27 files changed, 414 insertions, 172 deletions

diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index 23bb0ceabe3..ce8ef610772 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -919,6 +919,10 @@ static int context_is_writeable_or_written(struct inode *inode,
 /*
  * We are only allowed to write into/dirty the page if the page is
  * clean, or already dirty within the same snap context.
+ *
+ * called with page locked.
+ * return success with page locked,
+ * or any failure (incl -EAGAIN) with page unlocked.
  */
 static int ceph_update_writeable_page(struct file *file,
 			    loff_t pos, unsigned len,
@@ -961,9 +965,11 @@ retry_locked:
 			snapc = ceph_get_snap_context((void *)page->private);
 			unlock_page(page);
 			ceph_queue_writeback(inode);
-			wait_event_interruptible(ci->i_cap_wq,
+			r = wait_event_interruptible(ci->i_cap_wq,
 			       context_is_writeable_or_written(inode, snapc));
 			ceph_put_snap_context(snapc);
+			if (r == -ERESTARTSYS)
+				return r;
 			return -EAGAIN;
 		}
 
@@ -1035,7 +1041,7 @@ static int ceph_write_begin(struct file *file, struct address_space *mapping,
 	int r;
 
 	do {
-		/* get a page*/
+		/* get a page */
 		page = grab_cache_page_write_begin(mapping, index, 0);
 		if (!page)
 			return -ENOMEM;
diff --git a/fs/ceph/auth_x.c b/fs/ceph/auth_x.c
index f0318427b6d..8d8a8496476 100644
--- a/fs/ceph/auth_x.c
+++ b/fs/ceph/auth_x.c
@@ -28,6 +28,12 @@ static int ceph_x_is_authenticated(struct ceph_auth_client *ac)
 	return (ac->want_keys & xi->have_keys) == ac->want_keys;
 }
 
+static int ceph_x_encrypt_buflen(int ilen)
+{
+	return sizeof(struct ceph_x_encrypt_header) + ilen + 16 +
+		sizeof(u32);
+}
+
 static int ceph_x_encrypt(struct ceph_crypto_key *secret,
 			  void *ibuf, int ilen, void *obuf, size_t olen)
 {
@@ -150,6 +156,11 @@ static int ceph_x_proc_ticket_reply(struct ceph_auth_client *ac,
 		struct timespec validity;
 		struct ceph_crypto_key old_key;
 		void *tp, *tpend;
+		struct ceph_timespec new_validity;
+		struct ceph_crypto_key new_session_key;
+		struct ceph_buffer *new_ticket_blob;
+		unsigned long new_expires, new_renew_after;
+		u64 new_secret_id;
 
 		ceph_decode_need(&p, end, sizeof(u32) + 1, bad);
 
@@ -182,16 +193,16 @@ static int ceph_x_proc_ticket_reply(struct ceph_auth_client *ac,
 			goto bad;
 
 		memcpy(&old_key, &th->session_key, sizeof(old_key));
-		ret = ceph_crypto_key_decode(&th->session_key, &dp, dend);
+		ret = ceph_crypto_key_decode(&new_session_key, &dp, dend);
 		if (ret)
 			goto out;
 
-		ceph_decode_copy(&dp, &th->validity, sizeof(th->validity));
-		ceph_decode_timespec(&validity, &th->validity);
-		th->expires = get_seconds() + validity.tv_sec;
-		th->renew_after = th->expires - (validity.tv_sec / 4);
-		dout(" expires=%lu renew_after=%lu\n", th->expires,
-		     th->renew_after);
+		ceph_decode_copy(&dp, &new_validity, sizeof(new_validity));
+		ceph_decode_timespec(&validity, &new_validity);
+		new_expires = get_seconds() + validity.tv_sec;
+		new_renew_after = new_expires - (validity.tv_sec / 4);
+		dout(" expires=%lu renew_after=%lu\n", new_expires,
+		     new_renew_after);
 
 		/* ticket blob for service */
 		ceph_decode_8_safe(&p, end, is_enc, bad);
@@ -216,10 +227,21 @@ static int ceph_x_proc_ticket_reply(struct ceph_auth_client *ac,
 		dout(" ticket blob is %d bytes\n", dlen);
 		ceph_decode_need(&tp, tpend, 1 + sizeof(u64), bad);
 		struct_v = ceph_decode_8(&tp);
-		th->secret_id = ceph_decode_64(&tp);
-		ret = ceph_decode_buffer(&th->ticket_blob, &tp, tpend);
+		new_secret_id = ceph_decode_64(&tp);
+		ret = ceph_decode_buffer(&new_ticket_blob, &tp, tpend);
 		if (ret)
 			goto out;
+
+		/* all is well, update our ticket */
+		ceph_crypto_key_destroy(&th->session_key);
+		if (th->ticket_blob)
+			ceph_buffer_put(th->ticket_blob);
+		th->session_key = new_session_key;
+		th->ticket_blob = new_ticket_blob;
+		th->validity = new_validity;
+		th->secret_id = new_secret_id;
+		th->expires = new_expires;
+		th->renew_after = new_renew_after;
 		dout(" got ticket service %d (%s) secret_id %lld len %d\n",
 		     type, ceph_entity_type_name(type), th->secret_id,
 		     (int)th->ticket_blob->vec.iov_len);
@@ -242,7 +264,7 @@ static int ceph_x_build_authorizer(struct ceph_auth_client *ac,
 				   struct ceph_x_ticket_handler *th,
 				   struct ceph_x_authorizer *au)
 {
-	int len;
+	int maxlen;
 	struct ceph_x_authorize_a *msg_a;
 	struct ceph_x_authorize_b msg_b;
 	void *p, *end;
@@ -253,15 +275,15 @@ static int ceph_x_build_authorizer(struct ceph_auth_client *ac,
 	dout("build_authorizer for %s %p\n",
 	     ceph_entity_type_name(th->service), au);
 
-	len = sizeof(*msg_a) + sizeof(msg_b) + sizeof(u32) +
-		ticket_blob_len + 16;
-	dout("  need len %d\n", len);
-	if (au->buf && au->buf->alloc_len < len) {
+	maxlen = sizeof(*msg_a) + sizeof(msg_b) +
+		ceph_x_encrypt_buflen(ticket_blob_len);
+	dout("  need len %d\n", maxlen);
+	if (au->buf && au->buf->alloc_len < maxlen) {
 		ceph_buffer_put(au->buf);
 		au->buf = NULL;
 	}
 	if (!au->buf) {
-		au->buf = ceph_buffer_new(len, GFP_NOFS);
+		au->buf = ceph_buffer_new(maxlen, GFP_NOFS);
 		if (!au->buf)
 			return -ENOMEM;
 	}
@@ -296,6 +318,7 @@ static int ceph_x_build_authorizer(struct ceph_auth_client *ac,
 	au->buf->vec.iov_len = p - au->buf->vec.iov_base;
 	dout(" built authorizer nonce %llx len %d\n", au->nonce,
 	     (int)au->buf->vec.iov_len);
+	BUG_ON(au->buf->vec.iov_len > maxlen);
 	return 0;
 
 out_buf:
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index db122bb357b..7d0a0d0adc1 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -1407,6 +1407,7 @@ static int try_nonblocking_invalidate(struct inode *inode)
  */
 void ceph_check_caps(struct ceph_inode_info *ci, int flags,
 		     struct ceph_mds_session *session)
+	__releases(session->s_mutex)
 {
 	struct ceph_client *client = ceph_inode_to_client(&ci->vfs_inode);
 	struct ceph_mds_client *mdsc = &client->mdsc;
@@ -1414,7 +1415,6 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags,
 	struct ceph_cap *cap;
 	int file_wanted, used;
 	int took_snap_rwsem = 0;             /* true if mdsc->snap_rwsem held */
-	int drop_session_lock = session ? 0 : 1;
 	int issued, implemented, want, retain, revoking, flushing = 0;
 	int mds = -1;   /* keep track of how far we've gone through i_caps list
 			   to avoid an infinite loop on retry */
@@ -1639,7 +1639,7 @@ ack:
 	if (queue_invalidate)
 		ceph_queue_invalidate(inode);
 
-	if (session && drop_session_lock)
+	if (session)
 		mutex_unlock(&session->s_mutex);
 	if (took_snap_rwsem)
 		up_read(&mdsc->snap_rwsem);
@@ -2195,18 +2195,19 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr,
  * Handle a cap GRANT message from the MDS.  (Note that a GRANT may
  * actually be a revocation if it specifies a smaller cap set.)
  *
- * caller holds s_mutex.
+ * caller holds s_mutex and i_lock, we drop both.
+ *
  * return value:
  *  0 - ok
  *  1 - check_caps on auth cap only (writeback)
  *  2 - check_caps (ack revoke)
  */
-static int handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
-			    struct ceph_mds_session *session,
-			    struct ceph_cap *cap,
-			    struct ceph_buffer *xattr_buf)
+static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
+			     struct ceph_mds_session *session,
+			     struct ceph_cap *cap,
+			     struct ceph_buffer *xattr_buf)
 	__releases(inode->i_lock)
-
+	__releases(session->s_mutex)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int mds = session->s_mds;
@@ -2216,7 +2217,7 @@ static int handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 	u64 size = le64_to_cpu(grant->size);
 	u64 max_size = le64_to_cpu(grant->max_size);
 	struct timespec mtime, atime, ctime;
-	int reply = 0;
+	int check_caps = 0;
 	int wake = 0;
 	int writeback = 0;
 	int revoked_rdcache = 0;
@@ -2329,11 +2330,12 @@ static int handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 		if ((used & ~newcaps) & CEPH_CAP_FILE_BUFFER)
 			writeback = 1; /* will delay ack */
 		else if (dirty & ~newcaps)
-			reply = 1;     /* initiate writeback in check_caps */
+			check_caps = 1;  /* initiate writeback in check_caps */
 		else if (((used & ~newcaps) & CEPH_CAP_FILE_CACHE) == 0 ||
 			   revoked_rdcache)
-			reply = 2;     /* send revoke ack in check_caps */
+			check_caps = 2;     /* send revoke ack in check_caps */
 		cap->issued = newcaps;
+		cap->implemented |= newcaps;
 	} else if (cap->issued == newcaps) {
 		dout("caps unchanged: %s -> %s\n",
 		     ceph_cap_string(cap->issued), ceph_cap_string(newcaps));
@@ -2346,6 +2348,7 @@ static int handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 					      * pending revocation */
 		wake = 1;
 	}
+	BUG_ON(cap->issued & ~cap->implemented);
 
 	spin_unlock(&inode->i_lock);
 	if (writeback)
@@ -2359,7 +2362,14 @@ static int handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 		ceph_queue_invalidate(inode);
 	if (wake)
 		wake_up(&ci->i_cap_wq);
-	return reply;
+
+	if (check_caps == 1)
+		ceph_check_caps(ci, CHECK_CAPS_NODELAY|CHECK_CAPS_AUTHONLY,
+				session);
+	else if (check_caps == 2)
+		ceph_check_caps(ci, CHECK_CAPS_NODELAY, session);
+	else
+		mutex_unlock(&session->s_mutex);
 }
 
 /*
@@ -2548,9 +2558,8 @@ static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex,
 			ci->i_cap_exporting_issued = cap->issued;
 		}
 		__ceph_remove_cap(cap);
-	} else {
-		WARN_ON(!cap);
 	}
+	/* else, we already released it */
 
 	spin_unlock(&inode->i_lock);
 }
@@ -2621,9 +2630,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 	u64 cap_id;
 	u64 size, max_size;
 	u64 tid;
-	int check_caps = 0;
 	void *snaptrace;
-	int r;
 
 	dout("handle_caps from mds%d\n", mds);
 
@@ -2668,8 +2675,9 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 	case CEPH_CAP_OP_IMPORT:
 		handle_cap_import(mdsc, inode, h, session,
 				  snaptrace, le32_to_cpu(h->snap_trace_len));
-		check_caps = 1; /* we may have sent a RELEASE to the old auth */
-		goto done;
+		ceph_check_caps(ceph_inode(inode), CHECK_CAPS_NODELAY,
+				session);
+		goto done_unlocked;
 	}
 
 	/* the rest require a cap */
@@ -2686,16 +2694,8 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 	switch (op) {
 	case CEPH_CAP_OP_REVOKE:
 	case CEPH_CAP_OP_GRANT:
-		r = handle_cap_grant(inode, h, session, cap, msg->middle);
-		if (r == 1)
-			ceph_check_caps(ceph_inode(inode),
-					CHECK_CAPS_NODELAY|CHECK_CAPS_AUTHONLY,
-					session);
-		else if (r == 2)
-			ceph_check_caps(ceph_inode(inode),
-					CHECK_CAPS_NODELAY,
-					session);
-		break;
+		handle_cap_grant(inode, h, session, cap, msg->middle);
+		goto done_unlocked;
 
 	case CEPH_CAP_OP_FLUSH_ACK:
 		handle_cap_flush_ack(inode, tid, h, session, cap);
@@ -2713,9 +2713,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 
 done:
 	mutex_unlock(&session->s_mutex);
-
-	if (check_caps)
-		ceph_check_caps(ceph_inode(inode), CHECK_CAPS_NODELAY, NULL);
+done_unlocked:
 	if (inode)
 		iput(inode);
 	return;
@@ -2838,11 +2836,18 @@ int ceph_encode_inode_release(void **p, struct inode *inode,
 	struct ceph_cap *cap;
 	struct ceph_mds_request_release *rel = *p;
 	int ret = 0;
-
-	dout("encode_inode_release %p mds%d drop %s unless %s\n", inode,
-	     mds, ceph_cap_string(drop), ceph_cap_string(unless));
+	int used = 0;
 
 	spin_lock(&inode->i_lock);
+	used = __ceph_caps_used(ci);
+
+	dout("encode_inode_release %p mds%d used %s drop %s unless %s\n", inode,
+	     mds, ceph_cap_string(used), ceph_cap_string(drop),
+	     ceph_cap_string(unless));
+
+	/* only drop unused caps */
+	drop &= ~used;
+
 	cap = __get_cap_for_mds(ci, mds);
 	if (cap && __cap_is_valid(cap)) {
 		if (force ||
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index 5107384ee02..8a9116e15b7 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -288,8 +288,10 @@ more:
 			CEPH_MDS_OP_LSSNAP : CEPH_MDS_OP_READDIR;
 
 		/* discard old result, if any */
-		if (fi->last_readdir)
+		if (fi->last_readdir) {
 			ceph_mdsc_put_request(fi->last_readdir);
+			fi->last_readdir = NULL;
+		}
 
 		/* requery frag tree, as the frag topology may have changed */
 		frag = ceph_choose_frag(ceph_inode(inode), frag, NULL, NULL);
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 7abe1aed819..aca82d55cc5 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -378,6 +378,22 @@ void ceph_destroy_inode(struct inode *inode)
 
 	ceph_queue_caps_release(inode);
 
+	/*
+	 * we may still have a snap_realm reference if there are stray
+	 * caps in i_cap_exporting_issued or i_snap_caps.
+	 */
+	if (ci->i_snap_realm) {
+		struct ceph_mds_client *mdsc =
+			&ceph_client(ci->vfs_inode.i_sb)->mdsc;
+		struct ceph_snap_realm *realm = ci->i_snap_realm;
+
+		dout(" dropping residual ref to snap realm %p\n", realm);
+		spin_lock(&realm->inodes_with_caps_lock);
+		list_del_init(&ci->i_snap_realm_item);
+		spin_unlock(&realm->inodes_with_caps_lock);
+		ceph_put_snap_realm(mdsc, realm);
+	}
+
 	kfree(ci->i_symlink);
 	while ((n = rb_first(&ci->i_fragtree)) != NULL) {
 		frag = rb_entry(n, struct ceph_inode_frag, node);
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index a2600101ec2..5c7920be642 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -328,6 +328,8 @@ static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc,
 	struct ceph_mds_session *s;
 
 	s = kzalloc(sizeof(*s), GFP_NOFS);
+	if (!s)
+		return ERR_PTR(-ENOMEM);
 	s->s_mdsc = mdsc;
 	s->s_mds = mds;
 	s->s_state = CEPH_MDS_SESSION_NEW;
@@ -529,7 +531,7 @@ static void __unregister_request(struct ceph_mds_client *mdsc,
 {
 	dout("__unregister_request %p tid %lld\n", req, req->r_tid);
 	rb_erase(&req->r_node, &mdsc->request_tree);
-	ceph_mdsc_put_request(req);
+	RB_CLEAR_NODE(&req->r_node);
 
 	if (req->r_unsafe_dir) {
 		struct ceph_inode_info *ci = ceph_inode(req->r_unsafe_dir);
@@ -538,6 +540,8 @@ static void __unregister_request(struct ceph_mds_client *mdsc,
 		list_del_init(&req->r_unsafe_dir_item);
 		spin_unlock(&ci->i_unsafe_lock);
 	}
+
+	ceph_mdsc_put_request(req);
 }
 
 /*
@@ -862,6 +866,7 @@ static int send_renew_caps(struct ceph_mds_client *mdsc,
 	if (time_after_eq(jiffies, session->s_cap_ttl) &&
 	    time_after_eq(session->s_cap_ttl, session->s_renew_requested))
 		pr_info("mds%d caps stale\n", session->s_mds);
+	session->s_renew_requested = jiffies;
 
 	/* do not try to renew caps until a recovering mds has reconnected
 	 * with its clients. */
@@ -874,7 +879,6 @@ static int send_renew_caps(struct ceph_mds_client *mdsc,
 
 	dout("send_renew_caps to mds%d (%s)\n", session->s_mds,
 		ceph_mds_state_name(state));
-	session->s_renew_requested = jiffies;
 	msg = create_session_msg(CEPH_SESSION_REQUEST_RENEWCAPS,
 				 ++session->s_renew_seq);
 	if (IS_ERR(msg))
@@ -1566,8 +1570,13 @@ static int __do_request(struct ceph_mds_client *mdsc,
 
 	/* get, open session */
 	session = __ceph_lookup_mds_session(mdsc, mds);
-	if (!session)
+	if (!session) {
 		session = register_session(mdsc, mds);
+		if (IS_ERR(session)) {
+			err = PTR_ERR(session);
+			goto finish;
+		}
+	}
 	dout("do_request mds%d session %p state %s\n", mds, session,
 	     session_state_name(session->s_state));
 	if (session->s_state != CEPH_MDS_SESSION_OPEN &&
@@ -1770,7 +1779,7 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
 	dout("handle_reply %p\n", req);
 
 	/* correct session? */
-	if (!req->r_session && req->r_session != session) {
+	if (req->r_session != session) {
 		pr_err("mdsc_handle_reply got %llu on session mds%d"
 		       " not mds%d\n", tid, session->s_mds,
 		       req->r_session ? req->r_session->s_mds : -1);
@@ -2682,29 +2691,41 @@ void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc)
  */
 static void wait_unsafe_requests(struct ceph_mds_client *mdsc, u64 want_tid)
 {
-	struct ceph_mds_request *req = NULL;
+	struct ceph_mds_request *req = NULL, *nextreq;
 	struct rb_node *n;
 
 	mutex_lock(&mdsc->mutex);
 	dout("wait_unsafe_requests want %lld\n", want_tid);
+restart:
 	req = __get_oldest_req(mdsc);
 	while (req && req->r_tid <= want_tid) {
+		/* find next request */
+		n = rb_next(&req->r_node);
+		if (n)
+			nextreq = rb_entry(n, struct ceph_mds_request, r_node);
+		else
+			nextreq = NULL;
 		if ((req->r_op & CEPH_MDS_OP_WRITE)) {
 			/* write op */
 			ceph_mdsc_get_request(req);
+			if (nextreq)
+				ceph_mdsc_get_request(nextreq);
 			mutex_unlock(&mdsc->mutex);
 			dout("wait_unsafe_requests  wait on %llu (want %llu)\n",
 			     req->r_tid, want_tid);
 			wait_for_completion(&req->r_safe_completion);
 			mutex_lock(&mdsc->mutex);
-			n = rb_next(&req->r_node);
 			ceph_mdsc_put_request(req);
-		} else {
-			n = rb_next(&req->r_node);
+			if (!nextreq)
+				break;  /* next dne before, so we're done! */
+			if (RB_EMPTY_NODE(&nextreq->r_node)) {
+				/* next request was removed from tree */
+				ceph_mdsc_put_request(nextreq);
+				goto restart;
+			}
+			ceph_mdsc_put_request(nextreq);  /* won't go away */
 		}
-		if (!n)
-			break;
-		req = rb_entry(n, struct ceph_mds_request, r_node);
+		req = nextreq;
 	}
 	mutex_unlock(&mdsc->mutex);
 	dout("wait_unsafe_requests done\n");
diff --git a/fs/ceph/messenger.c b/fs/ceph/messenger.c
index 781656a49bf..a32f0f896d9 100644
--- a/fs/ceph/messenger.c
+++ b/fs/ceph/messenger.c
@@ -366,6 +366,14 @@ void ceph_con_open(struct ceph_connection *con, struct ceph_entity_addr *addr)
 }
 
 /*
+ * return true if this connection ever successfully opened
+ */
+bool ceph_con_opened(struct ceph_connection *con)
+{
+	return con->connect_seq > 0;
+}
+
+/*
  * generic get/put
  */
 struct ceph_connection *ceph_con_get(struct ceph_connection *con)
@@ -830,13 +838,6 @@ static void prepare_read_connect(struct ceph_connection *con)
 	con->in_base_pos = 0;
 }
 
-static void prepare_read_connect_retry(struct ceph_connection *con)
-{
-	dout("prepare_read_connect_retry %p\n", con);
-	con->in_base_pos = strlen(CEPH_BANNER) + sizeof(con->actual_peer_addr)
-		+ sizeof(con->peer_addr_for_me);
-}
-
 static void prepare_read_ack(struct ceph_connection *con)
 {
 	dout("prepare_read_ack %p\n", con);
@@ -1146,7 +1147,7 @@ static int process_connect(struct ceph_connection *con)
 		}
 		con->auth_retry = 1;
 		prepare_write_connect(con->msgr, con, 0);
-		prepare_read_connect_retry(con);
+		prepare_read_connect(con);
 		break;
 
 	case CEPH_MSGR_TAG_RESETSESSION:
@@ -1843,8 +1844,6 @@ static void ceph_fault(struct ceph_connection *con)
 		goto out;
 	}
 
-	clear_bit(BUSY, &con->state);  /* to avoid an improbable race */
-
 	mutex_lock(&con->mutex);
 	if (test_bit(CLOSED, &con->state))
 		goto out_unlock;
diff --git a/fs/ceph/messenger.h b/fs/ceph/messenger.h
index 4caaa591111..a343dae73cd 100644
--- a/fs/ceph/messenger.h
+++ b/fs/ceph/messenger.h
@@ -223,6 +223,7 @@ extern void ceph_con_init(struct ceph_messenger *msgr,
 			  struct ceph_connection *con);
 extern void ceph_con_open(struct ceph_connection *con,
 			  struct ceph_entity_addr *addr);
+extern bool ceph_con_opened(struct ceph_connection *con);
 extern void ceph_con_close(struct ceph_connection *con);
 extern void ceph_con_send(struct ceph_connection *con, struct ceph_msg *msg);
 extern void ceph_con_revoke(struct ceph_connection *con, struct ceph_msg *msg);
diff --git a/fs/ceph/osd_client.c b/fs/ceph/osd_client.c
index dbe63db9762..c7b4dedaace 100644
--- a/fs/ceph/osd_client.c
+++ b/fs/ceph/osd_client.c
@@ -413,11 +413,22 @@ static void remove_old_osds(struct ceph_osd_client *osdc, int remove_all)
  */
 static int __reset_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
 {
+	struct ceph_osd_request *req;
 	int ret = 0;
 
 	dout("__reset_osd %p osd%d\n", osd, osd->o_osd);
 	if (list_empty(&osd->o_requests)) {
 		__remove_osd(osdc, osd);
+	} else if (memcmp(&osdc->osdmap->osd_addr[osd->o_osd],
+			  &osd->o_con.peer_addr,
+			  sizeof(osd->o_con.peer_addr)) == 0 &&
+		   !ceph_con_opened(&osd->o_con)) {
+		dout(" osd addr hasn't changed and connection never opened,"
+		     " letting msgr retry");
+		/* touch each r_stamp for handle_timeout()'s benfit */
+		list_for_each_entry(req, &osd->o_requests, r_osd_item)
+			req->r_stamp = jiffies;
+		ret = -EAGAIN;
 	} else {
 		ceph_con_close(&osd->o_con);
 		ceph_con_open(&osd->o_con, &osdc->osdmap->osd_addr[osd->o_osd]);
@@ -633,7 +644,7 @@ static int __send_request(struct ceph_osd_client *osdc,
 	reqhead->flags |= cpu_to_le32(req->r_flags);  /* e.g., RETRY */
 	reqhead->reassert_version = req->r_reassert_version;
 
-	req->r_sent_stamp = jiffies;
+	req->r_stamp = jiffies;
 	list_move_tail(&osdc->req_lru, &req->r_req_lru_item);
 
 	ceph_msg_get(req->r_request); /* send consumes a ref */
@@ -660,7 +671,7 @@ static void handle_timeout(struct work_struct *work)
 	unsigned long timeout = osdc->client->mount_args->osd_timeout * HZ;
 	unsigned long keepalive =
 		osdc->client->mount_args->osd_keepalive_timeout * HZ;
-	unsigned long last_sent = 0;
+	unsigned long last_stamp = 0;
 	struct rb_node *p;
 	struct list_head slow_osds;
 
@@ -697,12 +708,12 @@ static void handle_timeout(struct work_struct *work)
 		req = list_entry(osdc->req_lru.next, struct ceph_osd_request,
 				 r_req_lru_item);
 
-		if (time_before(jiffies, req->r_sent_stamp + timeout))
+		if (time_before(jiffies, req->r_stamp + timeout))
 			break;
 
-		BUG_ON(req == last_req && req->r_sent_stamp == last_sent);
+		BUG_ON(req == last_req && req->r_stamp == last_stamp);
 		last_req = req;
-		last_sent = req->r_sent_stamp;
+		last_stamp = req->r_stamp;
 
 		osd = req->r_osd;
 		BUG_ON(!osd);
@@ -718,7 +729,7 @@ static void handle_timeout(struct work_struct *work)
 	 */
 	INIT_LIST_HEAD(&slow_osds);
 	list_for_each_entry(req, &osdc->req_lru, r_req_lru_item) {
-		if (time_before(jiffies, req->r_sent_stamp + keepalive))
+		if (time_before(jiffies, req->r_stamp + keepalive))
 			break;
 
 		osd = req->r_osd;
@@ -862,7 +873,9 @@ static int __kick_requests(struct ceph_osd_client *osdc,
 
 	dout("kick_requests osd%d\n", kickosd ? kickosd->o_osd : -1);
 	if (kickosd) {
-		__reset_osd(osdc, kickosd);
+		err = __reset_osd(osdc, kickosd);
+		if (err == -EAGAIN)
+			return 1;
 	} else {
 		for (p = rb_first(&osdc->osds); p; p = n) {
 			struct ceph_osd *osd =
@@ -913,7 +926,7 @@ static int __kick_requests(struct ceph_osd_client *osdc,
 
 kick:
 		dout("kicking %p tid %llu osd%d\n", req, req->r_tid,
-		     req->r_osd->o_osd);
+		     req->r_osd ? req->r_osd->o_osd : -1);
 		req->r_flags |= CEPH_OSD_FLAG_RETRY;
 		err = __send_request(osdc, req);
 		if (err) {
diff --git a/fs/ceph/osd_client.h b/fs/ceph/osd_client.h
index 1b1a3ca43af..b0759911e7c 100644
--- a/fs/ceph/osd_client.h
+++ b/fs/ceph/osd_client.h
@@ -70,7 +70,7 @@ struct ceph_osd_request {
 
 	char              r_oid[40];          /* object name */
 	int               r_oid_len;
-	unsigned long     r_sent_stamp;
+	unsigned long     r_stamp;            /* send OR check time */
 	bool              r_resend;           /* msg send failed, needs retry */
 
 	struct ceph_file_layout r_file_layout;
diff --git a/fs/ceph/osdmap.c b/fs/ceph/osdmap.c
index b83f2692b83..d82fe87c2a6 100644
--- a/fs/ceph/osdmap.c
+++ b/fs/ceph/osdmap.c
@@ -480,6 +480,14 @@ static struct ceph_pg_pool_info *__lookup_pg_pool(struct rb_root *root, int id)
 	return NULL;
 }
 
+void __decode_pool(void **p, struct ceph_pg_pool_info *pi)
+{
+	ceph_decode_copy(p, &pi->v, sizeof(pi->v));
+	calc_pg_masks(pi);
+	*p += le32_to_cpu(pi->v.num_snaps) * sizeof(u64);
+	*p += le32_to_cpu(pi->v.num_removed_snap_intervals) * sizeof(u64) * 2;
+}
+
 /*
  * decode a full map.
  */
@@ -526,12 +534,8 @@ struct ceph_osdmap *osdmap_decode(void **p, void *end)
 				   ev, CEPH_PG_POOL_VERSION);
 			goto bad;
 		}
-		ceph_decode_copy(p, &pi->v, sizeof(pi->v));
+		__decode_pool(p, pi);
 		__insert_pg_pool(&map->pg_pools, pi);
-		calc_pg_masks(pi);
-		*p += le32_to_cpu(pi->v.num_snaps) * sizeof(u64);
-		*p += le32_to_cpu(pi->v.num_removed_snap_intervals)
-			* sizeof(u64) * 2;
 	}
 	ceph_decode_32_safe(p, end, map->pool_max, bad);
 
@@ -714,8 +718,7 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end,
 			pi->id = pool;
 			__insert_pg_pool(&map->pg_pools, pi);
 		}
-		ceph_decode_copy(p, &pi->v, sizeof(pi->v));
-		calc_pg_masks(pi);
+		__decode_pool(p, pi);
 	}
 
 	/* old_pool */
diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c
index bf2a5f3846a..df04e210a05 100644
--- a/fs/ceph/snap.c
+++ b/fs/ceph/snap.c
@@ -314,9 +314,9 @@ static int build_snap_context(struct ceph_snap_realm *realm)
 	   because we rebuild_snap_realms() works _downward_ in
 	   hierarchy after each update.) */
 	if (realm->cached_context &&
-	    realm->cached_context->seq <= realm->seq &&
+	    realm->cached_context->seq == realm->seq &&
 	    (!parent ||
-	     realm->cached_context->seq <= parent->cached_context->seq)) {
+	     realm->cached_context->seq >= parent->cached_context->seq)) {
 		dout("build_snap_context %llx %p: %p seq %lld (%d snaps)"
 		     " (unchanged)\n",
 		     realm->ino, realm, realm->cached_context,
@@ -818,7 +818,9 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
 			 * queued (again) by ceph_update_snap_trace()
 			 * below.  Queue it _now_, under the old context.
 			 */
+			spin_lock(&realm->inodes_with_caps_lock);
 			list_del_init(&ci->i_snap_realm_item);
+			spin_unlock(&realm->inodes_with_caps_lock);
 			spin_unlock(&inode->i_lock);
 
 			ceph_queue_cap_snap(ci,
diff --git a/fs/ext3/ialloc.c b/fs/ext3/ialloc.c
index ef9008b885b..0d0e97ed3ff 100644
--- a/fs/ext3/ialloc.c
+++ b/fs/ext3/ialloc.c
@@ -582,7 +582,9 @@ got:
 	inode->i_generation = sbi->s_next_generation++;
 	spin_unlock(&sbi->s_next_gen_lock);
 
-	ei->i_state = EXT3_STATE_NEW;
+	ei->i_state_flags = 0;
+	ext3_set_inode_state(inode, EXT3_STATE_NEW);
+
 	ei->i_extra_isize =
 		(EXT3_INODE_SIZE(inode->i_sb) > EXT3_GOOD_OLD_INODE_SIZE) ?
 		sizeof(struct ext3_inode) - EXT3_GOOD_OLD_INODE_SIZE : 0;
diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c
index 7f920b7263a..ea33bdf0a30 100644
--- a/fs/ext3/inode.c
+++ b/fs/ext3/inode.c
@@ -2811,7 +2811,7 @@ struct inode *ext3_iget(struct super_block *sb, unsigned long ino)
 	inode->i_mtime.tv_sec = (signed)le32_to_cpu(raw_inode->i_mtime);
 	inode->i_atime.tv_nsec = inode->i_ctime.tv_nsec = inode->i_mtime.tv_nsec = 0;
 
-	ei->i_state = 0;
+	ei->i_state_flags = 0;
 	ei->i_dir_start_lookup = 0;
 	ei->i_dtime = le32_to_cpu(raw_inode->i_dtime);
 	/* We now have enough fields to check if the inode was active or not.
diff --git a/fs/fscache/object.c b/fs/fscache/object.c
index e513ac599c8..0b589a9b4ff 100644
--- a/fs/fscache/object.c
+++ b/fs/fscache/object.c
@@ -53,7 +53,7 @@ const char fscache_object_states_short[FSCACHE_OBJECT__NSTATES][5] = {
 static void fscache_object_slow_work_put_ref(struct slow_work *);
 static int  fscache_object_slow_work_get_ref(struct slow_work *);
 static void fscache_object_slow_work_execute(struct slow_work *);
-#ifdef CONFIG_SLOW_WORK_PROC
+#ifdef CONFIG_SLOW_WORK_DEBUG
 static void fscache_object_slow_work_desc(struct slow_work *, struct seq_file *);
 #endif
 static void fscache_initialise_object(struct fscache_object *);
@@ -69,7 +69,7 @@ const struct slow_work_ops fscache_object_slow_work_ops = {
 	.get_ref	= fscache_object_slow_work_get_ref,
 	.put_ref	= fscache_object_slow_work_put_ref,
 	.execute	= fscache_object_slow_work_execute,
-#ifdef CONFIG_SLOW_WORK_PROC
+#ifdef CONFIG_SLOW_WORK_DEBUG
 	.desc		= fscache_object_slow_work_desc,
 #endif
 };
@@ -364,7 +364,7 @@ static void fscache_object_slow_work_execute(struct slow_work *work)
 /*
  * describe an object for slow-work debugging
  */
-#ifdef CONFIG_SLOW_WORK_PROC
+#ifdef CONFIG_SLOW_WORK_DEBUG
 static void fscache_object_slow_work_desc(struct slow_work *work,
 					  struct seq_file *m)
 {
diff --git a/fs/fscache/operation.c b/fs/fscache/operation.c
index 313e79a1426..9f6c928d458 100644
--- a/fs/fscache/operation.c
+++ b/fs/fscache/operation.c
@@ -500,7 +500,7 @@ static void fscache_op_execute(struct slow_work *work)
 /*
  * describe an operation for slow-work debugging
  */
-#ifdef CONFIG_SLOW_WORK_PROC
+#ifdef CONFIG_SLOW_WORK_DEBUG
 static void fscache_op_desc(struct slow_work *work, struct seq_file *m)
 {
 	struct fscache_operation *op =
@@ -517,7 +517,7 @@ const struct slow_work_ops fscache_op_slow_work_ops = {
 	.get_ref	= fscache_op_get_ref,
 	.put_ref	= fscache_op_put_ref,
 	.execute	= fscache_op_execute,
-#ifdef CONFIG_SLOW_WORK_PROC
+#ifdef CONFIG_SLOW_WORK_DEBUG
 	.desc		= fscache_op_desc,
 #endif
 };
diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c
index 0501974bedd..8ccf0f8c9cc 100644
--- a/fs/ocfs2/acl.c
+++ b/fs/ocfs2/acl.c
@@ -30,6 +30,8 @@
 #include "alloc.h"
 #include "dlmglue.h"
 #include "file.h"
+#include "inode.h"
+#include "journal.h"
 #include "ocfs2_fs.h"
 
 #include "xattr.h"
@@ -166,6 +168,60 @@ static struct posix_acl *ocfs2_get_acl(struct inode *inode, int type)
 }
 
 /*
+ * Helper function to set i_mode in memory and disk. Some call paths
+ * will not have di_bh or a journal handle to pass, in which case it
+ * will create it's own.
+ */
+static int ocfs2_acl_set_mode(struct inode *inode, struct buffer_head *di_bh,
+			      handle_t *handle, umode_t new_mode)
+{
+	int ret, commit_handle = 0;
+	struct ocfs2_dinode *di;
+
+	if (di_bh == NULL) {
+		ret = ocfs2_read_inode_block(inode, &di_bh);
+		if (ret) {
+			mlog_errno(ret);
+			goto out;
+		}
+	} else
+		get_bh(di_bh);
+
+	if (handle == NULL) {
+		handle = ocfs2_start_trans(OCFS2_SB(inode->i_sb),
+					   OCFS2_INODE_UPDATE_CREDITS);
+		if (IS_ERR(handle)) {
+			ret = PTR_ERR(handle);
+			mlog_errno(ret);
+			goto out_brelse;
+		}
+
+		commit_handle = 1;
+	}
+
+	di = (struct ocfs2_dinode *)di_bh->b_data;
+	ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret) {
+		mlog_errno(ret);
+		goto out_commit;
+	}
+
+	inode->i_mode = new_mode;
+	di->i_mode = cpu_to_le16(inode->i_mode);
+
+	ocfs2_journal_dirty(handle, di_bh);
+
+out_commit:
+	if (commit_handle)
+		ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
+out_brelse:
+	brelse(di_bh);
+out:
+	return ret;
+}
+
+/*
  * Set the access or default ACL of an inode.
  */
 static int ocfs2_set_acl(handle_t *handle,
@@ -193,9 +249,14 @@ static int ocfs2_set_acl(handle_t *handle,
 			if (ret < 0)
 				return ret;
 			else {
-				inode->i_mode = mode;
 				if (ret == 0)
 					acl = NULL;
+
+				ret = ocfs2_acl_set_mode(inode, di_bh,
+							 handle, mode);
+				if (ret)
+					return ret;
+
 			}
 		}
 		break;
@@ -283,6 +344,7 @@ int ocfs2_init_acl(handle_t *handle,
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	struct posix_acl *acl = NULL;
 	int ret = 0;
+	mode_t mode;
 
 	if (!S_ISLNK(inode->i_mode)) {
 		if (osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) {
@@ -291,12 +353,17 @@ int ocfs2_init_acl(handle_t *handle,
 			if (IS_ERR(acl))
 				return PTR_ERR(acl);
 		}
-		if (!acl)
-			inode->i_mode &= ~current_umask();
+		if (!acl) {
+			mode = inode->i_mode & ~current_umask();
+			ret = ocfs2_acl_set_mode(inode, di_bh, handle, mode);
+			if (ret) {
+				mlog_errno(ret);
+				goto cleanup;
+			}
+		}
 	}
 	if ((osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) && acl) {
 		struct posix_acl *clone;
-		mode_t mode;
 
 		if (S_ISDIR(inode->i_mode)) {
 			ret = ocfs2_set_acl(handle, inode, di_bh,
@@ -313,7 +380,7 @@ int ocfs2_init_acl(handle_t *handle,
 		mode = inode->i_mode;
 		ret = posix_acl_create_masq(clone, &mode);
 		if (ret >= 0) {
-			inode->i_mode = mode;
+			ret = ocfs2_acl_set_mode(inode, di_bh, handle, mode);
 			if (ret > 0) {
 				ret = ocfs2_set_acl(handle, inode,
 						    di_bh, ACL_TYPE_ACCESS,
diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
index a659606dcb9..9289b4357d2 100644
--- a/fs/ocfs2/dlm/dlmmaster.c
+++ b/fs/ocfs2/dlm/dlmmaster.c
@@ -1875,7 +1875,6 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data,
 ok:
 		spin_unlock(&res->spinlock);
 	}
-	spin_unlock(&dlm->spinlock);
 
 	// mlog(0, "woo!  got an assert_master from node %u!\n",
 	// 	     assert->node_idx);
@@ -1926,7 +1925,6 @@ ok:
 		/* master is known, detach if not already detached.
 		 * ensures that only one assert_master call will happen
 		 * on this mle. */
-		spin_lock(&dlm->spinlock);
 		spin_lock(&dlm->master_lock);
 
 		rr = atomic_read(&mle->mle_refs.refcount);
@@ -1959,7 +1957,6 @@ ok:
 			__dlm_put_mle(mle);
 		}
 		spin_unlock(&dlm->master_lock);
-		spin_unlock(&dlm->spinlock);
 	} else if (res) {
 		if (res->owner != assert->node_idx) {
 			mlog(0, "assert_master from %u, but current "
@@ -1967,6 +1964,7 @@ ok:
 			     res->owner, namelen, name);
 		}
 	}
+	spin_unlock(&dlm->spinlock);
 
 done:
 	ret = 0;
diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index 278a223aae1..ab207901d32 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -891,6 +891,21 @@ static int ocfs2_query_inode_wipe(struct inode *inode,
 	/* Do some basic inode verification... */
 	di = (struct ocfs2_dinode *) di_bh->b_data;
 	if (!(di->i_flags & cpu_to_le32(OCFS2_ORPHANED_FL))) {
+		/*
+		 * Inodes in the orphan dir must have ORPHANED_FL.  The only
+		 * inodes that come back out of the orphan dir are reflink
+		 * targets. A reflink target may be moved out of the orphan
+		 * dir between the time we scan the directory and the time we
+		 * process it. This would lead to HAS_REFCOUNT_FL being set but
+		 * ORPHANED_FL not.
+		 */
+		if (di->i_dyn_features & cpu_to_le16(OCFS2_HAS_REFCOUNT_FL)) {
+			mlog(0, "Reflinked inode %llu is no longer orphaned.  "
+			     "it shouldn't be deleted\n",
+			     (unsigned long long)oi->ip_blkno);
+			goto bail;
+		}
+
 		/* for lack of a better error? */
 		status = -EEXIST;
 		mlog(ML_ERROR,
diff --git a/fs/ocfs2/localalloc.c b/fs/ocfs2/localalloc.c
index ca992d91f51..c983715d8d8 100644
--- a/fs/ocfs2/localalloc.c
+++ b/fs/ocfs2/localalloc.c
@@ -872,8 +872,10 @@ static int ocfs2_sync_local_to_main(struct ocfs2_super *osb,
 			     (unsigned long long)la_start_blk,
 			     (unsigned long long)blkno);
 
-			status = ocfs2_free_clusters(handle, main_bm_inode,
-						     main_bm_bh, blkno, count);
+			status = ocfs2_release_clusters(handle,
+							main_bm_inode,
+							main_bm_bh, blkno,
+							count);
 			if (status < 0) {
 				mlog_errno(status);
 				goto bail;
@@ -984,8 +986,7 @@ static int ocfs2_local_alloc_reserve_for_window(struct ocfs2_super *osb,
 	}
 
 retry_enospc:
-	(*ac)->ac_bits_wanted = osb->local_alloc_bits;
-
+	(*ac)->ac_bits_wanted = osb->local_alloc_default_bits;
 	status = ocfs2_reserve_cluster_bitmap_bits(osb, *ac);
 	if (status == -ENOSPC) {
 		if (ocfs2_recalc_la_window(osb, OCFS2_LA_EVENT_ENOSPC) ==
@@ -1061,6 +1062,7 @@ retry_enospc:
 		    OCFS2_LA_DISABLED)
 			goto bail;
 
+		ac->ac_bits_wanted = osb->local_alloc_default_bits;
 		status = ocfs2_claim_clusters(osb, handle, ac,
 					      osb->local_alloc_bits,
 					      &cluster_off,
diff --git a/fs/ocfs2/locks.c b/fs/ocfs2/locks.c
index 544ac624517..b5cb3ede940 100644
--- a/fs/ocfs2/locks.c
+++ b/fs/ocfs2/locks.c
@@ -133,7 +133,7 @@ int ocfs2_lock(struct file *file, int cmd, struct file_lock *fl)
 
 	if (!(fl->fl_flags & FL_POSIX))
 		return -ENOLCK;
-	if (__mandatory_lock(inode))
+	if (__mandatory_lock(inode) && fl->fl_type != F_UNLCK)
 		return -ENOLCK;
 
 	return ocfs2_plock(osb->cconn, OCFS2_I(inode)->ip_blkno, file, cmd, fl);
diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c
index d9cd4e373a5..b1eb50ae409 100644
--- a/fs/ocfs2/namei.c
+++ b/fs/ocfs2/namei.c
@@ -84,7 +84,7 @@ static int ocfs2_prepare_orphan_dir(struct ocfs2_super *osb,
 static int ocfs2_orphan_add(struct ocfs2_super *osb,
 			    handle_t *handle,
 			    struct inode *inode,
-			    struct ocfs2_dinode *fe,
+			    struct buffer_head *fe_bh,
 			    char *name,
 			    struct ocfs2_dir_lookup_result *lookup,
 			    struct inode *orphan_dir_inode);
@@ -879,7 +879,7 @@ static int ocfs2_unlink(struct inode *dir,
 	fe = (struct ocfs2_dinode *) fe_bh->b_data;
 
 	if (inode_is_unlinkable(inode)) {
-		status = ocfs2_orphan_add(osb, handle, inode, fe, orphan_name,
+		status = ocfs2_orphan_add(osb, handle, inode, fe_bh, orphan_name,
 					  &orphan_insert, orphan_dir);
 		if (status < 0) {
 			mlog_errno(status);
@@ -1300,7 +1300,7 @@ static int ocfs2_rename(struct inode *old_dir,
 		if (S_ISDIR(new_inode->i_mode) ||
 		    (ocfs2_read_links_count(newfe) == 1)) {
 			status = ocfs2_orphan_add(osb, handle, new_inode,
-						  newfe, orphan_name,
+						  newfe_bh, orphan_name,
 						  &orphan_insert, orphan_dir);
 			if (status < 0) {
 				mlog_errno(status);
@@ -1911,7 +1911,7 @@ leave:
 static int ocfs2_orphan_add(struct ocfs2_super *osb,
 			    handle_t *handle,
 			    struct inode *inode,
-			    struct ocfs2_dinode *fe,
+			    struct buffer_head *fe_bh,
 			    char *name,
 			    struct ocfs2_dir_lookup_result *lookup,
 			    struct inode *orphan_dir_inode)
@@ -1919,6 +1919,7 @@ static int ocfs2_orphan_add(struct ocfs2_super *osb,
 	struct buffer_head *orphan_dir_bh = NULL;
 	int status = 0;
 	struct ocfs2_dinode *orphan_fe;
+	struct ocfs2_dinode *fe = (struct ocfs2_dinode *) fe_bh->b_data;
 
 	mlog_entry("(inode->i_ino = %lu)\n", inode->i_ino);
 
@@ -1959,6 +1960,21 @@ static int ocfs2_orphan_add(struct ocfs2_super *osb,
 		goto leave;
 	}
 
+	/*
+	 * We're going to journal the change of i_flags and i_orphaned_slot.
+	 * It's safe anyway, though some callers may duplicate the journaling.
+	 * Journaling within the func just make the logic look more
+	 * straightforward.
+	 */
+	status = ocfs2_journal_access_di(handle,
+					 INODE_CACHE(inode),
+					 fe_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
+	if (status < 0) {
+		mlog_errno(status);
+		goto leave;
+	}
+
 	le32_add_cpu(&fe->i_flags, OCFS2_ORPHANED_FL);
 
 	/* Record which orphan dir our inode now resides
@@ -1966,6 +1982,8 @@ static int ocfs2_orphan_add(struct ocfs2_super *osb,
 	 * dir to lock. */
 	fe->i_orphaned_slot = cpu_to_le16(osb->slot_num);
 
+	ocfs2_journal_dirty(handle, fe_bh);
+
 	mlog(0, "Inode %llu orphaned in slot %d\n",
 	     (unsigned long long)OCFS2_I(inode)->ip_blkno, osb->slot_num);
 
@@ -2123,7 +2141,7 @@ int ocfs2_create_inode_in_orphan(struct inode *dir,
 	}
 
 	di = (struct ocfs2_dinode *)new_di_bh->b_data;
-	status = ocfs2_orphan_add(osb, handle, inode, di, orphan_name,
+	status = ocfs2_orphan_add(osb, handle, inode, new_di_bh, orphan_name,
 				  &orphan_insert, orphan_dir);
 	if (status < 0) {
 		mlog_errno(status);
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index 1238b491db9..adf5e2ebc2c 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -763,8 +763,18 @@ static inline unsigned int ocfs2_megabytes_to_clusters(struct super_block *sb,
 	return megs << (20 - OCFS2_SB(sb)->s_clustersize_bits);
 }
 
-#define ocfs2_set_bit ext2_set_bit
-#define ocfs2_clear_bit ext2_clear_bit
+static inline void _ocfs2_set_bit(unsigned int bit, unsigned long *bitmap)
+{
+	ext2_set_bit(bit, bitmap);
+}
+#define ocfs2_set_bit(bit, addr) _ocfs2_set_bit((bit), (unsigned long *)(addr))
+
+static inline void _ocfs2_clear_bit(unsigned int bit, unsigned long *bitmap)
+{
+	ext2_clear_bit(bit, bitmap);
+}
+#define ocfs2_clear_bit(bit, addr) _ocfs2_clear_bit((bit), (unsigned long *)(addr))
+
 #define ocfs2_test_bit ext2_test_bit
 #define ocfs2_find_next_zero_bit ext2_find_next_zero_bit
 #define ocfs2_find_next_bit ext2_find_next_bit
diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index 9e96921dffd..29405f2ff61 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -4075,6 +4075,7 @@ static int ocfs2_complete_reflink(struct inode *s_inode,
 	OCFS2_I(t_inode)->ip_dyn_features = OCFS2_I(s_inode)->ip_dyn_features;
 	spin_unlock(&OCFS2_I(t_inode)->ip_lock);
 	i_size_write(t_inode, size);
+	t_inode->i_blocks = s_inode->i_blocks;
 
 	di->i_xattr_inline_size = s_di->i_xattr_inline_size;
 	di->i_clusters = s_di->i_clusters;
diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c
index c3c60bc3e07..19ba00f2854 100644
--- a/fs/ocfs2/suballoc.c
+++ b/fs/ocfs2/suballoc.c
@@ -95,13 +95,6 @@ static inline int ocfs2_block_group_set_bits(handle_t *handle,
 					     struct buffer_head *group_bh,
 					     unsigned int bit_off,
 					     unsigned int num_bits);
-static inline int ocfs2_block_group_clear_bits(handle_t *handle,
-					       struct inode *alloc_inode,
-					       struct ocfs2_group_desc *bg,
-					       struct buffer_head *group_bh,
-					       unsigned int bit_off,
-					       unsigned int num_bits);
-
 static int ocfs2_relink_block_group(handle_t *handle,
 				    struct inode *alloc_inode,
 				    struct buffer_head *fe_bh,
@@ -152,7 +145,7 @@ static u32 ocfs2_bits_per_group(struct ocfs2_chain_list *cl)
 
 #define do_error(fmt, ...)						\
 	do{								\
-		if (clean_error)					\
+		if (resize)					\
 			mlog(ML_ERROR, fmt "\n", ##__VA_ARGS__);	\
 		else							\
 			ocfs2_error(sb, fmt, ##__VA_ARGS__);		\
@@ -160,7 +153,7 @@ static u32 ocfs2_bits_per_group(struct ocfs2_chain_list *cl)
 
 static int ocfs2_validate_gd_self(struct super_block *sb,
 				  struct buffer_head *bh,
-				  int clean_error)
+				  int resize)
 {
 	struct ocfs2_group_desc *gd = (struct ocfs2_group_desc *)bh->b_data;
 
@@ -211,7 +204,7 @@ static int ocfs2_validate_gd_self(struct super_block *sb,
 static int ocfs2_validate_gd_parent(struct super_block *sb,
 				    struct ocfs2_dinode *di,
 				    struct buffer_head *bh,
-				    int clean_error)
+				    int resize)
 {
 	unsigned int max_bits;
 	struct ocfs2_group_desc *gd = (struct ocfs2_group_desc *)bh->b_data;
@@ -233,8 +226,11 @@ static int ocfs2_validate_gd_parent(struct super_block *sb,
 		return -EINVAL;
 	}
 
-	if (le16_to_cpu(gd->bg_chain) >=
-	    le16_to_cpu(di->id2.i_chain.cl_next_free_rec)) {
+	/* In resize, we may meet the case bg_chain == cl_next_free_rec. */
+	if ((le16_to_cpu(gd->bg_chain) >
+	     le16_to_cpu(di->id2.i_chain.cl_next_free_rec)) ||
+	    ((le16_to_cpu(gd->bg_chain) ==
+	     le16_to_cpu(di->id2.i_chain.cl_next_free_rec)) && !resize)) {
 		do_error("Group descriptor #%llu has bad chain %u",
 			 (unsigned long long)bh->b_blocknr,
 			 le16_to_cpu(gd->bg_chain));
@@ -1975,18 +1971,18 @@ int ocfs2_claim_clusters(struct ocfs2_super *osb,
 				      bits_wanted, cluster_start, num_clusters);
 }
 
-static inline int ocfs2_block_group_clear_bits(handle_t *handle,
-					       struct inode *alloc_inode,
-					       struct ocfs2_group_desc *bg,
-					       struct buffer_head *group_bh,
-					       unsigned int bit_off,
-					       unsigned int num_bits)
+static int ocfs2_block_group_clear_bits(handle_t *handle,
+					struct inode *alloc_inode,
+					struct ocfs2_group_desc *bg,
+					struct buffer_head *group_bh,
+					unsigned int bit_off,
+					unsigned int num_bits,
+					void (*undo_fn)(unsigned int bit,
+							unsigned long *bmap))
 {
 	int status;
 	unsigned int tmp;
-	int journal_type = OCFS2_JOURNAL_ACCESS_WRITE;
 	struct ocfs2_group_desc *undo_bg = NULL;
-	int cluster_bitmap = 0;
 
 	mlog_entry_void();
 
@@ -1996,20 +1992,18 @@ static inline int ocfs2_block_group_clear_bits(handle_t *handle,
 
 	mlog(0, "off = %u, num = %u\n", bit_off, num_bits);
 
-	if (ocfs2_is_cluster_bitmap(alloc_inode))
-		journal_type = OCFS2_JOURNAL_ACCESS_UNDO;
-
+	BUG_ON(undo_fn && !ocfs2_is_cluster_bitmap(alloc_inode));
 	status = ocfs2_journal_access_gd(handle, INODE_CACHE(alloc_inode),
-					 group_bh, journal_type);
+					 group_bh,
+					 undo_fn ?
+					 OCFS2_JOURNAL_ACCESS_UNDO :
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
 	}
 
-	if (ocfs2_is_cluster_bitmap(alloc_inode))
-		cluster_bitmap = 1;
-
-	if (cluster_bitmap) {
+	if (undo_fn) {
 		jbd_lock_bh_state(group_bh);
 		undo_bg = (struct ocfs2_group_desc *)
 					bh2jh(group_bh)->b_committed_data;
@@ -2020,13 +2014,13 @@ static inline int ocfs2_block_group_clear_bits(handle_t *handle,
 	while(tmp--) {
 		ocfs2_clear_bit((bit_off + tmp),
 				(unsigned long *) bg->bg_bitmap);
-		if (cluster_bitmap)
-			ocfs2_set_bit(bit_off + tmp,
-				      (unsigned long *) undo_bg->bg_bitmap);
+		if (undo_fn)
+			undo_fn(bit_off + tmp,
+				(unsigned long *) undo_bg->bg_bitmap);
 	}
 	le16_add_cpu(&bg->bg_free_bits_count, num_bits);
 
-	if (cluster_bitmap)
+	if (undo_fn)
 		jbd_unlock_bh_state(group_bh);
 
 	status = ocfs2_journal_dirty(handle, group_bh);
@@ -2039,12 +2033,14 @@ bail:
 /*
  * expects the suballoc inode to already be locked.
  */
-int ocfs2_free_suballoc_bits(handle_t *handle,
-			     struct inode *alloc_inode,
-			     struct buffer_head *alloc_bh,
-			     unsigned int start_bit,
-			     u64 bg_blkno,
-			     unsigned int count)
+static int _ocfs2_free_suballoc_bits(handle_t *handle,
+				     struct inode *alloc_inode,
+				     struct buffer_head *alloc_bh,
+				     unsigned int start_bit,
+				     u64 bg_blkno,
+				     unsigned int count,
+				     void (*undo_fn)(unsigned int bit,
+						     unsigned long *bitmap))
 {
 	int status = 0;
 	u32 tmp_used;
@@ -2079,7 +2075,7 @@ int ocfs2_free_suballoc_bits(handle_t *handle,
 
 	status = ocfs2_block_group_clear_bits(handle, alloc_inode,
 					      group, group_bh,
-					      start_bit, count);
+					      start_bit, count, undo_fn);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -2110,6 +2106,17 @@ bail:
 	return status;
 }
 
+int ocfs2_free_suballoc_bits(handle_t *handle,
+			     struct inode *alloc_inode,
+			     struct buffer_head *alloc_bh,
+			     unsigned int start_bit,
+			     u64 bg_blkno,
+			     unsigned int count)
+{
+	return _ocfs2_free_suballoc_bits(handle, alloc_inode, alloc_bh,
+					 start_bit, bg_blkno, count, NULL);
+}
+
 int ocfs2_free_dinode(handle_t *handle,
 		      struct inode *inode_alloc_inode,
 		      struct buffer_head *inode_alloc_bh,
@@ -2123,11 +2130,13 @@ int ocfs2_free_dinode(handle_t *handle,
 					inode_alloc_bh, bit, bg_blkno, 1);
 }
 
-int ocfs2_free_clusters(handle_t *handle,
-		       struct inode *bitmap_inode,
-		       struct buffer_head *bitmap_bh,
-		       u64 start_blk,
-		       unsigned int num_clusters)
+static int _ocfs2_free_clusters(handle_t *handle,
+				struct inode *bitmap_inode,
+				struct buffer_head *bitmap_bh,
+				u64 start_blk,
+				unsigned int num_clusters,
+				void (*undo_fn)(unsigned int bit,
+						unsigned long *bitmap))
 {
 	int status;
 	u16 bg_start_bit;
@@ -2154,9 +2163,9 @@ int ocfs2_free_clusters(handle_t *handle,
 	mlog(0, "bg_blkno = %llu, bg_start_bit = %u\n",
 	     (unsigned long long)bg_blkno, bg_start_bit);
 
-	status = ocfs2_free_suballoc_bits(handle, bitmap_inode, bitmap_bh,
-					  bg_start_bit, bg_blkno,
-					  num_clusters);
+	status = _ocfs2_free_suballoc_bits(handle, bitmap_inode, bitmap_bh,
+					   bg_start_bit, bg_blkno,
+					   num_clusters, undo_fn);
 	if (status < 0) {
 		mlog_errno(status);
 		goto out;
@@ -2170,6 +2179,32 @@ out:
 	return status;
 }
 
+int ocfs2_free_clusters(handle_t *handle,
+			struct inode *bitmap_inode,
+			struct buffer_head *bitmap_bh,
+			u64 start_blk,
+			unsigned int num_clusters)
+{
+	return _ocfs2_free_clusters(handle, bitmap_inode, bitmap_bh,
+				    start_blk, num_clusters,
+				    _ocfs2_set_bit);
+}
+
+/*
+ * Give never-used clusters back to the global bitmap.  We don't need
+ * to protect these bits in the undo buffer.
+ */
+int ocfs2_release_clusters(handle_t *handle,
+			   struct inode *bitmap_inode,
+			   struct buffer_head *bitmap_bh,
+			   u64 start_blk,
+			   unsigned int num_clusters)
+{
+	return _ocfs2_free_clusters(handle, bitmap_inode, bitmap_bh,
+				    start_blk, num_clusters,
+				    _ocfs2_clear_bit);
+}
+
 static inline void ocfs2_debug_bg(struct ocfs2_group_desc *bg)
 {
 	printk("Block Group:\n");
diff --git a/fs/ocfs2/suballoc.h b/fs/ocfs2/suballoc.h
index fa60723c43e..e0f46df357e 100644
--- a/fs/ocfs2/suballoc.h
+++ b/fs/ocfs2/suballoc.h
@@ -127,6 +127,11 @@ int ocfs2_free_clusters(handle_t *handle,
 			struct buffer_head *bitmap_bh,
 			u64 start_blk,
 			unsigned int num_clusters);
+int ocfs2_release_clusters(handle_t *handle,
+			   struct inode *bitmap_inode,
+			   struct buffer_head *bitmap_bh,
+			   u64 start_blk,
+			   unsigned int num_clusters);
 
 static inline u64 ocfs2_which_suballoc_group(u64 block, unsigned int bit)
 {
diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c
index d1b0d386f6d..3e7773089b9 100644
--- a/fs/ocfs2/xattr.c
+++ b/fs/ocfs2/xattr.c
@@ -1622,7 +1622,7 @@ static void ocfs2_xa_block_wipe_namevalue(struct ocfs2_xa_loc *loc)
 	/* Now tell xh->xh_entries about it */
 	for (i = 0; i < count; i++) {
 		offset = le16_to_cpu(xh->xh_entries[i].xe_name_offset);
-		if (offset < namevalue_offset)
+		if (offset <= namevalue_offset)
 			le16_add_cpu(&xh->xh_entries[i].xe_name_offset,
 				     namevalue_size);
 	}
@@ -6528,13 +6528,11 @@ static int ocfs2_create_empty_xattr_block(struct inode *inode,
 					  int indexed)
 {
 	int ret;
-	struct ocfs2_alloc_context *meta_ac;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	struct ocfs2_xattr_set_ctxt ctxt = {
-		.meta_ac = meta_ac,
-	};
+	struct ocfs2_xattr_set_ctxt ctxt;
 
-	ret = ocfs2_reserve_new_metadata_blocks(osb, 1, &meta_ac);
+	memset(&ctxt, 0, sizeof(ctxt));
+	ret = ocfs2_reserve_new_metadata_blocks(osb, 1, &ctxt.meta_ac);
 	if (ret < 0) {
 		mlog_errno(ret);
 		return ret;
@@ -6556,7 +6554,7 @@ static int ocfs2_create_empty_xattr_block(struct inode *inode,
 
 	ocfs2_commit_trans(osb, ctxt.handle);
 out:
-	ocfs2_free_alloc_context(meta_ac);
+	ocfs2_free_alloc_context(ctxt.meta_ac);
 	return ret;
 }