From 762a80d9fc9f690a3a35983f3b4619a220650808 Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 2 Aug 2011 12:32:00 +0100
Subject: dm snapshot: flush disk cache when merging

This patch makes dm-snapshot flush disk cache when writing metadata for
merging snapshot.

Without cache flushing the disk may reorder metadata write and other
data writes and there is a possibility of data corruption in case of
power fault.

Cc: stable@kernel.org
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-snap-persistent.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/md/dm-snap-persistent.c b/drivers/md/dm-snap-persistent.c
index 135c2f1fdbfc..e4ecadf0548a 100644
--- a/drivers/md/dm-snap-persistent.c
+++ b/drivers/md/dm-snap-persistent.c
@@ -753,7 +753,7 @@ static int persistent_commit_merge(struct dm_exception_store *store,
 	for (i = 0; i < nr_merged; i++)
 		clear_exception(ps, ps->current_committed - 1 - i);
 
-	r = area_io(ps, WRITE);
+	r = area_io(ps, WRITE_FLUSH_FUA);
 	if (r < 0)
 		return r;
 
-- 
cgit v1.2.3


From 286f367dad40beb3234a18c17391d03ba939a7f3 Mon Sep 17 00:00:00 2001
From: Mike Snitzer <snitzer@redhat.com>
Date: Tue, 2 Aug 2011 12:32:00 +0100
Subject: dm mpath: fix potential NULL pointer in feature arg processing

Avoid dereferencing a NULL pointer if the number of feature arguments
supplied is fewer than indicated.

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Cc: stable@kernel.org
---
 drivers/md/dm-mpath.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index c3547016f0f1..adf851a081bd 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -807,6 +807,11 @@ static int parse_features(struct arg_set *as, struct multipath *m)
 	if (!argc)
 		return 0;
 
+	if (argc > as->argc) {
+		ti->error = "not enough arguments for features";
+		return -EINVAL;
+	}
+
 	do {
 		param_name = shift(as);
 		argc--;
-- 
cgit v1.2.3


From bb91bc7bacb906c9f3a9b22744c53fa7564b51ba Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 2 Aug 2011 12:32:01 +0100
Subject: dm io: flush cpu cache with vmapped io

For normal kernel pages, CPU cache is synchronized by the dma layer.
However, this is not done for pages allocated with vmalloc. If we do I/O
to/from vmallocated pages, we must synchronize CPU cache explicitly.

Prior to doing I/O on vmallocated page we must call
flush_kernel_vmap_range to flush dirty cache on the virtual address.
After finished read we must call invalidate_kernel_vmap_range to
invalidate cache on the virtual address, so that accesses to the virtual
address return newly read data and not stale data from CPU cache.

This patch fixes metadata corruption on dm-snapshots on PA-RISC and
possibly other architectures with caches indexed by virtual address.

Cc: stable <stable@kernel.org>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-io.c | 29 +++++++++++++++++++++++++++--
 1 file changed, 27 insertions(+), 2 deletions(-)

diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
index 2067288f61f9..ad2eba40e319 100644
--- a/drivers/md/dm-io.c
+++ b/drivers/md/dm-io.c
@@ -38,6 +38,8 @@ struct io {
 	struct dm_io_client *client;
 	io_notify_fn callback;
 	void *context;
+	void *vma_invalidate_address;
+	unsigned long vma_invalidate_size;
 } __attribute__((aligned(DM_IO_MAX_REGIONS)));
 
 static struct kmem_cache *_dm_io_cache;
@@ -116,6 +118,10 @@ static void dec_count(struct io *io, unsigned int region, int error)
 		set_bit(region, &io->error_bits);
 
 	if (atomic_dec_and_test(&io->count)) {
+		if (io->vma_invalidate_size)
+			invalidate_kernel_vmap_range(io->vma_invalidate_address,
+						     io->vma_invalidate_size);
+
 		if (io->sleeper)
 			wake_up_process(io->sleeper);
 
@@ -159,6 +165,9 @@ struct dpages {
 
 	unsigned context_u;
 	void *context_ptr;
+
+	void *vma_invalidate_address;
+	unsigned long vma_invalidate_size;
 };
 
 /*
@@ -377,6 +386,9 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,
 	io->sleeper = current;
 	io->client = client;
 
+	io->vma_invalidate_address = dp->vma_invalidate_address;
+	io->vma_invalidate_size = dp->vma_invalidate_size;
+
 	dispatch_io(rw, num_regions, where, dp, io, 1);
 
 	while (1) {
@@ -415,13 +427,21 @@ static int async_io(struct dm_io_client *client, unsigned int num_regions,
 	io->callback = fn;
 	io->context = context;
 
+	io->vma_invalidate_address = dp->vma_invalidate_address;
+	io->vma_invalidate_size = dp->vma_invalidate_size;
+
 	dispatch_io(rw, num_regions, where, dp, io, 0);
 	return 0;
 }
 
-static int dp_init(struct dm_io_request *io_req, struct dpages *dp)
+static int dp_init(struct dm_io_request *io_req, struct dpages *dp,
+		   unsigned long size)
 {
 	/* Set up dpages based on memory type */
+
+	dp->vma_invalidate_address = NULL;
+	dp->vma_invalidate_size = 0;
+
 	switch (io_req->mem.type) {
 	case DM_IO_PAGE_LIST:
 		list_dp_init(dp, io_req->mem.ptr.pl, io_req->mem.offset);
@@ -432,6 +452,11 @@ static int dp_init(struct dm_io_request *io_req, struct dpages *dp)
 		break;
 
 	case DM_IO_VMA:
+		flush_kernel_vmap_range(io_req->mem.ptr.vma, size);
+		if ((io_req->bi_rw & RW_MASK) == READ) {
+			dp->vma_invalidate_address = io_req->mem.ptr.vma;
+			dp->vma_invalidate_size = size;
+		}
 		vm_dp_init(dp, io_req->mem.ptr.vma);
 		break;
 
@@ -460,7 +485,7 @@ int dm_io(struct dm_io_request *io_req, unsigned num_regions,
 	int r;
 	struct dpages dp;
 
-	r = dp_init(io_req, &dp);
+	r = dp_init(io_req, &dp, (unsigned long)where->count << SECTOR_SHIFT);
 	if (r)
 		return r;
 
-- 
cgit v1.2.3


From d15b774c2920d55e3d58275c97fbe3adc3afde38 Mon Sep 17 00:00:00 2001
From: Alasdair G Kergon <agk@redhat.com>
Date: Tue, 2 Aug 2011 12:32:01 +0100
Subject: dm: fix idr leak on module removal

Destroy _minor_idr when unloading the core dm module.  (Found by kmemleak.)

Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 0cf68b478878..41abc6dd481b 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -37,6 +37,8 @@ static const char *_name = DM_NAME;
 static unsigned int major = 0;
 static unsigned int _major = 0;
 
+static DEFINE_IDR(_minor_idr);
+
 static DEFINE_SPINLOCK(_minor_lock);
 /*
  * For bio-based dm.
@@ -313,6 +315,12 @@ static void __exit dm_exit(void)
 
 	while (i--)
 		_exits[i]();
+
+	/*
+	 * Should be empty by this point.
+	 */
+	idr_remove_all(&_minor_idr);
+	idr_destroy(&_minor_idr);
 }
 
 /*
@@ -1705,8 +1713,6 @@ static int dm_any_congested(void *congested_data, int bdi_bits)
 /*-----------------------------------------------------------------
  * An IDR is used to keep track of allocated minor numbers.
  *---------------------------------------------------------------*/
-static DEFINE_IDR(_minor_idr);
-
 static void free_minor(int minor)
 {
 	spin_lock(&_minor_lock);
-- 
cgit v1.2.3


From 283a8328ca5b987e547848de8ff0e28edcfb9e08 Mon Sep 17 00:00:00 2001
From: Alasdair G Kergon <agk@redhat.com>
Date: Tue, 2 Aug 2011 12:32:01 +0100
Subject: dm: suppress endian warnings

Suppress sparse warnings about cpu_to_le32() by using __le32 types for
on-disk data etc.

Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-crypt.c           |  6 ++--
 drivers/md/dm-log.c             | 20 ++++++++----
 drivers/md/dm-snap-persistent.c | 71 ++++++++++++++++++++++-------------------
 3 files changed, 54 insertions(+), 43 deletions(-)

diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index bae6c4e23d3f..f5406766ece3 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -239,7 +239,7 @@ static int crypt_iv_plain_gen(struct crypt_config *cc, u8 *iv,
 			      struct dm_crypt_request *dmreq)
 {
 	memset(iv, 0, cc->iv_size);
-	*(u32 *)iv = cpu_to_le32(dmreq->iv_sector & 0xffffffff);
+	*(__le32 *)iv = cpu_to_le32(dmreq->iv_sector & 0xffffffff);
 
 	return 0;
 }
@@ -248,7 +248,7 @@ static int crypt_iv_plain64_gen(struct crypt_config *cc, u8 *iv,
 				struct dm_crypt_request *dmreq)
 {
 	memset(iv, 0, cc->iv_size);
-	*(u64 *)iv = cpu_to_le64(dmreq->iv_sector);
+	*(__le64 *)iv = cpu_to_le64(dmreq->iv_sector);
 
 	return 0;
 }
@@ -415,7 +415,7 @@ static int crypt_iv_essiv_gen(struct crypt_config *cc, u8 *iv,
 	struct crypto_cipher *essiv_tfm = this_crypt_config(cc)->iv_private;
 
 	memset(iv, 0, cc->iv_size);
-	*(u64 *)iv = cpu_to_le64(dmreq->iv_sector);
+	*(__le64 *)iv = cpu_to_le64(dmreq->iv_sector);
 	crypto_cipher_encrypt_one(essiv_tfm, iv, iv);
 
 	return 0;
diff --git a/drivers/md/dm-log.c b/drivers/md/dm-log.c
index 948e3f4925bf..5f06fb687408 100644
--- a/drivers/md/dm-log.c
+++ b/drivers/md/dm-log.c
@@ -197,15 +197,21 @@ EXPORT_SYMBOL(dm_dirty_log_destroy);
 #define MIRROR_DISK_VERSION 2
 #define LOG_OFFSET 2
 
-struct log_header {
-	uint32_t magic;
+struct log_header_disk {
+	__le32 magic;
 
 	/*
 	 * Simple, incrementing version. no backward
 	 * compatibility.
 	 */
+	__le32 version;
+	__le64 nr_regions;
+} __packed;
+
+struct log_header_core {
+	uint32_t magic;
 	uint32_t version;
-	sector_t nr_regions;
+	uint64_t nr_regions;
 };
 
 struct log_c {
@@ -239,10 +245,10 @@ struct log_c {
 	int log_dev_failed;
 	int log_dev_flush_failed;
 	struct dm_dev *log_dev;
-	struct log_header header;
+	struct log_header_core header;
 
 	struct dm_io_region header_location;
-	struct log_header *disk_header;
+	struct log_header_disk *disk_header;
 };
 
 /*
@@ -271,14 +277,14 @@ static inline void log_clear_bit(struct log_c *l,
 /*----------------------------------------------------------------
  * Header IO
  *--------------------------------------------------------------*/
-static void header_to_disk(struct log_header *core, struct log_header *disk)
+static void header_to_disk(struct log_header_core *core, struct log_header_disk *disk)
 {
 	disk->magic = cpu_to_le32(core->magic);
 	disk->version = cpu_to_le32(core->version);
 	disk->nr_regions = cpu_to_le64(core->nr_regions);
 }
 
-static void header_from_disk(struct log_header *core, struct log_header *disk)
+static void header_from_disk(struct log_header_core *core, struct log_header_disk *disk)
 {
 	core->magic = le32_to_cpu(disk->magic);
 	core->version = le32_to_cpu(disk->version);
diff --git a/drivers/md/dm-snap-persistent.c b/drivers/md/dm-snap-persistent.c
index e4ecadf0548a..39becbec4dfe 100644
--- a/drivers/md/dm-snap-persistent.c
+++ b/drivers/md/dm-snap-persistent.c
@@ -58,25 +58,30 @@
 #define NUM_SNAPSHOT_HDR_CHUNKS 1
 
 struct disk_header {
-	uint32_t magic;
+	__le32 magic;
 
 	/*
 	 * Is this snapshot valid.  There is no way of recovering
 	 * an invalid snapshot.
 	 */
-	uint32_t valid;
+	__le32 valid;
 
 	/*
 	 * Simple, incrementing version. no backward
 	 * compatibility.
 	 */
-	uint32_t version;
+	__le32 version;
 
 	/* In sectors */
-	uint32_t chunk_size;
-};
+	__le32 chunk_size;
+} __packed;
 
 struct disk_exception {
+	__le64 old_chunk;
+	__le64 new_chunk;
+} __packed;
+
+struct core_exception {
 	uint64_t old_chunk;
 	uint64_t new_chunk;
 };
@@ -396,32 +401,32 @@ static struct disk_exception *get_exception(struct pstore *ps, uint32_t index)
 }
 
 static void read_exception(struct pstore *ps,
-			   uint32_t index, struct disk_exception *result)
+			   uint32_t index, struct core_exception *result)
 {
-	struct disk_exception *e = get_exception(ps, index);
+	struct disk_exception *de = get_exception(ps, index);
 
 	/* copy it */
-	result->old_chunk = le64_to_cpu(e->old_chunk);
-	result->new_chunk = le64_to_cpu(e->new_chunk);
+	result->old_chunk = le64_to_cpu(de->old_chunk);
+	result->new_chunk = le64_to_cpu(de->new_chunk);
 }
 
 static void write_exception(struct pstore *ps,
-			    uint32_t index, struct disk_exception *de)
+			    uint32_t index, struct core_exception *e)
 {
-	struct disk_exception *e = get_exception(ps, index);
+	struct disk_exception *de = get_exception(ps, index);
 
 	/* copy it */
-	e->old_chunk = cpu_to_le64(de->old_chunk);
-	e->new_chunk = cpu_to_le64(de->new_chunk);
+	de->old_chunk = cpu_to_le64(e->old_chunk);
+	de->new_chunk = cpu_to_le64(e->new_chunk);
 }
 
 static void clear_exception(struct pstore *ps, uint32_t index)
 {
-	struct disk_exception *e = get_exception(ps, index);
+	struct disk_exception *de = get_exception(ps, index);
 
 	/* clear it */
-	e->old_chunk = 0;
-	e->new_chunk = 0;
+	de->old_chunk = 0;
+	de->new_chunk = 0;
 }
 
 /*
@@ -437,13 +442,13 @@ static int insert_exceptions(struct pstore *ps,
 {
 	int r;
 	unsigned int i;
-	struct disk_exception de;
+	struct core_exception e;
 
 	/* presume the area is full */
 	*full = 1;
 
 	for (i = 0; i < ps->exceptions_per_area; i++) {
-		read_exception(ps, i, &de);
+		read_exception(ps, i, &e);
 
 		/*
 		 * If the new_chunk is pointing at the start of
@@ -451,7 +456,7 @@ static int insert_exceptions(struct pstore *ps,
 		 * is we know that we've hit the end of the
 		 * exceptions.  Therefore the area is not full.
 		 */
-		if (de.new_chunk == 0LL) {
+		if (e.new_chunk == 0LL) {
 			ps->current_committed = i;
 			*full = 0;
 			break;
@@ -460,13 +465,13 @@ static int insert_exceptions(struct pstore *ps,
 		/*
 		 * Keep track of the start of the free chunks.
 		 */
-		if (ps->next_free <= de.new_chunk)
-			ps->next_free = de.new_chunk + 1;
+		if (ps->next_free <= e.new_chunk)
+			ps->next_free = e.new_chunk + 1;
 
 		/*
 		 * Otherwise we add the exception to the snapshot.
 		 */
-		r = callback(callback_context, de.old_chunk, de.new_chunk);
+		r = callback(callback_context, e.old_chunk, e.new_chunk);
 		if (r)
 			return r;
 	}
@@ -641,12 +646,12 @@ static void persistent_commit_exception(struct dm_exception_store *store,
 {
 	unsigned int i;
 	struct pstore *ps = get_info(store);
-	struct disk_exception de;
+	struct core_exception ce;
 	struct commit_callback *cb;
 
-	de.old_chunk = e->old_chunk;
-	de.new_chunk = e->new_chunk;
-	write_exception(ps, ps->current_committed++, &de);
+	ce.old_chunk = e->old_chunk;
+	ce.new_chunk = e->new_chunk;
+	write_exception(ps, ps->current_committed++, &ce);
 
 	/*
 	 * Add the callback to the back of the array.  This code
@@ -701,7 +706,7 @@ static int persistent_prepare_merge(struct dm_exception_store *store,
 				    chunk_t *last_new_chunk)
 {
 	struct pstore *ps = get_info(store);
-	struct disk_exception de;
+	struct core_exception ce;
 	int nr_consecutive;
 	int r;
 
@@ -722,9 +727,9 @@ static int persistent_prepare_merge(struct dm_exception_store *store,
 		ps->current_committed = ps->exceptions_per_area;
 	}
 
-	read_exception(ps, ps->current_committed - 1, &de);
-	*last_old_chunk = de.old_chunk;
-	*last_new_chunk = de.new_chunk;
+	read_exception(ps, ps->current_committed - 1, &ce);
+	*last_old_chunk = ce.old_chunk;
+	*last_new_chunk = ce.new_chunk;
 
 	/*
 	 * Find number of consecutive chunks within the current area,
@@ -733,9 +738,9 @@ static int persistent_prepare_merge(struct dm_exception_store *store,
 	for (nr_consecutive = 1; nr_consecutive < ps->current_committed;
 	     nr_consecutive++) {
 		read_exception(ps, ps->current_committed - 1 - nr_consecutive,
-			       &de);
-		if (de.old_chunk != *last_old_chunk - nr_consecutive ||
-		    de.new_chunk != *last_new_chunk - nr_consecutive)
+			       &ce);
+		if (ce.old_chunk != *last_old_chunk - nr_consecutive ||
+		    ce.new_chunk != *last_new_chunk - nr_consecutive)
 			break;
 	}
 
-- 
cgit v1.2.3


From 936688d7eb0f39be96c5791be1a04994cc8d6aa0 Mon Sep 17 00:00:00 2001
From: Mike Snitzer <snitzer@redhat.com>
Date: Tue, 2 Aug 2011 12:32:01 +0100
Subject: dm table: fix discard support

Remove 'discards_supported' from the dm_table structure.  The same
information can be easily discovered from the table's target(s) in
dm_table_supports_discards().

Before this fix dm_table_supports_discards() would skip checking the
individual targets' 'discards_supported' flag if any one target in the
table didn't set num_discard_requests > 0.  Now the per-target
'discards_supported' flag is effective at insuring the final DM device
advertises discard support.  But, to be clear, targets that don't
support discards (!num_discard_requests) will not receive discard
requests.

Also DMWARN if a target sets 'discards_supported' override but forgets
to set 'num_discard_requests'.

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-table.c | 15 +++++++--------
 drivers/md/dm.c       |  3 ++-
 2 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index bfe9c2333cea..3909fa259f55 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -54,7 +54,6 @@ struct dm_table {
 	sector_t *highs;
 	struct dm_target *targets;
 
-	unsigned discards_supported:1;
 	unsigned integrity_supported:1;
 
 	/*
@@ -209,7 +208,6 @@ int dm_table_create(struct dm_table **result, fmode_t mode,
 	INIT_LIST_HEAD(&t->devices);
 	INIT_LIST_HEAD(&t->target_callbacks);
 	atomic_set(&t->holders, 0);
-	t->discards_supported = 1;
 
 	if (!num_targets)
 		num_targets = KEYS_PER_NODE;
@@ -791,8 +789,9 @@ int dm_table_add_target(struct dm_table *t, const char *type,
 
 	t->highs[t->num_targets++] = tgt->begin + tgt->len - 1;
 
-	if (!tgt->num_discard_requests)
-		t->discards_supported = 0;
+	if (!tgt->num_discard_requests && tgt->discards_supported)
+		DMWARN("%s: %s: ignoring discards_supported because num_discard_requests is zero.",
+		       dm_device_name(t->md), type);
 
 	return 0;
 
@@ -1359,19 +1358,19 @@ bool dm_table_supports_discards(struct dm_table *t)
 	struct dm_target *ti;
 	unsigned i = 0;
 
-	if (!t->discards_supported)
-		return 0;
-
 	/*
 	 * Unless any target used by the table set discards_supported,
 	 * require at least one underlying device to support discards.
 	 * t->devices includes internal dm devices such as mirror logs
 	 * so we need to use iterate_devices here, which targets
-	 * supporting discard must provide.
+	 * supporting discard selectively must provide.
 	 */
 	while (i < dm_table_get_num_targets(t)) {
 		ti = dm_table_get_target(t, i++);
 
+		if (!ti->num_discard_requests)
+			continue;
+
 		if (ti->discards_supported)
 			return 1;
 
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 41abc6dd481b..aeb0fa1ccfe4 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -1179,7 +1179,8 @@ static int __clone_and_map_discard(struct clone_info *ci)
 
 		/*
 		 * Even though the device advertised discard support,
-		 * reconfiguration might have changed that since the
+		 * that does not mean every target supports it, and
+		 * reconfiguration might also have changed that since the
 		 * check was performed.
 		 */
 		if (!ti->num_discard_requests)
-- 
cgit v1.2.3


From c8f543e0786785d8c7118fc2878e42bc0193a88d Mon Sep 17 00:00:00 2001
From: Akinobu Mita <akinobu.mita@gmail.com>
Date: Tue, 2 Aug 2011 12:32:01 +0100
Subject: dm log: clean up bit little endian bitops

Using __test_and_{set,clear}_bit_le() with ignoring its return value
can be replaced with __{set,clear}_bit_le().

This also removes unnecessary casts.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-log.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/drivers/md/dm-log.c b/drivers/md/dm-log.c
index 5f06fb687408..306ce123d6b4 100644
--- a/drivers/md/dm-log.c
+++ b/drivers/md/dm-log.c
@@ -257,20 +257,20 @@ struct log_c {
  */
 static inline int log_test_bit(uint32_t *bs, unsigned bit)
 {
-	return test_bit_le(bit, (unsigned long *) bs) ? 1 : 0;
+	return test_bit_le(bit, bs) ? 1 : 0;
 }
 
 static inline void log_set_bit(struct log_c *l,
 			       uint32_t *bs, unsigned bit)
 {
-	__test_and_set_bit_le(bit, (unsigned long *) bs);
+	__set_bit_le(bit, bs);
 	l->touched_cleaned = 1;
 }
 
 static inline void log_clear_bit(struct log_c *l,
 				 uint32_t *bs, unsigned bit)
 {
-	__test_and_clear_bit_le(bit, (unsigned long *) bs);
+	__clear_bit_le(bit, bs);
 	l->touched_dirtied = 1;
 }
 
@@ -745,8 +745,7 @@ static int core_get_resync_work(struct dm_dirty_log *log, region_t *region)
 		return 0;
 
 	do {
-		*region = find_next_zero_bit_le(
-					     (unsigned long *) lc->sync_bits,
+		*region = find_next_zero_bit_le(lc->sync_bits,
 					     lc->region_count,
 					     lc->sync_search);
 		lc->sync_search = *region + 1;
-- 
cgit v1.2.3


From 6c9b27ab08aaf46426515b8b858ad9c60731c7a1 Mon Sep 17 00:00:00 2001
From: "Kirill A. Shutemov" <kirill@shutemov.name>
Date: Tue, 2 Aug 2011 12:32:02 +0100
Subject: dm log: userspace use list_move

Replace list_del() followed by list_add() with list_move().

Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-log-userspace-base.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/md/dm-log-userspace-base.c b/drivers/md/dm-log-userspace-base.c
index aa2e0c374ab3..1021c8986011 100644
--- a/drivers/md/dm-log-userspace-base.c
+++ b/drivers/md/dm-log-userspace-base.c
@@ -394,8 +394,7 @@ static int flush_by_group(struct log_c *lc, struct list_head *flush_list)
 			group[count] = fe->region;
 			count++;
 
-			list_del(&fe->list);
-			list_add(&fe->list, &tmp_list);
+			list_move(&fe->list, &tmp_list);
 
 			type = fe->type;
 			if (count >= MAX_FLUSH_GROUP_COUNT)
-- 
cgit v1.2.3


From e29e65aacbd9e628378084905cbcf62a9fa4a8cc Mon Sep 17 00:00:00 2001
From: Joe Perches <joe@perches.com>
Date: Tue, 2 Aug 2011 12:32:02 +0100
Subject: dm: use vzalloc

Use vzalloc() instead of vmalloc()+memset().

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-log.c             | 3 +--
 drivers/md/dm-snap-persistent.c | 3 +--
 drivers/md/dm-table.c           | 4 +---
 3 files changed, 3 insertions(+), 7 deletions(-)

diff --git a/drivers/md/dm-log.c b/drivers/md/dm-log.c
index 306ce123d6b4..3b52bb72bd1f 100644
--- a/drivers/md/dm-log.c
+++ b/drivers/md/dm-log.c
@@ -492,7 +492,7 @@ static int create_log_context(struct dm_dirty_log *log, struct dm_target *ti,
 	memset(lc->sync_bits, (sync == NOSYNC) ? -1 : 0, bitset_size);
 	lc->sync_count = (sync == NOSYNC) ? region_count : 0;
 
-	lc->recovering_bits = vmalloc(bitset_size);
+	lc->recovering_bits = vzalloc(bitset_size);
 	if (!lc->recovering_bits) {
 		DMWARN("couldn't allocate sync bitset");
 		vfree(lc->sync_bits);
@@ -504,7 +504,6 @@ static int create_log_context(struct dm_dirty_log *log, struct dm_target *ti,
 		kfree(lc);
 		return -ENOMEM;
 	}
-	memset(lc->recovering_bits, 0, bitset_size);
 	lc->sync_search = 0;
 	log->context = lc;
 
diff --git a/drivers/md/dm-snap-persistent.c b/drivers/md/dm-snap-persistent.c
index 39becbec4dfe..1a0acb8abdf6 100644
--- a/drivers/md/dm-snap-persistent.c
+++ b/drivers/md/dm-snap-persistent.c
@@ -174,10 +174,9 @@ static int alloc_area(struct pstore *ps)
 	if (!ps->area)
 		goto err_area;
 
-	ps->zero_area = vmalloc(len);
+	ps->zero_area = vzalloc(len);
 	if (!ps->zero_area)
 		goto err_zero_area;
-	memset(ps->zero_area, 0, len);
 
 	ps->header_area = vmalloc(len);
 	if (!ps->header_area)
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index 3909fa259f55..8dc67555e736 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -153,9 +153,7 @@ void *dm_vcalloc(unsigned long nmemb, unsigned long elem_size)
 		return NULL;
 
 	size = nmemb * elem_size;
-	addr = vmalloc(size);
-	if (addr)
-		memset(addr, 0, size);
+	addr = vzalloc(size);
 
 	return addr;
 }
-- 
cgit v1.2.3


From 4622afb3f50e03ce6da42002e7ed3675dfafc187 Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 2 Aug 2011 12:32:02 +0100
Subject: dm kcopyd: remove offset field from job structure

The offset field in struct kcopyd_job is always zero so remove it.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-kcopyd.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/md/dm-kcopyd.c b/drivers/md/dm-kcopyd.c
index 320401dec104..e7926fa1eef2 100644
--- a/drivers/md/dm-kcopyd.c
+++ b/drivers/md/dm-kcopyd.c
@@ -224,7 +224,6 @@ struct kcopyd_job {
 	unsigned int num_dests;
 	struct dm_io_region dests[DM_KCOPYD_MAX_REGIONS];
 
-	sector_t offset;
 	unsigned int nr_pages;
 	struct page_list *pages;
 
@@ -380,7 +379,7 @@ static int run_io_job(struct kcopyd_job *job)
 		.bi_rw = job->rw,
 		.mem.type = DM_IO_PAGE_LIST,
 		.mem.ptr.pl = job->pages,
-		.mem.offset = job->offset,
+		.mem.offset = 0,
 		.notify.fn = complete_io,
 		.notify.context = job,
 		.client = job->kc->io_client,
@@ -398,8 +397,7 @@ static int run_pages_job(struct kcopyd_job *job)
 {
 	int r;
 
-	job->nr_pages = dm_div_up(job->dests[0].count + job->offset,
-				  PAGE_SIZE >> 9);
+	job->nr_pages = dm_div_up(job->dests[0].count, PAGE_SIZE >> 9);
 	r = kcopyd_get_pages(job->kc, job->nr_pages, &job->pages);
 	if (!r) {
 		/* this job is ready for io */
@@ -602,7 +600,6 @@ int dm_kcopyd_copy(struct dm_kcopyd_client *kc, struct dm_io_region *from,
 	job->num_dests = num_dests;
 	memcpy(&job->dests, dests, sizeof(*dests) * num_dests);
 
-	job->offset = 0;
 	job->nr_pages = 0;
 	job->pages = NULL;
 
-- 
cgit v1.2.3


From 5bf45a3dcdba9ff43959f7b5b44523fab254c19c Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 2 Aug 2011 12:32:02 +0100
Subject: dm kcopyd: remove nr_pages field from job structure

The nr_pages field in struct kcopyd_job is only used temporarily in
run_pages_job() to count the number of required pages.
We can use a local variable instead.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-kcopyd.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/md/dm-kcopyd.c b/drivers/md/dm-kcopyd.c
index e7926fa1eef2..98725e119324 100644
--- a/drivers/md/dm-kcopyd.c
+++ b/drivers/md/dm-kcopyd.c
@@ -224,7 +224,6 @@ struct kcopyd_job {
 	unsigned int num_dests;
 	struct dm_io_region dests[DM_KCOPYD_MAX_REGIONS];
 
-	unsigned int nr_pages;
 	struct page_list *pages;
 
 	/*
@@ -396,9 +395,9 @@ static int run_io_job(struct kcopyd_job *job)
 static int run_pages_job(struct kcopyd_job *job)
 {
 	int r;
+	unsigned nr_pages = dm_div_up(job->dests[0].count, PAGE_SIZE >> 9);
 
-	job->nr_pages = dm_div_up(job->dests[0].count, PAGE_SIZE >> 9);
-	r = kcopyd_get_pages(job->kc, job->nr_pages, &job->pages);
+	r = kcopyd_get_pages(job->kc, nr_pages, &job->pages);
 	if (!r) {
 		/* this job is ready for io */
 		push(&job->kc->io_jobs, job);
@@ -600,7 +599,6 @@ int dm_kcopyd_copy(struct dm_kcopyd_client *kc, struct dm_io_region *from,
 	job->num_dests = num_dests;
 	memcpy(&job->dests, dests, sizeof(*dests) * num_dests);
 
-	job->nr_pages = 0;
 	job->pages = NULL;
 
 	job->fn = fn;
-- 
cgit v1.2.3


From aa3f0794d279cd154ac100f92ff3904ea1f56de2 Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 2 Aug 2011 12:32:03 +0100
Subject: dm snapshot: remove unused definitions

Remove a couple of unused #defines.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-snap.c | 10 ----------
 1 file changed, 10 deletions(-)

diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index 9ecff5f3023a..9d6daa0eaff6 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -29,16 +29,6 @@ static const char dm_snapshot_merge_target_name[] = "snapshot-merge";
 #define dm_target_is_snapshot_merge(ti) \
 	((ti)->type->name == dm_snapshot_merge_target_name)
 
-/*
- * The percentage increment we will wake up users at
- */
-#define WAKE_UP_PERCENT 5
-
-/*
- * kcopyd priority of snapshot operations
- */
-#define SNAPSHOT_COPY_PRIORITY 2
-
 /*
  * The size of the mempool used to track chunks in use.
  */
-- 
cgit v1.2.3


From a2d2b0345a0f30c169b7d08b8cebdd4853fcb0f8 Mon Sep 17 00:00:00 2001
From: Jonathan Brassow <jbrassow@redhat.com>
Date: Tue, 2 Aug 2011 12:32:03 +0100
Subject: dm snapshot: style cleanups

Coding style cleanups.

Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
---
 drivers/md/dm-snap-persistent.c |  4 ++--
 drivers/md/dm-snap.c            | 14 ++++++--------
 2 files changed, 8 insertions(+), 10 deletions(-)

diff --git a/drivers/md/dm-snap-persistent.c b/drivers/md/dm-snap-persistent.c
index 1a0acb8abdf6..d1f1d7017103 100644
--- a/drivers/md/dm-snap-persistent.c
+++ b/drivers/md/dm-snap-persistent.c
@@ -567,7 +567,7 @@ static int persistent_read_metadata(struct dm_exception_store *store,
 	ps->exceptions_per_area = (ps->store->chunk_size << SECTOR_SHIFT) /
 				  sizeof(struct disk_exception);
 	ps->callbacks = dm_vcalloc(ps->exceptions_per_area,
-			sizeof(*ps->callbacks));
+				   sizeof(*ps->callbacks));
 	if (!ps->callbacks)
 		return -ENOMEM;
 
@@ -674,7 +674,7 @@ static void persistent_commit_exception(struct dm_exception_store *store,
 	 * If we completely filled the current area, then wipe the next one.
 	 */
 	if ((ps->current_committed == ps->exceptions_per_area) &&
-	     zero_disk_area(ps, ps->current_area + 1))
+	    zero_disk_area(ps, ps->current_area + 1))
 		ps->valid = 0;
 
 	/*
diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index 9d6daa0eaff6..94dee05dd28e 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -1045,8 +1045,7 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 
 	s = kmalloc(sizeof(*s), GFP_KERNEL);
 	if (!s) {
-		ti->error = "Cannot allocate snapshot context private "
-		    "structure";
+		ti->error = "Cannot allocate private snapshot structure";
 		r = -ENOMEM;
 		goto bad;
 	}
@@ -1405,7 +1404,7 @@ static void pending_complete(struct dm_snap_pending_exception *pe, int success)
 	 */
 	dm_insert_exception(&s->complete, e);
 
- out:
+out:
 	dm_remove_exception(&pe->e);
 	snapshot_bios = bio_list_get(&pe->snapshot_bios);
 	origin_bios = bio_list_get(&pe->origin_bios);
@@ -1470,8 +1469,7 @@ static void start_copy(struct dm_snap_pending_exception *pe)
 	dest.count = src.count;
 
 	/* Hand over to kcopyd */
-	dm_kcopyd_copy(s->kcopyd_client,
-		    &src, 1, &dest, 0, copy_callback, pe);
+	dm_kcopyd_copy(s->kcopyd_client, &src, 1, &dest, 0, copy_callback, pe);
 }
 
 static struct dm_snap_pending_exception *
@@ -1618,9 +1616,9 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio,
 		map_context->ptr = track_chunk(s, chunk);
 	}
 
- out_unlock:
+out_unlock:
 	up_write(&s->lock);
- out:
+out:
 	return r;
 }
 
@@ -1964,7 +1962,7 @@ static int __origin_write(struct list_head *snapshots, sector_t sector,
 			pe_to_start_now = pe;
 		}
 
- next_snapshot:
+next_snapshot:
 		up_write(&snap->lock);
 
 		if (pe_to_start_now) {
-- 
cgit v1.2.3


From 13c87583ea4e867211fc3e7edab750c353c47c95 Mon Sep 17 00:00:00 2001
From: Jonathan Brassow <jbrassow@redhat.com>
Date: Tue, 2 Aug 2011 12:32:03 +0100
Subject: dm raid: cleanup parameter handling

Re-order the parameters so they are handled consistently in the same order
where defined, parsed and output.

Only include rebuild parameters in the STATUSTYPE_TABLE output if they were
supplied in the original table line.

Correct the parameter count when outputting rebuild: there are two words,
not one.

Use case-independent checks for keywords (as in other device-mapper targets).

Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-raid.c | 42 +++++++++++++++++++++++-------------------
 1 file changed, 23 insertions(+), 19 deletions(-)

diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c
index e5d8904fc8f6..efa960ff5ba4 100644
--- a/drivers/md/dm-raid.c
+++ b/drivers/md/dm-raid.c
@@ -43,13 +43,14 @@ struct raid_dev {
 /*
  * Flags for rs->print_flags field.
  */
-#define DMPF_DAEMON_SLEEP      0x1
-#define DMPF_MAX_WRITE_BEHIND  0x2
-#define DMPF_SYNC              0x4
-#define DMPF_NOSYNC            0x8
-#define DMPF_STRIPE_CACHE      0x10
-#define DMPF_MIN_RECOVERY_RATE 0x20
-#define DMPF_MAX_RECOVERY_RATE 0x40
+#define DMPF_SYNC              0x1
+#define DMPF_NOSYNC            0x2
+#define DMPF_REBUILD           0x4
+#define DMPF_DAEMON_SLEEP      0x8
+#define DMPF_MIN_RECOVERY_RATE 0x10
+#define DMPF_MAX_RECOVERY_RATE 0x20
+#define DMPF_MAX_WRITE_BEHIND  0x40
+#define DMPF_STRIPE_CACHE      0x80
 
 struct raid_set {
 	struct dm_target *ti;
@@ -275,13 +276,13 @@ static int parse_raid_params(struct raid_set *rs, char **argv,
 		set_bit(In_sync, &rs->dev[i].rdev.flags);
 
 	for (i = 0; i < num_raid_params; i++) {
-		if (!strcmp(argv[i], "nosync")) {
+		if (!strcasecmp(argv[i], "nosync")) {
 			rs->md.recovery_cp = MaxSector;
 			rs->print_flags |= DMPF_NOSYNC;
 			rs->md.flags |= MD_SYNC_STATE_FORCED;
 			continue;
 		}
-		if (!strcmp(argv[i], "sync")) {
+		if (!strcasecmp(argv[i], "sync")) {
 			rs->md.recovery_cp = 0;
 			rs->print_flags |= DMPF_SYNC;
 			rs->md.flags |= MD_SYNC_STATE_FORCED;
@@ -300,7 +301,7 @@ static int parse_raid_params(struct raid_set *rs, char **argv,
 			return -EINVAL;
 		}
 
-		if (!strcmp(key, "rebuild")) {
+		if (!strcasecmp(key, "rebuild")) {
 			if (++rebuild_cnt > rs->raid_type->parity_devs) {
 				rs->ti->error = "Too many rebuild drives given";
 				return -EINVAL;
@@ -311,7 +312,8 @@ static int parse_raid_params(struct raid_set *rs, char **argv,
 			}
 			clear_bit(In_sync, &rs->dev[value].rdev.flags);
 			rs->dev[value].rdev.recovery_offset = 0;
-		} else if (!strcmp(key, "max_write_behind")) {
+			rs->print_flags |= DMPF_REBUILD;
+		} else if (!strcasecmp(key, "max_write_behind")) {
 			rs->print_flags |= DMPF_MAX_WRITE_BEHIND;
 
 			/*
@@ -324,14 +326,14 @@ static int parse_raid_params(struct raid_set *rs, char **argv,
 				return -EINVAL;
 			}
 			rs->md.bitmap_info.max_write_behind = value;
-		} else if (!strcmp(key, "daemon_sleep")) {
+		} else if (!strcasecmp(key, "daemon_sleep")) {
 			rs->print_flags |= DMPF_DAEMON_SLEEP;
 			if (!value || (value > MAX_SCHEDULE_TIMEOUT)) {
 				rs->ti->error = "daemon sleep period out of range";
 				return -EINVAL;
 			}
 			rs->md.bitmap_info.daemon_sleep = value;
-		} else if (!strcmp(key, "stripe_cache")) {
+		} else if (!strcasecmp(key, "stripe_cache")) {
 			rs->print_flags |= DMPF_STRIPE_CACHE;
 
 			/*
@@ -348,14 +350,14 @@ static int parse_raid_params(struct raid_set *rs, char **argv,
 				rs->ti->error = "Bad stripe_cache size";
 				return -EINVAL;
 			}
-		} else if (!strcmp(key, "min_recovery_rate")) {
+		} else if (!strcasecmp(key, "min_recovery_rate")) {
 			rs->print_flags |= DMPF_MIN_RECOVERY_RATE;
 			if (value > INT_MAX) {
 				rs->ti->error = "min_recovery_rate out of range";
 				return -EINVAL;
 			}
 			rs->md.sync_speed_min = (int)value;
-		} else if (!strcmp(key, "max_recovery_rate")) {
+		} else if (!strcasecmp(key, "max_recovery_rate")) {
 			rs->print_flags |= DMPF_MAX_RECOVERY_RATE;
 			if (value > INT_MAX) {
 				rs->ti->error = "max_recovery_rate out of range";
@@ -547,11 +549,12 @@ static int raid_status(struct dm_target *ti, status_type_t type,
 	case STATUSTYPE_TABLE:
 		/* The string you would use to construct this array */
 		for (i = 0; i < rs->md.raid_disks; i++)
-			if (rs->dev[i].data_dev &&
+			if ((rs->print_flags & DMPF_REBUILD) &&
+			    rs->dev[i].data_dev &&
 			    !test_bit(In_sync, &rs->dev[i].rdev.flags))
-				raid_param_cnt++; /* for rebuilds */
+				raid_param_cnt += 2; /* for rebuilds */
 
-		raid_param_cnt += (hweight64(rs->print_flags) * 2);
+		raid_param_cnt += (hweight64(rs->print_flags & ~DMPF_REBUILD) * 2);
 		if (rs->print_flags & (DMPF_SYNC | DMPF_NOSYNC))
 			raid_param_cnt--;
 
@@ -565,7 +568,8 @@ static int raid_status(struct dm_target *ti, status_type_t type,
 			DMEMIT(" nosync");
 
 		for (i = 0; i < rs->md.raid_disks; i++)
-			if (rs->dev[i].data_dev &&
+			if ((rs->print_flags & DMPF_REBUILD) &&
+			    rs->dev[i].data_dev &&
 			    !test_bit(In_sync, &rs->dev[i].rdev.flags))
 				DMEMIT(" rebuild %u", i);
 
-- 
cgit v1.2.3


From 2ca4c92f58f9386e080b26f9ccd78c9ca5825a42 Mon Sep 17 00:00:00 2001
From: Alasdair G Kergon <agk@redhat.com>
Date: Tue, 2 Aug 2011 12:32:03 +0100
Subject: dm ioctl: prevent empty message

Detect invalid empty messages in core dm instead of requiring every target to
check this.

Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-ioctl.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index 4cacdad2270a..1622a6bc0bfd 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -1402,6 +1402,11 @@ static int target_message(struct dm_ioctl *param, size_t param_size)
 		goto out;
 	}
 
+	if (!argc) {
+		DMWARN("Empty message received.");
+		goto out;
+	}
+
 	table = dm_get_live_table(md);
 	if (!table)
 		goto out_argv;
-- 
cgit v1.2.3


From 3e8dbb7f3966c80d77272f8b4f430babc99f6595 Mon Sep 17 00:00:00 2001
From: Alasdair G Kergon <agk@redhat.com>
Date: Tue, 2 Aug 2011 12:32:03 +0100
Subject: dm raid: tidy includes

A dm target only needs to use include/linux dm headers.

Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-raid.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c
index efa960ff5ba4..d4e95b2e39f6 100644
--- a/drivers/md/dm-raid.c
+++ b/drivers/md/dm-raid.c
@@ -9,9 +9,10 @@
 
 #include "md.h"
 #include "raid5.h"
-#include "dm.h"
 #include "bitmap.h"
 
+#include <linux/device-mapper.h>
+
 #define DM_MSG_PREFIX "raid"
 
 /*
-- 
cgit v1.2.3


From 08649012545cfb116798260352547cf4d47064ec Mon Sep 17 00:00:00 2001
From: Mike Snitzer <snitzer@redhat.com>
Date: Tue, 2 Aug 2011 12:32:04 +0100
Subject: dm table: clean dm_get_device and move exports

There is no need for __table_get_device to be factored out.
Also move the exports to the end of their respective functions.

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-table.c | 33 +++++++++++++--------------------
 1 file changed, 13 insertions(+), 20 deletions(-)

diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index 8dc67555e736..d58b200a3531 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -157,6 +157,7 @@ void *dm_vcalloc(unsigned long nmemb, unsigned long elem_size)
 
 	return addr;
 }
+EXPORT_SYMBOL(dm_vcalloc);
 
 /*
  * highs, and targets are managed as dynamic arrays during a
@@ -277,6 +278,7 @@ void dm_table_get(struct dm_table *t)
 {
 	atomic_inc(&t->holders);
 }
+EXPORT_SYMBOL(dm_table_get);
 
 void dm_table_put(struct dm_table *t)
 {
@@ -286,6 +288,7 @@ void dm_table_put(struct dm_table *t)
 	smp_mb__before_atomic_dec();
 	atomic_dec(&t->holders);
 }
+EXPORT_SYMBOL(dm_table_put);
 
 /*
  * Checks to see if we need to extend highs or targets.
@@ -451,13 +454,14 @@ static int upgrade_mode(struct dm_dev_internal *dd, fmode_t new_mode,
  * Add a device to the list, or just increment the usage count if
  * it's already present.
  */
-static int __table_get_device(struct dm_table *t, struct dm_target *ti,
-		      const char *path, fmode_t mode, struct dm_dev **result)
+int dm_get_device(struct dm_target *ti, const char *path, fmode_t mode,
+		  struct dm_dev **result)
 {
 	int r;
 	dev_t uninitialized_var(dev);
 	struct dm_dev_internal *dd;
 	unsigned int major, minor;
+	struct dm_table *t = ti->table;
 
 	BUG_ON(!t);
 
@@ -505,6 +509,7 @@ static int __table_get_device(struct dm_table *t, struct dm_target *ti,
 	*result = &dd->dm_dev;
 	return 0;
 }
+EXPORT_SYMBOL(dm_get_device);
 
 int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev,
 			 sector_t start, sector_t len, void *data)
@@ -543,15 +548,8 @@ int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev,
 }
 EXPORT_SYMBOL_GPL(dm_set_device_limits);
 
-int dm_get_device(struct dm_target *ti, const char *path, fmode_t mode,
-		  struct dm_dev **result)
-{
-	return __table_get_device(ti->table, ti, path, mode, result);
-}
-
-
 /*
- * Decrement a devices use count and remove it if necessary.
+ * Decrement a device's use count and remove it if necessary.
  */
 void dm_put_device(struct dm_target *ti, struct dm_dev *d)
 {
@@ -564,6 +562,7 @@ void dm_put_device(struct dm_target *ti, struct dm_dev *d)
 		kfree(dd);
 	}
 }
+EXPORT_SYMBOL(dm_put_device);
 
 /*
  * Checks to see if the target joins onto the end of the table.
@@ -1074,11 +1073,13 @@ void dm_table_event(struct dm_table *t)
 		t->event_fn(t->event_context);
 	mutex_unlock(&_event_lock);
 }
+EXPORT_SYMBOL(dm_table_event);
 
 sector_t dm_table_get_size(struct dm_table *t)
 {
 	return t->num_targets ? (t->highs[t->num_targets - 1] + 1) : 0;
 }
+EXPORT_SYMBOL(dm_table_get_size);
 
 struct dm_target *dm_table_get_target(struct dm_table *t, unsigned int index)
 {
@@ -1234,6 +1235,7 @@ fmode_t dm_table_get_mode(struct dm_table *t)
 {
 	return t->mode;
 }
+EXPORT_SYMBOL(dm_table_get_mode);
 
 static void suspend_targets(struct dm_table *t, unsigned postsuspend)
 {
@@ -1342,6 +1344,7 @@ struct mapped_device *dm_table_get_md(struct dm_table *t)
 {
 	return t->md;
 }
+EXPORT_SYMBOL(dm_table_get_md);
 
 static int device_discard_capable(struct dm_target *ti, struct dm_dev *dev,
 				  sector_t start, sector_t len, void *data)
@@ -1379,13 +1382,3 @@ bool dm_table_supports_discards(struct dm_table *t)
 
 	return 0;
 }
-
-EXPORT_SYMBOL(dm_vcalloc);
-EXPORT_SYMBOL(dm_get_device);
-EXPORT_SYMBOL(dm_put_device);
-EXPORT_SYMBOL(dm_table_event);
-EXPORT_SYMBOL(dm_table_get_size);
-EXPORT_SYMBOL(dm_table_get_mode);
-EXPORT_SYMBOL(dm_table_get_md);
-EXPORT_SYMBOL(dm_table_put);
-EXPORT_SYMBOL(dm_table_get);
-- 
cgit v1.2.3


From d5b9dd04bd74b774b8e8d93ced7a0d15ad403fa9 Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 2 Aug 2011 12:32:04 +0100
Subject: dm: ignore merge_bvec for snapshots when safe

Add a new flag DMF_MERGE_IS_OPTIONAL to struct mapped_device to indicate
whether the device can accept bios larger than the size its merge
function returns.  When set, use this to send large bios to snapshots
which can split them if necessary.  Snapshot I/O may be significantly
fragmented and this approach seems to improve peformance.

Before the patch, dm_set_device_limits restricted bio size to page size
if the underlying device had a merge function and the target didn't
provide a merge function.  After the patch, dm_set_device_limits
restricts bio size to page size if the underlying device has a merge
function, doesn't have DMF_MERGE_IS_OPTIONAL flag and the target doesn't
provide a merge function.

The snapshot target can't provide a merge function because when the merge
function is called, it is impossible to determine where the bio will be
remapped.  Previously this led us to impose a 4k limit, which we can
now remove if the snapshot store is located on a device without a merge
function.  Together with another patch for optimizing full chunk writes,
it improves performance from 29MB/s to 40MB/s when writing to the
filesystem on snapshot store.

If the snapshot store is placed on a non-dm device with a merge function
(such as md-raid), device mapper still limits all bios to page size.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-table.c |  3 +--
 drivers/md/dm.c       | 61 +++++++++++++++++++++++++++++++++++++++++++++++++++
 drivers/md/dm.h       |  2 ++
 3 files changed, 64 insertions(+), 2 deletions(-)

diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index d58b200a3531..03516c7f5be1 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -540,8 +540,7 @@ int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev,
 	 * If not we'll force DM to use PAGE_SIZE or
 	 * smaller I/O, just to be safe.
 	 */
-
-	if (q->merge_bvec_fn && !ti->type->merge)
+	if (dm_queue_merge_is_compulsory(q) && !ti->type->merge)
 		blk_limits_max_hw_sectors(limits,
 					  (unsigned int) (PAGE_SIZE >> 9));
 	return 0;
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index aeb0fa1ccfe4..1000eaf984ef 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -111,6 +111,7 @@ EXPORT_SYMBOL_GPL(dm_get_rq_mapinfo);
 #define DMF_FREEING 3
 #define DMF_DELETING 4
 #define DMF_NOFLUSH_SUSPENDING 5
+#define DMF_MERGE_IS_OPTIONAL 6
 
 /*
  * Work processed by per-device workqueue.
@@ -1992,6 +1993,59 @@ static void __set_size(struct mapped_device *md, sector_t size)
 	i_size_write(md->bdev->bd_inode, (loff_t)size << SECTOR_SHIFT);
 }
 
+/*
+ * Return 1 if the queue has a compulsory merge_bvec_fn function.
+ *
+ * If this function returns 0, then the device is either a non-dm
+ * device without a merge_bvec_fn, or it is a dm device that is
+ * able to split any bios it receives that are too big.
+ */
+int dm_queue_merge_is_compulsory(struct request_queue *q)
+{
+	struct mapped_device *dev_md;
+
+	if (!q->merge_bvec_fn)
+		return 0;
+
+	if (q->make_request_fn == dm_request) {
+		dev_md = q->queuedata;
+		if (test_bit(DMF_MERGE_IS_OPTIONAL, &dev_md->flags))
+			return 0;
+	}
+
+	return 1;
+}
+
+static int dm_device_merge_is_compulsory(struct dm_target *ti,
+					 struct dm_dev *dev, sector_t start,
+					 sector_t len, void *data)
+{
+	struct block_device *bdev = dev->bdev;
+	struct request_queue *q = bdev_get_queue(bdev);
+
+	return dm_queue_merge_is_compulsory(q);
+}
+
+/*
+ * Return 1 if it is acceptable to ignore merge_bvec_fn based
+ * on the properties of the underlying devices.
+ */
+static int dm_table_merge_is_optional(struct dm_table *table)
+{
+	unsigned i = 0;
+	struct dm_target *ti;
+
+	while (i < dm_table_get_num_targets(table)) {
+		ti = dm_table_get_target(table, i++);
+
+		if (ti->type->iterate_devices &&
+		    ti->type->iterate_devices(ti, dm_device_merge_is_compulsory, NULL))
+			return 0;
+	}
+
+	return 1;
+}
+
 /*
  * Returns old map, which caller must destroy.
  */
@@ -2002,6 +2056,7 @@ static struct dm_table *__bind(struct mapped_device *md, struct dm_table *t,
 	struct request_queue *q = md->queue;
 	sector_t size;
 	unsigned long flags;
+	int merge_is_optional;
 
 	size = dm_table_get_size(t);
 
@@ -2027,10 +2082,16 @@ static struct dm_table *__bind(struct mapped_device *md, struct dm_table *t,
 
 	__bind_mempools(md, t);
 
+	merge_is_optional = dm_table_merge_is_optional(t);
+
 	write_lock_irqsave(&md->map_lock, flags);
 	old_map = md->map;
 	md->map = t;
 	dm_table_set_restrictions(t, q, limits);
+	if (merge_is_optional)
+		set_bit(DMF_MERGE_IS_OPTIONAL, &md->flags);
+	else
+		clear_bit(DMF_MERGE_IS_OPTIONAL, &md->flags);
 	write_unlock_irqrestore(&md->map_lock, flags);
 
 	return old_map;
diff --git a/drivers/md/dm.h b/drivers/md/dm.h
index 1aaf16746da8..6745dbd278a4 100644
--- a/drivers/md/dm.h
+++ b/drivers/md/dm.h
@@ -66,6 +66,8 @@ int dm_table_alloc_md_mempools(struct dm_table *t);
 void dm_table_free_md_mempools(struct dm_table *t);
 struct dm_md_mempools *dm_table_get_md_mempools(struct dm_table *t);
 
+int dm_queue_merge_is_compulsory(struct request_queue *q);
+
 void dm_lock_md_type(struct mapped_device *md);
 void dm_unlock_md_type(struct mapped_device *md);
 void dm_set_md_type(struct mapped_device *md, unsigned type);
-- 
cgit v1.2.3


From a6e50b409d3f9e0833e69c3c9cca822e8fa4adbb Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 2 Aug 2011 12:32:04 +0100
Subject: dm snapshot: skip reading origin when overwriting complete chunk

If we write a full chunk in the snapshot, skip reading the origin device
because the whole chunk will be overwritten anyway.

This patch changes the snapshot write logic when a full chunk is written.
In this case:
  1. allocate the exception
  2. dispatch the bio (but don't report the bio completion to device mapper)
  3. write the exception record
  4. report bio completed

Callbacks must be done through the kcopyd thread, because callbacks must not
race with each other.  So we create two new functions:

  dm_kcopyd_prepare_callback: allocate a job structure and prepare the callback.
  (This function must not be called from interrupt context.)

  dm_kcopyd_do_callback: submit callback.
  (This function may be called from interrupt context.)

Performance test (on snapshots with 4k chunk size):
  without the patch:
    non-direct-io sequential write (dd):    17.7MB/s
    direct-io sequential write (dd):        20.9MB/s
    non-direct-io random write (mkfs.ext2): 0.44s

  with the patch:
    non-direct-io sequential write (dd):    26.5MB/s
    direct-io sequential write (dd):        33.2MB/s
    non-direct-io random write (mkfs.ext2): 0.27s

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-kcopyd.c    | 31 ++++++++++++++++++++++++
 drivers/md/dm-snap.c      | 60 ++++++++++++++++++++++++++++++++++++++++++++---
 include/linux/dm-kcopyd.h | 15 ++++++++++++
 3 files changed, 103 insertions(+), 3 deletions(-)

diff --git a/drivers/md/dm-kcopyd.c b/drivers/md/dm-kcopyd.c
index 98725e119324..f82147029636 100644
--- a/drivers/md/dm-kcopyd.c
+++ b/drivers/md/dm-kcopyd.c
@@ -617,6 +617,37 @@ int dm_kcopyd_copy(struct dm_kcopyd_client *kc, struct dm_io_region *from,
 }
 EXPORT_SYMBOL(dm_kcopyd_copy);
 
+void *dm_kcopyd_prepare_callback(struct dm_kcopyd_client *kc,
+				 dm_kcopyd_notify_fn fn, void *context)
+{
+	struct kcopyd_job *job;
+
+	job = mempool_alloc(kc->job_pool, GFP_NOIO);
+
+	memset(job, 0, sizeof(struct kcopyd_job));
+	job->kc = kc;
+	job->fn = fn;
+	job->context = context;
+
+	atomic_inc(&kc->nr_jobs);
+
+	return job;
+}
+EXPORT_SYMBOL(dm_kcopyd_prepare_callback);
+
+void dm_kcopyd_do_callback(void *j, int read_err, unsigned long write_err)
+{
+	struct kcopyd_job *job = j;
+	struct dm_kcopyd_client *kc = job->kc;
+
+	job->read_err = read_err;
+	job->write_err = write_err;
+
+	push(&kc->complete_jobs, job);
+	wake(kc);
+}
+EXPORT_SYMBOL(dm_kcopyd_do_callback);
+
 /*
  * Cancels a kcopyd job, eg. someone might be deactivating a
  * mirror.
diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index 94dee05dd28e..6f758870fc19 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -170,6 +170,13 @@ struct dm_snap_pending_exception {
 	 * kcopyd.
 	 */
 	int started;
+
+	/*
+	 * For writing a complete chunk, bypassing the copy.
+	 */
+	struct bio *full_bio;
+	bio_end_io_t *full_bio_end_io;
+	void *full_bio_private;
 };
 
 /*
@@ -1369,6 +1376,7 @@ static void pending_complete(struct dm_snap_pending_exception *pe, int success)
 	struct dm_snapshot *s = pe->snap;
 	struct bio *origin_bios = NULL;
 	struct bio *snapshot_bios = NULL;
+	struct bio *full_bio = NULL;
 	int error = 0;
 
 	if (!success) {
@@ -1408,6 +1416,11 @@ out:
 	dm_remove_exception(&pe->e);
 	snapshot_bios = bio_list_get(&pe->snapshot_bios);
 	origin_bios = bio_list_get(&pe->origin_bios);
+	full_bio = pe->full_bio;
+	if (full_bio) {
+		full_bio->bi_end_io = pe->full_bio_end_io;
+		full_bio->bi_private = pe->full_bio_private;
+	}
 	free_pending_exception(pe);
 
 	increment_pending_exceptions_done_count();
@@ -1415,10 +1428,15 @@ out:
 	up_write(&s->lock);
 
 	/* Submit any pending write bios */
-	if (error)
+	if (error) {
+		if (full_bio)
+			bio_io_error(full_bio);
 		error_bios(snapshot_bios);
-	else
+	} else {
+		if (full_bio)
+			bio_endio(full_bio, 0);
 		flush_bios(snapshot_bios);
+	}
 
 	retry_origin_bios(s, origin_bios);
 }
@@ -1472,6 +1490,32 @@ static void start_copy(struct dm_snap_pending_exception *pe)
 	dm_kcopyd_copy(s->kcopyd_client, &src, 1, &dest, 0, copy_callback, pe);
 }
 
+static void full_bio_end_io(struct bio *bio, int error)
+{
+	void *callback_data = bio->bi_private;
+
+	dm_kcopyd_do_callback(callback_data, 0, error ? 1 : 0);
+}
+
+static void start_full_bio(struct dm_snap_pending_exception *pe,
+			   struct bio *bio)
+{
+	struct dm_snapshot *s = pe->snap;
+	void *callback_data;
+
+	pe->full_bio = bio;
+	pe->full_bio_end_io = bio->bi_end_io;
+	pe->full_bio_private = bio->bi_private;
+
+	callback_data = dm_kcopyd_prepare_callback(s->kcopyd_client,
+						   copy_callback, pe);
+
+	bio->bi_end_io = full_bio_end_io;
+	bio->bi_private = callback_data;
+
+	generic_make_request(bio);
+}
+
 static struct dm_snap_pending_exception *
 __lookup_pending_exception(struct dm_snapshot *s, chunk_t chunk)
 {
@@ -1507,6 +1551,7 @@ __find_pending_exception(struct dm_snapshot *s,
 	bio_list_init(&pe->origin_bios);
 	bio_list_init(&pe->snapshot_bios);
 	pe->started = 0;
+	pe->full_bio = NULL;
 
 	if (s->store->type->prepare_exception(s->store, &pe->e)) {
 		free_pending_exception(pe);
@@ -1600,10 +1645,19 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio,
 		}
 
 		remap_exception(s, &pe->e, bio, chunk);
-		bio_list_add(&pe->snapshot_bios, bio);
 
 		r = DM_MAPIO_SUBMITTED;
 
+		if (!pe->started &&
+		    bio->bi_size == (s->store->chunk_size << SECTOR_SHIFT)) {
+			pe->started = 1;
+			up_write(&s->lock);
+			start_full_bio(pe, bio);
+			goto out;
+		}
+
+		bio_list_add(&pe->snapshot_bios, bio);
+
 		if (!pe->started) {
 			/* this is protected by snap->lock */
 			pe->started = 1;
diff --git a/include/linux/dm-kcopyd.h b/include/linux/dm-kcopyd.h
index 298d587e349b..5e54458e920f 100644
--- a/include/linux/dm-kcopyd.h
+++ b/include/linux/dm-kcopyd.h
@@ -42,5 +42,20 @@ int dm_kcopyd_copy(struct dm_kcopyd_client *kc, struct dm_io_region *from,
 		   unsigned num_dests, struct dm_io_region *dests,
 		   unsigned flags, dm_kcopyd_notify_fn fn, void *context);
 
+/*
+ * Prepare a callback and submit it via the kcopyd thread.
+ *
+ * dm_kcopyd_prepare_callback allocates a callback structure and returns it.
+ * It must not be called from interrupt context.
+ * The returned value should be passed into dm_kcopyd_do_callback.
+ *
+ * dm_kcopyd_do_callback submits the callback.
+ * It may be called from interrupt context.
+ * The callback is issued from the kcopyd thread.
+ */
+void *dm_kcopyd_prepare_callback(struct dm_kcopyd_client *kc,
+				 dm_kcopyd_notify_fn fn, void *context);
+void dm_kcopyd_do_callback(void *job, int read_err, unsigned long write_err);
+
 #endif	/* __KERNEL__ */
 #endif	/* _LINUX_DM_KCOPYD_H */
-- 
cgit v1.2.3


From 498f0103ea13123e007660def9072a0b7dd1c599 Mon Sep 17 00:00:00 2001
From: Mike Snitzer <snitzer@redhat.com>
Date: Tue, 2 Aug 2011 12:32:04 +0100
Subject: dm table: share target argument parsing functions

Move multipath target argument parsing code into dm-table so other
targets can share it.

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-crypt.c         |   7 +-
 drivers/md/dm-mpath.c         | 152 ++++++++++++------------------------------
 drivers/md/dm-table.c         |  57 ++++++++++++++++
 include/linux/device-mapper.h |  43 ++++++++++++
 4 files changed, 147 insertions(+), 112 deletions(-)

diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index f5406766ece3..b79e7472a9b3 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -30,7 +30,6 @@
 #include <linux/device-mapper.h>
 
 #define DM_MSG_PREFIX "crypt"
-#define MESG_STR(x) x, sizeof(x)
 
 /*
  * context holding the current state of a multi-part conversion
@@ -1770,12 +1769,12 @@ static int crypt_message(struct dm_target *ti, unsigned argc, char **argv)
 	if (argc < 2)
 		goto error;
 
-	if (!strnicmp(argv[0], MESG_STR("key"))) {
+	if (!strcasecmp(argv[0], "key")) {
 		if (!test_bit(DM_CRYPT_SUSPENDED, &cc->flags)) {
 			DMWARN("not suspended during key manipulation.");
 			return -EINVAL;
 		}
-		if (argc == 3 && !strnicmp(argv[1], MESG_STR("set"))) {
+		if (argc == 3 && !strcasecmp(argv[1], "set")) {
 			ret = crypt_set_key(cc, argv[2]);
 			if (ret)
 				return ret;
@@ -1783,7 +1782,7 @@ static int crypt_message(struct dm_target *ti, unsigned argc, char **argv)
 				ret = cc->iv_gen_ops->init(cc);
 			return ret;
 		}
-		if (argc == 2 && !strnicmp(argv[1], MESG_STR("wipe"))) {
+		if (argc == 2 && !strcasecmp(argv[1], "wipe")) {
 			if (cc->iv_gen_ops && cc->iv_gen_ops->wipe) {
 				ret = cc->iv_gen_ops->wipe(cc);
 				if (ret)
diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index adf851a081bd..5e0090ef4182 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -22,7 +22,6 @@
 #include <linux/atomic.h>
 
 #define DM_MSG_PREFIX "multipath"
-#define MESG_STR(x) x, sizeof(x)
 #define DM_PG_INIT_DELAY_MSECS 2000
 #define DM_PG_INIT_DELAY_DEFAULT ((unsigned) -1)
 
@@ -505,80 +504,29 @@ static void trigger_event(struct work_struct *work)
  *      <#paths> <#per-path selector args>
  *         [<path> [<arg>]* ]+ ]+
  *---------------------------------------------------------------*/
-struct param {
-	unsigned min;
-	unsigned max;
-	char *error;
-};
-
-static int read_param(struct param *param, char *str, unsigned *v, char **error)
-{
-	if (!str ||
-	    (sscanf(str, "%u", v) != 1) ||
-	    (*v < param->min) ||
-	    (*v > param->max)) {
-		*error = param->error;
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
-struct arg_set {
-	unsigned argc;
-	char **argv;
-};
-
-static char *shift(struct arg_set *as)
-{
-	char *r;
-
-	if (as->argc) {
-		as->argc--;
-		r = *as->argv;
-		as->argv++;
-		return r;
-	}
-
-	return NULL;
-}
-
-static void consume(struct arg_set *as, unsigned n)
-{
-	BUG_ON (as->argc < n);
-	as->argc -= n;
-	as->argv += n;
-}
-
-static int parse_path_selector(struct arg_set *as, struct priority_group *pg,
+static int parse_path_selector(struct dm_arg_set *as, struct priority_group *pg,
 			       struct dm_target *ti)
 {
 	int r;
 	struct path_selector_type *pst;
 	unsigned ps_argc;
 
-	static struct param _params[] = {
+	static struct dm_arg _args[] = {
 		{0, 1024, "invalid number of path selector args"},
 	};
 
-	pst = dm_get_path_selector(shift(as));
+	pst = dm_get_path_selector(dm_shift_arg(as));
 	if (!pst) {
 		ti->error = "unknown path selector type";
 		return -EINVAL;
 	}
 
-	r = read_param(_params, shift(as), &ps_argc, &ti->error);
+	r = dm_read_arg_group(_args, as, &ps_argc, &ti->error);
 	if (r) {
 		dm_put_path_selector(pst);
 		return -EINVAL;
 	}
 
-	if (ps_argc > as->argc) {
-		dm_put_path_selector(pst);
-		ti->error = "not enough arguments for path selector";
-		return -EINVAL;
-	}
-
 	r = pst->create(&pg->ps, ps_argc, as->argv);
 	if (r) {
 		dm_put_path_selector(pst);
@@ -587,12 +535,12 @@ static int parse_path_selector(struct arg_set *as, struct priority_group *pg,
 	}
 
 	pg->ps.type = pst;
-	consume(as, ps_argc);
+	dm_consume_args(as, ps_argc);
 
 	return 0;
 }
 
-static struct pgpath *parse_path(struct arg_set *as, struct path_selector *ps,
+static struct pgpath *parse_path(struct dm_arg_set *as, struct path_selector *ps,
 			       struct dm_target *ti)
 {
 	int r;
@@ -609,7 +557,7 @@ static struct pgpath *parse_path(struct arg_set *as, struct path_selector *ps,
 	if (!p)
 		return ERR_PTR(-ENOMEM);
 
-	r = dm_get_device(ti, shift(as), dm_table_get_mode(ti->table),
+	r = dm_get_device(ti, dm_shift_arg(as), dm_table_get_mode(ti->table),
 			  &p->path.dev);
 	if (r) {
 		ti->error = "error getting device";
@@ -660,16 +608,16 @@ static struct pgpath *parse_path(struct arg_set *as, struct path_selector *ps,
 	return ERR_PTR(r);
 }
 
-static struct priority_group *parse_priority_group(struct arg_set *as,
+static struct priority_group *parse_priority_group(struct dm_arg_set *as,
 						   struct multipath *m)
 {
-	static struct param _params[] = {
+	static struct dm_arg _args[] = {
 		{1, 1024, "invalid number of paths"},
 		{0, 1024, "invalid number of selector args"}
 	};
 
 	int r;
-	unsigned i, nr_selector_args, nr_params;
+	unsigned i, nr_selector_args, nr_args;
 	struct priority_group *pg;
 	struct dm_target *ti = m->ti;
 
@@ -693,26 +641,26 @@ static struct priority_group *parse_priority_group(struct arg_set *as,
 	/*
 	 * read the paths
 	 */
-	r = read_param(_params, shift(as), &pg->nr_pgpaths, &ti->error);
+	r = dm_read_arg(_args, as, &pg->nr_pgpaths, &ti->error);
 	if (r)
 		goto bad;
 
-	r = read_param(_params + 1, shift(as), &nr_selector_args, &ti->error);
+	r = dm_read_arg(_args + 1, as, &nr_selector_args, &ti->error);
 	if (r)
 		goto bad;
 
-	nr_params = 1 + nr_selector_args;
+	nr_args = 1 + nr_selector_args;
 	for (i = 0; i < pg->nr_pgpaths; i++) {
 		struct pgpath *pgpath;
-		struct arg_set path_args;
+		struct dm_arg_set path_args;
 
-		if (as->argc < nr_params) {
+		if (as->argc < nr_args) {
 			ti->error = "not enough path parameters";
 			r = -EINVAL;
 			goto bad;
 		}
 
-		path_args.argc = nr_params;
+		path_args.argc = nr_args;
 		path_args.argv = as->argv;
 
 		pgpath = parse_path(&path_args, &pg->ps, ti);
@@ -723,7 +671,7 @@ static struct priority_group *parse_priority_group(struct arg_set *as,
 
 		pgpath->pg = pg;
 		list_add_tail(&pgpath->list, &pg->pgpaths);
-		consume(as, nr_params);
+		dm_consume_args(as, nr_args);
 	}
 
 	return pg;
@@ -733,28 +681,23 @@ static struct priority_group *parse_priority_group(struct arg_set *as,
 	return ERR_PTR(r);
 }
 
-static int parse_hw_handler(struct arg_set *as, struct multipath *m)
+static int parse_hw_handler(struct dm_arg_set *as, struct multipath *m)
 {
 	unsigned hw_argc;
 	int ret;
 	struct dm_target *ti = m->ti;
 
-	static struct param _params[] = {
+	static struct dm_arg _args[] = {
 		{0, 1024, "invalid number of hardware handler args"},
 	};
 
-	if (read_param(_params, shift(as), &hw_argc, &ti->error))
+	if (dm_read_arg_group(_args, as, &hw_argc, &ti->error))
 		return -EINVAL;
 
 	if (!hw_argc)
 		return 0;
 
-	if (hw_argc > as->argc) {
-		ti->error = "not enough arguments for hardware handler";
-		return -EINVAL;
-	}
-
-	m->hw_handler_name = kstrdup(shift(as), GFP_KERNEL);
+	m->hw_handler_name = kstrdup(dm_shift_arg(as), GFP_KERNEL);
 	request_module("scsi_dh_%s", m->hw_handler_name);
 	if (scsi_dh_handler_exist(m->hw_handler_name) == 0) {
 		ti->error = "unknown hardware handler type";
@@ -778,7 +721,7 @@ static int parse_hw_handler(struct arg_set *as, struct multipath *m)
 		for (i = 0, p+=j+1; i <= hw_argc - 2; i++, p+=j+1)
 			j = sprintf(p, "%s", as->argv[i]);
 	}
-	consume(as, hw_argc - 1);
+	dm_consume_args(as, hw_argc - 1);
 
 	return 0;
 fail:
@@ -787,52 +730,45 @@ fail:
 	return ret;
 }
 
-static int parse_features(struct arg_set *as, struct multipath *m)
+static int parse_features(struct dm_arg_set *as, struct multipath *m)
 {
 	int r;
 	unsigned argc;
 	struct dm_target *ti = m->ti;
-	const char *param_name;
+	const char *arg_name;
 
-	static struct param _params[] = {
+	static struct dm_arg _args[] = {
 		{0, 5, "invalid number of feature args"},
 		{1, 50, "pg_init_retries must be between 1 and 50"},
 		{0, 60000, "pg_init_delay_msecs must be between 0 and 60000"},
 	};
 
-	r = read_param(_params, shift(as), &argc, &ti->error);
+	r = dm_read_arg_group(_args, as, &argc, &ti->error);
 	if (r)
 		return -EINVAL;
 
 	if (!argc)
 		return 0;
 
-	if (argc > as->argc) {
-		ti->error = "not enough arguments for features";
-		return -EINVAL;
-	}
-
 	do {
-		param_name = shift(as);
+		arg_name = dm_shift_arg(as);
 		argc--;
 
-		if (!strnicmp(param_name, MESG_STR("queue_if_no_path"))) {
+		if (!strcasecmp(arg_name, "queue_if_no_path")) {
 			r = queue_if_no_path(m, 1, 0);
 			continue;
 		}
 
-		if (!strnicmp(param_name, MESG_STR("pg_init_retries")) &&
+		if (!strcasecmp(arg_name, "pg_init_retries") &&
 		    (argc >= 1)) {
-			r = read_param(_params + 1, shift(as),
-				       &m->pg_init_retries, &ti->error);
+			r = dm_read_arg(_args + 1, as, &m->pg_init_retries, &ti->error);
 			argc--;
 			continue;
 		}
 
-		if (!strnicmp(param_name, MESG_STR("pg_init_delay_msecs")) &&
+		if (!strcasecmp(arg_name, "pg_init_delay_msecs") &&
 		    (argc >= 1)) {
-			r = read_param(_params + 2, shift(as),
-				       &m->pg_init_delay_msecs, &ti->error);
+			r = dm_read_arg(_args + 2, as, &m->pg_init_delay_msecs, &ti->error);
 			argc--;
 			continue;
 		}
@@ -847,15 +783,15 @@ static int parse_features(struct arg_set *as, struct multipath *m)
 static int multipath_ctr(struct dm_target *ti, unsigned int argc,
 			 char **argv)
 {
-	/* target parameters */
-	static struct param _params[] = {
+	/* target arguments */
+	static struct dm_arg _args[] = {
 		{0, 1024, "invalid number of priority groups"},
 		{0, 1024, "invalid initial priority group number"},
 	};
 
 	int r;
 	struct multipath *m;
-	struct arg_set as;
+	struct dm_arg_set as;
 	unsigned pg_count = 0;
 	unsigned next_pg_num;
 
@@ -876,11 +812,11 @@ static int multipath_ctr(struct dm_target *ti, unsigned int argc,
 	if (r)
 		goto bad;
 
-	r = read_param(_params, shift(&as), &m->nr_priority_groups, &ti->error);
+	r = dm_read_arg(_args, &as, &m->nr_priority_groups, &ti->error);
 	if (r)
 		goto bad;
 
-	r = read_param(_params + 1, shift(&as), &next_pg_num, &ti->error);
+	r = dm_read_arg(_args + 1, &as, &next_pg_num, &ti->error);
 	if (r)
 		goto bad;
 
@@ -1510,10 +1446,10 @@ static int mu