stru
 	 * 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;
 }
@@ -312,7 +268,6 @@ int ll_back_merge_fn(struct request_queu
 		     struct bio *bio)
 {
 	unsigned short max_sectors;
-	int len;
 
 	if (unlikely(blk_pc_request(req)))
 		max_sectors = q->max_hw_sectors;
@@ -329,20 +284,6 @@ int ll_back_merge_fn(struct request_queu
 		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 (!bio_has_data(bio) ||
-	    (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);
 }
@@ -351,7 +292,6 @@ int ll_front_merge_fn(struct request_que
 		      struct bio *bio)
 {
 	unsigned short max_sectors;
-	int len;
 
 	if (unlikely(blk_pc_request(req)))
 		max_sectors = q->max_hw_sectors;
@@ -365,24 +305,10 @@ int ll_front_merge_fn(struct request_que
 			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 (!bio_has_data(bio) ||
-	    (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);
 }
@@ -391,7 +317,6 @@ static int ll_merge_requests_fn(struct r
 				struct request *next)
 {
 	int total_phys_segments;
-	int total_hw_segments;
 	unsigned int seg_size =
 		req->biotail->bi_seg_back_size + next->bio->bi_seg_front_size;
 
@@ -420,26 +345,11 @@ static int ll_merge_requests_fn(struct r
 	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;
 }
 
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -144,7 +144,7 @@ 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)
 {
@@ -229,7 +229,7 @@ EXPORT_SYMBOL(blk_queue_max_phys_segment
  * 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,
@@ -410,7 +410,7 @@ 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.
  *
  **/
@@ -426,7 +426,7 @@ 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
--- /dev/null
+++ b/block/blk-softirq.c
@@ -0,0 +1,103 @@
+/*
+ * 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);
+
+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,
+};
+
+/*
+ * 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);
+	}
+}
+
+/**
+ * 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);
+
+int __init 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);
--- 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 reques
  * 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.
--- 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
@@ -654,15 +662,6 @@ static void cfq_activate_request(struct 
 	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 re
 	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--;
@@ -1833,6 +1833,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, s
 {
 	struct cfq_io_context *cic = RQ_CIC(rq);
 
+	cfqd->rq_queued++;
 	if (rq_is_meta(rq))
 		cfqq->meta_pending++;
 
@@ -1880,6 +1881,31 @@ static void cfq_insert_request(struct re
 	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 +1916,8 @@ static void cfq_completed_request(struct
 	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 +2228,7 @@ static void *cfq_init_queue(struct reque
 	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;
 }
--- 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 *
 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 
 	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_queu
 	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(stru
 	 * 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_da
 	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_da
 }
 
 /*
- * 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(st
 	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
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -34,8 +34,7 @@
 #include <linux/delay.h>
 #include <linux/blktrace_api.h>
 #include <linux/hash.h>
-
-#include <asm/uaccess.h>
+#include <linux/uaccess.h>
 
 static DEFINE_SPINLOCK(elv_list_lock);
 static LIST_HEAD(elv_list);
@@ -790,7 +789,6 @@ struct request *elv_next_request(struct 
 			 * device can handle
 			 */
 			rq->nr_phys_segments++;
-			rq->nr_hw_segments++;
 		}
 
 		if (!q->prep_rq_fn)
@@ -813,7 +811,6 @@ struct request *elv_next_request(struct 
 				 * so that we don't add it again
 				 */
 				--rq->nr_phys_segments;
-				--rq->nr_hw_segments;
 			}
 
 			rq = NULL;
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -211,10 +211,11 @@ void unlink_gendisk(struct gendisk *disk
 
 /**
  * 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)
 {
--- 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 ioctl.o genhd.o \
+			scsi_ioctl.o cmd-filter.o
 
 obj-$(CONFIG_BLK_DEV_BSG)	+= bsg.o
 obj-$(CONFIG_IOSCHED_NOOP)	+= noop-iosched.o
--- a/Documentation/block/deadline-iosched.txt
+++ b/Documentation/block/deadline-iosched.txt
@@ -30,12 +30,18 @@ write_expire	(in ms)
 Similar to read_expire mentioned above, but for writes.
 
 
-fifo_batch
+fifo_batch	(number of requests)
 ----------
 
-When a read request expires its deadline, we must move some requests from
-the sorted io scheduler list to the block device dispatch queue. fifo_batch
-controls how many requests we move.
+Requests are grouped into ``batches'' of a particular data direction (read or
+write) which are serviced in increasing sector order.  To limit extra seeking,
+deadline expiries are only checked between batches.  fifo_batch controls the
+maximum number of requests per batch.
+
+This parameter tunes the balance between per-request latency and aggregate
+throughput.  When low latency is the primary concern, smaller is better (where
+a value of 1 yields first-come first-served behaviour).  Increasing fifo_batch
+generally improves throughput, at the cost of latency variation.
 
 
 writes_starved	(number of dispatches)
--- a/Documentation/DocBook/kernel-api.tmpl
+++ b/Documentation/DocBook/kernel-api.tmpl
@@ -364,6 +364,10 @@ X!Edrivers/pnp/system.c
 !Eblock/blk-barrier.c
 !Eblock/blk-tag.c
 !Iblock/blk-tag.c
+!Eblock/blk-integrity.c
+!Iblock/blktrace.c
+!Iblock/genhd.c
+!Eblock/genhd.c
   </chapter>
 
   <chapter id="chrdev">
--- a/drivers/block/ps3disk.c
+++ b/drivers/block/ps3disk.c
@@ -199,7 +199,8 @@ static void ps3disk_do_request(struct ps
 		if (blk_fs_request(req)) {
 			if (ps3disk_submit_request_sg(dev, req))
 				break;
-		} else if (req->cmd_type == REQ_TYPE_FLUSH) {
+		} else if (req->cmd_type == REQ_TYPE_LINUX_BLOCK &&
+			   req->cmd[0] == REQ_LB_OP_FLUSH) {
 			if (ps3disk_submit_flush_request(dev, req))
 				break;
 		} else {
@@ -257,7 +258,8 @@ static irqreturn_t ps3disk_interrupt(int
 		return IRQ_HANDLED;
 	}
 
-	if (req->cmd_type == REQ_TYPE_FLUSH) {
+	if (req->cmd_type == REQ_TYPE_LINUX_BLOCK &&
+	    req->cmd[0] == REQ_LB_OP_FLUSH) {
 		read = 0;
 		num_sectors = req->hard_cur_sectors;
 		op = "flush";
@@ -405,7 +407,8 @@ static void ps3disk_prepare_flush(struct
 
 	dev_dbg(&dev->sbd.core, "%s:%u\n", __func__, __LINE__);
 
-	req->cmd_type = REQ_TYPE_FLUSH;
+	req->cmd_type = REQ_TYPE_LINUX_BLOCK;
+	req->cmd[0] = REQ_LB_OP_FLUSH;
 }
 
 static unsigned long ps3disk_mask;
--- a/drivers/block/virtio_blk.c
+++ b/drivers/block/virtio_blk.c
@@ -84,11 +84,11 @@ static bool do_req(struct request_queue 
 	if (blk_fs_request(vbr->req)) {
 		vbr->out_hdr.type = 0;
 		vbr->out_hdr.sector = vbr->req->sector;
-		vbr->out_hdr.ioprio = vbr->req->ioprio;
+		vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
 	} else if (blk_pc_request(vbr->req)) {
 		vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD;
 		vbr->out_hdr.sector = 0;
-		vbr->out_hdr.ioprio = vbr->req->ioprio;
+		vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
 	} else {
 		/* We don't put anything else in the queue. */
 		BUG();
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -1346,9 +1346,6 @@ static void sync_request_write(mddev_t *
 		tbio->bi_size = r10_bio->sectors << 9;
 		tbio->bi_idx = 0;
 		tbio->bi_phys_segments = 0;
-		tbio->bi_hw_segments = 0;
-		tbio->bi_hw_front_size = 0;
-		tbio->bi_hw_back_size = 0;
 		tbio->bi_flags &= ~(BIO_POOL_MASK - 1);
 		tbio->bi_flags |= 1 << BIO_UPTODATE;
 		tbio->bi_next = NULL;
@@ -1948,7 +1945,6 @@ static sector_t sync_request(mddev_t *md
 		bio->bi_vcnt = 0;
 		bio->bi_idx = 0;
 		bio->bi_phys_segments = 0;
-		bio->bi_hw_segments = 0;
 		bio->bi_size = 0;
 	}
 
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1303,9 +1303,6 @@ static void sync_request_write(mddev_t *
 					sbio->bi_size = r1_bio->sectors << 9;
 					sbio->bi_idx = 0;
 					sbio->bi_phys_segments = 0;
-					sbio->bi_hw_segments = 0;
-					sbio->bi_hw_front_size = 0;
-					sbio->bi_hw_back_size = 0;
 					sbio->bi_flags &= ~(BIO_POOL_MASK - 1);
 					sbio->bi_flags |= 1 << BIO_UPTODATE;
 					sbio->bi_next = NULL;
@@ -1791,7 +1788,6 @@ static sector_t sync_request(mddev_t *md
 		bio->bi_vcnt = 0;
 		bio->bi_idx = 0;
 		bio->bi_phys_segments = 0;
-		bio->bi_hw_segments = 0;
 		bio->bi_size = 0;
 		bio->bi_end_io = NULL;
 		bio->bi_private = NULL;
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -101,6 +101,40 @@
 const char raid6_empty_zero_page[PAGE_SIZE] __attribute__((aligned(256)));
 #endif
 
+/*
+ * We maintain a biased count of active stripes in the bottom 16 bits of
+ * bi_phys_segments, and a count of processed stripes in the upper 16 bits
+ */
+static inline int raid5_bi_phys_segments(struct bio *bio)
+{
+	return bio->bi_phys_segments & 0xffff;
+}
+
+static inline int raid5_bi_hw_segments(struct bio *bio)
+{
+	return (bio->bi_phys_segments >> 16) & 0xffff;
+}
+
+static inline int raid5_dec_bi_phys_segments(struct bio *bio)
+{
+	--bio->bi_phys_segments;
+	return raid5_bi_phys_segments(bio);
+}
+
+static inline int raid5_dec_bi_hw_segments(struct bio *bio)
+{
+	unsigned short val = raid5_bi_hw_segments(bio);
+
+	--val;
+	bio->bi_phys_segments = (val << 16) | raid5_bi_phys_segments(bio);
+	return val;
+}
+
+static inline void raid5_set_bi_hw_segments(struct bio *bio, unsigned int cnt)
+{
+	bio->bi_phys_segments = raid5_bi_phys_segments(bio) || (cnt << 16);
+}
+
 static inline int raid6_next_disk(int disk, int raid_disks)
 {
 	disk++;
@@ -507,7 +541,7 @@ static void ops_complete_biofill(void *s
 			while (rbi && rbi->bi_sector <
 				dev->sector + STRIPE_SECTORS) {
 				rbi2 = r5_next_bio(rbi, dev->sector);
-				if (--rbi->bi_phys_segments == 0) {
+				if (!raid5_dec_bi_phys_segments(rbi)) {
 					rbi->bi_next = return_bi;
 					return_bi = rbi;
 				}
@@ -1725,7 +1759,7 @@ static int add_stripe_bio(struct stripe_
 	if (*bip)
 		bi->bi_next = *bip;
 	*bip = bi;
-	bi->bi_phys_segments ++;
+	bi->bi_phys_segments++;
 	spin_unlock_irq(&conf->device_lock);
 	spin_unlock(&sh->lock);
 
@@ -1819,7 +1853,7 @@ handle_failed_stripe(raid5_conf_t *conf,
 			sh->dev[i].sector + STRIPE_SECTORS) {
 			struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
 			clear_bit(BIO_UPTODATE, &bi->bi_flags);
-			if (--bi->bi_phys_segments == 0) {
+			if (!raid5_dec_bi_phys_segments(bi)) {
 				md_write_end(conf->mddev);
 				bi->bi_next = *return_bi;
 				*return_bi = bi;
@@ -1834,7 +1868,7 @@ handle_failed_stripe(raid5_conf_t *conf,
 		       sh->dev[i].sector + STRIPE_SECTORS) {
 			struct bio *bi2 = r5_next_bio(bi, sh->dev[i].sector);
 			clear_bit(BIO_UPTODATE, &bi->bi_flags);
-			if (--bi->bi_phys_segments == 0) {
+			if (!raid5_dec_bi_phys_segments(bi)) {
 				md_write_end(conf->mddev);
 				bi->bi_next = *return_bi;
 				*return_bi = bi;
@@ -1858,7 +1892,7 @@ handle_failed_stripe(raid5_conf_t *conf,
 				struct bio *nextbi =
 					r5_next_bio(bi, sh->dev[i].sector);
 				clear_bit(BIO_UPTODATE, &bi->bi_flags);
-				if (--bi->bi_phys_segments == 0) {
+				if (!raid5_dec_bi_phys_segments(bi)) {
 					bi->bi_next = *return_bi;
 					*return_bi = bi;
 				}
@@ -2033,7 +2067,7 @@ static void handle_stripe_clean_event(ra
 				while (wbi && wbi->bi_sector <
 					dev->sector + STRIPE_SECTORS) {
 					wbi2 = r5_next_bio(wbi, dev->sector);
-					if (--wbi->bi_phys_segments == 0) {
+					if (!raid5_dec_bi_phys_segments(wbi)) {
 						md_write_end(conf->mddev);
 						wbi->bi_next = *return_bi;
 						*return_bi = wbi;
@@ -2814,7 +2848,7 @@ static bool handle_stripe6(struct stripe
 				copy_data(0, rbi, dev->page, dev->sector);
 				rbi2 = r5_next_bio(rbi, dev->sector);
 				spin_lock_irq(&conf->device_lock);
-				if (--rbi->bi_phys_segments == 0) {
+				if (!raid5_dec_bi_phys_segments(rbi)) {
 					rbi->bi_next = return_bi;
 					return_bi = rbi;
 				}
@@ -3155,8 +3189,11 @@ static struct bio *remove_bio_from_retry
 	if(bi) {
 		conf->retry_read_aligned_list = bi->bi_next;
 		bi->bi_next = NULL;
+		/*
+		 * this sets the active strip count to 1 and the processed
+		 * strip count to zero (upper 8 bits)
+		 */
 		bi->bi_phys_segments = 1; /* biased count of active stripes */
-		bi->bi_hw_segments = 0; /* count of processed stripes */
 	}
 
 	return bi;
@@ -3206,8 +3243,7 @@ static int bio_fits_rdev(struct bio *bi)
 	if ((bi->bi_size>>9) > q->max_sectors)
 		return 0;
 	blk_recount_segments(q, bi);
-	if (bi->bi_phys_segments > q->max_phys_segments ||
-	    bi->bi_hw_segments > q->max_hw_segments)
+	if (bi->bi_phys_segments > q->max_phys_segments)
 		return 0;
 
 	if (q->merge_bvec_fn)
@@ -3468,7 +3504,7 @@ static int make_request(struct request_q
 			
 	}
 	spin_lock_irq(&conf->device_lock);
-	remaining = --bi->bi_phys_segments;
+	remaining = raid5_dec_bi_phys_segments(bi);
 	spin_unlock_irq(&conf->device_lock);
 	if (remaining == 0) {
 
@@ -3752,7 +3788,7 @@ static int  retry_aligned_read(raid5_con
 		     sector += STRIPE_SECTORS,
 		     scnt++) {
 
-		if (scnt < raid_bio->bi_hw_segments)
+		if (scnt < raid5_bi_hw_segments(raid_bio))
 			/* already done this stripe */
 			continue;
 
@@ -3760,7 +3796,7 @@ static int  retry_aligned_read(raid5_con
 
 		if (!sh) {
 			/* failed to get a stripe - must wait */
-			raid_bio->bi_hw_segments = scnt;
+			raid5_set_bi_hw_segments(raid_bio, scnt);
 			conf->retry_read_aligned = raid_bio;
 			return handled;
 		}
@@ -3768,7 +3804,7 @@ static int  retry_aligned_read(raid5_con
 		set_bit(R5_ReadError, &sh->dev[dd_idx].flags);
 		if (!add_stripe_bio(sh, raid_bio, dd_idx, 0)) {
 			release_stripe(sh);
-			raid_bio->bi_hw_segments = scnt;
+			raid5_set_bi_hw_segments(raid_bio, scnt);
 			conf->retry_read_aligned = raid_bio;
 			return handled;
 		}
@@ -3778,7 +3814,7 @@ static int  retry_aligned_read(raid5_con
 		handled++;
 	}
 	spin_lock_irq(&conf->device_lock);
-	remaining = --raid_bio->bi_phys_segments;
+	remaining = raid5_dec_bi_phys_segments(raid_bio);
 	spin_unlock_irq(&conf->device_lock);
 	if (remaining == 0)
 		bio_endio(raid_bio, 0);
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -208,14 +208,6 @@ inline int bio_phys_segments(struct requ
 	return bio->bi_phys_segments;
 }
 
-inline int bio_hw_segments(struct request_queue *q, struct bio *bio)
-{
-	if (unlikely(!bio_flagged(bio, BIO_SEG_VALID)))
-		blk_recount_segments(q, bio);
-
-	return bio->bi_hw_segments;
-}
-
 /**
  * 	__bio_clone	-	clone a bio
  * 	@bio: destination bio
@@ -350,8 +342,7 @@ static int __bio_add_page(struct request
 	 */
 
 	while (bio->bi_phys_segments >= q->max_phys_segments
-	       || bio->bi_hw_segments >= q->max_hw_segments
-	       || BIOVEC_VIRT_OVERSIZE(bio->bi_size)) {
+	       || bio->bi_phys_segments >= q->max_hw_segments) {
 
 		if (retried_segments)
 			return 0;
@@ -395,13 +386,11 @@ static int __bio_add_page(struct request
 	}
 
 	/* If we may be able to merge these biovecs, force a recount */
-	if (bio->bi_vcnt && (BIOVEC_PHYS_MERGEABLE(bvec-1, bvec) ||
-	    BIOVEC_VIRT_MERGEABLE(bvec-1, bvec)))
+	if (bio->bi_vcnt && (BIOVEC_PHYS_MERGEABLE(bvec-1, bvec)))
 		bio->bi_flags &= ~(1 << BIO_SEG_VALID);
 
 	bio->bi_vcnt++;
 	bio->bi_phys_segments++;
-	bio->bi_hw_segments++;
  done:
 	bio->bi_size += len;
 	return len;
@@ -1393,7 +1382,6 @@ EXPORT_SYMBOL(bio_init);
 EXPORT_SYMBOL(__bio_clone);
 EXPORT_SYMBOL(bio_clone);
 EXPORT_SYMBOL(bio_phys_segments);
-EXPORT_SYMBOL(bio_hw_segments);
 EXPORT_SYMBOL(bio_add_page);
 EXPORT_SYMBOL(bio_add_pc_page);
 EXPORT_SYMBOL(bio_get_nr_vecs);
--- a/include/linux/bio.h
+++ b/include/linux/bio.h
@@ -26,21 +26,8 @@
 
 #ifdef CONFIG_BLOCK
 
-/* Platforms may set this to teach the BIO layer about IOMMU hardware. */
 #include <asm/io.h>
 
-#if defined(BIO_VMERGE_MAX_SIZE) && defined(BIO_VMERGE_BOUNDARY)
-#define BIOVEC_VIRT_START_SIZE(x) (bvec_to_phys(x) & (BIO_VMERGE_BOUNDARY - 1))
-#define BIOVEC_VIRT_OVERSIZE(x)	((x) > BIO_VMERGE_MAX_SIZE)
-#else
-#define BIOVEC_VIRT_START_SIZE(x)	0
-#define BIOVEC_VIRT_OVERSIZE(x)		0
-#endif
-
-#ifndef BIO_VMERGE_BOUNDARY
-#define BIO_VMERGE_BOUNDARY	0
-#endif
-
 #define BIO_DEBUG
 
 #ifdef BIO_DEBUG
@@ -88,12 +75,7 @@ struct bio {
 	/* Number of segments in this BIO after
 	 * physical address coalescing is performed.
 	 */
-	unsigned short		bi_phys_segments;
-
-	/* Number of segments after physical and DMA remapping
-	 * hardware coalescing is performed.
-	 */
-	unsigned short		bi_hw_segments;
+	unsigned int		bi_phys_segments;
 
 	unsigned int		bi_size;	/* residual I/O count */
 
@@ -104,14 +86,6 @@ struct bio {
 	unsigned int            bi_seg_front_size;
 	unsigned int            bi_seg_back_size;
 
-	/*
-	 * To keep track of the max hw size, we account for the
-	 * sizes of the first and last virtually mergeable segments
-	 * in this bio
-	 */
-	unsigned int		bi_hw_front_size;
-	unsigned int		bi_hw_back_size;
-
 	unsigned int		bi_max_vecs;	/* max bvl_vecs we can hold */
 
 	struct bio_vec		*bi_io_vec;	/* the actual vec list */
@@ -133,7 +107,7 @@ struct bio {
 #define BIO_UPTODATE	0	/* ok after I/O completion */
 #define BIO_RW_BLOCK	1	/* RW_AHEAD set, and read/write would block */
 #define BIO_EOF		2	/* out-out-bounds error */
-#define BIO_SEG_VALID	3	/* nr_hw_seg valid */
+#define BIO_SEG_VALID	3	/* bi_phys_segments valid */
 #define BIO_CLONED	4	/* doesn't own data */
 #define BIO_BOUNCED	5	/* bio is a bounce bio */
 #define BIO_USER_MAPPED 6	/* contains user pages */
@@ -247,8 +221,6 @@ static inline void *bio_data(struct bio 
 	((bvec_to_phys((vec1)) + (vec1)->bv_len) == bvec_to_phys((vec2)))
 #endif
 
-#define BIOVEC_VIRT_MERGEABLE(vec1, vec2)	\
-	((((bvec_to_phys((vec1)) + (vec1)->bv_len) | bvec_to_phys((vec2))) & (BIO_VMERGE_BOUNDARY - 1)) == 0)
 #define __BIO_SEG_BOUNDARY(addr1, addr2, mask) \
 	(((addr1) | (mask)) == (((addr2) - 1) | (mask)))
 #define BIOVEC_SEG_BOUNDARY(q, b1, b2) \
@@ -346,7 +318,6 @@ extern void bio_free(struct bio *, struc
 extern void bio_endio(struct bio *, int);
 struct request_queue;
 extern int bio_phys_segments(struct request_queue *, struct bio *);
-extern int bio_hw_segments(struct request_queue *, struct bio *);
 
 extern void __bio_clone(struct bio *, struct bio *);
 extern struct bio *bio_clone(struct bio *, gfp_t);
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -54,7 +54,6 @@ enum rq_cmd_type_bits {
 	REQ_TYPE_PM_SUSPEND,		/* suspend request */
 	REQ_TYPE_PM_RESUME,		/* resume request */
 	REQ_TYPE_PM_SHUTDOWN,		/* shutdown request */
-	REQ_TYPE_FLUSH,			/* flush request */
 	REQ_TYPE_SPECIAL,		/* driver defined type */
 	REQ_TYPE_LINUX_BLOCK,		/* generic block layer message */
 	/*
@@ -76,11 +75,8 @@ enum rq_cmd_type_bits {
  *
  */
 enum {
-	/*
-	 * just examples for now
-	 */
 	REQ_LB_OP_EJECT	= 0x40,		/* eject request */
-	REQ_LB_OP_FLUSH = 0x41,		/* flush device */
+	REQ_LB_OP_FLUSH = 0x41,		/* flush request */
 	REQ_LB_OP_DISCARD = 0x42,	/* discard sectors */
 };
 
@@ -193,13 +189,6 @@ struct request {
 	 */
 	unsigned short nr_phys_segments;
 
-	/* Number of scatter-gather addr+len pairs after
-	 * physical and DMA remapping hardware coalescing is performed.
-	 * This is the number of scatter-gather entries the driver
-	 * will actually have to deal with after DMA mapping is done.
-	 */
-	unsigned short nr_hw_segments;
-
 	unsigned short ioprio;
 
 	void *special;
@@ -236,6 +225,11 @@ struct request {
 	struct request *next_rq;
 };
 
+static inline unsigned short req_get_ioprio(struct request *req)
+{
+	return req->ioprio;
+}
+
 /*
  * State information carried for REQ_TYPE_PM_SUSPEND and REQ_TYPE_PM_RESUME
  * requests. Some step values could eventually be made generic.