Merge tag 'for-5.19/dm-changes' of...

Merge tag 'for-5.19/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm

Pull device mapper updates from Mike Snitzer:

 - Enable DM core bioset's per-cpu bio cache if QUEUE_FLAG_POLL set.
   This change improves DM's hipri bio polling (REQ_POLLED) performance
   by 7 - 20% depending on the system.

 - Update DM core to use jump_labels to further reduce cost of unlikely
   branches for zoned block devices, dm-stats and swap_bios throttling.

 - Various DM core changes to reduce bio-based DM overhead and simplify
   IO accounting.

 - Fundamental DM core improvements to dm_io reference counting and the
   elimination of using bio_split()+bio_chain() -- instead DM's
   bio-based IO accounting is updated to account that a split occurred.

 - Improve DM core's abnormal bio processing to do less work.

 - Improve DM core's hipri polling support to use a single list rather
   than an hlist.

 - Update DM core to pass NULL bdev to bio_alloc_clone() so that
   initialization that isn't useful for DM can be elided.

 - Add cond_resched to DM stats' various loops that loop over all
   entries.

 - Fix incorrect error code return from DM integrity's constructor.

 - Make DM crypt's printing of the key constant-time.

 - Update bio-based DM multipath to provide high-resolution timer to the
   Historical Service Time (HST) path selector.

* tag 'for-5.19/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm: (26 commits)
  dm: pass NULL bdev to bio_alloc_clone
  dm cache metadata: remove unnecessary variable in __dump_mapping
  dm mpath: provide high-resolution timer to HST for bio-based
  dm crypt: make printing of the key constant-time
  dm integrity: fix error code in dm_integrity_ctr()
  dm stats: add cond_resched when looping over entries
  dm: improve abnormal bio processing
  dm: simplify bio-based IO accounting further
  dm: put all polled dm_io instances into a single list
  dm: improve dm_io reference counting
  dm: don't grab target io reference in dm_zone_map_bio
  dm: improve bio splitting and associated IO accounting
  dm: switch to bdev based IO accounting interfaces
  dm: pass dm_io instance to dm_io_acct directly
  dm: don't pass bio to __dm_start_io_acct and dm_end_io_acct
  dm: use bio_sectors in dm_aceept_partial_bio
  dm: simplify basic targets
  dm: conditionally enable branching for less used features
  dm: introduce dm_{get,put}_live_table_bio called from dm_submit_bio
  dm: move hot dm_io members to same cacheline as dm_target_io
  ...

drivers/md/dm-cache-metadata.c

@@ -1509,7 +1509,6 @@ int dm_cache_load_mappings(struct dm_cache_metadata *cmd,
 
 static int __dump_mapping(void *context, uint64_t cblock, void *leaf)
 {
-	int r = 0;
 	__le64 value;
 	dm_oblock_t oblock;
 	unsigned flags;
@@ -1517,7 +1516,7 @@ static int __dump_mapping(void *context, uint64_t cblock, void *leaf)
 	memcpy(&value, leaf, sizeof(value));
 	unpack_value(value, &oblock, &flags);
 
-	return r;
+	return 0;
 }
 
 static int __dump_mappings(struct dm_cache_metadata *cmd)

drivers/md/dm-core.h

@@ -13,6 +13,7 @@
 #include <linux/ktime.h>
 #include <linux/blk-mq.h>
 #include <linux/blk-crypto-profile.h>
+#include <linux/jump_label.h>
 
 #include <trace/events/block.h>
 
@@ -154,6 +155,10 @@ static inline struct dm_stats *dm_get_stats(struct mapped_device *md)
 	return &md->stats;
 }
 
+DECLARE_STATIC_KEY_FALSE(stats_enabled);
+DECLARE_STATIC_KEY_FALSE(swap_bios_enabled);
+DECLARE_STATIC_KEY_FALSE(zoned_enabled);
+
 static inline bool dm_emulate_zone_append(struct mapped_device *md)
 {
 	if (blk_queue_is_zoned(md->queue))
@@ -237,6 +242,12 @@ static inline void dm_tio_set_flag(struct dm_target_io *tio, unsigned int bit)
 	tio->flags |= (1U << bit);
 }
 
+static inline bool dm_tio_is_normal(struct dm_target_io *tio)
+{
+	return (dm_tio_flagged(tio, DM_TIO_INSIDE_DM_IO) &&
+		!dm_tio_flagged(tio, DM_TIO_IS_DUPLICATE_BIO));
+}
+
 /*
  * One of these is allocated per original bio.
  * It contains the first clone used for that original.
@@ -245,16 +256,20 @@ static inline void dm_tio_set_flag(struct dm_target_io *tio, unsigned int bit)
 struct dm_io {
 	unsigned short magic;
 	blk_short_t flags;
-	atomic_t io_count;
-	struct mapped_device *md;
-	struct bio *orig_bio;
-	blk_status_t status;
 	spinlock_t lock;
 	unsigned long start_time;
 	void *data;
-	struct hlist_node node;
-	struct task_struct *map_task;
+	struct dm_io *next;
 	struct dm_stats_aux stats_aux;
+	blk_status_t status;
+	atomic_t io_count;
+	struct mapped_device *md;
+
+	/* The three fields represent mapped part of original bio */
+	struct bio *orig_bio;
+	unsigned int sector_offset; /* offset to end of orig_bio */
+	unsigned int sectors;
+
 	/* last member of dm_target_io is 'struct bio' */
 	struct dm_target_io tio;
 };
@@ -263,8 +278,8 @@ struct dm_io {
  * dm_io flags
  */
 enum {
-	DM_IO_START_ACCT,
-	DM_IO_ACCOUNTED
+	DM_IO_ACCOUNTED,
+	DM_IO_WAS_SPLIT
 };
 
 static inline bool dm_io_flagged(struct dm_io *io, unsigned int bit)
@@ -277,13 +292,6 @@ static inline void dm_io_set_flag(struct dm_io *io, unsigned int bit)
 	io->flags |= (1U << bit);
 }
 
-static inline void dm_io_inc_pending(struct dm_io *io)
-{
-	atomic_inc(&io->io_count);
-}
-
-void dm_io_dec_pending(struct dm_io *io, blk_status_t error);
-
 static inline struct completion *dm_get_completion_from_kobject(struct kobject *kobj)
 {
 	return &container_of(kobj, struct dm_kobject_holder, kobj)->completion;

drivers/md/dm-crypt.c

@@ -3439,6 +3439,11 @@ static int crypt_map(struct dm_target *ti, struct bio *bio)
 	return DM_MAPIO_SUBMITTED;
 }
 
+static char hex2asc(unsigned char c)
+{
+	return c + '0' + ((unsigned)(9 - c) >> 4 & 0x27);
+}
+
 static void crypt_status(struct dm_target *ti, status_type_t type,
 			 unsigned status_flags, char *result, unsigned maxlen)
 {
@@ -3457,9 +3462,12 @@ static void crypt_status(struct dm_target *ti, status_type_t type,
 		if (cc->key_size > 0) {
 			if (cc->key_string)
 				DMEMIT(":%u:%s", cc->key_size, cc->key_string);
-			else
-				for (i = 0; i < cc->key_size; i++)
-					DMEMIT("%02x", cc->key[i]);
+			else {
+				for (i = 0; i < cc->key_size; i++) {
+					DMEMIT("%c%c", hex2asc(cc->key[i] >> 4),
+					       hex2asc(cc->key[i] & 0xf));
+				}
+			}
 		} else
 			DMEMIT("-");
 

drivers/md/dm-delay.c

@@ -296,7 +296,6 @@ static int delay_map(struct dm_target *ti, struct bio *bio)
 	}
 	delayed->class = c;
 	bio_set_dev(bio, c->dev->bdev);
-	if (bio_sectors(bio))
 	bio->bi_iter.bi_sector = c->start + dm_target_offset(ti, bio->bi_iter.bi_sector);
 
 	return delay_bio(dc, c, bio);

drivers/md/dm-flakey.c

@@ -280,9 +280,7 @@ static void flakey_map_bio(struct dm_target *ti, struct bio *bio)
 	struct flakey_c *fc = ti->private;
 
 	bio_set_dev(bio, fc->dev->bdev);
-	if (bio_sectors(bio) || op_is_zone_mgmt(bio_op(bio)))
-		bio->bi_iter.bi_sector =
-			flakey_map_sector(ti, bio->bi_iter.bi_sector);
+	bio->bi_iter.bi_sector = flakey_map_sector(ti, bio->bi_iter.bi_sector);
 }
 
 static void corrupt_bio_data(struct bio *bio, struct flakey_c *fc)

drivers/md/dm-integrity.c

@@ -4494,8 +4494,6 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv)
 	}
 
 	if (should_write_sb) {
-		int r;
-
 		init_journal(ic, 0, ic->journal_sections, 0);
 		r = dm_integrity_failed(ic);
 		if (unlikely(r)) {

drivers/md/dm-linear.c

@@ -84,19 +84,12 @@ static sector_t linear_map_sector(struct dm_target *ti, sector_t bi_sector)
 	return lc->start + dm_target_offset(ti, bi_sector);
 }
 
-static void linear_map_bio(struct dm_target *ti, struct bio *bio)
+static int linear_map(struct dm_target *ti, struct bio *bio)
 {
 	struct linear_c *lc = ti->private;
 
 	bio_set_dev(bio, lc->dev->bdev);
-	if (bio_sectors(bio) || op_is_zone_mgmt(bio_op(bio)))
-		bio->bi_iter.bi_sector =
-			linear_map_sector(ti, bio->bi_iter.bi_sector);
-}
-
-static int linear_map(struct dm_target *ti, struct bio *bio)
-{
-	linear_map_bio(ti, bio);
+	bio->bi_iter.bi_sector = linear_map_sector(ti, bio->bi_iter.bi_sector);
 
 	return DM_MAPIO_REMAPPED;
 }

drivers/md/dm-mpath.c

@@ -105,6 +105,7 @@ struct multipath {
 struct dm_mpath_io {
 	struct pgpath *pgpath;
 	size_t nr_bytes;
+	u64 start_time_ns;
 };
 
 typedef int (*action_fn) (struct pgpath *pgpath);
@@ -295,6 +296,7 @@ static void multipath_init_per_bio_data(struct bio *bio, struct dm_mpath_io **mp
 
 	mpio->nr_bytes = bio->bi_iter.bi_size;
 	mpio->pgpath = NULL;
+	mpio->start_time_ns = 0;
 	*mpio_p = mpio;
 
 	dm_bio_record(bio_details, bio);
@@ -647,6 +649,9 @@ static int __multipath_map_bio(struct multipath *m, struct bio *bio,
 
 	mpio->pgpath = pgpath;
 
+	if (dm_ps_use_hr_timer(pgpath->pg->ps.type))
+		mpio->start_time_ns = ktime_get_ns();
+
 	bio->bi_status = 0;
 	bio_set_dev(bio, pgpath->path.dev->bdev);
 	bio->bi_opf |= REQ_FAILFAST_TRANSPORT;
@@ -1713,7 +1718,8 @@ static int multipath_end_io_bio(struct dm_target *ti, struct bio *clone,
 
 		if (ps->type->end_io)
 			ps->type->end_io(ps, &pgpath->path, mpio->nr_bytes,
-					 dm_start_time_ns_from_clone(clone));
+					 (mpio->start_time_ns ?:
+					  dm_start_time_ns_from_clone(clone)));
 	}
 
 	return r;

drivers/md/dm-path-selector.h

@@ -26,11 +26,26 @@ struct path_selector {
 	void *context;
 };
 
+/*
+ * If a path selector uses this flag, a high resolution timer is used
+ * (via ktime_get_ns) to account for IO start time in BIO-based mpath.
+ * This improves performance of some path selectors (i.e. HST), in
+ * exchange for slightly higher overhead when submitting the BIO.
+ * The extra cost is usually offset by improved path selection for
+ * some benchmarks.
+ *
+ * This has no effect for request-based mpath, since it already uses a
+ * higher precision timer by default.
+ */
+#define DM_PS_USE_HR_TIMER		0x00000001
+#define dm_ps_use_hr_timer(type)	((type)->features & DM_PS_USE_HR_TIMER)
+
 /* Information about a path selector type */
 struct path_selector_type {
 	char *name;
 	struct module *module;
 
+	unsigned int features;
 	unsigned int table_args;
 	unsigned int info_args;
 

drivers/md/dm-ps-historical-service-time.c

@@ -523,6 +523,7 @@ static int hst_end_io(struct path_selector *ps, struct dm_path *path,
 static struct path_selector_type hst_ps = {
 	.name		= "historical-service-time",
 	.module		= THIS_MODULE,
+	.features	= DM_PS_USE_HR_TIMER,
 	.table_args	= 1,
 	.info_args	= 3,
 	.create		= hst_create,

drivers/md/dm-stats.c

@@ -225,6 +225,7 @@ void dm_stats_cleanup(struct dm_stats *stats)
 				       atomic_read(&shared->in_flight[READ]),
 				       atomic_read(&shared->in_flight[WRITE]));
 			}
+			cond_resched();
 		}
 		dm_stat_free(&s->rcu_head);
 	}
@@ -330,6 +331,7 @@ static int dm_stats_create(struct dm_stats *stats, sector_t start, sector_t end,
 	for (ni = 0; ni < n_entries; ni++) {
 		atomic_set(&s->stat_shared[ni].in_flight[READ], 0);
 		atomic_set(&s->stat_shared[ni].in_flight[WRITE], 0);
+		cond_resched();
 	}
 
 	if (s->n_histogram_entries) {
@@ -342,6 +344,7 @@ static int dm_stats_create(struct dm_stats *stats, sector_t start, sector_t end,
 		for (ni = 0; ni < n_entries; ni++) {
 			s->stat_shared[ni].tmp.histogram = hi;
 			hi += s->n_histogram_entries + 1;
+			cond_resched();
 		}
 	}
 
@@ -362,6 +365,7 @@ static int dm_stats_create(struct dm_stats *stats, sector_t start, sector_t end,
 			for (ni = 0; ni < n_entries; ni++) {
 				p[ni].histogram = hi;
 				hi += s->n_histogram_entries + 1;
+				cond_resched();
 			}
 		}
 	}
@@ -396,6 +400,9 @@ static int dm_stats_create(struct dm_stats *stats, sector_t start, sector_t end,
 
 	dm_stats_recalc_precise_timestamps(stats);
 
+	if (!static_key_enabled(&stats_enabled.key))
+		static_branch_enable(&stats_enabled);
+
 	mutex_unlock(&stats->mutex);
 
 	resume_callback(md);
@@ -497,6 +504,7 @@ static int dm_stats_list(struct dm_stats *stats, const char *program,
 			}
 			DMEMIT("\n");
 		}
+		cond_resched();
 	}
 	mutex_unlock(&stats->mutex);
 
@@ -774,6 +782,7 @@ static void __dm_stat_clear(struct dm_stat *s, size_t idx_start, size_t idx_end,
 				local_irq_enable();
 			}
 		}
+		cond_resched();
 	}
 }
 
@@ -889,6 +898,8 @@ static int dm_stats_print(struct dm_stats *stats, int id,
 
 		if (unlikely(sz + 1 >= maxlen))
 			goto buffer_overflow;
+
+		cond_resched();
 	}
 
 	if (clear)

drivers/md/dm-table.c

@@ -719,6 +719,9 @@ int dm_table_add_target(struct dm_table *t, const char *type,
 		DMWARN("%s: %s: ignoring discards_supported because num_discard_bios is zero.",
 		       dm_device_name(t->md), type);
 
+	if (tgt->limit_swap_bios && !static_key_enabled(&swap_bios_enabled.key))
+		static_branch_enable(&swap_bios_enabled);
+
 	return 0;
 
  bad:
@@ -1002,6 +1005,8 @@ bool dm_table_request_based(struct dm_table *t)
 	return __table_type_request_based(dm_table_get_type(t));
 }
 
+static int dm_table_supports_poll(struct dm_table *t);
+
 static int dm_table_alloc_md_mempools(struct dm_table *t, struct mapped_device *md)
 {
 	enum dm_queue_mode type = dm_table_get_type(t);
@@ -1009,21 +1014,24 @@ static int dm_table_alloc_md_mempools(struct dm_table *t, struct mapped_device *
 	unsigned min_pool_size = 0;
 	struct dm_target *ti;
 	unsigned i;
+	bool poll_supported = false;
 
 	if (unlikely(type == DM_TYPE_NONE)) {
 		DMWARN("no table type is set, can't allocate mempools");
 		return -EINVAL;
 	}
 
-	if (__table_type_bio_based(type))
+	if (__table_type_bio_based(type)) {
 		for (i = 0; i < t->num_targets; i++) {
 			ti = t->targets + i;
 			per_io_data_size = max(per_io_data_size, ti->per_io_data_size);
 			min_pool_size = max(min_pool_size, ti->num_flush_bios);
 		}
+		poll_supported = !!dm_table_supports_poll(t);
+	}
 
-	t->mempools = dm_alloc_md_mempools(md, type, t->integrity_supported,
-					   per_io_data_size, min_pool_size);
+	t->mempools = dm_alloc_md_mempools(md, type, per_io_data_size, min_pool_size,
+					   t->integrity_supported, poll_supported);
 	if (!t->mempools)
 		return -ENOMEM;
 
@@ -2035,6 +2043,8 @@ int dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 		r = dm_set_zones_restrictions(t, q);
 		if (r)
 			return r;
+		if (!static_key_enabled(&zoned_enabled.key))
+			static_branch_enable(&zoned_enabled);
 	}
 
 	dm_update_crypto_profile(q, t);

drivers/md/dm-zone.c

@@ -550,13 +550,6 @@ int dm_zone_map_bio(struct dm_target_io *tio)
 		return DM_MAPIO_KILL;
 	}
 
-	/*
-	 * The target map function may issue and complete the IO quickly.
-	 * Take an extra reference on the IO to make sure it does disappear
-	 * until we run dm_zone_map_bio_end().
-	 */
-	dm_io_inc_pending(io);
-
 	/* Let the target do its work */
 	r = ti->type->map(ti, clone);
 	switch (r) {
@@ -587,9 +580,6 @@ int dm_zone_map_bio(struct dm_target_io *tio)
 		break;
 	}
 
-	/* Drop the extra reference on the IO */
-	dm_io_dec_pending(io, sts);
-
 	if (sts != BLK_STS_OK)
 		return DM_MAPIO_KILL;
 

drivers/md/dm.c

@@ -71,6 +71,10 @@ void dm_issue_global_event(void)
 	wake_up(&dm_global_eventq);
 }
 
+DEFINE_STATIC_KEY_FALSE(stats_enabled);
+DEFINE_STATIC_KEY_FALSE(swap_bios_enabled);
+DEFINE_STATIC_KEY_FALSE(zoned_enabled);
+
 /*
  * One of these is allocated (on-stack) per original bio.
  */
@@ -80,7 +84,8 @@ struct clone_info {
 	struct dm_io *io;
 	sector_t sector;
 	unsigned sector_count;
-	bool submit_as_polled;
+	bool is_abnormal_io:1;
+	bool submit_as_polled:1;
 };
 
 #define DM_TARGET_IO_BIO_OFFSET (offsetof(struct dm_target_io, clone))
@@ -498,69 +503,76 @@ static bool bio_is_flush_with_data(struct bio *bio)
 	return ((bio->bi_opf & REQ_PREFLUSH) && bio->bi_iter.bi_size);
 }
 
-static void dm_io_acct(bool end, struct mapped_device *md, struct bio *bio,
-		       unsigned long start_time, struct dm_stats_aux *stats_aux)
+static void dm_io_acct(struct dm_io *io, bool end)
 {
-	bool is_flush_with_data;
-	unsigned int bi_size;
+	struct dm_stats_aux *stats_aux = &io->stats_aux;
+	unsigned long start_time = io->start_time;
+	struct mapped_device *md = io->md;
+	struct bio *bio = io->orig_bio;
+	unsigned int sectors;
 
-	/* If REQ_PREFLUSH set save any payload but do not account it */
-	is_flush_with_data = bio_is_flush_with_data(bio);
-	if (is_flush_with_data) {
-		bi_size = bio->bi_iter.bi_size;
-		bio->bi_iter.bi_size = 0;
-	}
+	/*
+	 * If REQ_PREFLUSH set, don't account payload, it will be
+	 * submitted (and accounted) after this flush completes.
+	 */
+	if (bio_is_flush_with_data(bio))
+		sectors = 0;
+	else if (likely(!(dm_io_flagged(io, DM_IO_WAS_SPLIT))))
+		sectors = bio_sectors(bio);
+	else
+		sectors = io->sectors;
 
 	if (!end)
-		bio_start_io_acct_time(bio, start_time);
+		bdev_start_io_acct(bio->bi_bdev, sectors, bio_op(bio),
+				   start_time);
 	else
-		bio_end_io_acct(bio, start_time);
+		bdev_end_io_acct(bio->bi_bdev, bio_op(bio), start_time);
+
+	if (static_branch_unlikely(&stats_enabled) &&
+	    unlikely(dm_stats_used(&md->stats))) {
+		sector_t sector;
+
+		if (likely(!dm_io_flagged(io, DM_IO_WAS_SPLIT)))
+			sector = bio->bi_iter.bi_sector;
+		else
+			sector = bio_end_sector(bio) - io->sector_offset;
 
-	if (unlikely(dm_stats_used(&md->stats)))
 		dm_stats_account_io(&md->stats, bio_data_dir(bio),
-				    bio->bi_iter.bi_sector, bio_sectors(bio),
+				    sector, sectors,
 				    end, start_time, stats_aux);
-
-	/* Restore bio's payload so it does get accounted upon requeue */
-	if (is_flush_with_data)
-		bio->bi_iter.bi_size = bi_size;
+	}
 }
 
-static void __dm_start_io_acct(struct dm_io *io, struct bio *bio)
+static void __dm_start_io_acct(struct dm_io *io)
 {
-	dm_io_acct(false, io->md, bio, io->start_time, &io->stats_aux);
+	dm_io_acct(io, false);
 }
 
 static void dm_start_io_acct(struct dm_io *io, struct bio *clone)
 {
-	/* Must account IO to DM device in terms of orig_bio */
-	struct bio *bio = io->orig_bio;
-
 	/*
 	 * Ensure IO accounting is only ever started once.
-	 * Expect no possibility for race unless DM_TIO_IS_DUPLICATE_BIO.
 	 */
-	if (!clone ||
-	    likely(!dm_tio_flagged(clone_to_tio(clone), DM_TIO_IS_DUPLICATE_BIO))) {
-		if (WARN_ON_ONCE(dm_io_flagged(io, DM_IO_ACCOUNTED)))
+	if (dm_io_flagged(io, DM_IO_ACCOUNTED))
 		return;
+
+	/* Expect no possibility for race unless DM_TIO_IS_DUPLICATE_BIO. */
+	if (!clone || likely(dm_tio_is_normal(clone_to_tio(clone)))) {
 		dm_io_set_flag(io, DM_IO_ACCOUNTED);
 	} else {
 		unsigned long flags;
-		if (dm_io_flagged(io, DM_IO_ACCOUNTED))
-			return;
 		/* Can afford locking given DM_TIO_IS_DUPLICATE_BIO */
 		spin_lock_irqsave(&io->lock, flags);
 		dm_io_set_flag(io, DM_IO_ACCOUNTED);
 		spin_unlock_irqrestore(&io->lock, flags);
 	}
 
-	__dm_start_io_acct(io, bio);
+	__dm_start_io_acct(io);
 }
 
-static void dm_end_io_acct(struct dm_io *io, struct bio *bio)
+static void dm_end_io_acct(struct dm_io *io)
 {
-	dm_io_acct(true, io->md, bio, io->start_time, &io->stats_aux);
+	dm_io_acct(io, true);
 }
 
 static struct dm_io *alloc_io(struct mapped_device *md, struct bio *bio)
@@ -569,7 +581,9 @@ static struct dm_io *alloc_io(struct mapped_device *md, struct bio *bio)
 	struct dm_target_io *tio;
 	struct bio *clone;
 
-	clone = bio_alloc_clone(bio->bi_bdev, bio, GFP_NOIO, &md->io_bs);
+	clone = bio_alloc_clone(NULL, bio, GFP_NOIO, &md->io_bs);
+	/* Set default bdev, but target must bio_set_dev() before issuing IO */
+	clone->bi_bdev = md->disk->part0;
 
 	tio = clone_to_tio(clone);
 	tio->flags = 0;
@@ -578,16 +592,18 @@ static struct dm_io *alloc_io(struct mapped_device *md, struct bio *bio)
 
 	io = container_of(tio, struct dm_io, tio);
 	io->magic = DM_IO_MAGIC;
-	io->status = 0;
-	atomic_set(&io->io_count, 1);
+	io->status = BLK_STS_OK;
+
+	/* one ref is for submission, the other is for completion */
+	atomic_set(&io->io_count, 2);
 	this_cpu_inc(*md->pending_io);
-	io->orig_bio = NULL;
+	io->orig_bio = bio;
 	io->md = md;
-	io->map_task = current;
 	spin_lock_init(&io->lock);
 	io->start_time = jiffies;
 	io->flags = 0;
 
+	if (static_branch_unlikely(&stats_enabled))
 		dm_stats_record_start(&md->stats, &io->stats_aux);
 
 	return io;
@@ -610,10 +626,13 @@ static struct bio *alloc_tio(struct clone_info *ci, struct dm_target *ti,
 		/* alloc_io() already initialized embedded clone */
 		clone = &tio->clone;
 	} else {
-		clone = bio_alloc_clone(ci->bio->bi_bdev, ci->bio,
-					gfp_mask, &ci->io->md->bs);
+		struct mapped_device *md = ci->io->md;
+
+		clone = bio_alloc_clone(NULL, ci->bio, gfp_mask, &md->bs);
 		if (!clone)
 			return NULL;
+		/* Set default bdev, but target must bio_set_dev() before issuing IO */
+		clone->bi_bdev = md->disk->part0;
 
 		/* REQ_DM_POLL_LIST shouldn't be inherited */
 		clone->bi_opf &= ~REQ_DM_POLL_LIST;
@@ -663,14 +682,16 @@ static void queue_io(struct mapped_device *md, struct bio *bio)
  * function to access the md->map field, and make sure they call
  * dm_put_live_table() when finished.
  */
-struct dm_table *dm_get_live_table(struct mapped_device *md, int *srcu_idx) __acquires(md->io_barrier)
+struct dm_table *dm_get_live_table(struct mapped_device *md,
+				   int *srcu_idx) __acquires(md->io_barrier)
 {
 	*srcu_idx = srcu_read_lock(&md->io_barrier);
 
 	return srcu_dereference(md->map, &md->io_barrier);
 }
 
-void dm_put_live_table(struct mapped_device *md, int srcu_idx) __releases(md->io_barrier)
+void dm_put_live_table(struct mapped_device *md,
+		       int srcu_idx) __releases(md->io_barrier)
 {
 	srcu_read_unlock(&md->io_barrier, srcu_idx);
 }
@@ -696,6 +717,24 @@ static void dm_put_live_table_fast(struct mapped_device *md) __releases(RCU)
 	rcu_read_unlock();
 }
 
+static inline struct dm_table *dm_get_live_table_bio(struct mapped_device *md,
+						     int *srcu_idx, struct bio *bio)
+{
+	if (bio->bi_opf & REQ_NOWAIT)
+		return dm_get_live_table_fast(md);
+	else
+		return dm_get_live_table(md, srcu_idx);
+}
+
+static inline void dm_put_live_table_bio(struct mapped_device *md, int srcu_idx,
+					 struct bio *bio)
+{
+	if (bio->bi_opf & REQ_NOWAIT)
+		dm_put_live_table_fast(md);
+	else
+		dm_put_live_table(md, srcu_idx);
+}
+
 static char *_dm_claim_ptr = "I belong to device-mapper";
 
 /*
@@ -875,14 +914,14 @@ static void dm_io_complete(struct dm_io *io)
 
 	io_error = io->status;
 	if (dm_io_flagged(io, DM_IO_ACCOUNTED))
-		dm_end_io_acct(io, bio);
+		dm_end_io_acct(io);
 	else if (!io_error) {
 		/*
 		 * Must handle target that DM_MAPIO_SUBMITTED only to
 		 * then bio_endio() rather than dm_submit_bio_remap()
 		 */
-		__dm_start_io_acct(io, bio);
-		dm_end_io_acct(io, bio);
+		__dm_start_io_acct(io);
+		dm_end_io_acct(io);
 	}
 	free_io(io);
 	smp_wmb();
@@ -899,7 +938,7 @@ static void dm_io_complete(struct dm_io *io)
 			 * may only reflect a subset of the pre-split original)
 			 * so clear REQ_POLLED in case of requeue.
 			 */
-			bio->bi_opf &= ~REQ_POLLED;
+			bio_clear_polled(bio);
 			if (io_error == BLK_STS_AGAIN) {
 				/* io_uring doesn't handle BLK_STS_AGAIN (yet) */
 				queue_io(md, bio);
@@ -923,30 +962,35 @@ static void dm_io_complete(struct dm_io *io)
 	}
 }
 
-static inline bool dm_tio_is_normal(struct dm_target_io *tio)
-{
-	return (dm_tio_flagged(tio, DM_TIO_INSIDE_DM_IO) &&
-		!dm_tio_flagged(tio, DM_TIO_IS_DUPLICATE_BIO));
-}
-
 /*
  * Decrements the number of outstanding ios that a bio has been
  * cloned into, completing the original io if necc.
  */
-void dm_io_dec_pending(struct dm_io *io, blk_status_t error)
+static inline void __dm_io_dec_pending(struct dm_io *io)
+{
+	if (atomic_dec_and_test(&io->io_count))
+		dm_io_complete(io);
+}
+
+static void dm_io_set_error(struct dm_io *io, blk_status_t error)
 {
-	/* Push-back supersedes any I/O errors */
-	if (unlikely(error)) {
 	unsigned long flags;
+
+	/* Push-back supersedes any I/O errors */
 	spin_lock_irqsave(&io->lock, flags);
 	if (!(io->status == BLK_STS_DM_REQUEUE &&
-		      __noflush_suspending(io->md)))
+	      __noflush_suspending(io->md))) {
 		io->status = error;
-		spin_unlock_irqrestore(&io->lock, flags);
 	}
+	spin_unlock_irqrestore(&io->lock, flags);
+}
 
-	if (atomic_dec_and_test(&io->io_count))
-		dm_io_complete(io);
+static void dm_io_dec_pending(struct dm_io *io, blk_status_t error)
+{
+	if (unlikely(error))
+		dm_io_set_error(io, error);
+
+	__dm_io_dec_pending(io);
 }
 
 void disable_discard(struct mapped_device *md)
@@ -974,10 +1018,13 @@ static void clone_endio(struct bio *bio)
 {
 	blk_status_t error = bio->bi_status;
 	struct dm_target_io *tio = clone_to_tio(bio);
+	struct dm_target *ti = tio->ti;
+	dm_endio_fn endio = ti->type->end_io;
 	struct dm_io *io = tio->io;
-	struct mapped_device *md = tio->io->md;
-	dm_endio_fn endio = tio->ti->type->end_io;
-	struct request_queue *q = bio->bi_bdev->bd_disk->queue;
+	struct mapped_device *md = io->md;
+
+	if (likely(bio->bi_bdev != md->disk->part0)) {
+		struct request_queue *q = bdev_get_queue(bio->bi_bdev);
 
 		if (unlikely(error == BLK_STS_TARGET)) {
 			if (bio_op(bio) == REQ_OP_DISCARD &&
@@ -988,13 +1035,16 @@ static void clone_endio(struct bio *bio)
 				disable_write_zeroes(md);
 		}
 
-	if (blk_queue_is_zoned(q))
+		if (static_branch_unlikely(&zoned_enabled) &&
+		    unlikely(blk_queue_is_zoned(q)))
 			dm_zone_endio(io, bio);
+	}
 
 	if (endio) {
-		int r = endio(tio->ti, bio, &error);
+		int r = endio(ti, bio, &error);
 		switch (r) {
 		case DM_ENDIO_REQUEUE:
+			if (static_branch_unlikely(&zoned_enabled)) {
 				/*
 				 * Requeuing writes to a sequential zone of a zoned
 				 * target will break the sequential write pattern:
@@ -1004,6 +1054,8 @@ static void clone_endio(struct bio *bio)
 					error = BLK_STS_IOERR;
 				else
 					error = BLK_STS_DM_REQUEUE;
+			} else
+				error = BLK_STS_DM_REQUEUE;
 			fallthrough;
 		case DM_ENDIO_DONE:
 			break;
@@ -1016,10 +1068,9 @@ static void clone_endio(struct bio *bio)
 		}
 	}
 
-	if (unlikely(swap_bios_limit(tio->ti, bio))) {
-		struct mapped_device *md = io->md;
+	if (static_branch_unlikely(&swap_bios_enabled) &&
+	    unlikely(swap_bios_limit(ti, bio)))
 		up(&md->swap_bios_semaphore);
-	}
 
 	free_tio(bio);
 	dm_io_dec_pending(io, error);
@@ -1161,7 +1212,7 @@ static int dm_dax_zero_page_range(struct dax_device *dax_dev, pgoff_t pgoff,
  * +--------------------+---------------+-------+
  *
  * <-------------- *tio->len_ptr --------------->
- *                      <------- bi_size ------->
+ *                      <----- bio_sectors ----->
  *                      <-- n_sectors -->
  *
  * Region 1 was already iterated over with bio_advance or similar function.
@@ -1178,25 +1229,25 @@ static int dm_dax_zero_page_range(struct dax_device *dax_dev, pgoff_t pgoff,
 void dm_accept_partial_bio(struct bio *bio, unsigned n_sectors)
 {
 	struct dm_target_io *tio = clone_to_tio(bio);
-	unsigned bi_size = bio->bi_iter.bi_size >> SECTOR_SHIFT;
+	unsigned bio_sectors = bio_sectors(bio);
 
 	BUG_ON(dm_tio_flagged(tio, DM_TIO_IS_DUPLICATE_BIO));
 	BUG_ON(op_is_zone_mgmt(bio_op(bio)));
 	BUG_ON(bio_op(bio) == REQ_OP_ZONE_APPEND);
-	BUG_ON(bi_size > *tio->len_ptr);
-	BUG_ON(n_sectors > bi_size);
+	BUG_ON(bio_sectors > *tio->len_ptr);
+	BUG_ON(n_sectors > bio_sectors);
 
-	*tio->len_ptr -= bi_size - n_sectors;
+	*tio->len_ptr -= bio_sectors - n_sectors;
 	bio->bi_iter.bi_size = n_sectors << SECTOR_SHIFT;
-}
-EXPORT_SYMBOL_GPL(dm_accept_partial_bio);
 
-static inline void __dm_submit_bio_remap(struct bio *clone,
-					 dev_t dev, sector_t old_sector)
-{
-	trace_block_bio_remap(clone, dev, old_sector);
-	submit_bio_noacct(clone);
+	/*
+	 * __split_and_process_bio() may have already saved mapped part
+	 * for accounting but it is being reduced so update accordingly.
+	 */
+	dm_io_set_flag(tio->io, DM_IO_WAS_SPLIT);
+	tio->io->sectors = n_sectors;
 }
+EXPORT_SYMBOL_GPL(dm_accept_partial_bio);
 
 /*
  * @clone: clone bio that DM core passed to target's .map function
@@ -1212,8 +1263,6 @@ void dm_submit_bio_remap(struct bio *clone, struct bio *tgt_clone)
 	struct dm_target_io *tio = clone_to_tio(clone);
 	struct dm_io *io = tio->io;
 
-	WARN_ON_ONCE(!tio->ti->accounts_remapped_io);
-
 	/* establish bio that will get submitted */
 	if (!tgt_clone)
 		tgt_clone = clone;
@@ -1222,22 +1271,11 @@ void dm_submit_bio_remap(struct bio *clone, struct bio *tgt_clone)
 	 * Account io->origin_bio to DM dev on behalf of target
 	 * that took ownership of IO with DM_MAPIO_SUBMITTED.
 	 */
-	if (io->map_task == current) {
-		/* Still in target's map function */
-		dm_io_set_flag(io, DM_IO_START_ACCT);
-	} else {
-		/*
-		 * Called by another thread, managed by DM target,
-		 * wait for dm_split_and_process_bio() to store
-		 * io->orig_bio
-		 */
-		while (unlikely(!smp_load_acquire(&io->orig_bio)))
-			msleep(1);
 	dm_start_io_acct(io, clone);
-	}
 
-	__dm_submit_bio_remap(tgt_clone, disk_devt(io->md->disk),
+	trace_block_bio_remap(tgt_clone, disk_devt(io->md->disk),
 			      tio->old_sector);
+	submit_bio_noacct(tgt_clone);
 }
 EXPORT_SYMBOL_GPL(dm_submit_bio_remap);
 
@@ -1260,55 +1298,53 @@ static noinline void __set_swap_bios_limit(struct mapped_device *md, int latch)
 static void __map_bio(struct bio *clone)
 {
 	struct dm_target_io *tio = clone_to_tio(clone);
-	int r;
-	struct dm_io *io = tio->io;
 	struct dm_target *ti = tio->ti;
+	struct dm_io *io = tio->io;
+	struct mapped_device *md = io->md;
+	int r;
 
 	clone->bi_end_io = clone_endio;
 
 	/*
 	 * Map the clone.
 	 */
-	dm_io_inc_pending(io);
 	tio->old_sector = clone->bi_iter.bi_sector;
 
-	if (unlikely(swap_bios_limit(ti, clone))) {
-		struct mapped_device *md = io->md;
+	if (static_branch_unlikely(&swap_bios_enabled) &&
+	    unlikely(swap_bios_limit(ti, clone))) {
 		int latch = get_swap_bios();
 		if (unlikely(latch != md->swap_bios))
 			__set_swap_bios_limit(md, latch);
 		down(&md->swap_bios_semaphore);
 	}
 
+	if (static_branch_unlikely(&zoned_enabled)) {
 		/*
-	 * Check if the IO needs a special mapping due to zone append emulation
-	 * on zoned target. In this case, dm_zone_map_bio() calls the target
-	 * map operation.
+		 * Check if the IO needs a special mapping due to zone append
+		 * emulation on zoned target. In this case, dm_zone_map_bio()
+		 * calls the target map operation.
 		 */
-	if (dm_emulate_zone_append(io->md))
+		if (unlikely(dm_emulate_zone_append(md)))
 			r = dm_zone_map_bio(tio);
 		else
 			r = ti->type->map(ti, clone);
+	} else
+		r = ti->type->map(ti, clone);
 
 	switch (r) {
 	case DM_MAPIO_SUBMITTED:
 		/* target has assumed ownership of this io */
 		if (!ti->accounts_remapped_io)
-			dm_io_set_flag(io, DM_IO_START_ACCT);
+			dm_start_io_acct(io, clone);
 		break;
 	case DM_MAPIO_REMAPPED:
-		/*
-		 * the bio has been remapped so dispatch it, but defer
-		 * dm_start_io_acct() until after possible bio_split().
-		 */
-		__dm_submit_bio_remap(clone, disk_devt(io->md->disk),
-				      tio->old_sector);
-		dm_io_set_flag(io, DM_IO_START_ACCT);
+		dm_submit_bio_remap(clone, NULL);
 		break;
 	case DM_MAPIO_KILL:
 	case DM_MAPIO_REQUEUE:
-		if (unlikely(swap_bios_limit(ti, clone)))
-			up(&io->md->swap_bios_semaphore);
+		if (static_branch_unlikely(&swap_bios_enabled) &&
+		    unlikely(swap_bios_limit(ti, clone)))
+			up(&md->swap_bios_semaphore);
 		free_tio(clone);
 		if (r == DM_MAPIO_KILL)
 			dm_io_dec_pending(io, BLK_STS_IOERR);
@@ -1321,6 +1357,31 @@ static void __map_bio(struct bio *clone)
 	}
 }
 
+static void setup_split_accounting(struct clone_info *ci, unsigned len)
+{
+	struct dm_io *io = ci->io;
+
+	if (ci->sector_count > len) {
+		/*
+		 * Split needed, save the mapped part for accounting.
+		 * NOTE: dm_accept_partial_bio() will update accordingly.
+		 */
+		dm_io_set_flag(io, DM_IO_WAS_SPLIT);
+		io->sectors = len;
+	}
+
+	if (static_branch_unlikely(&stats_enabled) &&
+	    unlikely(dm_stats_used(&io->md->stats))) {
+		/*
+		 * Save bi_sector in terms of its offset from end of
+		 * original bio, only needed for DM-stats' benefit.
+		 * - saved regardless of whether split needed so that
+		 *   dm_accept_partial_bio() doesn't need to.
+		 */
+		io->sector_offset = bio_end_sector(ci->bio) - ci->sector;
+	}
+}
+
 static void alloc_multiple_bios(struct bio_list *blist, struct clone_info *ci,
 				struct dm_target *ti, unsigned num_bios)
 {
@@ -1350,18 +1411,22 @@ static void alloc_multiple_bios(struct bio_list *blist, struct clone_info *ci,
 	}
 }
 
-static void __send_duplicate_bios(struct clone_info *ci, struct dm_target *ti,
+static int __send_duplicate_bios(struct clone_info *ci, struct dm_target *ti,
 				  unsigned num_bios, unsigned *len)
 {
 	struct bio_list blist = BIO_EMPTY_LIST;
 	struct bio *clone;
+	int ret = 0;
 
 	switch (num_bios) {
 	case 0:
 		break;
 	case 1:
+		if (len)
+			setup_split_accounting(ci, *len);
 		clone = alloc_tio(ci, ti, 0, len, GFP_NOIO);
 		__map_bio(clone);
+		ret = 1;
 		break;
 	default:
 		/* dm_accept_partial_bio() is not supported with shared tio->len_ptr */
@@ -1369,9 +1434,12 @@ static void __send_duplicate_bios(struct clone_info *ci, struct dm_target *ti,
 		while ((clone = bio_list_pop(&blist))) {
 			dm_tio_set_flag(clone_to_tio(clone), DM_TIO_IS_DUPLICATE_BIO);
 			__map_bio(clone);
+			ret += 1;
 		}
 		break;
 	}
+
+	return ret;
 }
 
 static void __send_empty_flush(struct clone_info *ci)
@@ -1392,8 +1460,19 @@ static void __send_empty_flush(struct clone_info *ci)
 	ci->sector_count = 0;
 	ci->io->tio.clone.bi_iter.bi_size = 0;
 
-	while ((ti = dm_table_get_target(ci->map, target_nr++)))
-		__send_duplicate_bios(ci, ti, ti->num_flush_bios, NULL);
+	while ((ti = dm_table_get_target(ci->map, target_nr++))) {
+		int bios;
+
+		atomic_add(ti->num_flush_bios, &ci->io->io_count);
+		bios = __send_duplicate_bios(ci, ti, ti->num_flush_bios, NULL);
+		atomic_sub(ti->num_flush_bios - bios, &ci->io->io_count);
+	}
+
+	/*
+	 * alloc_io() takes one extra reference for submission, so the
+	 * reference won't reach 0 without the following subtraction
+	 */
+	atomic_sub(1, &ci->io->io_count);
 
 	bio_uninit(ci->bio);
 }
@@ -1402,11 +1481,18 @@ static void __send_changing_extent_only(struct clone_info *ci, struct dm_target
 					unsigned num_bios)
 {
 	unsigned len;
+	int bios;
 
 	len = min_t(sector_t, ci->sector_count,
 		    max_io_len_target_boundary(ti, dm_target_offset(ti, ci->sector)));
 
-	__send_duplicate_bios(ci, ti, num_bios, &len);
+	atomic_add(num_bios, &ci->io->io_count);
+	bios = __send_duplicate_bios(ci, ti, num_bios, &len);
+	/*
+	 * alloc_io() takes one extra reference for submission, so the
+	 * reference won't reach 0 without the following (+1) subtraction
+	 */
+	atomic_sub(num_bios - bios + 1, &ci->io->io_count);
 
 	ci->sector += len;
 	ci->sector_count -= len;
@@ -1414,21 +1500,24 @@ static void __send_changing_extent_only(struct clone_info *ci, struct dm_target
 
 static bool is_abnormal_io(struct bio *bio)
 {
-	bool r = false;
+	unsigned int op = bio_op(bio);
 
-	switch (bio_op(bio)) {
+	if (op != REQ_OP_READ && op != REQ_OP_WRITE && op != REQ_OP_FLUSH) {
+		switch (op) {
 		case REQ_OP_DISCARD:
 		case REQ_OP_SECURE_ERASE:
 		case REQ_OP_WRITE_ZEROES:
-		r = true;
+			return true;
+		default:
 			break;
 		}
+	}
 
-	return r;
+	return false;
 }
 
-static bool __process_abnormal_io(struct clone_info *ci, struct dm_target *ti,
-				  int *result)
+static blk_status_t __process_abnormal_io(struct clone_info *ci,
+					  struct dm_target *ti)
 {
 	unsigned num_bios = 0;
 
@@ -1442,8 +1531,6 @@ static bool __process_abnormal_io(struct clone_info *ci, struct dm_target *ti,
 	case REQ_OP_WRITE_ZEROES:
 		num_bios = ti->num_write_zeroes_bios;
 		break;
-	default:
-		return false;
 	}
 
 	/*
@@ -1452,17 +1539,15 @@ static bool __process_abnormal_io(struct clone_info *ci, struct dm_target *ti,
 	 * reconfiguration might also have changed that since the
 	 * check was performed.
 	 */
-	if (!num_bios)
-		*result = -EOPNOTSUPP;
-	else {
+	if (unlikely(!num_bios))
+		return BLK_STS_NOTSUPP;
+
 	__send_changing_extent_only(ci, ti, num_bios);
-		*result = 0;
-	}
-	return true;
+	return BLK_STS_OK;
 }
 
 /*
- * Reuse ->bi_private as hlist head for storing all dm_io instances
+ * Reuse ->bi_private as dm_io list head for storing all dm_io instances
  * associated with this bio, and this bio's bi_private needs to be
  * stored in dm_io->data before the reuse.
  *
@@ -1470,54 +1555,53 @@ static bool __process_abnormal_io(struct clone_info *ci, struct dm_target *ti,
  * touch it after splitting. Meantime it won't be changed by anyone after
  * bio is submitted. So this reuse is safe.
  */
-static inline struct hlist_head *dm_get_bio_hlist_head(struct bio *bio)
+static inline struct dm_io **dm_poll_list_head(struct bio *bio)
 {
-	return (struct hlist_head *)&bio->bi_private;
+	return (struct dm_io **)&bio->bi_private;
 }
 
 static void dm_queue_poll_io(struct bio *bio, struct dm_io *io)
 {
-	struct hlist_head *head = dm_get_bio_hlist_head(bio);
+	struct dm_io **head = dm_poll_list_head(bio);
 
 	if (!(bio->bi_opf & REQ_DM_POLL_LIST)) {
 		bio->bi_opf |= REQ_DM_POLL_LIST;
 		/*
 		 * Save .bi_private into dm_io, so that we can reuse
-		 * .bi_private as hlist head for storing dm_io list
+		 * .bi_private as dm_io list head for storing dm_io list
 		 */
 		io->data = bio->bi_private;
 
-		INIT_HLIST_HEAD(head);
-
 		/* tell block layer to poll for completion */
 		bio->bi_cookie = ~BLK_QC_T_NONE;
+
+		io->next = NULL;
 	} else {
 		/*
 		 * bio recursed due to split, reuse original poll list,
 		 * and save bio->bi_private too.
 		 */
-		io->data = hlist_entry(head->first, struct dm_io, node)->data;
+		io->data = (*head)->data;
+		io->next = *head;
 	}
 
-	hlist_add_head(&io->node, head);
+	*head = io;
 }
 
 /*
  * Select the correct strategy for processing a non-flush bio.
  */
-static int __split_and_process_bio(struct clone_info *ci)
+static blk_status_t __split_and_process_bio(struct clone_info *ci)
 {
 	struct bio *clone;
 	struct dm_target *ti;
 	unsigned len;
-	int r;
 
 	ti = dm_table_find_target(ci->map, ci->sector);
-	if (!ti)
-		return -EIO;
-
-	if (__process_abnormal_io(ci, ti, &r))
-		return r;
+	if (unlikely(!ti))
+		return BLK_STS_IOERR;
+	else if (unlikely(ci->is_abnormal_io))
+		return __process_abnormal_io(ci, ti);
 
 	/*
 	 * Only support bio polling for normal IO, and the target io is
@@ -1526,27 +1610,30 @@ static int __split_and_process_bio(struct clone_info *ci)
 	ci->submit_as_polled = ci->bio->bi_opf & REQ_POLLED;
 
 	len = min_t(sector_t, max_io_len(ti, ci->sector), ci->sector_count);
+	setup_split_accounting(ci, len);
 	clone = alloc_tio(ci, ti, 0, &len, GFP_NOIO);
 	__map_bio(clone);
 
 	ci->sector += len;
 	ci->sector_count -= len;
 
-	return 0;
+	return BLK_STS_OK;
 }
 
 static void init_clone_info(struct clone_info *ci, struct mapped_device *md,
-			    struct dm_table *map, struct bio *bio)
+			    struct dm_table *map, struct bio *bio, bool is_abnormal)
 {
 	ci->map = map;
 	ci->io = alloc_io(md, bio);
 	ci->bio = bio;
+	ci->is_abnormal_io = is_abnormal;
 	ci->submit_as_polled = false;
 	ci->sector = bio->bi_iter.bi_sector;
 	ci->sector_count = bio_sectors(bio);
 
 	/* Shouldn't happen but sector_count was being set to 0 so... */
-	if (WARN_ON_ONCE(op_is_zone_mgmt(bio_op(bio)) && ci->sector_count))
+	if (static_branch_unlikely(&zoned_enabled) &&
+	    WARN_ON_ONCE(op_is_zone_mgmt(bio_op(bio)) && ci->sector_count))
 		ci->sector_count = 0;
 }
 
@@ -1557,10 +1644,21 @@ static void dm_split_and_process_bio(struct mapped_device *md,
 				     struct dm_table *map, struct bio *bio)
 {
 	struct clone_info ci;
-	struct bio *orig_bio = NULL;
-	int error = 0;
+	struct dm_io *io;
+	blk_status_t error = BLK_STS_OK;
+	bool is_abnormal;
+
+	is_abnormal = is_abnormal_io(bio);
+	if (unlikely(is_abnormal)) {
+		/*
+		 * Use blk_queue_split() for abnormal IO (e.g. discard, etc)
+		 * otherwise associated queue_limits won't be imposed.
+		 */
+		blk_queue_split(&bio);
+	}
 
-	init_clone_info(&ci, md, map, bio);
+	init_clone_info(&ci, md, map, bio, is_abnormal);
+	io = ci.io;
 
 	if (bio->bi_opf & REQ_PREFLUSH) {
 		__send_empty_flush(&ci);
@@ -1569,40 +1667,34 @@ static void dm_split_and_process_bio(struct mapped_device *md,
 	}
 
 	error = __split_and_process_bio(&ci);
-	ci.io->map_task = NULL;
 	if (error || !ci.sector_count)
 		goto out;
-
 	/*
 	 * Remainder must be passed to submit_bio_noacct() so it gets handled
 	 * *after* bios already submitted have been completely processed.
-	 * We take a clone of the original to store in ci.io->orig_bio to be
-	 * used by dm_end_io_acct() and for dm_io_complete() to use for
-	 * completion handling.
-	 */
-	orig_bio = bio_split(bio, bio_sectors(bio) - ci.sector_count,
-			     GFP_NOIO, &md->queue->bio_split);
-	bio_chain(orig_bio, bio);
-	trace_block_split(orig_bio, bio->bi_iter.bi_sector);
+	 */
+	bio_trim(bio, io->sectors, ci.sector_count);
+	trace_block_split(bio, bio->bi_iter.bi_sector);
+	bio_inc_remaining(bio);
 	submit_bio_noacct(bio);
 out:
-	if (!orig_bio)
-		orig_bio = bio;
-	smp_store_release(&ci.io->orig_bio, orig_bio);
-	if (dm_io_flagged(ci.io, DM_IO_START_ACCT))
-		dm_start_io_acct(ci.io, NULL);
-
 	/*
 	 * Drop the extra reference count for non-POLLED bio, and hold one
 	 * reference for POLLED bio, which will be released in dm_poll_bio
 	 *
-	 * Add every dm_io instance into the hlist_head which is stored in
-	 * bio->bi_private, so that dm_poll_bio can poll them all.
+	 * Add every dm_io instance into the dm_io list head which is stored
+	 * in bio->bi_private, so that dm_poll_bio can poll them all.
 	 */
-	if (error || !ci.submit_as_polled)
-		dm_io_dec_pending(ci.io, errno_to_blk_status(error));
-	else
-		dm_queue_poll_io(bio, ci.io);
+	if (error || !ci.submit_as_polled) {
+		/*
+		 * In case of submission failure, the extra reference for
+		 * submitting io isn't consumed yet
+		 */
+		if (error)
+			atomic_dec(&io->io_count);
+		dm_io_dec_pending(io, error);
+	} else
+		dm_queue_poll_io(bio, io);
 }
 
 static void dm_submit_bio(struct bio *bio)
@@ -1611,7 +1703,7 @@ static void dm_submit_bio(struct bio *bio)
 	int srcu_idx;
 	struct dm_table *map;
 
-	map = dm_get_live_table(md, &srcu_idx);
+	map = dm_get_live_table_bio(md, &srcu_idx, bio);
 
 	/* If suspended, or map not yet available, queue this IO for later */
 	if (unlikely(test_bit(DMF_BLOCK_IO_FOR_SUSPEND, &md->flags)) ||
@@ -1625,16 +1717,9 @@ static void dm_submit_bio(struct bio *bio)
 		goto out;
 	}
 
-	/*
-	 * Use blk_queue_split() for abnormal IO (e.g. discard, writesame, etc)
-	 * otherwise associated queue_limits won't be imposed.
-	 */
-	if (is_abnormal_io(bio))
-		blk_queue_split(&bio);
-
 	dm_split_and_process_bio(md, map, bio);
 out:
-	dm_put_live_table(md, srcu_idx);
+	dm_put_live_table_bio(md, srcu_idx, bio);
 }
 
 static bool dm_poll_dm_io(struct dm_io *io, struct io_comp_batch *iob,
@@ -1653,18 +1738,16 @@ static bool dm_poll_dm_io(struct dm_io *io, struct io_comp_batch *iob,
 static int dm_poll_bio(struct bio *bio, struct io_comp_batch *iob,
 		       unsigned int flags)
 {
-	struct hlist_head *head = dm_get_bio_hlist_head(bio);
-	struct hlist_head tmp = HLIST_HEAD_INIT;
-	struct hlist_node *next;
-	struct dm_io *io;
+	struct dm_io **head = dm_poll_list_head(bio);
+	struct dm_io *list = *head;
+	struct dm_io *tmp = NULL;
+	struct dm_io *curr, *next;
 
 	/* Only poll normal bio which was marked as REQ_DM_POLL_LIST */
 	if (!(bio->bi_opf & REQ_DM_POLL_LIST))
 		return 0;
 
-	WARN_ON_ONCE(hlist_empty(head));
-
-	hlist_move_list(head, &tmp);
+	WARN_ON_ONCE(!list);
 
 	/*
 	 * Restore .bi_private before possibly completing dm_io.
@@ -1675,24 +1758,27 @@ static int dm_poll_bio(struct bio *bio, struct io_comp_batch *iob,
 	 * clearing REQ_DM_POLL_LIST here.
 	 */
 	bio->bi_opf &= ~REQ_DM_POLL_LIST;
-	bio->bi_private = hlist_entry(tmp.first, struct dm_io, node)->data;
+	bio->bi_private = list->data;
 
-	hlist_for_each_entry_safe(io, next, &tmp, node) {
-		if (dm_poll_dm_io(io, iob, flags)) {
-			hlist_del_init(&io->node);
+	for (curr = list, next = curr->next; curr; curr = next, next =
+			curr ? curr->next : NULL) {
+		if (dm_poll_dm_io(curr, iob, flags)) {
 			/*
-			 * clone_endio() has already occurred, so passing
-			 * error as 0 here doesn't override io->status
+			 * clone_endio() has already occurred, so no
+			 * error handling is needed here.
 			 */
-			dm_io_dec_pending(io, 0);
+			__dm_io_dec_pending(curr);
+		} else {
+			curr->next = tmp;
+			tmp = curr;
 		}
 	}
 
 	/* Not done? */
-	if (!hlist_empty(&tmp)) {
+	if (tmp) {
 		bio->bi_opf |= REQ_DM_POLL_LIST;
 		/* Reset bio->bi_private to dm_io list head */
-		hlist_move_list(&tmp, head);
+		*head = tmp;
 		return 0;
 	}
 	return 1;
@@ -2901,8 +2987,8 @@ int dm_noflush_suspending(struct dm_target *ti)
 EXPORT_SYMBOL_GPL(dm_noflush_suspending);
 
 struct dm_md_mempools *dm_alloc_md_mempools(struct mapped_device *md, enum dm_queue_mode type,
-					    unsigned integrity, unsigned per_io_data_size,
-					    unsigned min_pool_size)
+					    unsigned per_io_data_size, unsigned min_pool_size,
+					    bool integrity, bool poll)
 {
 	struct dm_md_mempools *pools = kzalloc_node(sizeof(*pools), GFP_KERNEL, md->numa_node_id);
 	unsigned int pool_size = 0;
@@ -2918,7 +3004,7 @@ struct dm_md_mempools *dm_alloc_md_mempools(struct mapped_device *md, enum dm_qu
 		pool_size = max(dm_get_reserved_bio_based_ios(), min_pool_size);
 		front_pad = roundup(per_io_data_size, __alignof__(struct dm_target_io)) + DM_TARGET_IO_BIO_OFFSET;
 		io_front_pad = roundup(per_io_data_size,  __alignof__(struct dm_io)) + DM_IO_BIO_OFFSET;
-		ret = bioset_init(&pools->io_bs, pool_size, io_front_pad, 0);
+		ret = bioset_init(&pools->io_bs, pool_size, io_front_pad, poll ? BIOSET_PERCPU_CACHE : 0);
 		if (ret)
 			goto out;
 		if (integrity && bioset_integrity_create(&pools->io_bs, pool_size))

drivers/md/dm.h

@@ -221,8 +221,8 @@ void dm_kcopyd_exit(void);
  * Mempool operations
  */
 struct dm_md_mempools *dm_alloc_md_mempools(struct mapped_device *md, enum dm_queue_mode type,
-					    unsigned integrity, unsigned per_bio_data_size,
-					    unsigned min_pool_size);
+					    unsigned per_io_data_size, unsigned min_pool_size,
+					    bool integrity, bool poll);
 void dm_free_md_mempools(struct dm_md_mempools *pools);
 
 /*