1 files changed, 131 insertions, 45 deletions

diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index c434e5aab2d..5bd2290cfb1 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -66,6 +66,18 @@ struct dm_snapshot {
 
 	atomic_t pending_exceptions_count;
 
+	/* Protected by "lock" */
+	sector_t exception_start_sequence;
+
+	/* Protected by kcopyd single-threaded callback */
+	sector_t exception_complete_sequence;
+
+	/*
+	 * A list of pending exceptions that completed out of order.
+	 * Protected by kcopyd single-threaded callback.
+	 */
+	struct list_head out_of_order_list;
+
 	mempool_t *pending_pool;
 
 	struct dm_exception_table pending;
@@ -173,6 +185,14 @@ struct dm_snap_pending_exception {
 	 */
 	int started;
 
+	/* There was copying error. */
+	int copy_error;
+
+	/* A sequence number, it is used for in-order completion. */
+	sector_t exception_sequence;
+
+	struct list_head out_of_order_entry;
+
 	/*
 	 * For writing a complete chunk, bypassing the copy.
 	 */
@@ -590,12 +610,12 @@ static struct dm_exception *dm_lookup_exception(struct dm_exception_table *et,
 	return NULL;
 }
 
-static struct dm_exception *alloc_completed_exception(void)
+static struct dm_exception *alloc_completed_exception(gfp_t gfp)
 {
 	struct dm_exception *e;
 
-	e = kmem_cache_alloc(exception_cache, GFP_NOIO);
-	if (!e)
+	e = kmem_cache_alloc(exception_cache, gfp);
+	if (!e && gfp == GFP_NOIO)
 		e = kmem_cache_alloc(exception_cache, GFP_ATOMIC);
 
 	return e;
@@ -622,7 +642,7 @@ static void free_pending_exception(struct dm_snap_pending_exception *pe)
 	struct dm_snapshot *s = pe->snap;
 
 	mempool_free(pe, s->pending_pool);
-	smp_mb__before_atomic_dec();
+	smp_mb__before_atomic();
 	atomic_dec(&s->pending_exceptions_count);
 }
 
@@ -677,7 +697,7 @@ static int dm_add_exception(void *context, chunk_t old, chunk_t new)
 	struct dm_snapshot *s = context;
 	struct dm_exception *e;
 
-	e = alloc_completed_exception();
+	e = alloc_completed_exception(GFP_KERNEL);
 	if (!e)
 		return -ENOMEM;
 
@@ -725,17 +745,16 @@ static int calc_max_buckets(void)
  */
 static int init_hash_tables(struct dm_snapshot *s)
 {
-	sector_t hash_size, cow_dev_size, origin_dev_size, max_buckets;
+	sector_t hash_size, cow_dev_size, max_buckets;
 
 	/*
 	 * Calculate based on the size of the original volume or
 	 * the COW volume...
 	 */
 	cow_dev_size = get_dev_size(s->cow->bdev);
-	origin_dev_size = get_dev_size(s->origin->bdev);
 	max_buckets = calc_max_buckets();
 
-	hash_size = min(origin_dev_size, cow_dev_size) >> s->store->chunk_shift;
+	hash_size = cow_dev_size >> s->store->chunk_shift;
 	hash_size = min(hash_size, max_buckets);
 
 	if (hash_size < 64)
@@ -764,7 +783,7 @@ static int init_hash_tables(struct dm_snapshot *s)
 static void merge_shutdown(struct dm_snapshot *s)
 {
 	clear_bit_unlock(RUNNING_MERGE, &s->state_bits);
-	smp_mb__after_clear_bit();
+	smp_mb__after_atomic();
 	wake_up_bit(&s->state_bits, RUNNING_MERGE);
 }
 
@@ -1095,6 +1114,9 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	s->valid = 1;
 	s->active = 0;
 	atomic_set(&s->pending_exceptions_count, 0);
+	s->exception_start_sequence = 0;
+	s->exception_complete_sequence = 0;
+	INIT_LIST_HEAD(&s->out_of_order_list);
 	init_rwsem(&s->lock);
 	INIT_LIST_HEAD(&s->list);
 	spin_lock_init(&s->pe_lock);
@@ -1383,7 +1405,7 @@ static void pending_complete(struct dm_snap_pending_exception *pe, int success)
 		goto out;
 	}
 
-	e = alloc_completed_exception();
+	e = alloc_completed_exception(GFP_NOIO);
 	if (!e) {
 		down_write(&s->lock);
 		__invalidate_snapshot(s, -ENOMEM);
@@ -1416,6 +1438,7 @@ out:
 	if (full_bio) {
 		full_bio->bi_end_io = pe->full_bio_end_io;
 		full_bio->bi_private = pe->full_bio_private;
+		atomic_inc(&full_bio->bi_remaining);
 	}
 	free_pending_exception(pe);
 
@@ -1444,6 +1467,19 @@ static void commit_callback(void *context, int success)
 	pending_complete(pe, success);
 }
 
+static void complete_exception(struct dm_snap_pending_exception *pe)
+{
+	struct dm_snapshot *s = pe->snap;
+
+	if (unlikely(pe->copy_error))
+		pending_complete(pe, 0);
+
+	else
+		/* Update the metadata if we are persistent */
+		s->store->type->commit_exception(s->store, &pe->e,
+						 commit_callback, pe);
+}
+
 /*
  * Called when the copy I/O has finished.  kcopyd actually runs
  * this code so don't block.
@@ -1453,13 +1489,32 @@ static void copy_callback(int read_err, unsigned long write_err, void *context)
 	struct dm_snap_pending_exception *pe = context;
 	struct dm_snapshot *s = pe->snap;
 
-	if (read_err || write_err)
-		pending_complete(pe, 0);
+	pe->copy_error = read_err || write_err;
 
-	else
-		/* Update the metadata if we are persistent */
-		s->store->type->commit_exception(s->store, &pe->e,
-						 commit_callback, pe);
+	if (pe->exception_sequence == s->exception_complete_sequence) {
+		s->exception_complete_sequence++;
+		complete_exception(pe);
+
+		while (!list_empty(&s->out_of_order_list)) {
+			pe = list_entry(s->out_of_order_list.next,
+					struct dm_snap_pending_exception, out_of_order_entry);
+			if (pe->exception_sequence != s->exception_complete_sequence)
+				break;
+			s->exception_complete_sequence++;
+			list_del(&pe->out_of_order_entry);
+			complete_exception(pe);
+		}
+	} else {
+		struct list_head *lh;
+		struct dm_snap_pending_exception *pe2;
+
+		list_for_each_prev(lh, &s->out_of_order_list) {
+			pe2 = list_entry(lh, struct dm_snap_pending_exception, out_of_order_entry);
+			if (pe2->exception_sequence < pe->exception_sequence)
+				break;
+		}
+		list_add(&pe->out_of_order_entry, lh);
+	}
 }
 
 /*
@@ -1554,6 +1609,8 @@ __find_pending_exception(struct dm_snapshot *s,
 		return NULL;
 	}
 
+	pe->exception_sequence = s->exception_start_sequence++;
+
 	dm_insert_exception(&s->pending, &pe->e);
 
 	return pe;
@@ -1563,11 +1620,10 @@ static void remap_exception(struct dm_snapshot *s, struct dm_exception *e,
 			    struct bio *bio, chunk_t chunk)
 {
 	bio->bi_bdev = s->cow->bdev;
-	bio->bi_sector = chunk_to_sector(s->store,
-					 dm_chunk_number(e->new_chunk) +
-					 (chunk - e->old_chunk)) +
-					 (bio->bi_sector &
-					  s->store->chunk_mask);
+	bio->bi_iter.bi_sector =
+		chunk_to_sector(s->store, dm_chunk_number(e->new_chunk) +
+				(chunk - e->old_chunk)) +
+		(bio->bi_iter.bi_sector & s->store->chunk_mask);
 }
 
 static int snapshot_map(struct dm_target *ti, struct bio *bio)
@@ -1585,7 +1641,7 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio)
 		return DM_MAPIO_REMAPPED;
 	}
 
-	chunk = sector_to_chunk(s->store, bio->bi_sector);
+	chunk = sector_to_chunk(s->store, bio->bi_iter.bi_sector);
 
 	/* Full snapshots are not usable */
 	/* To get here the table must be live so s->active is always set. */
@@ -1646,7 +1702,8 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio)
 		r = DM_MAPIO_SUBMITTED;
 
 		if (!pe->started &&
-		    bio->bi_size == (s->store->chunk_size << SECTOR_SHIFT)) {
+		    bio->bi_iter.bi_size ==
+		    (s->store->chunk_size << SECTOR_SHIFT)) {
 			pe->started = 1;
 			up_write(&s->lock);
 			start_full_bio(pe, bio);
@@ -1702,7 +1759,7 @@ static int snapshot_merge_map(struct dm_target *ti, struct bio *bio)
 		return DM_MAPIO_REMAPPED;
 	}
 
-	chunk = sector_to_chunk(s->store, bio->bi_sector);
+	chunk = sector_to_chunk(s->store, bio->bi_iter.bi_sector);
 
 	down_write(&s->lock);
 
@@ -2039,7 +2096,7 @@ static int do_origin(struct dm_dev *origin, struct bio *bio)
 	down_read(&_origins_lock);
 	o = __lookup_origin(origin->bdev);
 	if (o)
-		r = __origin_write(&o->snapshots, bio->bi_sector, bio);
+		r = __origin_write(&o->snapshots, bio->bi_iter.bi_sector, bio);
 	up_read(&_origins_lock);
 
 	return r;
@@ -2084,6 +2141,11 @@ static int origin_write_extent(struct dm_snapshot *merging_snap,
  * Origin: maps a linear range of a device, with hooks for snapshotting.
  */
 
+struct dm_origin {
+	struct dm_dev *dev;
+	unsigned split_boundary;
+};
+
 /*
  * Construct an origin mapping: <dev_path>
  * The context for an origin is merely a 'struct dm_dev *'
@@ -2092,41 +2154,65 @@ static int origin_write_extent(struct dm_snapshot *merging_snap,
 static int origin_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 {
 	int r;
-	struct dm_dev *dev;
+	struct dm_origin *o;
 
 	if (argc != 1) {
 		ti->error = "origin: incorrect number of arguments";
 		return -EINVAL;
 	}
 
-	r = dm_get_device(ti, argv[0], dm_table_get_mode(ti->table), &dev);
+	o = kmalloc(sizeof(struct dm_origin), GFP_KERNEL);
+	if (!o) {
+		ti->error = "Cannot allocate private origin structure";
+		r = -ENOMEM;
+		goto bad_alloc;
+	}
+
+	r = dm_get_device(ti, argv[0], dm_table_get_mode(ti->table), &o->dev);
 	if (r) {
 		ti->error = "Cannot get target device";
-		return r;
+		goto bad_open;
 	}
 
-	ti->private = dev;
+	ti->private = o;
 	ti->num_flush_bios = 1;
 
 	return 0;
+
+bad_open:
+	kfree(o);
+bad_alloc:
+	return r;
 }
 
 static void origin_dtr(struct dm_target *ti)
 {
-	struct dm_dev *dev = ti->private;
-	dm_put_device(ti, dev);
+	struct dm_origin *o = ti->private;
+	dm_put_device(ti, o->dev);
+	kfree(o);
 }
 
 static int origin_map(struct dm_target *ti, struct bio *bio)
 {
-	struct dm_dev *dev = ti->private;
-	bio->bi_bdev = dev->bdev;
+	struct dm_origin *o = ti->private;
+	unsigned available_sectors;
 
-	if (bio->bi_rw & REQ_FLUSH)
+	bio->bi_bdev = o->dev->bdev;
+
+	if (unlikely(bio->bi_rw & REQ_FLUSH))
+		return DM_MAPIO_REMAPPED;
+
+	if (bio_rw(bio) != WRITE)
 		return DM_MAPIO_REMAPPED;
 
+	available_sectors = o->split_boundary -
+		((unsigned)bio->bi_iter.bi_sector & (o->split_boundary - 1));
+
+	if (bio_sectors(bio) > available_sectors)
+		dm_accept_partial_bio(bio, available_sectors);
+
 	/* Only tell snapshots if this is a write */
-	return (bio_rw(bio) == WRITE) ? do_origin(dev, bio) : DM_MAPIO_REMAPPED;
+	return do_origin(o->dev, bio);
 }
 
 /*
@@ -2135,15 +2221,15 @@ static int origin_map(struct dm_target *ti, struct bio *bio)
  */
 static void origin_resume(struct dm_target *ti)
 {
-	struct dm_dev *dev = ti->private;
+	struct dm_origin *o = ti->private;
 
-	ti->max_io_len = get_origin_minimum_chunksize(dev->bdev);
+	o->split_boundary = get_origin_minimum_chunksize(o->dev->bdev);
 }
 
 static void origin_status(struct dm_target *ti, status_type_t type,
 			  unsigned status_flags, char *result, unsigned maxlen)
 {
-	struct dm_dev *dev = ti->private;
+	struct dm_origin *o = ti->private;
 
 	switch (type) {
 	case STATUSTYPE_INFO:
@@ -2151,7 +2237,7 @@ static void origin_status(struct dm_target *ti, status_type_t type,
 		break;
 
 	case STATUSTYPE_TABLE:
-		snprintf(result, maxlen, "%s", dev->name);
+		snprintf(result, maxlen, "%s", o->dev->name);
 		break;
 	}
 }
@@ -2159,13 +2245,13 @@ static void origin_status(struct dm_target *ti, status_type_t type,
 static int origin_merge(struct dm_target *ti, struct bvec_merge_data *bvm,
 			struct bio_vec *biovec, int max_size)
 {
-	struct dm_dev *dev = ti->private;
-	struct request_queue *q = bdev_get_queue(dev->bdev);
+	struct dm_origin *o = ti->private;
+	struct request_queue *q = bdev_get_queue(o->dev->bdev);
 
 	if (!q->merge_bvec_fn)
 		return max_size;
 
-	bvm->bi_bdev = dev->bdev;
+	bvm->bi_bdev = o->dev->bdev;
 
 	return min(max_size, q->merge_bvec_fn(q, bvm, biovec));
 }
@@ -2173,9 +2259,9 @@ static int origin_merge(struct dm_target *ti, struct bvec_merge_data *bvm,
 static int origin_iterate_devices(struct dm_target *ti,
 				  iterate_devices_callout_fn fn, void *data)
 {
-	struct dm_dev *dev = ti->private;
+	struct dm_origin *o = ti->private;
 
-	return fn(ti, dev, 0, ti->len, data);
+	return fn(ti, o->dev, 0, ti->len, data);
 }
 
 static struct target_type origin_target = {
@@ -2193,7 +2279,7 @@ static struct target_type origin_target = {
 
 static struct target_type snapshot_target = {
 	.name    = "snapshot",
-	.version = {1, 11, 1},
+	.version = {1, 12, 0},
 	.module  = THIS_MODULE,
 	.ctr     = snapshot_ctr,
 	.dtr     = snapshot_dtr,