5 files changed, 29 insertions, 12 deletions

diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index e4a3f692057b..a3763d664a67 100644
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -548,6 +548,7 @@ struct cache_set {
 	 */
 	wait_queue_head_t	btree_cache_wait;
 	struct task_struct	*btree_cache_alloc_lock;
+	spinlock_t		btree_cannibalize_lock;
 
 	/*
 	 * When we free a btree node, we increment the gen of the bucket the
diff --git a/drivers/md/bcache/bset.c b/drivers/md/bcache/bset.c
index 1edf9515345e..03fb06c61e1c 100644
--- a/drivers/md/bcache/bset.c
+++ b/drivers/md/bcache/bset.c
@@ -319,7 +319,7 @@ int bch_btree_keys_alloc(struct btree_keys *b, unsigned page_order, gfp_t gfp)
 
 	b->page_order = page_order;
 
-	t->data = (void *) __get_free_pages(gfp, b->page_order);
+	t->data = (void *) __get_free_pages(__GFP_COMP|gfp, b->page_order);
 	if (!t->data)
 		goto err;
 
diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c
index 96a6583e7b52..fba0fff8040d 100644
--- a/drivers/md/bcache/btree.c
+++ b/drivers/md/bcache/btree.c
@@ -794,7 +794,7 @@ int bch_btree_cache_alloc(struct cache_set *c)
 	mutex_init(&c->verify_lock);
 
 	c->verify_ondisk = (void *)
-		__get_free_pages(GFP_KERNEL, ilog2(bucket_pages(c)));
+		__get_free_pages(GFP_KERNEL|__GFP_COMP, ilog2(bucket_pages(c)));
 
 	c->verify_data = mca_bucket_alloc(c, &ZERO_KEY, GFP_KERNEL);
 
@@ -840,15 +840,17 @@ out:
 
 static int mca_cannibalize_lock(struct cache_set *c, struct btree_op *op)
 {
-	struct task_struct *old;
-
-	old = cmpxchg(&c->btree_cache_alloc_lock, NULL, current);
-	if (old && old != current) {
+	spin_lock(&c->btree_cannibalize_lock);
+	if (likely(c->btree_cache_alloc_lock == NULL)) {
+		c->btree_cache_alloc_lock = current;
+	} else if (c->btree_cache_alloc_lock != current) {
 		if (op)
 			prepare_to_wait(&c->btree_cache_wait, &op->wait,
 					TASK_UNINTERRUPTIBLE);
+		spin_unlock(&c->btree_cannibalize_lock);
 		return -EINTR;
 	}
+	spin_unlock(&c->btree_cannibalize_lock);
 
 	return 0;
 }
@@ -883,10 +885,12 @@ static struct btree *mca_cannibalize(struct cache_set *c, struct btree_op *op,
  */
 static void bch_cannibalize_unlock(struct cache_set *c)
 {
+	spin_lock(&c->btree_cannibalize_lock);
 	if (c->btree_cache_alloc_lock == current) {
 		c->btree_cache_alloc_lock = NULL;
 		wake_up(&c->btree_cache_wait);
 	}
+	spin_unlock(&c->btree_cannibalize_lock);
 }
 
 static struct btree *mca_alloc(struct cache_set *c, struct btree_op *op,
@@ -1374,7 +1378,7 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,
 			if (__set_blocks(n1, n1->keys + n2->keys,
 					 block_bytes(b->c)) >
 			    btree_blocks(new_nodes[i]))
-				goto out_nocoalesce;
+				goto out_unlock_nocoalesce;
 
 			keys = n2->keys;
 			/* Take the key of the node we're getting rid of */
@@ -1403,7 +1407,7 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,
 
 		if (__bch_keylist_realloc(&keylist,
 					  bkey_u64s(&new_nodes[i]->key)))
-			goto out_nocoalesce;
+			goto out_unlock_nocoalesce;
 
 		bch_btree_node_write(new_nodes[i], &cl);
 		bch_keylist_add(&keylist, &new_nodes[i]->key);
@@ -1449,6 +1453,10 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,
 	/* Invalidated our iterator */
 	return -EINTR;
 
+out_unlock_nocoalesce:
+	for (i = 0; i < nodes; i++)
+		mutex_unlock(&new_nodes[i]->write_lock);
+
 out_nocoalesce:
 	closure_sync(&cl);
 	bch_keylist_free(&keylist);
diff --git a/drivers/md/bcache/journal.c b/drivers/md/bcache/journal.c
index 6394be5ee9a8..6aafda26903c 100644
--- a/drivers/md/bcache/journal.c
+++ b/drivers/md/bcache/journal.c
@@ -838,8 +838,8 @@ int bch_journal_alloc(struct cache_set *c)
 	j->w[1].c = c;
 
 	if (!(init_fifo(&j->pin, JOURNAL_PIN, GFP_KERNEL)) ||
-	    !(j->w[0].data = (void *) __get_free_pages(GFP_KERNEL, JSET_BITS)) ||
-	    !(j->w[1].data = (void *) __get_free_pages(GFP_KERNEL, JSET_BITS)))
+	    !(j->w[0].data = (void *) __get_free_pages(GFP_KERNEL|__GFP_COMP, JSET_BITS)) ||
+	    !(j->w[1].data = (void *) __get_free_pages(GFP_KERNEL|__GFP_COMP, JSET_BITS)))
 		return -ENOMEM;
 
 	return 0;
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index 690aeb09bbf5..6bf1559a1f0d 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1468,7 +1468,7 @@ void bch_cache_set_unregister(struct cache_set *c)
 }
 
 #define alloc_bucket_pages(gfp, c)			\
-	((void *) __get_free_pages(__GFP_ZERO|gfp, ilog2(bucket_pages(c))))
+	((void *) __get_free_pages(__GFP_ZERO|__GFP_COMP|gfp, ilog2(bucket_pages(c))))
 
 struct cache_set *bch_cache_set_alloc(struct cache_sb *sb)
 {
@@ -1510,6 +1510,7 @@ struct cache_set *bch_cache_set_alloc(struct cache_sb *sb)
 	sema_init(&c->sb_write_mutex, 1);
 	mutex_init(&c->bucket_lock);
 	init_waitqueue_head(&c->btree_cache_wait);
+	spin_lock_init(&c->btree_cannibalize_lock);
 	init_waitqueue_head(&c->bucket_wait);
 	init_waitqueue_head(&c->gc_wait);
 	sema_init(&c->uuid_write_mutex, 1);
@@ -1780,7 +1781,14 @@ found:
 	    sysfs_create_link(&c->kobj, &ca->kobj, buf))
 		goto err;
 
-	if (ca->sb.seq > c->sb.seq) {
+	/*
+	 * A special case is both ca->sb.seq and c->sb.seq are 0,
+	 * such condition happens on a new created cache device whose
+	 * super block is never flushed yet. In this case c->sb.version
+	 * and other members should be updated too, otherwise we will
+	 * have a mistaken super block version in cache set.
+	 */
+	if (ca->sb.seq > c->sb.seq || c->sb.seq == 0) {
 		c->sb.version		= ca->sb.version;
 		memcpy(c->sb.set_uuid, ca->sb.set_uuid, 16);
 		c->sb.flags             = ca->sb.flags;