Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6

27 files changed, 718 insertions, 256 deletions

diff --git a/drivers/md/Kconfig b/drivers/md/Kconfig
index c92c1521546..4540ade6b6b 100644
--- a/drivers/md/Kconfig
+++ b/drivers/md/Kconfig
@@ -215,6 +215,7 @@ config DM_CRYPT
 	tristate "Crypt target support"
 	depends on BLK_DEV_DM && EXPERIMENTAL
 	select CRYPTO
+	select CRYPTO_CBC
 	---help---
 	  This device-mapper target allows you to create a device that
 	  transparently encrypts the data on it. You'll need to activate
diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c
index d6f614738bb..5432d07c074 100644
--- a/drivers/md/bitmap.c
+++ b/drivers/md/bitmap.c
@@ -212,8 +212,8 @@ char *file_path(struct file *file, char *buf, int count)
 	if (!buf)
 		return NULL;
 
-	d = file->f_dentry;
-	v = file->f_vfsmnt;
+	d = file->f_path.dentry;
+	v = file->f_path.mnt;
 
 	buf = d_path(d, v, buf, count);
 
@@ -349,7 +349,7 @@ static struct page *read_page(struct file *file, unsigned long index,
 			      unsigned long count)
 {
 	struct page *page = NULL;
-	struct inode *inode = file->f_dentry->d_inode;
+	struct inode *inode = file->f_path.dentry->d_inode;
 	struct buffer_head *bh;
 	sector_t block;
 
@@ -662,7 +662,7 @@ static void bitmap_file_put(struct bitmap *bitmap)
 	bitmap_file_unmap(bitmap);
 
 	if (file) {
-		struct inode *inode = file->f_dentry->d_inode;
+		struct inode *inode = file->f_path.dentry->d_inode;
 		invalidate_inode_pages(inode->i_mapping);
 		fput(file);
 	}
diff --git a/drivers/md/dm-bio-list.h b/drivers/md/dm-bio-list.h
index bbf4615f0e3..da4349649f7 100644
--- a/drivers/md/dm-bio-list.h
+++ b/drivers/md/dm-bio-list.h
@@ -44,6 +44,20 @@ static inline void bio_list_merge(struct bio_list *bl, struct bio_list *bl2)
 	bl->tail = bl2->tail;
 }
 
+static inline void bio_list_merge_head(struct bio_list *bl,
+				       struct bio_list *bl2)
+{
+	if (!bl2->head)
+		return;
+
+	if (bl->head)
+		bl2->tail->bi_next = bl->head;
+	else
+		bl->tail = bl2->tail;
+
+	bl->head = bl2->head;
+}
+
 static inline struct bio *bio_list_pop(struct bio_list *bl)
 {
 	struct bio *bio = bl->head;
diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index 08a40f4e4f6..4c2471ee054 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -20,6 +20,7 @@
 #include <asm/atomic.h>
 #include <linux/scatterlist.h>
 #include <asm/page.h>
+#include <asm/unaligned.h>
 
 #include "dm.h"
 
@@ -85,7 +86,10 @@ struct crypt_config {
 	 */
 	struct crypt_iv_operations *iv_gen_ops;
 	char *iv_mode;
-	struct crypto_cipher *iv_gen_private;
+	union {
+		struct crypto_cipher *essiv_tfm;
+		int benbi_shift;
+	} iv_gen_private;
 	sector_t iv_offset;
 	unsigned int iv_size;
 
@@ -101,7 +105,7 @@ struct crypt_config {
 #define MIN_POOL_PAGES 32
 #define MIN_BIO_PAGES  8
 
-static kmem_cache_t *_crypt_io_pool;
+static struct kmem_cache *_crypt_io_pool;
 
 /*
  * Different IV generation algorithms:
@@ -113,6 +117,9 @@ static kmem_cache_t *_crypt_io_pool;
  *        encrypted with the bulk cipher using a salt as key. The salt
  *        should be derived from the bulk cipher's key via hashing.
  *
+ * benbi: the 64-bit "big-endian 'narrow block'-count", starting at 1
+ *        (needed for LRW-32-AES and possible other narrow block modes)
+ *
  * plumb: unimplemented, see:
  * http://article.gmane.org/gmane.linux.kernel.device-mapper.dm-crypt/454
  */
@@ -191,21 +198,61 @@ static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti,
 	}
 	kfree(salt);
 
-	cc->iv_gen_private = essiv_tfm;
+	cc->iv_gen_private.essiv_tfm = essiv_tfm;
 	return 0;
 }
 
 static void crypt_iv_essiv_dtr(struct crypt_config *cc)
 {
-	crypto_free_cipher(cc->iv_gen_private);
-	cc->iv_gen_private = NULL;
+	crypto_free_cipher(cc->iv_gen_private.essiv_tfm);
+	cc->iv_gen_private.essiv_tfm = NULL;
 }
 
 static int crypt_iv_essiv_gen(struct crypt_config *cc, u8 *iv, sector_t sector)
 {
 	memset(iv, 0, cc->iv_size);
 	*(u64 *)iv = cpu_to_le64(sector);
-	crypto_cipher_encrypt_one(cc->iv_gen_private, iv, iv);
+	crypto_cipher_encrypt_one(cc->iv_gen_private.essiv_tfm, iv, iv);
+	return 0;
+}
+
+static int crypt_iv_benbi_ctr(struct crypt_config *cc, struct dm_target *ti,
+			      const char *opts)
+{
+	unsigned int bs = crypto_blkcipher_blocksize(cc->tfm);
+	int log = ilog2(bs);
+
+	/* we need to calculate how far we must shift the sector count
+	 * to get the cipher block count, we use this shift in _gen */
+
+	if (1 << log != bs) {
+		ti->error = "cypher blocksize is not a power of 2";
+		return -EINVAL;
+	}
+
+	if (log > 9) {
+		ti->error = "cypher blocksize is > 512";
+		return -EINVAL;
+	}
+
+	cc->iv_gen_private.benbi_shift = 9 - log;
+
+	return 0;
+}
+
+static void crypt_iv_benbi_dtr(struct crypt_config *cc)
+{
+}
+
+static int crypt_iv_benbi_gen(struct crypt_config *cc, u8 *iv, sector_t sector)
+{
+	__be64 val;
+
+	memset(iv, 0, cc->iv_size - sizeof(u64)); /* rest is cleared below */
+
+	val = cpu_to_be64(((u64)sector << cc->iv_gen_private.benbi_shift) + 1);
+	put_unaligned(val, (__be64 *)(iv + cc->iv_size - sizeof(u64)));
+
 	return 0;
 }
 
@@ -219,13 +266,18 @@ static struct crypt_iv_operations crypt_iv_essiv_ops = {
 	.generator = crypt_iv_essiv_gen
 };
 
+static struct crypt_iv_operations crypt_iv_benbi_ops = {
+	.ctr	   = crypt_iv_benbi_ctr,
+	.dtr	   = crypt_iv_benbi_dtr,
+	.generator = crypt_iv_benbi_gen
+};
 
 static int
 crypt_convert_scatterlist(struct crypt_config *cc, struct scatterlist *out,
                           struct scatterlist *in, unsigned int length,
                           int write, sector_t sector)
 {
-	u8 iv[cc->iv_size];
+	u8 iv[cc->iv_size] __attribute__ ((aligned(__alignof__(u64))));
 	struct blkcipher_desc desc = {
 		.tfm = cc->tfm,
 		.info = iv,
@@ -458,11 +510,11 @@ static void dec_pending(struct crypt_io *io, int error)
  * interrupt context.
  */
 static struct workqueue_struct *_kcryptd_workqueue;
-static void kcryptd_do_work(void *data);
+static void kcryptd_do_work(struct work_struct *work);
 
 static void kcryptd_queue_io(struct crypt_io *io)
 {
-	INIT_WORK(&io->work, kcryptd_do_work, io);
+	INIT_WORK(&io->work, kcryptd_do_work);
 	queue_work(_kcryptd_workqueue, &io->work);
 }
 
@@ -618,9 +670,9 @@ static void process_read_endio(struct crypt_io *io)
 	dec_pending(io, crypt_convert(cc, &ctx));
 }
 
-static void kcryptd_do_work(void *data)
+static void kcryptd_do_work(struct work_struct *work)
 {
-	struct crypt_io *io = data;
+	struct crypt_io *io = container_of(work, struct crypt_io, work);
 
 	if (io->post_process)
 		process_read_endio(io);
@@ -768,7 +820,7 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	cc->tfm = tfm;
 
 	/*
-	 * Choose ivmode. Valid modes: "plain", "essiv:<esshash>".
+	 * Choose ivmode. Valid modes: "plain", "essiv:<esshash>", "benbi".
 	 * See comments at iv code
 	 */
 
@@ -778,6 +830,8 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 		cc->iv_gen_ops = &crypt_iv_plain_ops;
 	else if (strcmp(ivmode, "essiv") == 0)
 		cc->iv_gen_ops = &crypt_iv_essiv_ops;
+	else if (strcmp(ivmode, "benbi") == 0)
+		cc->iv_gen_ops = &crypt_iv_benbi_ops;
 	else {
 		ti->error = "Invalid IV mode";
 		goto bad2;
@@ -908,7 +962,7 @@ static int crypt_map(struct dm_target *ti, struct bio *bio,
 	atomic_set(&io->pending, 0);
 	kcryptd_queue_io(io);
 
-	return 0;
+	return DM_MAPIO_SUBMITTED;
 }
 
 static int crypt_status(struct dm_target *ti, status_type_t type,
diff --git a/drivers/md/dm-emc.c b/drivers/md/dm-emc.c
index 2b2d45d7baa..265c467854d 100644
--- a/drivers/md/dm-emc.c
+++ b/drivers/md/dm-emc.c
@@ -40,7 +40,7 @@ static inline void free_bio(struct bio *bio)
 
 static int emc_endio(struct bio *bio, unsigned int bytes_done, int error)
 {
-	struct path *path = bio->bi_private;
+	struct dm_path *path = bio->bi_private;
 
 	if (bio->bi_size)
 		return 1;
@@ -61,7 +61,7 @@ static int emc_endio(struct bio *bio, unsigned int bytes_done, int error)
 	return 0;
 }
 
-static struct bio *get_failover_bio(struct path *path, unsigned data_size)
+static struct bio *get_failover_bio(struct dm_path *path, unsigned data_size)
 {
 	struct bio *bio;
 	struct page *page;
@@ -96,7 +96,7 @@ static struct bio *get_failover_bio(struct path *path, unsigned data_size)
 }
 
 static struct request *get_failover_req(struct emc_handler *h,
-					struct bio *bio, struct path *path)
+					struct bio *bio, struct dm_path *path)
 {
 	struct request *rq;
 	struct block_device *bdev = bio->bi_bdev;
@@ -133,7 +133,7 @@ static struct request *get_failover_req(struct emc_handler *h,
 }
 
 static struct request *emc_trespass_get(struct emc_handler *h,
-					struct path *path)
+					struct dm_path *path)
 {
 	struct bio *bio;
 	struct request *rq;
@@ -191,7 +191,7 @@ static struct request *emc_trespass_get(struct emc_handler *h,
 }
 
 static void emc_pg_init(struct hw_handler *hwh, unsigned bypassed,
-			struct path *path)
+			struct dm_path *path)
 {
 	struct request *rq;
 	struct request_queue *q = bdev_get_queue(path->dev->bdev);
diff --git a/drivers/md/dm-hw-handler.h b/drivers/md/dm-hw-handler.h
index 15f5629e231..32eff28e4ad 100644
--- a/drivers/md/dm-hw-handler.h
+++ b/drivers/md/dm-hw-handler.h
@@ -32,7 +32,7 @@ struct hw_handler_type {
 	void (*destroy) (struct hw_handler *hwh);
 
 	void (*pg_init) (struct hw_handler *hwh, unsigned bypassed,
-			 struct path *path);
+			 struct dm_path *path);
 	unsigned (*error) (struct hw_handler *hwh, struct bio *bio);
 	int (*status) (struct hw_handler *hwh, status_type_t type,
 		       char *result, unsigned int maxlen);
diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
index da663d2ff55..4eb73d39521 100644
--- a/drivers/md/dm-io.c
+++ b/drivers/md/dm-io.c
@@ -92,12 +92,12 @@ void dm_io_put(unsigned int num_pages)
  *---------------------------------------------------------------*/
 static inline void bio_set_region(struct bio *bio, unsigned region)
 {
-	bio->bi_io_vec[bio->bi_max_vecs - 1].bv_len = region;
+	bio->bi_io_vec[bio->bi_max_vecs].bv_len = region;
 }
 
 static inline unsigned bio_get_region(struct bio *bio)
 {
-	return bio->bi_io_vec[bio->bi_max_vecs - 1].bv_len;
+	return bio->bi_io_vec[bio->bi_max_vecs].bv_len;
 }
 
 /*-----------------------------------------------------------------
@@ -136,6 +136,7 @@ static int endio(struct bio *bio, unsigned int done, int error)
 		zero_fill_bio(bio);
 
 	dec_count(io, bio_get_region(bio), error);
+	bio->bi_max_vecs++;
 	bio_put(bio);
 
 	return 0;
@@ -250,16 +251,18 @@ static void do_region(int rw, unsigned int region, struct io_region *where,
 
 	while (remaining) {
 		/*
-		 * Allocate a suitably sized bio, we add an extra
-		 * bvec for bio_get/set_region().
+		 * Allocate a suitably sized-bio: we add an extra
+		 * bvec for bio_get/set_region() and decrement bi_max_vecs
+		 * to hide it from bio_add_page().
 		 */
-		num_bvecs = (remaining / (PAGE_SIZE >> 9)) + 2;
+		num_bvecs = (remaining / (PAGE_SIZE >> SECTOR_SHIFT)) + 2;
 		bio = bio_alloc_bioset(GFP_NOIO, num_bvecs, _bios);
 		bio->bi_sector = where->sector + (where->count - remaining);
 		bio->bi_bdev = where->bdev;
 		bio->bi_end_io = endio;
 		bio->bi_private = io;
 		bio->bi_destructor = dm_bio_destructor;
+		bio->bi_max_vecs--;
 		bio_set_region(bio, region);
 
 		/*
@@ -302,7 +305,7 @@ static void dispatch_io(int rw, unsigned int num_regions,
 	}
 
 	/*
-	 * Drop the extra refence that we were holding to avoid
+	 * Drop the extra reference that we were holding to avoid
 	 * the io being completed too early.
 	 */
 	dec_count(io, 0, 0);
diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index 4510ad8f971..cd6a184536a 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -765,7 +765,7 @@ out:
 static int do_suspend(struct dm_ioctl *param)
 {
 	int r = 0;
-	int do_lockfs = 1;
+	unsigned suspend_flags = DM_SUSPEND_LOCKFS_FLAG;
 	struct mapped_device *md;
 
 	md = find_device(param);
@@ -773,10 +773,12 @@ static int do_suspend(struct dm_ioctl *param)
 		return -ENXIO;
 
 	if (param->flags & DM_SKIP_LOCKFS_FLAG)
-		do_lockfs = 0;
+		suspend_flags &= ~DM_SUSPEND_LOCKFS_FLAG;
+	if (param->flags & DM_NOFLUSH_FLAG)
+		suspend_flags |= DM_SUSPEND_NOFLUSH_FLAG;
 
 	if (!dm_suspended(md))
-		r = dm_suspend(md, do_lockfs);
+		r = dm_suspend(md, suspend_flags);
 
 	if (!r)
 		r = __dev_status(md, param);
@@ -788,7 +790,7 @@ static int do_suspend(struct dm_ioctl *param)
 static int do_resume(struct dm_ioctl *param)
 {
 	int r = 0;
-	int do_lockfs = 1;
+	unsigned suspend_flags = DM_SUSPEND_LOCKFS_FLAG;
 	struct hash_cell *hc;
 	struct mapped_device *md;
 	struct dm_table *new_map;
@@ -814,9 +816,11 @@ static int do_resume(struct dm_ioctl *param)
 	if (new_map) {
 		/* Suspend if it isn't already suspended */
 		if (param->flags & DM_SKIP_LOCKFS_FLAG)
-			do_lockfs = 0;
+			suspend_flags &= ~DM_SUSPEND_LOCKFS_FLAG;
+		if (param->flags & DM_NOFLUSH_FLAG)
+			suspend_flags |= DM_SUSPEND_NOFLUSH_FLAG;
 		if (!dm_suspended(md))
-			dm_suspend(md, do_lockfs);
+			dm_suspend(md, suspend_flags);
 
 		r = dm_swap_table(md, new_map);
 		if (r) {
diff --git a/drivers/md/dm-linear.c b/drivers/md/dm-linear.c
index 00234909b3d..17753d80ad2 100644
--- a/drivers/md/dm-linear.c
+++ b/drivers/md/dm-linear.c
@@ -77,7 +77,7 @@ static int linear_map(struct dm_target *ti, struct bio *bio,
 	bio->bi_bdev = lc->dev->bdev;
 	bio->bi_sector = lc->start + (bio->bi_sector - ti->begin);
 
-	return 1;
+	return DM_MAPIO_REMAPPED;
 }
 
 static int linear_status(struct dm_target *ti, status_type_t type,
@@ -108,7 +108,7 @@ static int linear_ioctl(struct dm_target *ti, struct inode *inode,
 	struct dentry fake_dentry = {};
 
 	fake_file.f_mode = lc->dev->mode;
-	fake_file.f_dentry = &fake_dentry;
+	fake_file.f_path.dentry = &fake_dentry;
 	fake_dentry.d_inode = bdev->bd_inode;
 
 	return blkdev_driver_ioctl(bdev->bd_inode, &fake_file, bdev->bd_disk, cmd, arg);
diff --git a/drivers/md/dm-log.c b/drivers/md/dm-log.c
index 64b764bd02c..6a926135184 100644
--- a/drivers/md/dm-log.c
+++ b/drivers/md/dm-log.c
@@ -466,6 +466,7 @@ static int disk_resume(struct dirty_log *log)
 	/* copy clean across to sync */
 	memcpy(lc->sync_bits, lc->clean_bits, size);
 	lc->sync_count = count_bits32(lc->clean_bits, lc->bitset_uint32_count);
+	lc->sync_search = 0;
 
 	/* set the correct number of regions in the header */
 	lc->header.nr_regions = lc->region_count;
@@ -480,6 +481,13 @@ static uint32_t core_get_region_size(struct dirty_log *log)
 	return lc->region_size;
 }
 
+static int core_resume(struct dirty_log *log)
+{
+	struct log_c *lc = (struct log_c *) log->context;
+	lc->sync_search = 0;
+	return 0;
+}
+
 static int core_is_clean(struct dirty_log *log, region_t region)
 {
 	struct log_c *lc = (struct log_c *) log->context;
@@ -549,16 +557,19 @@ static int core_get_resync_work(struct dirty_log *log, region_t *region)
 	return 1;
 }
 
-static void core_complete_resync_work(struct dirty_log *log, region_t region,
-				      int success)
+static void core_set_region_sync(struct dirty_log *log, region_t region,
+				 int in_sync)
 {
 	struct log_c *lc = (struct log_c *) log->context;
 
 	log_clear_bit(lc, lc->recovering_bits, region);
-	if (success) {
+	if (in_sync) {
 		log_set_bit(lc, lc->sync_bits, region);
                 lc->sync_count++;
-        }
+        } else if (log_test_bit(lc->sync_bits, region)) {
+		lc->sync_count--;
+		log_clear_bit(lc, lc->sync_bits, region);
+	}
 }
 
 static region_t core_get_sync_count(struct dirty_log *log)
@@ -618,6 +629,7 @@ static struct dirty_log_type _core_type = {
 	.module = THIS_MODULE,
 	.ctr = core_ctr,
 	.dtr = core_dtr,
+	.resume = core_resume,
 	.get_region_size = core_get_region_size,
 	.is_clean = core_is_clean,
 	.in_sync = core_in_sync,
@@ -625,7 +637,7 @@ static struct dirty_log_type _core_type = {
 	.mark_region = core_mark_region,
 	.clear_region = core_clear_region,
 	.get_resync_work = core_get_resync_work,
-	.complete_resync_work = core_complete_resync_work,
+	.set_region_sync = core_set_region_sync,
 	.get_sync_count = core_get_sync_count,
 	.status = core_status,
 };
@@ -644,7 +656,7 @@ static struct dirty_log_type _disk_type = {
 	.mark_region = core_mark_region,
 	.clear_region = core_clear_region,
 	.get_resync_work = core_get_resync_work,
-	.complete_resync_work = core_complete_resync_work,
+	.set_region_sync = core_set_region_sync,
 	.get_sync_count = core_get_sync_count,
 	.status = disk_status,
 };
diff --git a/drivers/md/dm-log.h b/drivers/md/dm-log.h
index 5ae5309ebf2..86a301c8daf 100644
--- a/drivers/md/dm-log.h
+++ b/drivers/md/dm-log.h
@@ -90,12 +90,12 @@ struct dirty_log_type {
 	int (*get_resync_work)(struct dirty_log *log, region_t *region);
 
 	/*
-	 * This notifies the log that the resync of an area has
-	 * been completed.  The log should then mark this region
-	 * as CLEAN.
+	 * This notifies the log that the resync status of a region
+	 * has changed.  It also clears the region from the recovering
+	 * list (if present).
 	 */
-	void (*complete_resync_work)(struct dirty_log *log,
-				     region_t region, int success);
+	void (*set_region_sync)(struct dirty_log *log,
+				region_t region, int in_sync);
 
         /*
 	 * Returns the number of regions that are in sync.
diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index d754e0bc6e9..3aa01350696 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -31,7 +31,7 @@ struct pgpath {
 	struct priority_group *pg;	/* Owning PG */
 	unsigned fail_count;		/* Cumulative failure count */
 
-	struct path path;
+	struct dm_path path;
 };
 
 #define path_to_pgpath(__pgp) container_of((__pgp), struct pgpath, path)
@@ -101,11 +101,11 @@ typedef int (*action_fn) (struct pgpath *pgpath);
 
 #define MIN_IOS 256	/* Mempool size */
 
-static kmem_cache_t *_mpio_cache;
+static struct kmem_cache *_mpio_cache;
 
 struct workqueue_struct *kmultipathd;
-static void process_queued_ios(void *data);
-static void trigger_event(void *data);
+static void process_queued_ios(struct work_struct *work);
+static void trigger_event(struct work_struct *work);
 
 
 /*-----------------------------------------------
@@ -173,8 +173,8 @@ static struct multipath *alloc_multipath(struct dm_target *ti)
 		INIT_LIST_HEAD(&m->priority_groups);
 		spin_lock_init(&m->lock);
 		m->queue_io = 1;
-		INIT_WORK(&m->process_queued_ios, process_queued_ios, m);
-		INIT_WORK(&m->trigger_event, trigger_event, m);
+		INIT_WORK(&m->process_queued_ios, process_queued_ios);
+		INIT_WORK(&m->trigger_event, trigger_event);
 		m->mpio_pool = mempool_create_slab_pool(MIN_IOS, _mpio_cache);
 		if (!m->mpio_pool) {
 			kfree(m);
@@ -229,7 +229,7 @@ static void __switch_pg(struct multipath *m, struct pgpath *pgpath)
 
 static int __choose_path_in_pg(struct multipath *m, struct priority_group *pg)
 {
-	struct path *path;
+	struct dm_path *path;
 
 	path = pg->ps.type->select_path(&pg->ps, &m->repeat_count);
 	if (!path)
@@ -282,10 +282,27 @@ failed:
 	m->current_pg = NULL;
 }
 
+/*
+ * Check whether bios must be queued in the device-mapper core rather
+ * than here in the target.
+ *
+ * m->lock must be held on entry.
+ *
+ * If m->queue_if_no_path and m->saved_queue_if_no_path hold the
+ * same value then we are not between multipath_presuspend()
+ * and multipath_resume() calls and we have no need to check
+ * for the DMF_NOFLUSH_SUSPENDING flag.
+ */
+static int __must_push_back(struct multipath *m)
+{
+	return (m->queue_if_no_path != m->saved_queue_if_no_path &&
+		dm_noflush_suspending(m->ti));
+}
+
 static int map_io(struct multipath *m, struct bio *bio, struct mpath_io *mpio,
 		  unsigned was_queued)
 {
-	int r = 1;
+	int r = DM_MAPIO_REMAPPED;
 	unsigned long flags;
 	struct pgpath *pgpath;
 
@@ -310,11 +327,13 @@ static int map_io(struct multipath *m, struct bio *bio, struct mpath_io *mpio,
 		    !m->queue_io)
 			queue_work(kmultipathd, &m->process_queued_ios);
 		pgpath = NULL;
-		r = 0;
-	} else if (!pgpath)
-		r = -EIO;		/* Failed */
-	else
+		r = DM_MAPIO_SUBMITTED;
+	} else if (pgpath)
 		bio->bi_bdev = pgpath->path.dev->bdev;
+	else if (__must_push_back(m))
+		r = DM_MAPIO_REQUEUE;
+	else
+		r = -EIO;	/* Failed */
 
 	mpio->pgpath = pgpath;
 
@@ -372,16 +391,19 @@ static void dispatch_queued_ios(struct multipath *m)
 		r = map_io(m, bio, mpio, 1);
 		if (r < 0)
 			bio_endio(bio, bio->bi_size, r);
-		else if (r == 1)
+		else if (r == DM_MAPIO_REMAPPED)
 			generic_make_request(bio);
+		else if (r == DM_MAPIO_REQUEUE)
+			bio_endio(bio, bio->bi_size, -EIO);
 
 		bio = next;
 	}
 }
 
-static void process_queued_ios(void *data)
+static void process_queued_ios(struct work_struct *work)
 {
-	struct multipath *m = (struct multipath *) data;
+	struct multipath *m =
+		container_of(work, struct multipath, process_queued_ios);
 	struct hw_handler *hwh = &m->hw_handler;
 	struct pgpath *pgpath = NULL;
 	unsigned init_required = 0, must_queue = 1;
@@ -421,9 +443,10 @@ out:
  * An event is triggered whenever a path is taken out of use.
  * Includes path failure and PG bypass.
  */
-static void trigger_event(void *data)
+static void trigger_event(struct work_struct *work)
 {
-	struct multipath *m = (struct multipath *) data;
+	struct multipath *m =
+		container_of(work, struct multipath, trigger_event);
 
 	dm_table_event(m->ti->table);
 }
@@ -781,7 +804,7 @@ static int multipath_map(struct dm_target *ti, struct bio *bio,
 	map_context->ptr = mpio;
 	bio->bi_rw |= (1 << BIO_RW_FAILFAST);
 	r = map_io(m, bio, mpio, 0);
-	if (r < 0)
+	if (r < 0 || r == DM_MAPIO_REQUEUE)
 		mempool_free(mpio, m->mpio_pool);
 
 	return r;
@@ -955,7 +978,7 @@ static int bypass_pg_num(struct multipath *m, const char *pgstr, int bypassed)
 /*
  * pg_init must call this when it has completed its initialisation
  */
-void dm_pg_init_complete(struct path *path, unsigned err_flags)
+void dm_pg_init_complete(struct dm_path *path, unsigned err_flags)
 {
 	struct pgpath *pgpath = path_to_pgpath(path);
 	struct priority_group *pg = pgpath->pg;
@@ -1005,7 +1028,10 @@ static int do_end_io(struct multipath *m, struct bio *bio,
 
 	spin_lock_irqsave(&m->lock, flags);
 	if (!m->nr_valid_paths) {
-		if (!m->queue_if_no_path) {
+		if (__must_push_back(m)) {
+			spin_unlock_irqrestore(&m->lock, flags);
+			return DM_ENDIO_REQUEUE;
+		} else if (!m->queue_if_no_path) {
 			spin_unlock_irqrestore(&m->lock, flags);
 			return -EIO;
 		} else {
@@ -1040,7 +1066,7 @@ static int do_end_io(struct multipath *m, struct bio *bio,
 		queue_work(kmultipathd, &m->process_queued_ios);
 	spin_unlock_irqrestore(&m->lock, flags);
 
-	return 1;	/* io not complete */
+	return DM_ENDIO_INCOMPLETE;	/* io not complete */
 }
 
 static int multipath_end_io(struct dm_target *ti, struct bio *bio,
@@ -1058,7 +1084,7 @@ static int multipath_end_io(struct dm_target *ti, struct bio *bio,
 		if (ps->type->end_io)
 			ps->type->end_io(ps, &pgpath->path);
 	}
-	if (r <= 0)
+	if (r != DM_ENDIO_INCOMPLETE)
 		mempool_free(mpio, m->mpio_pool);
 
 	return r;
@@ -1270,7 +1296,7 @@ static int multipath_ioctl(struct dm_target *ti, struct inode *inode,
 	struct dentry fake_dentry = {};
 	int r = 0;
 
-	fake_file.f_dentry = &fake_dentry;
+	fake_file.f_path.dentry = &fake_dentry;
 
 	spin_lock_irqsave(&m->lock, flags);
 
diff --git a/drivers/md/dm-mpath.h b/drivers/md/dm-mpath.h
index 8a4bf2b6d52..b9cdcbb3ed5 100644
--- a/drivers/md/dm-mpath.h
+++ b/drivers/md/dm-mpath.h
@@ -11,7 +11,7 @@
 
 struct dm_dev;
 
-struct path {
+struct dm_path {
 	struct dm_dev *dev;	/* Read-only */
 	unsigned is_active;	/* Read-only */
 
@@ -20,6 +20,6 @@ struct path {
 };
 
 /* Callback for hwh_pg_init_fn to use when complete */
-void dm_pg_init_complete(struct path *path, unsigned err_flags);
+void dm_pg_init_complete(struct dm_path *path, unsigned err_flags);
 
 #endif
diff --git a/drivers/md/dm-path-selector.h b/drivers/md/dm-path-selector.h
index 732d06a84f8..27357b85d73 100644
--- a/drivers/md/dm-path-selector.h
+++ b/drivers/md/dm-path-selector.h
@@ -44,7 +44,7 @@ struct path_selector_type {
 	 * Add an opaque path object, along with some selector specific
 	 * path args (eg, path priority).
 	 */
-	int (*add_path) (struct path_selector *ps, struct path *path,
+	int (*add_path) (struct path_selector *ps, struct dm_path *path,
 			 int argc, char **argv, char **error);
 
 	/*
@@ -55,27 +55,27 @@ struct path_selector_type {
 	 * calling the function again.  0 means don't call it again unless
 	 * the path fails.
 	 */
-	struct path *(*select_path) (struct path_selector *ps,
+	struct dm_path *(*select_path) (struct path_selector *ps,
 				     unsigned *repeat_count);
 
 	/*
 	 * Notify the selector that a path has failed.
 	 */
-	void (*fail_path) (struct path_selector *ps, struct path *p);
+	void (*fail_path) (struct path_selector *ps, struct dm_path *p);
 
 	/*
 	 * Ask selector to reinstate a path.
 	 */
-	int (*reinstate_path) (struct path_selector *ps, struct path *p);
+	int (*reinstate_path) (struct path_selector *ps, struct dm_path *p);
 
 	/*
 	 * Table content based on parameters added in ps_add_path_fn
 	 * or path selector status
 	 */
-	int (*status) (struct path_selector *ps, struct path *path,
+	int (*status) (struct path_selector *ps, struct dm_path *path,
 		       status_type_t type, char *result, unsigned int maxlen);
 
-	int (*end_io) (struct path_selector *ps, struct path *path);
+	int (*end_io) (struct path_selector *ps, struct dm_path *path);
 };
 
 /* Register a path selector */
diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c
index 48a653b3f51..23a642619be 100644
--- a/drivers/md/dm-raid1.c
+++ b/drivers/md/dm-raid1.c
@@ -344,6 +344,17 @@ static void dispatch_bios(struct mirror_set *ms, struct bio_list *bio_list)
 	}
 }
 
+static void complete_resync_work(struct region *reg, int success)
+{
+	struct region_hash *rh = reg->rh;
+
+	rh->log->type->set_region_sync(rh->log, reg->key, success);
+	dispatch_bios(rh->ms, &reg->delayed_bios);
+	if (atomic_dec_and_test(&rh->recovery_in_flight))
+		wake_up_all(&_kmirrord_recovery_stopped);
+	up(&rh->recovery_count);
+}
+
 static void rh_update_states(struct region_hash *rh)
 {
 	struct region *reg, *next;
@@ -383,11 +394,7 @@ static void rh_update_states(struct region_hash *rh)
 	 */
 	list_for_each_entry_safe (reg, next, &recovered, list) {
 		rh->log->type->clear_region(rh->log, reg->key);
-		rh->log->type->complete_resync_work(rh->log, reg->key, 1);
-		dispatch_bios(rh->ms, &reg->delayed_bios);
-		if (atomic_dec_and_test(&rh->recovery_in_flight))
-			wake_up_all(&_kmirrord_recovery_stopped);
-		up(&rh->recovery_count);
+		complete_resync_work(reg, 1);
 		mempool_free(reg, rh->region_pool);
 	}
 
@@ -883,7 +890,7 @@ static void do_mirror(struct mirror_set *ms)
 	do_writes(ms, &writes);
 }
 
-static void do_work(void *ignored)
+static void do_work(struct work_struct *ignored)
 {
 	struct mirror_set *ms;
 
@@ -1137,7 +1144,7 @@ static int mirror_map(struct dm_target *ti, struct bio *bio,
 
 	if (rw == WRITE) {
 		queue_bio(ms, bio, rw);
-		return 0;
+		return DM_MAPIO_SUBMITTED;
 	}
 
 	r = ms->rh.log->type->in_sync(ms->rh.log,
@@ -1146,7 +1153,7 @@ static int mirror_map(struct dm_target *ti, struct bio *bio,
 		return r;
 
 	if (r == -EWOULDBLOCK)	/* FIXME: ugly */
-		r = 0;
+		r = DM_MAPIO_SUBMITTED;
 
 	/*
 	 * We don't want to fast track a recovery just for a read
@@ -1159,7 +1166,7 @@ static int mirror_map(struct dm_target *ti, struct bio *bio,
 	if (!r) {
 		/* Pass this io over to the daemon */
 		queue_bio(ms, bio, rw);
-		return 0;
+		return DM_MAPIO_SUBMITTED;
 	}
 
 	m = choose_mirror(ms, bio->bi_sector);
@@ -1167,7 +1174,7 @@ static int mirror_map(struct dm_target *ti, struct bio *bio,
 		return -EIO;
 
 	map_bio(ms, m, bio);
-	return 1;
+	return DM_MAPIO_REMAPPED;
 }
 
 static int mirror_end_io(struct dm_target *ti, struct bio *bio,
@@ -1269,7 +1276,7 @@ static int __init dm_mirror_init(void)
 		dm_dirty_log_exit();
 		return r;
 	}
-	INIT_WORK(&_kmirrord_work, do_work, NULL);
+	INIT_WORK(&_kmirrord_work, do_work);
 
 	r = dm_register_target(&mirror_target);
 	if (r < 0) {
diff --git a/drivers/md/dm-round-robin.c b/drivers/md/dm-round-robin.c
index 6f9fcd4db9b..a348a97b65a 100644
--- a/drivers/md/dm-round-robin.c
+++ b/drivers/md/dm-round-robin.c
@@ -21,7 +21,7 @@
  *---------------------------------------------------------------*/
 struct path_info {
 	struct list_head list;
-	struct path *path;
+	struct dm_path *path;
 	unsigned repeat_count;
 };
 
@@ -80,7 +80,7 @@ static void rr_destroy(struct path_selector *ps)
 	ps->context = NULL;
 }
 
-static int rr_status(struct path_selector *ps, struct path *path,
+static int rr_status(struct path_selector *ps, struct dm_path *path,
 		     status_type_t type, char *result, unsigned int maxlen)
 {
 	struct path_info *pi;
@@ -106,7 +106,7 @@ static int rr_status(struct path_selector *ps, struct path *path,
  * Called during initialisation to register each path with an
  * optional repeat_count.
  */
-static int rr_add_path(struct path_selector *ps, struct path *path,
+static int rr_add_path(struct path_selector *ps, struct dm_path *path,
 		       int argc, char **argv, char **error)
 {
 	struct selector *s = (struct selector *) ps->context;
@@ -141,7 +141,7 @@ static int rr_add_path(struct path_selector *ps, struct path *path,
 	return 0;
 }
 
-static void rr_fail_path(struct path_selector *ps, struct path *p)
+static void rr_fail_path(struct path_selector *ps, struct dm_path *p)
 {
 	struct selector *s = (struct selector *) ps->context;
 	struct path_info *pi = p->pscontext;
@@ -149,7 +149,7 @@ static void rr_fail_path(struct path_selector *ps, struct path *p)
 	list_move(&pi->list, &s->invalid_paths);
 }
 
-static int rr_reinstate_path(struct path_selector *ps, struct path *p)
+static int rr_reinstate_path(struct path_selector *ps, struct dm_path *p)
 {
 	struct selector *s = (struct selector *) ps->context;
 	struct path_info *pi = p->pscontext;
@@ -159,7 +159,7 @@ static int rr_reinstate_path(struct path_selector *ps, struct path *p)
 	return 0;
 }
 
-static struct path *rr_select_path(struct path_selector *ps,
+static struct dm_path *rr_select_path(struct path_selector *ps,
 				   unsigned *repeat_count)
 {
 	struct selector *s = (struct selector *) ps->context;
diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index 5281e009407..0821a2b68a7 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -39,8 +39,8 @@
  */
 #define SNAPSHOT_PAGES 256
 
-struct workqueue_struct *ksnapd;
-static void flush_queued_bios(void *data);
+static struct workqueue_struct *ksnapd;
+static void flush_queued_bios(struct work_struct *work);
 
 struct pending_exception {
 	struct exception e;
@@ -88,8 +88,8 @@ struct pending_exception {
  * Hash table mapping origin volumes to lists of snapshots and
  * a lock to protect it
  */
-static kmem_cache_t *exception_cache;
-static kmem_cache_t *pending_cache;
+static struct kmem_cache *exception_cache;
+static struct kmem_cache *pending_cache;
 static mempool_t *pending_pool;
 
 /*
@@ -228,7 +228,7 @@ static int init_exception_table(struct exception_table *et, uint32_t size)
 	return 0;
 }
 
-static void exit_exception_table(struct exception_table *et, kmem_cache_t *mem)
+static void exit_exception_table(struct exception_table *et, struct kmem_cache *mem)
 {
 	struct list_head *slot;
 	struct exception *ex, *next;
@@ -528,7 +528,7 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	}
 
 	bio_list_init(&s->queued_bios);
-	INIT_WORK(&s->queued_bios_work, flush_queued_bios, s);
+	INIT_WORK(&s->queued_bios_work, flush_queued_bios);
 
 	/* Add snapshot to the list of snapshots for this origin */
 	/* Exceptions aren't triggered till snapshot_resume() is called */
@@ -564,6 +564,17 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	return r;
 }
 
+static void __free_exceptions(struct dm_snapshot *s)
+{
+	kcopyd_client_destroy(s->kcopyd_client);
+	s->kcopyd_client = NULL;
+
+	exit_exception_table(&s->pending, pending_cache);
+	exit_exception_table(&s->complete, exception_cache);
+
+	s->store.destroy(&s->store);
+}
+
 static void snapshot_dtr(struct dm_target *ti)
 {
 	struct dm_snapshot *s = (struct dm_snapshot *) ti->private;
@@ -574,13 +585,7 @@ static void snapshot_dtr(struct dm_target *ti)
 	/* After this returns there can be no new kcopyd jobs. */
 	unregister_snapshot(s);
 
-	kcopyd_client_destroy(s->kcopyd_client);
-
-	exit_exception_table(&s->pending, pending_cache);
-	exit_exception_table(&s->complete, exception_cache);
-
-	/* Deallocate memory used */
-	s->store.destroy(&s->store);
+	__free_exceptions(s);
 
 	dm_put_device(ti, s->origin);
 	dm_put_device(ti, s->cow);
@@ -603,9 +608,10 @@ static void flush_bios(struct bio *bio)
 	}
 }
 
-static void flush_queued_bios(void *data)
+static void flush_queued_bios(struct work_struct *work)
 {
-	struct dm_snapshot *s = (struct dm_snapshot *) data;
+	struct dm_snapshot *s =
+		container_of(work, struct dm_snapshot, queued_bios_work);
 	struct bio *queued_bios;
 	unsigned long flags;
 
@@ -867,7 +873,7 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio,
 {
 	struct exception *e;
 	struct dm_snapshot *s = (struct dm_snapshot *) ti->private;
-	int r = 1;
+	int r = DM_MAPIO_REMAPPED;
 	chunk_t chunk;
 	struct pending_exception *pe = NULL;
 
@@ -913,7 +919,7 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio,
 		remap_exception(s, &pe->e, bio);
 		bio_list_add(&pe->snapshot_bios, bio);
 
-		r = 0;
+		r = DM_MAPIO_SUBMITTED;
 
 		if (!pe->started) {
 			/* this is protected by snap->lock */
@@ -991,7 +997,7 @@ static int snapshot_status(struct dm_target *ti, status_type_t type,
  *---------------------------------------------------------------*/
 static int __origin_write(struct list_head *snapshots, struct bio *bio)
 {
-	int r = 1, first = 0;
+	int r = DM_MAPIO_REMAPPED, first = 0;
 	struct dm_snapshot *snap;
 	struct exception *e;
 	struct pending_exception *pe, *next_pe, *primary_pe = NULL;
@@ -1049,7 +1055,7 @@ static int __origin_write(struct list_head *snapshots, struct bio *bio)
 
 			bio_list_add(&primary_pe->origin_bios, bio);
 
-			r = 0;
+			r = DM_MAPIO_SUBMITTED;
 		}
 
 		if (!pe->primary_pe) {
@@ -1098,7 +1104,7 @@ static int __origin_write(struct list_head *snapshots, struct bio *bio)
 static int do_origin(struct dm_dev *origin, struct bio *bio)
 {
 	struct origin *o;
-	int r = 1;
+	int r = DM_MAPIO_REMAPPED;
 
 	down_read(&_origins_lock);
 	o = __lookup_origin(origin->bdev);
@@ -1155,7 +1161,7 @@ static int origin_map(struct dm_target *ti, struct bio *bio,
 		return -EOPNOTSUPP;
 
 	/* Only tell snapshots if this is a write */
-	return (bio_rw(bio) == WRITE) ? do_origin(dev, bio) : 1;
+	return (bio_rw(bio) == WRITE) ? do_origin(dev, bio) : DM_MAPIO_REMAPPED;
 }
 
 #define min_not_zero(l, r) (l == 0) ? r : ((r == 0) ? l : min(l, r))
diff --git a/drivers/md/dm-stripe.c b/drivers/md/dm-stripe.c
index 6c29fcecd89..51f5e076001 100644
--- a/drivers/md/dm-stripe.c
+++ b/drivers/md/dm-stripe.c
@@ -186,7 +186,7 @@ static int stripe_map(struct dm_target *ti, struct bio *bio,
 	bio->bi_bdev = sc->stripe[stripe].dev->bdev;
 	bio->bi_sector = sc->stripe[stripe].physical_start +
 	    (chunk << sc->chunk_shift) + (offset & sc->chunk_mask);
-	return 1;
+	return DM_MAPIO_REMAPPED;
 }
 
 static int stripe_status(struct dm_target *ti,
diff --git a/drivers/md/dm-zero.c b/drivers/md/dm-zero.c
index ea569f7348d..f314d7dc9c2 100644
--- a/drivers/md/dm-zero.c
+++ b/drivers/md/dm-zero.c
@@ -46,7 +46,7 @@ static int zero_map(struct dm_target *ti, struct bio *bio,
 	bio_endio(bio, bio->bi_size, 0);
 
 	/* accepted bio, don't make new request */
-	return 0;
+	return DM_MAPIO_SUBMITTED;
 }
 
 static struct target_type zero_target = {
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index fc4f743f3b5..fe7c56e1043 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -68,10 +68,12 @@ union map_info *dm_get_mapinfo(struct bio *bio)
 #define DMF_FROZEN 2
 #define DMF_FREEING 3
 #define DMF_DELETING 4
+#define DMF_NOFLUSH_SUSPENDING 5
 
 struct mapped_device {
 	struct rw_semaphore io_lock;
 	struct semaphore suspend_lock;
+	spinlock_t pushback_lock;
 	rwlock_t map_lock;
 	atomic_t holders;
 	atomic_t open_count;
@@ -89,7 +91,8 @@ struct mapped_device {
 	 */
 	atomic_t pending;
 	wait_queue_head_t wait;
- 	struct bio_list deferred;
+	struct bio_list deferred;
+	struct bio_list pushback;
 
 	/*
 	 * The current mapping.
@@ -121,8 +124,8 @@ struct mapped_device {
 };
 
 #define MIN_IOS 256
-static kmem_cache_t *_io_cache;
-static kmem_cache_t *_tio_cache;
+static struct kmem_cache *_io_cache;
+static struct kmem_cache *_tio_cache;
 
 static int __init local_init(void)
 {
@@ -444,23 +447,50 @@ int dm_set_geometry(struct mapped_device *md, struct hd_geometry *geo)
  *   you this clearly demarcated crap.
  *---------------------------------------------------------------*/
 
+static int __noflush_suspending(struct mapped_device *md)
+{
+	return test_bit(DMF_NOFLUSH_SUSPENDING, &md->flags);
+}
+
 /*
  * Decrements the number of outstanding ios that a bio has been
  * cloned into, completing the original io if necc.
  */
 static void dec_pending(struct dm_io *io, int error)
 {
-	if (error)
+	unsigned long flags;
+
+	/* Push-back supersedes any I/O errors */
+	if (error && !(io->error > 0 && __noflush_suspending(io->md)))
 		io->error = error;
 
 	if (atomic_dec_and_test(&io->io_count)) {
+		if (io->error == DM_ENDIO_REQUEUE) {
+			/*
+			 * Target requested pushing back the I/O.
+			 * This must be handled before the sleeper on
+			 * suspend queue merges the pushback list.
+			 */
+			spin_lock_irqsave(&io->md->pushback_lock, flags);
+			if (__noflush_suspending(io->md))
+				bio_list_add(&io->md->pushback, io->bio);
+			else
+				/* noflush suspend was interrupted. */
+				io->error = -EIO;
+			spin_unlock_irqrestore(&io->md->pushback_lock, flags);
+		}
+
 		if (end_io_acct(io))
 			/* nudge anyone waiting on suspend queue */
 			wake_up(&io->md->wait);
 
-		blk_add_trace_bio(io->md->queue, io->bio, BLK_TA_COMPLETE);
+		if (io->error != DM_ENDIO_REQUEUE) {
+			blk_add_trace_bio(io->md->queue, io->bio,
+					  BLK_TA_COMPLETE);
+
+			bio_endio(io->bio, io->bio->bi_size, io->error);
+		}
 
-		bio_endio(io->bio, io->bio->bi_size, io->error);
 		free_io(io->md, io);
 	}
 }
@@ -480,12 +510,19 @@ static int clone_endio(struct bio *bio, unsigned int done, int error)
 
 	if (endio) {
 		r = endio(tio->ti, bio, error, &tio->info);
-		if (r < 0)
+		if (r < 0 || r == DM_ENDIO_REQUEUE)
+			/*
+			 * error and requeue request are handled
+			 * in dec_pending().
+			 */
 			error = r;
-
-		else if (r > 0)
-			/* the target wants another shot at the io */
+		else if (r == DM_ENDIO_INCOMPLETE)
+			/* The target will handle the io */
 			return 1;
+		else if (r) {
+			DMWARN("unimplemented target endio return value: %d", r);
+			BUG();
+		}
 	}
 
 	dec_pending(tio->io, error);
@@ -543,7 +580,7 @@ static void __map_bio(struct dm_target *ti, struct bio *clone,
 	atomic_inc(&tio->io->io_count);
 	sector = clone->bi_sector;
 	r = ti->type->map(ti, clone, &tio->info);
-	if (r > 0) {
+	if (r == DM_MAPIO_REMAPPED) {
 		/* the bio has been remapped so dispatch it */
 
 		blk_add_trace_remap(bdev_get_queue(clone->bi_bdev), clone,
@@ -551,10 +588,8 @@ static void __map_bio(struct dm_target *ti, struct bio *clone,
 				    clone->bi_sector);
 
 		generic_make_request(clone);
-	}
-
-	else if (r < 0) {
-		/* error the io and bail out */
+	} else if (r < 0 || r == DM_MAPIO_REQUEUE) {
+		/* error the io and bail out, or requeue it if needed */
 		md = tio->io->md;
 		dec_pending(tio->io, r);
 		/*
@@ -563,6 +598,9 @@ static void __map_bio(struct dm_target *ti, struct bio *clone,
 		clone->bi_private = md->bs;
 		bio_put(clone);
 		free_tio(md, tio);
+	} else if (r) {
+		DMWARN("unimplemented target map return value: %d", r);
+		BUG();
 	}
 }
 
@@ -948,6 +986,7 @@ static struct mapped_device *alloc_dev(int minor)
 	memset(md, 0, sizeof(*md));
 	init_rwsem(&md->io_lock);
 	init_MUTEX(&md->suspend_lock);
+	spin_lock_init(&md->pushback_lock);
 	rwlock_init(&md->map_lock);
 	atomic_set(&md->holders, 1);
 	atomic_set(&md->open_count, 0);
@@ -966,8 +1005,8 @@ static struct mapped_device *alloc_dev(int minor)
 	md->queue->issue_flush_fn = dm_flush_all;
 
 	md->io_pool = mempool_create_slab_pool(MIN_IOS, _io_cache);
- 	if (!md->io_pool)
- 		goto bad2;
+	if (!md->io_pool)
+		goto bad2;
 
 	md->tio_pool = mempool_create_slab_pool(MIN_IOS, _tio_cache);
 	if (!md->tio_pool)
@@ -1275,12 +1314,15 @@ static void unlock_fs(struct mapped_device *md)
  * dm_bind_table, dm_suspend must be called to flush any in
  * flight bios and ensure that any further io gets deferred.
  */
-int dm_suspend(struct mapped_device *md, int do_lockfs)
+int dm_suspend(struct mapped_device *md, unsigned suspend_flags)
 {
 	struct dm_table *map = NULL;
+	unsigned long flags;
 	DECLARE_WAITQUEUE(wait, current);
 	struct bio *def;
 	int r = -EINVAL;
+	int do_lockfs = suspend_flags & DM_SUSPEND_LOCKFS_FLAG ? 1 : 0;
+	int noflush = suspend_flags & DM_SUSPEND_NOFLUSH_FLAG ? 1 : 0;
 
 	down(&md->suspend_lock);
 
@@ -1289,6 +1331,13 @@ int dm_suspend(struct mapped_device *md, int do_lockfs)
 
 	map = dm_get_table(md);
 
+	/*
+	 * DMF_NOFLUSH_SUSPENDING must be set before presuspend.
+	 * This flag is cleared before dm_suspend returns.
+	 */
+	if (noflush)
+		set_bit(DMF_NOFLUSH_SUSPENDING, &md->flags);
+
 	/* This does not get reverted if there's an error later. */
 	dm_table_presuspend_targets(map);
 
@@ -1296,11 +1345,14 @@ int dm_suspend(struct mapped_device *md, int do_lockfs)
 	if (!md->suspended_bdev) {
 		DMWARN("bdget failed in dm_suspend");
 		r = -ENOMEM;
-		goto out;
+		goto flush_and_out;
 	}
 
-	/* Flush I/O to the device. */
-	if (do_lockfs) {
+	/*
+	 * Flush I/O to the device.
+	 * noflush supersedes do_lockfs, because lock_fs() needs to flush I/Os.
+	 */
+	if (do_lockfs && !noflush) {
 		r = lock_fs(md);
 		if (r)
 			goto out;
@@ -1336,6 +1388,14 @@ int dm_suspend(struct mapped_device *md, int do_lockfs)
 	down_write(&md->io_lock);
 	remove_wait_queue(&md->wait, &wait);
 
+	if (noflush) {
+		spin_lock_irqsave(&md->pushback_lock, flags);
+		clear_bit(DMF_NOFLUSH_SUSPENDING, &md->flags);
+		bio_list_merge_head(&md->deferred, &md->pushback);
+		bio_list_init(&md->pushback);
+		spin_unlock_irqrestore(&md->pushback_lock, flags);
+	}
+
 	/* were we interrupted ? */
 	r = -EINTR;
 	if (atomic_read(&md->pending)) {
@@ -1344,7 +1404,7 @@ int dm_suspend(struct mapped_device *md, int do_lockfs)
 		__flush_deferred_io(md, def);
 		up_write(&md->io_lock);
 		unlock_fs(md);
-		goto out;
+		goto out; /* pushback list is already flushed, so skip flush */
 	}
 	up_write(&md->io_lock);
 
@@ -1354,6 +1414,25 @@ int dm_suspend(struct mapped_device *md, int do_lockfs)
 
 	r = 0;
 
+flush_and_out:
+	if (r && noflush) {
+		/*
+		 * Because there may be already I/Os in the pushback list,
+		 * flush them before return.
+		 */
+		down_write(&md->io_lock);
+
+		spin_lock_irqsave(&md->pushback_lock, flags);
+		clear_bit(DMF_NOFLUSH_SUSPENDING, &md->flags);
+		bio_list_merge_head(&md->deferred, &md->pushback);
+		bio_list_init(&md->pushback);
+		spin_unlock_irqrestore(&md->pushback_lock, flags);
+
+		def = bio_list_get(&md->deferred);
+		__flush_deferred_io(md, def);
+		up_write(&md->io_lock);
+	}
+
 out:
 	if (r && md->suspended_bdev) {
 		bdput(md->suspended_bdev);
@@ -1440,6 +1519,17 @@ int dm_suspended(struct mapped_device *md)
 	return test_bit(DMF_SUSPENDED, &md->flags);
 }
 
+int dm_noflush_suspending(struct dm_target *ti)
+{
+	struct mapped_device *md = dm_table_get_md(ti->table);
+	int r = __noflush_suspending(md);
+
+	dm_put(md);
+
+	return r;
+}
+EXPORT_SYMBOL_GPL(dm_noflush_suspending);
+
 static struct block_device_operations dm_blk_dops = {
 	.open = dm_blk_open,
 	.release = dm_blk_close,
diff --git a/drivers/md/dm.h b/drivers/md/dm.h
index a48ec5e3c1f..2f796b1436b 100644
--- a/drivers/md/dm.h
+++ b/drivers/md/dm.h
@@ -33,6 +33,25 @@
 #define SECTOR_SHIFT 9
 
 /*
+ * Definitions of return values from target end_io function.
+ */
+#define DM_ENDIO_INCOMPLETE	1
+#define DM_ENDIO_REQUEUE	2
+
+/*
+ * Definitions of return values from target map function.
+ */
+#define DM_MAPIO_SUBMITTED	0
+#define DM_MAPIO_REMAPPED	1
+#define DM_MAPIO_REQUEUE	DM_ENDIO_REQUEUE
+
+/*
+ * Suspend feature flags
+ */
+#define DM_SUSPEND_LOCKFS_FLAG		(1 << 0)
+#define DM_SUSPEND_NOFLUSH_FLAG		(1 << 1)
+
+/*
  * List of devices that a metadevice uses and should open/close.
  */
 struct dm_dev {
diff --git a/drivers/md/faulty.c b/drivers/md/faulty.c
index a7a5ab55433..4ebd0f2a75e 100644
--- a/drivers/md/faulty.c
+++ b/drivers/md/faulty.c
@@ -173,7 +173,7 @@ static int make_request(request_queue_t *q, struct bio *bio)
 	conf_t *conf = (conf_t*)mddev->private;
 	int failit = 0;
 
-	if (bio->bi_rw & 1) {
+	if (bio_data_dir(bio) == WRITE) {
 		/* write request */
 		if (atomic_read(&conf->counters[WriteAll])) {
 			/* special case - don't decrement, don't generic_make_request,
diff --git a/drivers/md/kcopyd.c b/drivers/md/kcopyd.c
index f1db6eff485..b46f6c575f7 100644
--- a/drivers/md/kcopyd.c
+++ b/drivers/md/kcopyd.c
@@ -203,7 +203,7 @@ struct kcopyd_job {
 /* FIXME: this should scale with the number of pages */
 #define MIN_JOBS 512
 
-static kmem_cache_t *_job_cache;
+static struct kmem_cache *_job_cache;
 static mempool_t *_job_pool;
 
 /*
@@ -417,7 +417,7 @@ static int process_jobs(struct list_head *jobs, int (*fn) (struct kcopyd_job *))
 /*
  * kcopyd does this every time it's woken up.
  */
-static void do_work(void *ignored)
+static void do_work(struct work_struct *ignored)
 {
 	/*
 	 * The order that these are called is *very* important.
@@ -628,7 +628,7 @@ static int kcopyd_init(void)
 	}
 
 	kcopyd_clients++;
-	INIT_WORK(&_kcopyd_work, do_work, NULL);
+	INIT_WORK(&_kcopyd_work, do_work);
 	mutex_unlock(&kcopyd_init_lock);
 	return 0;
 }
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 8cbf9c9df1c..d1cb45f6d6a 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -39,10 +39,10 @@
 #include <linux/raid/bitmap.h>
 #include <linux/sysctl.h>
 #include <linux/buffer_head.h> /* for invalidate_bdev */
-#include <linux/suspend.h>
 #include <linux/poll.h>
 #include <linux/mutex.h>
 #include <linux/ctype.h>
+#include <linux/freezer.h>
 
 #include <linux/init.h>
 
@@ -1413,7 +1413,7 @@ static int lock_rdev(mdk_rdev_t *rdev, dev_t dev)
 	struct block_device *bdev;
 	char b[BDEVNAME_SIZE];
 
-	bdev = open_partition_by_devnum(dev, FMODE_READ|FMODE_WRITE);
+	bdev = open_by_devnum(dev, FMODE_READ|FMODE_WRITE);
 	if (IS_ERR(bdev)) {
 		printk(KERN_ERR "md: could not open %s.\n",
 			__bdevname(dev, b));
@@ -1423,7 +1423,7 @@ static int lock_rdev(mdk_rdev_t *rdev, dev_t dev)
 	if (err) {
 		printk(KERN_ERR "md: could not bd_claim %s.\n",
 			bdevname(bdev, b));
-		blkdev_put_partition(bdev);
+		blkdev_put(bdev);
 		return err;
 	}
 	rdev->bdev = bdev;
@@ -1437,7 +1437,7 @@ static void unlock_rdev(mdk_rdev_t *rdev)
 	if (!bdev)
 		MD_BUG();
 	bd_release(bdev);
-	blkdev_put_partition(bdev);
+	blkdev_put(bdev);
 }
 
 void md_autodetect_dev(dev_t dev);
@@ -1792,7 +1792,8 @@ state_store(mdk_rdev_t *rdev, const char *buf, size_t len)
 		else {
 			mddev_t *mddev = rdev->mddev;
 			kick_rdev_from_array(rdev);
-			md_update_sb(mddev, 1);
+			if (mddev->pers)
+				md_update_sb(mddev, 1);
 			md_new_event(mddev);
 			err = 0;
 		}
@@ -2004,6 +2005,7 @@ static mdk_rdev_t *md_import_device(dev_t newdev, int super_format, int super_mi
 
 	rdev->desc_nr = -1;
 	rdev->saved_raid_disk = -1;
+	rdev->raid_disk = -1;
 	rdev->flags = 0;
 	rdev->data_offset = 0;
 	rdev->sb_events = 0;
@@ -2233,7 +2235,6 @@ static int update_raid_disks(mddev_t *mddev, int raid_disks);
 static ssize_t
 raid_disks_store(mddev_t *mddev, const char *buf, size_t len)
 {
-	/* can only set raid_disks if array is not yet active */
 	char *e;
 	int rv = 0;
 	unsigned long n = simple_strtoul(buf, &e, 10);
@@ -2631,7 +2632,7 @@ metadata_store(mddev_t *mddev, const char *buf, size_t len)
 		return -EINVAL;
 	buf = e+1;
 	minor = simple_strtoul(buf, &e, 10);
-	if (e==buf || *e != '\n')
+	if (e==buf || (*e && *e != '\n') )
 		return -EINVAL;
 	if (major >= sizeof(super_types)/sizeof(super_types[0]) ||
 	    super_types[major].name == NULL)
@@ -3314,6 +3315,10 @@ static int do_md_stop(mddev_t * mddev, int mode)
 
 			module_put(mddev->pers->owner);
 			mddev->pers = NULL;
+
+			set_capacity(disk, 0);
+			mddev->changed = 1;
+
 			if (mddev->ro)
 				mddev->ro = 0;
 		}
@@ -3333,7 +3338,7 @@ static int do_md_stop(mddev_t * mddev, int mode)
 	if (mode == 0) {
 		mdk_rdev_t *rdev;
 		struct list_head *tmp;
-		struct gendisk *disk;
+
 		printk(KERN_INFO "md: %s stopped.\n", mdname(mddev));
 
 		bitmap_destroy(mddev);
@@ -3358,10 +3363,6 @@ static int do_md_stop(mddev_t * mddev, int mode)
 		mddev->raid_disks = 0;
 		mddev->recovery_cp = 0;
 
-		disk = mddev->gendisk;
-		if (disk)
-			set_capacity(disk, 0);
-		mddev->changed = 1;
 	} else if (mddev->pers)
 		printk(KERN_INFO "md: %s switched to read-only mode.\n",
 			mdname(mddev));
@@ -3371,6 +3372,7 @@ out:
 	return err;
 }
 
+#ifndef MODULE
 static void autorun_array(mddev_t *mddev)
 {
 	mdk_rdev_t *rdev;
@@ -3485,6 +3487,7 @@ static void autorun_devices(int part)
 	}
 	printk(KERN_INFO "md: ... autorun DONE.\n");
 }
+#endif /* !MODULE */
 
 static int get_version(void __user * arg)
 {
@@ -3722,6 +3725,7 @@ static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 		if (err)
 			export_rdev(rdev);
 
+		md_update_sb(mddev, 1);
 		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 		md_wakeup_thread(mddev->thread);
 		return err;
@@ -3977,6 +3981,7 @@ static int set_array_info(mddev_t * mddev, mdu_array_info_t *info)
 		mddev->major_version = info->major_version;
 		mddev->minor_version = info->minor_version;
 		mddev->patch_version = info->patch_version;
+		mddev->persistent = !info->not_persistent;
 		return 0;
 	}
 	mddev->major_version = MD_MAJOR_VERSION;
@@ -4301,9 +4306,10 @@ static int md_ioctl(struct inode *inode, struct file *file,
 	 * Commands querying/configuring an existing array:
 	 */
 	/* if we are not initialised yet, only ADD_NEW_DISK, STOP_ARRAY,
-	 * RUN_ARRAY, and SET_BITMAP_FILE are allowed */
+	 * RUN_ARRAY, and GET_ and SET_BITMAP_FILE are allowed */
 	if (!mddev->raid_disks && cmd != ADD_NEW_DISK && cmd != STOP_ARRAY
-			&& cmd != RUN_ARRAY && cmd != SET_BITMAP_FILE) {
+			&& cmd != RUN_ARRAY && cmd != SET_BITMAP_FILE
+	    		&& cmd != GET_BITMAP_FILE) {
 		err = -ENODEV;
 		goto abort_unlock;
 	}
@@ -4423,7 +4429,7 @@ static int md_open(struct inode *inode, struct file *file)
 	mddev_t *mddev = inode->i_bdev->bd_disk->private_data;
 	int err;
 
-	if ((err = mddev_lock(mddev)))
+	if ((err = mutex_lock_interruptible_nested(&mddev->reconfig_mutex, 1)))
 		goto out;
 
 	err = 0;
@@ -4846,8 +4852,8 @@ static int md_seq_show(struct seq_file *seq, void *v)
 				chunk_kb ? "KB" : "B");
 			if (bitmap->file) {
 				seq_printf(seq, ", file: ");
-				seq_path(seq, bitmap->file->f_vfsmnt,
-					 bitmap->file->f_dentry," \t\n");
+				seq_path(seq, bitmap->file->f_path.mnt,
+					 bitmap->file->f_path.dentry," \t\n");
 			}
 
 			seq_printf(seq, "\n");
@@ -5273,7 +5279,6 @@ void md_do_sync(mddev_t *mddev)
 	mddev->pers->sync_request(mddev, max_sectors, &skipped, 1);
 
 	if (!test_bit(MD_RECOVERY_ERR, &mddev->recovery) &&
-	    test_bit(MD_RECOVERY_SYNC, &mddev->recovery) &&
 	    !test_bit(MD_RECOVERY_CHECK, &mddev->recovery) &&
 	    mddev->curr_resync > 2) {
 		if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
@@ -5297,6 +5302,7 @@ void md_do_sync(mddev_t *mddev)
 					rdev->recovery_offset = mddev->curr_resync;
 		}
 	}
+	set_bit(MD_CHANGE_DEVS, &mddev->flags);
 
  skip:
 	mddev->curr_resync = 0;
@@ -5593,7 +5599,7 @@ static void autostart_arrays(int part)
 	autorun_devices(part);
 }
 
-#endif
+#endif /* !MODULE */
 
 static __exit void md_exit(void)
 {
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 656fae912fe..164b25dca10 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -775,6 +775,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 	struct bio_list bl;
 	struct page **behind_pages = NULL;
 	const int rw = bio_data_dir(bio);
+	const int do_sync = bio_sync(bio);
 	int do_barriers;
 
 	/*
@@ -835,7 +836,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 		read_bio->bi_sector = r1_bio->sector + mirror->rdev->data_offset;
 		read_bio->bi_bdev = mirror->rdev->bdev;
 		read_bio->bi_end_io = raid1_end_read_request;
-		read_bio->bi_rw = READ;
+		read_bio->bi_rw = READ | do_sync;
 		read_bio->bi_private = r1_bio;
 
 		generic_make_request(read_bio);
@@ -906,7 +907,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 		mbio->bi_sector	= r1_bio->sector + conf->mirrors[i].rdev->data_offset;
 		mbio->bi_bdev = conf->mirrors[i].rdev->bdev;
 		mbio->bi_end_io	= raid1_end_write_request;
-		mbio->bi_rw = WRITE | do_barriers;
+		mbio->bi_rw = WRITE | do_barriers | do_sync;
 		mbio->bi_private = r1_bio;
 
 		if (behind_pages) {
@@ -941,6 +942,8 @@ static int make_request(request_queue_t *q, struct bio * bio)
 	blk_plug_device(mddev->queue);
 	spin_unlock_irqrestore(&conf->device_lock, flags);
 
+	if (do_sync)
+		md_wakeup_thread(mddev->thread);
 #if 0
 	while ((bio = bio_list_pop(&bl)) != NULL)
 		generic_make_request(bio);
@@ -1541,6 +1544,7 @@ static void raid1d(mddev_t *mddev)
 			 * We already have a nr_pending reference on these rdevs.
 			 */
 			int i;
+			const int do_sync = bio_sync(r1_bio->master_bio);
 			clear_bit(R1BIO_BarrierRetry, &r1_bio->state);
 			clear_bit(R1BIO_Barrier, &r1_bio->state);
 			for (i=0; i < conf->raid_disks; i++)
@@ -1561,7 +1565,7 @@ static void raid1d(mddev_t *mddev)
 						conf->mirrors[i].rdev->data_offset;
 					bio->bi_bdev = conf->mirrors[i].rdev->bdev;
 					bio->bi_end_io = raid1_end_write_request;
-					bio->bi_rw = WRITE;
+					bio->bi_rw = WRITE | do_sync;
 					bio->bi_private = r1_bio;
 					r1_bio->bios[i] = bio;
 					generic_make_request(bio);
@@ -1593,6 +1597,7 @@ static void raid1d(mddev_t *mddev)
 				       (unsigned long long)r1_bio->sector);
 				raid_end_bio_io(r1_bio);
 			} else {
+				const int do_sync = bio_sync(r1_bio->master_bio);
 				r1_bio->bios[r1_bio->read_disk] =
 					mddev->ro ? IO_BLOCKED : NULL;
 				r1_bio->read_disk = disk;
@@ -1608,7 +1613,7 @@ static void raid1d(mddev_t *mddev)
 				bio->bi_sector = r1_bio->sector + rdev->data_offset;
 				bio->bi_bdev = rdev->bdev;
 				bio->bi_end_io = raid1_end_read_request;
-				bio->bi_rw = READ;
+				bio->bi_rw = READ | do_sync;
 				bio->bi_private = r1_bio;
 				unplug = 1;
 				generic_make_request(bio);
@@ -1736,7 +1741,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 		/* take from bio_init */
 		bio->bi_next = NULL;
 		bio->bi_flags |= 1 << BIO_UPTODATE;
-		bio->bi_rw = 0;
+		bio->bi_rw = READ;
 		bio->bi_vcnt = 0;
 		bio->bi_idx = 0;
 		bio->bi_phys_segments = 0;
@@ -1951,6 +1956,7 @@ static int run(mddev_t *mddev)
 		    !test_bit(In_sync, &disk->rdev->flags)) {
 			disk->head_position = 0;
 			mddev->degraded++;
+			conf->fullsync = 1;
 		}
 	}
 	if (mddev->degraded == conf->raid_disks) {
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 7492d6033ac..a9401c017e3 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -782,6 +782,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 	int i;
 	int chunk_sects = conf->chunk_mask + 1;
 	const int rw = bio_data_dir(bio);
+	const int do_sync = bio_sync(bio);
 	struct bio_list bl;
 	unsigned long flags;
 
@@ -863,7 +864,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 			mirror->rdev->data_offset;
 		read_bio->bi_bdev = mirror->rdev->bdev;
 		read_bio->bi_end_io = raid10_end_read_request;
-		read_bio->bi_rw = READ;
+		read_bio->bi_rw = READ | do_sync;
 		read_bio->bi_private = r10_bio;
 
 		generic_make_request(read_bio);
@@ -909,7 +910,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 			conf->mirrors[d].rdev->data_offset;
 		mbio->bi_bdev = conf->mirrors[d].rdev->bdev;
 		mbio->bi_end_io	= raid10_end_write_request;
-		mbio->bi_rw = WRITE;
+		mbio->bi_rw = WRITE | do_sync;
 		mbio->bi_private = r10_bio;
 
 		atomic_inc(&r10_bio->remaining);
@@ -922,6 +923,9 @@ static int make_request(request_queue_t *q, struct bio * bio)
 	blk_plug_device(mddev->queue);
 	spin_unlock_irqrestore(&conf->device_lock, flags);
 
+	if (do_sync)
+		md_wakeup_thread(mddev->thread);
+
 	return 0;
 }
 
@@ -1563,6 +1567,7 @@ static void raid10d(mddev_t *mddev)
 				       (unsigned long long)r10_bio->sector);
 				raid_end_bio_io(r10_bio);
 			} else {
+				const int do_sync = bio_sync(r10_bio->master_bio);
 				rdev = conf->mirrors[mirror].rdev;
 				if (printk_ratelimit())
 					printk(KERN_ERR "raid10: %s: redirecting sector %llu to"
@@ -1574,7 +1579,7 @@ static void raid10d(mddev_t *mddev)
 				bio->bi_sector = r10_bio->devs[r10_bio->read_slot].addr
 					+ rdev->data_offset;
 				bio->bi_bdev = rdev->bdev;
-				bio->bi_rw = READ;
+				bio->bi_rw = READ | do_sync;
 				bio->bi_private = r10_bio;
 				bio->bi_end_io = raid10_end_read_request;
 				unplug = 1;
@@ -1785,7 +1790,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 						biolist = bio;
 						bio->bi_private = r10_bio;
 						bio->bi_end_io = end_sync_read;
-						bio->bi_rw = 0;
+						bio->bi_rw = READ;
 						bio->bi_sector = r10_bio->devs[j].addr +
 							conf->mirrors[d].rdev->data_offset;
 						bio->bi_bdev = conf->mirrors[d].rdev->bdev;
@@ -1801,7 +1806,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 						biolist = bio;
 						bio->bi_private = r10_bio;
 						bio->bi_end_io = end_sync_write;
-						bio->bi_rw = 1;
+						bio->bi_rw = WRITE;
 						bio->bi_sector = r10_bio->devs[k].addr +
 							conf->mirrors[i].rdev->data_offset;
 						bio->bi_bdev = conf->mirrors[i].rdev->bdev;
@@ -1870,7 +1875,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
 			biolist = bio;
 			bio->bi_private = r10_bio;
 			bio->bi_end_io = end_sync_read;
-			bio->bi_rw = 0;
+			bio->bi_rw = READ;
 			bio->bi_sector = r10_bio->devs[i].addr +
 				conf->mirrors[d].rdev->data_offset;
 			bio->bi_bdev = conf->mirrors[d].rdev->bdev;
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 69c3e201fa3..be008f034ad 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -134,6 +134,8 @@ static void __release_stripe(raid5_conf_t *conf, struct stripe_head *sh)
 			if (!test_bit(STRIPE_EXPANDING, &sh->state)) {
 				list_add_tail(&sh->lru, &conf->inactive_list);
 				wake_up(&conf->wait_for_stripe);
+				if (conf->retry_read_aligned)
+					md_wakeup_thread(conf->mddev->thread);
 			}
 		}
 	}
@@ -348,7 +350,7 @@ static int grow_one_stripe(raid5_conf_t *conf)
 
 static int grow_stripes(raid5_conf_t *conf, int num)
 {
-	kmem_cache_t *sc;
+	struct kmem_cache *sc;
 	int devs = conf->raid_disks;
 
 	sprintf(conf->cache_name[0], "raid5/%s", mdname(conf->mddev));
@@ -397,7 +399,7 @@ static int resize_stripes(raid5_conf_t *conf, int newsize)
 	LIST_HEAD(newstripes);
 	struct disk_info *ndisks;
 	int err = 0;
-	kmem_cache_t *sc;
+	struct kmem_cache *sc;
 	int i;
 
 	if (newsize <= conf->pool_size)
@@ -542,35 +544,7 @@ static int raid5_end_read_request(struct bio * bi, unsigned int bytes_done,
 	}
 
 	if (uptodate) {
-#if 0
-		struct bio *bio;
-		unsigned long flags;
-		spin_lock_irqsave(&conf->device_lock, flags);
-		/* we can return a buffer if we bypassed the cache or
-		 * if the top buffer is not in highmem.  If there are
-		 * multiple buffers, leave the extra work to
-		 * handle_stripe
-		 */
-		buffer = sh->bh_read[i];
-		if (buffer &&
-		    (!PageHighMem(buffer->b_page)
-		     || buffer->b_page == bh->b_page )
-			) {
-			sh->bh_read[i] = buffer->b_reqnext;
-			buffer->b_reqnext = NULL;
-		} else
-			buffer = NULL;
-		spin_unlock_irqrestore(&conf->device_lock, flags);
-		if (sh->bh_page[i]==bh->b_page)
-			set_buffer_uptodate(bh);
-		if (buffer) {
-			if (buffer->b_page != bh->b_page)
-				memcpy(buffer->b_data, bh->b_data, bh->b_size);
-			buffer->b_end_io(buffer, 1);
-		}
-#else
 		set_bit(R5_UPTODATE, &sh->dev[i].flags);
-#endif
 		if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
 			rdev = conf->disks[i].rdev;
 			printk(KERN_INFO "raid5:%s: read error corrected (%lu sectors at %llu on %s)\n",
@@ -616,14 +590,6 @@ static int raid5_end_read_request(struct bio * bi, unsigned int bytes_done,
 		}
 	}
 	rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
-#if 0
-	/* must restore b_page before unlocking buffer... */
-	if (sh->bh_page[i] != bh->b_page) {
-		bh->b_page = sh->bh_page[i];
-		bh->b_data = page_address(bh->b_page);
-		clear_buffer_uptodate(bh);
-	}
-#endif
 	clear_bit(R5_LOCKED, &sh->dev[i].flags);
 	set_bit(STRIPE_HANDLE, &sh->state);
 	release_stripe(sh);
@@ -821,7 +787,8 @@ static sector_t raid5_compute_sector(sector_t r_sector, unsigned int raid_disks,
 static sector_t compute_blocknr(struct stripe_head *sh, int i)
 {
 	raid5_conf_t *conf = sh->raid_conf;
-	int raid_disks = sh->disks, data_disks = raid_disks - 1;
+	int raid_disks = sh->disks;
+	int data_disks = raid_disks - conf->max_degraded;
 	sector_t new_sector = sh->sector, check;
 	int sectors_per_chunk = conf->chunk_size >> 9;
 	sector_t stripe;
@@ -857,7 +824,6 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i)
 		}
 		break;
 	case 6:
-		data_disks = raid_disks - 2;
 		if (i == raid6_next_disk(sh->pd_idx, raid_disks))
 			return 0; /* It is the Q disk */
 		switch (conf->algorithm) {
@@ -1353,8 +1319,10 @@ static int stripe_to_pdidx(sector_t stripe, raid5_conf_t *conf, int disks)
 	int pd_idx, dd_idx;
 	int chunk_offset = sector_div(stripe, sectors_per_chunk);
 
-	raid5_compute_sector(stripe*(disks-1)*sectors_per_chunk
-			     + chunk_offset, disks, disks-1, &dd_idx, &pd_idx, conf);
+	raid5_compute_sector(stripe * (disks - conf->max_degraded)
+			     *sectors_per_chunk + chunk_offset,
+			     disks, disks - conf->max_degraded,
+			     &dd_idx, &pd_idx, conf);
 	return pd_idx;
 }
 
@@ -1615,15 +1583,6 @@ static void handle_stripe5(struct stripe_head *sh)
 				} else if (test_bit(R5_Insync, &dev->flags)) {
 					set_bit(R5_LOCKED, &dev->flags);
 					set_bit(R5_Wantread, &dev->flags);
-#if 0
-					/* if I am just reading this block and we don't have
-					   a failed drive, or any pending writes then sidestep the cache */
-					if (sh->bh_read[i] && !sh->bh_read[i]->b_reqnext &&
-					    ! syncing && !failed && !to_write) {
-						sh->bh_cache[i]->b_page =  sh->bh_read[i]->b_page;
-						sh->bh_cache[i]->b_data =  sh->bh_read[i]->b_data;
-					}
-#endif
 					locked++;
 					PRINTK("Reading block %d (sync=%d)\n", 
 						i, syncing);
@@ -1641,9 +1600,6 @@ static void handle_stripe5(struct stripe_head *sh)
 			dev = &sh->dev[i];
 			if ((dev->towrite || i == sh->pd_idx) &&
 			    (!test_bit(R5_LOCKED, &dev->flags) 
-#if 0
-|| sh->bh_page[i]!=bh->b_page
-#endif
 				    ) &&
 			    !test_bit(R5_UPTODATE, &dev->flags)) {
 				if (test_bit(R5_Insync, &dev->flags)
@@ -1655,9 +1611,6 @@ static void handle_stripe5(struct stripe_head *sh)
 			/* Would I have to read this buffer for reconstruct_write */
 			if (!test_bit(R5_OVERWRITE, &dev->flags) && i != sh->pd_idx &&
 			    (!test_bit(R5_LOCKED, &dev->flags) 
-#if 0
-|| sh->bh_page[i] != bh->b_page
-#endif
 				    ) &&
 			    !test_bit(R5_UPTODATE, &dev->flags)) {
 				if (test_bit(R5_Insync, &dev->flags)) rcw++;
@@ -1865,23 +1818,25 @@ static void handle_stripe5(struct stripe_head *sh)
 		return_bi = bi->bi_next;
 		bi->bi_next = NULL;
 		bi->bi_size = 0;
-		bi->bi_end_io(bi, bytes, 0);
+		bi->bi_end_io(bi, bytes,
+			      test_bit(BIO_UPTODATE, &bi->bi_flags)
+			        ? 0 : -EIO);
 	}
 	for (i=disks; i-- ;) {
 		int rw;
 		struct bio *bi;
 		mdk_rdev_t *rdev;
 		if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags))
-			rw = 1;
+			rw = WRITE;
 		else if (test_and_clear_bit(R5_Wantread, &sh->dev[i].flags))
-			rw = 0;
+			rw = READ;
 		else
 			continue;
  
 		bi = &sh->dev[i].req;
  
 		bi->bi_rw = rw;
-		if (rw)
+		if (rw == WRITE)
 			bi->bi_end_io = raid5_end_write_request;
 		else
 			bi->bi_end_io = raid5_end_read_request;
@@ -1917,7 +1872,7 @@ static void handle_stripe5(struct stripe_head *sh)
 				atomic_add(STRIPE_SECTORS, &rdev->corrected_errors);
 			generic_make_request(bi);
 		} else {
-			if (rw == 1)
+			if (rw == WRITE)
 				set_bit(STRIPE_DEGRADED, &sh->state);
 			PRINTK("skip op %ld on disc %d for sector %llu\n",
 				bi->bi_rw, i, (unsigned long long)sh->sector);
@@ -2193,15 +2148,6 @@ static void handle_stripe6(struct stripe_head *sh, struct page *tmp_page)
 				} else if (test_bit(R5_Insync, &dev->flags)) {
 					set_bit(R5_LOCKED, &dev->flags);
 					set_bit(R5_Wantread, &dev->flags);
-#if 0
-					/* if I am just reading this block and we don't have
-					   a failed drive, or any pending writes then sidestep the cache */
-					if (sh->bh_read[i] && !sh->bh_read[i]->b_reqnext &&
-					    ! syncing && !failed && !to_write) {
-						sh->bh_cache[i]->b_page =  sh->bh_read[i]->b_page;
-						sh->bh_cache[i]->b_data =  sh->bh_read[i]->b_data;
-					}
-#endif
 					locked++;
 					PRINTK("Reading block %d (sync=%d)\n",
 						i, syncing);
@@ -2220,9 +2166,6 @@ static void handle_stripe6(struct stripe_head *sh, struct page *tmp_page)
 			if (!test_bit(R5_OVERWRITE, &dev->flags)
 			    && i != pd_idx && i != qd_idx
 			    && (!test_bit(R5_LOCKED, &dev->flags)
-#if 0
-				|| sh->bh_page[i] != bh->b_page
-#endif
 				    ) &&
 			    !test_bit(R5_UPTODATE, &dev->flags)) {
 				if (test_bit(R5_Insync, &dev->flags)) rcw++;
@@ -2418,23 +2361,25 @@ static void handle_stripe6(struct stripe_head *sh, struct page *tmp_page)
 		return_bi = bi->bi_next;
 		bi->bi_next = NULL;
 		bi->bi_size = 0;
-		bi->bi_end_io(bi, bytes, 0);
+		bi->bi_end_io(bi, bytes,
+			      test_bit(BIO_UPTODATE, &bi->bi_flags)
+			        ? 0 : -EIO);
 	}
 	for (i=disks; i-- ;) {
 		int rw;
 		struct bio *bi;
 		mdk_rdev_t *rdev;
 		if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags))
-			rw = 1;
+			rw = WRITE;
 		else if (test_and_clear_bit(R5_Wantread, &sh->dev[i].flags))
-			rw = 0;
+			rw = READ;
 		else
 			continue;
 
 		bi = &sh->dev[i].req;
 
 		bi->bi_rw = rw;
-		if (rw)
+		if (rw == WRITE)
 			bi->bi_end_io = raid5_end_write_request;
 		else
 			bi->bi_end_io = raid5_end_read_request;
@@ -2470,7 +2415,7 @@ static void handle_stripe6(struct stripe_head *sh, struct page *tmp_page)
 				atomic_add(STRIPE_SECTORS, &rdev->corrected_errors);
 			generic_make_request(bi);
 		} else {
-			if (rw == 1)
+			if (rw == WRITE)
 				set_bit(STRIPE_DEGRADED, &sh->state);
 			PRINTK("skip op %ld on disc %d for sector %llu\n",
 				bi->bi_rw, i, (unsigned long long)sh->sector);
@@ -2611,6 +2556,180 @@ static int raid5_congested(void *data, int bits)
 	return 0;
 }
 
+/* We want read requests to align with chunks where possible,
+ * but write requests don't need to.
+ */
+static int raid5_mergeable_bvec(request_queue_t *q, struct bio *bio, struct bio_vec *biovec)
+{
+	mddev_t *mddev = q->queuedata;
+	sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev);
+	int max;
+	unsigned int chunk_sectors = mddev->chunk_size >> 9;
+	unsigned int bio_sectors = bio->bi_size >> 9;
+
+	if (bio_data_dir(bio) == WRITE)
+		return biovec->bv_len; /* always allow writes to be mergeable */
+
+	max =  (chunk_sectors - ((sector & (chunk_sectors - 1)) + bio_sectors)) << 9;
+	if (max < 0) max = 0;
+	if (max <= biovec->bv_len && bio_sectors == 0)
+		return biovec->bv_len;
+	else
+		return max;
+}
+
+
+static int in_chunk_boundary(mddev_t *mddev, struct bio *bio)
+{
+	sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev);
+	unsigned int chunk_sectors = mddev->chunk_size >> 9;
+	unsigned int bio_sectors = bio->bi_size >> 9;
+
+	return  chunk_sectors >=
+		((sector & (chunk_sectors - 1)) + bio_sectors);
+}
+
+/*
+ *  add bio to the retry LIFO  ( in O(1) ... we are in interrupt )
+ *  later sampled by raid5d.
+ */
+static void add_bio_to_retry(struct bio *bi,raid5_conf_t *conf)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&conf->device_lock, flags);
+
+	bi->bi_next = conf->retry_read_aligned_list;
+	conf->retry_read_aligned_list = bi;
+
+	spin_unlock_irqrestore(&conf->device_lock, flags);
+	md_wakeup_thread(conf->mddev->thread);
+}
+
+
+static struct bio *remove_bio_from_retry(raid5_conf_t *conf)
+{
+	struct bio *bi;
+
+	bi = conf->retry_read_aligned;
+	if (bi) {
+		conf->retry_read_aligned = NULL;
+		return bi;
+	}
+	bi = conf->retry_read_aligned_list;
+	if(bi) {
+		conf->retry_read_aligned = bi->bi_next;
+		bi->bi_next = NULL;
+		bi->bi_phys_segments = 1; /* biased count of active stripes */
+		bi->bi_hw_segments = 0; /* count of processed stripes */
+	}
+
+	return bi;
+}
+
+
+/*
+ *  The "raid5_align_endio" should check if the read succeeded and if it
+ *  did, call bio_endio on the original bio (having bio_put the new bio
+ *  first).
+ *  If the read failed..
+ */
+static int raid5_align_endio(struct bio *bi, unsigned int bytes, int error)
+{
+	struct bio* raid_bi  = bi->bi_private;
+	mddev_t *mddev;
+	raid5_conf_t *conf;
+	int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
+	mdk_rdev_t *rdev;
+
+	if (bi->bi_size)
+		return 1;
+	bio_put(bi);
+
+	mddev = raid_bi->bi_bdev->bd_disk->queue->queuedata;
+	conf = mddev_to_conf(mddev);
+	rdev = (void*)raid_bi->bi_next;
+	raid_bi->bi_next = NULL;
+
+	rdev_dec_pending(rdev, conf->mddev);
+
+	if (!error && uptodate) {
+		bio_endio(raid_bi, bytes, 0);
+		if (atomic_dec_and_test(&conf->active_aligned_reads))
+			wake_up(&conf->wait_for_stripe);
+		return 0;
+	}
+
+
+	PRINTK("raid5_align_endio : io error...handing IO for a retry\n");
+
+	add_bio_to_retry(raid_bi, conf);
+	return 0;
+}
+
+static int chunk_aligned_read(request_queue_t *q, struct bio * raid_bio)
+{
+	mddev_t *mddev = q->queuedata;
+	raid5_conf_t *conf = mddev_to_conf(mddev);
+	const unsigned int raid_disks = conf->raid_disks;
+	const unsigned int data_disks = raid_disks - conf->max_degraded;
+	unsigned int dd_idx, pd_idx;
+	struct bio* align_bi;
+	mdk_rdev_t *rdev;
+
+	if (!in_chunk_boundary(mddev, raid_bio)) {
+		printk("chunk_aligned_read : non aligned\n");
+		return 0;
+	}
+	/*
+ 	 * use bio_clone to make a copy of the bio
+	 */
+	align_bi = bio_clone(raid_bio, GFP_NOIO);
+	if (!align_bi)
+		return 0;
+	/*
+	 *   set bi_end_io to a new function, and set bi_private to the
+	 *     original bio.
+	 */
+	align_bi->bi_end_io  = raid5_align_endio;
+	align_bi->bi_private = raid_bio;
+	/*
+	 *	compute position
+	 */
+	align_bi->bi_sector =  raid5_compute_sector(raid_bio->bi_sector,
+					raid_disks,
+					data_disks,
+					&dd_idx,
+					&pd_idx,
+					conf);
+
+	rcu_read_lock();
+	rdev = rcu_dereference(conf->disks[dd_idx].rdev);
+	if (rdev && test_bit(In_sync, &rdev->flags)) {
+		atomic_inc(&rdev->nr_pending);
+		rcu_read_unlock();
+		raid_bio->bi_next = (void*)rdev;
+		align_bi->bi_bdev =  rdev->bdev;
+		align_bi->bi_flags &= ~(1 << BIO_SEG_VALID);
+		align_bi->bi_sector += rdev->data_offset;
+
+		spin_lock_irq(&conf->device_lock);
+		wait_event_lock_irq(conf->wait_for_stripe,
+				    conf->quiesce == 0,
+				    conf->device_lock, /* nothing */);
+		atomic_inc(&conf->active_aligned_reads);
+		spin_unlock_irq(&conf->device_lock);
+
+		generic_make_request(align_bi);
+		return 1;
+	} else {
+		rcu_read_unlock();
+		bio_put(align_bi);
+		return 0;
+	}
+}
+
+
 static int make_request(request_queue_t *q, struct bio * bi)
 {
 	mddev_t *mddev = q->queuedata;
@@ -2632,6 +2751,11 @@ static int make_request(request_queue_t *q, struct bio * bi)
 	disk_stat_inc(mddev->gendisk, ios[rw]);
 	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bi));
 
+	if (rw == READ &&
+	     mddev->reshape_position == MaxSector &&
+	     chunk_aligned_read(q,bi))
+            	return 0;
+
 	logical_sector = bi->bi_sector & ~((sector_t)STRIPE_SECTORS-1);
 	last_sector = bi->bi_sector + (bi->bi_size>>9);
 	bi->bi_next = NULL;
@@ -2739,7 +2863,9 @@ static int make_request(request_queue_t *q, struct bio * bi)
 		if ( rw == WRITE )
 			md_write_end(mddev);
 		bi->bi_size = 0;
-		bi->bi_end_io(bi, bytes, 0);
+		bi->bi_end_io(bi, bytes,
+			      test_bit(BIO_UPTODATE, &bi->bi_flags)
+			        ? 0 : -EIO);
 	}
 	return 0;
 }
@@ -2950,6 +3076,74 @@ static inline sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *ski
 	return STRIPE_SECTORS;
 }
 
+static int  retry_aligned_read(raid5_conf_t *conf, struct bio *raid_bio)
+{
+	/* We may not be able to submit a whole bio at once as there
+	 * may not be enough stripe_heads available.
+	 * We cannot pre-allocate enough stripe_heads as we may need
+	 * more than exist in the cache (if we allow ever large chunks).
+	 * So we do one stripe head at a time and record in
+	 * ->bi_hw_segments how many have been done.
+	 *
+	 * We *know* that this entire raid_bio is in one chunk, so
+	 * it will be only one 'dd_idx' and only need one call to raid5_compute_sector.
+	 */
+	struct stripe_head *sh;
+	int dd_idx, pd_idx;
+	sector_t sector, logical_sector, last_sector;
+	int scnt = 0;
+	int remaining;
+	int handled = 0;
+
+	logical_sector = raid_bio->bi_sector & ~((sector_t)STRIPE_SECTORS-1);
+	sector = raid5_compute_sector(	logical_sector,
+					conf->raid_disks,
+					conf->raid_disks - conf->max_degraded,
+					&dd_idx,
+					&pd_idx,
+					conf);
+	last_sector = raid_bio->bi_sector + (raid_bio->bi_size>>9);
+
+	for (; logical_sector < last_sector;
+	     logical_sector += STRIPE_SECTORS, scnt++) {
+
+		if (scnt < raid_bio->bi_hw_segments)
+			/* already done this stripe */
+			continue;
+
+		sh = get_active_stripe(conf, sector, conf->raid_disks, pd_idx, 1);
+
+		if (!sh) {
+			/* failed to get a stripe - must wait */
+			raid_bio->bi_hw_segments = scnt;
+			conf->retry_read_aligned = raid_bio;
+			return handled;
+		}
+
+		set_bit(R5_ReadError, &sh->dev[dd_idx].flags);
+		add_stripe_bio(sh, raid_bio, dd_idx, 0);
+		handle_stripe(sh, NULL);
+		release_stripe(sh);
+		handled++;
+	}
+	spin_lock_irq(&conf->device_lock);
+	remaining = --raid_bio->bi_phys_segments;
+	spin_unlock_irq(&conf->device_lock);
+	if (remaining == 0) {
+		int bytes = raid_bio->bi_size;
+
+		raid_bio->bi_size = 0;
+		raid_bio->bi_end_io(raid_bio, bytes,
+			      test_bit(BIO_UPTODATE, &raid_bio->bi_flags)
+			        ? 0 : -EIO);
+	}
+	if (atomic_dec_and_test(&conf->active_aligned_reads))
+		wake_up(&conf->wait_for_stripe);
+	return handled;
+}
+
+
+
 /*
  * This is our raid5 kernel thread.
  *
@@ -2971,6 +3165,7 @@ static void raid5d (mddev_t *mddev)
 	spin_lock_irq(&conf->device_lock);
 	while (1) {
 		struct list_head *first;
+		struct bio *bio;
 
 		if (conf->seq_flush != conf->seq_write) {
 			int seq = conf->seq_flush;
@@ -2987,6 +3182,16 @@ static void raid5d (mddev_t *mddev)
 		    !list_empty(&conf->delayed_list))
 			raid5_activate_delayed(conf);
 
+		while ((bio = remove_bio_from_retry(conf))) {
+			int ok;
+			spin_unlock_irq(&conf->device_lock);
+			ok = retry_aligned_read(conf, bio);
+			spin_lock_irq(&conf->device_lock);
+			if (!ok)
+				break;
+			handled++;
+		}
+
 		if (list_empty(&conf->handle_list))
 			break;
 
@@ -3174,6 +3379,7 @@ static int run(mddev_t *mddev)
 	INIT_LIST_HEAD(&conf->inactive_list);
 	atomic_set(&conf->active_stripes, 0);
 	atomic_set(&conf->preread_active_stripes, 0);
+	atomic_set(&conf->active_aligned_reads, 0);
 
 	PRINTK("raid5: run(%s) called.\n", mdname(mddev));
 
@@ -3320,6 +3526,8 @@ static int run(mddev_t *mddev)
 	mddev->array_size =  mddev->size * (conf->previous_raid_disks -
 					    conf->max_degraded);
 
+	blk_queue_merge_bvec(mddev->queue, raid5_mergeable_bvec);
+
 	return 0;
 abort:
 	if (conf) {
@@ -3694,7 +3902,8 @@ static void raid5_quiesce(mddev_t *mddev, int state)
 		spin_lock_irq(&conf->device_lock);
 		conf->quiesce = 1;
 		wait_event_lock_irq(conf->wait_for_stripe,
-				    atomic_read(&conf->active_stripes) == 0,
+				    atomic_read(&conf->active_stripes) == 0 &&
+				    atomic_read(&conf->active_aligned_reads) == 0,
 				    conf->device_lock, /* nothing */);
 		spin_unlock_irq(&conf->device_lock);
 		break;