Merge master.kernel.org:/pub/scm/linux/kernel/git/agk/linux-2.6-dm

* master.kernel.org:/pub/scm/linux/kernel/git/agk/linux-2.6-dm: (29 commits)
  dm crypt: tidy pending
  dm mpath: send uevents
  dm: uevent generate events
  dm: add uevent to core
  dm: export name and uuid
  dm raid1: add mirror_set to struct mirror
  dm log: split suspend
  dm mpath: hp retry if not ready
  dm mpath: add hp handler
  dm mpath: add retry pg init
  dm crypt: tidy labels
  dm crypt: tidy whitespace
  dm crypt: add post processing queue
  dm crypt: use per device singlethread workqueues
  dm mpath: emc fix an error message
  dm: bio_list macro renaming
  dm io:ctl remove vmalloc void cast
  dm: tidy bio_io_error usage
  kcopyd use mutex instead of semaphore
  dm: use kzalloc
  ...

Documentation/device-mapper/dm-uevent.txt

@@ -0,0 +1,97 @@
+The device-mapper uevent code adds the capability to device-mapper to create
+and send kobject uevents (uevents).  Previously device-mapper events were only
+available through the ioctl interface.  The advantage of the uevents interface
+is the event contains environment attributes providing increased context for
+the event avoiding the need to query the state of the device-mapper device after
+the event is received.
+
+There are two functions currently for device-mapper events.  The first function
+listed creates the event and the second function sends the event(s).
+
+void dm_path_uevent(enum dm_uevent_type event_type, struct dm_target *ti,
+                    const char *path, unsigned nr_valid_paths)
+
+void dm_send_uevents(struct list_head *events, struct kobject *kobj)
+
+
+The variables added to the uevent environment are:
+
+Variable Name: DM_TARGET
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description:
+Value: Name of device-mapper target that generated the event.
+
+Variable Name: DM_ACTION
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description:
+Value: Device-mapper specific action that caused the uevent action.
+	PATH_FAILED - A path has failed.
+	PATH_REINSTATED - A path has been reinstated.
+
+Variable Name: DM_SEQNUM
+Uevent Action(s): KOBJ_CHANGE
+Type: unsigned integer
+Description: A sequence number for this specific device-mapper device.
+Value: Valid unsigned integer range.
+
+Variable Name: DM_PATH
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description: Major and minor number of the path device pertaining to this
+event.
+Value: Path name in the form of "Major:Minor"
+
+Variable Name: DM_NR_VALID_PATHS
+Uevent Action(s): KOBJ_CHANGE
+Type: unsigned integer
+Description:
+Value: Valid unsigned integer range.
+
+Variable Name: DM_NAME
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description: Name of the device-mapper device.
+Value: Name
+
+Variable Name: DM_UUID
+Uevent Action(s): KOBJ_CHANGE
+Type: string
+Description: UUID of the device-mapper device.
+Value: UUID. (Empty string if there isn't one.)
+
+An example of the uevents generated as captured by udevmonitor is shown
+below.
+
+1.) Path failure.
+UEVENT[1192521009.711215] change@/block/dm-3
+ACTION=change
+DEVPATH=/block/dm-3
+SUBSYSTEM=block
+DM_TARGET=multipath
+DM_ACTION=PATH_FAILED
+DM_SEQNUM=1
+DM_PATH=8:32
+DM_NR_VALID_PATHS=0
+DM_NAME=mpath2
+DM_UUID=mpath-35333333000002328
+MINOR=3
+MAJOR=253
+SEQNUM=1130
+
+2.) Path reinstate.
+UEVENT[1192521132.989927] change@/block/dm-3
+ACTION=change
+DEVPATH=/block/dm-3
+SUBSYSTEM=block
+DM_TARGET=multipath
+DM_ACTION=PATH_REINSTATED
+DM_SEQNUM=2
+DM_PATH=8:32
+DM_NR_VALID_PATHS=1
+DM_NAME=mpath2
+DM_UUID=mpath-35333333000002328
+MINOR=3
+MAJOR=253
+SEQNUM=1131

drivers/md/Kconfig

@@ -267,6 +267,12 @@ config DM_MULTIPATH_RDAC
 	---help---
 	  Multipath support for LSI/Engenio RDAC.
 
+config DM_MULTIPATH_HP
+        tristate "HP MSA multipath support (EXPERIMENTAL)"
+        depends on DM_MULTIPATH && BLK_DEV_DM && EXPERIMENTAL
+        ---help---
+          Multipath support for HP MSA (Active/Passive) series hardware.
+
 config DM_DELAY
 	tristate "I/O delaying target (EXPERIMENTAL)"
 	depends on BLK_DEV_DM && EXPERIMENTAL
@@ -276,4 +282,10 @@ config DM_DELAY
 
 	If unsure, say N.
 
+config DM_UEVENT
+	bool "DM uevents (EXPERIMENTAL)"
+	depends on BLK_DEV_DM && EXPERIMENTAL
+	---help---
+	Generate udev events for DM events.
+
 endif # MD

drivers/md/Makefile

@@ -8,6 +8,7 @@ dm-multipath-objs := dm-hw-handler.o dm-path-selector.o dm-mpath.o
 dm-snapshot-objs := dm-snap.o dm-exception-store.o
 dm-mirror-objs	:= dm-log.o dm-raid1.o
 dm-rdac-objs	:= dm-mpath-rdac.o
+dm-hp-sw-objs	:= dm-mpath-hp-sw.o
 md-mod-objs     := md.o bitmap.o
 raid456-objs	:= raid5.o raid6algos.o raid6recov.o raid6tables.o \
 		   raid6int1.o raid6int2.o raid6int4.o \
@@ -35,6 +36,7 @@ obj-$(CONFIG_DM_CRYPT)		+= dm-crypt.o
 obj-$(CONFIG_DM_DELAY)		+= dm-delay.o
 obj-$(CONFIG_DM_MULTIPATH)	+= dm-multipath.o dm-round-robin.o
 obj-$(CONFIG_DM_MULTIPATH_EMC)	+= dm-emc.o
+obj-$(CONFIG_DM_MULTIPATH_HP)	+= dm-hp-sw.o
 obj-$(CONFIG_DM_MULTIPATH_RDAC)	+= dm-rdac.o
 obj-$(CONFIG_DM_SNAPSHOT)	+= dm-snapshot.o
 obj-$(CONFIG_DM_MIRROR)		+= dm-mirror.o
@@ -48,6 +50,10 @@ ifeq ($(CONFIG_ALTIVEC),y)
 altivec_flags := -maltivec -mabi=altivec
 endif
 
+ifeq ($(CONFIG_DM_UEVENT),y)
+dm-mod-objs			+= dm-uevent.o
+endif
+
 targets += raid6int1.c
 $(obj)/raid6int1.c:   UNROLL := 1
 $(obj)/raid6int1.c:   $(src)/raid6int.uc $(src)/unroll.pl FORCE

drivers/md/dm-bio-list.h

@@ -21,11 +21,6 @@ static inline int bio_list_empty(const struct bio_list *bl)
 	return bl->head == NULL;
 }
 
-#define BIO_LIST_INIT { .head = NULL, .tail = NULL }
-
-#define BIO_LIST(bl) \
-	struct bio_list bl = BIO_LIST_INIT
-
 static inline void bio_list_init(struct bio_list *bl)
 {
 	bl->head = bl->tail = NULL;

drivers/md/dm-crypt.c

@@ -36,7 +36,6 @@ struct dm_crypt_io {
 	struct work_struct work;
 	atomic_t pending;
 	int error;
-	int post_process;
 };
 
 /*
@@ -57,7 +56,7 @@ struct crypt_config;
 
 struct crypt_iv_operations {
 	int (*ctr)(struct crypt_config *cc, struct dm_target *ti,
-	           const char *opts);
+		   const char *opts);
 	void (*dtr)(struct crypt_config *cc);
 	const char *(*status)(struct crypt_config *cc);
 	int (*generator)(struct crypt_config *cc, u8 *iv, sector_t sector);
@@ -80,6 +79,8 @@ struct crypt_config {
 	mempool_t *page_pool;
 	struct bio_set *bs;
 
+	struct workqueue_struct *io_queue;
+	struct workqueue_struct *crypt_queue;
 	/*
 	 * crypto related data
 	 */
@@ -137,7 +138,7 @@ static int crypt_iv_plain_gen(struct crypt_config *cc, u8 *iv, sector_t sector)
 }
 
 static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti,
-	                      const char *opts)
+			      const char *opts)
 {
 	struct crypto_cipher *essiv_tfm;
 	struct crypto_hash *hash_tfm;
@@ -175,6 +176,7 @@ static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti,
 
 	if (err) {
 		ti->error = "Error calculating hash in ESSIV";
+		kfree(salt);
 		return err;
 	}
 
@@ -188,7 +190,7 @@ static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti,
 	if (crypto_cipher_blocksize(essiv_tfm) !=
 	    crypto_blkcipher_ivsize(cc->tfm)) {
 		ti->error = "Block size of ESSIV cipher does "
-			        "not match IV size of block cipher";
+			    "not match IV size of block cipher";
 		crypto_free_cipher(essiv_tfm);
 		kfree(salt);
 		return -EINVAL;
@@ -319,10 +321,10 @@ crypt_convert_scatterlist(struct crypt_config *cc, struct scatterlist *out,
 	return r;
 }
 
-static void
-crypt_convert_init(struct crypt_config *cc, struct convert_context *ctx,
-                   struct bio *bio_out, struct bio *bio_in,
-                   sector_t sector, int write)
+static void crypt_convert_init(struct crypt_config *cc,
+			       struct convert_context *ctx,
+			       struct bio *bio_out, struct bio *bio_in,
+			       sector_t sector, int write)
 {
 	ctx->bio_in = bio_in;
 	ctx->bio_out = bio_out;
@@ -338,7 +340,7 @@ crypt_convert_init(struct crypt_config *cc, struct convert_context *ctx,
  * Encrypt / decrypt data from one bio to another one (can be the same one)
  */
 static int crypt_convert(struct crypt_config *cc,
-                         struct convert_context *ctx)
+			 struct convert_context *ctx)
 {
 	int r = 0;
 
@@ -370,7 +372,7 @@ static int crypt_convert(struct crypt_config *cc,
 		}
 
 		r = crypt_convert_scatterlist(cc, &sg_out, &sg_in, sg_in.length,
-		                              ctx->write, ctx->sector);
+					      ctx->write, ctx->sector);
 		if (r < 0)
 			break;
 
@@ -380,13 +382,13 @@ static int crypt_convert(struct crypt_config *cc,
 	return r;
 }
 
- static void dm_crypt_bio_destructor(struct bio *bio)
- {
+static void dm_crypt_bio_destructor(struct bio *bio)
+{
 	struct dm_crypt_io *io = bio->bi_private;
 	struct crypt_config *cc = io->target->private;
 
 	bio_free(bio, cc->bs);
- }
+}
 
 /*
  * Generate a new unfragmented bio with the given size
@@ -458,7 +460,7 @@ static void crypt_free_buffer_pages(struct crypt_config *cc, struct bio *clone)
  * One of the bios was finished. Check for completion of
  * the whole request and correctly clean up the buffer.
  */
-static void dec_pending(struct dm_crypt_io *io, int error)
+static void crypt_dec_pending(struct dm_crypt_io *io, int error)
 {
 	struct crypt_config *cc = (struct crypt_config *) io->target->private;
 
@@ -474,18 +476,36 @@ static void dec_pending(struct dm_crypt_io *io, int error)
 }
 
 /*
- * kcryptd:
+ * kcryptd/kcryptd_io:
  *
  * Needed because it would be very unwise to do decryption in an
  * interrupt context.
+ *
+ * kcryptd performs the actual encryption or decryption.
+ *
+ * kcryptd_io performs the IO submission.
+ *
+ * They must be separated as otherwise the final stages could be
+ * starved by new requests which can block in the first stages due
+ * to memory allocation.
  */
-static struct workqueue_struct *_kcryptd_workqueue;
 static void kcryptd_do_work(struct work_struct *work);
+static void kcryptd_do_crypt(struct work_struct *work);
 
 static void kcryptd_queue_io(struct dm_crypt_io *io)
 {
+	struct crypt_config *cc = io->target->private;
+
 	INIT_WORK(&io->work, kcryptd_do_work);
-	queue_work(_kcryptd_workqueue, &io->work);
+	queue_work(cc->io_queue, &io->work);
+}
+
+static void kcryptd_queue_crypt(struct dm_crypt_io *io)
+{
+	struct crypt_config *cc = io->target->private;
+
+	INIT_WORK(&io->work, kcryptd_do_crypt);
+	queue_work(cc->crypt_queue, &io->work);
 }
 
 static void crypt_endio(struct bio *clone, int error)
@@ -508,13 +528,12 @@ static void crypt_endio(struct bio *clone, int error)
 	}
 
 	bio_put(clone);
-	io->post_process = 1;
-	kcryptd_queue_io(io);
+	kcryptd_queue_crypt(io);
 	return;
 
 out:
 	bio_put(clone);
-	dec_pending(io, error);
+	crypt_dec_pending(io, error);
 }
 
 static void clone_init(struct dm_crypt_io *io, struct bio *clone)
@@ -544,7 +563,7 @@ static void process_read(struct dm_crypt_io *io)
 	 */
 	clone = bio_alloc_bioset(GFP_NOIO, bio_segments(base_bio), cc->bs);
 	if (unlikely(!clone)) {
-		dec_pending(io, -ENOMEM);
+		crypt_dec_pending(io, -ENOMEM);
 		return;
 	}
 
@@ -579,7 +598,7 @@ static void process_write(struct dm_crypt_io *io)
 	while (remaining) {
 		clone = crypt_alloc_buffer(io, remaining);
 		if (unlikely(!clone)) {
-			dec_pending(io, -ENOMEM);
+			crypt_dec_pending(io, -ENOMEM);
 			return;
 		}
 
@@ -589,7 +608,7 @@ static void process_write(struct dm_crypt_io *io)
 		if (unlikely(crypt_convert(cc, &ctx) < 0)) {
 			crypt_free_buffer_pages(cc, clone);
 			bio_put(clone);
-			dec_pending(io, -EIO);
+			crypt_dec_pending(io, -EIO);
 			return;
 		}
 
@@ -624,17 +643,23 @@ static void process_read_endio(struct dm_crypt_io *io)
 	crypt_convert_init(cc, &ctx, io->base_bio, io->base_bio,
 			   io->base_bio->bi_sector - io->target->begin, 0);
 
-	dec_pending(io, crypt_convert(cc, &ctx));
+	crypt_dec_pending(io, crypt_convert(cc, &ctx));
 }
 
 static void kcryptd_do_work(struct work_struct *work)
 {
 	struct dm_crypt_io *io = container_of(work, struct dm_crypt_io, work);
 
-	if (io->post_process)
-		process_read_endio(io);
-	else if (bio_data_dir(io->base_bio) == READ)
+	if (bio_data_dir(io->base_bio) == READ)
 		process_read(io);
+}
+
+static void kcryptd_do_crypt(struct work_struct *work)
+{
+	struct dm_crypt_io *io = container_of(work, struct dm_crypt_io, work);
+
+	if (bio_data_dir(io->base_bio) == READ)
+		process_read_endio(io);
 	else
 		process_write(io);
 }
@@ -690,7 +715,7 @@ static int crypt_set_key(struct crypt_config *cc, char *key)
 	cc->key_size = key_size; /* initial settings */
 
 	if ((!key_size && strcmp(key, "-")) ||
-	    (key_size && crypt_decode_key(cc->key, key, key_size) < 0))
+	   (key_size && crypt_decode_key(cc->key, key, key_size) < 0))
 		return -EINVAL;
 
 	set_bit(DM_CRYPT_KEY_VALID, &cc->flags);
@@ -746,7 +771,7 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 
  	if (crypt_set_key(cc, argv[1])) {
 		ti->error = "Error decoding key";
-		goto bad1;
+		goto bad_cipher;
 	}
 
 	/* Compatiblity mode for old dm-crypt cipher strings */
@@ -757,19 +782,19 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 
 	if (strcmp(chainmode, "ecb") && !ivmode) {
 		ti->error = "This chaining mode requires an IV mechanism";
-		goto bad1;
+		goto bad_cipher;
 	}
 
-	if (snprintf(cc->cipher, CRYPTO_MAX_ALG_NAME, "%s(%s)", chainmode, 
-		     cipher) >= CRYPTO_MAX_ALG_NAME) {
+	if (snprintf(cc->cipher, CRYPTO_MAX_ALG_NAME, "%s(%s)",
+		     chainmode, cipher) >= CRYPTO_MAX_ALG_NAME) {
 		ti->error = "Chain mode + cipher name is too long";
-		goto bad1;
+		goto bad_cipher;
 	}
 
 	tfm = crypto_alloc_blkcipher(cc->cipher, 0, CRYPTO_ALG_ASYNC);
 	if (IS_ERR(tfm)) {
 		ti->error = "Error allocating crypto tfm";
-		goto bad1;
+		goto bad_cipher;
 	}
 
 	strcpy(cc->cipher, cipher);
@@ -793,18 +818,18 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 		cc->iv_gen_ops = &crypt_iv_null_ops;
 	else {
 		ti->error = "Invalid IV mode";
-		goto bad2;
+		goto bad_ivmode;
 	}
 
 	if (cc->iv_gen_ops && cc->iv_gen_ops->ctr &&
 	    cc->iv_gen_ops->ctr(cc, ti, ivopts) < 0)
-		goto bad2;
+		goto bad_ivmode;
 
 	cc->iv_size = crypto_blkcipher_ivsize(tfm);
 	if (cc->iv_size)
 		/* at least a 64 bit sector number should fit in our buffer */
 		cc->iv_size = max(cc->iv_size,
-		                  (unsigned int)(sizeof(u64) / sizeof(u8)));
+				  (unsigned int)(sizeof(u64) / sizeof(u8)));
 	else {
 		if (cc->iv_gen_ops) {
 			DMWARN("Selected cipher does not support IVs");
@@ -817,13 +842,13 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	cc->io_pool = mempool_create_slab_pool(MIN_IOS, _crypt_io_pool);
 	if (!cc->io_pool) {
 		ti->error = "Cannot allocate crypt io mempool";
-		goto bad3;
+		goto bad_slab_pool;
 	}
 
 	cc->page_pool = mempool_create_page_pool(MIN_POOL_PAGES, 0);
 	if (!cc->page_pool) {
 		ti->error = "Cannot allocate page mempool";
-		goto bad4;
+		goto bad_page_pool;
 	}
 
 	cc->bs = bioset_create(MIN_IOS, MIN_IOS);
@@ -834,25 +859,25 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 
 	if (crypto_blkcipher_setkey(tfm, cc->key, key_size) < 0) {
 		ti->error = "Error setting key";
-		goto bad5;
+		goto bad_device;
 	}
 
 	if (sscanf(argv[2], "%llu", &tmpll) != 1) {
 		ti->error = "Invalid iv_offset sector";
-		goto bad5;
+		goto bad_device;
 	}
 	cc->iv_offset = tmpll;
 
 	if (sscanf(argv[4], "%llu", &tmpll) != 1) {
 		ti->error = "Invalid device sector";
-		goto bad5;
+		goto bad_device;
 	}
 	cc->start = tmpll;
 
 	if (dm_get_device(ti, argv[3], cc->start, ti->len,
-	                  dm_table_get_mode(ti->table), &cc->dev)) {
+			  dm_table_get_mode(ti->table), &cc->dev)) {
 		ti->error = "Device lookup failed";
-		goto bad5;
+		goto bad_device;
 	}
 
 	if (ivmode && cc->iv_gen_ops) {
@@ -861,27 +886,45 @@ static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 		cc->iv_mode = kmalloc(strlen(ivmode) + 1, GFP_KERNEL);
 		if (!cc->iv_mode) {
 			ti->error = "Error kmallocing iv_mode string";
-			goto bad5;
+			goto bad_ivmode_string;
 		}
 		strcpy(cc->iv_mode, ivmode);
 	} else
 		cc->iv_mode = NULL;
 
+	cc->io_queue = create_singlethread_workqueue("kcryptd_io");
+	if (!cc->io_queue) {
+		ti->error = "Couldn't create kcryptd io queue";
+		goto bad_io_queue;
+	}
+
+	cc->crypt_queue = create_singlethread_workqueue("kcryptd");
+	if (!cc->crypt_queue) {
+		ti->error = "Couldn't create kcryptd queue";
+		goto bad_crypt_queue;
+	}
+
 	ti->private = cc;
 	return 0;
 
-bad5:
+bad_crypt_queue:
+	destroy_workqueue(cc->io_queue);
+bad_io_queue:
+	kfree(cc->iv_mode);
+bad_ivmode_string:
+	dm_put_device(ti, cc->dev);
+bad_device:
 	bioset_free(cc->bs);
 bad_bs:
 	mempool_destroy(cc->page_pool);
-bad4:
+bad_page_pool:
 	mempool_destroy(cc->io_pool);
-bad3:
+bad_slab_pool:
 	if (cc->iv_gen_ops && cc->iv_gen_ops->dtr)
 		cc->iv_gen_ops->dtr(cc);
-bad2:
+bad_ivmode:
 	crypto_free_blkcipher(tfm);
-bad1:
+bad_cipher:
 	/* Must zero key material before freeing */
 	memset(cc, 0, sizeof(*cc) + cc->key_size * sizeof(u8));
 	kfree(cc);
@@ -892,7 +935,8 @@ static void crypt_dtr(struct dm_target *ti)
 {
 	struct crypt_config *cc = (struct crypt_config *) ti->private;
 
-	flush_workqueue(_kcryptd_workqueue);
+	destroy_workqueue(cc->io_queue);
+	destroy_workqueue(cc->crypt_queue);
 
 	bioset_free(cc->bs);
 	mempool_destroy(cc->page_pool);
@@ -918,9 +962,13 @@ static int crypt_map(struct dm_target *ti, struct bio *bio,
 	io = mempool_alloc(cc->io_pool, GFP_NOIO);
 	io->target = ti;
 	io->base_bio = bio;
-	io->error = io->post_process = 0;
+	io->error = 0;
 	atomic_set(&io->pending, 0);
-	kcryptd_queue_io(io);
+
+	if (bio_data_dir(io->base_bio) == READ)
+		kcryptd_queue_io(io);
+	else
+		kcryptd_queue_crypt(io);
 
 	return DM_MAPIO_SUBMITTED;
 }
@@ -1037,25 +1085,12 @@ static int __init dm_crypt_init(void)
 	if (!_crypt_io_pool)
 		return -ENOMEM;
 
-	_kcryptd_workqueue = create_workqueue("kcryptd");
-	if (!_kcryptd_workqueue) {
-		r = -ENOMEM;
-		DMERR("couldn't create kcryptd");
-		goto bad1;
-	}
-
 	r = dm_register_target(&crypt_target);
 	if (r < 0) {
 		DMERR("register failed %d", r);
-		goto bad2;
+		kmem_cache_destroy(_crypt_io_pool);
 	}
 
-	return 0;
-
-bad2:
-	destroy_workqueue(_kcryptd_workqueue);
-bad1:
-	kmem_cache_destroy(_crypt_io_pool);
 	return r;
 }
 
@@ -1066,7 +1101,6 @@ static void __exit dm_crypt_exit(void)
 	if (r < 0)
 		DMERR("unregister failed %d", r);
 
-	destroy_workqueue(_kcryptd_workqueue);
 	kmem_cache_destroy(_crypt_io_pool);
 }
 

drivers/md/dm-delay.c

@@ -83,7 +83,7 @@ static struct bio *flush_delayed_bios(struct delay_c *dc, int flush_all)
 	struct dm_delay_info *delayed, *next;
 	unsigned long next_expires = 0;
 	int start_timer = 0;
-	BIO_LIST(flush_bios);
+	struct bio_list flush_bios = { };
 
 	mutex_lock(&delayed_bios_lock);
 	list_for_each_entry_safe(delayed, next, &dc->delayed_bios, list) {
@@ -163,34 +163,32 @@ static int delay_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 		goto bad;
 	}
 
-	if (argc == 3) {
-		dc->dev_write = NULL;
+	dc->dev_write = NULL;
+	if (argc == 3)
 		goto out;
-	}
 
 	if (sscanf(argv[4], "%llu", &tmpll) != 1) {
 		ti->error = "Invalid write device sector";
-		goto bad;
+		goto bad_dev_read;
 	}
 	dc->start_write = tmpll;
 
 	if (sscanf(argv[5], "%u", &dc->write_delay) != 1) {
 		ti->error = "Invalid write delay";
-		goto bad;
+		goto bad_dev_read;
 	}
 
 	if (dm_get_device(ti, argv[3], dc->start_write, ti->len,
 			  dm_table_get_mode(ti->table), &dc->dev_write)) {
 		ti->error = "Write device lookup failed";
-		dm_put_device(ti, dc->dev_read);
-		goto bad;
+		goto bad_dev_read;
 	}
 
 out:
 	dc->delayed_pool = mempool_create_slab_pool(128, delayed_cache);
 	if (!dc->delayed_pool) {
 		DMERR("Couldn't create delayed bio pool.");
-		goto bad;
+		goto bad_dev_write;
 	}
 
 	setup_timer(&dc->delay_timer, handle_delayed_timer, (unsigned long)dc);
@@ -203,6 +201,11 @@ out:
 	ti->private = dc;
 	return 0;
 
+bad_dev_write:
+	if (dc->dev_write)
+		dm_put_device(ti, dc->dev_write);
+bad_dev_read:
+	dm_put_device(ti, dc->dev_read);
 bad:
 	kfree(dc);
 	return -EINVAL;
@@ -305,7 +308,7 @@ static int delay_status(struct dm_target *ti, status_type_t type,
 		       (unsigned long long) dc->start_read,
 		       dc->read_delay);
 		if (dc->dev_write)
-			DMEMIT("%s %llu %u", dc->dev_write->name,
+			DMEMIT(" %s %llu %u", dc->dev_write->name,
 			       (unsigned long long) dc->start_write,
 			       dc->write_delay);
 		break;

drivers/md/dm-emc.c

@@ -81,7 +81,7 @@ static struct bio *get_failover_bio(struct dm_path *path, unsigned data_size)
 	}
 
 	if (bio_add_page(bio, page, data_size, 0) != data_size) {
-		DMERR("get_failover_bio: alloc_page() failed.");
+		DMERR("get_failover_bio: bio_add_page() failed.");
 		__free_page(page);
 		bio_put(bio);
 		return NULL;
@@ -211,12 +211,10 @@ fail_path:
 
 static struct emc_handler *alloc_emc_handler(void)
 {
-	struct emc_handler *h = kmalloc(sizeof(*h), GFP_KERNEL);
+	struct emc_handler *h = kzalloc(sizeof(*h), GFP_KERNEL);
 
-	if (h) {
-		memset(h, 0, sizeof(*h));
+	if (h)
 		spin_lock_init(&h->lock);
-	}
 
 	return h;
 }

drivers/md/dm-hw-handler.c

@@ -91,12 +91,10 @@ void dm_put_hw_handler(struct hw_handler_type *hwht)
 
 static struct hwh_internal *_alloc_hw_handler(struct hw_handler_type *hwht)
 {
-	struct hwh_internal *hwhi = kmalloc(sizeof(*hwhi), GFP_KERNEL);
+	struct hwh_internal *hwhi = kzalloc(sizeof(*hwhi), GFP_KERNEL);
 
-	if (hwhi) {
-		memset(hwhi, 0, sizeof(*hwhi));
+	if (hwhi)
 		hwhi->hwht = *hwht;
-	}
 
 	return hwhi;
 }

drivers/md/dm-hw-handler.h

@@ -58,5 +58,6 @@ unsigned dm_scsi_err_handler(struct hw_handler *hwh, struct bio *bio);
 #define MP_FAIL_PATH 1
 #define MP_BYPASS_PG 2
 #define MP_ERROR_IO  4	/* Don't retry this I/O */
+#define MP_RETRY 8
 
 #endif

drivers/md/dm-ioctl.c

@@ -700,7 +700,7 @@ static int dev_rename(struct dm_ioctl *param, size_t param_size)
 	int r;
 	char *new_name = (char *) param + param->data_start;
 
-	if (new_name < (char *) (param + 1) ||
+	if (new_name < (char *) param->data ||
 	    invalid_str(new_name, (void *) param + param_size)) {
 		DMWARN("Invalid new logical volume name supplied.");
 		return -EINVAL;
@@ -726,7 +726,7 @@ static int dev_set_geometry(struct dm_ioctl *param, size_t param_size)
 	if (!md)
 		return -ENXIO;
 
-	if (geostr < (char *) (param + 1) ||
+	if (geostr < (char *) param->data ||
 	    invalid_str(geostr, (void *) param + param_size)) {
 		DMWARN("Invalid geometry supplied.");
 		goto out;
@@ -1233,7 +1233,7 @@ static int target_message(struct dm_ioctl *param, size_t param_size)
 	if (r)
 		goto out;
 
-	if (tmsg < (struct dm_target_msg *) (param + 1) ||
+	if (tmsg < (struct dm_target_msg *) param->data ||
 	    invalid_str(tmsg->message, (void *) param + param_size)) {
 		DMWARN("Invalid target message parameters.");
 		r = -EINVAL;
@@ -1358,7 +1358,7 @@ static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl **param)
 	if (tmp.data_size < sizeof(tmp))
 		return -EINVAL;
 
-	dmi = (struct dm_ioctl *) vmalloc(tmp.data_size);
+	dmi = vmalloc(tmp.data_size);
 	if (!dmi)
 		return -ENOMEM;
 
@@ -1515,3 +1515,35 @@ void dm_interface_exit(void)
 
 	dm_hash_exit();
 }
+
+/**
+ * dm_copy_name_and_uuid - Copy mapped device name & uuid into supplied buffers
+ * @md: Pointer to mapped_device
+ * @name: Buffer (size DM_NAME_LEN) for name
+ * @uuid: Buffer (size DM_UUID_LEN) for uuid or empty string if uuid not defined
+ */
+int dm_copy_name_and_uuid(struct mapped_device *md, char *name, char *uuid)
+{
+	int r = 0;
+	struct hash_cell *hc;
+
+	if (!md)
+		return -ENXIO;
+
+	dm_get(md);
+	down_read(&_hash_lock);
+	hc = dm_get_mdptr(md);
+	if (!hc || hc->md != md) {
+		r = -ENXIO;
+		goto out;
+	}
+
+	strcpy(name, hc->name);
+	strcpy(uuid, hc->uuid ? : "");
+
+out:
+	up_read(&_hash_lock);
+	dm_put(md);
+
+	return r;
+}

drivers/md/dm-log.c

@@ -696,7 +696,7 @@ static struct dirty_log_type _disk_type = {
 	.module = THIS_MODULE,
 	.ctr = disk_ctr,
 	.dtr = disk_dtr,
-	.suspend = disk_flush,
+	.postsuspend = disk_flush,
 	.resume = disk_resume,
 	.get_region_size = core_get_region_size,
 	.is_clean = core_is_clean,

drivers/md/dm-log.h

@@ -32,7 +32,8 @@ struct dirty_log_type {
 	 * There are times when we don't want the log to touch
 	 * the disk.
 	 */
-	int (*suspend)(struct dirty_log *log);
+	int (*presuspend)(struct dirty_log *log);
+	int (*postsuspend)(struct dirty_log *log);
 	int (*resume)(struct dirty_log *log);
 
 	/*

drivers/md/dm-mpath-hp-sw.c

@@ -0,0 +1,248 @@
+/*
+ * Copyright (C) 2005 Mike Christie, All rights reserved.
+ * Copyright (C) 2007 Red Hat, Inc. All rights reserved.
+ * Authors: Mike Christie
+ *          Dave Wysochanski
+ *
+ * This file is released under the GPL.
+ *
+ * This module implements the specific path activation code for
+ * HP StorageWorks and FSC FibreCat Asymmetric (Active/Passive)
+ * storage arrays.
+ * These storage arrays have controller-based failover, not
+ * LUN-based failover.  However, LUN-based failover is the design
+ * of dm-multipath. Thus, this module is written for LUN-based failover.
+ */
+#include <linux/blkdev.h>
+#include <linux/list.h>
+#include <linux/types.h>
+#include <scsi/scsi.h>
+#include <scsi/scsi_cmnd.h>
+#include <scsi/scsi_dbg.h>
+
+#include "dm.h"
+#include "dm-hw-handler.h"
+
+#define DM_MSG_PREFIX "multipath hp-sw"
+#define DM_HP_HWH_NAME "hp-sw"
+#define DM_HP_HWH_VER "1.0.0"
+
+struct hp_sw_context {
+	unsigned char sense[SCSI_SENSE_BUFFERSIZE];
+};
+
+/*
+ * hp_sw_error_is_retryable - Is an HP-specific check condition retryable?
+ * @req: path activation request
+ *
+ * Examine error codes of request and determine whether the error is retryable.
+ * Some error codes are already retried by scsi-ml (see
+ * scsi_decide_disposition), but some HP specific codes are not.
+ * The intent of this routine is to supply the logic for the HP specific
+ * check conditions.
+ *
+ * Returns:
+ *  1 - command completed with retryable error
+ *  0 - command completed with non-retryable error
+ *
+ * Possible optimizations
+ * 1. More hardware-specific error codes
+ */
+static int hp_sw_error_is_retryable(struct request *req)
+{
+	/*
+	 * NOT_READY is known to be retryable
+	 * For now we just dump out the sense data and call it retryable
+	 */
+	if (status_byte(req->errors) == CHECK_CONDITION)
+		__scsi_print_sense(DM_HP_HWH_NAME, req->sense, req->sense_len);
+
+	/*
+	 * At this point we don't have complete information about all the error
+	 * codes from this hardware, so we are just conservative and retry
+	 * when in doubt.
+	 */
+	return 1;
+}
+
+/*
+ * hp_sw_end_io - Completion handler for HP path activation.
+ * @req: path activation request
+ * @error: scsi-ml error
+ *
+ *  Check sense data, free request structure, and notify dm that
+ *  pg initialization has completed.
+ *
+ * Context: scsi-ml softirq
+ *
+ */
+static void hp_sw_end_io(struct request *req, int error)
+{
+	struct dm_path *path = req->end_io_data;
+	unsigned err_flags = 0;
+
+	if (!error) {
+		DMDEBUG("%s path activation command - success",
+			path->dev->name);
+		goto out;
+	}
+
+	if (hp_sw_error_is_retryable(req)) {
+		DMDEBUG("%s path activation command - retry",
+			path->dev->name);
+		err_flags = MP_RETRY;
+		goto out;
+	}
+
+	DMWARN("%s path activation fail - error=0x%x",
+	       path->dev->name, error);
+	err_flags = MP_FAIL_PATH;
+
+out:
+	req->end_io_data = NULL;
+	__blk_put_request(req->q, req);
+	dm_pg_init_complete(path, err_flags);
+}
+
+/*
+ * hp_sw_get_request - Allocate an HP specific path activation request
+ * @path: path on which request will be sent (needed for request queue)
+ *
+ * The START command is used for path activation request.
+ * These arrays are controller-based failover, not LUN based.
+ * One START command issued to a single path will fail over all
+ * LUNs for the same controller.
+ *
+ * Possible optimizations
+ * 1. Make timeout configurable
+ * 2. Preallocate request
+ */
+static struct request *hp_sw_get_request(struct dm_path *path)
+{
+	struct request *req;
+	struct block_device *bdev = path->dev->bdev;
+	struct request_queue *q = bdev_get_queue(bdev);
+	struct hp_sw_context *h = path->hwhcontext;
+
+	req = blk_get_request(q, WRITE, GFP_NOIO);
+	if (!req)
+		goto out;
+
+	req->timeout = 60 * HZ;
+
+	req->errors = 0;
+	req->cmd_type = REQ_TYPE_BLOCK_PC;
+	req->cmd_flags |= REQ_FAILFAST | REQ_NOMERGE;
+	req->end_io_data = path;
+	req->sense = h->sense;
+	memset(req->sense, 0, SCSI_SENSE_BUFFERSIZE);
+
+	memset(&req->cmd, 0, BLK_MAX_CDB);
+	req->cmd[0] = START_STOP;
+	req->cmd[4] = 1;
+	req->cmd_len = COMMAND_SIZE(req->cmd[0]);
+
+out:
+	return req;
+}
+
+/*
+ * hp_sw_pg_init - HP path activation implementation.
+ * @hwh: hardware handler specific data
+ * @bypassed: unused; is the path group bypassed? (see dm-mpath.c)
+ * @path: path to send initialization command
+ *
+ * Send an HP-specific path activation command on 'path'.
+ * Do not try to optimize in any way, just send the activation command.
+ * More than one path activation command may be sent to the same controller.
+ * This seems to work fine for basic failover support.
+ *
+ * Possible optimizations
+ * 1. Detect an in-progress activation request and avoid submitting another one
+ * 2. Model the controller and only send a single activation request at a time
+ * 3. Determine the state of a path before sending an activation request
+ *
+ * Context: kmpathd (see process_queued_ios() in dm-mpath.c)
+ */
+static void hp_sw_pg_init(struct hw_handler *hwh, unsigned bypassed,
+			  struct dm_path *path)
+{
+	struct request *req;
+	struct hp_sw_context *h;
+
+	path->hwhcontext = hwh->context;
+	h = hwh->context;
+
+	req = hp_sw_get_request(path);
+	if (!req) {
+		DMERR("%s path activation command - allocation fail",
+		      path->dev->name);
+		goto retry;
+	}
+
+	DMDEBUG("%s path activation command - sent", path->dev->name);
+
+	blk_execute_rq_nowait(req->q, NULL, req, 1, hp_sw_end_io);
+	return;
+
+retry:
+	dm_pg_init_complete(path, MP_RETRY);
+}
+
+static int hp_sw_create(struct hw_handler *hwh, unsigned argc, char **argv)
+{
+	struct hp_sw_context *h;
+
+	h = kmalloc(sizeof(*h), GFP_KERNEL);
+	if (!h)
+		return -ENOMEM;
+
+	hwh->context = h;
+
+	return 0;
+}
+
+static void hp_sw_destroy(struct hw_handler *hwh)
+{
+	struct hp_sw_context *h = hwh->context;
+
+	kfree(h);
+}
+
+static struct hw_handler_type hp_sw_hwh = {
+	.name = DM_HP_HWH_NAME,
+	.module = THIS_MODULE,
+	.create = hp_sw_create,
+	.destroy = hp_sw_destroy,
+	.pg_init = hp_sw_pg_init,
+};
+
+static int __init hp_sw_init(void)
+{
+	int r;
+
+	r = dm_register_hw_handler(&hp_sw_hwh);
+	if (r < 0)
+		DMERR("register failed %d", r);
+	else
+		DMINFO("version " DM_HP_HWH_VER " loaded");
+
+	return r;
+}
+
+static void __exit hp_sw_exit(void)
+{
+	int r;
+
+	r = dm_unregister_hw_handler(&hp_sw_hwh);
+	if (r < 0)
+		DMERR("unregister failed %d", r);
+}
+
+module_init(hp_sw_init);
+module_exit(hp_sw_exit);
+
+MODULE_DESCRIPTION("DM Multipath HP StorageWorks / FSC FibreCat (A/P) support");
+MODULE_AUTHOR("Mike Christie, Dave Wysochanski <dm-devel@redhat.com>");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DM_HP_HWH_VER);

drivers/md/dm-mpath-rdac.c

@@ -664,20 +664,21 @@ static struct hw_handler_type rdac_handler = {
 
 static int __init rdac_init(void)
 {
-	int r = dm_register_hw_handler(&rdac_handler);
-
-	if (r < 0) {
-		DMERR("%s: register failed %d", RDAC_DM_HWH_NAME, r);
-		return r;
-	}
+	int r;
 
 	rdac_wkqd = create_singlethread_workqueue("rdac_wkqd");
 	if (!rdac_wkqd) {
 		DMERR("Failed to create workqueue rdac_wkqd.");
-		dm_unregister_hw_handler(&rdac_handler);
 		return -ENOMEM;
 	}
 
+	r = dm_register_hw_handler(&rdac_handler);
+	if (r < 0) {
+		DMERR("%s: register failed %d", RDAC_DM_HWH_NAME, r);
+		destroy_workqueue(rdac_wkqd);
+		return r;
+	}
+
 	DMINFO("%s: version %s loaded", RDAC_DM_HWH_NAME, RDAC_DM_HWH_VER);
 	return 0;
 }

drivers/md/dm-mpath.c

@@ -10,6 +10,7 @@
 #include "dm-hw-handler.h"
 #include "dm-bio-list.h"
 #include "dm-bio-record.h"
+#include "dm-uevent.h"
 
 #include <linux/ctype.h>
 #include <linux/init.h>
@@ -75,6 +76,8 @@ struct multipath {
 	unsigned queue_io;		/* Must we queue all I/O? */
 	unsigned queue_if_no_path;	/* Queue I/O if last path fails? */
 	unsigned saved_queue_if_no_path;/* Saved state during suspension */
+	unsigned pg_init_retries;	/* Number of times to retry pg_init */
+	unsigned pg_init_count;		/* Number of times pg_init called */
 
 	struct work_struct process_queued_ios;
 	struct bio_list queued_ios;
@@ -225,6 +228,8 @@ static void __switch_pg(struct multipath *m, struct pgpath *pgpath)
 		m->pg_init_required = 0;
 		m->queue_io = 0;
 	}
+
+	m->pg_init_count = 0;
 }
 
 static int __choose_path_in_pg(struct multipath *m, struct priority_group *pg)
@@ -424,6 +429,7 @@ static void process_queued_ios(struct work_struct *work)
 		must_queue = 0;
 
 	if (m->pg_init_required && !m->pg_init_in_progress) {
+		m->pg_init_count++;
 		m->pg_init_required = 0;
 		m->pg_init_in_progress = 1;
 		init_required = 1;
@@ -689,9 +695,11 @@ static int parse_features(struct arg_set *as, struct multipath *m)
 	int r;
 	unsigned argc;
 	struct dm_target *ti = m->ti;
+	const char *param_name;
 
 	static struct param _params[] = {
-		{0, 1, "invalid number of feature args"},
+		{0, 3, "invalid number of feature args"},
+		{1, 50, "pg_init_retries must be between 1 and 50"},
 	};
 
 	r = read_param(_params, shift(as), &argc, &ti->error);
@@ -701,12 +709,28 @@ static int parse_features(struct arg_set *as, struct multipath *m)
 	if (!argc)
 		return 0;
 
-	if (!strnicmp(shift(as), MESG_STR("queue_if_no_path")))
-		return queue_if_no_path(m, 1, 0);
-	else {
+	do {
+		param_name = shift(as);
+		argc--;
+
+		if (!strnicmp(param_name, MESG_STR("queue_if_no_path"))) {
+			r = queue_if_no_path(m, 1, 0);
+			continue;
+		}
+
+		if (!strnicmp(param_name, MESG_STR("pg_init_retries")) &&
+		    (argc >= 1)) {
+			r = read_param(_params + 1, shift(as),
+				       &m->pg_init_retries, &ti->error);
+			argc--;
+			continue;
+		}
+
 		ti->error = "Unrecognised multipath feature request";
-		return -EINVAL;
-	}
+		r = -EINVAL;
+	} while (argc && !r);
+
+	return r;
 }
 
 static int multipath_ctr(struct dm_target *ti, unsigned int argc,
@@ -834,6 +858,9 @@ static int fail_path(struct pgpath *pgpath)
 	if (pgpath == m->current_pgpath)
 		m->current_pgpath = NULL;
 
+	dm_path_uevent(DM_UEVENT_PATH_FAILED, m->ti,
+		      pgpath->path.dev->name, m->nr_valid_paths);
+
 	queue_work(kmultipathd, &m->trigger_event);
 
 out:
@@ -873,6 +900,9 @@ static int reinstate_path(struct pgpath *pgpath)
 	if (!m->nr_valid_paths++ && m->queue_size)
 		queue_work(kmultipathd, &m->process_queued_ios);
 
+	dm_path_uevent(DM_UEVENT_PATH_REINSTATED, m->ti,
+		      pgpath->path.dev->name, m->nr_valid_paths);
+
 	queue_work(kmultipathd, &m->trigger_event);
 
 out:
@@ -975,6 +1005,26 @@ static int bypass_pg_num(struct multipath *m, const char *pgstr, int bypassed)
 	return 0;
 }
 
+/*
+ * Should we retry pg_init immediately?
+ */
+static int pg_init_limit_reached(struct multipath *m, struct pgpath *pgpath)
+{
+	unsigned long flags;
+	int limit_reached = 0;
+
+	spin_lock_irqsave(&m->lock, flags);
+
+	if (m->pg_init_count <= m->pg_init_retries)
+		m->pg_init_required = 1;
+	else
+		limit_reached = 1;
+
+	spin_unlock_irqrestore(&m->lock, flags);
+
+	return limit_reached;
+}
+
 /*
  * pg_init must call this when it has completed its initialisation
  */
@@ -985,8 +1035,14 @@ void dm_pg_init_complete(struct dm_path *path, unsigned err_flags)
 	struct multipath *m = pg->m;
 	unsigned long flags;
 
-	/* We insist on failing the path if the PG is already bypassed. */
-	if (err_flags && pg->bypassed)
+	/*
+	 * If requested, retry pg_init until maximum number of retries exceeded.
+	 * If retry not requested and PG already bypassed, always fail the path.
+	 */
+	if (err_flags & MP_RETRY) {
+		if (pg_init_limit_reached(m, pgpath))
+			err_flags |= MP_FAIL_PATH;
+	} else if (err_flags && pg->bypassed)
 		err_flags |= MP_FAIL_PATH;
 
 	if (err_flags & MP_FAIL_PATH)
@@ -996,7 +1052,7 @@ void dm_pg_init_complete(struct dm_path *path, unsigned err_flags)
 		bypass_pg(m, pg, 1);
 
 	spin_lock_irqsave(&m->lock, flags);
-	if (err_flags) {
+	if (err_flags & ~MP_RETRY) {
 		m->current_pgpath = NULL;
 		m->current_pg = NULL;
 	} else if (!m->pg_init_required)
@@ -1148,11 +1204,15 @@ static int multipath_status(struct dm_target *ti, status_type_t type,
 
 	/* Features */
 	if (type == STATUSTYPE_INFO)
-		DMEMIT("1 %u ", m->queue_size);
-	else if (m->queue_if_no_path)
-		DMEMIT("1 queue_if_no_path ");
-	else
-		DMEMIT("0 ");
+		DMEMIT("2 %u %u ", m->queue_size, m->pg_init_count);
+	else {
+		DMEMIT("%u ", m->queue_if_no_path +
+			      (m->pg_init_retries > 0) * 2);
+		if (m->queue_if_no_path)
+			DMEMIT("queue_if_no_path ");
+		if (m->pg_init_retries)
+			DMEMIT("pg_init_retries %u ", m->pg_init_retries);
+	}
 
 	if (hwh->type && hwh->type->status)
 		sz += hwh->type->status(hwh, type, result + sz, maxlen - sz);

drivers/md/dm-path-selector.c

@@ -94,12 +94,10 @@ out:
 
 static struct ps_internal *_alloc_path_selector(struct path_selector_type *pst)
 {
-	struct ps_internal *psi = kmalloc(sizeof(*psi), GFP_KERNEL);
+	struct ps_internal *psi = kzalloc(sizeof(*psi), GFP_KERNEL);
 
-	if (psi) {
-		memset(psi, 0, sizeof(*psi));
+	if (psi)
 		psi->pst = *pst;
-	}
 
 	return psi;
 }

drivers/md/dm-raid1.c

@@ -19,6 +19,7 @@
 #include <linux/time.h>
 #include <linux/vmalloc.h>
 #include <linux/workqueue.h>
+#include <linux/log2.h>
 
 #define DM_MSG_PREFIX "raid1"
 #define DM_IO_PAGES 64
@@ -113,6 +114,7 @@ struct region {
  * Mirror set structures.
  *---------------------------------------------------------------*/
 struct mirror {
+	struct mirror_set *ms;
 	atomic_t error_count;
 	struct dm_dev *dev;
 	sector_t offset;
@@ -974,6 +976,7 @@ static struct mirror_set *alloc_context(unsigned int nr_mirrors,
 
 	if (rh_init(&ms->rh, ms, dl, region_size, ms->nr_regions)) {
 		ti->error = "Error creating dirty region hash";
+		dm_io_client_destroy(ms->io_client);
 		kfree(ms);
 		return NULL;
 	}
@@ -994,7 +997,7 @@ static void free_context(struct mirror_set *ms, struct dm_target *ti,
 
 static inline int _check_region_size(struct dm_target *ti, uint32_t size)
 {
-	return !(size % (PAGE_SIZE >> 9) || (size & (size - 1)) ||
+	return !(size % (PAGE_SIZE >> 9) || !is_power_of_2(size) ||
 		 size > ti->len);
 }
 
@@ -1015,6 +1018,7 @@ static int get_mirror(struct mirror_set *ms, struct dm_target *ti,
 		return -ENXIO;
 	}
 
+	ms->mirror[mirror].ms = ms;
 	ms->mirror[mirror].offset = offset;
 
 	return 0;
@@ -1163,16 +1167,14 @@ static int mirror_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	ms->kmirrord_wq = create_singlethread_workqueue("kmirrord");
 	if (!ms->kmirrord_wq) {
 		DMERR("couldn't start kmirrord");
-		free_context(ms, ti, m);
-		return -ENOMEM;
+		r = -ENOMEM;
+		goto err_free_context;
 	}
 	INIT_WORK(&ms->kmirrord_work, do_mirror);
 
 	r = parse_features(ms, argc, argv, &args_used);
-	if (r) {
-		free_context(ms, ti, ms->nr_mirrors);
-		return r;
-	}
+	if (r)
+		goto err_destroy_wq;
 
 	argv += args_used;
 	argc -= args_used;
@@ -1188,19 +1190,22 @@ static int mirror_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 
 	if (argc) {
 		ti->error = "Too many mirror arguments";
-		free_context(ms, ti, ms->nr_mirrors);
-		return -EINVAL;
+		r = -EINVAL;
+		goto err_destroy_wq;
 	}
 
 	r = kcopyd_client_create(DM_IO_PAGES, &ms->kcopyd_client);
-	if (r) {
-		destroy_workqueue(ms->kmirrord_wq);
-		free_context(ms, ti, ms->nr_mirrors);
-		return r;
-	}
+	if (r)
+		goto err_destroy_wq;
 
 	wake(ms);
 	return 0;
+
+err_destroy_wq:
+	destroy_workqueue(ms->kmirrord_wq);
+err_free_context:
+	free_context(ms, ti, ms->nr_mirrors);
+	return r;
 }
 
 static void mirror_dtr(struct dm_target *ti)
@@ -1302,7 +1307,7 @@ static void mirror_postsuspend(struct dm_target *ti)
 	wait_event(_kmirrord_recovery_stopped,
 		   !atomic_read(&ms->rh.recovery_in_flight));
 
-	if (log->type->suspend && log->type->suspend(log))
+	if (log->type->postsuspend && log->type->postsuspend(log))
 		/* FIXME: need better error handling */
 		DMWARN("log suspend failed");
 }

drivers/md/dm-snap.c

@@ -17,6 +17,7 @@
 #include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/vmalloc.h>
+#include <linux/log2.h>
 
 #include "dm-snap.h"
 #include "dm-bio-list.h"
@@ -415,7 +416,7 @@ static int set_chunk_size(struct dm_snapshot *s, const char *chunk_size_arg,
 	chunk_size = round_up(chunk_size, PAGE_SIZE >> 9);
 
 	/* Check chunk_size is a power of 2 */
-	if (chunk_size & (chunk_size - 1)) {
+	if (!is_power_of_2(chunk_size)) {
 		*error = "Chunk size is not a power of 2";
 		return -EINVAL;
 	}

drivers/md/dm-stripe.c

@@ -11,6 +11,7 @@
 #include <linux/blkdev.h>
 #include <linux/bio.h>
 #include <linux/slab.h>
+#include <linux/log2.h>
 
 #define DM_MSG_PREFIX "striped"
 
@@ -99,7 +100,7 @@ static int stripe_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	/*
 	 * chunk_size is a power of two
 	 */
-	if (!chunk_size || (chunk_size & (chunk_size - 1)) ||
+	if (!is_power_of_2(chunk_size) ||
 	    (chunk_size < (PAGE_SIZE >> SECTOR_SHIFT))) {
 		ti->error = "Invalid chunk size";
 		return -EINVAL;

drivers/md/dm-table.c

@@ -213,12 +213,11 @@ static int alloc_targets(struct dm_table *t, unsigned int num)
 int dm_table_create(struct dm_table **result, int mode,
 		    unsigned num_targets, struct mapped_device *md)
 {
-	struct dm_table *t = kmalloc(sizeof(*t), GFP_KERNEL);
+	struct dm_table *t = kzalloc(sizeof(*t), GFP_KERNEL);
 
 	if (!t)
 		return -ENOMEM;
 
-	memset(t, 0, sizeof(*t));
 	INIT_LIST_HEAD(&t->devices);
 	atomic_set(&t->holders, 1);
 

drivers/md/dm-target.c

@@ -88,12 +88,10 @@ void dm_put_target_type(struct target_type *t)
 
 static struct tt_internal *alloc_target(struct target_type *t)
 {
-	struct tt_internal *ti = kmalloc(sizeof(*ti), GFP_KERNEL);
+	struct tt_internal *ti = kzalloc(sizeof(*ti), GFP_KERNEL);
 
-	if (ti) {
-		memset(ti, 0, sizeof(*ti));
+	if (ti)
 		ti->tt = *t;
-	}
 
 	return ti;
 }

drivers/md/dm-uevent.c

@@ -0,0 +1,222 @@
+/*
+ * Device Mapper Uevent Support (dm-uevent)
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright IBM Corporation, 2007
+ * 	Author: Mike Anderson <andmike@linux.vnet.ibm.com>
+ */
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/kobject.h>
+#include <linux/dm-ioctl.h>
+
+#include "dm.h"
+#include "dm-uevent.h"
+
+#define DM_MSG_PREFIX "uevent"
+
+static const struct {
+	enum dm_uevent_type type;
+	enum kobject_action action;
+	char *name;
+} _dm_uevent_type_names[] = {
+	{DM_UEVENT_PATH_FAILED, KOBJ_CHANGE, "PATH_FAILED"},
+	{DM_UEVENT_PATH_REINSTATED, KOBJ_CHANGE, "PATH_REINSTATED"},
+};
+
+static struct kmem_cache *_dm_event_cache;
+
+struct dm_uevent {
+	struct mapped_device *md;
+	enum kobject_action action;
+	struct kobj_uevent_env ku_env;
+	struct list_head elist;
+	char name[DM_NAME_LEN];
+	char uuid[DM_UUID_LEN];
+};
+
+static void dm_uevent_free(struct dm_uevent *event)
+{
+	kmem_cache_free(_dm_event_cache, event);
+}
+
+static struct dm_uevent *dm_uevent_alloc(struct mapped_device *md)
+{
+	struct dm_uevent *event;
+
+	event = kmem_cache_zalloc(_dm_event_cache, GFP_ATOMIC);
+	if (!event)
+		return NULL;
+
+	INIT_LIST_HEAD(&event->elist);
+	event->md = md;
+
+	return event;
+}
+
+static struct dm_uevent *dm_build_path_uevent(struct mapped_device *md,
+					      struct dm_target *ti,
+					      enum kobject_action action,
+					      const char *dm_action,
+					      const char *path,
+					      unsigned nr_valid_paths)
+{
+	struct dm_uevent *event;
+
+	event = dm_uevent_alloc(md);
+	if (!event) {
+		DMERR("%s: dm_uevent_alloc() failed", __FUNCTION__);
+		goto err_nomem;
+	}
+
+	event->action = action;
+
+	if (add_uevent_var(&event->ku_env, "DM_TARGET=%s", ti->type->name)) {
+		DMERR("%s: add_uevent_var() for DM_TARGET failed",
+		      __FUNCTION__);
+		goto err_add;
+	}
+
+	if (add_uevent_var(&event->ku_env, "DM_ACTION=%s", dm_action)) {
+		DMERR("%s: add_uevent_var() for DM_ACTION failed",
+		      __FUNCTION__);
+		goto err_add;
+	}
+
+	if (add_uevent_var(&event->ku_env, "DM_SEQNUM=%u",
+			   dm_next_uevent_seq(md))) {
+		DMERR("%s: add_uevent_var() for DM_SEQNUM failed",
+		      __FUNCTION__);
+		goto err_add;
+	}
+
+	if (add_uevent_var(&event->ku_env, "DM_PATH=%s", path)) {
+		DMERR("%s: add_uevent_var() for DM_PATH failed", __FUNCTION__);
+		goto err_add;
+	}
+
+	if (add_uevent_var(&event->ku_env, "DM_NR_VALID_PATHS=%d",
+			   nr_valid_paths)) {
+		DMERR("%s: add_uevent_var() for DM_NR_VALID_PATHS failed",
+		      __FUNCTION__);
+		goto err_add;
+	}
+
+	return event;
+
+err_add:
+	dm_uevent_free(event);
+err_nomem:
+	return ERR_PTR(-ENOMEM);
+}
+
+/**
+ * dm_send_uevents - send uevents for given list
+ *
+ * @events:	list of events to send
+ * @kobj:	kobject generating event
+ *
+ */
+void dm_send_uevents(struct list_head *events, struct kobject *kobj)
+{
+	int r;
+	struct dm_uevent *event, *next;
+
+	list_for_each_entry_safe(event, next, events, elist) {
+		list_del_init(&event->elist);
+
+		/*
+		 * Need to call dm_copy_name_and_uuid from here for now.
+		 * Context of previous var adds and locking used for
+		 * hash_cell not compatable.
+		 */
+		if (dm_copy_name_and_uuid(event->md, event->name,
+					  event->uuid)) {
+			DMERR("%s: dm_copy_name_and_uuid() failed",
+			      __FUNCTION__);
+			goto uevent_free;
+		}
+
+		if (add_uevent_var(&event->ku_env, "DM_NAME=%s", event->name)) {
+			DMERR("%s: add_uevent_var() for DM_NAME failed",
+			      __FUNCTION__);
+			goto uevent_free;
+		}
+
+		if (add_uevent_var(&event->ku_env, "DM_UUID=%s", event->uuid)) {
+			DMERR("%s: add_uevent_var() for DM_UUID failed",
+			      __FUNCTION__);
+			goto uevent_free;
+		}
+
+		r = kobject_uevent_env(kobj, event->action, event->ku_env.envp);
+		if (r)
+			DMERR("%s: kobject_uevent_env failed", __FUNCTION__);
+uevent_free:
+		dm_uevent_free(event);
+	}
+}
+EXPORT_SYMBOL_GPL(dm_send_uevents);
+
+/**
+ * dm_path_uevent - called to create a new path event and queue it
+ *
+ * @event_type:	path event type enum
+ * @ti:			pointer to a dm_target
+ * @path:		string containing pathname
+ * @nr_valid_paths:	number of valid paths remaining
+ *
+ */
+void dm_path_uevent(enum dm_uevent_type event_type, struct dm_target *ti,
+		   const char *path, unsigned nr_valid_paths)
+{
+	struct mapped_device *md = dm_table_get_md(ti->table);
+	struct dm_uevent *event;
+
+	if (event_type >= ARRAY_SIZE(_dm_uevent_type_names)) {
+		DMERR("%s: Invalid event_type %d", __FUNCTION__, event_type);
+		goto out;
+	}
+
+	event = dm_build_path_uevent(md, ti,
+				     _dm_uevent_type_names[event_type].action,
+				     _dm_uevent_type_names[event_type].name,
+				     path, nr_valid_paths);
+	if (IS_ERR(event))
+		goto out;
+
+	dm_uevent_add(md, &event->elist);
+
+out:
+	dm_put(md);
+}
+EXPORT_SYMBOL_GPL(dm_path_uevent);
+
+int dm_uevent_init(void)
+{
+	_dm_event_cache = KMEM_CACHE(dm_uevent, 0);
+	if (!_dm_event_cache)
+		return -ENOMEM;
+
+	DMINFO("version 1.0.3");
+
+	return 0;
+}
+
+void dm_uevent_exit(void)
+{
+	kmem_cache_destroy(_dm_event_cache);
+}

drivers/md/dm-uevent.h

@@ -0,0 +1,59 @@
+/*
+ * Device Mapper Uevent Support
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright IBM Corporation, 2007
+ * 	Author: Mike Anderson <andmike@linux.vnet.ibm.com>
+ */
+#ifndef DM_UEVENT_H
+#define DM_UEVENT_H
+
+enum dm_uevent_type {
+	DM_UEVENT_PATH_FAILED,
+	DM_UEVENT_PATH_REINSTATED,
+};
+
+#ifdef CONFIG_DM_UEVENT
+
+extern int dm_uevent_init(void);
+extern void dm_uevent_exit(void);
+extern void dm_send_uevents(struct list_head *events, struct kobject *kobj);
+extern void dm_path_uevent(enum dm_uevent_type event_type,
+			   struct dm_target *ti, const char *path,
+			   unsigned nr_valid_paths);
+
+#else
+
+static inline int dm_uevent_init(void)
+{
+	return 0;
+}
+static inline void dm_uevent_exit(void)
+{
+}
+static inline void dm_send_uevents(struct list_head *events,
+				   struct kobject *kobj)
+{
+}
+static inline void dm_path_uevent(enum dm_uevent_type event_type,
+				  struct dm_target *ti, const char *path,
+				  unsigned nr_valid_paths)
+{
+}
+
+#endif	/* CONFIG_DM_UEVENT */
+
+#endif	/* DM_UEVENT_H */

drivers/md/dm.c

@@ -7,6 +7,7 @@
 
 #include "dm.h"
 #include "dm-bio-list.h"
+#include "dm-uevent.h"
 
 #include <linux/init.h>
 #include <linux/module.h>
@@ -112,6 +113,9 @@ struct mapped_device {
 	 */
 	atomic_t event_nr;
 	wait_queue_head_t eventq;
+	atomic_t uevent_seq;
+	struct list_head uevent_list;
+	spinlock_t uevent_lock; /* Protect access to uevent_list */
 
 	/*
 	 * freeze/thaw support require holding onto a super block
@@ -143,11 +147,19 @@ static int __init local_init(void)
 		return -ENOMEM;
 	}
 
+	r = dm_uevent_init();
+	if (r) {
+		kmem_cache_destroy(_tio_cache);
+		kmem_cache_destroy(_io_cache);
+		return r;
+	}
+
 	_major = major;
 	r = register_blkdev(_major, _name);
 	if (r < 0) {
 		kmem_cache_destroy(_tio_cache);
 		kmem_cache_destroy(_io_cache);
+		dm_uevent_exit();
 		return r;
 	}
 
@@ -162,6 +174,7 @@ static void local_exit(void)
 	kmem_cache_destroy(_tio_cache);
 	kmem_cache_destroy(_io_cache);
 	unregister_blkdev(_major, _name);
+	dm_uevent_exit();
 
 	_major = 0;
 
@@ -751,15 +764,13 @@ static void __clone_and_map(struct clone_info *ci)
 /*
  * Split the bio into several clones.
  */
-static void __split_bio(struct mapped_device *md, struct bio *bio)
+static int __split_bio(struct mapped_device *md, struct bio *bio)
 {
 	struct clone_info ci;
 
 	ci.map = dm_get_table(md);
-	if (!ci.map) {
-		bio_io_error(bio);
-		return;
-	}
+	if (unlikely(!ci.map))
+		return -EIO;
 
 	ci.md = md;
 	ci.bio = bio;
@@ -779,6 +790,8 @@ static void __split_bio(struct mapped_device *md, struct bio *bio)
 	/* drop the extra reference count */
 	dec_pending(ci.io, 0);
 	dm_table_put(ci.map);
+
+	return 0;
 }
 /*-----------------------------------------------------------------
  * CRUD END
@@ -790,7 +803,7 @@ static void __split_bio(struct mapped_device *md, struct bio *bio)
  */
 static int dm_request(struct request_queue *q, struct bio *bio)
 {
-	int r;
+	int r = -EIO;
 	int rw = bio_data_dir(bio);
 	struct mapped_device *md = q->queuedata;
 
@@ -815,18 +828,11 @@ static int dm_request(struct request_queue *q, struct bio *bio)
 	while (test_bit(DMF_BLOCK_IO, &md->flags)) {
 		up_read(&md->io_lock);
 
-		if (bio_rw(bio) == READA) {
-			bio_io_error(bio);
-			return 0;
-		}
-
-		r = queue_io(md, bio);
-		if (r < 0) {
-			bio_io_error(bio);
-			return 0;
+		if (bio_rw(bio) != READA)
+			r = queue_io(md, bio);
 
-		} else if (r == 0)
-			return 0;	/* deferred successfully */
+		if (r <= 0)
+			goto out_req;
 
 		/*
 		 * We're in a while loop, because someone could suspend
@@ -835,8 +841,13 @@ static int dm_request(struct request_queue *q, struct bio *bio)
 		down_read(&md->io_lock);
 	}
 
-	__split_bio(md, bio);
+	r = __split_bio(md, bio);
 	up_read(&md->io_lock);
+
+out_req:
+	if (r < 0)
+		bio_io_error(bio);
+
 	return 0;
 }
 
@@ -977,6 +988,9 @@ static struct mapped_device *alloc_dev(int minor)
 	atomic_set(&md->holders, 1);
 	atomic_set(&md->open_count, 0);
 	atomic_set(&md->event_nr, 0);
+	atomic_set(&md->uevent_seq, 0);
+	INIT_LIST_HEAD(&md->uevent_list);
+	spin_lock_init(&md->uevent_lock);
 
 	md->queue = blk_alloc_queue(GFP_KERNEL);
 	if (!md->queue)
@@ -1044,12 +1058,14 @@ static struct mapped_device *alloc_dev(int minor)
 	return NULL;
 }
 
+static void unlock_fs(struct mapped_device *md);
+
 static void free_dev(struct mapped_device *md)
 {
 	int minor = md->disk->first_minor;
 
 	if (md->suspended_bdev) {
-		thaw_bdev(md->suspended_bdev, NULL);
+		unlock_fs(md);
 		bdput(md->suspended_bdev);
 	}
 	mempool_destroy(md->tio_pool);
@@ -1073,8 +1089,16 @@ static void free_dev(struct mapped_device *md)
  */
 static void event_callback(void *context)
 {
+	unsigned long flags;
+	LIST_HEAD(uevents);
 	struct mapped_device *md = (struct mapped_device *) context;
 
+	spin_lock_irqsave(&md->uevent_lock, flags);
+	list_splice_init(&md->uevent_list, &uevents);
+	spin_unlock_irqrestore(&md->uevent_lock, flags);
+
+	dm_send_uevents(&uevents, &md->disk->kobj);
+
 	atomic_inc(&md->event_nr);
 	wake_up(&md->eventq);
 }
@@ -1233,7 +1257,8 @@ static void __flush_deferred_io(struct mapped_device *md, struct bio *c)
 	while (c) {
 		n = c->bi_next;
 		c->bi_next = NULL;
-		__split_bio(md, c);
+		if (__split_bio(md, c))
+			bio_io_error(c);
 		c = n;
 	}
 }
@@ -1491,6 +1516,11 @@ out:
 /*-----------------------------------------------------------------
  * Event notification.
  *---------------------------------------------------------------*/
+uint32_t dm_next_uevent_seq(struct mapped_device *md)
+{
+	return atomic_add_return(1, &md->uevent_seq);
+}
+
 uint32_t dm_get_event_nr(struct mapped_device *md)
 {
 	return atomic_read(&md->event_nr);
@@ -1502,6 +1532,15 @@ int dm_wait_event(struct mapped_device *md, int event_nr)
 			(event_nr != atomic_read(&md->event_nr)));
 }
 
+void dm_uevent_add(struct mapped_device *md, struct list_head *elist)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&md->uevent_lock, flags);
+	list_add(elist, &md->uevent_list);
+	spin_unlock_irqrestore(&md->uevent_lock, flags);
+}
+
 /*
  * The gendisk is only valid as long as you have a reference
  * count on 'md'.

drivers/md/kcopyd.c

@@ -198,7 +198,7 @@ struct kcopyd_job {
 	 * These fields are only used if the job has been split
 	 * into more manageable parts.
 	 */
-	struct semaphore lock;
+	struct mutex lock;
 	atomic_t sub_jobs;
 	sector_t progress;
 };
@@ -456,7 +456,7 @@ static void segment_complete(int read_err,
 	sector_t count = 0;
 	struct kcopyd_job *job = (struct kcopyd_job *) context;
 
-	down(&job->lock);
+	mutex_lock(&job->lock);
 
 	/* update the error */
 	if (read_err)
@@ -480,7 +480,7 @@ static void segment_complete(int read_err,
 			job->progress += count;
 		}
 	}
-	up(&job->lock);
+	mutex_unlock(&job->lock);
 
 	if (count) {
 		int i;
@@ -562,7 +562,7 @@ int kcopyd_copy(struct kcopyd_client *kc, struct io_region *from,
 		dispatch_job(job);
 
 	else {
-		init_MUTEX(&job->lock);
+		mutex_init(&job->lock);
 		job->progress = 0;
 		split_job(job);
 	}

include/linux/device-mapper.h

@@ -183,11 +183,14 @@ int dm_resume(struct mapped_device *md);
  */
 uint32_t dm_get_event_nr(struct mapped_device *md);
 int dm_wait_event(struct mapped_device *md, int event_nr);
+uint32_t dm_next_uevent_seq(struct mapped_device *md);
+void dm_uevent_add(struct mapped_device *md, struct list_head *elist);
 
 /*
  * Info functions.
  */
 const char *dm_device_name(struct mapped_device *md);
+int dm_copy_name_and_uuid(struct mapped_device *md, char *name, char *uuid);
 struct gendisk *dm_disk(struct mapped_device *md);
 int dm_suspended(struct mapped_device *md);
 int dm_noflush_suspending(struct dm_target *ti);

include/linux/dm-ioctl.h

@@ -131,6 +131,7 @@ struct dm_ioctl {
 	char name[DM_NAME_LEN];	/* device name */
 	char uuid[DM_UUID_LEN];	/* unique identifier for
 				 * the block device */
+	char data[7];		/* padding or data */
 };
 
 /*
@@ -285,9 +286,9 @@ typedef char ioctl_struct[308];
 #define DM_DEV_SET_GEOMETRY	_IOWR(DM_IOCTL, DM_DEV_SET_GEOMETRY_CMD, struct dm_ioctl)
 
 #define DM_VERSION_MAJOR	4
-#define DM_VERSION_MINOR	11
+#define DM_VERSION_MINOR	12
 #define DM_VERSION_PATCHLEVEL	0
-#define DM_VERSION_EXTRA	"-ioctl (2006-10-12)"
+#define DM_VERSION_EXTRA	"-ioctl (2007-10-02)"
 
 /* Status bits */
 #define DM_READONLY_FLAG	(1 << 0) /* In/Out */