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

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mfasheh/ocfs2: (32 commits)
  ocfs2: recover orphans in offline slots during recovery and mount
  ocfs2: Pagecache usage optimization on ocfs2
  ocfs2: fix rare stale inode errors when exporting via nfs
  ocfs2/dlm: Tweak mle_state output
  ocfs2/dlm: Do not purge lockres that is being migrated dlm_purge_lockres()
  ocfs2/dlm: Remove struct dlm_lock_name in struct dlm_master_list_entry
  ocfs2/dlm: Show the number of lockres/mles in dlm_state
  ocfs2/dlm: dlm_set_lockres_owner() and dlm_change_lockres_owner() inlined
  ocfs2/dlm: Improve lockres counts
  ocfs2/dlm: Track number of mles
  ocfs2/dlm: Indent dlm_cleanup_master_list()
  ocfs2/dlm: Activate dlm->master_hash for master list entries
  ocfs2/dlm: Create and destroy the dlm->master_hash
  ocfs2/dlm: Refactor dlm_clean_master_list()
  ocfs2/dlm: Clean up struct dlm_lock_name
  ocfs2/dlm: Encapsulate adding and removing of mle from dlm->master_list
  ocfs2: Optimize inode group allocation by recording last used group.
  ocfs2: Allocate inode groups from global_bitmap.
  ocfs2: Optimize inode allocation by remembering last group
  ocfs2: fix leaf start calculation in ocfs2_dx_dir_rebalance()
  ...

fs/ocfs2/alloc.c

@@ -294,6 +294,55 @@ static struct ocfs2_extent_tree_operations ocfs2_xattr_tree_et_ops = {
 	.eo_fill_max_leaf_clusters = ocfs2_xattr_tree_fill_max_leaf_clusters,
 };
 
+static void ocfs2_dx_root_set_last_eb_blk(struct ocfs2_extent_tree *et,
+					  u64 blkno)
+{
+	struct ocfs2_dx_root_block *dx_root = et->et_object;
+
+	dx_root->dr_last_eb_blk = cpu_to_le64(blkno);
+}
+
+static u64 ocfs2_dx_root_get_last_eb_blk(struct ocfs2_extent_tree *et)
+{
+	struct ocfs2_dx_root_block *dx_root = et->et_object;
+
+	return le64_to_cpu(dx_root->dr_last_eb_blk);
+}
+
+static void ocfs2_dx_root_update_clusters(struct inode *inode,
+					  struct ocfs2_extent_tree *et,
+					  u32 clusters)
+{
+	struct ocfs2_dx_root_block *dx_root = et->et_object;
+
+	le32_add_cpu(&dx_root->dr_clusters, clusters);
+}
+
+static int ocfs2_dx_root_sanity_check(struct inode *inode,
+				      struct ocfs2_extent_tree *et)
+{
+	struct ocfs2_dx_root_block *dx_root = et->et_object;
+
+	BUG_ON(!OCFS2_IS_VALID_DX_ROOT(dx_root));
+
+	return 0;
+}
+
+static void ocfs2_dx_root_fill_root_el(struct ocfs2_extent_tree *et)
+{
+	struct ocfs2_dx_root_block *dx_root = et->et_object;
+
+	et->et_root_el = &dx_root->dr_list;
+}
+
+static struct ocfs2_extent_tree_operations ocfs2_dx_root_et_ops = {
+	.eo_set_last_eb_blk	= ocfs2_dx_root_set_last_eb_blk,
+	.eo_get_last_eb_blk	= ocfs2_dx_root_get_last_eb_blk,
+	.eo_update_clusters	= ocfs2_dx_root_update_clusters,
+	.eo_sanity_check	= ocfs2_dx_root_sanity_check,
+	.eo_fill_root_el	= ocfs2_dx_root_fill_root_el,
+};
+
 static void __ocfs2_init_extent_tree(struct ocfs2_extent_tree *et,
 				     struct inode *inode,
 				     struct buffer_head *bh,
@@ -339,6 +388,14 @@ void ocfs2_init_xattr_value_extent_tree(struct ocfs2_extent_tree *et,
 				 &ocfs2_xattr_value_et_ops);
 }
 
+void ocfs2_init_dx_root_extent_tree(struct ocfs2_extent_tree *et,
+				    struct inode *inode,
+				    struct buffer_head *bh)
+{
+	__ocfs2_init_extent_tree(et, inode, bh, ocfs2_journal_access_dr,
+				 NULL, &ocfs2_dx_root_et_ops);
+}
+
 static inline void ocfs2_et_set_last_eb_blk(struct ocfs2_extent_tree *et,
 					    u64 new_last_eb_blk)
 {

fs/ocfs2/alloc.h

@@ -75,6 +75,9 @@ struct ocfs2_xattr_value_buf;
 void ocfs2_init_xattr_value_extent_tree(struct ocfs2_extent_tree *et,
 					struct inode *inode,
 					struct ocfs2_xattr_value_buf *vb);
+void ocfs2_init_dx_root_extent_tree(struct ocfs2_extent_tree *et,
+				    struct inode *inode,
+				    struct buffer_head *bh);
 
 /*
  * Read an extent block into *bh.  If *bh is NULL, a bh will be

fs/ocfs2/aops.c

@@ -1956,15 +1956,16 @@ static int ocfs2_write_end(struct file *file, struct address_space *mapping,
 }
 
 const struct address_space_operations ocfs2_aops = {
-	.readpage	= ocfs2_readpage,
-	.readpages	= ocfs2_readpages,
-	.writepage	= ocfs2_writepage,
-	.write_begin	= ocfs2_write_begin,
-	.write_end	= ocfs2_write_end,
-	.bmap		= ocfs2_bmap,
-	.sync_page	= block_sync_page,
-	.direct_IO	= ocfs2_direct_IO,
-	.invalidatepage	= ocfs2_invalidatepage,
-	.releasepage	= ocfs2_releasepage,
-	.migratepage	= buffer_migrate_page,
+	.readpage		= ocfs2_readpage,
+	.readpages		= ocfs2_readpages,
+	.writepage		= ocfs2_writepage,
+	.write_begin		= ocfs2_write_begin,
+	.write_end		= ocfs2_write_end,
+	.bmap			= ocfs2_bmap,
+	.sync_page		= block_sync_page,
+	.direct_IO		= ocfs2_direct_IO,
+	.invalidatepage		= ocfs2_invalidatepage,
+	.releasepage		= ocfs2_releasepage,
+	.migratepage		= buffer_migrate_page,
+	.is_partially_uptodate	= block_is_partially_uptodate,
 };

fs/ocfs2/cluster/heartbeat.c

@@ -33,6 +33,7 @@
 #include <linux/random.h>
 #include <linux/crc32.h>
 #include <linux/time.h>
+#include <linux/debugfs.h>
 
 #include "heartbeat.h"
 #include "tcp.h"
@@ -60,6 +61,11 @@ static unsigned long o2hb_live_node_bitmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
 static LIST_HEAD(o2hb_node_events);
 static DECLARE_WAIT_QUEUE_HEAD(o2hb_steady_queue);
 
+#define O2HB_DEBUG_DIR			"o2hb"
+#define O2HB_DEBUG_LIVENODES		"livenodes"
+static struct dentry *o2hb_debug_dir;
+static struct dentry *o2hb_debug_livenodes;
+
 static LIST_HEAD(o2hb_all_regions);
 
 static struct o2hb_callback {
@@ -905,7 +911,77 @@ static int o2hb_thread(void *data)
 	return 0;
 }
 
-void o2hb_init(void)
+#ifdef CONFIG_DEBUG_FS
+static int o2hb_debug_open(struct inode *inode, struct file *file)
+{
+	unsigned long map[BITS_TO_LONGS(O2NM_MAX_NODES)];
+	char *buf = NULL;
+	int i = -1;
+	int out = 0;
+
+	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!buf)
+		goto bail;
+
+	o2hb_fill_node_map(map, sizeof(map));
+
+	while ((i = find_next_bit(map, O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES)
+		out += snprintf(buf + out, PAGE_SIZE - out, "%d ", i);
+	out += snprintf(buf + out, PAGE_SIZE - out, "\n");
+
+	i_size_write(inode, out);
+
+	file->private_data = buf;
+
+	return 0;
+bail:
+	return -ENOMEM;
+}
+
+static int o2hb_debug_release(struct inode *inode, struct file *file)
+{
+	kfree(file->private_data);
+	return 0;
+}
+
+static ssize_t o2hb_debug_read(struct file *file, char __user *buf,
+				 size_t nbytes, loff_t *ppos)
+{
+	return simple_read_from_buffer(buf, nbytes, ppos, file->private_data,
+				       i_size_read(file->f_mapping->host));
+}
+#else
+static int o2hb_debug_open(struct inode *inode, struct file *file)
+{
+	return 0;
+}
+static int o2hb_debug_release(struct inode *inode, struct file *file)
+{
+	return 0;
+}
+static ssize_t o2hb_debug_read(struct file *file, char __user *buf,
+			       size_t nbytes, loff_t *ppos)
+{
+	return 0;
+}
+#endif  /* CONFIG_DEBUG_FS */
+
+static struct file_operations o2hb_debug_fops = {
+	.open =		o2hb_debug_open,
+	.release =	o2hb_debug_release,
+	.read =		o2hb_debug_read,
+	.llseek =	generic_file_llseek,
+};
+
+void o2hb_exit(void)
+{
+	if (o2hb_debug_livenodes)
+		debugfs_remove(o2hb_debug_livenodes);
+	if (o2hb_debug_dir)
+		debugfs_remove(o2hb_debug_dir);
+}
+
+int o2hb_init(void)
 {
 	int i;
 
@@ -918,6 +994,24 @@ void o2hb_init(void)
 	INIT_LIST_HEAD(&o2hb_node_events);
 
 	memset(o2hb_live_node_bitmap, 0, sizeof(o2hb_live_node_bitmap));
+
+	o2hb_debug_dir = debugfs_create_dir(O2HB_DEBUG_DIR, NULL);
+	if (!o2hb_debug_dir) {
+		mlog_errno(-ENOMEM);
+		return -ENOMEM;
+	}
+
+	o2hb_debug_livenodes = debugfs_create_file(O2HB_DEBUG_LIVENODES,
+						   S_IFREG|S_IRUSR,
+						   o2hb_debug_dir, NULL,
+						   &o2hb_debug_fops);
+	if (!o2hb_debug_livenodes) {
+		mlog_errno(-ENOMEM);
+		debugfs_remove(o2hb_debug_dir);
+		return -ENOMEM;
+	}
+
+	return 0;
 }
 
 /* if we're already in a callback then we're already serialized by the sem */

fs/ocfs2/cluster/heartbeat.h

@@ -75,7 +75,8 @@ void o2hb_unregister_callback(const char *region_uuid,
 			      struct o2hb_callback_func *hc);
 void o2hb_fill_node_map(unsigned long *map,
 			unsigned bytes);
-void o2hb_init(void);
+void o2hb_exit(void);
+int o2hb_init(void);
 int o2hb_check_node_heartbeating(u8 node_num);
 int o2hb_check_node_heartbeating_from_callback(u8 node_num);
 int o2hb_check_local_node_heartbeating(void);

fs/ocfs2/cluster/nodemanager.c

@@ -881,6 +881,7 @@ static void __exit exit_o2nm(void)
 	o2cb_sys_shutdown();
 
 	o2net_exit();
+	o2hb_exit();
 }
 
 static int __init init_o2nm(void)
@@ -889,11 +890,13 @@ static int __init init_o2nm(void)
 
 	cluster_print_version();
 
-	o2hb_init();
+	ret = o2hb_init();
+	if (ret)
+		goto out;
 
 	ret = o2net_init();
 	if (ret)
-		goto out;
+		goto out_o2hb;
 
 	ret = o2net_register_hb_callbacks();
 	if (ret)
@@ -916,6 +919,8 @@ static int __init init_o2nm(void)
 	o2net_unregister_hb_callbacks();
 out_o2net:
 	o2net_exit();
+out_o2hb:
+	o2hb_exit();
 out:
 	return ret;
 }

fs/ocfs2/dir.h

@@ -26,44 +26,70 @@
 #ifndef OCFS2_DIR_H
 #define OCFS2_DIR_H
 
-struct buffer_head *ocfs2_find_entry(const char *name,
-				     int namelen,
-				     struct inode *dir,
-				     struct ocfs2_dir_entry **res_dir);
+struct ocfs2_dx_hinfo {
+	u32	major_hash;
+	u32	minor_hash;
+};
+
+struct ocfs2_dir_lookup_result {
+	struct buffer_head		*dl_leaf_bh;	/* Unindexed leaf
+							 * block */
+	struct ocfs2_dir_entry		*dl_entry;	/* Target dirent in
+							 * unindexed leaf */
+
+	struct buffer_head		*dl_dx_root_bh;	/* Root of indexed
+							 * tree */
+
+	struct buffer_head		*dl_dx_leaf_bh;	/* Indexed leaf block */
+	struct ocfs2_dx_entry		*dl_dx_entry;	/* Target dx_entry in
+							 * indexed leaf */
+	struct ocfs2_dx_hinfo		dl_hinfo;	/* Name hash results */
+
+	struct buffer_head		*dl_prev_leaf_bh;/* Previous entry in
+							  * dir free space
+							  * list. NULL if
+							  * previous entry is
+							  * dx root block. */
+};
+
+void ocfs2_free_dir_lookup_result(struct ocfs2_dir_lookup_result *res);
+
+int ocfs2_find_entry(const char *name, int namelen,
+		     struct inode *dir,
+		     struct ocfs2_dir_lookup_result *lookup);
 int ocfs2_delete_entry(handle_t *handle,
 		       struct inode *dir,
-		       struct ocfs2_dir_entry *de_del,
-		       struct buffer_head *bh);
+		       struct ocfs2_dir_lookup_result *res);
 int __ocfs2_add_entry(handle_t *handle,
 		      struct inode *dir,
 		      const char *name, int namelen,
 		      struct inode *inode, u64 blkno,
 		      struct buffer_head *parent_fe_bh,
-		      struct buffer_head *insert_bh);
+		      struct ocfs2_dir_lookup_result *lookup);
 static inline int ocfs2_add_entry(handle_t *handle,
 				  struct dentry *dentry,
 				  struct inode *inode, u64 blkno,
 				  struct buffer_head *parent_fe_bh,
-				  struct buffer_head *insert_bh)
+				  struct ocfs2_dir_lookup_result *lookup)
 {
 	return __ocfs2_add_entry(handle, dentry->d_parent->d_inode,
 				 dentry->d_name.name, dentry->d_name.len,
-				 inode, blkno, parent_fe_bh, insert_bh);
+				 inode, blkno, parent_fe_bh, lookup);
 }
 int ocfs2_update_entry(struct inode *dir, handle_t *handle,
-		       struct buffer_head *de_bh, struct ocfs2_dir_entry *de,
+		       struct ocfs2_dir_lookup_result *res,
 		       struct inode *new_entry_inode);
 
 int ocfs2_check_dir_for_entry(struct inode *dir,
 			      const char *name,
 			      int namelen);
 int ocfs2_empty_dir(struct inode *inode);
+
 int ocfs2_find_files_on_disk(const char *name,
 			     int namelen,
 			     u64 *blkno,
 			     struct inode *inode,
-			     struct buffer_head **dirent_bh,
-			     struct ocfs2_dir_entry **dirent);
+			     struct ocfs2_dir_lookup_result *res);
 int ocfs2_lookup_ino_from_name(struct inode *dir, const char *name,
 			       int namelen, u64 *blkno);
 int ocfs2_readdir(struct file *filp, void *dirent, filldir_t filldir);
@@ -74,14 +100,17 @@ int ocfs2_prepare_dir_for_insert(struct ocfs2_super *osb,
 				 struct buffer_head *parent_fe_bh,
 				 const char *name,
 				 int namelen,
-				 struct buffer_head **ret_de_bh);
+				 struct ocfs2_dir_lookup_result *lookup);
 struct ocfs2_alloc_context;
 int ocfs2_fill_new_dir(struct ocfs2_super *osb,
 		       handle_t *handle,
 		       struct inode *parent,
 		       struct inode *inode,
 		       struct buffer_head *fe_bh,
-		       struct ocfs2_alloc_context *data_ac);
+		       struct ocfs2_alloc_context *data_ac,
+		       struct ocfs2_alloc_context *meta_ac);
+
+int ocfs2_dx_dir_truncate(struct inode *dir, struct buffer_head *di_bh);
 
 struct ocfs2_dir_block_trailer *ocfs2_dir_trailer_from_size(int blocksize,
 							    void *data);

fs/ocfs2/dlm/dlmcommon.h

@@ -52,16 +52,12 @@
 enum dlm_mle_type {
 	DLM_MLE_BLOCK,
 	DLM_MLE_MASTER,
-	DLM_MLE_MIGRATION
-};
-
-struct dlm_lock_name {
-	u8 len;
-	u8 name[DLM_LOCKID_NAME_MAX];
+	DLM_MLE_MIGRATION,
+	DLM_MLE_NUM_TYPES
 };
 
 struct dlm_master_list_entry {
-	struct list_head list;
+	struct hlist_node master_hash_node;
 	struct list_head hb_events;
 	struct dlm_ctxt *dlm;
 	spinlock_t spinlock;
@@ -78,10 +74,10 @@ struct dlm_master_list_entry {
 	enum dlm_mle_type type;
 	struct o2hb_callback_func mle_hb_up;
 	struct o2hb_callback_func mle_hb_down;
-	union {
-		struct dlm_lock_resource *res;
-		struct dlm_lock_name name;
-	} u;
+	struct dlm_lock_resource *mleres;
+	unsigned char mname[DLM_LOCKID_NAME_MAX];
+	unsigned int mnamelen;
+	unsigned int mnamehash;
 };
 
 enum dlm_ast_type {
@@ -151,13 +147,14 @@ struct dlm_ctxt
 	unsigned long recovery_map[BITS_TO_LONGS(O2NM_MAX_NODES)];
 	struct dlm_recovery_ctxt reco;
 	spinlock_t master_lock;
-	struct list_head master_list;
+	struct hlist_head **master_hash;
 	struct list_head mle_hb_events;
 
 	/* these give a really vague idea of the system load */
-	atomic_t local_resources;
-	atomic_t remote_resources;
-	atomic_t unknown_resources;
+	atomic_t mle_tot_count[DLM_MLE_NUM_TYPES];
+	atomic_t mle_cur_count[DLM_MLE_NUM_TYPES];
+	atomic_t res_tot_count;
+	atomic_t res_cur_count;
 
 	struct dlm_debug_ctxt *dlm_debug_ctxt;
 	struct dentry *dlm_debugfs_subroot;
@@ -195,6 +192,13 @@ static inline struct hlist_head *dlm_lockres_hash(struct dlm_ctxt *dlm, unsigned
 	return dlm->lockres_hash[(i / DLM_BUCKETS_PER_PAGE) % DLM_HASH_PAGES] + (i % DLM_BUCKETS_PER_PAGE);
 }
 
+static inline struct hlist_head *dlm_master_hash(struct dlm_ctxt *dlm,
+						 unsigned i)
+{
+	return dlm->master_hash[(i / DLM_BUCKETS_PER_PAGE) % DLM_HASH_PAGES] +
+			(i % DLM_BUCKETS_PER_PAGE);
+}
+
 /* these keventd work queue items are for less-frequently
  * called functions that cannot be directly called from the
  * net message handlers for some reason, usually because
@@ -848,9 +852,7 @@ struct dlm_lock_resource * dlm_lookup_lockres(struct dlm_ctxt *dlm,
 					      unsigned int len);
 
 int dlm_is_host_down(int errno);
-void dlm_change_lockres_owner(struct dlm_ctxt *dlm,
-			      struct dlm_lock_resource *res,
-			      u8 owner);
+
 struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
 						 const char *lockid,
 						 int namelen,
@@ -1008,6 +1010,9 @@ static inline void __dlm_wait_on_lockres(struct dlm_lock_resource *res)
 					  DLM_LOCK_RES_MIGRATING));
 }
 
+void __dlm_unlink_mle(struct dlm_ctxt *dlm, struct dlm_master_list_entry *mle);
+void __dlm_insert_mle(struct dlm_ctxt *dlm, struct dlm_master_list_entry *mle);
+
 /* create/destroy slab caches */
 int dlm_init_master_caches(void);
 void dlm_destroy_master_caches(void);
@@ -1110,6 +1115,23 @@ static inline int dlm_node_iter_next(struct dlm_node_iter *iter)
 	return bit;
 }
 
+static inline void dlm_set_lockres_owner(struct dlm_ctxt *dlm,
+					 struct dlm_lock_resource *res,
+					 u8 owner)
+{
+	assert_spin_locked(&res->spinlock);
+
+	res->owner = owner;
+}
 
+static inline void dlm_change_lockres_owner(struct dlm_ctxt *dlm,
+					    struct dlm_lock_resource *res,
+					    u8 owner)
+{
+	assert_spin_locked(&res->spinlock);
+
+	if (owner != res->owner)
+		dlm_set_lockres_owner(dlm, res, owner);
+}
 
 #endif /* DLMCOMMON_H */

fs/ocfs2/dlm/dlmdebug.c

@@ -287,18 +287,8 @@ static int stringify_nodemap(unsigned long *nodemap, int maxnodes,
 static int dump_mle(struct dlm_master_list_entry *mle, char *buf, int len)
 {
 	int out = 0;
-	unsigned int namelen;
-	const char *name;
 	char *mle_type;
 
-	if (mle->type != DLM_MLE_MASTER) {
-		namelen = mle->u.name.len;
-		name = mle->u.name.name;
-	} else {
-		namelen = mle->u.res->lockname.len;
-		name = mle->u.res->lockname.name;
-	}
-
 	if (mle->type == DLM_MLE_BLOCK)
 		mle_type = "BLK";
 	else if (mle->type == DLM_MLE_MASTER)
@@ -306,7 +296,7 @@ static int dump_mle(struct dlm_master_list_entry *mle, char *buf, int len)
 	else
 		mle_type = "MIG";
 
-	out += stringify_lockname(name, namelen, buf + out, len - out);
+	out += stringify_lockname(mle->mname, mle->mnamelen, buf + out, len - out);
 	out += snprintf(buf + out, len - out,
 			"\t%3s\tmas=%3u\tnew=%3u\tevt=%1d\tuse=%1d\tref=%3d\n",
 			mle_type, mle->master, mle->new_master,
@@ -501,23 +491,33 @@ static struct file_operations debug_purgelist_fops = {
 static int debug_mle_print(struct dlm_ctxt *dlm, struct debug_buffer *db)
 {
 	struct dlm_master_list_entry *mle;
-	int out = 0;
-	unsigned long total = 0;
+	struct hlist_head *bucket;
+	struct hlist_node *list;
+	int i, out = 0;
+	unsigned long total = 0, longest = 0, bktcnt;
 
 	out += snprintf(db->buf + out, db->len - out,
 			"Dumping MLEs for Domain: %s\n", dlm->name);
 
 	spin_lock(&dlm->master_lock);
-	list_for_each_entry(mle, &dlm->master_list, list) {
-		++total;
-		if (db->len - out < 200)
-			continue;
-		out += dump_mle(mle, db->buf + out, db->len - out);
+	for (i = 0; i < DLM_HASH_BUCKETS; i++) {
+		bucket = dlm_master_hash(dlm, i);
+		hlist_for_each(list, bucket) {
+			mle = hlist_entry(list, struct dlm_master_list_entry,
+					  master_hash_node);
+			++total;
+			++bktcnt;
+			if (db->len - out < 200)
+				continue;
+			out += dump_mle(mle, db->buf + out, db->len - out);
+		}
+		longest = max(longest, bktcnt);
+		bktcnt = 0;
 	}
 	spin_unlock(&dlm->master_lock);
 
 	out += snprintf(db->buf + out, db->len - out,
-			"Total on list: %ld\n", total);
+			"Total: %ld, Longest: %ld\n", total, longest);
 	return out;
 }
 
@@ -756,12 +756,8 @@ static int debug_state_print(struct dlm_ctxt *dlm, struct debug_buffer *db)
 	int out = 0;
 	struct dlm_reco_node_data *node;
 	char *state;
-	int lres, rres, ures, tres;
-
-	lres = atomic_read(&dlm->local_resources);
-	rres = atomic_read(&dlm->remote_resources);
-	ures = atomic_read(&dlm->unknown_resources);
-	tres = lres + rres + ures;
+	int cur_mles = 0, tot_mles = 0;
+	int i;
 
 	spin_lock(&dlm->spinlock);
 
@@ -804,21 +800,48 @@ static int debug_state_print(struct dlm_ctxt *dlm, struct debug_buffer *db)
 				 db->buf + out, db->len - out);
 	out += snprintf(db->buf + out, db->len - out, "\n");
 
-	/* Mastered Resources Total: xxx  Locally: xxx  Remotely: ... */
+	/* Lock Resources: xxx (xxx) */
+	out += snprintf(db->buf + out, db->len - out,
+			"Lock Resources: %d (%d)\n",
+			atomic_read(&dlm->res_cur_count),
+			atomic_read(&dlm->res_tot_count));
+
+	for (i = 0; i < DLM_MLE_NUM_TYPES; ++i)
+		tot_mles += atomic_read(&dlm->mle_tot_count[i]);
+
+	for (i = 0; i < DLM_MLE_NUM_TYPES; ++i)
+		cur_mles += atomic_read(&dlm->mle_cur_count[i]);
+
+	/* MLEs: xxx (xxx) */
+	out += snprintf(db->buf + out, db->len - out,
+			"MLEs: %d (%d)\n", cur_mles, tot_mles);
+
+	/*  Blocking: xxx (xxx) */
+	out += snprintf(db->buf + out, db->len - out,
+			"  Blocking: %d (%d)\n",
+			atomic_read(&dlm->mle_cur_count[DLM_MLE_BLOCK]),
+			atomic_read(&dlm->mle_tot_count[DLM_MLE_BLOCK]));
+
+	/*  Mastery: xxx (xxx) */
+	out += snprintf(db->buf + out, db->len - out,
+			"  Mastery: %d (%d)\n",
+			atomic_read(&dlm->mle_cur_count[DLM_MLE_MASTER]),
+			atomic_read(&dlm->mle_tot_count[DLM_MLE_MASTER]));
+
+	/*  Migration: xxx (xxx) */
 	out += snprintf(db->buf + out, db->len - out,
-			"Mastered Resources Total: %d  Locally: %d  "
-			"Remotely: %d  Unknown: %d\n",
-			tres, lres, rres, ures);
+			"  Migration: %d (%d)\n",
+			atomic_read(&dlm->mle_cur_count[DLM_MLE_MIGRATION]),
+			atomic_read(&dlm->mle_tot_count[DLM_MLE_MIGRATION]));
 
 	/* Lists: Dirty=Empty  Purge=InUse  PendingASTs=Empty  ... */
 	out += snprintf(db->buf + out, db->len - out,
 			"Lists: Dirty=%s  Purge=%s  PendingASTs=%s  "
-			"PendingBASTs=%s  Master=%s\n",
+			"PendingBASTs=%s\n",
 			(list_empty(&dlm->dirty_list) ? "Empty" : "InUse"),
 			(list_empty(&dlm->purge_list) ? "Empty" : "InUse"),
 			(list_empty(&dlm->pending_asts) ? "Empty" : "InUse"),
-			(list_empty(&dlm->pending_basts) ? "Empty" : "InUse"),
-			(list_empty(&dlm->master_list) ? "Empty" : "InUse"));
+			(list_empty(&dlm->pending_basts) ? "Empty" : "InUse"));
 
 	/* Purge Count: xxx  Refs: xxx */
 	out += snprintf(db->buf + out, db->len - out,

fs/ocfs2/dlm/dlmdomain.c

@@ -304,6 +304,9 @@ static void dlm_free_ctxt_mem(struct dlm_ctxt *dlm)
 	if (dlm->lockres_hash)
 		dlm_free_pagevec((void **)dlm->lockres_hash, DLM_HASH_PAGES);
 
+	if (dlm->master_hash)
+		dlm_free_pagevec((void **)dlm->master_hash, DLM_HASH_PAGES);
+
 	if (dlm->name)
 		kfree(dlm->name);
 
@@ -1534,12 +1537,27 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain,
 	for (i = 0; i < DLM_HASH_BUCKETS; i++)
 		INIT_HLIST_HEAD(dlm_lockres_hash(dlm, i));
 
+	dlm->master_hash = (struct hlist_head **)
+				dlm_alloc_pagevec(DLM_HASH_PAGES);
+	if (!dlm->master_hash) {
+		mlog_errno(-ENOMEM);
+		dlm_free_pagevec((void **)dlm->lockres_hash, DLM_HASH_PAGES);
+		kfree(dlm->name);
+		kfree(dlm);
+		dlm = NULL;
+		goto leave;
+	}
+
+	for (i = 0; i < DLM_HASH_BUCKETS; i++)
+		INIT_HLIST_HEAD(dlm_master_hash(dlm, i));
+
 	strcpy(dlm->name, domain);
 	dlm->key = key;
 	dlm->node_num = o2nm_this_node();
 
 	ret = dlm_create_debugfs_subroot(dlm);
 	if (ret < 0) {
+		dlm_free_pagevec((void **)dlm->master_hash, DLM_HASH_PAGES);
 		dlm_free_pagevec((void **)dlm->lockres_hash, DLM_HASH_PAGES);
 		kfree(dlm->name);
 		kfree(dlm);
@@ -1579,7 +1597,6 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain,
 	init_waitqueue_head(&dlm->reco.event);
 	init_waitqueue_head(&dlm->ast_wq);
 	init_waitqueue_head(&dlm->migration_wq);
-	INIT_LIST_HEAD(&dlm->master_list);
 	INIT_LIST_HEAD(&dlm->mle_hb_events);
 
 	dlm->joining_node = DLM_LOCK_RES_OWNER_UNKNOWN;
@@ -1587,9 +1604,13 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain,
 
 	dlm->reco.new_master = O2NM_INVALID_NODE_NUM;
 	dlm->reco.dead_node = O2NM_INVALID_NODE_NUM;
-	atomic_set(&dlm->local_resources, 0);
-	atomic_set(&dlm->remote_resources, 0);
-	atomic_set(&dlm->unknown_resources, 0);
+
+	atomic_set(&dlm->res_tot_count, 0);
+	atomic_set(&dlm->res_cur_count, 0);
+	for (i = 0; i < DLM_MLE_NUM_TYPES; ++i) {
+		atomic_set(&dlm->mle_tot_count[i], 0);
+		atomic_set(&dlm->mle_cur_count[i], 0);
+	}
 
 	spin_lock_init(&dlm->work_lock);
 	INIT_LIST_HEAD(&dlm->work_list);

fs/ocfs2/dlm/dlmthread.c

@@ -162,12 +162,28 @@ static int dlm_purge_lockres(struct dlm_ctxt *dlm,
 
 	spin_lock(&res->spinlock);
 	if (!__dlm_lockres_unused(res)) {
-		spin_unlock(&res->spinlock);
 		mlog(0, "%s:%.*s: tried to purge but not unused\n",
 		     dlm->name, res->lockname.len, res->lockname.name);
-		return -ENOTEMPTY;
+		__dlm_print_one_lock_resource(res);
+		spin_unlock(&res->spinlock);
+		BUG();
 	}
+
+	if (res->state & DLM_LOCK_RES_MIGRATING) {
+		mlog(0, "%s:%.*s: Delay dropref as this lockres is "
+		     "being remastered\n", dlm->name, res->lockname.len,
+		     res->lockname.name);
+		/* Re-add the lockres to the end of the purge list */
+		if (!list_empty(&res->purge)) {
+			list_del_init(&res->purge);
+			list_add_tail(&res->purge, &dlm->purge_list);
+		}
+		spin_unlock(&res->spinlock);
+		return 0;
+	}
+
 	master = (res->owner == dlm->node_num);
+
 	if (!master)
 		res->state |= DLM_LOCK_RES_DROPPING_REF;
 	spin_unlock(&res->spinlock);

fs/ocfs2/dlmglue.c

@@ -244,6 +244,10 @@ static struct ocfs2_lock_res_ops ocfs2_rename_lops = {
 	.flags		= 0,
 };
 
+static struct ocfs2_lock_res_ops ocfs2_nfs_sync_lops = {
+	.flags		= 0,
+};
+
 static struct ocfs2_lock_res_ops ocfs2_dentry_lops = {
 	.get_osb	= ocfs2_get_dentry_osb,
 	.post_unlock	= ocfs2_dentry_post_unlock,
@@ -622,6 +626,17 @@ static void ocfs2_rename_lock_res_init(struct ocfs2_lock_res *res,
 				   &ocfs2_rename_lops, osb);
 }
 
+static void ocfs2_nfs_sync_lock_res_init(struct ocfs2_lock_res *res,
+					 struct ocfs2_super *osb)
+{
+	/* nfs_sync lockres doesn't come from a slab so we call init
+	 * once on it manually.  */
+	ocfs2_lock_res_init_once(res);
+	ocfs2_build_lock_name(OCFS2_LOCK_TYPE_NFS_SYNC, 0, 0, res->l_name);
+	ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_NFS_SYNC,
+				   &ocfs2_nfs_sync_lops, osb);
+}
+
 void ocfs2_file_lock_res_init(struct ocfs2_lock_res *lockres,
 			      struct ocfs2_file_private *fp)
 {
@@ -2417,6 +2432,34 @@ void ocfs2_rename_unlock(struct ocfs2_super *osb)
 		ocfs2_cluster_unlock(osb, lockres, DLM_LOCK_EX);
 }
 
+int ocfs2_nfs_sync_lock(struct ocfs2_super *osb, int ex)
+{
+	int status;
+	struct ocfs2_lock_res *lockres = &osb->osb_nfs_sync_lockres;
+
+	if (ocfs2_is_hard_readonly(osb))
+		return -EROFS;
+
+	if (ocfs2_mount_local(osb))
+		return 0;
+
+	status = ocfs2_cluster_lock(osb, lockres, ex ? LKM_EXMODE : LKM_PRMODE,
+				    0, 0);
+	if (status < 0)
+		mlog(ML_ERROR, "lock on nfs sync lock failed %d\n", status);
+
+	return status;
+}
+
+void ocfs2_nfs_sync_unlock(struct ocfs2_super *osb, int ex)
+{
+	struct ocfs2_lock_res *lockres = &osb->osb_nfs_sync_lockres;
+
+	if (!ocfs2_mount_local(osb))
+		ocfs2_cluster_unlock(osb, lockres,
+				     ex ? LKM_EXMODE : LKM_PRMODE);
+}
+
 int ocfs2_dentry_lock(struct dentry *dentry, int ex)
 {
 	int ret;
@@ -2798,6 +2841,7 @@ int ocfs2_dlm_init(struct ocfs2_super *osb)
 local:
 	ocfs2_super_lock_res_init(&osb->osb_super_lockres, osb);
 	ocfs2_rename_lock_res_init(&osb->osb_rename_lockres, osb);
+	ocfs2_nfs_sync_lock_res_init(&osb->osb_nfs_sync_lockres, osb);
 
 	osb->cconn = conn;
 
@@ -2833,6 +2877,7 @@ void ocfs2_dlm_shutdown(struct ocfs2_super *osb,
 
 	ocfs2_lock_res_free(&osb->osb_super_lockres);
 	ocfs2_lock_res_free(&osb->osb_rename_lockres);
+	ocfs2_lock_res_free(&osb->osb_nfs_sync_lockres);
 
 	ocfs2_cluster_disconnect(osb->cconn, hangup_pending);
 	osb->cconn = NULL;
@@ -3015,6 +3060,7 @@ static void ocfs2_drop_osb_locks(struct ocfs2_super *osb)
 {
 	ocfs2_simple_drop_lockres(osb, &osb->osb_super_lockres);
 	ocfs2_simple_drop_lockres(osb, &osb->osb_rename_lockres);
+	ocfs2_simple_drop_lockres(osb, &osb->osb_nfs_sync_lockres);
 }
 
 int ocfs2_drop_inode_locks(struct inode *inode)

fs/ocfs2/dlmglue.h

@@ -115,6 +115,8 @@ void ocfs2_super_unlock(struct ocfs2_super *osb,
 			int ex);
 int ocfs2_rename_lock(struct ocfs2_super *osb);
 void ocfs2_rename_unlock(struct ocfs2_super *osb);
+int ocfs2_nfs_sync_lock(struct ocfs2_super *osb, int ex);
+void ocfs2_nfs_sync_unlock(struct ocfs2_super *osb, int ex);
 int ocfs2_dentry_lock(struct dentry *dentry, int ex);
 void ocfs2_dentry_unlock(struct dentry *dentry, int ex);
 int ocfs2_file_lock(struct file *file, int ex, int trylock);

fs/ocfs2/export.c

@@ -31,6 +31,7 @@
 
 #include "ocfs2.h"
 
+#include "alloc.h"
 #include "dir.h"
 #include "dlmglue.h"
 #include "dcache.h"
@@ -38,6 +39,7 @@
 #include "inode.h"
 
 #include "buffer_head_io.h"
+#include "suballoc.h"
 
 struct ocfs2_inode_handle
 {
@@ -49,29 +51,97 @@ static struct dentry *ocfs2_get_dentry(struct super_block *sb,
 		struct ocfs2_inode_handle *handle)
 {
 	struct inode *inode;
+	struct ocfs2_super *osb = OCFS2_SB(sb);
+	u64 blkno = handle->ih_blkno;
+	int status, set;
 	struct dentry *result;
 
 	mlog_entry("(0x%p, 0x%p)\n", sb, handle);
 
-	if (handle->ih_blkno == 0) {
-		mlog_errno(-ESTALE);
-		return ERR_PTR(-ESTALE);
+	if (blkno == 0) {
+		mlog(0, "nfs wants inode with blkno: 0\n");
+		result = ERR_PTR(-ESTALE);
+		goto bail;
+	}
+
+	inode = ocfs2_ilookup(sb, blkno);
+	/*
+	 * If the inode exists in memory, we only need to check it's
+	 * generation number
+	 */
+	if (inode)
+		goto check_gen;
+
+	/*
+	 * This will synchronize us against ocfs2_delete_inode() on
+	 * all nodes
+	 */
+	status = ocfs2_nfs_sync_lock(osb, 1);
+	if (status < 0) {
+		mlog(ML_ERROR, "getting nfs sync lock(EX) failed %d\n", status);
+		goto check_err;
+	}
+
+	status = ocfs2_test_inode_bit(osb, blkno, &set);
+	if (status < 0) {
+		if (status == -EINVAL) {
+			/*
+			 * The blkno NFS gave us doesn't even show up
+			 * as an inode, we return -ESTALE to be
+			 * nice
+			 */
+			mlog(0, "test inode bit failed %d\n", status);
+			status = -ESTALE;
+		} else {
+			mlog(ML_ERROR, "test inode bit failed %d\n", status);
+		}
+		goto unlock_nfs_sync;
+	}
+
+	/* If the inode allocator bit is clear, this inode must be stale */
+	if (!set) {
+		mlog(0, "inode %llu suballoc bit is clear\n", blkno);
+		status = -ESTALE;
+		goto unlock_nfs_sync;
 	}
 
-	inode = ocfs2_iget(OCFS2_SB(sb), handle->ih_blkno, 0, 0);
+	inode = ocfs2_iget(osb, blkno, 0, 0);
 
-	if (IS_ERR(inode))
-		return (void *)inode;
+unlock_nfs_sync:
+	ocfs2_nfs_sync_unlock(osb, 1);
 
+check_err:
+	if (status < 0) {
+		if (status == -ESTALE) {
+			mlog(0, "stale inode ino: %llu generation: %u\n",
+			     blkno, handle->ih_generation);
+		}
+		result = ERR_PTR(status);
+		goto bail;
+	}
+
+	if (IS_ERR(inode)) {
+		mlog_errno(PTR_ERR(inode));
+		result = (void *)inode;
+		goto bail;
+	}
+
+check_gen:
 	if (handle->ih_generation != inode->i_generation) {
 		iput(inode);
-		return ERR_PTR(-ESTALE);
+		mlog(0, "stale inode ino: %llu generation: %u\n", blkno,
+		     handle->ih_generation);
+		result = ERR_PTR(-ESTALE);
+		goto bail;
 	}
 
 	result = d_obtain_alias(inode);
 	if (!IS_ERR(result))
 		result->d_op = &ocfs2_dentry_ops;
+	else
+		mlog_errno(PTR_ERR(result));
 
+bail:
 	mlog_exit_ptr(result);
 	return result;
 }

fs/ocfs2/inode.c

@@ -38,6 +38,7 @@
 #include "ocfs2.h"
 
 #include "alloc.h"
+#include "dir.h"
 #include "blockcheck.h"
 #include "dlmglue.h"
 #include "extent_map.h"
@@ -112,6 +113,17 @@ void ocfs2_get_inode_flags(struct ocfs2_inode_info *oi)
 		oi->ip_attr |= OCFS2_DIRSYNC_FL;
 }
 
+struct inode *ocfs2_ilookup(struct super_block *sb, u64 blkno)
+{
+	struct ocfs2_find_inode_args args;
+
+	args.fi_blkno = blkno;
+	args.fi_flags = 0;
+	args.fi_ino = ino_from_blkno(sb, blkno);
+	args.fi_sysfile_type = 0;
+
+	return ilookup5(sb, blkno, ocfs2_find_actor, &args);
+}
 struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno, unsigned flags,
 			 int sysfile_type)
 {
@@ -275,7 +287,7 @@ void ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
 		     (unsigned long long)OCFS2_I(inode)->ip_blkno,
 		     (unsigned long long)le64_to_cpu(fe->i_blkno));
 
-	inode->i_nlink = le16_to_cpu(fe->i_links_count);
+	inode->i_nlink = ocfs2_read_links_count(fe);
 
 	if (fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL)) {
 		OCFS2_I(inode)->ip_flags |= OCFS2_INODE_SYSTEM_FILE;
@@ -351,6 +363,8 @@ void ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
 
 	ocfs2_set_inode_flags(inode);
 
+	OCFS2_I(inode)->ip_last_used_slot = 0;
+	OCFS2_I(inode)->ip_last_used_group = 0;
 	mlog_exit_void();
 }
 
@@ -606,7 +620,7 @@ static int ocfs2_remove_inode(struct inode *inode,
 	}
 
 	handle = ocfs2_start_trans(osb, OCFS2_DELETE_INODE_CREDITS +
-					ocfs2_quota_trans_credits(inode->i_sb));
+				   ocfs2_quota_trans_credits(inode->i_sb));
 	if (IS_ERR(handle)) {
 		status = PTR_ERR(handle);
 		mlog_errno(status);
@@ -740,6 +754,15 @@ static int ocfs2_wipe_inode(struct inode *inode,
 		goto bail_unlock_dir;
 	}
 
+	/* Remove any dir index tree */
+	if (S_ISDIR(inode->i_mode)) {
+		status = ocfs2_dx_dir_truncate(inode, di_bh);
+		if (status) {
+			mlog_errno(status);
+			goto bail_unlock_dir;
+		}
+	}
+
 	/*Free extended attribute resources associated with this inode.*/
 	status = ocfs2_xattr_remove(inode, di_bh);
 	if (status < 0) {
@@ -949,6 +972,17 @@ void ocfs2_delete_inode(struct inode *inode)
 		goto bail;
 	}
 
+	/*
+	 * Synchronize us against ocfs2_get_dentry. We take this in
+	 * shared mode so that all nodes can still concurrently
+	 * process deletes.
+	 */
+	status = ocfs2_nfs_sync_lock(OCFS2_SB(inode->i_sb), 0);
+	if (status < 0) {
+		mlog(ML_ERROR, "getting nfs sync lock(PR) failed %d\n", status);
+		ocfs2_cleanup_delete_inode(inode, 0);
+		goto bail_unblock;
+	}
 	/* Lock down the inode. This gives us an up to date view of
 	 * it's metadata (for verification), and allows us to
 	 * serialize delete_inode on multiple nodes.
@@ -962,7 +996,7 @@ void ocfs2_delete_inode(struct inode *inode)
 		if (status != -ENOENT)
 			mlog_errno(status);
 		ocfs2_cleanup_delete_inode(inode, 0);
-		goto bail_unblock;
+		goto bail_unlock_nfs_sync;
 	}
 
 	/* Query the cluster. This will be the final decision made
@@ -1005,6 +1039,10 @@ void ocfs2_delete_inode(struct inode *inode)
 bail_unlock_inode:
 	ocfs2_inode_unlock(inode, 1);
 	brelse(di_bh);
+
+bail_unlock_nfs_sync:
+	ocfs2_nfs_sync_unlock(OCFS2_SB(inode->i_sb), 0);
+
 bail_unblock:
 	status = sigprocmask(SIG_SETMASK, &oldset, NULL);
 	if (status < 0)
@@ -1205,7 +1243,7 @@ int ocfs2_mark_inode_dirty(handle_t *handle,
 	spin_unlock(&OCFS2_I(inode)->ip_lock);
 
 	fe->i_size = cpu_to_le64(i_size_read(inode));
-	fe->i_links_count = cpu_to_le16(inode->i_nlink);
+	ocfs2_set_links_count(fe, inode->i_nlink);
 	fe->i_uid = cpu_to_le32(inode->i_uid);
 	fe->i_gid = cpu_to_le32(inode->i_gid);
 	fe->i_mode = cpu_to_le16(inode->i_mode);
@@ -1242,7 +1280,7 @@ void ocfs2_refresh_inode(struct inode *inode,
 	OCFS2_I(inode)->ip_dyn_features = le16_to_cpu(fe->i_dyn_features);
 	ocfs2_set_inode_flags(inode);
 	i_size_write(inode, le64_to_cpu(fe->i_size));
-	inode->i_nlink = le16_to_cpu(fe->i_links_count);
+	inode->i_nlink = ocfs2_read_links_count(fe);
 	inode->i_uid = le32_to_cpu(fe->i_uid);
 	inode->i_gid = le32_to_cpu(fe->i_gid);
 	inode->i_mode = le16_to_cpu(fe->i_mode);

fs/ocfs2/inode.h

@@ -72,6 +72,10 @@ struct ocfs2_inode_info
 
 	struct inode			vfs_inode;
 	struct jbd2_inode		ip_jinode;
+
+	/* Only valid if the inode is the dir. */
+	u32				ip_last_used_slot;
+	u64				ip_last_used_group;
 };
 
 /*
@@ -124,6 +128,7 @@ void ocfs2_drop_inode(struct inode *inode);
 /* Flags for ocfs2_iget() */
 #define OCFS2_FI_FLAG_SYSFILE		0x1
 #define OCFS2_FI_FLAG_ORPHAN_RECOVERY	0x2
+struct inode *ocfs2_ilookup(struct super_block *sb, u64 feoff);
 struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 feoff, unsigned flags,
 			 int sysfile_type);
 int ocfs2_inode_init_private(struct inode *inode);

fs/ocfs2/journal.c

@@ -65,6 +65,11 @@ static int ocfs2_trylock_journal(struct ocfs2_super *osb,
 static int ocfs2_recover_orphans(struct ocfs2_super *osb,
 				 int slot);
 static int ocfs2_commit_thread(void *arg);
+static void ocfs2_queue_recovery_completion(struct ocfs2_journal *journal,
+					    int slot_num,
+					    struct ocfs2_dinode *la_dinode,
+					    struct ocfs2_dinode *tl_dinode,
+					    struct ocfs2_quota_recovery *qrec);
 
 static inline int ocfs2_wait_on_mount(struct ocfs2_super *osb)
 {
@@ -76,18 +81,97 @@ static inline int ocfs2_wait_on_quotas(struct ocfs2_super *osb)
 	return __ocfs2_wait_on_mount(osb, 1);
 }
 
-
-
 /*
- * The recovery_list is a simple linked list of node numbers to recover.
- * It is protected by the recovery_lock.
+ * This replay_map is to track online/offline slots, so we could recover
+ * offline slots during recovery and mount
  */
 
-struct ocfs2_recovery_map {
-	unsigned int rm_used;
-	unsigned int *rm_entries;
+enum ocfs2_replay_state {
+	REPLAY_UNNEEDED = 0,	/* Replay is not needed, so ignore this map */
+	REPLAY_NEEDED, 		/* Replay slots marked in rm_replay_slots */
+	REPLAY_DONE 		/* Replay was already queued */
 };
 
+struct ocfs2_replay_map {
+	unsigned int rm_slots;
+	enum ocfs2_replay_state rm_state;
+	unsigned char rm_replay_slots[0];
+};
+
+void ocfs2_replay_map_set_state(struct ocfs2_super *osb, int state)
+{
+	if (!osb->replay_map)
+		return;
+
+	/* If we've already queued the replay, we don't have any more to do */
+	if (osb->replay_map->rm_state == REPLAY_DONE)
+		return;
+
+	osb->replay_map->rm_state = state;
+}
+
+int ocfs2_compute_replay_slots(struct ocfs2_super *osb)
+{
+	struct ocfs2_replay_map *replay_map;
+	int i, node_num;
+
+	/* If replay map is already set, we don't do it again */
+	if (osb->replay_map)
+		return 0;
+
+	replay_map = kzalloc(sizeof(struct ocfs2_replay_map) +
+			     (osb->max_slots * sizeof(char)), GFP_KERNEL);
+
+	if (!replay_map) {
+		mlog_errno(-ENOMEM);
+		return -ENOMEM;
+	}
+
+	spin_lock(&osb->osb_lock);
+
+	replay_map->rm_slots = osb->max_slots;
+	replay_map->rm_state = REPLAY_UNNEEDED;
+
+	/* set rm_replay_slots for offline slot(s) */
+	for (i = 0; i < replay_map->rm_slots; i++) {
+		if (ocfs2_slot_to_node_num_locked(osb, i, &node_num) == -ENOENT)
+			replay_map->rm_replay_slots[i] = 1;
+	}
+
+	osb->replay_map = replay_map;
+	spin_unlock(&osb->osb_lock);
+	return 0;
+}
+
+void ocfs2_queue_replay_slots(struct ocfs2_super *osb)
+{
+	struct ocfs2_replay_map *replay_map = osb->replay_map;
+	int i;
+
+	if (!replay_map)
+		return;
+
+	if (replay_map->rm_state != REPLAY_NEEDED)
+		return;
+
+	for (i = 0; i < replay_map->rm_slots; i++)
+		if (replay_map->rm_replay_slots[i])
+			ocfs2_queue_recovery_completion(osb->journal, i, NULL,
+							NULL, NULL);
+	replay_map->rm_state = REPLAY_DONE;
+}
+
+void ocfs2_free_replay_slots(struct ocfs2_super *osb)
+{
+	struct ocfs2_replay_map *replay_map = osb->replay_map;
+
+	if (!osb->replay_map)
+		return;
+
+	kfree(replay_map);
+	osb->replay_map = NULL;
+}
+
 int ocfs2_recovery_init(struct ocfs2_super *osb)
 {
 	struct ocfs2_recovery_map *rm;
@@ -496,6 +580,22 @@ static struct ocfs2_triggers dq_triggers = {
 	},
 };
 
+static struct ocfs2_triggers dr_triggers = {
+	.ot_triggers = {
+		.t_commit = ocfs2_commit_trigger,
+		.t_abort = ocfs2_abort_trigger,
+	},
+	.ot_offset	= offsetof(struct ocfs2_dx_root_block, dr_check),
+};
+
+static struct ocfs2_triggers dl_triggers = {
+	.ot_triggers = {
+		.t_commit = ocfs2_commit_trigger,
+		.t_abort = ocfs2_abort_trigger,
+	},
+	.ot_offset	= offsetof(struct ocfs2_dx_leaf, dl_check),
+};
+
 static int __ocfs2_journal_access(handle_t *handle,
 				  struct inode *inode,
 				  struct buffer_head *bh,
@@ -600,6 +700,20 @@ int ocfs2_journal_access_dq(handle_t *handle, struct inode *inode,
 				      type);
 }
 
+int ocfs2_journal_access_dr(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, &dr_triggers,
+				      type);
+}
+
+int ocfs2_journal_access_dl(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, &dl_triggers,
+				      type);
+}
+
 int ocfs2_journal_access(handle_t *handle, struct inode *inode,
 			 struct buffer_head *bh, int type)
 {
@@ -1176,24 +1290,24 @@ static void ocfs2_queue_recovery_completion(struct ocfs2_journal *journal,
 }
 
 /* Called by the mount code to queue recovery the last part of
- * recovery for it's own slot. */
+ * recovery for it's own and offline slot(s). */
 void ocfs2_complete_mount_recovery(struct ocfs2_super *osb)
 {
 	struct ocfs2_journal *journal = osb->journal;
 
-	if (osb->dirty) {
-		/* No need to queue up our truncate_log as regular
-		 * cleanup will catch that. */
-		ocfs2_queue_recovery_completion(journal,
-						osb->slot_num,
-						osb->local_alloc_copy,
-						NULL,
-						NULL);
-		ocfs2_schedule_truncate_log_flush(osb, 0);
+	/* No need to queue up our truncate_log as regular cleanup will catch
+	 * that */
+	ocfs2_queue_recovery_completion(journal, osb->slot_num,
+					osb->local_alloc_copy, NULL, NULL);
+	ocfs2_schedule_truncate_log_flush(osb, 0);
 
-		osb->local_alloc_copy = NULL;
-		osb->dirty = 0;
-	}
+	osb->local_alloc_copy = NULL;
+	osb->dirty = 0;
+
+	/* queue to recover orphan slots for all offline slots */
+	ocfs2_replay_map_set_state(osb, REPLAY_NEEDED);
+	ocfs2_queue_replay_slots(osb);
+	ocfs2_free_replay_slots(osb);
 }
 
 void ocfs2_complete_quota_recovery(struct ocfs2_super *osb)
@@ -1236,6 +1350,14 @@ static int __ocfs2_recovery_thread(void *arg)
 		goto bail;
 	}
 
+	status = ocfs2_compute_replay_slots(osb);
+	if (status < 0)
+		mlog_errno(status);
+
+	/* queue recovery for our own slot */
+	ocfs2_queue_recovery_completion(osb->journal, osb->slot_num, NULL,
+					NULL, NULL);
+
 	spin_lock(&osb->osb_lock);
 	while (rm->rm_used) {
 		/* It's always safe to remove entry zero, as we won't
@@ -1301,11 +1423,8 @@ static int __ocfs2_recovery_thread(void *arg)
 
 	ocfs2_super_unlock(osb, 1);
 
-	/* We always run recovery on our own orphan dir - the dead
-	 * node(s) may have disallowd a previos inode delete. Re-processing
-	 * is therefore required. */
-	ocfs2_queue_recovery_completion(osb->journal, osb->slot_num, NULL,
-					NULL, NULL);
+	/* queue recovery for offline slots */
+	ocfs2_queue_replay_slots(osb);
 
 bail:
 	mutex_lock(&osb->recovery_lock);
@@ -1314,6 +1433,7 @@ static int __ocfs2_recovery_thread(void *arg)
 		goto restart;
 	}
 
+	ocfs2_free_replay_slots(osb);
 	osb->recovery_thread_task = NULL;
 	mb(); /* sync with ocfs2_recovery_thread_running */
 	wake_up(&osb->recovery_event);
@@ -1465,6 +1585,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
 		goto done;
 	}
 
+	/* we need to run complete recovery for offline orphan slots */
+	ocfs2_replay_map_set_state(osb, REPLAY_NEEDED);
+
 	mlog(ML_NOTICE, "Recovering node %d from slot %d on device (%u,%u)\n",
 	     node_num, slot_num,
 	     MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev));

fs/ocfs2/journal.h

@@ -38,6 +38,17 @@ enum ocfs2_journal_state {
 struct ocfs2_super;
 struct ocfs2_dinode;
 
+/*
+ * The recovery_list is a simple linked list of node numbers to recover.
+ * It is protected by the recovery_lock.
+ */
+
+struct ocfs2_recovery_map {
+	unsigned int rm_used;
+	unsigned int *rm_entries;
+};
+
+
 struct ocfs2_journal {
 	enum ocfs2_journal_state   j_state;    /* Journals current state   */
 
@@ -139,6 +150,7 @@ void ocfs2_wait_for_recovery(struct ocfs2_super *osb);
 int ocfs2_recovery_init(struct ocfs2_super *osb);
 void ocfs2_recovery_exit(struct ocfs2_super *osb);
 
+int ocfs2_compute_replay_slots(struct ocfs2_super *osb);
 /*
  *  Journal Control:
  *  Initialize, Load, Shutdown, Wipe a journal.
@@ -266,6 +278,12 @@ int ocfs2_journal_access_dq(handle_t *handle, struct inode *inode,
 /* dirblock */
 int ocfs2_journal_access_db(handle_t *handle, struct inode *inode,
 			    struct buffer_head *bh, int type);
+/* ocfs2_dx_root_block */
+int ocfs2_journal_access_dr(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type);
+/* ocfs2_dx_leaf */
+int ocfs2_journal_access_dl(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type);
 /* Anything that has no ecc */
 int ocfs2_journal_access(handle_t *handle, struct inode *inode,
 			 struct buffer_head *bh, int type);
@@ -368,14 +386,29 @@ static inline int ocfs2_remove_extent_credits(struct super_block *sb)
 }
 
 /* data block for new dir/symlink, 2 for bitmap updates (bitmap fe +
- * bitmap block for the new bit) */
-#define OCFS2_DIR_LINK_ADDITIONAL_CREDITS (1 + 2)
+ * bitmap block for the new bit) dx_root update for free list */
+#define OCFS2_DIR_LINK_ADDITIONAL_CREDITS (1 + 2 + 1)
+
+static inline int ocfs2_add_dir_index_credits(struct super_block *sb)
+{
+	/* 1 block for index, 2 allocs (data, metadata), 1 clusters
+	 * worth of blocks for initial extent. */
+	return 1 + 2 * OCFS2_SUBALLOC_ALLOC +
+		ocfs2_clusters_to_blocks(sb, 1);
+}
 
-/* parent fe, parent block, new file entry, inode alloc fe, inode alloc
- * group descriptor + mkdir/symlink blocks + quota update */
-static inline int ocfs2_mknod_credits(struct super_block *sb)
+/* parent fe, parent block, new file entry, index leaf, inode alloc fe, inode
+ * alloc group descriptor + mkdir/symlink blocks + dir blocks + xattr
+ * blocks + quota update */
+static inline int ocfs2_mknod_credits(struct super_block *sb, int is_dir,
+				      int xattr_credits)
 {
-	return 3 + OCFS2_SUBALLOC_ALLOC + OCFS2_DIR_LINK_ADDITIONAL_CREDITS +
+	int dir_credits = OCFS2_DIR_LINK_ADDITIONAL_CREDITS;
+
+	if (is_dir)
+		dir_credits += ocfs2_add_dir_index_credits(sb);
+
+	return 4 + OCFS2_SUBALLOC_ALLOC + dir_credits + xattr_credits +
 	       ocfs2_quota_trans_credits(sb);
 }
 
@@ -388,31 +421,31 @@ static inline int ocfs2_mknod_credits(struct super_block *sb)
 #define OCFS2_SIMPLE_DIR_EXTEND_CREDITS (2)
 
 /* file update (nlink, etc) + directory mtime/ctime + dir entry block + quota
- * update on dir */
+ * update on dir + index leaf + dx root update for free list */
 static inline int ocfs2_link_credits(struct super_block *sb)
 {
-	return 2*OCFS2_INODE_UPDATE_CREDITS + 1 +
+	return 2*OCFS2_INODE_UPDATE_CREDITS + 3 +
 	       ocfs2_quota_trans_credits(sb);
 }
 
 /* inode + dir inode (if we unlink a dir), + dir entry block + orphan
- * dir inode link */
+ * dir inode link + dir inode index leaf + dir index root */
 static inline int ocfs2_unlink_credits(struct super_block *sb)
 {
 	/* The quota update from ocfs2_link_credits is unused here... */
-	return 2 * OCFS2_INODE_UPDATE_CREDITS + 1 + ocfs2_link_credits(sb);
+	return 2 * OCFS2_INODE_UPDATE_CREDITS + 3 + ocfs2_link_credits(sb);
 }
 
 /* dinode + orphan dir dinode + inode alloc dinode + orphan dir entry +
- * inode alloc group descriptor */
-#define OCFS2_DELETE_INODE_CREDITS (3 * OCFS2_INODE_UPDATE_CREDITS + 1 + 1)
+ * inode alloc group descriptor + orphan dir index leaf */
+#define OCFS2_DELETE_INODE_CREDITS (3 * OCFS2_INODE_UPDATE_CREDITS + 3)
 
 /* dinode update, old dir dinode update, new dir dinode update, old
  * dir dir entry, new dir dir entry, dir entry update for renaming
- * directory + target unlink */
+ * directory + target unlink + 3 x dir index leaves */
 static inline int ocfs2_rename_credits(struct super_block *sb)
 {
-	return 3 * OCFS2_INODE_UPDATE_CREDITS + 3 + ocfs2_unlink_credits(sb);
+	return 3 * OCFS2_INODE_UPDATE_CREDITS + 6 + ocfs2_unlink_credits(sb);
 }
 
 /* global bitmap dinode, group desc., relinked group,
@@ -422,6 +455,20 @@ static inline int ocfs2_rename_credits(struct super_block *sb)
 					  + OCFS2_INODE_UPDATE_CREDITS \
 					  + OCFS2_XATTR_BLOCK_UPDATE_CREDITS)
 
+/* inode update, removal of dx root block from allocator */
+#define OCFS2_DX_ROOT_REMOVE_CREDITS (OCFS2_INODE_UPDATE_CREDITS +	\
+				      OCFS2_SUBALLOC_FREE)
+
+static inline int ocfs2_calc_dxi_expand_credits(struct super_block *sb)
+{
+	int credits = 1 + OCFS2_SUBALLOC_ALLOC;
+
+	credits += ocfs2_clusters_to_blocks(sb, 1);
+	credits += ocfs2_quota_trans_credits(sb);
+
+	return credits;
+}
+
 /*
  * Please note that the caller must make sure that root_el is the root
  * of extent tree. So for an inode, it should be &fe->id2.i_list. Otherwise
@@ -457,7 +504,7 @@ static inline int ocfs2_calc_extend_credits(struct super_block *sb,
 
 static inline int ocfs2_calc_symlink_credits(struct super_block *sb)
 {
-	int blocks = ocfs2_mknod_credits(sb);
+	int blocks = ocfs2_mknod_credits(sb, 0, 0);
 
 	/* links can be longer than one block so we may update many
 	 * within our single allocated extent. */

fs/ocfs2/localalloc.c

@@ -28,7 +28,6 @@
 #include <linux/slab.h>
 #include <linux/highmem.h>
 #include <linux/bitops.h>
-#include <linux/debugfs.h>
 
 #define MLOG_MASK_PREFIX ML_DISK_ALLOC
 #include <cluster/masklog.h>
@@ -75,84 +74,6 @@ static int ocfs2_local_alloc_new_window(struct ocfs2_super *osb,
 static int ocfs2_local_alloc_slide_window(struct ocfs2_super *osb,
 					  struct inode *local_alloc_inode);
 
-#ifdef CONFIG_OCFS2_FS_STATS
-
-static int ocfs2_la_debug_open(struct inode *inode, struct file *file)
-{
-	file->private_data = inode->i_private;
-	return 0;
-}
-
-#define LA_DEBUG_BUF_SZ	PAGE_CACHE_SIZE
-#define LA_DEBUG_VER	1
-static ssize_t ocfs2_la_debug_read(struct file *file, char __user *userbuf,
-				   size_t count, loff_t *ppos)
-{
-	static DEFINE_MUTEX(la_debug_mutex);
-	struct ocfs2_super *osb = file->private_data;
-	int written, ret;
-	char *buf = osb->local_alloc_debug_buf;
-
-	mutex_lock(&la_debug_mutex);
-	memset(buf, 0, LA_DEBUG_BUF_SZ);
-
-	written = snprintf(buf, LA_DEBUG_BUF_SZ,
-			   "0x%x\t0x%llx\t%u\t%u\t0x%x\n",
-			   LA_DEBUG_VER,
-			   (unsigned long long)osb->la_last_gd,
-			   osb->local_alloc_default_bits,
-			   osb->local_alloc_bits, osb->local_alloc_state);
-
-	ret = simple_read_from_buffer(userbuf, count, ppos, buf, written);
-
-	mutex_unlock(&la_debug_mutex);
-	return ret;
-}
-
-static const struct file_operations ocfs2_la_debug_fops = {
-	.open =		ocfs2_la_debug_open,
-	.read =		ocfs2_la_debug_read,
-};
-
-static void ocfs2_init_la_debug(struct ocfs2_super *osb)
-{
-	osb->local_alloc_debug_buf = kmalloc(LA_DEBUG_BUF_SZ, GFP_NOFS);
-	if (!osb->local_alloc_debug_buf)
-		return;
-
-	osb->local_alloc_debug = debugfs_create_file("local_alloc_stats",
-						     S_IFREG|S_IRUSR,
-						     osb->osb_debug_root,
-						     osb,
-						     &ocfs2_la_debug_fops);
-	if (!osb->local_alloc_debug) {
-		kfree(osb->local_alloc_debug_buf);
-		osb->local_alloc_debug_buf = NULL;
-	}
-}
-
-static void ocfs2_shutdown_la_debug(struct ocfs2_super *osb)
-{
-	if (osb->local_alloc_debug)
-		debugfs_remove(osb->local_alloc_debug);
-
-	if (osb->local_alloc_debug_buf)
-		kfree(osb->local_alloc_debug_buf);
-
-	osb->local_alloc_debug_buf = NULL;
-	osb->local_alloc_debug = NULL;
-}
-#else	/* CONFIG_OCFS2_FS_STATS */
-static void ocfs2_init_la_debug(struct ocfs2_super *osb)
-{
-	return;
-}
-static void ocfs2_shutdown_la_debug(struct ocfs2_super *osb)
-{
-	return;
-}
-#endif
-
 static inline int ocfs2_la_state_enabled(struct ocfs2_super *osb)
 {
 	return (osb->local_alloc_state == OCFS2_LA_THROTTLED ||
@@ -226,8 +147,6 @@ int ocfs2_load_local_alloc(struct ocfs2_super *osb)
 
 	mlog_entry_void();
 
-	ocfs2_init_la_debug(osb);
-
 	if (osb->local_alloc_bits == 0)
 		goto bail;
 
@@ -299,9 +218,6 @@ int ocfs2_load_local_alloc(struct ocfs2_super *osb)
 	if (inode)
 		iput(inode);
 
-	if (status < 0)
-		ocfs2_shutdown_la_debug(osb);
-
 	mlog(0, "Local alloc window bits = %d\n", osb->local_alloc_bits);
 
 	mlog_exit(status);
@@ -331,8 +247,6 @@ void ocfs2_shutdown_local_alloc(struct ocfs2_super *osb)
 	cancel_delayed_work(&osb->la_enable_wq);
 	flush_workqueue(ocfs2_wq);
 
-	ocfs2_shutdown_la_debug(osb);
-
 	if (osb->local_alloc_state == OCFS2_LA_UNUSED)
 		goto out;
 

fs/ocfs2/ocfs2.h

@@ -209,6 +209,7 @@ enum ocfs2_mount_options
 struct ocfs2_journal;
 struct ocfs2_slot_info;
 struct ocfs2_recovery_map;
+struct ocfs2_replay_map;
 struct ocfs2_quota_recovery;
 struct ocfs2_dentry_lock;
 struct ocfs2_super
@@ -264,6 +265,7 @@ struct ocfs2_super
 	atomic_t vol_state;
 	struct mutex recovery_lock;
 	struct ocfs2_recovery_map *recovery_map;
+	struct ocfs2_replay_map *replay_map;
 	struct task_struct *recovery_thread_task;
 	int disable_recovery;
 	wait_queue_head_t checkpoint_event;
@@ -287,11 +289,6 @@ struct ocfs2_super
 
 	u64 la_last_gd;
 
-#ifdef CONFIG_OCFS2_FS_STATS
-	struct dentry *local_alloc_debug;
-	char *local_alloc_debug_buf;
-#endif
-
 	/* Next three fields are for local node slot recovery during
 	 * mount. */
 	int dirty;
@@ -305,9 +302,11 @@ struct ocfs2_super
 	struct ocfs2_cluster_connection *cconn;
 	struct ocfs2_lock_res osb_super_lockres;
 	struct ocfs2_lock_res osb_rename_lockres;
+	struct ocfs2_lock_res osb_nfs_sync_lockres;
 	struct ocfs2_dlm_debug *osb_dlm_debug;
 
 	struct dentry *osb_debug_root;
+	struct dentry *osb_ctxt;
 
 	wait_queue_head_t recovery_event;
 
@@ -344,6 +343,12 @@ struct ocfs2_super
 
 	/* used to protect metaecc calculation check of xattr. */
 	spinlock_t osb_xattr_lock;
+
+	unsigned int			osb_dx_mask;
+	u32				osb_dx_seed[4];
+
+	/* the group we used to allocate inodes. */
+	u64				osb_inode_alloc_group;
 };
 
 #define OCFS2_SB(sb)	    ((struct ocfs2_super *)(sb)->s_fs_info)
@@ -402,6 +407,51 @@ static inline int ocfs2_meta_ecc(struct ocfs2_super *osb)
 	return 0;
 }
 
+static inline int ocfs2_supports_indexed_dirs(struct ocfs2_super *osb)
+{
+	if (osb->s_feature_incompat & OCFS2_FEATURE_INCOMPAT_INDEXED_DIRS)
+		return 1;
+	return 0;
+}
+
+static inline unsigned int ocfs2_link_max(struct ocfs2_super *osb)
+{
+	if (ocfs2_supports_indexed_dirs(osb))
+		return OCFS2_DX_LINK_MAX;
+	return OCFS2_LINK_MAX;
+}
+
+static inline unsigned int ocfs2_read_links_count(struct ocfs2_dinode *di)
+{
+	u32 nlink = le16_to_cpu(di->i_links_count);
+	u32 hi = le16_to_cpu(di->i_links_count_hi);
+
+	if (di->i_dyn_features & cpu_to_le16(OCFS2_INDEXED_DIR_FL))
+		nlink |= (hi << OCFS2_LINKS_HI_SHIFT);
+
+	return nlink;
+}
+
+static inline void ocfs2_set_links_count(struct ocfs2_dinode *di, u32 nlink)
+{
+	u16 lo, hi;
+
+	lo = nlink;
+	hi = nlink >> OCFS2_LINKS_HI_SHIFT;
+
+	di->i_links_count = cpu_to_le16(lo);
+	di->i_links_count_hi = cpu_to_le16(hi);
+}
+
+static inline void ocfs2_add_links_count(struct ocfs2_dinode *di, int n)
+{
+	u32 links = ocfs2_read_links_count(di);
+
+	links += n;
+
+	ocfs2_set_links_count(di, links);
+}
+
 /* set / clear functions because cluster events can make these happen
  * in parallel so we want the transitions to be atomic. this also
  * means that any future flags osb_flags must be protected by spinlock
@@ -482,6 +532,12 @@ static inline int ocfs2_uses_extended_slot_map(struct ocfs2_super *osb)
 #define OCFS2_IS_VALID_DIR_TRAILER(ptr)					\
 	(!strcmp((ptr)->db_signature, OCFS2_DIR_TRAILER_SIGNATURE))
 
+#define OCFS2_IS_VALID_DX_ROOT(ptr)					\
+	(!strcmp((ptr)->dr_signature, OCFS2_DX_ROOT_SIGNATURE))
+
+#define OCFS2_IS_VALID_DX_LEAF(ptr)					\
+	(!strcmp((ptr)->dl_signature, OCFS2_DX_LEAF_SIGNATURE))
+
 static inline unsigned long ino_from_blkno(struct super_block *sb,
 					   u64 blkno)
 {
@@ -532,6 +588,16 @@ static inline u64 ocfs2_clusters_to_bytes(struct super_block *sb,
 	return (u64)clusters << OCFS2_SB(sb)->s_clustersize_bits;
 }
 
+static inline u64 ocfs2_block_to_cluster_start(struct super_block *sb,
+					       u64 blocks)
+{
+	int bits = OCFS2_SB(sb)->s_clustersize_bits - sb->s_blocksize_bits;
+	unsigned int clusters;
+
+	clusters = ocfs2_blocks_to_clusters(sb, blocks);
+	return (u64)clusters << bits;
+}
+
 static inline u64 ocfs2_align_bytes_to_clusters(struct super_block *sb,
 						u64 bytes)
 {

fs/ocfs2/ocfs2_fs.h

@@ -66,6 +66,8 @@
 #define OCFS2_GROUP_DESC_SIGNATURE      "GROUP01"
 #define OCFS2_XATTR_BLOCK_SIGNATURE	"XATTR01"
 #define OCFS2_DIR_TRAILER_SIGNATURE	"DIRTRL1"
+#define OCFS2_DX_ROOT_SIGNATURE		"DXDIR01"
+#define OCFS2_DX_LEAF_SIGNATURE		"DXLEAF1"
 
 /* Compatibility flags */
 #define OCFS2_HAS_COMPAT_FEATURE(sb,mask)			\
@@ -95,7 +97,8 @@
 					 | OCFS2_FEATURE_INCOMPAT_EXTENDED_SLOT_MAP \
 					 | OCFS2_FEATURE_INCOMPAT_USERSPACE_STACK \
 					 | OCFS2_FEATURE_INCOMPAT_XATTR \
-					 | OCFS2_FEATURE_INCOMPAT_META_ECC)
+					 | OCFS2_FEATURE_INCOMPAT_META_ECC \
+					 | OCFS2_FEATURE_INCOMPAT_INDEXED_DIRS)
 #define OCFS2_FEATURE_RO_COMPAT_SUPP	(OCFS2_FEATURE_RO_COMPAT_UNWRITTEN \
 					 | OCFS2_FEATURE_RO_COMPAT_USRQUOTA \
 					 | OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)
@@ -151,6 +154,9 @@
 /* Support for extended attributes */
 #define OCFS2_FEATURE_INCOMPAT_XATTR		0x0200
 
+/* Support for indexed directores */
+#define OCFS2_FEATURE_INCOMPAT_INDEXED_DIRS	0x0400
+
 /* Metadata checksum and error correction */
 #define OCFS2_FEATURE_INCOMPAT_META_ECC		0x0800
 
@@ -411,8 +417,12 @@ static struct ocfs2_system_inode_info ocfs2_system_inodes[NUM_SYSTEM_INODES] = {
 #define OCFS2_DIR_REC_LEN(name_len)	(((name_len) + OCFS2_DIR_MEMBER_LEN + \
                                           OCFS2_DIR_ROUND) & \
 					 ~OCFS2_DIR_ROUND)
+#define OCFS2_DIR_MIN_REC_LEN	OCFS2_DIR_REC_LEN(1)
 
 #define OCFS2_LINK_MAX		32000
+#define	OCFS2_DX_LINK_MAX	((1U << 31) - 1U)
+#define	OCFS2_LINKS_HI_SHIFT	16
+#define	OCFS2_DX_ENTRIES_MAX	(0xffffffffU)
 
 #define S_SHIFT			12
 static unsigned char ocfs2_type_by_mode[S_IFMT >> S_SHIFT] = {
@@ -628,8 +638,9 @@ struct ocfs2_super_block {
 /*B8*/	__le16 s_xattr_inline_size;	/* extended attribute inline size
 					   for this fs*/
 	__le16 s_reserved0;
-	__le32 s_reserved1;
-/*C0*/  __le64 s_reserved2[16];		/* Fill out superblock */
+	__le32 s_dx_seed[3];		/* seed[0-2] for dx dir hash.
+					 * s_uuid_hash serves as seed[3]. */
+/*C0*/  __le64 s_reserved2[15];		/* Fill out superblock */
 /*140*/
 
 	/*
@@ -679,7 +690,7 @@ struct ocfs2_dinode {
 					   belongs to */
 	__le16 i_suballoc_bit;		/* Bit offset in suballocator
 					   block group */
-/*10*/	__le16 i_reserved0;
+/*10*/	__le16 i_links_count_hi;	/* High 16 bits of links count */
 	__le16 i_xattr_inline_size;
 	__le32 i_clusters;		/* Cluster count */
 	__le32 i_uid;			/* Owner UID */
@@ -705,7 +716,8 @@ struct ocfs2_dinode {
 	__le16 i_dyn_features;
 	__le64 i_xattr_loc;
 /*80*/	struct ocfs2_block_check i_check;	/* Error checking */
-/*88*/	__le64 i_reserved2[6];
+/*88*/	__le64 i_dx_root;		/* Pointer to dir index root block */
+	__le64 i_reserved2[5];
 /*B8*/	union {
 		__le64 i_pad1;		/* Generic way to refer to this
 					   64bit union */
@@ -781,6 +793,90 @@ struct ocfs2_dir_block_trailer {
 /*40*/
 };
 
+ /*
+ * A directory entry in the indexed tree. We don't store the full name here,
+ * but instead provide a pointer to the full dirent in the unindexed tree.
+ *
+ * We also store name_len here so as to reduce the number of leaf blocks we
+ * need to search in case of collisions.
+ */
+struct ocfs2_dx_entry {
+	__le32		dx_major_hash;	/* Used to find logical
+					 * cluster in index */
+	__le32		dx_minor_hash;	/* Lower bits used to find
+					 * block in cluster */
+	__le64		dx_dirent_blk;	/* Physical block in unindexed
+					 * tree holding this dirent. */
+};
+
+struct ocfs2_dx_entry_list {
+	__le32		de_reserved;
+	__le16		de_count;	/* Maximum number of entries
+					 * possible in de_entries */
+	__le16		de_num_used;	/* Current number of
+					 * de_entries entries */
+	struct	ocfs2_dx_entry		de_entries[0];	/* Indexed dir entries
+							 * in a packed array of
+							 * length de_num_used */
+};
+
+#define OCFS2_DX_FLAG_INLINE	0x01
+
+/*
+ * A directory indexing block. Each indexed directory has one of these,
+ * pointed to by ocfs2_dinode.
+ *
+ * This block stores an indexed btree root, and a set of free space
+ * start-of-list pointers.
+ */
+struct ocfs2_dx_root_block {
+	__u8		dr_signature[8];	/* Signature for verification */
+	struct ocfs2_block_check dr_check;	/* Error checking */
+	__le16		dr_suballoc_slot;	/* Slot suballocator this
+						 * block belongs to. */
+	__le16		dr_suballoc_bit;	/* Bit offset in suballocator
+						 * block group */
+	__le32		dr_fs_generation;	/* Must match super block */
+	__le64		dr_blkno;		/* Offset on disk, in blocks */
+	__le64		dr_last_eb_blk;		/* Pointer to last
+						 * extent block */
+	__le32		dr_clusters;		/* Clusters allocated
+						 * to the indexed tree. */
+	__u8		dr_flags;		/* OCFS2_DX_FLAG_* flags */
+	__u8		dr_reserved0;
+	__le16		dr_reserved1;
+	__le64		dr_dir_blkno;		/* Pointer to parent inode */
+	__le32		dr_num_entries;		/* Total number of
+						 * names stored in
+						 * this directory.*/
+	__le32		dr_reserved2;
+	__le64		dr_free_blk;		/* Pointer to head of free
+						 * unindexed block list. */
+	__le64		dr_reserved3[15];
+	union {
+		struct ocfs2_extent_list dr_list; /* Keep this aligned to 128
+						   * bits for maximum space
+						   * efficiency. */
+		struct ocfs2_dx_entry_list dr_entries; /* In-root-block list of
+							* entries. We grow out
+							* to extents if this
+							* gets too big. */
+	};
+};
+
+/*
+ * The header of a leaf block in the indexed tree.
+ */
+struct ocfs2_dx_leaf {
+	__u8		dl_signature[8];/* Signature for verification */
+	struct ocfs2_block_check dl_check;	/* Error checking */
+	__le64		dl_blkno;	/* Offset on disk, in blocks */
+	__le32		dl_fs_generation;/* Must match super block */
+	__le32		dl_reserved0;
+	__le64		dl_reserved1;
+	struct ocfs2_dx_entry_list	dl_list;
+};
+
 /*
  * On disk allocator group structure for OCFS2
  */
@@ -1112,6 +1208,16 @@ static inline int ocfs2_extent_recs_per_inode_with_xattr(
 	return size / sizeof(struct ocfs2_extent_rec);
 }
 
+static inline int ocfs2_extent_recs_per_dx_root(struct super_block *sb)
+{
+	int size;
+
+	size = sb->s_blocksize -
+		offsetof(struct ocfs2_dx_root_block, dr_list.l_recs);
+
+	return size / sizeof(struct ocfs2_extent_rec);
+}
+
 static inline int ocfs2_chain_recs_per_inode(struct super_block *sb)
 {
 	int size;
@@ -1132,6 +1238,26 @@ static inline u16 ocfs2_extent_recs_per_eb(struct super_block *sb)
 	return size / sizeof(struct ocfs2_extent_rec);
 }
 
+static inline int ocfs2_dx_entries_per_leaf(struct super_block *sb)
+{
+	int size;
+
+	size = sb->s_blocksize -
+		offsetof(struct ocfs2_dx_leaf, dl_list.de_entries);
+
+	return size / sizeof(struct ocfs2_dx_entry);
+}
+
+static inline int ocfs2_dx_entries_per_root(struct super_block *sb)
+{
+	int size;
+
+	size = sb->s_blocksize -
+		offsetof(struct ocfs2_dx_root_block, dr_entries.de_entries);
+
+	return size / sizeof(struct ocfs2_dx_entry);
+}
+
 static inline u16 ocfs2_local_alloc_size(struct super_block *sb)
 {
 	u16 size;

fs/ocfs2/ocfs2_lockid.h

@@ -47,6 +47,7 @@ enum ocfs2_lock_type {
 	OCFS2_LOCK_TYPE_OPEN,
 	OCFS2_LOCK_TYPE_FLOCK,
 	OCFS2_LOCK_TYPE_QINFO,
+	OCFS2_LOCK_TYPE_NFS_SYNC,
 	OCFS2_NUM_LOCK_TYPES
 };
 
@@ -81,6 +82,9 @@ static inline char ocfs2_lock_type_char(enum ocfs2_lock_type type)
 		case OCFS2_LOCK_TYPE_QINFO:
 			c = 'Q';
 			break;
+		case OCFS2_LOCK_TYPE_NFS_SYNC:
+			c = 'Y';
+			break;
 		default:
 			c = '\0';
 	}

fs/ocfs2/suballoc.h

@@ -88,6 +88,8 @@ int ocfs2_claim_metadata(struct ocfs2_super *osb,
 			 u64 *blkno_start);
 int ocfs2_claim_new_inode(struct ocfs2_super *osb,
 			  handle_t *handle,
+			  struct inode *dir,
+			  struct buffer_head *parent_fe_bh,
 			  struct ocfs2_alloc_context *ac,
 			  u16 *suballoc_bit,
 			  u64 *fe_blkno);
@@ -186,4 +188,6 @@ int ocfs2_lock_allocators(struct inode *inode, struct ocfs2_extent_tree *et,
 			  u32 clusters_to_add, u32 extents_to_split,
 			  struct ocfs2_alloc_context **data_ac,
 			  struct ocfs2_alloc_context **meta_ac);
+
+int ocfs2_test_inode_bit(struct ocfs2_super *osb, u64 blkno, int *res);
 #endif /* _CHAINALLOC_H_ */

fs/ocfs2/super.c

@@ -201,6 +201,170 @@ static const match_table_t tokens = {
 	{Opt_err, NULL}
 };
 
+#ifdef CONFIG_DEBUG_FS
+static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len)
+{
+	int out = 0;
+	int i;
+	struct ocfs2_cluster_connection *cconn = osb->cconn;
+	struct ocfs2_recovery_map *rm = osb->recovery_map;
+
+	out += snprintf(buf + out, len - out,
+			"%10s => Id: %-s  Uuid: %-s  Gen: 0x%X  Label: %-s\n",
+			"Device", osb->dev_str, osb->uuid_str,
+			osb->fs_generation, osb->vol_label);
+
+	out += snprintf(buf + out, len - out,
+			"%10s => State: %d  Flags: 0x%lX\n", "Volume",
+			atomic_read(&osb->vol_state), osb->osb_flags);
+
+	out += snprintf(buf + out, len - out,
+			"%10s => Block: %lu  Cluster: %d\n", "Sizes",
+			osb->sb->s_blocksize, osb->s_clustersize);
+
+	out += snprintf(buf + out, len - out,
+			"%10s => Compat: 0x%X  Incompat: 0x%X  "
+			"ROcompat: 0x%X\n",
+			"Features", osb->s_feature_compat,
+			osb->s_feature_incompat, osb->s_feature_ro_compat);
+
+	out += snprintf(buf + out, len - out,
+			"%10s => Opts: 0x%lX  AtimeQuanta: %u\n", "Mount",
+			osb->s_mount_opt, osb->s_atime_quantum);
+
+	out += snprintf(buf + out, len - out,
+			"%10s => Stack: %s  Name: %*s  Version: %d.%d\n",
+			"Cluster",
+			(*osb->osb_cluster_stack == '\0' ?
+			 "o2cb" : osb->osb_cluster_stack),
+			cconn->cc_namelen, cconn->cc_name,
+			cconn->cc_version.pv_major, cconn->cc_version.pv_minor);
+
+	spin_lock(&osb->dc_task_lock);
+	out += snprintf(buf + out, len - out,
+			"%10s => Pid: %d  Count: %lu  WakeSeq: %lu  "
+			"WorkSeq: %lu\n", "DownCnvt",
+			task_pid_nr(osb->dc_task), osb->blocked_lock_count,
+			osb->dc_wake_sequence, osb->dc_work_sequence);
+	spin_unlock(&osb->dc_task_lock);
+
+	spin_lock(&osb->osb_lock);
+	out += snprintf(buf + out, len - out, "%10s => Pid: %d  Nodes:",
+			"Recovery",
+			(osb->recovery_thread_task ?
+			 task_pid_nr(osb->recovery_thread_task) : -1));
+	if (rm->rm_used == 0)
+		out += snprintf(buf + out, len - out, " None\n");
+	else {
+		for (i = 0; i < rm->rm_used; i++)
+			out += snprintf(buf + out, len - out, " %d",
+					rm->rm_entries[i]);
+		out += snprintf(buf + out, len - out, "\n");
+	}
+	spin_unlock(&osb->osb_lock);
+
+	out += snprintf(buf + out, len - out,
+			"%10s => Pid: %d  Interval: %lu  Needs: %d\n", "Commit",
+			task_pid_nr(osb->commit_task), osb->osb_commit_interval,
+			atomic_read(&osb->needs_checkpoint));
+
+	out += snprintf(buf + out, len - out,
+			"%10s => State: %d  NumTxns: %d  TxnId: %lu\n",
+			"Journal", osb->journal->j_state,
+			atomic_read(&osb->journal->j_num_trans),
+			osb->journal->j_trans_id);
+
+	out += snprintf(buf + out, len - out,
+			"%10s => GlobalAllocs: %d  LocalAllocs: %d  "
+			"SubAllocs: %d  LAWinMoves: %d  SAExtends: %d\n",
+			"Stats",
+			atomic_read(&osb->alloc_stats.bitmap_data),
+			atomic_read(&osb->alloc_stats.local_data),
+			atomic_read(&osb->alloc_stats.bg_allocs),
+			atomic_read(&osb->alloc_stats.moves),
+			atomic_read(&osb->alloc_stats.bg_extends));
+
+	out += snprintf(buf + out, len - out,
+			"%10s => State: %u  Descriptor: %llu  Size: %u bits  "
+			"Default: %u bits\n",
+			"LocalAlloc", osb->local_alloc_state,
+			(unsigned long long)osb->la_last_gd,
+			osb->local_alloc_bits, osb->local_alloc_default_bits);
+
+	spin_lock(&osb->osb_lock);
+	out += snprintf(buf + out, len - out,
+			"%10s => Slot: %d  NumStolen: %d\n", "Steal",
+			osb->s_inode_steal_slot,
+			atomic_read(&osb->s_num_inodes_stolen));
+	spin_unlock(&osb->osb_lock);
+
+	out += snprintf(buf + out, len - out, "%10s => %3s  %10s\n",
+			"Slots", "Num", "RecoGen");
+
+	for (i = 0; i < osb->max_slots; ++i) {
+		out += snprintf(buf + out, len - out,
+				"%10s  %c %3d  %10d\n",
+				" ",
+				(i == osb->slot_num ? '*' : ' '),
+				i, osb->slot_recovery_generations[i]);
+	}
+
+	return out;
+}
+
+static int ocfs2_osb_debug_open(struct inode *inode, struct file *file)
+{
+	struct ocfs2_super *osb = inode->i_private;
+	char *buf = NULL;
+
+	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!buf)
+		goto bail;
+
+	i_size_write(inode, ocfs2_osb_dump(osb, buf, PAGE_SIZE));
+
+	file->private_data = buf;
+
+	return 0;
+bail:
+	return -ENOMEM;
+}
+
+static int ocfs2_debug_release(struct inode *inode, struct file *file)
+{
+	kfree(file->private_data);
+	return 0;
+}
+
+static ssize_t ocfs2_debug_read(struct file *file, char __user *buf,
+				size_t nbytes, loff_t *ppos)
+{
+	return simple_read_from_buffer(buf, nbytes, ppos, file->private_data,
+				       i_size_read(file->f_mapping->host));
+}
+#else
+static int ocfs2_osb_debug_open(struct inode *inode, struct file *file)
+{
+	return 0;
+}
+static int ocfs2_debug_release(struct inode *inode, struct file *file)
+{
+	return 0;
+}
+static ssize_t ocfs2_debug_read(struct file *file, char __user *buf,
+				size_t nbytes, loff_t *ppos)
+{
+	return 0;
+}
+#endif	/* CONFIG_DEBUG_FS */
+
+static struct file_operations ocfs2_osb_debug_fops = {
+	.open =		ocfs2_osb_debug_open,
+	.release =	ocfs2_debug_release,
+	.read =		ocfs2_debug_read,
+	.llseek =	generic_file_llseek,
+};
+
 /*
  * write_super and sync_fs ripped right out of ext3.
  */
@@ -926,6 +1090,16 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
 		goto read_super_error;
 	}
 
+	osb->osb_ctxt = debugfs_create_file("fs_state", S_IFREG|S_IRUSR,
+					    osb->osb_debug_root,
+					    osb,
+					    &ocfs2_osb_debug_fops);
+	if (!osb->osb_ctxt) {
+		status = -EINVAL;
+		mlog_errno(status);
+		goto read_super_error;
+	}
+
 	status = ocfs2_mount_volume(sb);
 	if (osb->root_inode)
 		inode = igrab(osb->root_inode);
@@ -1620,6 +1794,8 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err)
 	osb = OCFS2_SB(sb);
 	BUG_ON(!osb);
 
+	debugfs_remove(osb->osb_ctxt);
+
 	ocfs2_disable_quotas(osb);
 
 	ocfs2_shutdown_local_alloc(osb);
@@ -1742,6 +1918,12 @@ static int ocfs2_initialize_super(struct super_block *sb,
 	bbits = le32_to_cpu(di->id2.i_super.s_blocksize_bits);
 	sb->s_maxbytes = ocfs2_max_file_offset(bbits, cbits);
 
+	osb->osb_dx_mask = (1 << (cbits - bbits)) - 1;
+
+	for (i = 0; i < 3; i++)
+		osb->osb_dx_seed[i] = le32_to_cpu(di->id2.i_super.s_dx_seed[i]);
+	osb->osb_dx_seed[3] = le32_to_cpu(di->id2.i_super.s_uuid_hash);
+
 	osb->sb = sb;
 	/* Save off for ocfs2_rw_direct */
 	osb->s_sectsize_bits = blksize_bits(sector_size);
@@ -2130,6 +2312,12 @@ static int ocfs2_check_volume(struct ocfs2_super *osb)
 	 * lock, and it's marked as dirty, set the bit in the recover
 	 * map and launch a recovery thread for it. */
 	status = ocfs2_mark_dead_nodes(osb);
+	if (status < 0) {
+		mlog_errno(status);
+		goto finally;
+	}
+
+	status = ocfs2_compute_replay_slots(osb);
 	if (status < 0)
 		mlog_errno(status);
 

fs/ocfs2/xattr.c

@@ -512,7 +512,7 @@ int ocfs2_calc_xattr_init(struct inode *dir,
 			  struct ocfs2_security_xattr_info *si,
 			  int *want_clusters,
 			  int *xattr_credits,
-			  struct ocfs2_alloc_context **xattr_ac)
+			  int *want_meta)
 {
 	int ret = 0;
 	struct ocfs2_super *osb = OCFS2_SB(dir->i_sb);
@@ -554,11 +554,7 @@ int ocfs2_calc_xattr_init(struct inode *dir,
 	if (dir->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE ||
 	    (S_ISDIR(mode) && ocfs2_supports_inline_data(osb)) ||
 	    (s_size + a_size) > OCFS2_XATTR_FREE_IN_IBODY) {
-		ret = ocfs2_reserve_new_metadata_blocks(osb, 1, xattr_ac);
-		if (ret) {
-			mlog_errno(ret);
-			return ret;
-		}
+		*want_meta = *want_meta + 1;
 		*xattr_credits += OCFS2_XATTR_BLOCK_CREATE_CREDITS;
 	}
 

fs/ocfs2/xattr.h

@@ -68,7 +68,7 @@ int ocfs2_calc_security_init(struct inode *,
 			     int *, int *, struct ocfs2_alloc_context **);
 int ocfs2_calc_xattr_init(struct inode *, struct buffer_head *,
 			  int, struct ocfs2_security_xattr_info *,
-			  int *, int *, struct ocfs2_alloc_context **);
+			  int *, int *, int *);
 
 /*
  * xattrs can live inside an inode, as part of an external xattr block,