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: (47 commits)
  ocfs2: Silence a gcc warning.
  ocfs2: Don't retry xattr set in case value extension fails.
  ocfs2:dlm: avoid dlm->ast_lock lockres->spinlock dependency break
  ocfs2: Reset xattr value size after xa_cleanup_value_truncate().
  fs/ocfs2/dlm: Use kstrdup
  fs/ocfs2/dlm: Drop memory allocation cast
  Ocfs2: Optimize punching-hole code.
  Ocfs2: Make ocfs2_find_cpos_for_left_leaf() public.
  Ocfs2: Fix hole punching to correctly do CoW during cluster zeroing.
  Ocfs2: Optimize ocfs2 truncate to use ocfs2_remove_btree_range() instead.
  ocfs2: Block signals for mkdir/link/symlink/O_CREAT.
  ocfs2: Wrap signal blocking in void functions.
  ocfs2/dlm: Increase o2dlm lockres hash size
  ocfs2: Make ocfs2_extend_trans() really extend.
  ocfs2/trivial: Code cleanup for allocation reservation.
  ocfs2: make ocfs2_adjust_resv_from_alloc simple.
  ocfs2: Make nointr a default mount option
  ocfs2/dlm: Make o2dlm domain join/leave messages KERN_NOTICE
  o2net: log socket state changes
  ocfs2: print node # when tcp fails
  ...

Documentation/filesystems/ocfs2.txt

@@ -80,3 +80,10 @@ user_xattr	(*)	Enables Extended User Attributes.
 nouser_xattr		Disables Extended User Attributes.
 acl			Enables POSIX Access Control Lists support.
 noacl		(*)	Disables POSIX Access Control Lists support.
+resv_level=2	(*)	Set how agressive allocation reservations will be.
+			Valid values are between 0 (reservations off) to 8
+			(maximum space for reservations).
+dir_resv_level=	(*)	By default, directory reservations will scale with file
+			reservations - users should rarely need to change this
+			value. If allocation reservations are turned off, this
+			option will have no effect.

fs/ocfs2/Makefile

@@ -29,6 +29,7 @@ ocfs2-objs := \
 	mmap.o 			\
 	namei.o 		\
 	refcounttree.o		\
+	reservations.o		\
 	resize.o		\
 	slot_map.o 		\
 	suballoc.o 		\

fs/ocfs2/alloc.h

@@ -140,8 +140,9 @@ int ocfs2_remove_extent(handle_t *handle, struct ocfs2_extent_tree *et,
 			struct ocfs2_cached_dealloc_ctxt *dealloc);
 int ocfs2_remove_btree_range(struct inode *inode,
 			     struct ocfs2_extent_tree *et,
-			     u32 cpos, u32 phys_cpos, u32 len,
-			     struct ocfs2_cached_dealloc_ctxt *dealloc);
+			     u32 cpos, u32 phys_cpos, u32 len, int flags,
+			     struct ocfs2_cached_dealloc_ctxt *dealloc,
+			     u64 refcount_loc);
 
 int ocfs2_num_free_extents(struct ocfs2_super *osb,
 			   struct ocfs2_extent_tree *et);
@@ -209,7 +210,7 @@ static inline void ocfs2_init_dealloc_ctxt(struct ocfs2_cached_dealloc_ctxt *c)
 int ocfs2_cache_cluster_dealloc(struct ocfs2_cached_dealloc_ctxt *ctxt,
 				u64 blkno, unsigned int bit);
 int ocfs2_cache_block_dealloc(struct ocfs2_cached_dealloc_ctxt *ctxt,
-			      int type, int slot, u64 blkno,
+			      int type, int slot, u64 suballoc, u64 blkno,
 			      unsigned int bit);
 static inline int ocfs2_dealloc_has_cluster(struct ocfs2_cached_dealloc_ctxt *c)
 {
@@ -233,8 +234,7 @@ int ocfs2_prepare_truncate(struct ocfs2_super *osb,
 			   struct ocfs2_truncate_context **tc);
 int ocfs2_commit_truncate(struct ocfs2_super *osb,
 			  struct inode *inode,
-			  struct buffer_head *fe_bh,
-			  struct ocfs2_truncate_context *tc);
+			  struct buffer_head *di_bh);
 int ocfs2_truncate_inline(struct inode *inode, struct buffer_head *di_bh,
 			  unsigned int start, unsigned int end, int trunc);
 
@@ -319,6 +319,8 @@ int ocfs2_journal_access_path(struct ocfs2_caching_info *ci,
 			      struct ocfs2_path *path);
 int ocfs2_find_cpos_for_right_leaf(struct super_block *sb,
 				   struct ocfs2_path *path, u32 *cpos);
+int ocfs2_find_cpos_for_left_leaf(struct super_block *sb,
+				  struct ocfs2_path *path, u32 *cpos);
 int ocfs2_find_subtree_root(struct ocfs2_extent_tree *et,
 			    struct ocfs2_path *left,
 			    struct ocfs2_path *right);

fs/ocfs2/aops.c

@@ -1735,6 +1735,9 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 			goto out;
 		}
 
+		if (data_ac)
+			data_ac->ac_resv = &OCFS2_I(inode)->ip_la_data_resv;
+
 		credits = ocfs2_calc_extend_credits(inode->i_sb,
 						    &di->id2.i_list,
 						    clusters_to_alloc);

fs/ocfs2/cluster/masklog.c

@@ -116,6 +116,7 @@ static struct mlog_attribute mlog_attrs[MLOG_MAX_BITS] = {
 	define_mask(ERROR),
 	define_mask(NOTICE),
 	define_mask(KTHREAD),
+	define_mask(RESERVATIONS),
 };
 
 static struct attribute *mlog_attr_ptrs[MLOG_MAX_BITS] = {NULL, };

fs/ocfs2/cluster/masklog.h

@@ -119,6 +119,7 @@
 #define ML_ERROR	0x0000000100000000ULL /* sent to KERN_ERR */
 #define ML_NOTICE	0x0000000200000000ULL /* setn to KERN_NOTICE */
 #define ML_KTHREAD	0x0000000400000000ULL /* kernel thread activity */
+#define	ML_RESERVATIONS	0x0000000800000000ULL /* ocfs2 alloc reservations */
 
 #define MLOG_INITIAL_AND_MASK (ML_ERROR|ML_NOTICE)
 #define MLOG_INITIAL_NOT_MASK (ML_ENTRY|ML_EXIT)

fs/ocfs2/cluster/tcp.c

@@ -583,6 +583,9 @@ static void o2net_state_change(struct sock *sk)
 			o2net_sc_queue_work(sc, &sc->sc_connect_work);
 			break;
 		default:
+			printk(KERN_INFO "o2net: connection to " SC_NODEF_FMT
+			      " shutdown, state %d\n",
+			      SC_NODEF_ARGS(sc), sk->sk_state);
 			o2net_sc_queue_work(sc, &sc->sc_shutdown_work);
 			break;
 	}

fs/ocfs2/dir.c

@@ -1194,7 +1194,7 @@ static int __ocfs2_delete_entry(handle_t *handle, struct inode *dir,
 			else
 				de->inode = 0;
 			dir->i_version++;
-			status = ocfs2_journal_dirty(handle, bh);
+			ocfs2_journal_dirty(handle, bh);
 			goto bail;
 		}
 		i += le16_to_cpu(de->rec_len);
@@ -1752,7 +1752,7 @@ int __ocfs2_add_entry(handle_t *handle,
 				ocfs2_recalc_free_list(dir, handle, lookup);
 
 			dir->i_version++;
-			status = ocfs2_journal_dirty(handle, insert_bh);
+			ocfs2_journal_dirty(handle, insert_bh);
 			retval = 0;
 			goto bail;
 		}
@@ -2297,12 +2297,7 @@ static int ocfs2_fill_new_dir_id(struct ocfs2_super *osb,
 	}
 
 	ocfs2_fill_initial_dirents(inode, parent, data->id_data, size);
-
 	ocfs2_journal_dirty(handle, di_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
 
 	i_size_write(inode, size);
 	inode->i_nlink = 2;
@@ -2366,11 +2361,7 @@ static int ocfs2_fill_new_dir_el(struct ocfs2_super *osb,
 		ocfs2_init_dir_trailer(inode, new_bh, size);
 	}
 
-	status = ocfs2_journal_dirty(handle, new_bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto bail;
-	}
+	ocfs2_journal_dirty(handle, new_bh);
 
 	i_size_write(inode, inode->i_sb->s_blocksize);
 	inode->i_nlink = 2;
@@ -2404,15 +2395,15 @@ static int ocfs2_dx_dir_attach_index(struct ocfs2_super *osb,
 	int ret;
 	struct ocfs2_dinode *di = (struct ocfs2_dinode *) di_bh->b_data;
 	u16 dr_suballoc_bit;
-	u64 dr_blkno;
+	u64 suballoc_loc, dr_blkno;
 	unsigned int num_bits;
 	struct buffer_head *dx_root_bh = NULL;
 	struct ocfs2_dx_root_block *dx_root;
 	struct ocfs2_dir_block_trailer *trailer =
 		ocfs2_trailer_from_bh(dirdata_bh, dir->i_sb);
 
-	ret = ocfs2_claim_metadata(osb, handle, meta_ac, 1, &dr_suballoc_bit,
-				   &num_bits, &dr_blkno);
+	ret = ocfs2_claim_metadata(handle, meta_ac, 1, &suballoc_loc,
+				   &dr_suballoc_bit, &num_bits, &dr_blkno);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -2440,6 +2431,7 @@ static int ocfs2_dx_dir_attach_index(struct ocfs2_super *osb,
 	memset(dx_root, 0, osb->sb->s_blocksize);
 	strcpy(dx_root->dr_signature, OCFS2_DX_ROOT_SIGNATURE);
 	dx_root->dr_suballoc_slot = cpu_to_le16(meta_ac->ac_alloc_slot);
+	dx_root->dr_suballoc_loc = cpu_to_le64(suballoc_loc);
 	dx_root->dr_suballoc_bit = cpu_to_le16(dr_suballoc_bit);
 	dx_root->dr_fs_generation = cpu_to_le32(osb->fs_generation);
 	dx_root->dr_blkno = cpu_to_le64(dr_blkno);
@@ -2458,10 +2450,7 @@ static int ocfs2_dx_dir_attach_index(struct ocfs2_super *osb,
 		dx_root->dr_list.l_count =
 			cpu_to_le16(ocfs2_extent_recs_per_dx_root(osb->sb));
 	}
-
-	ret = ocfs2_journal_dirty(handle, dx_root_bh);
-	if (ret)
-		mlog_errno(ret);
+	ocfs2_journal_dirty(handle, dx_root_bh);
 
 	ret = ocfs2_journal_access_di(handle, INODE_CACHE(dir), di_bh,
 				      OCFS2_JOURNAL_ACCESS_CREATE);
@@ -2475,9 +2464,7 @@ static int ocfs2_dx_dir_attach_index(struct ocfs2_super *osb,
 	OCFS2_I(dir)->ip_dyn_features |= OCFS2_INDEXED_DIR_FL;
 	di->i_dyn_features = cpu_to_le16(OCFS2_I(dir)->ip_dyn_features);
 
-	ret = ocfs2_journal_dirty(handle, di_bh);
-	if (ret)
-		mlog_errno(ret);
+	ocfs2_journal_dirty(handle, di_bh);
 
 	*ret_dx_root_bh = dx_root_bh;
 	dx_root_bh = NULL;
@@ -2558,7 +2545,7 @@ static int __ocfs2_dx_dir_new_cluster(struct inode *dir,
 	 * chance of contiguousness as the directory grows in number
 	 * of entries.
 	 */
-	ret = __ocfs2_claim_clusters(osb, handle, data_ac, 1, 1, &phys, &num);
+	ret = __ocfs2_claim_clusters(handle, data_ac, 1, 1, &phys, &num);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -2991,7 +2978,9 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 	 * if we only get one now, that's enough to continue. The rest
 	 * will be claimed after the conversion to extents.
 	 */
-	ret = ocfs2_claim_clusters(osb, handle, data_ac, 1, &bit_off, &len);
+	if (ocfs2_dir_resv_allowed(osb))
+		data_ac->ac_resv = &oi->ip_la_data_resv;
+	ret = ocfs2_claim_clusters(handle, data_ac, 1, &bit_off, &len);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_commit;
@@ -3034,11 +3023,7 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 		ocfs2_init_dir_trailer(dir, dirdata_bh, i);
 	}
 
-	ret = ocfs2_journal_dirty(handle, dirdata_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
+	ocfs2_journal_dirty(handle, dirdata_bh);
 
 	if (ocfs2_supports_indexed_dirs(osb) && !dx_inline) {
 		/*
@@ -3104,11 +3089,7 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 	 */
 	dir->i_blocks = ocfs2_inode_sector_count(dir);
 
-	ret = ocfs2_journal_dirty(handle, di_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
+	ocfs2_journal_dirty(handle, di_bh);
 
 	if (ocfs2_supports_indexed_dirs(osb)) {
 		ret = ocfs2_dx_dir_attach_index(osb, handle, dir, di_bh,
@@ -3138,7 +3119,7 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 	 * pass. Claim the 2nd cluster as a separate extent.
 	 */
 	if (alloc > len) {
-		ret = ocfs2_claim_clusters(osb, handle, data_ac, 1, &bit_off,
+		ret = ocfs2_claim_clusters(handle, data_ac, 1, &bit_off,
 					   &len);
 		if (ret) {
 			mlog_errno(ret);
@@ -3369,6 +3350,9 @@ static int ocfs2_extend_dir(struct ocfs2_super *osb,
 			goto bail;
 		}
 
+		if (ocfs2_dir_resv_allowed(osb))
+			data_ac->ac_resv = &OCFS2_I(dir)->ip_la_data_resv;
+
 		credits = ocfs2_calc_extend_credits(sb, el, 1);
 	} else {
 		spin_unlock(&OCFS2_I(dir)->ip_lock);
@@ -3423,11 +3407,7 @@ static int ocfs2_extend_dir(struct ocfs2_super *osb,
 	} else {
 		de->rec_len = cpu_to_le16(sb->s_blocksize);
 	}
-	status = ocfs2_journal_dirty(handle, new_bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto bail;
-	}
+	ocfs2_journal_dirty(handle, new_bh);
 
 	dir_i_size += dir->i_sb->s_blocksize;
 	i_size_write(dir, dir_i_size);
@@ -3906,11 +3886,7 @@ static int ocfs2_dx_dir_rebalance(struct ocfs2_super *osb, struct inode *dir,
 	     sizeof(struct ocfs2_dx_entry), dx_leaf_sort_cmp,
 	     dx_leaf_sort_swap);
 
-	ret = ocfs2_journal_dirty(handle, dx_leaf_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
+	ocfs2_journal_dirty(handle, dx_leaf_bh);
 
 	ret = ocfs2_dx_dir_find_leaf_split(dx_leaf, leaf_cpos, insert_hash,
 					   &split_hash);
@@ -4490,7 +4466,10 @@ static int ocfs2_dx_dir_remove_index(struct inode *dir,
 
 	blk = le64_to_cpu(dx_root->dr_blkno);
 	bit = le16_to_cpu(dx_root->dr_suballoc_bit);
-	bg_blkno = ocfs2_which_suballoc_group(blk, bit);
+	if (dx_root->dr_suballoc_loc)
+		bg_blkno = le64_to_cpu(dx_root->dr_suballoc_loc);
+	else
+		bg_blkno = ocfs2_which_suballoc_group(blk, bit);
 	ret = ocfs2_free_suballoc_bits(handle, dx_alloc_inode, dx_alloc_bh,
 				       bit, bg_blkno, 1);
 	if (ret)
@@ -4551,8 +4530,8 @@ int ocfs2_dx_dir_truncate(struct inode *dir, struct buffer_head *di_bh)
 
 		p_cpos = ocfs2_blocks_to_clusters(dir->i_sb, blkno);
 
-		ret = ocfs2_remove_btree_range(dir, &et, cpos, p_cpos, clen,
-					       &dealloc);
+		ret = ocfs2_remove_btree_range(dir, &et, cpos, p_cpos, clen, 0,
+					       &dealloc, 0);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;

fs/ocfs2/dlm/dlmast.c

@@ -88,7 +88,7 @@ static int dlm_should_cancel_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 	return 0;
 }
 
-static 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();
 
@@ -145,7 +145,7 @@ void dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock)
 }
 
 
-static 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();
 
@@ -451,7 +451,9 @@ 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_errno(ret);
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", ret, DLM_PROXY_AST_MSG, dlm->key,
+		     lock->ml.node);
 	else {
 		if (status == DLM_RECOVERING) {
 			mlog(ML_ERROR, "sent AST to node %u, it thinks this "

fs/ocfs2/dlm/dlmcommon.h

@@ -37,7 +37,7 @@
 #define DLM_THREAD_SHUFFLE_INTERVAL    5     // flush everything every 5 passes
 #define DLM_THREAD_MS                  200   // flush at least every 200 ms
 
-#define DLM_HASH_SIZE_DEFAULT	(1 << 14)
+#define DLM_HASH_SIZE_DEFAULT	(1 << 17)
 #if DLM_HASH_SIZE_DEFAULT < PAGE_SIZE
 # define DLM_HASH_PAGES		1
 #else
@@ -904,6 +904,8 @@ void __dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
 
 void dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
 void dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
+void __dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
+void __dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
 void dlm_do_local_ast(struct dlm_ctxt *dlm,
 		      struct dlm_lock_resource *res,
 		      struct dlm_lock *lock);

fs/ocfs2/dlm/dlmconvert.c

@@ -390,7 +390,9 @@ static enum dlm_status dlm_send_remote_convert_request(struct dlm_ctxt *dlm,
 		} else if (ret != DLM_NORMAL && ret != DLM_NOTQUEUED)
 			dlm_error(ret);
 	} else {
-		mlog_errno(tmpret);
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", tmpret, DLM_CONVERT_LOCK_MSG, dlm->key,
+		     res->owner);
 		if (dlm_is_host_down(tmpret)) {
 			/* instead of logging the same network error over
 			 * and over, sleep here and wait for the heartbeat

fs/ocfs2/dlm/dlmdomain.c

@@ -511,7 +511,7 @@ static void __dlm_print_nodes(struct dlm_ctxt *dlm)
 
 	assert_spin_locked(&dlm->spinlock);
 
-	printk(KERN_INFO "ocfs2_dlm: Nodes in domain (\"%s\"): ", dlm->name);
+	printk(KERN_NOTICE "o2dlm: Nodes in domain %s: ", dlm->name);
 
 	while ((node = find_next_bit(dlm->domain_map, O2NM_MAX_NODES,
 				     node + 1)) < O2NM_MAX_NODES) {
@@ -534,7 +534,7 @@ static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data,
 
 	node = exit_msg->node_idx;
 
-	printk(KERN_INFO "ocfs2_dlm: Node %u leaves domain %s\n", node, dlm->name);
+	printk(KERN_NOTICE "o2dlm: Node %u leaves domain %s\n", node, dlm->name);
 
 	spin_lock(&dlm->spinlock);
 	clear_bit(node, dlm->domain_map);
@@ -565,7 +565,9 @@ static int dlm_send_one_domain_exit(struct dlm_ctxt *dlm,
 	status = o2net_send_message(DLM_EXIT_DOMAIN_MSG, dlm->key,
 				    &leave_msg, sizeof(leave_msg), node,
 				    NULL);
-
+	if (status < 0)
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", status, DLM_EXIT_DOMAIN_MSG, dlm->key, node);
 	mlog(0, "status return %d from o2net_send_message\n", status);
 
 	return status;
@@ -904,7 +906,7 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data,
 		set_bit(assert->node_idx, dlm->domain_map);
 		__dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN);
 
-		printk(KERN_INFO "ocfs2_dlm: Node %u joins domain %s\n",
+		printk(KERN_NOTICE "o2dlm: Node %u joins domain %s\n",
 		       assert->node_idx, dlm->name);
 		__dlm_print_nodes(dlm);
 
@@ -962,7 +964,9 @@ static int dlm_send_one_join_cancel(struct dlm_ctxt *dlm,
 				    &cancel_msg, sizeof(cancel_msg), node,
 				    NULL);
 	if (status < 0) {
-		mlog_errno(status);
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", status, DLM_CANCEL_JOIN_MSG, DLM_MOD_KEY,
+		     node);
 		goto bail;
 	}
 
@@ -1029,10 +1033,11 @@ static int dlm_request_join(struct dlm_ctxt *dlm,
 	byte_copymap(join_msg.node_map, dlm->live_nodes_map, O2NM_MAX_NODES);
 
 	status = o2net_send_message(DLM_QUERY_JOIN_MSG, DLM_MOD_KEY, &join_msg,
-				    sizeof(join_msg), node,
-				    &join_resp);
+				    sizeof(join_msg), node, &join_resp);
 	if (status < 0 && status != -ENOPROTOOPT) {
-		mlog_errno(status);
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", status, DLM_QUERY_JOIN_MSG, DLM_MOD_KEY,
+		     node);
 		goto bail;
 	}
 	dlm_query_join_wire_to_packet(join_resp, &packet);
@@ -1103,7 +1108,9 @@ static int dlm_send_one_join_assert(struct dlm_ctxt *dlm,
 				    &assert_msg, sizeof(assert_msg), node,
 				    NULL);
 	if (status < 0)
-		mlog_errno(status);
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", status, DLM_ASSERT_JOINED_MSG, DLM_MOD_KEY,
+		     node);
 
 	return status;
 }
@@ -1516,7 +1523,7 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain,
 		goto leave;
 	}
 
-	dlm->name = kmalloc(strlen(domain) + 1, GFP_KERNEL);
+	dlm->name = kstrdup(domain, GFP_KERNEL);
 	if (dlm->name == NULL) {
 		mlog_errno(-ENOMEM);
 		kfree(dlm);
@@ -1550,7 +1557,6 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain,
 	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();
 

fs/ocfs2/dlm/dlmlock.c

@@ -329,7 +329,9 @@ static enum dlm_status dlm_send_remote_lock_request(struct dlm_ctxt *dlm,
 			BUG();
 		}
 	} else {
-		mlog_errno(tmpret);
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", tmpret, DLM_CREATE_LOCK_MSG, dlm->key,
+		     res->owner);
 		if (dlm_is_host_down(tmpret)) {
 			ret = DLM_RECOVERING;
 			mlog(0, "node %u died so returning DLM_RECOVERING "
@@ -429,7 +431,7 @@ struct dlm_lock * dlm_new_lock(int type, u8 node, u64 cookie,
 	struct dlm_lock *lock;
 	int kernel_allocated = 0;
 
-	lock = (struct dlm_lock *) kmem_cache_zalloc(dlm_lock_cache, GFP_NOFS);
+	lock = kmem_cache_zalloc(dlm_lock_cache, GFP_NOFS);
 	if (!lock)
 		return NULL;
 

fs/ocfs2/dlm/dlmmaster.c

@@ -617,13 +617,11 @@ struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm,
 {
 	struct dlm_lock_resource *res = NULL;
 
-	res = (struct dlm_lock_resource *)
-				kmem_cache_zalloc(dlm_lockres_cache, GFP_NOFS);
+	res = kmem_cache_zalloc(dlm_lockres_cache, GFP_NOFS);
 	if (!res)
 		goto error;
 
-	res->lockname.name = (char *)
-				kmem_cache_zalloc(dlm_lockname_cache, GFP_NOFS);
+	res->lockname.name = kmem_cache_zalloc(dlm_lockname_cache, GFP_NOFS);
 	if (!res->lockname.name)
 		goto error;
 
@@ -757,8 +755,7 @@ struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
 		spin_unlock(&dlm->spinlock);
 		mlog(0, "allocating a new resource\n");
 		/* nothing found and we need to allocate one. */
-		alloc_mle = (struct dlm_master_list_entry *)
-			kmem_cache_alloc(dlm_mle_cache, GFP_NOFS);
+		alloc_mle = kmem_cache_alloc(dlm_mle_cache, GFP_NOFS);
 		if (!alloc_mle)
 			goto leave;
 		res = dlm_new_lockres(dlm, lockid, namelen);
@@ -1542,8 +1539,7 @@ int dlm_master_request_handler(struct o2net_msg *msg, u32 len, void *data,
 			spin_unlock(&dlm->master_lock);
 			spin_unlock(&dlm->spinlock);
 
-			mle = (struct dlm_master_list_entry *)
-				kmem_cache_alloc(dlm_mle_cache, GFP_NOFS);
+			mle = kmem_cache_alloc(dlm_mle_cache, GFP_NOFS);
 			if (!mle) {
 				response = DLM_MASTER_RESP_ERROR;
 				mlog_errno(-ENOMEM);
@@ -1666,7 +1662,9 @@ static int dlm_do_assert_master(struct dlm_ctxt *dlm,
 		tmpret = o2net_send_message(DLM_ASSERT_MASTER_MSG, dlm->key,
 					    &assert, sizeof(assert), to, &r);
 		if (tmpret < 0) {
-			mlog(0, "assert_master returned %d!\n", tmpret);
+			mlog(ML_ERROR, "Error %d when sending message %u (key "
+			     "0x%x) to node %u\n", tmpret,
+			     DLM_ASSERT_MASTER_MSG, dlm->key, to);
 			if (!dlm_is_host_down(tmpret)) {
 				mlog(ML_ERROR, "unhandled error=%d!\n", tmpret);
 				BUG();
@@ -2205,7 +2203,9 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 	ret = o2net_send_message(DLM_DEREF_LOCKRES_MSG, dlm->key,
 				 &deref, sizeof(deref), res->owner, &r);
 	if (ret < 0)
-		mlog_errno(ret);
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", ret, DLM_DEREF_LOCKRES_MSG, dlm->key,
+		     res->owner);
 	else if (r < 0) {
 		/* BAD.  other node says I did not have a ref. */
 		mlog(ML_ERROR,"while dropping ref on %s:%.*s "
@@ -2452,8 +2452,7 @@ static int dlm_migrate_lockres(struct dlm_ctxt *dlm,
 		goto leave;
 	}
 
-	mle = (struct dlm_master_list_entry *) kmem_cache_alloc(dlm_mle_cache,
-								GFP_NOFS);
+	mle = kmem_cache_alloc(dlm_mle_cache, GFP_NOFS);
 	if (!mle) {
 		mlog_errno(ret);
 		goto leave;
@@ -2975,7 +2974,9 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
 					 &migrate, sizeof(migrate), nodenum,
 					 &status);
 		if (ret < 0) {
-			mlog(0, "migrate_request returned %d!\n", ret);
+			mlog(ML_ERROR, "Error %d when sending message %u (key "
+			     "0x%x) to node %u\n", ret, DLM_MIGRATE_REQUEST_MSG,
+			     dlm->key, nodenum);
 			if (!dlm_is_host_down(ret)) {
 				mlog(ML_ERROR, "unhandled error=%d!\n", ret);
 				BUG();
@@ -3033,8 +3034,7 @@ int dlm_migrate_request_handler(struct o2net_msg *msg, u32 len, void *data,
 	hash = dlm_lockid_hash(name, namelen);
 
 	/* preallocate.. if this fails, abort */
-	mle = (struct dlm_master_list_entry *) kmem_cache_alloc(dlm_mle_cache,
-							 GFP_NOFS);
+	mle = kmem_cache_alloc(dlm_mle_cache, GFP_NOFS);
 
 	if (!mle) {
 		ret = -ENOMEM;

fs/ocfs2/dlm/dlmrecovery.c

@@ -803,7 +803,9 @@ static int dlm_request_all_locks(struct dlm_ctxt *dlm, u8 request_from,
 
 	/* negative status is handled by caller */
 	if (ret < 0)
-		mlog_errno(ret);
+		mlog(ML_ERROR, "Error %d when sending message %u (key "
+		     "0x%x) to node %u\n", ret, DLM_LOCK_REQUEST_MSG,
+		     dlm->key, request_from);
 
 	// return from here, then
 	// sleep until all received or error
@@ -955,10 +957,10 @@ static int dlm_send_all_done_msg(struct dlm_ctxt *dlm, u8 dead_node, u8 send_to)
 	ret = o2net_send_message(DLM_RECO_DATA_DONE_MSG, dlm->key, &done_msg,
 				 sizeof(done_msg), send_to, &tmpret);
 	if (ret < 0) {
+		mlog(ML_ERROR, "Error %d when sending message %u (key "
+		     "0x%x) to node %u\n", ret, DLM_RECO_DATA_DONE_MSG,
+		     dlm->key, send_to);
 		if (!dlm_is_host_down(ret)) {
-			mlog_errno(ret);
-			mlog(ML_ERROR, "%s: unknown error sending data-done "
-			     "to %u\n", dlm->name, send_to);
 			BUG();
 		}
 	} else
@@ -1126,7 +1128,9 @@ static int dlm_send_mig_lockres_msg(struct dlm_ctxt *dlm,
 	if (ret < 0) {
 		/* XXX: negative status is not handled.
 		 * this will end up killing this node. */
-		mlog_errno(ret);
+		mlog(ML_ERROR, "Error %d when sending message %u (key "
+		     "0x%x) to node %u\n", ret, DLM_MIG_LOCKRES_MSG,
+		     dlm->key, send_to);
 	} else {
 		/* might get an -ENOMEM back here */
 		ret = status;
@@ -1642,7 +1646,9 @@ int dlm_do_master_requery(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 				 &req, sizeof(req), nodenum, &status);
 	/* XXX: negative status not handled properly here. */
 	if (ret < 0)
-		mlog_errno(ret);
+		mlog(ML_ERROR, "Error %d when sending message %u (key "
+		     "0x%x) to node %u\n", ret, DLM_MASTER_REQUERY_MSG,
+		     dlm->key, nodenum);
 	else {
 		BUG_ON(status < 0);
 		BUG_ON(status > DLM_LOCK_RES_OWNER_UNKNOWN);
@@ -2640,7 +2646,7 @@ static int dlm_send_begin_reco_message(struct dlm_ctxt *dlm, u8 dead_node)
 		if (dlm_is_host_down(ret)) {
 			/* node is down.  not involved in recovery
 			 * so just keep going */
-			mlog(0, "%s: node %u was down when sending "
+			mlog(ML_NOTICE, "%s: node %u was down when sending "
 			     "begin reco msg (%d)\n", dlm->name, nodenum, ret);
 			ret = 0;
 		}
@@ -2660,11 +2666,12 @@ static int dlm_send_begin_reco_message(struct dlm_ctxt *dlm, u8 dead_node)
 		}
 		if (ret < 0) {
 			struct dlm_lock_resource *res;
+
 			/* this is now a serious problem, possibly ENOMEM
 			 * in the network stack.  must retry */
 			mlog_errno(ret);
 			mlog(ML_ERROR, "begin reco of dlm %s to node %u "
-			    " returned %d\n", dlm->name, nodenum, ret);
+			     "returned %d\n", dlm->name, nodenum, ret);
 			res = dlm_lookup_lockres(dlm, DLM_RECOVERY_LOCK_NAME,
 						 DLM_RECOVERY_LOCK_NAME_LEN);
 			if (res) {
@@ -2789,7 +2796,9 @@ static int dlm_send_finalize_reco_message(struct dlm_ctxt *dlm)
 		if (ret >= 0)
 			ret = status;
 		if (ret < 0) {
-			mlog_errno(ret);
+			mlog(ML_ERROR, "Error %d when sending message %u (key "
+			     "0x%x) to node %u\n", ret, DLM_FINALIZE_RECO_MSG,
+			     dlm->key, nodenum);
 			if (dlm_is_host_down(ret)) {
 				/* this has no effect on this recovery
 				 * session, so set the status to zero to

fs/ocfs2/dlm/dlmthread.c

@@ -309,6 +309,7 @@ static void dlm_shuffle_lists(struct dlm_ctxt *dlm,
 	 * spinlock, and because we know that it is not migrating/
 	 * recovering/in-progress, it is fine to reserve asts and
 	 * basts right before queueing them all throughout */
+	assert_spin_locked(&dlm->ast_lock);
 	assert_spin_locked(&res->spinlock);
 	BUG_ON((res->state & (DLM_LOCK_RES_MIGRATING|
 			      DLM_LOCK_RES_RECOVERING|
@@ -337,7 +338,7 @@ static void dlm_shuffle_lists(struct dlm_ctxt *dlm,
 			/* queue the BAST if not already */
 			if (lock->ml.highest_blocked == LKM_IVMODE) {
 				__dlm_lockres_reserve_ast(res);
-				dlm_queue_bast(dlm, lock);
+				__dlm_queue_bast(dlm, lock);
 			}
 			/* update the highest_blocked if needed */
 			if (lock->ml.highest_blocked < target->ml.convert_type)
@@ -355,7 +356,7 @@ static void dlm_shuffle_lists(struct dlm_ctxt *dlm,
 			can_grant = 0;
 			if (lock->ml.highest_blocked == LKM_IVMODE) {
 				__dlm_lockres_reserve_ast(res);
-				dlm_queue_bast(dlm, lock);
+				__dlm_queue_bast(dlm, lock);
 			}
 			if (lock->ml.highest_blocked < target->ml.convert_type)
 				lock->ml.highest_blocked =
@@ -383,7 +384,7 @@ static void dlm_shuffle_lists(struct dlm_ctxt *dlm,
 		spin_unlock(&target->spinlock);
 
 		__dlm_lockres_reserve_ast(res);
-		dlm_queue_ast(dlm, target);
+		__dlm_queue_ast(dlm, target);
 		/* go back and check for more */
 		goto converting;
 	}
@@ -402,7 +403,7 @@ static void dlm_shuffle_lists(struct dlm_ctxt *dlm,
 			can_grant = 0;
 			if (lock->ml.highest_blocked == LKM_IVMODE) {
 				__dlm_lockres_reserve_ast(res);
-				dlm_queue_bast(dlm, lock);
+				__dlm_queue_bast(dlm, lock);
 			}
 			if (lock->ml.highest_blocked < target->ml.type)
 				lock->ml.highest_blocked = target->ml.type;
@@ -418,7 +419,7 @@ static void dlm_shuffle_lists(struct dlm_ctxt *dlm,
 			can_grant = 0;
 			if (lock->ml.highest_blocked == LKM_IVMODE) {
 				__dlm_lockres_reserve_ast(res);
-				dlm_queue_bast(dlm, lock);
+				__dlm_queue_bast(dlm, lock);
 			}
 			if (lock->ml.highest_blocked < target->ml.type)
 				lock->ml.highest_blocked = target->ml.type;
@@ -444,7 +445,7 @@ static void dlm_shuffle_lists(struct dlm_ctxt *dlm,
 		spin_unlock(&target->spinlock);
 
 		__dlm_lockres_reserve_ast(res);
-		dlm_queue_ast(dlm, target);
+		__dlm_queue_ast(dlm, target);
 		/* go back and check for more */
 		goto converting;
 	}
@@ -674,6 +675,7 @@ static int dlm_thread(void *data)
 		 	/* lockres can be re-dirtied/re-added to the
 			 * dirty_list in this gap, but that is ok */
 
+			spin_lock(&dlm->ast_lock);
 			spin_lock(&res->spinlock);
 			if (res->owner != dlm->node_num) {
 				__dlm_print_one_lock_resource(res);
@@ -694,6 +696,7 @@ static int dlm_thread(void *data)
 				/* move it to the tail and keep going */
 				res->state &= ~DLM_LOCK_RES_DIRTY;
 				spin_unlock(&res->spinlock);
+				spin_unlock(&dlm->ast_lock);
 				mlog(0, "delaying list shuffling for in-"
 				     "progress lockres %.*s, state=%d\n",
 				     res->lockname.len, res->lockname.name,
@@ -715,6 +718,7 @@ static int dlm_thread(void *data)
 			dlm_shuffle_lists(dlm, res);
 			res->state &= ~DLM_LOCK_RES_DIRTY;
 			spin_unlock(&res->spinlock);
+			spin_unlock(&dlm->ast_lock);
 
 			dlm_lockres_calc_usage(dlm, res);
 

fs/ocfs2/dlm/dlmunlock.c

@@ -354,7 +354,8 @@ static enum dlm_status dlm_send_remote_unlock_request(struct dlm_ctxt *dlm,
 			mlog(0, "master was in-progress.  retry\n");
 		ret = status;
 	} else {
-		mlog_errno(tmpret);
+		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
+		     "node %u\n", tmpret, DLM_UNLOCK_LOCK_MSG, dlm->key, owner);
 		if (dlm_is_host_down(tmpret)) {
 			/* NOTE: this seems strange, but it is what we want.
 			 * when the master goes down during a cancel or

fs/ocfs2/file.c

@@ -278,10 +278,7 @@ int ocfs2_update_inode_atime(struct inode *inode,
 	inode->i_atime = CURRENT_TIME;
 	di->i_atime = cpu_to_le64(inode->i_atime.tv_sec);
 	di->i_atime_nsec = cpu_to_le32(inode->i_atime.tv_nsec);
-
-	ret = ocfs2_journal_dirty(handle, bh);
-	if (ret < 0)
-		mlog_errno(ret);
+	ocfs2_journal_dirty(handle, bh);
 
 out_commit:
 	ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
@@ -430,9 +427,7 @@ static int ocfs2_orphan_for_truncate(struct ocfs2_super *osb,
 	di->i_ctime = di->i_mtime = cpu_to_le64(inode->i_ctime.tv_sec);
 	di->i_ctime_nsec = di->i_mtime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec);
 
-	status = ocfs2_journal_dirty(handle, fe_bh);
-	if (status < 0)
-		mlog_errno(status);
+	ocfs2_journal_dirty(handle, fe_bh);
 
 out_commit:
 	ocfs2_commit_trans(osb, handle);
@@ -449,7 +444,6 @@ static int ocfs2_truncate_file(struct inode *inode,
 	int status = 0;
 	struct ocfs2_dinode *fe = NULL;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	struct ocfs2_truncate_context *tc = NULL;
 
 	mlog_entry("(inode = %llu, new_i_size = %llu\n",
 		   (unsigned long long)OCFS2_I(inode)->ip_blkno,
@@ -488,6 +482,9 @@ static int ocfs2_truncate_file(struct inode *inode,
 
 	down_write(&OCFS2_I(inode)->ip_alloc_sem);
 
+	ocfs2_resv_discard(&osb->osb_la_resmap,
+			   &OCFS2_I(inode)->ip_la_data_resv);
+
 	/*
 	 * The inode lock forced other nodes to sync and drop their
 	 * pages, which (correctly) happens even if we have a truncate
@@ -517,13 +514,7 @@ static int ocfs2_truncate_file(struct inode *inode,
 		goto bail_unlock_sem;
 	}
 
-	status = ocfs2_prepare_truncate(osb, inode, di_bh, &tc);
-	if (status < 0) {
-		mlog_errno(status);
-		goto bail_unlock_sem;
-	}
-
-	status = ocfs2_commit_truncate(osb, inode, di_bh, tc);
+	status = ocfs2_commit_truncate(osb, inode, di_bh);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail_unlock_sem;
@@ -666,11 +657,7 @@ static int __ocfs2_extend_allocation(struct inode *inode, u32 logical_start,
 		goto leave;
 	}
 
-	status = ocfs2_journal_dirty(handle, bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto leave;
-	}
+	ocfs2_journal_dirty(handle, bh);
 
 	spin_lock(&OCFS2_I(inode)->ip_lock);
 	clusters_to_add -= (OCFS2_I(inode)->ip_clusters - prev_clusters);
@@ -1195,9 +1182,7 @@ static int __ocfs2_write_remove_suid(struct inode *inode,
 	di = (struct ocfs2_dinode *) bh->b_data;
 	di->i_mode = cpu_to_le16(inode->i_mode);
 
-	ret = ocfs2_journal_dirty(handle, bh);
-	if (ret < 0)
-		mlog_errno(ret);
+	ocfs2_journal_dirty(handle, bh);
 
 out_trans:
 	ocfs2_commit_trans(osb, handle);
@@ -1434,16 +1419,90 @@ static int ocfs2_zero_partial_clusters(struct inode *inode,
 	return ret;
 }
 
+static int ocfs2_find_rec(struct ocfs2_extent_list *el, u32 pos)
+{
+	int i;
+	struct ocfs2_extent_rec *rec = NULL;
+
+	for (i = le16_to_cpu(el->l_next_free_rec) - 1; i >= 0; i--) {
+
+		rec = &el->l_recs[i];
+
+		if (le32_to_cpu(rec->e_cpos) < pos)
+			break;
+	}
+
+	return i;
+}
+
+/*
+ * Helper to calculate the punching pos and length in one run, we handle the
+ * following three cases in order:
+ *
+ * - remove the entire record
+ * - remove a partial record
+ * - no record needs to be removed (hole-punching completed)
+*/
+static void ocfs2_calc_trunc_pos(struct inode *inode,
+				 struct ocfs2_extent_list *el,
+				 struct ocfs2_extent_rec *rec,
+				 u32 trunc_start, u32 *trunc_cpos,
+				 u32 *trunc_len, u32 *trunc_end,
+				 u64 *blkno, int *done)
+{
+	int ret = 0;
+	u32 coff, range;
+
+	range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec);
+
+	if (le32_to_cpu(rec->e_cpos) >= trunc_start) {
+		*trunc_cpos = le32_to_cpu(rec->e_cpos);
+		/*
+		 * Skip holes if any.
+		 */
+		if (range < *trunc_end)
+			*trunc_end = range;
+		*trunc_len = *trunc_end - le32_to_cpu(rec->e_cpos);
+		*blkno = le64_to_cpu(rec->e_blkno);
+		*trunc_end = le32_to_cpu(rec->e_cpos);
+	} else if (range > trunc_start) {
+		*trunc_cpos = trunc_start;
+		*trunc_len = *trunc_end - trunc_start;
+		coff = trunc_start - le32_to_cpu(rec->e_cpos);
+		*blkno = le64_to_cpu(rec->e_blkno) +
+				ocfs2_clusters_to_blocks(inode->i_sb, coff);
+		*trunc_end = trunc_start;
+	} else {
+		/*
+		 * It may have two following possibilities:
+		 *
+		 * - last record has been removed
+		 * - trunc_start was within a hole
+		 *
+		 * both two cases mean the completion of hole punching.
+		 */
+		ret = 1;
+	}
+
+	*done = ret;
+}
+
 static int ocfs2_remove_inode_range(struct inode *inode,
 				    struct buffer_head *di_bh, u64 byte_start,
 				    u64 byte_len)
 {
-	int ret = 0;
-	u32 trunc_start, trunc_len, cpos, phys_cpos, alloc_size;
+	int ret = 0, flags = 0, done = 0, i;
+	u32 trunc_start, trunc_len, trunc_end, trunc_cpos, phys_cpos;
+	u32 cluster_in_el;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	struct ocfs2_cached_dealloc_ctxt dealloc;
 	struct address_space *mapping = inode->i_mapping;
 	struct ocfs2_extent_tree et;
+	struct ocfs2_path *path = NULL;
+	struct ocfs2_extent_list *el = NULL;
+	struct ocfs2_extent_rec *rec = NULL;
+	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
+	u64 blkno, refcount_loc = le64_to_cpu(di->i_refcount_loc);
 
 	ocfs2_init_dinode_extent_tree(&et, INODE_CACHE(inode), di_bh);
 	ocfs2_init_dealloc_ctxt(&dealloc);
@@ -1469,17 +1528,35 @@ static int ocfs2_remove_inode_range(struct inode *inode,
 		goto out;
 	}
 
+	/*
+	 * For reflinks, we may need to CoW 2 clusters which might be
+	 * partially zero'd later, if hole's start and end offset were
+	 * within one cluster(means is not exactly aligned to clustersize).
+	 */
+
+	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_HAS_REFCOUNT_FL) {
+
+		ret = ocfs2_cow_file_pos(inode, di_bh, byte_start);
+		if (ret) {
+			mlog_errno(ret);
+			goto out;
+		}
+
+		ret = ocfs2_cow_file_pos(inode, di_bh, byte_start + byte_len);
+		if (ret) {
+			mlog_errno(ret);
+			goto out;
+		}
+	}
+
 	trunc_start = ocfs2_clusters_for_bytes(osb->sb, byte_start);
-	trunc_len = (byte_start + byte_len) >> osb->s_clustersize_bits;
-	if (trunc_len >= trunc_start)
-		trunc_len -= trunc_start;
-	else
-		trunc_len = 0;
+	trunc_end = (byte_start + byte_len) >> osb->s_clustersize_bits;
+	cluster_in_el = trunc_end;
 
-	mlog(0, "Inode: %llu, start: %llu, len: %llu, cstart: %u, clen: %u\n",
+	mlog(0, "Inode: %llu, start: %llu, len: %llu, cstart: %u, cend: %u\n",
 	     (unsigned long long)OCFS2_I(inode)->ip_blkno,
 	     (unsigned long long)byte_start,
-	     (unsigned long long)byte_len, trunc_start, trunc_len);
+	     (unsigned long long)byte_len, trunc_start, trunc_end);
 
 	ret = ocfs2_zero_partial_clusters(inode, byte_start, byte_len);
 	if (ret) {
@@ -1487,31 +1564,79 @@ static int ocfs2_remove_inode_range(struct inode *inode,
 		goto out;
 	}
 
-	cpos = trunc_start;
-	while (trunc_len) {
-		ret = ocfs2_get_clusters(inode, cpos, &phys_cpos,
-					 &alloc_size, NULL);
+	path = ocfs2_new_path_from_et(&et);
+	if (!path) {
+		ret = -ENOMEM;
+		mlog_errno(ret);
+		goto out;
+	}
+
+	while (trunc_end > trunc_start) {
+
+		ret = ocfs2_find_path(INODE_CACHE(inode), path,
+				      cluster_in_el);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
 
-		if (alloc_size > trunc_len)
-			alloc_size = trunc_len;
+		el = path_leaf_el(path);
 
-		/* Only do work for non-holes */
-		if (phys_cpos != 0) {
-			ret = ocfs2_remove_btree_range(inode, &et, cpos,
-						       phys_cpos, alloc_size,
-						       &dealloc);
+		i = ocfs2_find_rec(el, trunc_end);
+		/*
+		 * Need to go to previous extent block.
+		 */
+		if (i < 0) {
+			if (path->p_tree_depth == 0)
+				break;
+
+			ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb,
+							    path,
+							    &cluster_in_el);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;
 			}
+
+			/*
+			 * We've reached the leftmost extent block,
+			 * it's safe to leave.
+			 */
+			if (cluster_in_el == 0)
+				break;
+
+			/*
+			 * The 'pos' searched for previous extent block is
+			 * always one cluster less than actual trunc_end.
+			 */
+			trunc_end = cluster_in_el + 1;
+
+			ocfs2_reinit_path(path, 1);
+
+			continue;
+
+		} else
+			rec = &el->l_recs[i];
+
+		ocfs2_calc_trunc_pos(inode, el, rec, trunc_start, &trunc_cpos,
+				     &trunc_len, &trunc_end, &blkno, &done);
+		if (done)
+			break;
+
+		flags = rec->e_flags;
+		phys_cpos = ocfs2_blocks_to_clusters(inode->i_sb, blkno);
+
+		ret = ocfs2_remove_btree_range(inode, &et, trunc_cpos,
+					       phys_cpos, trunc_len, flags,
+					       &dealloc, refcount_loc);
+		if (ret < 0) {
+			mlog_errno(ret);
+			goto out;
 		}
 
-		cpos += alloc_size;
-		trunc_len -= alloc_size;
+		cluster_in_el = trunc_end;
+
+		ocfs2_reinit_path(path, 1);
 	}
 
 	ocfs2_truncate_cluster_pages(inode, byte_start, byte_len);

fs/ocfs2/inode.c

@@ -376,6 +376,10 @@ void ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
 
 	OCFS2_I(inode)->ip_last_used_slot = 0;
 	OCFS2_I(inode)->ip_last_used_group = 0;
+
+	if (S_ISDIR(inode->i_mode))
+		ocfs2_resv_set_type(&OCFS2_I(inode)->ip_la_data_resv,
+				    OCFS2_RESV_FLAG_DIR);
 	mlog_exit_void();
 }
 
@@ -539,7 +543,6 @@ static int ocfs2_truncate_for_delete(struct ocfs2_super *osb,
 				     struct buffer_head *fe_bh)
 {
 	int status = 0;
-	struct ocfs2_truncate_context *tc = NULL;
 	struct ocfs2_dinode *fe;
 	handle_t *handle = NULL;
 
@@ -582,13 +585,7 @@ static int ocfs2_truncate_for_delete(struct ocfs2_super *osb,
 		ocfs2_commit_trans(osb, handle);
 		handle = NULL;
 
-		status = ocfs2_prepare_truncate(osb, inode, fe_bh, &tc);
-		if (status < 0) {
-			mlog_errno(status);
-			goto out;
-		}
-
-		status = ocfs2_commit_truncate(osb, inode, fe_bh, tc);
+		status = ocfs2_commit_truncate(osb, inode, fe_bh);
 		if (status < 0) {
 			mlog_errno(status);
 			goto out;
@@ -659,12 +656,7 @@ static int ocfs2_remove_inode(struct inode *inode,
 
 	di->i_dtime = cpu_to_le64(CURRENT_TIME.tv_sec);
 	di->i_flags &= cpu_to_le32(~(OCFS2_VALID_FL | OCFS2_ORPHANED_FL));
-
-	status = ocfs2_journal_dirty(handle, di_bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto bail_commit;
-	}
+	ocfs2_journal_dirty(handle, di_bh);
 
 	ocfs2_remove_from_cache(INODE_CACHE(inode), di_bh);
 	dquot_free_inode(inode);
@@ -980,7 +972,7 @@ static void ocfs2_cleanup_delete_inode(struct inode *inode,
 void ocfs2_delete_inode(struct inode *inode)
 {
 	int wipe, status;
-	sigset_t blocked, oldset;
+	sigset_t oldset;
 	struct buffer_head *di_bh = NULL;
 
 	mlog_entry("(inode->i_ino = %lu)\n", inode->i_ino);
@@ -1007,13 +999,7 @@ void ocfs2_delete_inode(struct inode *inode)
 	 * messaging paths may return us -ERESTARTSYS. Which would
 	 * cause us to exit early, resulting in inodes being orphaned
 	 * forever. */
-	sigfillset(&blocked);
-	status = sigprocmask(SIG_BLOCK, &blocked, &oldset);
-	if (status < 0) {
-		mlog_errno(status);
-		ocfs2_cleanup_delete_inode(inode, 1);
-		goto bail;
-	}
+	ocfs2_block_signals(&oldset);
 
 	/*
 	 * Synchronize us against ocfs2_get_dentry. We take this in
@@ -1087,9 +1073,7 @@ void ocfs2_delete_inode(struct inode *inode)
 	ocfs2_nfs_sync_unlock(OCFS2_SB(inode->i_sb), 0);
 
 bail_unblock:
-	status = sigprocmask(SIG_SETMASK, &oldset, NULL);
-	if (status < 0)
-		mlog_errno(status);
+	ocfs2_unblock_signals(&oldset);
 bail:
 	clear_inode(inode);
 	mlog_exit_void();
@@ -1123,6 +1107,10 @@ void ocfs2_clear_inode(struct inode *inode)
 	ocfs2_mark_lockres_freeing(&oi->ip_inode_lockres);
 	ocfs2_mark_lockres_freeing(&oi->ip_open_lockres);
 
+	ocfs2_resv_discard(&OCFS2_SB(inode->i_sb)->osb_la_resmap,
+			   &oi->ip_la_data_resv);
+	ocfs2_resv_init_once(&oi->ip_la_data_resv);
+
 	/* We very well may get a clear_inode before all an inodes
 	 * metadata has hit disk. Of course, we can't drop any cluster
 	 * locks until the journal has finished with it. The only
@@ -1298,13 +1286,8 @@ int ocfs2_mark_inode_dirty(handle_t *handle,
 	fe->i_mtime = cpu_to_le64(inode->i_mtime.tv_sec);
 	fe->i_mtime_nsec = cpu_to_le32(inode->i_mtime.tv_nsec);
 
-	status = ocfs2_journal_dirty(handle, bh);
-	if (status < 0)
-		mlog_errno(status);
-
-	status = 0;
+	ocfs2_journal_dirty(handle, bh);
 leave:
-
 	mlog_exit(status);
 	return status;
 }

fs/ocfs2/inode.h

@@ -70,6 +70,8 @@ struct ocfs2_inode_info
 	/* Only valid if the inode is the dir. */
 	u32				ip_last_used_slot;
 	u64				ip_last_used_group;
+
+	struct ocfs2_alloc_reservation	ip_la_data_resv;
 };
 
 /*

fs/ocfs2/journal.c

@@ -402,9 +402,7 @@ int ocfs2_commit_trans(struct ocfs2_super *osb,
 }
 
 /*
- * 'nblocks' is what you want to add to the current
- * transaction. extend_trans will either extend the current handle by
- * nblocks, or commit it and start a new one with nblocks credits.
+ * 'nblocks' is what you want to add to the current transaction.
  *
  * This might call jbd2_journal_restart() which will commit dirty buffers
  * and then restart the transaction. Before calling
@@ -422,11 +420,15 @@ int ocfs2_commit_trans(struct ocfs2_super *osb,
  */
 int ocfs2_extend_trans(handle_t *handle, int nblocks)
 {
-	int status;
+	int status, old_nblocks;
 
 	BUG_ON(!handle);
-	BUG_ON(!nblocks);
+	BUG_ON(nblocks < 0);
+
+	if (!nblocks)
+		return 0;
 
+	old_nblocks = handle->h_buffer_credits;
 	mlog_entry_void();
 
 	mlog(0, "Trying to extend transaction by %d blocks\n", nblocks);
@@ -445,7 +447,8 @@ int ocfs2_extend_trans(handle_t *handle, int nblocks)
 		mlog(0,
 		     "jbd2_journal_extend failed, trying "
 		     "jbd2_journal_restart\n");
-		status = jbd2_journal_restart(handle, nblocks);
+		status = jbd2_journal_restart(handle,
+					      old_nblocks + nblocks);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
@@ -734,8 +737,7 @@ int ocfs2_journal_access(handle_t *handle, struct ocfs2_caching_info *ci,
 	return __ocfs2_journal_access(handle, ci, bh, NULL, type);
 }
 
-int ocfs2_journal_dirty(handle_t *handle,
-			struct buffer_head *bh)
+void ocfs2_journal_dirty(handle_t *handle, struct buffer_head *bh)
 {
 	int status;
 
@@ -743,13 +745,9 @@ int ocfs2_journal_dirty(handle_t *handle,
 		   (unsigned long long)bh->b_blocknr);
 
 	status = jbd2_journal_dirty_metadata(handle, bh);
-	if (status < 0)
-		mlog(ML_ERROR, "Could not dirty metadata buffer. "
-		     "(bh->b_blocknr=%llu)\n",
-		     (unsigned long long)bh->b_blocknr);
+	BUG_ON(status);
 
-	mlog_exit(status);
-	return status;
+	mlog_exit_void();
 }
 
 #define OCFS2_DEFAULT_COMMIT_INTERVAL	(HZ * JBD2_DEFAULT_MAX_COMMIT_AGE)

fs/ocfs2/journal.h

@@ -325,8 +325,7 @@ int ocfs2_journal_access(handle_t *handle, struct ocfs2_caching_info *ci,
  *	<modify the bh>
  * 	ocfs2_journal_dirty(handle, bh);
  */
-int                  ocfs2_journal_dirty(handle_t *handle,
-					 struct buffer_head *bh);
+void ocfs2_journal_dirty(handle_t *handle, struct buffer_head *bh);
 
 /*
  *  Credit Macros:
@@ -562,6 +561,18 @@ static inline int ocfs2_calc_group_alloc_credits(struct super_block *sb,
 	return blocks;
 }
 
+/*
+ * Allocating a discontiguous block group requires the credits from
+ * ocfs2_calc_group_alloc_credits() as well as enough credits to fill
+ * the group descriptor's extent list.  The caller already has started
+ * the transaction with ocfs2_calc_group_alloc_credits().  They extend
+ * it with these credits.
+ */
+static inline int ocfs2_calc_bg_discontig_credits(struct super_block *sb)
+{
+	return ocfs2_extent_recs_per_gd(sb);
+}
+
 static inline int ocfs2_calc_tree_trunc_credits(struct super_block *sb,
 						unsigned int clusters_to_del,
 						struct ocfs2_dinode *fe,

fs/ocfs2/localalloc.h

@@ -30,6 +30,9 @@ int ocfs2_load_local_alloc(struct ocfs2_super *osb);
 
 void ocfs2_shutdown_local_alloc(struct ocfs2_super *osb);
 
+void ocfs2_la_set_sizes(struct ocfs2_super *osb, int requested_mb);
+unsigned int ocfs2_la_default_mb(struct ocfs2_super *osb);
+
 int ocfs2_begin_local_alloc_recovery(struct ocfs2_super *osb,
 				     int node_num,
 				     struct ocfs2_dinode **alloc_copy);

fs/ocfs2/mmap.c

@@ -41,44 +41,20 @@
 #include "file.h"
 #include "inode.h"
 #include "mmap.h"
+#include "super.h"
 
-static inline int ocfs2_vm_op_block_sigs(sigset_t *blocked, sigset_t *oldset)
-{
-	/* The best way to deal with signals in the vm path is
-	 * to block them upfront, rather than allowing the
-	 * locking paths to return -ERESTARTSYS. */
-	sigfillset(blocked);
-
-	/* We should technically never get a bad return value
-	 * from sigprocmask */
-	return sigprocmask(SIG_BLOCK, blocked, oldset);
-}
-
-static inline int ocfs2_vm_op_unblock_sigs(sigset_t *oldset)
-{
-	return sigprocmask(SIG_SETMASK, oldset, NULL);
-}
 
 static int ocfs2_fault(struct vm_area_struct *area, struct vm_fault *vmf)
 {
-	sigset_t blocked, oldset;
-	int error, ret;
+	sigset_t oldset;
+	int ret;
 
 	mlog_entry("(area=%p, page offset=%lu)\n", area, vmf->pgoff);
 
-	error = ocfs2_vm_op_block_sigs(&blocked, &oldset);
-	if (error < 0) {
-		mlog_errno(error);
-		ret = VM_FAULT_SIGBUS;
-		goto out;
-	}
-
+	ocfs2_block_signals(&oldset);
 	ret = filemap_fault(area, vmf);
+	ocfs2_unblock_signals(&oldset);
 
-	error = ocfs2_vm_op_unblock_sigs(&oldset);
-	if (error < 0)
-		mlog_errno(error);
-out:
 	mlog_exit_ptr(vmf->page);
 	return ret;
 }
@@ -158,14 +134,10 @@ static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	struct page *page = vmf->page;
 	struct inode *inode = vma->vm_file->f_path.dentry->d_inode;
 	struct buffer_head *di_bh = NULL;
-	sigset_t blocked, oldset;
-	int ret, ret2;
+	sigset_t oldset;
+	int ret;
 
-	ret = ocfs2_vm_op_block_sigs(&blocked, &oldset);
-	if (ret < 0) {
-		mlog_errno(ret);
-		return ret;
-	}
+	ocfs2_block_signals(&oldset);
 
 	/*
 	 * The cluster locks taken will block a truncate from another
@@ -193,9 +165,7 @@ static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	ocfs2_inode_unlock(inode, 1);
 
 out:
-	ret2 = ocfs2_vm_op_unblock_sigs(&oldset);
-	if (ret2 < 0)
-		mlog_errno(ret2);
+	ocfs2_unblock_signals(&oldset);
 	if (ret)
 		ret = VM_FAULT_SIGBUS;
 	return ret;

fs/ocfs2/namei.c

@@ -239,6 +239,8 @@ static int ocfs2_mknod(struct inode *dir,
 	};
 	int did_quota_inode = 0;
 	struct ocfs2_dir_lookup_result lookup = { NULL, };
+	sigset_t oldset;
+	int did_block_signals = 0;
 
 	mlog_entry("(0x%p, 0x%p, %d, %lu, '%.*s')\n", dir, dentry, mode,
 		   (unsigned long)dev, dentry->d_name.len,
@@ -350,6 +352,10 @@ static int ocfs2_mknod(struct inode *dir,
 		goto leave;
 	}
 
+	/* Starting to change things, restart is no longer possible. */
+	ocfs2_block_signals(&oldset);
+	did_block_signals = 1;
+
 	status = dquot_alloc_inode(inode);
 	if (status)
 		goto leave;
@@ -384,11 +390,7 @@ static int ocfs2_mknod(struct inode *dir,
 			goto leave;
 		}
 		ocfs2_add_links_count(dirfe, 1);
-		status = ocfs2_journal_dirty(handle, parent_fe_bh);
-		if (status < 0) {
-			mlog_errno(status);
-			goto leave;
-		}
+		ocfs2_journal_dirty(handle, parent_fe_bh);
 		inc_nlink(dir);
 	}
 
@@ -439,6 +441,8 @@ static int ocfs2_mknod(struct inode *dir,
 		ocfs2_commit_trans(osb, handle);
 
 	ocfs2_inode_unlock(dir, 1);
+	if (did_block_signals)
+		ocfs2_unblock_signals(&oldset);
 
 	if (status == -ENOSPC)
 		mlog(0, "Disk is full\n");
@@ -487,14 +491,15 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 	int status = 0;
 	struct ocfs2_dinode *fe = NULL;
 	struct ocfs2_extent_list *fel;
-	u64 fe_blkno = 0;
+	u64 suballoc_loc, fe_blkno = 0;
 	u16 suballoc_bit;
 	u16 feat;
 
 	*new_fe_bh = NULL;
 
-	status = ocfs2_claim_new_inode(osb, handle, dir, parent_fe_bh,
-				       inode_ac, &suballoc_bit, &fe_blkno);
+	status = ocfs2_claim_new_inode(handle, dir, parent_fe_bh,
+				       inode_ac, &suballoc_loc,
+				       &suballoc_bit, &fe_blkno);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
@@ -531,6 +536,7 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 	fe->i_generation = cpu_to_le32(inode->i_generation);
 	fe->i_fs_generation = cpu_to_le32(osb->fs_generation);
 	fe->i_blkno = cpu_to_le64(fe_blkno);
+	fe->i_suballoc_loc = cpu_to_le64(suballoc_loc);
 	fe->i_suballoc_bit = cpu_to_le16(suballoc_bit);
 	fe->i_suballoc_slot = cpu_to_le16(inode_ac->ac_alloc_slot);
 	fe->i_uid = cpu_to_le32(inode->i_uid);
@@ -567,11 +573,7 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 		fel->l_count = cpu_to_le16(ocfs2_extent_recs_per_inode(osb->sb));
 	}
 
-	status = ocfs2_journal_dirty(handle, *new_fe_bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto leave;
-	}
+	ocfs2_journal_dirty(handle, *new_fe_bh);
 
 	ocfs2_populate_inode(inode, fe, 1);
 	ocfs2_ci_set_new(osb, INODE_CACHE(inode));
@@ -637,6 +639,7 @@ static int ocfs2_link(struct dentry *old_dentry,
 	struct ocfs2_dinode *fe = NULL;
 	struct ocfs2_super *osb = OCFS2_SB(dir->i_sb);
 	struct ocfs2_dir_lookup_result lookup = { NULL, };
+	sigset_t oldset;
 
 	mlog_entry("(inode=%lu, old='%.*s' new='%.*s')\n", inode->i_ino,
 		   old_dentry->d_name.len, old_dentry->d_name.name,
@@ -693,6 +696,9 @@ static int ocfs2_link(struct dentry *old_dentry,
 		goto out_unlock_inode;
 	}
 
+	/* Starting to change things, restart is no longer possible. */
+	ocfs2_block_signals(&oldset);
+
 	err = ocfs2_journal_access_di(handle, INODE_CACHE(inode), fe_bh,
 				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (err < 0) {
@@ -705,14 +711,7 @@ static int ocfs2_link(struct dentry *old_dentry,
 	ocfs2_set_links_count(fe, inode->i_nlink);
 	fe->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec);
 	fe->i_ctime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec);
-
-	err = ocfs2_journal_dirty(handle, fe_bh);
-	if (err < 0) {
-		ocfs2_add_links_count(fe, -1);
-		drop_nlink(inode);
-		mlog_errno(err);
-		goto out_commit;
-	}
+	ocfs2_journal_dirty(handle, fe_bh);
 
 	err = ocfs2_add_entry(handle, dentry, inode,
 			      OCFS2_I(inode)->ip_blkno,
@@ -736,6 +735,7 @@ static int ocfs2_link(struct dentry *old_dentry,
 
 out_commit:
 	ocfs2_commit_trans(osb, handle);
+	ocfs2_unblock_signals(&oldset);
 out_unlock_inode:
 	ocfs2_inode_unlock(inode, 1);
 
@@ -909,12 +909,7 @@ static int ocfs2_unlink(struct inode *dir,
 		drop_nlink(inode);
 	drop_nlink(inode);
 	ocfs2_set_links_count(fe, inode->i_nlink);
-
-	status = ocfs2_journal_dirty(handle, fe_bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto leave;
-	}
+	ocfs2_journal_dirty(handle, fe_bh);
 
 	dir->i_ctime = dir->i_mtime = CURRENT_TIME;
 	if (S_ISDIR(inode->i_mode))
@@ -1332,12 +1327,7 @@ static int ocfs2_rename(struct inode *old_dir,
 			ocfs2_set_links_count(newfe, 0);
 		else
 			ocfs2_add_links_count(newfe, -1);
-
-		status = ocfs2_journal_dirty(handle, newfe_bh);
-		if (status < 0) {
-			mlog_errno(status);
-			goto bail;
-		}
+		ocfs2_journal_dirty(handle, newfe_bh);
 	} else {
 		/* if the name was not found in new_dir, add it now */
 		status = ocfs2_add_entry(handle, new_dentry, old_inode,
@@ -1356,10 +1346,7 @@ static int ocfs2_rename(struct inode *old_dir,
 
 		old_di->i_ctime = cpu_to_le64(old_inode->i_ctime.tv_sec);
 		old_di->i_ctime_nsec = cpu_to_le32(old_inode->i_ctime.tv_nsec);
-
-		status = ocfs2_journal_dirty(handle, old_inode_bh);
-		if (status < 0)
-			mlog_errno(status);
+		ocfs2_journal_dirty(handle, old_inode_bh);
 	} else
 		mlog_errno(status);
 
@@ -1431,7 +1418,7 @@ static int ocfs2_rename(struct inode *old_dir,
 							 OCFS2_JOURNAL_ACCESS_WRITE);
 			fe = (struct ocfs2_dinode *) old_dir_bh->b_data;
 			ocfs2_set_links_count(fe, old_dir->i_nlink);
-			status = ocfs2_journal_dirty(handle, old_dir_bh);
+			ocfs2_journal_dirty(handle, old_dir_bh);
 		}
 	}
 	ocfs2_dentry_move(old_dentry, new_dentry, old_dir, new_dir);
@@ -1563,11 +1550,7 @@ static int ocfs2_create_symlink_data(struct ocfs2_super *osb,
 		       (bytes_left > sb->s_blocksize) ? sb->s_blocksize :
 		       bytes_left);
 
-		status = ocfs2_journal_dirty(handle, bhs[virtual]);
-		if (status < 0) {
-			mlog_errno(status);
-			goto bail;
-		}
+		ocfs2_journal_dirty(handle, bhs[virtual]);
 
 		virtual++;
 		p_blkno++;
@@ -1611,6 +1594,8 @@ static int ocfs2_symlink(struct inode *dir,
 	};
 	int did_quota = 0, did_quota_inode = 0;
 	struct ocfs2_dir_lookup_result lookup = { NULL, };
+	sigset_t oldset;
+	int did_block_signals = 0;
 
 	mlog_entry("(0x%p, 0x%p, symname='%s' actual='%.*s')\n", dir,
 		   dentry, symname, dentry->d_name.len, dentry->d_name.name);
@@ -1706,6 +1691,10 @@ static int ocfs2_symlink(struct inode *dir,
 		goto bail;
 	}
 
+	/* Starting to change things, restart is no longer possible. */
+	ocfs2_block_signals(&oldset);
+	did_block_signals = 1;
+
 	status = dquot_alloc_inode(inode);
 	if (status)
 		goto bail;
@@ -1814,6 +1803,8 @@ static int ocfs2_symlink(struct inode *dir,
 		ocfs2_commit_trans(osb, handle);
 
 	ocfs2_inode_unlock(dir, 1);
+	if (did_block_signals)
+		ocfs2_unblock_signals(&oldset);
 
 	brelse(new_fe_bh);
 	brelse(parent_fe_bh);
@@ -1961,12 +1952,7 @@ static int ocfs2_orphan_add(struct ocfs2_super *osb,
 	if (S_ISDIR(inode->i_mode))
 		ocfs2_add_links_count(orphan_fe, 1);
 	orphan_dir_inode->i_nlink = ocfs2_read_links_count(orphan_fe);
-
-	status = ocfs2_journal_dirty(handle, orphan_dir_bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto leave;
-	}
+	ocfs2_journal_dirty(handle, orphan_dir_bh);
 
 	status = __ocfs2_add_entry(handle, orphan_dir_inode, name,
 				   OCFS2_ORPHAN_NAMELEN, inode,
@@ -2065,12 +2051,7 @@ int ocfs2_orphan_del(struct ocfs2_super *osb,
 	if (S_ISDIR(inode->i_mode))
 		ocfs2_add_links_count(orphan_fe, -1);
 	orphan_dir_inode->i_nlink = ocfs2_read_links_count(orphan_fe);
-
-	status = ocfs2_journal_dirty(handle, orphan_dir_bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto leave;
-	}
+	ocfs2_journal_dirty(handle, orphan_dir_bh);
 
 leave:
 	ocfs2_free_dir_lookup_result(&lookup);

fs/ocfs2/ocfs2.h

@@ -47,6 +47,7 @@
 /* For struct ocfs2_blockcheck_stats */
 #include "blockcheck.h"
 
+#include "reservations.h"
 
 /* Caching of metadata buffers */
 
@@ -341,6 +342,9 @@ struct ocfs2_super
 	 */
 	unsigned int local_alloc_bits;
 	unsigned int local_alloc_default_bits;
+	/* osb_clusters_at_boot can become stale! Do not trust it to
+	 * be up to date. */
+	unsigned int osb_clusters_at_boot;
 
 	enum ocfs2_local_alloc_state local_alloc_state; /* protected
 							 * by osb_lock */
@@ -349,6 +353,11 @@ struct ocfs2_super
 
 	u64 la_last_gd;
 
+	struct ocfs2_reservation_map	osb_la_resmap;
+
+	unsigned int	osb_resv_level;
+	unsigned int	osb_dir_resv_level;
+
 	/* Next three fields are for local node slot recovery during
 	 * mount. */
 	int dirty;
@@ -482,6 +491,13 @@ static inline int ocfs2_supports_indexed_dirs(struct ocfs2_super *osb)
 	return 0;
 }
 
+static inline int ocfs2_supports_discontig_bg(struct ocfs2_super *osb)
+{
+	if (osb->s_feature_incompat & OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG)
+		return 1;
+	return 0;
+}
+
 static inline unsigned int ocfs2_link_max(struct ocfs2_super *osb)
 {
 	if (ocfs2_supports_indexed_dirs(osb))
@@ -763,6 +779,12 @@ static inline unsigned int ocfs2_megabytes_to_clusters(struct super_block *sb,
 	return megs << (20 - OCFS2_SB(sb)->s_clustersize_bits);
 }
 
+static inline unsigned int ocfs2_clusters_to_megabytes(struct super_block *sb,
+						       unsigned int clusters)
+{
+	return clusters >> (20 - OCFS2_SB(sb)->s_clustersize_bits);
+}
+
 static inline void _ocfs2_set_bit(unsigned int bit, unsigned long *bitmap)
 {
 	ext2_set_bit(bit, bitmap);

fs/ocfs2/ocfs2_fs.h

@@ -100,7 +100,8 @@
 					 | OCFS2_FEATURE_INCOMPAT_XATTR \
 					 | OCFS2_FEATURE_INCOMPAT_META_ECC \
 					 | OCFS2_FEATURE_INCOMPAT_INDEXED_DIRS \
-					 | OCFS2_FEATURE_INCOMPAT_REFCOUNT_TREE)
+					 | OCFS2_FEATURE_INCOMPAT_REFCOUNT_TREE \
+					 | OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG)
 #define OCFS2_FEATURE_RO_COMPAT_SUPP	(OCFS2_FEATURE_RO_COMPAT_UNWRITTEN \
 					 | OCFS2_FEATURE_RO_COMPAT_USRQUOTA \
 					 | OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)
@@ -165,6 +166,9 @@
 /* Refcount tree support */
 #define OCFS2_FEATURE_INCOMPAT_REFCOUNT_TREE	0x1000
 
+/* Discontigous block groups */
+#define OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG	0x2000
+
 /*
  * backup superblock flag is used to indicate that this volume
  * has backup superblocks.
@@ -282,14 +286,6 @@
 /* Journal limits (in bytes) */
 #define OCFS2_MIN_JOURNAL_SIZE		(4 * 1024 * 1024)
 
-/*
- * Default local alloc size (in megabytes)
- *
- * The value chosen should be such that most allocations, including new
- * block groups, use local alloc.
- */
-#define OCFS2_DEFAULT_LOCAL_ALLOC_SIZE	8
-
 /*
  * Inline extended attribute size (in bytes)
  * The value chosen should be aligned to 16 byte boundaries.
@@ -512,7 +508,10 @@ struct ocfs2_extent_block
 					   block group */
 	__le32 h_fs_generation;		/* Must match super block */
 	__le64 h_blkno;			/* Offset on disk, in blocks */
-/*20*/	__le64 h_reserved3;
+/*20*/	__le64 h_suballoc_loc;		/* Suballocator block group this
+					   eb belongs to.  Only valid
+					   if allocated from a
+					   discontiguous block group */
 	__le64 h_next_leaf_blk;		/* Offset on disk, in blocks,
 					   of next leaf header pointing
 					   to data */
@@ -679,7 +678,11 @@ struct ocfs2_dinode {
 /*80*/	struct ocfs2_block_check i_check;	/* Error checking */
 /*88*/	__le64 i_dx_root;		/* Pointer to dir index root block */
 /*90*/	__le64 i_refcount_loc;
-	__le64 i_reserved2[4];
+	__le64 i_suballoc_loc;		/* Suballocator block group this
+					   inode belongs to.  Only valid
+					   if allocated from a
+					   discontiguous block group */
+/*A0*/	__le64 i_reserved2[3];
 /*B8*/	union {
 		__le64 i_pad1;		/* Generic way to refer to this
 					   64bit union */
@@ -814,7 +817,12 @@ struct ocfs2_dx_root_block {
 	__le32		dr_reserved2;
 	__le64		dr_free_blk;		/* Pointer to head of free
 						 * unindexed block list. */
-	__le64		dr_reserved3[15];
+	__le64		dr_suballoc_loc;	/* Suballocator block group
+						   this root belongs to.
+						   Only valid if allocated
+						   from a discontiguous
+						   block group */
+	__le64		dr_reserved3[14];
 	union {
 		struct ocfs2_extent_list dr_list; /* Keep this aligned to 128
 						   * bits for maximum space
@@ -839,6 +847,13 @@ struct ocfs2_dx_leaf {
 	struct ocfs2_dx_entry_list	dl_list;
 };
 
+/*
+ * Largest bitmap for a block (suballocator) group in bytes.  This limit
+ * does not affect cluster groups (global allocator).  Cluster group
+ * bitmaps run to the end of the block.
+ */
+#define OCFS2_MAX_BG_BITMAP_SIZE	256
+
 /*
  * On disk allocator group structure for OCFS2
  */
@@ -860,7 +875,29 @@ struct ocfs2_group_desc
 	__le64   bg_blkno;               /* Offset on disk, in blocks */
 /*30*/	struct ocfs2_block_check bg_check;	/* Error checking */
 	__le64   bg_reserved2;
-/*40*/	__u8    bg_bitmap[0];
+/*40*/	union {
+		__u8    bg_bitmap[0];
+		struct {
+			/*
+			 * Block groups may be discontiguous when
+			 * OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG is set.
+			 * The extents of a discontigous block group are
+			 * stored in bg_list.  It is a flat list.
+			 * l_tree_depth must always be zero.  A
+			 * discontiguous group is signified by a non-zero
+			 * bg_list->l_next_free_rec.  Only block groups
+			 * can be discontiguous; Cluster groups cannot.
+			 * We've never made a block group with more than
+			 * 2048 blocks (256 bytes of bg_bitmap).  This
+			 * codifies that limit so that we can fit bg_list.
+			 * bg_size of a discontiguous block group will
+			 * be 256 to match bg_bitmap_filler.
+			 */
+			__u8 bg_bitmap_filler[OCFS2_MAX_BG_BITMAP_SIZE];
+/*140*/			struct ocfs2_extent_list bg_list;
+		};
+	};
+/* Actual on-disk size is one block */
 };
 
 struct ocfs2_refcount_rec {
@@ -905,7 +942,11 @@ struct ocfs2_refcount_block {
 /*40*/	__le32 rf_generation;		/* generation number. all be the same
 					 * for the same refcount tree. */
 	__le32 rf_reserved0;
-	__le64 rf_reserved1[7];
+	__le64 rf_suballoc_loc;		/* Suballocator block group this
+					   refcount block belongs to. Only
+					   valid if allocated from a
+					   discontiguous block group */
+/*50*/	__le64 rf_reserved1[6];
 /*80*/	union {
 		struct ocfs2_refcount_list rf_records;  /* List of refcount
 							  records */
@@ -1017,7 +1058,10 @@ struct ocfs2_xattr_block {
 					real xattr or a xattr tree. */
 	__le16	xb_reserved0;
 	__le32  xb_reserved1;
-	__le64	xb_reserved2;
+	__le64	xb_suballoc_loc;	/* Suballocator block group this
+					   xattr block belongs to. Only
+					   valid if allocated from a
+					   discontiguous block group */
 /*30*/	union {
 		struct ocfs2_xattr_header xb_header; /* xattr header if this
 							block contains xattr */
@@ -1254,6 +1298,16 @@ static inline u16 ocfs2_extent_recs_per_eb(struct super_block *sb)
 	return size / sizeof(struct ocfs2_extent_rec);
 }
 
+static inline u16 ocfs2_extent_recs_per_gd(struct super_block *sb)
+{
+	int size;
+
+	size = sb->s_blocksize -
+		offsetof(struct ocfs2_group_desc, bg_list.l_recs);
+
+	return size / sizeof(struct ocfs2_extent_rec);
+}
+
 static inline int ocfs2_dx_entries_per_leaf(struct super_block *sb)
 {
 	int size;
@@ -1284,13 +1338,23 @@ static inline u16 ocfs2_local_alloc_size(struct super_block *sb)
 	return size;
 }
 
-static inline int ocfs2_group_bitmap_size(struct super_block *sb)
+static inline int ocfs2_group_bitmap_size(struct super_block *sb,
+					  int suballocator,
+					  u32 feature_incompat)
 {
-	int size;
-
-	size = sb->s_blocksize -
+	int size = sb->s_blocksize -
 		offsetof(struct ocfs2_group_desc, bg_bitmap);
 
+	/*
+	 * The cluster allocator uses the entire block.  Suballocators have
+	 * never used more than OCFS2_MAX_BG_BITMAP_SIZE.  Unfortunately, older
+	 * code expects bg_size set to the maximum.  Thus we must keep
+	 * bg_size as-is unless discontig_bg is enabled.
+	 */
+	if (suballocator &&
+	    (feature_incompat & OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG))
+		size = OCFS2_MAX_BG_BITMAP_SIZE;
+
 	return size;
 }
 
@@ -1402,23 +1466,43 @@ static inline int ocfs2_extent_recs_per_eb(int blocksize)
 	return size / sizeof(struct ocfs2_extent_rec);
 }
 
-static inline int ocfs2_local_alloc_size(int blocksize)
+static inline int ocfs2_extent_recs_per_gd(int blocksize)
 {
 	int size;
 
 	size = blocksize -
-		offsetof(struct ocfs2_dinode, id2.i_lab.la_bitmap);
+		offsetof(struct ocfs2_group_desc, bg_list.l_recs);
 
-	return size;
+	return size / sizeof(struct ocfs2_extent_rec);
 }
 
-static inline int ocfs2_group_bitmap_size(int blocksize)
+static inline int ocfs2_local_alloc_size(int blocksize)
 {
 	int size;
 
 	size = blocksize -
+		offsetof(struct ocfs2_dinode, id2.i_lab.la_bitmap);
+
+	return size;
+}
+
+static inline int ocfs2_group_bitmap_size(int blocksize,
+					  int suballocator,
+					  uint32_t feature_incompat)
+{
+	int size = sb->s_blocksize -
 		offsetof(struct ocfs2_group_desc, bg_bitmap);
 
+	/*
+	 * The cluster allocator uses the entire block.  Suballocators have
+	 * never used more than OCFS2_MAX_BG_BITMAP_SIZE.  Unfortunately, older
+	 * code expects bg_size set to the maximum.  Thus we must keep
+	 * bg_size as-is unless discontig_bg is enabled.
+	 */
+	if (suballocator &&
+	    (feature_incompat & OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG))
+		size = OCFS2_MAX_BG_BITMAP_SIZE;
+
 	return size;
 }
 
@@ -1491,5 +1575,19 @@ static inline void ocfs2_set_de_type(struct ocfs2_dir_entry *de,
 	de->file_type = ocfs2_type_by_mode[(mode & S_IFMT)>>S_SHIFT];
 }
 
+static inline int ocfs2_gd_is_discontig(struct ocfs2_group_desc *gd)
+{
+	if ((offsetof(struct ocfs2_group_desc, bg_bitmap) +
+	     le16_to_cpu(gd->bg_size)) !=
+	    offsetof(struct ocfs2_group_desc, bg_list))
+		return 0;
+	/*
+	 * Only valid to check l_next_free_rec if
+	 * bg_bitmap + bg_size == bg_list.
+	 */
+	if (!gd->bg_list.l_next_free_rec)
+		return 0;
+	return 1;
+}
 #endif  /* _OCFS2_FS_H */
 

fs/ocfs2/quota_global.c

@@ -261,10 +261,8 @@ ssize_t ocfs2_quota_write(struct super_block *sb, int type,
 		brelse(bh);
 		goto out;
 	}
-	err = ocfs2_journal_dirty(handle, bh);
+	ocfs2_journal_dirty(handle, bh);
 	brelse(bh);
-	if (err < 0)
-		goto out;
 out:
 	if (err) {
 		mutex_unlock(&gqinode->i_mutex);

fs/ocfs2/quota_local.c

@@ -119,12 +119,8 @@ static int ocfs2_modify_bh(struct inode *inode, struct buffer_head *bh,
 	lock_buffer(bh);
 	modify(bh, private);
 	unlock_buffer(bh);
-	status = ocfs2_journal_dirty(handle, bh);
-	if (status < 0) {
-		mlog_errno(status);
-		ocfs2_commit_trans(OCFS2_SB(sb), handle);
-		return status;
-	}
+	ocfs2_journal_dirty(handle, bh);
+
 	status = ocfs2_commit_trans(OCFS2_SB(sb), handle);
 	if (status < 0) {
 		mlog_errno(status);
@@ -523,9 +519,7 @@ static int ocfs2_recover_local_quota_file(struct inode *lqinode,
 			ocfs2_clear_bit(bit, dchunk->dqc_bitmap);
 			le32_add_cpu(&dchunk->dqc_free, 1);
 			unlock_buffer(qbh);
-			status = ocfs2_journal_dirty(handle, qbh);
-			if (status < 0)
-				mlog_errno(status);
+			ocfs2_journal_dirty(handle, qbh);
 out_commit:
 			mutex_unlock(&sb_dqopt(sb)->dqio_mutex);
 			ocfs2_commit_trans(OCFS2_SB(sb), handle);
@@ -631,9 +625,7 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
 		lock_buffer(bh);
 		ldinfo->dqi_flags = cpu_to_le32(flags | OLQF_CLEAN);
 		unlock_buffer(bh);
-		status = ocfs2_journal_dirty(handle, bh);
-		if (status < 0)
-			mlog_errno(status);
+		ocfs2_journal_dirty(handle, bh);
 out_trans:
 		ocfs2_commit_trans(osb, handle);
 out_bh:
@@ -1009,11 +1001,7 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk(
 	       sb->s_blocksize - sizeof(struct ocfs2_local_disk_chunk) -
 	       OCFS2_QBLK_RESERVED_SPACE);
 	unlock_buffer(bh);
-	status = ocfs2_journal_dirty(handle, bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto out_trans;
-	}
+	ocfs2_journal_dirty(handle, bh);
 
 	/* Initialize new block with structures */
 	down_read(&OCFS2_I(lqinode)->ip_alloc_sem);
@@ -1040,11 +1028,7 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk(
 	lock_buffer(dbh);
 	memset(dbh->b_data, 0, sb->s_blocksize - OCFS2_QBLK_RESERVED_SPACE);
 	unlock_buffer(dbh);
-	status = ocfs2_journal_dirty(handle, dbh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto out_trans;
-	}
+	ocfs2_journal_dirty(handle, dbh);
 
 	/* Update local quotafile info */
 	oinfo->dqi_blocks += 2;
@@ -1155,11 +1139,8 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file(
 	lock_buffer(bh);
 	memset(bh->b_data, 0, sb->s_blocksize);
 	unlock_buffer(bh);
-	status = ocfs2_journal_dirty(handle, bh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto out_trans;
-	}
+	ocfs2_journal_dirty(handle, bh);
+
 	/* Update chunk header */
 	status = ocfs2_journal_access_dq(handle, INODE_CACHE(lqinode),
 					 chunk->qc_headerbh,
@@ -1173,11 +1154,8 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file(
 	lock_buffer(chunk->qc_headerbh);
 	le32_add_cpu(&dchunk->dqc_free, ol_quota_entries_per_block(sb));
 	unlock_buffer(chunk->qc_headerbh);
-	status = ocfs2_journal_dirty(handle, chunk->qc_headerbh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto out_trans;
-	}
+	ocfs2_journal_dirty(handle, chunk->qc_headerbh);
+
 	/* Update file header */
 	oinfo->dqi_blocks++;
 	status = ocfs2_local_write_info(sb, type);
@@ -1312,12 +1290,8 @@ static int ocfs2_local_release_dquot(struct dquot *dquot)
 	ocfs2_clear_bit(offset, dchunk->dqc_bitmap);
 	le32_add_cpu(&dchunk->dqc_free, 1);
 	unlock_buffer(od->dq_chunk->qc_headerbh);
-	status = ocfs2_journal_dirty(handle, od->dq_chunk->qc_headerbh);
-	if (status < 0) {
-		mlog_errno(status);
-		goto out;
-	}
-	status = 0;
+	ocfs2_journal_dirty(handle, od->dq_chunk->qc_headerbh);
+
 out:
 	/* Clear the read bit so that next time someone uses this
 	 * dquot he reads fresh info from disk and allocates local

fs/ocfs2/refcounttree.c

@@ -570,7 +570,7 @@ static int ocfs2_create_refcount_tree(struct inode *inode,
 	struct ocfs2_refcount_tree *new_tree = NULL, *tree = NULL;
 	u16 suballoc_bit_start;
 	u32 num_got;
-	u64 first_blkno;
+	u64 suballoc_loc, first_blkno;
 
 	BUG_ON(oi->ip_dyn_features & OCFS2_HAS_REFCOUNT_FL);
 
@@ -596,7 +596,7 @@ static int ocfs2_create_refcount_tree(struct inode *inode,
 		goto out_commit;
 	}
 
-	ret = ocfs2_claim_metadata(osb, handle, meta_ac, 1,
+	ret = ocfs2_claim_metadata(handle, meta_ac, 1, &suballoc_loc,
 				   &suballoc_bit_start, &num_got,
 				   &first_blkno);
 	if (ret) {
@@ -626,6 +626,7 @@ static int ocfs2_create_refcount_tree(struct inode *inode,
 	memset(rb, 0, inode->i_sb->s_blocksize);
 	strcpy((void *)rb, OCFS2_REFCOUNT_BLOCK_SIGNATURE);
 	rb->rf_suballoc_slot = cpu_to_le16(meta_ac->ac_alloc_slot);
+	rb->rf_suballoc_loc = cpu_to_le64(suballoc_loc);
 	rb->rf_suballoc_bit = cpu_to_le16(suballoc_bit_start);
 	rb->rf_fs_generation = cpu_to_le32(osb->fs_generation);
 	rb->rf_blkno = cpu_to_le64(first_blkno);
@@ -790,7 +791,10 @@ int ocfs2_remove_refcount_tree(struct inode *inode, struct buffer_head *di_bh)
 	if (le32_to_cpu(rb->rf_count) == 1) {
 		blk = le64_to_cpu(rb->rf_blkno);
 		bit = le16_to_cpu(rb->rf_suballoc_bit);
-		bg_blkno = ocfs2_which_suballoc_group(blk, bit);
+		if (rb->rf_suballoc_loc)
+			bg_blkno = le64_to_cpu(rb->rf_suballoc_loc);
+		else
+			bg_blkno = ocfs2_which_suballoc_group(blk, bit);
 
 		alloc_inode = ocfs2_get_system_file_inode(osb,
 					EXTENT_ALLOC_SYSTEM_INODE,
@@ -1268,9 +1272,7 @@ static int ocfs2_change_refcount_rec(handle_t *handle,
 	} else if (merge)
 		ocfs2_refcount_rec_merge(rb, index);
 
-	ret = ocfs2_journal_dirty(handle, ref_leaf_bh);
-	if (ret)
-		mlog_errno(ret);
+	ocfs2_journal_dirty(handle, ref_leaf_bh);
 out:
 	return ret;
 }
@@ -1284,7 +1286,7 @@ static int ocfs2_expand_inline_ref_root(handle_t *handle,
 	int ret;
 	u16 suballoc_bit_start;
 	u32 num_got;
-	u64 blkno;
+	u64 suballoc_loc, blkno;
 	struct super_block *sb = ocfs2_metadata_cache_get_super(ci);
 	struct buffer_head *new_bh = NULL;
 	struct ocfs2_refcount_block *new_rb;
@@ -1298,7 +1300,7 @@ static int ocfs2_expand_inline_ref_root(handle_t *handle,
 		goto out;
 	}
 
-	ret = ocfs2_claim_metadata(OCFS2_SB(sb), handle, meta_ac, 1,
+	ret = ocfs2_claim_metadata(handle, meta_ac, 1, &suballoc_loc,
 				   &suballoc_bit_start, &num_got,
 				   &blkno);
 	if (ret) {
@@ -1330,6 +1332,7 @@ static int ocfs2_expand_inline_ref_root(handle_t *handle,
 
 	new_rb = (struct ocfs2_refcount_block *)new_bh->b_data;
 	new_rb->rf_suballoc_slot = cpu_to_le16(meta_ac->ac_alloc_slot);
+	new_rb->rf_suballoc_loc = cpu_to_le64(suballoc_loc);
 	new_rb->rf_suballoc_bit = cpu_to_le16(suballoc_bit_start);
 	new_rb->rf_blkno = cpu_to_le64(blkno);
 	new_rb->rf_cpos = cpu_to_le32(0);
@@ -1524,7 +1527,7 @@ static int ocfs2_new_leaf_refcount_block(handle_t *handle,
 	int ret;
 	u16 suballoc_bit_start;
 	u32 num_got, new_cpos;
-	u64 blkno;
+	u64 suballoc_loc, blkno;
 	struct super_block *sb = ocfs2_metadata_cache_get_super(ci);
 	struct ocfs2_refcount_block *root_rb =
 			(struct ocfs2_refcount_block *)ref_root_bh->b_data;
@@ -1548,7 +1551,7 @@ static int ocfs2_new_leaf_refcount_block(handle_t *handle,
 		goto out;
 	}
 
-	ret = ocfs2_claim_metadata(OCFS2_SB(sb), handle, meta_ac, 1,
+	ret = ocfs2_claim_metadata(handle, meta_ac, 1, &suballoc_loc,
 				   &suballoc_bit_start, &num_got,
 				   &blkno);
 	if (ret) {
@@ -1576,6 +1579,7 @@ static int ocfs2_new_leaf_refcount_block(handle_t *handle,
 	memset(new_rb, 0, sb->s_blocksize);
 	strcpy((void *)new_rb, OCFS2_REFCOUNT_BLOCK_SIGNATURE);
 	new_rb->rf_suballoc_slot = cpu_to_le16(meta_ac->ac_alloc_slot);
+	new_rb->rf_suballoc_loc = cpu_to_le64(suballoc_loc);
 	new_rb->rf_suballoc_bit = cpu_to_le16(suballoc_bit_start);
 	new_rb->rf_fs_generation = cpu_to_le32(OCFS2_SB(sb)->fs_generation);
 	new_rb->rf_blkno = cpu_to_le64(blkno);
@@ -1694,7 +1698,7 @@ static int ocfs2_adjust_refcount_rec(handle_t *handle,
 	 * 2 more credits, one for the leaf refcount block, one for
 	 * the extent block contains the extent rec.
 	 */
-	ret = ocfs2_extend_trans(handle, handle->h_buffer_credits + 2);
+	ret = ocfs2_extend_trans(handle, 2);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out;
@@ -1802,11 +1806,7 @@ static int ocfs2_insert_refcount_rec(handle_t *handle,
 	if (merge)
 		ocfs2_refcount_rec_merge(rb, index);
 
-	ret = ocfs2_journal_dirty(handle, ref_leaf_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
+	ocfs2_journal_dirty(handle, ref_leaf_bh);
 
 	if (index == 0) {
 		ret = ocfs2_adjust_refcount_rec(handle, ci,
@@ -1977,9 +1977,7 @@ static int ocfs2_split_refcount_rec(handle_t *handle,
 			ocfs2_refcount_rec_merge(rb, index);
 	}
 
-	ret = ocfs2_journal_dirty(handle, ref_leaf_bh);
-	if (ret)
-		mlog_errno(ret);
+	ocfs2_journal_dirty(handle, ref_leaf_bh);
 
 out:
 	brelse(new_bh);
@@ -2112,6 +2110,7 @@ static int ocfs2_remove_refcount_extent(handle_t *handle,
 	 */
 	ret = ocfs2_cache_block_dealloc(dealloc, EXTENT_ALLOC_SYSTEM_INODE,
 					le16_to_cpu(rb->rf_suballoc_slot),
+					le64_to_cpu(rb->rf_suballoc_loc),
 					le64_to_cpu(rb->rf_blkno),
 					le16_to_cpu(rb->rf_suballoc_bit));
 	if (ret) {
@@ -2516,20 +2515,19 @@ static int ocfs2_calc_refcount_meta_credits(struct super_block *sb,
  *
  * Normally the refcount blocks store these refcount should be
  * contiguous also, so that we can get the number easily.
- * As for meta_ac, we will at most add split 2 refcount record and
- * 2 more refcount block, so just check it in a rough way.
+ * We will at most add split 2 refcount records and 2 more
+ * refcount blocks, so just check it in a rough way.
  *
  * Caller must hold refcount tree lock.
  */
 int ocfs2_prepare_refcount_change_for_del(struct inode *inode,
-					  struct buffer_head *di_bh,
+					  u64 refcount_loc,
 					  u64 phys_blkno,
 					  u32 clusters,
 					  int *credits,
-					  struct ocfs2_alloc_context **meta_ac)
+					  int *ref_blocks)
 {
-	int ret, ref_blocks = 0;
-	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
+	int ret;
 	struct ocfs2_inode_info *oi = OCFS2_I(inode);
 	struct buffer_head *ref_root_bh = NULL;
 	struct ocfs2_refcount_tree *tree;
@@ -2546,14 +2544,13 @@ int ocfs2_prepare_refcount_change_for_del(struct inode *inode,
 	BUG_ON(!(oi->ip_dyn_features & OCFS2_HAS_REFCOUNT_FL));
 
 	ret = ocfs2_get_refcount_tree(OCFS2_SB(inode->i_sb),
-				      le64_to_cpu(di->i_refcount_loc), &tree);
+				      refcount_loc, &tree);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_read_refcount_block(&tree->rf_ci,
-					le64_to_cpu(di->i_refcount_loc),
+	ret = ocfs2_read_refcount_block(&tree->rf_ci, refcount_loc,
 					&ref_root_bh);
 	if (ret) {
 		mlog_errno(ret);
@@ -2564,21 +2561,14 @@ int ocfs2_prepare_refcount_change_for_del(struct inode *inode,
 					       &tree->rf_ci,
 					       ref_root_bh,
 					       start_cpos, clusters,
-					       &ref_blocks, credits);
+					       ref_blocks, credits);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	mlog(0, "reserve new metadata %d, credits = %d\n",
-	     ref_blocks, *credits);
-
-	if (ref_blocks) {
-		ret = ocfs2_reserve_new_metadata_blocks(OCFS2_SB(inode->i_sb),
-							ref_blocks, meta_ac);
-		if (ret)
-			mlog_errno(ret);
-	}
+	mlog(0, "reserve new metadata %d blocks, credits = %d\n",
+	     *ref_blocks, *credits);
 
 out:
 	brelse(ref_root_bh);
@@ -3040,11 +3030,7 @@ static int ocfs2_duplicate_clusters_by_jbd(handle_t *handle,
 		}
 
 		memcpy(new_bh->b_data, old_bh->b_data, sb->s_blocksize);
-		ret = ocfs2_journal_dirty(handle, new_bh);
-		if (ret) {
-			mlog_errno(ret);
-			break;
-		}
+		ocfs2_journal_dirty(handle, new_bh);
 
 		brelse(new_bh);
 		brelse(old_bh);
@@ -3282,7 +3268,7 @@ static int ocfs2_make_clusters_writable(struct super_block *sb,
 		} else {
 			delete = 1;
 
-			ret = __ocfs2_claim_clusters(osb, handle,
+			ret = __ocfs2_claim_clusters(handle,
 						     context->data_ac,
 						     1, set_len,
 						     &new_bit, &new_len);

fs/ocfs2/refcounttree.h

@@ -47,11 +47,11 @@ int ocfs2_decrease_refcount(struct inode *inode,
 			    struct ocfs2_cached_dealloc_ctxt *dealloc,
 			    int delete);
 int ocfs2_prepare_refcount_change_for_del(struct inode *inode,
-					  struct buffer_head *di_bh,
+					  u64 refcount_loc,
 					  u64 phys_blkno,
 					  u32 clusters,
 					  int *credits,
-					  struct ocfs2_alloc_context **meta_ac);
+					  int *ref_blocks);
 int ocfs2_refcount_cow(struct inode *inode, struct buffer_head *di_bh,
 		       u32 cpos, u32 write_len, u32 max_cpos);
 

fs/ocfs2/reservations.h

+/* -*- mode: c; c-basic-offset: 8; -*-
+ * vim: noexpandtab sw=8 ts=8 sts=0:
+ *
+ * reservations.h
+ *
+ * Allocation reservations function prototypes and structures.
+ *
+ * Copyright (C) 2010 Novell.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef	OCFS2_RESERVATIONS_H
+#define	OCFS2_RESERVATIONS_H
+
+#include <linux/rbtree.h>
+
+#define OCFS2_DEFAULT_RESV_LEVEL	2
+#define OCFS2_MAX_RESV_LEVEL	9
+#define OCFS2_MIN_RESV_LEVEL	0
+
+struct ocfs2_alloc_reservation {
+	struct rb_node	r_node;
+
+	unsigned int	r_start;	/* Begining of current window */
+	unsigned int	r_len;		/* Length of the window */
+
+	unsigned int	r_last_len;	/* Length of most recent alloc */
+	unsigned int	r_last_start;	/* Start of most recent alloc */
+	struct list_head	r_lru;	/* LRU list head */
+
+	unsigned int	r_flags;
+};
+
+#define	OCFS2_RESV_FLAG_INUSE	0x01	/* Set when r_node is part of a btree */
+#define	OCFS2_RESV_FLAG_TMP	0x02	/* Temporary reservation, will be
+					 * destroyed immedately after use */
+#define	OCFS2_RESV_FLAG_DIR	0x04	/* Reservation is for an unindexed
+					 * directory btree */
+
+struct ocfs2_reservation_map {
+	struct rb_root		m_reservations;
+	char			*m_disk_bitmap;
+
+	struct ocfs2_super	*m_osb;
+
+	/* The following are not initialized to meaningful values until a disk
+	 * bitmap is provided. */
+	u32			m_bitmap_len;	/* Number of valid
+						 * bits available */
+
+	struct list_head	m_lru;		/* LRU of reservations
+						 * structures. */
+
+};
+
+void ocfs2_resv_init_once(struct ocfs2_alloc_reservation *resv);
+
+#define OCFS2_RESV_TYPES	(OCFS2_RESV_FLAG_TMP|OCFS2_RESV_FLAG_DIR)
+void ocfs2_resv_set_type(struct ocfs2_alloc_reservation *resv,
+			 unsigned int flags);
+
+int ocfs2_dir_resv_allowed(struct ocfs2_super *osb);
+
+/**
+ * ocfs2_resv_discard() - truncate a reservation
+ * @resmap:
+ * @resv: the reservation to truncate.
+ *
+ * After this function is called, the reservation will be empty, and
+ * unlinked from the rbtree.
+ */
+void ocfs2_resv_discard(struct ocfs2_reservation_map *resmap,
+			struct ocfs2_alloc_reservation *resv);
+
+
+/**
+ * ocfs2_resmap_init() - Initialize fields of a reservations bitmap
+ * @resmap: struct ocfs2_reservation_map to initialize
+ * @obj: unused for now
+ * @ops: unused for now
+ * @max_bitmap_bytes: Maximum size of the bitmap (typically blocksize)
+ *
+ * Only possible return value other than '0' is -ENOMEM for failure to
+ * allocation mirror bitmap.
+ */
+int ocfs2_resmap_init(struct ocfs2_super *osb,
+		      struct ocfs2_reservation_map *resmap);
+
+/**
+ * ocfs2_resmap_restart() - "restart" a reservation bitmap
+ * @resmap: reservations bitmap
+ * @clen: Number of valid bits in the bitmap
+ * @disk_bitmap: the disk bitmap this resmap should refer to.
+ *
+ * Re-initialize the parameters of a reservation bitmap. This is
+ * useful for local alloc window slides.
+ *
+ * This function will call ocfs2_trunc_resv against all existing
+ * reservations. A future version will recalculate existing
+ * reservations based on the new bitmap.
+ */
+void ocfs2_resmap_restart(struct ocfs2_reservation_map *resmap,
+			  unsigned int clen, char *disk_bitmap);
+
+/**
+ * ocfs2_resmap_uninit() - uninitialize a reservation bitmap structure
+ * @resmap: the struct ocfs2_reservation_map to uninitialize
+ */
+void ocfs2_resmap_uninit(struct ocfs2_reservation_map *resmap);
+
+/**
+ * ocfs2_resmap_resv_bits() - Return still-valid reservation bits
+ * @resmap: reservations bitmap
+ * @resv: reservation to base search from
+ * @cstart: start of proposed allocation
+ * @clen: length (in clusters) of proposed allocation
+ *
+ * Using the reservation data from resv, this function will compare
+ * resmap and resmap->m_disk_bitmap to determine what part (if any) of
+ * the reservation window is still clear to use. If resv is empty,
+ * this function will try to allocate a window for it.
+ *
+ * On success, zero is returned and the valid allocation area is set in cstart
+ * and clen.
+ *
+ * Returns -ENOSPC if reservations are disabled.
+ */
+int ocfs2_resmap_resv_bits(struct ocfs2_reservation_map *resmap,
+			   struct ocfs2_alloc_reservation *resv,
+			   int *cstart, int *clen);
+
+/**
+ * ocfs2_resmap_claimed_bits() - Tell the reservation code that bits were used.
+ * @resmap: reservations bitmap
+ * @resv: optional reservation to recalulate based on new bitmap
+ * @cstart: start of allocation in clusters
+ * @clen: end of allocation in clusters.
+ *
+ * Tell the reservation code that bits were used to fulfill allocation in
+ * resmap. The bits don't have to have been part of any existing
+ * reservation. But we must always call this function when bits are claimed.
+ * Internally, the reservations code will use this information to mark the
+ * reservations bitmap. If resv is passed, it's next allocation window will be
+ * calculated. It also expects that 'cstart' is the same as we passed back
+ * from ocfs2_resmap_resv_bits().
+ */
+void ocfs2_resmap_claimed_bits(struct ocfs2_reservation_map *resmap,
+			       struct ocfs2_alloc_reservation *resv,
+			       u32 cstart, u32 clen);
+
+#endif	/* OCFS2_RESERVATIONS_H */

fs/ocfs2/resize.c

@@ -134,11 +134,7 @@ static int ocfs2_update_last_group_and_inode(handle_t *handle,
 		le16_add_cpu(&group->bg_free_bits_count, -1 * backups);
 	}
 
-	ret = ocfs2_journal_dirty(handle, group_bh);
-	if (ret < 0) {
-		mlog_errno(ret);
-		goto out_rollback;
-	}
+	ocfs2_journal_dirty(handle, group_bh);
 
 	/* update the inode accordingly. */
 	ret = ocfs2_journal_access_di(handle, INODE_CACHE(bm_inode), bm_bh,
@@ -319,7 +315,8 @@ int ocfs2_group_extend(struct inode * inode, int new_clusters)
 	BUG_ON(!OCFS2_IS_VALID_DINODE(fe));
 
 	if (le16_to_cpu(fe->id2.i_chain.cl_cpg) !=
-				 ocfs2_group_bitmap_size(osb->sb) * 8) {
+		ocfs2_group_bitmap_size(osb->sb, 0,
+					osb->s_feature_incompat) * 8) {
 		mlog(ML_ERROR, "The disk is too old and small. "
 		     "Force to do offline resize.");
 		ret = -EINVAL;
@@ -500,7 +497,8 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input)
 	fe = (struct ocfs2_dinode *)main_bm_bh->b_data;
 
 	if (le16_to_cpu(fe->id2.i_chain.cl_cpg) !=
-				 ocfs2_group_bitmap_size(osb->sb) * 8) {
+		ocfs2_group_bitmap_size(osb->sb, 0,
+					osb->s_feature_incompat) * 8) {
 		mlog(ML_ERROR, "The disk is too old and small."
 		     " Force to do offline resize.");
 		ret = -EINVAL;
@@ -545,12 +543,7 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input)
 
 	group = (struct ocfs2_group_desc *)group_bh->b_data;
 	group->bg_next_group = cr->c_blkno;
-
-	ret = ocfs2_journal_dirty(handle, group_bh);
-	if (ret < 0) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
+	ocfs2_journal_dirty(handle, group_bh);
 
 	ret = ocfs2_journal_access_di(handle, INODE_CACHE(main_bm_inode),
 				      main_bm_bh, OCFS2_JOURNAL_ACCESS_WRITE);

fs/ocfs2/suballoc.h

@@ -26,13 +26,14 @@
 #ifndef _CHAINALLOC_H_
 #define _CHAINALLOC_H_
 
+struct ocfs2_suballoc_result;
 typedef int (group_search_t)(struct inode *,
 			     struct buffer_head *,
 			     u32,			/* bits_wanted */
 			     u32,			/* min_bits */
 			     u64,			/* max_block */
-			     u16 *,			/* *bit_off */
-			     u16 *);			/* *bits_found */
+			     struct ocfs2_suballoc_result *);
+							/* found bits */
 
 struct ocfs2_alloc_context {
 	struct inode *ac_inode;    /* which bitmap are we allocating from? */
@@ -54,6 +55,8 @@ struct ocfs2_alloc_context {
 	u64    ac_last_group;
 	u64    ac_max_block;  /* Highest block number to allocate. 0 is
 				 is the same as ~0 - unlimited */
+
+	struct ocfs2_alloc_reservation	*ac_resv;
 };
 
 void ocfs2_init_steal_slots(struct ocfs2_super *osb);
@@ -80,22 +83,21 @@ int ocfs2_reserve_clusters(struct ocfs2_super *osb,
 			   u32 bits_wanted,
 			   struct ocfs2_alloc_context **ac);
 
-int ocfs2_claim_metadata(struct ocfs2_super *osb,
-			 handle_t *handle,
+int ocfs2_claim_metadata(handle_t *handle,
 			 struct ocfs2_alloc_context *ac,
 			 u32 bits_wanted,
+			 u64 *suballoc_loc,
 			 u16 *suballoc_bit_start,
 			 u32 *num_bits,
 			 u64 *blkno_start);
-int ocfs2_claim_new_inode(struct ocfs2_super *osb,
-			  handle_t *handle,
+int ocfs2_claim_new_inode(handle_t *handle,
 			  struct inode *dir,
 			  struct buffer_head *parent_fe_bh,
 			  struct ocfs2_alloc_context *ac,
+			  u64 *suballoc_loc,
 			  u16 *suballoc_bit,
 			  u64 *fe_blkno);
-int ocfs2_claim_clusters(struct ocfs2_super *osb,
-			 handle_t *handle,
+int ocfs2_claim_clusters(handle_t *handle,
 			 struct ocfs2_alloc_context *ac,
 			 u32 min_clusters,
 			 u32 *cluster_start,
@@ -104,8 +106,7 @@ int ocfs2_claim_clusters(struct ocfs2_super *osb,
  * Use this variant of ocfs2_claim_clusters to specify a maxiumum
  * number of clusters smaller than the allocation reserved.
  */
-int __ocfs2_claim_clusters(struct ocfs2_super *osb,
-			   handle_t *handle,
+int __ocfs2_claim_clusters(handle_t *handle,
 			   struct ocfs2_alloc_context *ac,
 			   u32 min_clusters,
 			   u32 max_clusters,

fs/ocfs2/super.h

@@ -45,4 +45,11 @@ void __ocfs2_abort(struct super_block *sb,
 
 #define ocfs2_abort(sb, fmt, args...) __ocfs2_abort(sb, __PRETTY_FUNCTION__, fmt, ##args)
 
+/*
+ * Void signal blockers, because in-kernel sigprocmask() only fails
+ * when SIG_* is wrong.
+ */
+void ocfs2_block_signals(sigset_t *oldset);
+void ocfs2_unblock_signals(sigset_t *oldset);
+
 #endif /* OCFS2_SUPER_H */