Merge tag 'dm-3.6-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-dm

Pull device-mapper updates from Alasdair G Kergon:
- Flip the thin target into new read-only or failed modes if errors
  are detected;
- Handle chunk sizes that are not powers of two in the snapshot and
  thin targets;
- Provide a way for userspace to avoid replacing an already-loaded
  multipath hardware handler while booting;
- Reduce dm_thin_endio_hook slab size to avoid allocation failures;
- Numerous small changes and cleanups to the code.

* tag 'dm-3.6-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-dm: (63 commits)
  dm thin: commit before gathering status
  dm thin: add read only and fail io modes
  dm thin metadata: introduce dm_pool_abort_metadata
  dm thin metadata: introduce dm_pool_metadata_set_read_only
  dm persistent data: introduce dm_bm_set_read_only
  dm thin: reduce number of metadata commits
  dm thin metadata: add dm_thin_changed_this_transaction
  dm thin metadata: add format option to dm_pool_metadata_open
  dm thin metadata: tidy up open and format error paths
  dm thin metadata: only check incompat features on open
  dm thin metadata: remove duplicate pmd initialisation
  dm thin metadata: remove create parameter from __create_persistent_data_objects
  dm thin metadata: move __superblock_all_zeroes to __open_or_format_metadata
  dm thin metadata: remove nr_blocks arg from __create_persistent_data_objects
  dm thin metadata: split __open or format metadata
  dm thin metadata: use struct dm_pool_metadata members in __open_or_format_metadata
  dm thin metadata: zero unused superblock uuid
  dm thin metadata: lift __begin_transaction out of __write_initial_superblock
  dm thin metadata: move dm_commit_pool_metadata into __write_initial_superblock
  dm thin metadata: factor out __write_initial_superblock
  ...

29 files changed, 1321 insertions, 1301 deletions

diff --git a/drivers/md/Kconfig b/drivers/md/Kconfig
index 10f122a3a85..1eee45b69b7 100644
--- a/drivers/md/Kconfig
+++ b/drivers/md/Kconfig
@@ -260,15 +260,6 @@ config DM_DEBUG_BLOCK_STACK_TRACING
 
 	  If unsure, say N.
 
-config DM_DEBUG_SPACE_MAPS
-	boolean "Extra validation for thin provisioning space maps"
-	depends on DM_THIN_PROVISIONING
-	---help---
-	  Enable this for messages that may help debug problems with the
-	  space maps used by thin provisioning.
-
-          If unsure, say N.
-
 config DM_MIRROR
        tristate "Mirror target"
        depends on BLK_DEV_DM
diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index 3f06df59fd8..664743d6a6c 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -42,21 +42,21 @@ struct convert_context {
 	unsigned int offset_out;
 	unsigned int idx_in;
 	unsigned int idx_out;
-	sector_t sector;
-	atomic_t pending;
+	sector_t cc_sector;
+	atomic_t cc_pending;
 };
 
 /*
  * per bio private data
  */
 struct dm_crypt_io {
-	struct dm_target *target;
+	struct crypt_config *cc;
 	struct bio *base_bio;
 	struct work_struct work;
 
 	struct convert_context ctx;
 
-	atomic_t pending;
+	atomic_t io_pending;
 	int error;
 	sector_t sector;
 	struct dm_crypt_io *base_io;
@@ -109,9 +109,6 @@ enum flags { DM_CRYPT_SUSPENDED, DM_CRYPT_KEY_VALID };
  */
 struct crypt_cpu {
 	struct ablkcipher_request *req;
-	/* ESSIV: struct crypto_cipher *essiv_tfm */
-	void *iv_private;
-	struct crypto_ablkcipher *tfms[0];
 };
 
 /*
@@ -151,6 +148,10 @@ struct crypt_config {
 	 * per_cpu_ptr() only.
 	 */
 	struct crypt_cpu __percpu *cpu;
+
+	/* ESSIV: struct crypto_cipher *essiv_tfm */
+	void *iv_private;
+	struct crypto_ablkcipher **tfms;
 	unsigned tfms_count;
 
 	/*
@@ -193,7 +194,7 @@ static struct crypt_cpu *this_crypt_config(struct crypt_config *cc)
  */
 static struct crypto_ablkcipher *any_tfm(struct crypt_config *cc)
 {
-	return __this_cpu_ptr(cc->cpu)->tfms[0];
+	return cc->tfms[0];
 }
 
 /*
@@ -258,7 +259,7 @@ static int crypt_iv_essiv_init(struct crypt_config *cc)
 	struct hash_desc desc;
 	struct scatterlist sg;
 	struct crypto_cipher *essiv_tfm;
-	int err, cpu;
+	int err;
 
 	sg_init_one(&sg, cc->key, cc->key_size);
 	desc.tfm = essiv->hash_tfm;
@@ -268,14 +269,12 @@ static int crypt_iv_essiv_init(struct crypt_config *cc)
 	if (err)
 		return err;
 
-	for_each_possible_cpu(cpu) {
-		essiv_tfm = per_cpu_ptr(cc->cpu, cpu)->iv_private,
+	essiv_tfm = cc->iv_private;
 
-		err = crypto_cipher_setkey(essiv_tfm, essiv->salt,
-				    crypto_hash_digestsize(essiv->hash_tfm));
-		if (err)
-			return err;
-	}
+	err = crypto_cipher_setkey(essiv_tfm, essiv->salt,
+			    crypto_hash_digestsize(essiv->hash_tfm));
+	if (err)
+		return err;
 
 	return 0;
 }
@@ -286,16 +285,14 @@ static int crypt_iv_essiv_wipe(struct crypt_config *cc)
 	struct iv_essiv_private *essiv = &cc->iv_gen_private.essiv;
 	unsigned salt_size = crypto_hash_digestsize(essiv->hash_tfm);
 	struct crypto_cipher *essiv_tfm;
-	int cpu, r, err = 0;
+	int r, err = 0;
 
 	memset(essiv->salt, 0, salt_size);
 
-	for_each_possible_cpu(cpu) {
-		essiv_tfm = per_cpu_ptr(cc->cpu, cpu)->iv_private;
-		r = crypto_cipher_setkey(essiv_tfm, essiv->salt, salt_size);
-		if (r)
-			err = r;
-	}
+	essiv_tfm = cc->iv_private;
+	r = crypto_cipher_setkey(essiv_tfm, essiv->salt, salt_size);
+	if (r)
+		err = r;
 
 	return err;
 }
@@ -335,8 +332,6 @@ static struct crypto_cipher *setup_essiv_cpu(struct crypt_config *cc,
 
 static void crypt_iv_essiv_dtr(struct crypt_config *cc)
 {
-	int cpu;
-	struct crypt_cpu *cpu_cc;
 	struct crypto_cipher *essiv_tfm;
 	struct iv_essiv_private *essiv = &cc->iv_gen_private.essiv;
 
@@ -346,15 +341,12 @@ static void crypt_iv_essiv_dtr(struct crypt_config *cc)
 	kzfree(essiv->salt);
 	essiv->salt = NULL;
 
-	for_each_possible_cpu(cpu) {
-		cpu_cc = per_cpu_ptr(cc->cpu, cpu);
-		essiv_tfm = cpu_cc->iv_private;
+	essiv_tfm = cc->iv_private;
 
-		if (essiv_tfm)
-			crypto_free_cipher(essiv_tfm);
+	if (essiv_tfm)
+		crypto_free_cipher(essiv_tfm);
 
-		cpu_cc->iv_private = NULL;
-	}
+	cc->iv_private = NULL;
 }
 
 static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti,
@@ -363,7 +355,7 @@ static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti,
 	struct crypto_cipher *essiv_tfm = NULL;
 	struct crypto_hash *hash_tfm = NULL;
 	u8 *salt = NULL;
-	int err, cpu;
+	int err;
 
 	if (!opts) {
 		ti->error = "Digest algorithm missing for ESSIV mode";
@@ -388,15 +380,13 @@ static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti,
 	cc->iv_gen_private.essiv.salt = salt;
 	cc->iv_gen_private.essiv.hash_tfm = hash_tfm;
 
-	for_each_possible_cpu(cpu) {
-		essiv_tfm = setup_essiv_cpu(cc, ti, salt,
-					crypto_hash_digestsize(hash_tfm));
-		if (IS_ERR(essiv_tfm)) {
-			crypt_iv_essiv_dtr(cc);
-			return PTR_ERR(essiv_tfm);
-		}
-		per_cpu_ptr(cc->cpu, cpu)->iv_private = essiv_tfm;
+	essiv_tfm = setup_essiv_cpu(cc, ti, salt,
+				crypto_hash_digestsize(hash_tfm));
+	if (IS_ERR(essiv_tfm)) {
+		crypt_iv_essiv_dtr(cc);
+		return PTR_ERR(essiv_tfm);
 	}
+	cc->iv_private = essiv_tfm;
 
 	return 0;
 
@@ -410,7 +400,7 @@ bad:
 static int crypt_iv_essiv_gen(struct crypt_config *cc, u8 *iv,
 			      struct dm_crypt_request *dmreq)
 {
-	struct crypto_cipher *essiv_tfm = this_crypt_config(cc)->iv_private;
+	struct crypto_cipher *essiv_tfm = cc->iv_private;
 
 	memset(iv, 0, cc->iv_size);
 	*(__le64 *)iv = cpu_to_le64(dmreq->iv_sector);
@@ -664,7 +654,7 @@ static void crypt_convert_init(struct crypt_config *cc,
 	ctx->offset_out = 0;
 	ctx->idx_in = bio_in ? bio_in->bi_idx : 0;
 	ctx->idx_out = bio_out ? bio_out->bi_idx : 0;
-	ctx->sector = sector + cc->iv_offset;
+	ctx->cc_sector = sector + cc->iv_offset;
 	init_completion(&ctx->restart);
 }
 
@@ -695,12 +685,12 @@ static int crypt_convert_block(struct crypt_config *cc,
 	struct bio_vec *bv_out = bio_iovec_idx(ctx->bio_out, ctx->idx_out);
 	struct dm_crypt_request *dmreq;
 	u8 *iv;
-	int r = 0;
+	int r;
 
 	dmreq = dmreq_of_req(cc, req);
 	iv = iv_of_dmreq(cc, dmreq);
 
-	dmreq->iv_sector = ctx->sector;
+	dmreq->iv_sector = ctx->cc_sector;
 	dmreq->ctx = ctx;
 	sg_init_table(&dmreq->sg_in, 1);
 	sg_set_page(&dmreq->sg_in, bv_in->bv_page, 1 << SECTOR_SHIFT,
@@ -749,12 +739,12 @@ static void crypt_alloc_req(struct crypt_config *cc,
 			    struct convert_context *ctx)
 {
 	struct crypt_cpu *this_cc = this_crypt_config(cc);
-	unsigned key_index = ctx->sector & (cc->tfms_count - 1);
+	unsigned key_index = ctx->cc_sector & (cc->tfms_count - 1);
 
 	if (!this_cc->req)
 		this_cc->req = mempool_alloc(cc->req_pool, GFP_NOIO);
 
-	ablkcipher_request_set_tfm(this_cc->req, this_cc->tfms[key_index]);
+	ablkcipher_request_set_tfm(this_cc->req, cc->tfms[key_index]);
 	ablkcipher_request_set_callback(this_cc->req,
 	    CRYPTO_TFM_REQ_MAY_BACKLOG | CRYPTO_TFM_REQ_MAY_SLEEP,
 	    kcryptd_async_done, dmreq_of_req(cc, this_cc->req));
@@ -769,14 +759,14 @@ static int crypt_convert(struct crypt_config *cc,
 	struct crypt_cpu *this_cc = this_crypt_config(cc);
 	int r;
 
-	atomic_set(&ctx->pending, 1);
+	atomic_set(&ctx->cc_pending, 1);
 
 	while(ctx->idx_in < ctx->bio_in->bi_vcnt &&
 	      ctx->idx_out < ctx->bio_out->bi_vcnt) {
 
 		crypt_alloc_req(cc, ctx);
 
-		atomic_inc(&ctx->pending);
+		atomic_inc(&ctx->cc_pending);
 
 		r = crypt_convert_block(cc, ctx, this_cc->req);
 
@@ -788,19 +778,19 @@ static int crypt_convert(struct crypt_config *cc,
 			/* fall through*/
 		case -EINPROGRESS:
 			this_cc->req = NULL;
-			ctx->sector++;
+			ctx->cc_sector++;
 			continue;
 
 		/* sync */
 		case 0:
-			atomic_dec(&ctx->pending);
-			ctx->sector++;
+			atomic_dec(&ctx->cc_pending);
+			ctx->cc_sector++;
 			cond_resched();
 			continue;
 
 		/* error */
 		default:
-			atomic_dec(&ctx->pending);
+			atomic_dec(&ctx->cc_pending);
 			return r;
 		}
 	}
@@ -811,7 +801,7 @@ static int crypt_convert(struct crypt_config *cc,
 static void dm_crypt_bio_destructor(struct bio *bio)
 {
 	struct dm_crypt_io *io = bio->bi_private;
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 
 	bio_free(bio, cc->bs);
 }
@@ -825,7 +815,7 @@ static void dm_crypt_bio_destructor(struct bio *bio)
 static struct bio *crypt_alloc_buffer(struct dm_crypt_io *io, unsigned size,
 				      unsigned *out_of_pages)
 {
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 	struct bio *clone;
 	unsigned int nr_iovecs = (size + PAGE_SIZE - 1) >> PAGE_SHIFT;
 	gfp_t gfp_mask = GFP_NOIO | __GFP_HIGHMEM;
@@ -884,26 +874,25 @@ static void crypt_free_buffer_pages(struct crypt_config *cc, struct bio *clone)
 	}
 }
 
-static struct dm_crypt_io *crypt_io_alloc(struct dm_target *ti,
+static struct dm_crypt_io *crypt_io_alloc(struct crypt_config *cc,
 					  struct bio *bio, sector_t sector)
 {
-	struct crypt_config *cc = ti->private;
 	struct dm_crypt_io *io;
 
 	io = mempool_alloc(cc->io_pool, GFP_NOIO);
-	io->target = ti;
+	io->cc = cc;
 	io->base_bio = bio;
 	io->sector = sector;
 	io->error = 0;
 	io->base_io = NULL;
-	atomic_set(&io->pending, 0);
+	atomic_set(&io->io_pending, 0);
 
 	return io;
 }
 
 static void crypt_inc_pending(struct dm_crypt_io *io)
 {
-	atomic_inc(&io->pending);
+	atomic_inc(&io->io_pending);
 }
 
 /*
@@ -913,12 +902,12 @@ static void crypt_inc_pending(struct dm_crypt_io *io)
  */
 static void crypt_dec_pending(struct dm_crypt_io *io)
 {
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 	struct bio *base_bio = io->base_bio;
 	struct dm_crypt_io *base_io = io->base_io;
 	int error = io->error;
 
-	if (!atomic_dec_and_test(&io->pending))
+	if (!atomic_dec_and_test(&io->io_pending))
 		return;
 
 	mempool_free(io, cc->io_pool);
@@ -952,7 +941,7 @@ static void crypt_dec_pending(struct dm_crypt_io *io)
 static void crypt_endio(struct bio *clone, int error)
 {
 	struct dm_crypt_io *io = clone->bi_private;
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 	unsigned rw = bio_data_dir(clone);
 
 	if (unlikely(!bio_flagged(clone, BIO_UPTODATE) && !error))
@@ -979,7 +968,7 @@ static void crypt_endio(struct bio *clone, int error)
 
 static void clone_init(struct dm_crypt_io *io, struct bio *clone)
 {
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 
 	clone->bi_private = io;
 	clone->bi_end_io  = crypt_endio;
@@ -990,7 +979,7 @@ static void clone_init(struct dm_crypt_io *io, struct bio *clone)
 
 static int kcryptd_io_read(struct dm_crypt_io *io, gfp_t gfp)
 {
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 	struct bio *base_bio = io->base_bio;
 	struct bio *clone;
 
@@ -1038,7 +1027,7 @@ static void kcryptd_io(struct work_struct *work)
 
 static void kcryptd_queue_io(struct dm_crypt_io *io)
 {
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 
 	INIT_WORK(&io->work, kcryptd_io);
 	queue_work(cc->io_queue, &io->work);
@@ -1047,7 +1036,7 @@ static void kcryptd_queue_io(struct dm_crypt_io *io)
 static void kcryptd_crypt_write_io_submit(struct dm_crypt_io *io, int async)
 {
 	struct bio *clone = io->ctx.bio_out;
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 
 	if (unlikely(io->error < 0)) {
 		crypt_free_buffer_pages(cc, clone);
@@ -1069,7 +1058,7 @@ static void kcryptd_crypt_write_io_submit(struct dm_crypt_io *io, int async)
 
 static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)
 {
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 	struct bio *clone;
 	struct dm_crypt_io *new_io;
 	int crypt_finished;
@@ -1107,7 +1096,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)
 		if (r < 0)
 			io->error = -EIO;
 
-		crypt_finished = atomic_dec_and_test(&io->ctx.pending);
+		crypt_finished = atomic_dec_and_test(&io->ctx.cc_pending);
 
 		/* Encryption was already finished, submit io now */
 		if (crypt_finished) {
@@ -1135,7 +1124,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)
 		 * between fragments, so switch to a new dm_crypt_io structure.
 		 */
 		if (unlikely(!crypt_finished && remaining)) {
-			new_io = crypt_io_alloc(io->target, io->base_bio,
+			new_io = crypt_io_alloc(io->cc, io->base_bio,
 						sector);
 			crypt_inc_pending(new_io);
 			crypt_convert_init(cc, &new_io->ctx, NULL,
@@ -1169,7 +1158,7 @@ static void kcryptd_crypt_read_done(struct dm_crypt_io *io)
 
 static void kcryptd_crypt_read_convert(struct dm_crypt_io *io)
 {
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 	int r = 0;
 
 	crypt_inc_pending(io);
@@ -1181,7 +1170,7 @@ static void kcryptd_crypt_read_convert(struct dm_crypt_io *io)
 	if (r < 0)
 		io->error = -EIO;
 
-	if (atomic_dec_and_test(&io->ctx.pending))
+	if (atomic_dec_and_test(&io->ctx.cc_pending))
 		kcryptd_crypt_read_done(io);
 
 	crypt_dec_pending(io);
@@ -1193,7 +1182,7 @@ static void kcryptd_async_done(struct crypto_async_request *async_req,
 	struct dm_crypt_request *dmreq = async_req->data;
 	struct convert_context *ctx = dmreq->ctx;
 	struct dm_crypt_io *io = container_of(ctx, struct dm_crypt_io, ctx);
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 
 	if (error == -EINPROGRESS) {
 		complete(&ctx->restart);
@@ -1208,7 +1197,7 @@ static void kcryptd_async_done(struct crypto_async_request *async_req,
 
 	mempool_free(req_of_dmreq(cc, dmreq), cc->req_pool);
 
-	if (!atomic_dec_and_test(&ctx->pending))
+	if (!atomic_dec_and_test(&ctx->cc_pending))
 		return;
 
 	if (bio_data_dir(io->base_bio) == READ)
@@ -1229,7 +1218,7 @@ static void kcryptd_crypt(struct work_struct *work)
 
 static void kcryptd_queue_crypt(struct dm_crypt_io *io)
 {
-	struct crypt_config *cc = io->target->private;
+	struct crypt_config *cc = io->cc;
 
 	INIT_WORK(&io->work, kcryptd_crypt);
 	queue_work(cc->crypt_queue, &io->work);
@@ -1241,7 +1230,6 @@ static void kcryptd_queue_crypt(struct dm_crypt_io *io)
 static int crypt_decode_key(u8 *key, char *hex, unsigned int size)
 {
 	char buffer[3];
-	char *endp;
 	unsigned int i;
 
 	buffer[2] = '\0';
@@ -1250,9 +1238,7 @@ static int crypt_decode_key(u8 *key, char *hex, unsigned int size)
 		buffer[0] = *hex++;
 		buffer[1] = *hex++;
 
-		key[i] = (u8)simple_strtoul(buffer, &endp, 16);
-
-		if (endp != &buffer[2])
+		if (kstrtou8(buffer, 16, &key[i]))
 			return -EINVAL;
 	}
 
@@ -1276,29 +1262,38 @@ static void crypt_encode_key(char *hex, u8 *key, unsigned int size)
 	}
 }
 
-static void crypt_free_tfms(struct crypt_config *cc, int cpu)
+static void crypt_free_tfms(struct crypt_config *cc)
 {
-	struct crypt_cpu *cpu_cc = per_cpu_ptr(cc->cpu, cpu);
 	unsigned i;
 
+	if (!cc->tfms)
+		return;
+
 	for (i = 0; i < cc->tfms_count; i++)
-		if (cpu_cc->tfms[i] && !IS_ERR(cpu_cc->tfms[i])) {
-			crypto_free_ablkcipher(cpu_cc->tfms[i]);
-			cpu_cc->tfms[i] = NULL;
+		if (cc->tfms[i] && !IS_ERR(cc->tfms[i])) {
+			crypto_free_ablkcipher(cc->tfms[i]);
+			cc->tfms[i] = NULL;
 		}
+
+	kfree(cc->tfms);
+	cc->tfms = NULL;
 }
 
-static int crypt_alloc_tfms(struct crypt_config *cc, int cpu, char *ciphermode)
+static int crypt_alloc_tfms(struct crypt_config *cc, char *ciphermode)
 {
-	struct crypt_cpu *cpu_cc = per_cpu_ptr(cc->cpu, cpu);
 	unsigned i;
 	int err;
 
+	cc->tfms = kmalloc(cc->tfms_count * sizeof(struct crypto_ablkcipher *),
+			   GFP_KERNEL);
+	if (!cc->tfms)
+		return -ENOMEM;
+
 	for (i = 0; i < cc->tfms_count; i++) {
-		cpu_cc->tfms[i] = crypto_alloc_ablkcipher(ciphermode, 0, 0);
-		if (IS_ERR(cpu_cc->tfms[i])) {
-			err = PTR_ERR(cpu_cc->tfms[i]);
-			crypt_free_tfms(cc, cpu);
+		cc->tfms[i] = crypto_alloc_ablkcipher(ciphermode, 0, 0);
+		if (IS_ERR(cc->tfms[i])) {
+			err = PTR_ERR(cc->tfms[i]);
+			crypt_free_tfms(cc);
 			return err;
 		}
 	}
@@ -1309,15 +1304,14 @@ static int crypt_alloc_tfms(struct crypt_config *cc, int cpu, char *ciphermode)
 static int crypt_setkey_allcpus(struct crypt_config *cc)
 {
 	unsigned subkey_size = cc->key_size >> ilog2(cc->tfms_count);
-	int cpu, err = 0, i, r;
-
-	for_each_possible_cpu(cpu) {
-		for (i = 0; i < cc->tfms_count; i++) {
-			r = crypto_ablkcipher_setkey(per_cpu_ptr(cc->cpu, cpu)->tfms[i],
-						     cc->key + (i * subkey_size), subkey_size);
-			if (r)
-				err = r;
-		}
+	int err = 0, i, r;
+
+	for (i = 0; i < cc->tfms_count; i++) {
+		r = crypto_ablkcipher_setkey(cc->tfms[i],
+					     cc->key + (i * subkey_size),
+					     subkey_size);
+		if (r)
+			err = r;
 	}
 
 	return err;
@@ -1379,9 +1373,10 @@ static void crypt_dtr(struct dm_target *ti)
 			cpu_cc = per_cpu_ptr(cc->cpu, cpu);
 			if (cpu_cc->req)
 				mempool_free(cpu_cc->req, cc->req_pool);
-			crypt_free_tfms(cc, cpu);
 		}
 
+	crypt_free_tfms(cc);
+
 	if (cc->bs)
 		bioset_free(cc->bs);
 
@@ -1414,7 +1409,7 @@ static int crypt_ctr_cipher(struct dm_target *ti,
 	struct crypt_config *cc = ti->private;
 	char *tmp, *cipher, *chainmode, *ivmode, *ivopts, *keycount;
 	char *cipher_api = NULL;
-	int cpu, ret = -EINVAL;
+	int ret = -EINVAL;
 	char dummy;
 
 	/* Convert to crypto api definition? */
@@ -1455,8 +1450,7 @@ static int crypt_ctr_cipher(struct dm_target *ti,
 	if (tmp)
 		DMWARN("Ignoring unexpected additional cipher options");
 
-	cc->cpu = __alloc_percpu(sizeof(*(cc->cpu)) +
-				 cc->tfms_count * sizeof(*(cc->cpu->tfms)),
+	cc->cpu = __alloc_percpu(sizeof(*(cc->cpu)),
 				 __alignof__(struct crypt_cpu));
 	if (!cc->cpu) {
 		ti->error = "Cannot allocate per cpu state";
@@ -1489,12 +1483,10 @@ static int crypt_ctr_cipher(struct dm_target *ti,
 	}
 
 	/* Allocate cipher */
-	for_each_possible_cpu(cpu) {
-		ret = crypt_alloc_tfms(cc, cpu, cipher_api);
-		if (ret < 0) {
-			ti->error = "Error allocating crypto tfm";
-			goto bad;
-		}
+	ret = crypt_alloc_tfms(cc, cipher_api);
+	if (ret < 0) {
+		ti->error = "Error allocating crypto tfm";
+		goto bad;
 	}
 
 	/* Initialize and set key */
@@ -1702,7 +1694,7 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	}
 
 	ti->num_flush_requests = 1;
-	ti->discard_zeroes_data_unsupported = 1;
+	ti->discard_zeroes_data_unsupported = true;
 
 	return 0;
 
@@ -1715,7 +1707,7 @@ static int crypt_map(struct dm_target *ti, struct bio *bio,
 		     union map_info *map_context)
 {
 	struct dm_crypt_io *io;
-	struct crypt_config *cc;
+	struct crypt_config *cc = ti->private;
 
 	/*
 	 * If bio is REQ_FLUSH or REQ_DISCARD, just bypass crypt queues.
@@ -1723,14 +1715,13 @@ static int crypt_map(struct dm_target *ti, struct bio *bio,
 	 * - for REQ_DISCARD caller must use flush if IO ordering matters
 	 */
 	if (unlikely(bio->bi_rw & (REQ_FLUSH | REQ_DISCARD))) {
-		cc = ti->private;
 		bio->bi_bdev = cc->dev->bdev;
 		if (bio_sectors(bio))
 			bio->bi_sector = cc->start + dm_target_offset(ti, bio->bi_sector);
 		return DM_MAPIO_REMAPPED;
 	}
 
-	io = crypt_io_alloc(ti, bio, dm_target_offset(ti, bio->bi_sector));
+	io = crypt_io_alloc(cc, bio, dm_target_offset(ti, bio->bi_sector));
 
 	if (bio_data_dir(io->base_bio) == READ) {
 		if (kcryptd_io_read(io, GFP_NOWAIT))
@@ -1742,7 +1733,7 @@ static int crypt_map(struct dm_target *ti, struct bio *bio,
 }
 
 static int crypt_status(struct dm_target *ti, status_type_t type,
-			char *result, unsigned int maxlen)
+			unsigned status_flags, char *result, unsigned maxlen)
 {
 	struct crypt_config *cc = ti->private;
 	unsigned int sz = 0;
diff --git a/drivers/md/dm-delay.c b/drivers/md/dm-delay.c
index 2dc22dddb2a..f53846f9ab5 100644
--- a/drivers/md/dm-delay.c
+++ b/drivers/md/dm-delay.c
@@ -295,7 +295,7 @@ static int delay_map(struct dm_target *ti, struct bio *bio,
 }
 
 static int delay_status(struct dm_target *ti, status_type_t type,
-			char *result, unsigned maxlen)
+			unsigned status_flags, char *result, unsigned maxlen)
 {
 	struct delay_c *dc = ti->private;
 	int sz = 0;
diff --git a/drivers/md/dm-exception-store.c b/drivers/md/dm-exception-store.c
index aa70f7d43a1..ebaa4f803ee 100644
--- a/drivers/md/dm-exception-store.c
+++ b/drivers/md/dm-exception-store.c
@@ -142,24 +142,19 @@ EXPORT_SYMBOL(dm_exception_store_type_unregister);
 static int set_chunk_size(struct dm_exception_store *store,
 			  const char *chunk_size_arg, char **error)
 {
-	unsigned long chunk_size_ulong;
-	char *value;
+	unsigned chunk_size;
 
-	chunk_size_ulong = simple_strtoul(chunk_size_arg, &value, 10);
-	if (*chunk_size_arg == '\0' || *value != '\0' ||
-	    chunk_size_ulong > UINT_MAX) {
+	if (kstrtouint(chunk_size_arg, 10, &chunk_size)) {
 		*error = "Invalid chunk size";
 		return -EINVAL;
 	}
 
-	if (!chunk_size_ulong) {
+	if (!chunk_size) {
 		store->chunk_size = store->chunk_mask = store->chunk_shift = 0;
 		return 0;
 	}
 
-	return dm_exception_store_set_chunk_size(store,
-						 (unsigned) chunk_size_ulong,
-						 error);
+	return dm_exception_store_set_chunk_size(store, chunk_size, error);
 }
 
 int dm_exception_store_set_chunk_size(struct dm_exception_store *store,
diff --git a/drivers/md/dm-flakey.c b/drivers/md/dm-flakey.c
index ac49c01f1a4..cc15543a6ad 100644
--- a/drivers/md/dm-flakey.c
+++ b/drivers/md/dm-flakey.c
@@ -333,7 +333,7 @@ static int flakey_end_io(struct dm_target *ti, struct bio *bio,
 }
 
 static int flakey_status(struct dm_target *ti, status_type_t type,
-			 char *result, unsigned int maxlen)
+			 unsigned status_flags, char *result, unsigned maxlen)
 {
 	unsigned sz = 0;
 	struct flakey_c *fc = ti->private;
diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index a1a3e6df17b..afd95986d09 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -1054,6 +1054,7 @@ static void retrieve_status(struct dm_table *table,
 	char *outbuf, *outptr;
 	status_type_t type;
 	size_t remaining, len, used = 0;
+	unsigned status_flags = 0;
 
 	outptr = outbuf = get_result_buffer(param, param_size, &len);
 
@@ -1090,7 +1091,9 @@ static void retrieve_status(struct dm_table *table,
 
 		/* Get the status/table string from the target driver */
 		if (ti->type->status) {
-			if (ti->type->status(ti, type, outptr, remaining)) {
+			if (param->flags & DM_NOFLUSH_FLAG)
+				status_flags |= DM_STATUS_NOFLUSH_FLAG;
+			if (ti->type->status(ti, type, status_flags, outptr, remaining)) {
 				param->flags |= DM_BUFFER_FULL_FLAG;
 				break;
 			}
diff --git a/drivers/md/dm-linear.c b/drivers/md/dm-linear.c
index 3639eeab604..1bf19a93eef 100644
--- a/drivers/md/dm-linear.c
+++ b/drivers/md/dm-linear.c
@@ -96,7 +96,7 @@ static int linear_map(struct dm_target *ti, struct bio *bio,
 }
 
 static int linear_status(struct dm_target *ti, status_type_t type,
-			 char *result, unsigned int maxlen)
+			 unsigned status_flags, char *result, unsigned maxlen)
 {
 	struct linear_c *lc = (struct linear_c *) ti->private;
 
diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index 638dae048b4..d8abb90a6c2 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -85,6 +85,7 @@ struct multipath {
 	unsigned queue_io:1;		/* Must we queue all I/O? */
 	unsigned queue_if_no_path:1;	/* Queue I/O if last path fails? */
 	unsigned saved_queue_if_no_path:1; /* Saved state during suspension */
+	unsigned retain_attached_hw_handler:1; /* If there's already a hw_handler present, don't change it. */
 
 	unsigned pg_init_retries;	/* Number of times to retry pg_init */
 	unsigned pg_init_count;		/* Number of times pg_init called */
@@ -568,6 +569,8 @@ static struct pgpath *parse_path(struct dm_arg_set *as, struct path_selector *ps
 	int r;
 	struct pgpath *p;
 	struct multipath *m = ti->private;
+	struct request_queue *q = NULL;
+	const char *attached_handler_name;
 
 	/* we need at least a path arg */
 	if (as->argc < 1) {
@@ -586,13 +589,37 @@ static struct pgpath *parse_path(struct dm_arg_set *as, struct path_selector *ps
 		goto bad;
 	}
 
-	if (m->hw_handler_name) {
-		struct request_queue *q = bdev_get_queue(p->path.dev->bdev);
+	if (m->retain_attached_hw_handler || m->hw_handler_name)
+		q = bdev_get_queue(p->path.dev->bdev);
+
+	if (m->retain_attached_hw_handler) {
+		attached_handler_name = scsi_dh_attached_handler_name(q, GFP_KERNEL);
+		if (attached_handler_name) {
+			/*
+			 * Reset hw_handler_name to match the attached handler
+			 * and clear any hw_handler_params associated with the
+			 * ignored handler.
+			 *
+			 * NB. This modifies the table line to show the actual
+			 * handler instead of the original table passed in.
+			 */
+			kfree(m->hw_handler_name);
+			m->hw_handler_name = attached_handler_name;
+
+			kfree(m->hw_handler_params);
+			m->hw_handler_params = NULL;
+		}
+	}
 
+	if (m->hw_handler_name) {
+		/*
+		 * Increments scsi_dh reference, even when using an
+		 * already-attached handler.
+		 */
 		r = scsi_dh_attach(q, m->hw_handler_name);
 		if (r == -EBUSY) {
 			/*
-			 * Already attached to different hw_handler,
+			 * Already attached to different hw_handler:
 			 * try to reattach with c