Merge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/ocfs2

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/ocfs2: (22 commits)
  MAINTAINERS: Update Joel Becker's email address
  ocfs2: Remove unused truncate function from alloc.c
  ocfs2/cluster: dereferencing before checking in nst_seq_show()
  ocfs2: fix build for OCFS2_FS_STATS not enabled
  ocfs2/cluster: Show o2net timing statistics
  ocfs2/cluster: Track process message timing stats for each socket
  ocfs2/cluster: Track send message timing stats for each socket
  ocfs2/cluster: Use ktime instead of timeval in struct o2net_sock_container
  ocfs2/cluster: Replace timeval with ktime in struct o2net_send_tracking
  ocfs2: Add DEBUG_FS dependency
  ocfs2/dlm: Hard code the values for enums
  ocfs2/dlm: Minor cleanup
  ocfs2/dlm: Cleanup dlmdebug.c
  ocfs2: Release buffer_head in case of error in ocfs2_double_lock.
  ocfs2/cluster: Pin the local node when o2hb thread starts
  ocfs2/cluster: Show pin state for each o2hb region
  ocfs2/cluster: Pin/unpin o2hb regions
  ocfs2/cluster: Remove dropped region from o2hb quorum region bitmap
  ocfs2/cluster: Pin the remote node item in configfs
  ocfs2/dlm: make existing convertion precedent over new lock
  ...

MAINTAINERS

@@ -1785,7 +1785,8 @@ S:	Maintained
 F:	drivers/usb/atm/cxacru.c
 
 CONFIGFS
-M:	Joel Becker <joel.becker@oracle.com>
+M:	Joel Becker <jlbec@evilplan.org>
+T:	git git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/configfs.git
 S:	Supported
 F:	fs/configfs/
 F:	include/linux/configfs.h
@@ -4549,7 +4550,7 @@ F:	include/linux/oprofile.h
 
 ORACLE CLUSTER FILESYSTEM 2 (OCFS2)
 M:	Mark Fasheh <mfasheh@suse.com>
-M:	Joel Becker <joel.becker@oracle.com>
+M:	Joel Becker <jlbec@evilplan.org>
 L:	ocfs2-devel@oss.oracle.com (moderated for non-subscribers)
 W:	http://oss.oracle.com/projects/ocfs2/
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/ocfs2.git

fs/ocfs2/Kconfig

@@ -51,7 +51,7 @@ config OCFS2_FS_USERSPACE_CLUSTER
 
 config OCFS2_FS_STATS
 	bool "OCFS2 statistics"
-	depends on OCFS2_FS
+	depends on OCFS2_FS && DEBUG_FS
 	default y
 	help
 	  This option allows some fs statistics to be captured. Enabling

fs/ocfs2/alloc.c

@@ -565,7 +565,6 @@ static inline int ocfs2_et_sanity_check(struct ocfs2_extent_tree *et)
 	return ret;
 }
 
-static void ocfs2_free_truncate_context(struct ocfs2_truncate_context *tc);
 static int ocfs2_cache_extent_block_free(struct ocfs2_cached_dealloc_ctxt *ctxt,
 					 struct ocfs2_extent_block *eb);
 static void ocfs2_adjust_rightmost_records(handle_t *handle,
@@ -5858,6 +5857,7 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb,
 
 	ocfs2_journal_dirty(handle, tl_bh);
 
+	osb->truncated_clusters += num_clusters;
 bail:
 	mlog_exit(status);
 	return status;
@@ -5929,6 +5929,8 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb,
 		i--;
 	}
 
+	osb->truncated_clusters = 0;
+
 bail:
 	mlog_exit(status);
 	return status;
@@ -7138,64 +7140,6 @@ int ocfs2_commit_truncate(struct ocfs2_super *osb,
 	return status;
 }
 
-/*
- * Expects the inode to already be locked.
- */
-int ocfs2_prepare_truncate(struct ocfs2_super *osb,
-			   struct inode *inode,
-			   struct buffer_head *fe_bh,
-			   struct ocfs2_truncate_context **tc)
-{
-	int status;
-	unsigned int new_i_clusters;
-	struct ocfs2_dinode *fe;
-	struct ocfs2_extent_block *eb;
-	struct buffer_head *last_eb_bh = NULL;
-
-	mlog_entry_void();
-
-	*tc = NULL;
-
-	new_i_clusters = ocfs2_clusters_for_bytes(osb->sb,
-						  i_size_read(inode));
-	fe = (struct ocfs2_dinode *) fe_bh->b_data;
-
-	mlog(0, "fe->i_clusters = %u, new_i_clusters = %u, fe->i_size ="
-	     "%llu\n", le32_to_cpu(fe->i_clusters), new_i_clusters,
-	     (unsigned long long)le64_to_cpu(fe->i_size));
-
-	*tc = kzalloc(sizeof(struct ocfs2_truncate_context), GFP_KERNEL);
-	if (!(*tc)) {
-		status = -ENOMEM;
-		mlog_errno(status);
-		goto bail;
-	}
-	ocfs2_init_dealloc_ctxt(&(*tc)->tc_dealloc);
-
-	if (fe->id2.i_list.l_tree_depth) {
-		status = ocfs2_read_extent_block(INODE_CACHE(inode),
-						 le64_to_cpu(fe->i_last_eb_blk),
-						 &last_eb_bh);
-		if (status < 0) {
-			mlog_errno(status);
-			goto bail;
-		}
-		eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
-	}
-
-	(*tc)->tc_last_eb_bh = last_eb_bh;
-
-	status = 0;
-bail:
-	if (status < 0) {
-		if (*tc)
-			ocfs2_free_truncate_context(*tc);
-		*tc = NULL;
-	}
-	mlog_exit_void();
-	return status;
-}
-
 /*
  * 'start' is inclusive, 'end' is not.
  */
@@ -7270,18 +7214,3 @@ int ocfs2_truncate_inline(struct inode *inode, struct buffer_head *di_bh,
 out:
 	return ret;
 }
-
-static void ocfs2_free_truncate_context(struct ocfs2_truncate_context *tc)
-{
-	/*
-	 * The caller is responsible for completing deallocation
-	 * before freeing the context.
-	 */
-	if (tc->tc_dealloc.c_first_suballocator != NULL)
-		mlog(ML_NOTICE,
-		     "Truncate completion has non-empty dealloc context\n");
-
-	brelse(tc->tc_last_eb_bh);
-
-	kfree(tc);
-}

fs/ocfs2/alloc.h

@@ -228,10 +228,6 @@ struct ocfs2_truncate_context {
 
 int ocfs2_zero_range_for_truncate(struct inode *inode, handle_t *handle,
 				  u64 range_start, u64 range_end);
-int ocfs2_prepare_truncate(struct ocfs2_super *osb,
-			   struct inode *inode,
-			   struct buffer_head *fe_bh,
-			   struct ocfs2_truncate_context **tc);
 int ocfs2_commit_truncate(struct ocfs2_super *osb,
 			  struct inode *inode,
 			  struct buffer_head *di_bh);

fs/ocfs2/aops.c

@@ -1630,6 +1630,43 @@ static int ocfs2_zero_tail(struct inode *inode, struct buffer_head *di_bh,
 	return ret;
 }
 
+/*
+ * Try to flush truncate logs if we can free enough clusters from it.
+ * As for return value, "< 0" means error, "0" no space and "1" means
+ * we have freed enough spaces and let the caller try to allocate again.
+ */
+static int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb,
+					  unsigned int needed)
+{
+	tid_t target;
+	int ret = 0;
+	unsigned int truncated_clusters;
+
+	mutex_lock(&osb->osb_tl_inode->i_mutex);
+	truncated_clusters = osb->truncated_clusters;
+	mutex_unlock(&osb->osb_tl_inode->i_mutex);
+
+	/*
+	 * Check whether we can succeed in allocating if we free
+	 * the truncate log.
+	 */
+	if (truncated_clusters < needed)
+		goto out;
+
+	ret = ocfs2_flush_truncate_log(osb);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
+
+	if (jbd2_journal_start_commit(osb->journal->j_journal, &target)) {
+		jbd2_log_wait_commit(osb->journal->j_journal, target);
+		ret = 1;
+	}
+out:
+	return ret;
+}
+
 int ocfs2_write_begin_nolock(struct file *filp,
 			     struct address_space *mapping,
 			     loff_t pos, unsigned len, unsigned flags,
@@ -1637,7 +1674,7 @@ int ocfs2_write_begin_nolock(struct file *filp,
 			     struct buffer_head *di_bh, struct page *mmap_page)
 {
 	int ret, cluster_of_pages, credits = OCFS2_INODE_UPDATE_CREDITS;
-	unsigned int clusters_to_alloc, extents_to_split;
+	unsigned int clusters_to_alloc, extents_to_split, clusters_need = 0;
 	struct ocfs2_write_ctxt *wc;
 	struct inode *inode = mapping->host;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
@@ -1646,7 +1683,9 @@ int ocfs2_write_begin_nolock(struct file *filp,
 	struct ocfs2_alloc_context *meta_ac = NULL;
 	handle_t *handle;
 	struct ocfs2_extent_tree et;
+	int try_free = 1, ret1;
 
+try_again:
 	ret = ocfs2_alloc_write_ctxt(&wc, osb, pos, len, di_bh);
 	if (ret) {
 		mlog_errno(ret);
@@ -1681,6 +1720,7 @@ int ocfs2_write_begin_nolock(struct file *filp,
 		mlog_errno(ret);
 		goto out;
 	} else if (ret == 1) {
+		clusters_need = wc->w_clen;
 		ret = ocfs2_refcount_cow(inode, filp, di_bh,
 					 wc->w_cpos, wc->w_clen, UINT_MAX);
 		if (ret) {
@@ -1695,6 +1735,7 @@ int ocfs2_write_begin_nolock(struct file *filp,
 		mlog_errno(ret);
 		goto out;
 	}
+	clusters_need += clusters_to_alloc;
 
 	di = (struct ocfs2_dinode *)wc->w_di_bh->b_data;
 
@@ -1817,6 +1858,22 @@ int ocfs2_write_begin_nolock(struct file *filp,
 		ocfs2_free_alloc_context(data_ac);
 	if (meta_ac)
 		ocfs2_free_alloc_context(meta_ac);
+
+	if (ret == -ENOSPC && try_free) {
+		/*
+		 * Try to free some truncate log so that we can have enough
+		 * clusters to allocate.
+		 */
+		try_free = 0;
+
+		ret1 = ocfs2_try_to_free_truncate_log(osb, clusters_need);
+		if (ret1 == 1)
+			goto try_again;
+
+		if (ret1 < 0)
+			mlog_errno(ret1);
+	}
+
 	return ret;
 }
 

fs/ocfs2/cluster/tcp.c

@@ -153,63 +153,114 @@ static void o2net_init_nst(struct o2net_send_tracking *nst, u32 msgtype,
 	nst->st_node = node;
 }
 
-static void o2net_set_nst_sock_time(struct o2net_send_tracking *nst)
+static inline void o2net_set_nst_sock_time(struct o2net_send_tracking *nst)
 {
-	do_gettimeofday(&nst->st_sock_time);
+	nst->st_sock_time = ktime_get();
 }
 
-static void o2net_set_nst_send_time(struct o2net_send_tracking *nst)
+static inline void o2net_set_nst_send_time(struct o2net_send_tracking *nst)
 {
-	do_gettimeofday(&nst->st_send_time);
+	nst->st_send_time = ktime_get();
 }
 
-static void o2net_set_nst_status_time(struct o2net_send_tracking *nst)
+static inline void o2net_set_nst_status_time(struct o2net_send_tracking *nst)
 {
-	do_gettimeofday(&nst->st_status_time);
+	nst->st_status_time = ktime_get();
 }
 
-static void o2net_set_nst_sock_container(struct o2net_send_tracking *nst,
-					 struct o2net_sock_container *sc)
+static inline void o2net_set_nst_sock_container(struct o2net_send_tracking *nst,
+						struct o2net_sock_container *sc)
 {
 	nst->st_sc = sc;
 }
 
-static void o2net_set_nst_msg_id(struct o2net_send_tracking *nst, u32 msg_id)
+static inline void o2net_set_nst_msg_id(struct o2net_send_tracking *nst,
+					u32 msg_id)
 {
 	nst->st_id = msg_id;
 }
 
-#else  /* CONFIG_DEBUG_FS */
-
-static inline void o2net_init_nst(struct o2net_send_tracking *nst, u32 msgtype,
-				  u32 msgkey, struct task_struct *task, u8 node)
+static inline void o2net_set_sock_timer(struct o2net_sock_container *sc)
 {
+	sc->sc_tv_timer = ktime_get();
 }
 
-static inline void o2net_set_nst_sock_time(struct o2net_send_tracking *nst)
+static inline void o2net_set_data_ready_time(struct o2net_sock_container *sc)
 {
+	sc->sc_tv_data_ready = ktime_get();
 }
 
-static inline void o2net_set_nst_send_time(struct o2net_send_tracking *nst)
+static inline void o2net_set_advance_start_time(struct o2net_sock_container *sc)
 {
+	sc->sc_tv_advance_start = ktime_get();
 }
 
-static inline void o2net_set_nst_status_time(struct o2net_send_tracking *nst)
+static inline void o2net_set_advance_stop_time(struct o2net_sock_container *sc)
 {
+	sc->sc_tv_advance_stop = ktime_get();
 }
 
-static inline void o2net_set_nst_sock_container(struct o2net_send_tracking *nst,
-						struct o2net_sock_container *sc)
+static inline void o2net_set_func_start_time(struct o2net_sock_container *sc)
 {
+	sc->sc_tv_func_start = ktime_get();
 }
 
-static inline void o2net_set_nst_msg_id(struct o2net_send_tracking *nst,
-					u32 msg_id)
+static inline void o2net_set_func_stop_time(struct o2net_sock_container *sc)
 {
+	sc->sc_tv_func_stop = ktime_get();
 }
 
+static ktime_t o2net_get_func_run_time(struct o2net_sock_container *sc)
+{
+	return ktime_sub(sc->sc_tv_func_stop, sc->sc_tv_func_start);
+}
+#else  /* CONFIG_DEBUG_FS */
+# define o2net_init_nst(a, b, c, d, e)
+# define o2net_set_nst_sock_time(a)
+# define o2net_set_nst_send_time(a)
+# define o2net_set_nst_status_time(a)
+# define o2net_set_nst_sock_container(a, b)
+# define o2net_set_nst_msg_id(a, b)
+# define o2net_set_sock_timer(a)
+# define o2net_set_data_ready_time(a)
+# define o2net_set_advance_start_time(a)
+# define o2net_set_advance_stop_time(a)
+# define o2net_set_func_start_time(a)
+# define o2net_set_func_stop_time(a)
+# define o2net_get_func_run_time(a)		(ktime_t)0
 #endif /* CONFIG_DEBUG_FS */
 
+#ifdef CONFIG_OCFS2_FS_STATS
+static void o2net_update_send_stats(struct o2net_send_tracking *nst,
+				    struct o2net_sock_container *sc)
+{
+	sc->sc_tv_status_total = ktime_add(sc->sc_tv_status_total,
+					   ktime_sub(ktime_get(),
+						     nst->st_status_time));
+	sc->sc_tv_send_total = ktime_add(sc->sc_tv_send_total,
+					 ktime_sub(nst->st_status_time,
+						   nst->st_send_time));
+	sc->sc_tv_acquiry_total = ktime_add(sc->sc_tv_acquiry_total,
+					    ktime_sub(nst->st_send_time,
+						      nst->st_sock_time));
+	sc->sc_send_count++;
+}
+
+static void o2net_update_recv_stats(struct o2net_sock_container *sc)
+{
+	sc->sc_tv_process_total = ktime_add(sc->sc_tv_process_total,
+					    o2net_get_func_run_time(sc));
+	sc->sc_recv_count++;
+}
+
+#else
+
+# define o2net_update_send_stats(a, b)
+
+# define o2net_update_recv_stats(sc)
+
+#endif /* CONFIG_OCFS2_FS_STATS */
+
 static inline int o2net_reconnect_delay(void)
 {
 	return o2nm_single_cluster->cl_reconnect_delay_ms;
@@ -355,6 +406,7 @@ static void sc_kref_release(struct kref *kref)
 		sc->sc_sock = NULL;
 	}
 
+	o2nm_undepend_item(&sc->sc_node->nd_item);
 	o2nm_node_put(sc->sc_node);
 	sc->sc_node = NULL;
 
@@ -376,6 +428,7 @@ static struct o2net_sock_container *sc_alloc(struct o2nm_node *node)
 {
 	struct o2net_sock_container *sc, *ret = NULL;
 	struct page *page = NULL;
+	int status = 0;
 
 	page = alloc_page(GFP_NOFS);
 	sc = kzalloc(sizeof(*sc), GFP_NOFS);
@@ -386,6 +439,13 @@ static struct o2net_sock_container *sc_alloc(struct o2nm_node *node)
 	o2nm_node_get(node);
 	sc->sc_node = node;
 
+	/* pin the node item of the remote node */
+	status = o2nm_depend_item(&node->nd_item);
+	if (status) {
+		mlog_errno(status);
+		o2nm_node_put(node);
+		goto out;
+	}
 	INIT_WORK(&sc->sc_connect_work, o2net_sc_connect_completed);
 	INIT_WORK(&sc->sc_rx_work, o2net_rx_until_empty);
 	INIT_WORK(&sc->sc_shutdown_work, o2net_shutdown_sc);
@@ -546,7 +606,7 @@ static void o2net_data_ready(struct sock *sk, int bytes)
 	if (sk->sk_user_data) {
 		struct o2net_sock_container *sc = sk->sk_user_data;
 		sclog(sc, "data_ready hit\n");
-		do_gettimeofday(&sc->sc_tv_data_ready);
+		o2net_set_data_ready_time(sc);
 		o2net_sc_queue_work(sc, &sc->sc_rx_work);
 		ready = sc->sc_data_ready;
 	} else {
@@ -1070,6 +1130,8 @@ int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec,
 	o2net_set_nst_status_time(&nst);
 	wait_event(nsw.ns_wq, o2net_nsw_completed(nn, &nsw));
 
+	o2net_update_send_stats(&nst, sc);
+
 	/* Note that we avoid overwriting the callers status return
 	 * variable if a system error was reported on the other
 	 * side. Callers beware. */
@@ -1183,13 +1245,15 @@ static int o2net_process_message(struct o2net_sock_container *sc,
 	if (syserr != O2NET_ERR_NONE)
 		goto out_respond;
 
-	do_gettimeofday(&sc->sc_tv_func_start);
+	o2net_set_func_start_time(sc);
 	sc->sc_msg_key = be32_to_cpu(hdr->key);
 	sc->sc_msg_type = be16_to_cpu(hdr->msg_type);
 	handler_status = (nmh->nh_func)(hdr, sizeof(struct o2net_msg) +
 					     be16_to_cpu(hdr->data_len),
 					nmh->nh_func_data, &ret_data);
-	do_gettimeofday(&sc->sc_tv_func_stop);
+	o2net_set_func_stop_time(sc);
+
+	o2net_update_recv_stats(sc);
 
 out_respond:
 	/* this destroys the hdr, so don't use it after this */
@@ -1300,7 +1364,7 @@ static int o2net_advance_rx(struct o2net_sock_container *sc)
 	size_t datalen;
 
 	sclog(sc, "receiving\n");
-	do_gettimeofday(&sc->sc_tv_advance_start);
+	o2net_set_advance_start_time(sc);
 
 	if (unlikely(sc->sc_handshake_ok == 0)) {
 		if(sc->sc_page_off < sizeof(struct o2net_handshake)) {
@@ -1375,7 +1439,7 @@ static int o2net_advance_rx(struct o2net_sock_container *sc)
 
 out:
 	sclog(sc, "ret = %d\n", ret);
-	do_gettimeofday(&sc->sc_tv_advance_stop);
+	o2net_set_advance_stop_time(sc);
 	return ret;
 }
 
@@ -1475,27 +1539,28 @@ static void o2net_idle_timer(unsigned long data)
 {
 	struct o2net_sock_container *sc = (struct o2net_sock_container *)data;
 	struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
-	struct timeval now;
 
-	do_gettimeofday(&now);
+#ifdef CONFIG_DEBUG_FS
+	ktime_t now = ktime_get();
+#endif
 
 	printk(KERN_NOTICE "o2net: connection to " SC_NODEF_FMT " has been idle for %u.%u "
 	     "seconds, shutting it down.\n", SC_NODEF_ARGS(sc),
 		     o2net_idle_timeout() / 1000,
 		     o2net_idle_timeout() % 1000);
-	mlog(ML_NOTICE, "here are some times that might help debug the "
-	     "situation: (tmr %ld.%ld now %ld.%ld dr %ld.%ld adv "
-	     "%ld.%ld:%ld.%ld func (%08x:%u) %ld.%ld:%ld.%ld)\n",
-	     sc->sc_tv_timer.tv_sec, (long) sc->sc_tv_timer.tv_usec,
-	     now.tv_sec, (long) now.tv_usec,
-	     sc->sc_tv_data_ready.tv_sec, (long) sc->sc_tv_data_ready.tv_usec,
-	     sc->sc_tv_advance_start.tv_sec,
-	     (long) sc->sc_tv_advance_start.tv_usec,
-	     sc->sc_tv_advance_stop.tv_sec,
-	     (long) sc->sc_tv_advance_stop.tv_usec,
+
+#ifdef CONFIG_DEBUG_FS
+	mlog(ML_NOTICE, "Here are some times that might help debug the "
+	     "situation: (Timer: %lld, Now %lld, DataReady %lld, Advance %lld-%lld, "
+	     "Key 0x%08x, Func %u, FuncTime %lld-%lld)\n",
+	     (long long)ktime_to_us(sc->sc_tv_timer), (long long)ktime_to_us(now),
+	     (long long)ktime_to_us(sc->sc_tv_data_ready),
+	     (long long)ktime_to_us(sc->sc_tv_advance_start),
+	     (long long)ktime_to_us(sc->sc_tv_advance_stop),
 	     sc->sc_msg_key, sc->sc_msg_type,
-	     sc->sc_tv_func_start.tv_sec, (long) sc->sc_tv_func_start.tv_usec,
-	     sc->sc_tv_func_stop.tv_sec, (long) sc->sc_tv_func_stop.tv_usec);
+	     (long long)ktime_to_us(sc->sc_tv_func_start),
+	     (long long)ktime_to_us(sc->sc_tv_func_stop));
+#endif
 
 	/*
 	 * Initialize the nn_timeout so that the next connection attempt
@@ -1511,7 +1576,7 @@ static void o2net_sc_reset_idle_timer(struct o2net_sock_container *sc)
 	o2net_sc_cancel_delayed_work(sc, &sc->sc_keepalive_work);
 	o2net_sc_queue_delayed_work(sc, &sc->sc_keepalive_work,
 		      msecs_to_jiffies(o2net_keepalive_delay()));
-	do_gettimeofday(&sc->sc_tv_timer);
+	o2net_set_sock_timer(sc);
 	mod_timer(&sc->sc_idle_timeout,
 	       jiffies + msecs_to_jiffies(o2net_idle_timeout()));
 }

fs/ocfs2/cluster/tcp_internal.h

@@ -166,18 +166,27 @@ struct o2net_sock_container {
 	/* original handlers for the sockets */
 	void			(*sc_state_change)(struct sock *sk);
 	void			(*sc_data_ready)(struct sock *sk, int bytes);
-#ifdef CONFIG_DEBUG_FS
-	struct list_head        sc_net_debug_item;
-#endif
-	struct timeval 		sc_tv_timer;
-	struct timeval 		sc_tv_data_ready;
-	struct timeval 		sc_tv_advance_start;
-	struct timeval 		sc_tv_advance_stop;
-	struct timeval 		sc_tv_func_start;
-	struct timeval 		sc_tv_func_stop;
+
 	u32			sc_msg_key;
 	u16			sc_msg_type;
 
+#ifdef CONFIG_DEBUG_FS
+	struct list_head        sc_net_debug_item;
+	ktime_t			sc_tv_timer;
+	ktime_t			sc_tv_data_ready;
+	ktime_t			sc_tv_advance_start;
+	ktime_t			sc_tv_advance_stop;
+	ktime_t			sc_tv_func_start;
+	ktime_t			sc_tv_func_stop;
+#endif
+#ifdef CONFIG_OCFS2_FS_STATS
+	ktime_t			sc_tv_acquiry_total;
+	ktime_t			sc_tv_send_total;
+	ktime_t			sc_tv_status_total;
+	u32			sc_send_count;
+	u32			sc_recv_count;
+	ktime_t			sc_tv_process_total;
+#endif
 	struct mutex		sc_send_lock;
 };
 
@@ -220,9 +229,9 @@ struct o2net_send_tracking {
 	u32				st_msg_type;
 	u32				st_msg_key;
 	u8				st_node;
-	struct timeval			st_sock_time;
-	struct timeval			st_send_time;
-	struct timeval			st_status_time;
+	ktime_t				st_sock_time;
+	ktime_t				st_send_time;
+	ktime_t				st_status_time;
 };
 #else
 struct o2net_send_tracking {

fs/ocfs2/dlm/dlmast.c

@@ -90,19 +90,29 @@ static int dlm_should_cancel_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 
 void __dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 {
-	mlog_entry_void();
+	struct dlm_lock_resource *res;
 
 	BUG_ON(!dlm);
 	BUG_ON(!lock);
 
+	res = lock->lockres;
+
 	assert_spin_locked(&dlm->ast_lock);
+
 	if (!list_empty(&lock->ast_list)) {
-		mlog(ML_ERROR, "ast list not empty!!  pending=%d, newlevel=%d\n",
+		mlog(ML_ERROR, "%s: res %.*s, lock %u:%llu, "
+		     "AST list not empty, pending %d, newlevel %d\n",
+		     dlm->name, res->lockname.len, res->lockname.name,
+		     dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
+		     dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)),
 		     lock->ast_pending, lock->ml.type);
 		BUG();
 	}
 	if (lock->ast_pending)
-		mlog(0, "lock has an ast getting flushed right now\n");
+		mlog(0, "%s: res %.*s, lock %u:%llu, AST getting flushed\n",
+		     dlm->name, res->lockname.len, res->lockname.name,
+		     dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
+		     dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)));
 
 	/* putting lock on list, add a ref */
 	dlm_lock_get(lock);
@@ -110,9 +120,10 @@ void __dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 
 	/* check to see if this ast obsoletes the bast */
 	if (dlm_should_cancel_bast(dlm, lock)) {
-		struct dlm_lock_resource *res = lock->lockres;
-		mlog(0, "%s: cancelling bast for %.*s\n",
-		     dlm->name, res->lockname.len, res->lockname.name);
+		mlog(0, "%s: res %.*s, lock %u:%llu, Cancelling BAST\n",
+		     dlm->name, res->lockname.len, res->lockname.name,
+		     dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
+		     dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)));
 		lock->bast_pending = 0;
 		list_del_init(&lock->bast_list);
 		lock->ml.highest_blocked = LKM_IVMODE;
@@ -134,8 +145,6 @@ void __dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 
 void dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 {
-	mlog_entry_void();
-
 	BUG_ON(!dlm);
 	BUG_ON(!lock);
 
@@ -147,15 +156,21 @@ void dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 
 void __dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 {
-	mlog_entry_void();
+	struct dlm_lock_resource *res;
 
 	BUG_ON(!dlm);
 	BUG_ON(!lock);
+
 	assert_spin_locked(&dlm->ast_lock);
 
+	res = lock->lockres;
+
 	BUG_ON(!list_empty(&lock->bast_list));
 	if (lock->bast_pending)
-		mlog(0, "lock has a bast getting flushed right now\n");
+		mlog(0, "%s: res %.*s, lock %u:%llu, BAST getting flushed\n",
+		     dlm->name, res->lockname.len, res->lockname.name,
+		     dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
+		     dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)));
 
 	/* putting lock on list, add a ref */
 	dlm_lock_get(lock);
@@ -167,8 +182,6 @@ void __dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 
 void dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 {
-	mlog_entry_void();
-
 	BUG_ON(!dlm);
 	BUG_ON(!lock);
 
@@ -213,7 +226,10 @@ void dlm_do_local_ast(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 	dlm_astlockfunc_t *fn;
 	struct dlm_lockstatus *lksb;
 
-	mlog_entry_void();
+	mlog(0, "%s: res %.*s, lock %u:%llu, Local AST\n", dlm->name,
+	     res->lockname.len, res->lockname.name,
+	     dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
+	     dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)));
 
 	lksb = lock->lksb;
 	fn = lock->ast;
@@ -231,7 +247,10 @@ int dlm_do_remote_ast(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 	struct dlm_lockstatus *lksb;
 	int lksbflags;
 
-	mlog_entry_void();
+	mlog(0, "%s: res %.*s, lock %u:%llu, Remote AST\n", dlm->name,
+	     res->lockname.len, res->lockname.name,
+	     dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
+	     dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)));
 
 	lksb = lock->lksb;
 	BUG_ON(lock->ml.node == dlm->node_num);
@@ -250,9 +269,14 @@ void dlm_do_local_bast(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 {
 	dlm_bastlockfunc_t *fn = lock->bast;
 
-	mlog_entry_void();
 	BUG_ON(lock->ml.node != dlm->node_num);
 
+	mlog(0, "%s: res %.*s, lock %u:%llu, Local BAST, blocked %d\n",
+	     dlm->name, res->lockname.len, res->lockname.name,
+	     dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
+	     dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)),
+	     blocked_type);
+
 	(*fn)(lock->astdata, blocked_type);
 }
 
@@ -332,7 +356,8 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data,
 	/* cannot get a proxy ast message if this node owns it */
 	BUG_ON(res->owner == dlm->node_num);
 
-	mlog(0, "lockres %.*s\n", res->lockname.len, res->lockname.name);
+	mlog(0, "%s: res %.*s\n", dlm->name, res->lockname.len,
+	     res->lockname.name);
 
 	spin_lock(&res->spinlock);
 	if (res->state & DLM_LOCK_RES_RECOVERING) {
@@ -382,8 +407,12 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data,
 	if (past->type == DLM_AST) {
 		/* do not alter lock refcount.  switching lists. */
 		list_move_tail(&lock->list, &res->granted);
-		mlog(0, "ast: Adding to granted list... type=%d, "
-		     "convert_type=%d\n", lock->ml.type, lock->ml.convert_type);
+		mlog(0, "%s: res %.*s, lock %u:%llu, Granted type %d => %d\n",
+		     dlm->name, res->lockname.len, res->lockname.name,
+		     dlm_get_lock_cookie_node(be64_to_cpu(cookie)),
+		     dlm_get_lock_cookie_seq(be64_to_cpu(cookie)),
+		     lock->ml.type, lock->ml.convert_type);
+
 		if (lock->ml.convert_type != LKM_IVMODE) {
 			lock->ml.type = lock->ml.convert_type;
 			lock->ml.convert_type = LKM_IVMODE;
@@ -426,9 +455,9 @@ int dlm_send_proxy_ast_msg(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 	size_t veclen = 1;
 	int status;
 
-	mlog_entry("res %.*s, to=%u, type=%d, blocked_type=%d\n",
-		   res->lockname.len, res->lockname.name, lock->ml.node,
-		   msg_type, blocked_type);
+	mlog(0, "%s: res %.*s, to %u, type %d, blocked_type %d\n", dlm->name,
+	     res->lockname.len, res->lockname.name, lock->ml.node, msg_type,
+	     blocked_type);
 
 	memset(&past, 0, sizeof(struct dlm_proxy_ast));
 	past.node_idx = dlm->node_num;
@@ -441,7 +470,6 @@ int dlm_send_proxy_ast_msg(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 	vec[0].iov_len = sizeof(struct dlm_proxy_ast);
 	vec[0].iov_base = &past;
 	if (flags & DLM_LKSB_GET_LVB) {
-		mlog(0, "returning requested LVB data\n");
 		be32_add_cpu(&past.flags, LKM_GET_LVB);
 		vec[1].iov_len = DLM_LVB_LEN;
 		vec[1].iov_base = lock->lksb->lvb;
@@ -451,8 +479,8 @@ int dlm_send_proxy_ast_msg(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 	ret = o2net_send_message_vec(DLM_PROXY_AST_MSG, dlm->key, vec, veclen,
 				     lock->ml.node, &status);
 	if (ret < 0)
-		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
-		     "node %u\n", ret, DLM_PROXY_AST_MSG, dlm->key,
+		mlog(ML_ERROR, "%s: res %.*s, error %d send AST to node %u\n",
+		     dlm->name, res->lockname.len, res->lockname.name, ret,
 		     lock->ml.node);
 	else {
 		if (status == DLM_RECOVERING) {

fs/ocfs2/dlm/dlmcommon.h

@@ -50,10 +50,10 @@
 #define dlm_lockid_hash(_n, _l) full_name_hash(_n, _l)
 
 enum dlm_mle_type {
-	DLM_MLE_BLOCK,
-	DLM_MLE_MASTER,
-	DLM_MLE_MIGRATION,
-	DLM_MLE_NUM_TYPES
+	DLM_MLE_BLOCK = 0,
+	DLM_MLE_MASTER = 1,
+	DLM_MLE_MIGRATION = 2,
+	DLM_MLE_NUM_TYPES = 3,
 };
 
 struct dlm_master_list_entry {
@@ -82,8 +82,8 @@ struct dlm_master_list_entry {
 
 enum dlm_ast_type {
 	DLM_AST = 0,
-	DLM_BAST,
-	DLM_ASTUNLOCK
+	DLM_BAST = 1,
+	DLM_ASTUNLOCK = 2,
 };
 
 
@@ -119,9 +119,9 @@ struct dlm_recovery_ctxt
 
 enum dlm_ctxt_state {
 	DLM_CTXT_NEW = 0,
-	DLM_CTXT_JOINED,
-	DLM_CTXT_IN_SHUTDOWN,
-	DLM_CTXT_LEAVING,
+	DLM_CTXT_JOINED = 1,
+	DLM_CTXT_IN_SHUTDOWN = 2,
+	DLM_CTXT_LEAVING = 3,
 };
 
 struct dlm_ctxt
@@ -388,8 +388,8 @@ struct dlm_lock
 
 enum dlm_lockres_list {
 	DLM_GRANTED_LIST = 0,
-	DLM_CONVERTING_LIST,
-	DLM_BLOCKED_LIST
+	DLM_CONVERTING_LIST = 1,
+	DLM_BLOCKED_LIST = 2,
 };
 
 static inline int dlm_lvb_is_empty(char *lvb)
@@ -427,27 +427,27 @@ struct dlm_node_iter
 
 
 enum {
-	DLM_MASTER_REQUEST_MSG    = 500,
-	DLM_UNUSED_MSG1,         /* 501 */
-	DLM_ASSERT_MASTER_MSG,	 /* 502 */
-	DLM_CREATE_LOCK_MSG,	 /* 503 */
-	DLM_CONVERT_LOCK_MSG,	 /* 504 */
-	DLM_PROXY_AST_MSG,	 /* 505 */
-	DLM_UNLOCK_LOCK_MSG,	 /* 506 */
-	DLM_DEREF_LOCKRES_MSG,	 /* 507 */
-	DLM_MIGRATE_REQUEST_MSG, /* 508 */
-	DLM_MIG_LOCKRES_MSG, 	 /* 509 */
-	DLM_QUERY_JOIN_MSG,	 /* 510 */
-	DLM_ASSERT_JOINED_MSG,	 /* 511 */
-	DLM_CANCEL_JOIN_MSG,	 /* 512 */
-	DLM_EXIT_DOMAIN_MSG,	 /* 513 */
-	DLM_MASTER_REQUERY_MSG,	 /* 514 */
-	DLM_LOCK_REQUEST_MSG,	 /* 515 */
-	DLM_RECO_DATA_DONE_MSG,	 /* 516 */
-	DLM_BEGIN_RECO_MSG,	 /* 517 */
-	DLM_FINALIZE_RECO_MSG,	 /* 518 */
-	DLM_QUERY_REGION,	 /* 519 */
-	DLM_QUERY_NODEINFO,	 /* 520 */
+	DLM_MASTER_REQUEST_MSG		= 500,
+	DLM_UNUSED_MSG1			= 501,
+	DLM_ASSERT_MASTER_MSG		= 502,
+	DLM_CREATE_LOCK_MSG		= 503,
+	DLM_CONVERT_LOCK_MSG		= 504,
+	DLM_PROXY_AST_MSG		= 505,
+	DLM_UNLOCK_LOCK_MSG		= 506,
+	DLM_DEREF_LOCKRES_MSG		= 507,
+	DLM_MIGRATE_REQUEST_MSG		= 508,
+	DLM_MIG_LOCKRES_MSG		= 509,
+	DLM_QUERY_JOIN_MSG		= 510,
+	DLM_ASSERT_JOINED_MSG		= 511,
+	DLM_CANCEL_JOIN_MSG		= 512,
+	DLM_EXIT_DOMAIN_MSG		= 513,
+	DLM_MASTER_REQUERY_MSG		= 514,
+	DLM_LOCK_REQUEST_MSG		= 515,
+	DLM_RECO_DATA_DONE_MSG		= 516,
+	DLM_BEGIN_RECO_MSG		= 517,
+	DLM_FINALIZE_RECO_MSG		= 518,
+	DLM_QUERY_REGION		= 519,
+	DLM_QUERY_NODEINFO		= 520,
 };
 
 struct dlm_reco_node_data
@@ -460,19 +460,19 @@ struct dlm_reco_node_data
 enum {
 	DLM_RECO_NODE_DATA_DEAD = -1,
 	DLM_RECO_NODE_DATA_INIT = 0,
-	DLM_RECO_NODE_DATA_REQUESTING,
-	DLM_RECO_NODE_DATA_REQUESTED,
-	DLM_RECO_NODE_DATA_RECEIVING,
-	DLM_RECO_NODE_DATA_DONE,
-	DLM_RECO_NODE_DATA_FINALIZE_SENT,
+	DLM_RECO_NODE_DATA_REQUESTING = 1,
+	DLM_RECO_NODE_DATA_REQUESTED = 2,
+	DLM_RECO_NODE_DATA_RECEIVING = 3,
+	DLM_RECO_NODE_DATA_DONE = 4,
+	DLM_RECO_NODE_DATA_FINALIZE_SENT = 5,
 };
 
 
 enum {
 	DLM_MASTER_RESP_NO = 0,
-	DLM_MASTER_RESP_YES,
-	DLM_MASTER_RESP_MAYBE,
-	DLM_MASTER_RESP_ERROR
+	DLM_MASTER_RESP_YES = 1,
+	DLM_MASTER_RESP_MAYBE = 2,
+	DLM_MASTER_RESP_ERROR = 3,
 };
 
 
@@ -649,9 +649,9 @@ struct dlm_proxy_ast
 #define DLM_MOD_KEY (0x666c6172)
 enum dlm_query_join_response_code {
 	JOIN_DISALLOW = 0,
-	JOIN_OK,
-	JOIN_OK_NO_MAP,
-	JOIN_PROTOCOL_MISMATCH,
+	JOIN_OK = 1,
+	JOIN_OK_NO_MAP = 2,
+	JOIN_PROTOCOL_MISMATCH = 3,
 };
 
 struct dlm_query_join_packet {

fs/ocfs2/dlm/dlmdebug.h

@@ -37,11 +37,6 @@ struct dlm_debug_ctxt {
 	struct dentry *debug_purgelist_dentry;
 };
 
-struct debug_buffer {
-	int len;
-	char *buf;
-};
-
 struct debug_lockres {
 	int dl_len;
 	char *dl_buf;

fs/ocfs2/dlm/dlmdomain.c

@@ -460,8 +460,6 @@ static int dlm_migrate_all_locks(struct dlm_ctxt *dlm)
 		}
 		cond_resched_lock(&dlm->spinlock);
 		num += n;
-		mlog(0, "%s: touched %d lockreses in bucket %d "
-		     "(tot=%d)\n", dlm->name, n, i, num);
 	}
 	spin_unlock(&dlm->spinlock);
 	wake_up(&dlm->dlm_thread_wq);
@@ -1661,8 +1659,8 @@ static int dlm_try_to_join_domain(struct dlm_ctxt *dlm)
 
 static void dlm_unregister_domain_handlers(struct dlm_ctxt *dlm)
 {
-	o2hb_unregister_callback(NULL, &dlm->dlm_hb_up);
-	o2hb_unregister_callback(NULL, &dlm->dlm_hb_down);
+	o2hb_unregister_callback(dlm->name, &dlm->dlm_hb_up);
+	o2hb_unregister_callback(dlm->name, &dlm->dlm_hb_down);
 	o2net_unregister_handler_list(&dlm->dlm_domain_handlers);
 }
 
@@ -1674,13 +1672,13 @@ static int dlm_register_domain_handlers(struct dlm_ctxt *dlm)
 
 	o2hb_setup_callback(&dlm->dlm_hb_down, O2HB_NODE_DOWN_CB,
 			    dlm_hb_node_down_cb, dlm, DLM_HB_NODE_DOWN_PRI);
-	status = o2hb_register_callback(NULL, &dlm->dlm_hb_down);
+	status = o2hb_register_callback(dlm->name, &dlm->dlm_hb_down);
 	if (status)
 		goto bail;
 
 	o2hb_setup_callback(&dlm->dlm_hb_up, O2HB_NODE_UP_CB,
 			    dlm_hb_node_up_cb, dlm, DLM_HB_NODE_UP_PRI);
-	status = o2hb_register_callback(NULL, &dlm->dlm_hb_up);
+	status = o2hb_register_callback(dlm->name, &dlm->dlm_hb_up);
 	if (status)
 		goto bail;
 

fs/ocfs2/dlm/dlmlock.c

@@ -106,6 +106,9 @@ static int dlm_can_grant_new_lock(struct dlm_lock_resource *res,
 
 		if (!dlm_lock_compatible(tmplock->ml.type, lock->ml.type))
 			return 0;
+		if (!dlm_lock_compatible(tmplock->ml.convert_type,
+					 lock->ml.type))
+			return 0;
 	}
 
 	return 1;

fs/ocfs2/namei.c

@@ -1017,8 +1017,11 @@ static int ocfs2_double_lock(struct ocfs2_super *osb,
 		 * An error return must mean that no cluster locks
 		 * were held on function exit.
 		 */
-		if (oi1->ip_blkno != oi2->ip_blkno)
+		if (oi1->ip_blkno != oi2->ip_blkno) {
 			ocfs2_inode_unlock(inode2, 1);
+			brelse(*bh2);
+			*bh2 = NULL;
+		}
 
 		if (status != -ENOENT)
 			mlog_errno(status);

fs/ocfs2/ocfs2.h

@@ -420,6 +420,11 @@ struct ocfs2_super
 	struct inode			*osb_tl_inode;
 	struct buffer_head		*osb_tl_bh;
 	struct delayed_work		osb_truncate_log_wq;
+	/*
+	 * How many clusters in our truncate log.
+	 * It must be protected by osb_tl_inode->i_mutex.
+	 */
+	unsigned int truncated_clusters;
 
 	struct ocfs2_node_map		osb_recovering_orphan_dirs;
 	unsigned int			*osb_orphan_wipes;