1 files changed, 456 insertions, 176 deletions

diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index 4b54618b415..51521429fb5 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -36,6 +36,14 @@ struct hash_cell {
 	struct dm_table *new_map;
 };
 
+/*
+ * A dummy definition to make RCU happy.
+ * struct dm_table should never be dereferenced in this file.
+ */
+struct dm_table {
+	int undefined__;
+};
+
 struct vers_iter {
     size_t param_size;
     struct dm_target_versions *vers, *old_vers;
@@ -49,7 +57,7 @@ struct vers_iter {
 static struct list_head _name_buckets[NUM_BUCKETS];
 static struct list_head _uuid_buckets[NUM_BUCKETS];
 
-static void dm_hash_remove_all(int keep_open_devices);
+static void dm_hash_remove_all(bool keep_open_devices, bool mark_deferred, bool only_deferred);
 
 /*
  * Guards access to both hash tables.
@@ -78,7 +86,7 @@ static int dm_hash_init(void)
 
 static void dm_hash_exit(void)
 {
-	dm_hash_remove_all(0);
+	dm_hash_remove_all(false, false, false);
 }
 
 /*-----------------------------------------------------------------
@@ -128,6 +136,24 @@ static struct hash_cell *__get_uuid_cell(const char *str)
 	return NULL;
 }
 
+static struct hash_cell *__get_dev_cell(uint64_t dev)
+{
+	struct mapped_device *md;
+	struct hash_cell *hc;
+
+	md = dm_get_md(huge_decode_dev(dev));
+	if (!md)
+		return NULL;
+
+	hc = dm_get_mdptr(md);
+	if (!hc) {
+		dm_put(md);
+		return NULL;
+	}
+
+	return hc;
+}
+
 /*-----------------------------------------------------------------
  * Inserting, removing and renaming a device.
  *---------------------------------------------------------------*/
@@ -224,9 +250,10 @@ static int dm_hash_insert(const char *name, const char *uuid, struct mapped_devi
 	return -EBUSY;
 }
 
-static void __hash_remove(struct hash_cell *hc)
+static struct dm_table *__hash_remove(struct hash_cell *hc)
 {
 	struct dm_table *table;
+	int srcu_idx;
 
 	/* remove from the dev hash */
 	list_del(&hc->uuid_list);
@@ -235,23 +262,26 @@ static void __hash_remove(struct hash_cell *hc)
 	dm_set_mdptr(hc->md, NULL);
 	mutex_unlock(&dm_hash_cells_mutex);
 
-	table = dm_get_live_table(hc->md);
-	if (table) {
+	table = dm_get_live_table(hc->md, &srcu_idx);
+	if (table)
 		dm_table_event(table);
-		dm_table_put(table);
-	}
+	dm_put_live_table(hc->md, srcu_idx);
 
+	table = NULL;
 	if (hc->new_map)
-		dm_table_destroy(hc->new_map);
+		table = hc->new_map;
 	dm_put(hc->md);
 	free_cell(hc);
+
+	return table;
 }
 
-static void dm_hash_remove_all(int keep_open_devices)
+static void dm_hash_remove_all(bool keep_open_devices, bool mark_deferred, bool only_deferred)
 {
 	int i, dev_skipped;
 	struct hash_cell *hc;
 	struct mapped_device *md;
+	struct dm_table *t;
 
 retry:
 	dev_skipped = 0;
@@ -263,16 +293,21 @@ retry:
 			md = hc->md;
 			dm_get(md);
 
-			if (keep_open_devices && dm_lock_for_deletion(md)) {
+			if (keep_open_devices &&
+			    dm_lock_for_deletion(md, mark_deferred, only_deferred)) {
 				dm_put(md);
 				dev_skipped++;
 				continue;
 			}
 
-			__hash_remove(hc);
+			t = __hash_remove(hc);
 
 			up_write(&_hash_lock);
 
+			if (t) {
+				dm_sync_table(md);
+				dm_table_destroy(t);
+			}
 			dm_put(md);
 			if (likely(keep_open_devices))
 				dm_destroy(md);
@@ -295,19 +330,56 @@ retry:
 		DMWARN("remove_all left %d open device(s)", dev_skipped);
 }
 
+/*
+ * Set the uuid of a hash_cell that isn't already set.
+ */
+static void __set_cell_uuid(struct hash_cell *hc, char *new_uuid)
+{
+	mutex_lock(&dm_hash_cells_mutex);
+	hc->uuid = new_uuid;
+	mutex_unlock(&dm_hash_cells_mutex);
+
+	list_add(&hc->uuid_list, _uuid_buckets + hash_str(new_uuid));
+}
+
+/*
+ * Changes the name of a hash_cell and returns the old name for
+ * the caller to free.
+ */
+static char *__change_cell_name(struct hash_cell *hc, char *new_name)
+{
+	char *old_name;
+
+	/*
+	 * Rename and move the name cell.
+	 */
+	list_del(&hc->name_list);
+	old_name = hc->name;
+
+	mutex_lock(&dm_hash_cells_mutex);
+	hc->name = new_name;
+	mutex_unlock(&dm_hash_cells_mutex);
+
+	list_add(&hc->name_list, _name_buckets + hash_str(new_name));
+
+	return old_name;
+}
+
 static struct mapped_device *dm_hash_rename(struct dm_ioctl *param,
 					    const char *new)
 {
-	char *new_name, *old_name;
+	char *new_data, *old_name = NULL;
 	struct hash_cell *hc;
 	struct dm_table *table;
 	struct mapped_device *md;
+	unsigned change_uuid = (param->flags & DM_UUID_FLAG) ? 1 : 0;
+	int srcu_idx;
 
 	/*
 	 * duplicate new.
 	 */
-	new_name = kstrdup(new, GFP_KERNEL);
-	if (!new_name)
+	new_data = kstrdup(new, GFP_KERNEL);
+	if (!new_data)
 		return ERR_PTR(-ENOMEM);
 
 	down_write(&_hash_lock);
@@ -315,13 +387,19 @@ static struct mapped_device *dm_hash_rename(struct dm_ioctl *param,
 	/*
 	 * Is new free ?
 	 */
-	hc = __get_name_cell(new);
+	if (change_uuid)
+		hc = __get_uuid_cell(new);
+	else
+		hc = __get_name_cell(new);
+
 	if (hc) {
-		DMWARN("asked to rename to an already-existing name %s -> %s",
+		DMWARN("Unable to change %s on mapped device %s to one that "
+		       "already exists: %s",
+		       change_uuid ? "uuid" : "name",
 		       param->name, new);
 		dm_put(hc->md);
 		up_write(&_hash_lock);
-		kfree(new_name);
+		kfree(new_data);
 		return ERR_PTR(-EBUSY);
 	}
 
@@ -330,31 +408,38 @@ static struct mapped_device *dm_hash_rename(struct dm_ioctl *param,
 	 */
 	hc = __get_name_cell(param->name);
 	if (!hc) {
-		DMWARN("asked to rename a non-existent device %s -> %s",
-		       param->name, new);
+		DMWARN("Unable to rename non-existent device, %s to %s%s",
+		       param->name, change_uuid ? "uuid " : "", new);
 		up_write(&_hash_lock);
-		kfree(new_name);
+		kfree(new_data);
 		return ERR_PTR(-ENXIO);
 	}
 
 	/*
-	 * rename and move the name cell.
+	 * Does this device already have a uuid?
 	 */
-	list_del(&hc->name_list);
-	old_name = hc->name;
-	mutex_lock(&dm_hash_cells_mutex);
-	hc->name = new_name;
-	mutex_unlock(&dm_hash_cells_mutex);
-	list_add(&hc->name_list, _name_buckets + hash_str(new_name));
+	if (change_uuid && hc->uuid) {
+		DMWARN("Unable to change uuid of mapped device %s to %s "
+		       "because uuid is already set to %s",
+		       param->name, new, hc->uuid);
+		dm_put(hc->md);
+		up_write(&_hash_lock);
+		kfree(new_data);
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (change_uuid)
+		__set_cell_uuid(hc, new_data);
+	else
+		old_name = __change_cell_name(hc, new_data);
 
 	/*
 	 * Wake up any dm event waiters.
 	 */
-	table = dm_get_live_table(hc->md);
-	if (table) {
+	table = dm_get_live_table(hc->md, &srcu_idx);
+	if (table)
 		dm_table_event(table);
-		dm_table_put(table);
-	}
+	dm_put_live_table(hc->md, srcu_idx);
 
 	if (!dm_kobject_uevent(hc->md, KOBJ_CHANGE, param->event_nr))
 		param->flags |= DM_UEVENT_GENERATED_FLAG;
@@ -366,6 +451,11 @@ static struct mapped_device *dm_hash_rename(struct dm_ioctl *param,
 	return md;
 }
 
+void dm_deferred_remove(void)
+{
+	dm_hash_remove_all(true, false, true);
+}
+
 /*-----------------------------------------------------------------
  * Implementation of the ioctl commands
  *---------------------------------------------------------------*/
@@ -377,7 +467,7 @@ typedef int (*ioctl_fn)(struct dm_ioctl *param, size_t param_size);
 
 static int remove_all(struct dm_ioctl *param, size_t param_size)
 {
-	dm_hash_remove_all(1);
+	dm_hash_remove_all(true, !!(param->flags & DM_DEFERRED_REMOVE), false);
 	param->data_size = 0;
 	return 0;
 }
@@ -552,11 +642,14 @@ static int check_name(const char *name)
  * _hash_lock without first calling dm_table_put, because dm_table_destroy
  * waits for this dm_table_put and could be called under this lock.
  */
-static struct dm_table *dm_get_inactive_table(struct mapped_device *md)
+static struct dm_table *dm_get_inactive_table(struct mapped_device *md, int *srcu_idx)
 {
 	struct hash_cell *hc;
 	struct dm_table *table = NULL;
 
+	/* increment rcu count, we don't care about the table pointer */
+	dm_get_live_table(md, srcu_idx);
+
 	down_read(&_hash_lock);
 	hc = dm_get_mdptr(md);
 	if (!hc || hc->md != md) {
@@ -565,8 +658,6 @@ static struct dm_table *dm_get_inactive_table(struct mapped_device *md)
 	}
 
 	table = hc->new_map;
-	if (table)
-		dm_table_get(table);
 
 out:
 	up_read(&_hash_lock);
@@ -575,10 +666,11 @@ out:
 }
 
 static struct dm_table *dm_get_live_or_inactive_table(struct mapped_device *md,
-						      struct dm_ioctl *param)
+						      struct dm_ioctl *param,
+						      int *srcu_idx)
 {
 	return (param->flags & DM_QUERY_INACTIVE_TABLE_FLAG) ?
-		dm_get_inactive_table(md) : dm_get_live_table(md);
+		dm_get_inactive_table(md, srcu_idx) : dm_get_live_table(md, srcu_idx);
 }
 
 /*
@@ -589,6 +681,7 @@ static void __dev_status(struct mapped_device *md, struct dm_ioctl *param)
 {
 	struct gendisk *disk = dm_disk(md);
 	struct dm_table *table;
+	int srcu_idx;
 
 	param->flags &= ~(DM_SUSPEND_FLAG | DM_READONLY_FLAG |
 			  DM_ACTIVE_PRESENT_FLAG);
@@ -596,6 +689,9 @@ static void __dev_status(struct mapped_device *md, struct dm_ioctl *param)
 	if (dm_suspended_md(md))
 		param->flags |= DM_SUSPEND_FLAG;
 
+	if (dm_test_deferred_remove_flag(md))
+		param->flags |= DM_DEFERRED_REMOVE;
+
 	param->dev = huge_encode_dev(disk_devt(disk));
 
 	/*
@@ -608,26 +704,27 @@ static void __dev_status(struct mapped_device *md, struct dm_ioctl *param)
 	param->event_nr = dm_get_event_nr(md);
 	param->target_count = 0;
 
-	table = dm_get_live_table(md);
+	table = dm_get_live_table(md, &srcu_idx);
 	if (table) {
 		if (!(param->flags & DM_QUERY_INACTIVE_TABLE_FLAG)) {
 			if (get_disk_ro(disk))
 				param->flags |= DM_READONLY_FLAG;
 			param->target_count = dm_table_get_num_targets(table);
 		}
-		dm_table_put(table);
 
 		param->flags |= DM_ACTIVE_PRESENT_FLAG;
 	}
+	dm_put_live_table(md, srcu_idx);
 
 	if (param->flags & DM_QUERY_INACTIVE_TABLE_FLAG) {
-		table = dm_get_inactive_table(md);
+		int srcu_idx;
+		table = dm_get_inactive_table(md, &srcu_idx);
 		if (table) {
 			if (!(dm_table_get_mode(table) & FMODE_WRITE))
 				param->flags |= DM_READONLY_FLAG;
 			param->target_count = dm_table_get_num_targets(table);
-			dm_table_put(table);
 		}
+		dm_put_live_table(md, srcu_idx);
 	}
 }
 
@@ -668,25 +765,45 @@ static int dev_create(struct dm_ioctl *param, size_t param_size)
  */
 static struct hash_cell *__find_device_hash_cell(struct dm_ioctl *param)
 {
-	struct mapped_device *md;
-	void *mdptr = NULL;
+	struct hash_cell *hc = NULL;
 
-	if (*param->uuid)
-		return __get_uuid_cell(param->uuid);
+	if (*param->uuid) {
+		if (*param->name || param->dev)
+			return NULL;
 
-	if (*param->name)
-		return __get_name_cell(param->name);
+		hc = __get_uuid_cell(param->uuid);
+		if (!hc)
+			return NULL;
+	} else if (*param->name) {
+		if (param->dev)
+			return NULL;
 
-	md = dm_get_md(huge_decode_dev(param->dev));
-	if (!md)
-		goto out;
+		hc = __get_name_cell(param->name);
+		if (!hc)
+			return NULL;
+	} else if (param->dev) {
+		hc = __get_dev_cell(param->dev);
+		if (!hc)
+			return NULL;
+	} else
+		return NULL;
 
-	mdptr = dm_get_mdptr(md);
-	if (!mdptr)
-		dm_put(md);
+	/*
+	 * Sneakily write in both the name and the uuid
+	 * while we have the cell.
+	 */
+	strlcpy(param->name, hc->name, sizeof(param->name));
+	if (hc->uuid)
+		strlcpy(param->uuid, hc->uuid, sizeof(param->uuid));
+	else
+		param->uuid[0] = '\0';
 
-out:
-	return mdptr;
+	if (hc->new_map)
+		param->flags |= DM_INACTIVE_PRESENT_FLAG;
+	else
+		param->flags &= ~DM_INACTIVE_PRESENT_FLAG;
+
+	return hc;
 }
 
 static struct mapped_device *find_device(struct dm_ioctl *param)
@@ -696,24 +813,8 @@ static struct mapped_device *find_device(struct dm_ioctl *param)
 
 	down_read(&_hash_lock);
 	hc = __find_device_hash_cell(param);
-	if (hc) {
+	if (hc)
 		md = hc->md;
-
-		/*
-		 * Sneakily write in both the name and the uuid
-		 * while we have the cell.
-		 */
-		strlcpy(param->name, hc->name, sizeof(param->name));
-		if (hc->uuid)
-			strlcpy(param->uuid, hc->uuid, sizeof(param->uuid));
-		else
-			param->uuid[0] = '\0';
-
-		if (hc->new_map)
-			param->flags |= DM_INACTIVE_PRESENT_FLAG;
-		else
-			param->flags &= ~DM_INACTIVE_PRESENT_FLAG;
-	}
 	up_read(&_hash_lock);
 
 	return md;
@@ -724,12 +825,13 @@ static int dev_remove(struct dm_ioctl *param, size_t param_size)
 	struct hash_cell *hc;
 	struct mapped_device *md;
 	int r;
+	struct dm_table *t;
 
 	down_write(&_hash_lock);
 	hc = __find_device_hash_cell(param);
 
 	if (!hc) {
-		DMWARN("device doesn't appear to be in the dev hash table.");
+		DMDEBUG_LIMIT("device doesn't appear to be in the dev hash table.");
 		up_write(&_hash_lock);
 		return -ENXIO;
 	}
@@ -739,17 +841,29 @@ static int dev_remove(struct dm_ioctl *param, size_t param_size)
 	/*
 	 * Ensure the device is not open and nothing further can open it.
 	 */
-	r = dm_lock_for_deletion(md);
+	r = dm_lock_for_deletion(md, !!(param->flags & DM_DEFERRED_REMOVE), false);
 	if (r) {
-		DMWARN("unable to remove open device %s", hc->name);
+		if (r == -EBUSY && param->flags & DM_DEFERRED_REMOVE) {
+			up_write(&_hash_lock);
+			dm_put(md);
+			return 0;
+		}
+		DMDEBUG_LIMIT("unable to remove open device %s", hc->name);
 		up_write(&_hash_lock);
 		dm_put(md);
 		return r;
 	}
 
-	__hash_remove(hc);
+	t = __hash_remove(hc);
 	up_write(&_hash_lock);
 
+	if (t) {
+		dm_sync_table(md);
+		dm_table_destroy(t);
+	}
+
+	param->flags &= ~DM_DEFERRED_REMOVE;
+
 	if (!dm_kobject_uevent(md, KOBJ_REMOVE, param->event_nr))
 		param->flags |= DM_UEVENT_GENERATED_FLAG;
 
@@ -774,21 +888,24 @@ static int invalid_str(char *str, void *end)
 static int dev_rename(struct dm_ioctl *param, size_t param_size)
 {
 	int r;
-	char *new_name = (char *) param + param->data_start;
+	char *new_data = (char *) param + param->data_start;
 	struct mapped_device *md;
+	unsigned change_uuid = (param->flags & DM_UUID_FLAG) ? 1 : 0;
 
-	if (new_name < param->data ||
-	    invalid_str(new_name, (void *) param + param_size) ||
-	    strlen(new_name) > DM_NAME_LEN - 1) {
-		DMWARN("Invalid new logical volume name supplied.");
+	if (new_data < param->data ||
+	    invalid_str(new_data, (void *) param + param_size) || !*new_data ||
+	    strlen(new_data) > (change_uuid ? DM_UUID_LEN - 1 : DM_NAME_LEN - 1)) {
+		DMWARN("Invalid new mapped device name or uuid string supplied.");
 		return -EINVAL;
 	}
 
-	r = check_name(new_name);
-	if (r)
-		return r;
+	if (!change_uuid) {
+		r = check_name(new_data);
+		if (r)
+			return r;
+	}
 
-	md = dm_hash_rename(param, new_name);
+	md = dm_hash_rename(param, new_data);
 	if (IS_ERR(md))
 		return PTR_ERR(md);
 
@@ -805,6 +922,7 @@ static int dev_set_geometry(struct dm_ioctl *param, size_t param_size)
 	struct hd_geometry geometry;
 	unsigned long indata[4];
 	char *geostr = (char *) param + param->data_start;
+	char dummy;
 
 	md = find_device(param);
 	if (!md)
@@ -816,8 +934,8 @@ static int dev_set_geometry(struct dm_ioctl *param, size_t param_size)
 		goto out;
 	}
 
-	x = sscanf(geostr, "%lu %lu %lu %lu", indata,
-		   indata + 1, indata + 2, indata + 3);
+	x = sscanf(geostr, "%lu %lu %lu %lu%c", indata,
+		   indata + 1, indata + 2, indata + 3, &dummy);
 
 	if (x != 4) {
 		DMWARN("Unable to interpret geometry settings.");
@@ -885,7 +1003,7 @@ static int do_resume(struct dm_ioctl *param)
 
 	hc = __find_device_hash_cell(param);
 	if (!hc) {
-		DMWARN("device doesn't appear to be in the dev hash table.");
+		DMDEBUG_LIMIT("device doesn't appear to be in the dev hash table.");
 		up_write(&_hash_lock);
 		return -ENXIO;
 	}
@@ -910,6 +1028,7 @@ static int do_resume(struct dm_ioctl *param)
 
 		old_map = dm_swap_table(md, new_map);
 		if (IS_ERR(old_map)) {
+			dm_sync_table(md);
 			dm_table_destroy(new_map);
 			dm_put(md);
 			return PTR_ERR(old_map);
@@ -927,6 +1046,10 @@ static int do_resume(struct dm_ioctl *param)
 			param->flags |= DM_UEVENT_GENERATED_FLAG;
 	}
 
+	/*
+	 * Since dm_swap_table synchronizes RCU, nobody should be in
+	 * read-side critical section already.
+	 */
 	if (old_map)
 		dm_table_destroy(old_map);
 
@@ -978,6 +1101,7 @@ 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);
 
@@ -990,6 +1114,7 @@ static void retrieve_status(struct dm_table *table,
 	num_targets = dm_table_get_num_targets(table);
 	for (i = 0; i < num_targets; i++) {
 		struct dm_target *ti = dm_table_get_target(table, i);
+		size_t l;
 
 		remaining = len - (outptr - outbuf);
 		if (remaining <= sizeof(struct dm_target_spec)) {
@@ -1014,14 +1139,19 @@ 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)) {
-				param->flags |= DM_BUFFER_FULL_FLAG;
-				break;
-			}
+			if (param->flags & DM_NOFLUSH_FLAG)
+				status_flags |= DM_STATUS_NOFLUSH_FLAG;
+			ti->type->status(ti, type, status_flags, outptr, remaining);
 		} else
 			outptr[0] = '\0';
 
-		outptr += strlen(outptr) + 1;
+		l = strlen(outptr) + 1;
+		if (l == remaining) {
+			param->flags |= DM_BUFFER_FULL_FLAG;
+			break;
+		}
+
+		outptr += l;
 		used = param->data_start + (outptr - outbuf);
 
 		outptr = align_ptr(outptr);
@@ -1042,6 +1172,7 @@ static int dev_wait(struct dm_ioctl *param, size_t param_size)
 	int r = 0;
 	struct mapped_device *md;
 	struct dm_table *table;
+	int srcu_idx;
 
 	md = find_device(param);
 	if (!md)
@@ -1062,11 +1193,10 @@ static int dev_wait(struct dm_ioctl *param, size_t param_size)
 	 */
 	__dev_status(md, param);
 
-	table = dm_get_live_or_inactive_table(md, param);
-	if (table) {
+	table = dm_get_live_or_inactive_table(md, param, &srcu_idx);
+	if (table)
 		retrieve_status(table, param, param_size);
-		dm_table_put(table);
-	}
+	dm_put_live_table(md, srcu_idx);
 
 out:
 	dm_put(md);
@@ -1138,8 +1268,9 @@ static int table_load(struct dm_ioctl *param, size_t param_size)
 {
 	int r;
 	struct hash_cell *hc;
-	struct dm_table *t;
+	struct dm_table *t, *old_map = NULL;
 	struct mapped_device *md;
+	struct target_type *immutable_target_type;
 
 	md = find_device(param);
 	if (!md)
@@ -1147,34 +1278,37 @@ static int table_load(struct dm_ioctl *param, size_t param_size)
 
 	r = dm_table_create(&t, get_mode(param), param->target_count, md);
 	if (r)
-		goto out;
+		goto err;
 
+	/* Protect md->type and md->queue against concurrent table loads. */
+	dm_lock_md_type(md);
 	r = populate_table(t, param, param_size);
-	if (r) {
-		dm_table_destroy(t);
-		goto out;
+	if (r)
+		goto err_unlock_md_type;
+
+	immutable_target_type = dm_get_immutable_target_type(md);
+	if (immutable_target_type &&
+	    (immutable_target_type != dm_table_get_immutable_target_type(t))) {
+		DMWARN("can't replace immutable target type %s",
+		       immutable_target_type->name);
+		r = -EINVAL;
+		goto err_unlock_md_type;
 	}
 
-	/* Protect md->type and md->queue against concurrent table loads. */
-	dm_lock_md_type(md);
 	if (dm_get_md_type(md) == DM_TYPE_NONE)
 		/* Initial table load: acquire type of table. */
 		dm_set_md_type(md, dm_table_get_type(t));
 	else if (dm_get_md_type(md) != dm_table_get_type(t)) {
 		DMWARN("can't change device type after initial table load.");
-		dm_table_destroy(t);
-		dm_unlock_md_type(md);
 		r = -EINVAL;
-		goto out;
+		goto err_unlock_md_type;
 	}
 
 	/* setup md->queue to reflect md's type (may block) */
 	r = dm_setup_md_queue(md);
 	if (r) {
 		DMWARN("unable to set up device queue for new table.");
-		dm_table_destroy(t);
-		dm_unlock_md_type(md);
-		goto out;
+		goto err_unlock_md_type;
 	}
 	dm_unlock_md_type(md);
 
@@ -1183,21 +1317,33 @@ static int table_load(struct dm_ioctl *param, size_t param_size)
 	hc = dm_get_mdptr(md);
 	if (!hc || hc->md != md) {
 		DMWARN("device has been removed from the dev hash table.");
-		dm_table_destroy(t);
 		up_write(&_hash_lock);
 		r = -ENXIO;
-		goto out;
+		goto err_destroy_table;
 	}
 
 	if (hc->new_map)
-		dm_table_destroy(hc->new_map);
+		old_map = hc->new_map;
 	hc->new_map = t;
 	up_write(&_hash_lock);
 
 	param->flags |= DM_INACTIVE_PRESENT_FLAG;
 	__dev_status(md, param);
 
-out:
+	if (old_map) {
+		dm_sync_table(md);
+		dm_table_destroy(old_map);
+	}
+
+	dm_put(md);
+
+	return 0;
+
+err_unlock_md_type:
+	dm_unlock_md_type(md);
+err_destroy_table:
+	dm_table_destroy(t);
+err:
 	dm_put(md);
 
 	return r;
@@ -1207,18 +1353,19 @@ static int table_clear(struct dm_ioctl *param, size_t param_size)
 {
 	struct hash_cell *hc;
 	struct mapped_device *md;
+	struct dm_table *old_map = NULL;
 
 	down_write(&_hash_lock);
 
 	hc = __find_device_hash_cell(param);
 	if (!hc) {
-		DMWARN("device doesn't appear to be in the dev hash table.");
+		DMDEBUG_LIMIT("device doesn't appear to be in the dev hash table.");
 		up_write(&_hash_lock);
 		return -ENXIO;
 	}
 
 	if (hc->new_map) {
-		dm_table_destroy(hc->new_map);
+		old_map = hc->new_map;
 		hc->new_map = NULL;
 	}
 
@@ -1227,6 +1374,10 @@ static int table_clear(struct dm_ioctl *param, size_t param_size)
 	__dev_status(hc->md, param);
 	md = hc->md;
 	up_write(&_hash_lock);
+	if (old_map) {
+		dm_sync_table(md);
+		dm_table_destroy(old_map);
+	}
 	dm_put(md);
 
 	return 0;
@@ -1276,6 +1427,7 @@ static int table_deps(struct dm_ioctl *param, size_t param_size)
 {
 	struct mapped_device *md;
 	struct dm_table *table;
+	int srcu_idx;
 
 	md = find_device(param);
 	if (!md)
@@ -1283,11 +1435,10 @@ static int table_deps(struct dm_ioctl *param, size_t param_size)
 
 	__dev_status(md, param);
 
-	table = dm_get_live_or_inactive_table(md, param);
-	if (table) {
+	table = dm_get_live_or_inactive_table(md, param, &srcu_idx);
+	if (table)
 		retrieve_deps(table, param, param_size);
-		dm_table_put(table);
-	}
+	dm_put_live_table(md, srcu_idx);
 
 	dm_put(md);
 
@@ -1302,6 +1453,7 @@ static int table_status(struct dm_ioctl *param, size_t param_size)
 {
 	struct mapped_device *md;
 	struct dm_table *table;
+	int srcu_idx;
 
 	md = find_device(param);
 	if (!md)
@@ -1309,11 +1461,10 @@ static int table_status(struct dm_ioctl *param, size_t param_size)
 
 	__dev_status(md, param);
 
-	table = dm_get_live_or_inactive_table(md, param);
-	if (table) {
+	table = dm_get_live_or_inactive_table(md, param, &srcu_idx);
+	if (table)
 		retrieve_status(table, param, param_size);
-		dm_table_put(table);
-	}
+	dm_put_live_table(md, srcu_idx);
 
 	dm_put(md);
 
@@ -1321,6 +1472,36 @@ static int table_status(struct dm_ioctl *param, size_t param_size)
 }
 
 /*
+ * Process device-mapper dependent messages.  Messages prefixed with '@'
+ * are processed by the DM core.  All others are delivered to the target.
+ * Returns a number <= 1 if message was processed by device mapper.
+ * Returns 2 if message should be delivered to the target.
+ */
+static int message_for_md(struct mapped_device *md, unsigned argc, char **argv,
+			  char *result, unsigned maxlen)
+{
+	int r;
+
+	if (**argv != '@')
+		return 2; /* no '@' prefix, deliver to target */
+
+	if (!strcasecmp(argv[0], "@cancel_deferred_remove")) {
+		if (argc != 1) {
+			DMERR("Invalid arguments for @cancel_deferred_remove");
+			return -EINVAL;
+		}
+		return dm_cancel_deferred_remove(md);
+	}
+
+	r = dm_stats_message(md, argc, argv, result, maxlen);
+	if (r < 2)
+		return r;
+
+	DMERR("Unsupported message sent to DM core: %s", argv[0]);
+	return -EINVAL;
+}
+
+/*
  * Pass a message to the target that's at the supplied device offset.
  */
 static int target_message(struct dm_ioctl *param, size_t param_size)
@@ -1331,6 +1512,9 @@ static int target_message(struct dm_ioctl *param, size_t param_size)
 	struct dm_table *table;
 	struct dm_target *ti;
 	struct dm_target_msg *tmsg = (void *) param + param->data_start;
+	size_t maxlen;
+	char *result = get_result_buffer(param, param_size, &maxlen);
+	int srcu_idx;
 
 	md = find_device(param);
 	if (!md)
@@ -1349,10 +1533,19 @@ static int target_message(struct dm_ioctl *param, size_t param_size)
 		goto out;
 	}
 
-	table = dm_get_live_table(md);
-	if (!table)
+	if (!argc) {
+		DMWARN("Empty message received.");
+		goto out_argv;
+	}
+
+	r = message_for_md(md, argc, argv, result, maxlen);
+	if (r <= 1)
 		goto out_argv;
 
+	table = dm_get_live_table(md, &srcu_idx);
+	if (!table)
+		goto out_table;
+
 	if (dm_deleting_md(md)) {
 		r = -ENXIO;
 		goto out_table;
@@ -1370,48 +1563,72 @@ static int target_message(struct dm_ioctl *param, size_t param_size)
 	}
 
  out_table:
-	dm_table_put(table);
+	dm_put_live_table(md, srcu_idx);
  out_argv:
 	kfree(argv);
  out:
-	param->data_size = 0;
+	if (r >= 0)
+		__dev_status(md, param);
+
+	if (r == 1) {
+		param->flags |= DM_DATA_OUT_FLAG;
+		if (dm_message_test_buffer_overflow(result, maxlen))
+			param->flags |= DM_BUFFER_FULL_FLAG;
+		else
+			param->data_size = param->data_start + strlen(result) + 1;
+		r = 0;
+	}
+
 	dm_put(md);
 	return r;
 }
 
+/*
+ * The ioctl parameter block consists of two parts, a dm_ioctl struct
+ * followed by a data buffer.  This flag is set if the second part,
+ * which has a variable size, is not used by the function processing
+ * the ioctl.
+ */
+#define IOCTL_FLAGS_NO_PARAMS	1
+
 /*-----------------------------------------------------------------
  * Implementation of open/close/ioctl on the special char
  * device.
  *---------------------------------------------------------------*/
-static ioctl_fn lookup_ioctl(unsigned int cmd)
+static ioctl_fn lookup_ioctl(unsigned int cmd, int *ioctl_flags)
 {
 	static struct {
 		int cmd;
+		int flags;
 		ioctl_fn fn;
 	} _ioctls[] = {
-		{DM_VERSION_CMD, NULL},	/* version is dealt with elsewhere */
-		{DM_REMOVE_ALL_CMD, remove_all},
-		{DM_LIST_DEVICES_CMD, list_devices},
-
-		{DM_DEV_CREATE_CMD, dev_create},
-		{DM_DEV_REMOVE_CMD, dev_remove},
-		{DM_DEV_RENAME_CMD, dev_rename},
-		{DM_DEV_SUSPEND_CMD, dev_suspend},
-		{DM_DEV_STATUS_CMD, dev_status},
-		{DM_DEV_WAIT_CMD, dev_wait},
-
-		{DM_TABLE_LOAD_CMD, table_load},
-		{DM_TABLE_CLEAR_CMD, table_clear},
-		{DM_TABLE_DEPS_CMD, table_deps},
-		{DM_TABLE_STATUS_CMD, table_status},
-
-		{DM_LIST_VERSIONS_CMD, list_versions},
-
-		{DM_TARGET_MSG_CMD, target_message},
-		{DM_DEV_SET_GEOMETRY_CMD, dev_set_geometry}
+		{DM_VERSION_CMD, 0, NULL}, /* version is dealt with elsewhere */
+		{DM_REMOVE_ALL_CMD, IOCTL_FLAGS_NO_PARAMS, remove_all},
+		{DM_LIST_DEVICES_CMD, 0, list_devices},
+
+		{DM_DEV_CREATE_CMD, IOCTL_FLAGS_NO_PARAMS, dev_create},
+		{DM_DEV_REMOVE_CMD, IOCTL_FLAGS_NO_PARAMS, dev_remove},
+		{DM_DEV_RENAME_CMD, 0, dev_rename},
+		{DM_DEV_SUSPEND_CMD, IOCTL_FLAGS_NO_PARAMS, dev_suspend},
+		{DM_DEV_STATUS_CMD, IOCTL_FLAGS_NO_PARAMS, dev_status},
+		{DM_DEV_WAIT_CMD, 0, dev_wait},
+
+		{DM_TABLE_LOAD_CMD, 0, table_load},
+		{DM_TABLE_CLEAR_CMD, IOCTL_FLAGS_NO_PARAMS, table_clear},
+		{DM_TABLE_DEPS_CMD, 0, table_deps},
+		{DM_TABLE_STATUS_CMD, 0, table_status},
+
+		{DM_LIST_VERSIONS_CMD, 0, list_versions},
+
+		{DM_TARGET_MSG_CMD, 0, target_message},
+		{DM_DEV_SET_GEOMETRY_CMD, 0, dev_set_geometry}
 	};
 
-	return (cmd >= ARRAY_SIZE(_ioctls)) ? NULL : _ioctls[cmd].fn;
+	if (unlikely(cmd >= ARRAY_SIZE(_ioctls)))
+		return NULL;
+
+	*ioctl_flags = _ioctls[cmd].flags;
+	return _ioctls[cmd].fn;
 }
 
 /*
@@ -1448,32 +1665,94 @@ static int check_version(unsigned int cmd, struct dm_ioctl __user *user)
 	return r;
 }
 
-static void free_params(struct dm_ioctl *param)
+#define DM_PARAMS_KMALLOC	0x0001	/* Params alloced with kmalloc */
+#define DM_PARAMS_VMALLOC	0x0002	/* Params alloced with vmalloc */
+#define DM_WIPE_BUFFER		0x0010	/* Wipe input buffer before returning from ioctl */
+
+static void free_params(struct dm_ioctl *param, size_t param_size, int param_flags)
 {
-	vfree(param);
+	if (param_flags & DM_WIPE_BUFFER)
+		memset(param, 0, param_size);
+
+	if (param_flags & DM_PARAMS_KMALLOC)
+		kfree(param);
+	if (param_flags & DM_PARAMS_VMALLOC)
+		vfree(param);
 }
 
-static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl **param)
+static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl *param_kernel,
+		       int ioctl_flags,
+		       struct dm_ioctl **param, int *param_flags)
 {
-	struct dm_ioctl tmp, *dmi;
+	struct dm_ioctl *dmi;
+	int secure_data;
+	const size_t minimum_data_size = sizeof(*param_kernel) - sizeof(param_kernel->data);
 
-	if (copy_from_user(&tmp, user, sizeof(tmp) - sizeof(tmp.data)))
+	if (copy_from_user(param_kernel, user, minimum_data_size))
 		return -EFAULT;
 
-	if (tmp.data_size < (sizeof(tmp) - sizeof(tmp.data)))
+	if (param_kernel->data_size < minimum_data_size)
 		return -EINVAL;
 
-	dmi = vmalloc(tmp.data_size);
-	if (!dmi)
+	secure_data = param_kernel->flags & DM_SECURE_DATA_FLAG;
+
+	*param_flags = secure_data ? DM_WIPE_BUFFER : 0;
+
+	if (ioctl_flags & IOCTL_FLAGS_NO_PARAMS) {
+		dmi = param_kernel;
+		dmi->data_size = minimum_data_size;
+		goto data_copied;
+	}
+
+	/*
+	 * Try to avoid low memory issues when a device is suspended.
+	 * Use kmalloc() rather than vmalloc() when we can.
+	 */
+	dmi = NULL;
+	if (param_kernel->data_size <= KMALLOC_MAX_SIZE) {
+		dmi = kmalloc(param_kernel->data_size, GFP_NOIO | __GFP_NORETRY | __GFP_NOMEMALLOC | __GFP_NOWARN);
+		if (dmi)
+			*param_flags |= DM_PARAMS_KMALLOC;
+	}
+
+	if (!dmi) {
+		unsigned noio_flag;
+		noio_flag = memalloc_noio_save();
+		dmi = __vmalloc(param_kernel->data_size, GFP_NOIO | __GFP_REPEAT | __GFP_HIGH | __GFP_HIGHMEM, PAGE_KERNEL);
+		memalloc_noio_restore(noio_flag);
+		if (dmi)
+			*param_flags |= DM_PARAMS_VMALLOC;
+	}
+
+	if (!dmi) {
+		if (secure_data && clear_user(user, param_kernel->data_size))
+			return -EFAULT;
 		return -ENOMEM;
+	}
 
-	if (copy_from_user(dmi, user, tmp.data_size)) {
-		vfree(dmi);
-		return -EFAULT;
+	if (copy_from_user(dmi, user, param_kernel->data_size))
+		goto bad;
+
+data_copied:
+	/*
+	 * Abort if something changed the ioctl data while it was being copied.
+	 */
+	if (dmi->data_size != param_kernel->data_size) {
+		DMERR("rejecting ioctl: data size modified while processing parameters");
+		goto bad;
 	}
 
+	/* Wipe the user buffer so we do not return it to userspace */
+	if (secure_data && clear_user(user, param_kernel->data_size))
+		goto bad;
+
 	*param = dmi;
 	return 0;
+
+bad:
+	free_params(dmi, param_kernel->data_size, *param_flags);
+
+	return -EFAULT;
 }
 
 static int validate_params(uint cmd, struct dm_ioctl *param)
@@ -1481,6 +1760,8 @@ static int validate_params(uint cmd, struct dm_ioctl *param)
 	/* Always clear this flag */
 	param->flags &= ~DM_BUFFER_FULL_FLAG;
 	param->flags &= ~DM_UEVENT_GENERATED_FLAG;
+	param->flags &= ~DM_SECURE_DATA_FLAG;
+	param->flags &= ~DM_DATA_OUT_FLAG;
 
 	/* Ignores parameters */
 	if (cmd == DM_REMOVE_ALL_CMD ||
@@ -1508,10 +1789,13 @@ static int validate_params(uint cmd, struct dm_ioctl *param)
 static int ctl_ioctl(uint command, struct dm_ioctl __user *user)
 {
 	int r = 0;
+	int ioctl_flags;
+	int param_flags;
 	unsigned int cmd;
 	struct dm_ioctl *uninitialized_var(param);
 	ioctl_fn fn = NULL;
-	size_t param_size;
+	size_t input_param_size;
+	struct dm_ioctl param_kernel;
 
 	/* only root can play with this */
 	if (!capable(CAP_SYS_ADMIN))
@@ -1536,35 +1820,31 @@ static int ctl_ioctl(uint command, struct dm_ioctl __user *user)
 	if (cmd == DM_VERSION_CMD)
 		return 0;
 
-	fn = lookup_ioctl(cmd);
+	fn = lookup_ioctl(cmd, &ioctl_flags);
 	if (!fn) {
 		DMWARN("dm_ctl_ioctl: unknown command 0x%x", command);
 		return -ENOTTY;
 	}
 
 	/*
-	 * Trying to avoid low memory issues when a device is
-	 * suspended.
-	 */
-	current->flags |= PF_MEMALLOC;
-
-	/*
 	 * Copy the parameters into kernel space.
 	 */
-	r = copy_params(user, &param);
-
-	current->flags &= ~PF_MEMALLOC;
+	r = copy_params(user, &param_kernel, ioctl_flags, &param, &param_flags);
 
 	if (r)
 		return r;
 
+	input_param_size = param->data_size;
 	r = validate_params(cmd, param);
 	if (r)
 		goto out;
 
-	param_size = param->data_size;
 	param->data_size = sizeof(*param);
-	r = fn(param, param_size);
+	r = fn(param, input_param_size);
+
+	if (unlikely(param->flags & DM_BUFFER_FULL_FLAG) &&
+	    unlikely(ioctl_flags & IOCTL_FLAGS_NO_PARAMS))
+		DMERR("ioctl %d tried to output some data but has IOCTL_FLAGS_NO_PARAMS set", cmd);
 
 	/*
 	 * Copy the results back to userland.
@@ -1572,8 +1852,8 @@ static int ctl_ioctl(uint command, struct dm_ioctl __user *user)
 	if (!r && copy_to_user(user, param, param->data_size))
 		r = -EFAULT;
 
- out:
-	free_params(param);
+out:
+	free_params(param, input_param_size, param_flags);
 	return r;
 }