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: (27 commits)
  ocfs2: Cache extent records
  ocfs2: Remember rw lock level during direct io
  ocfs2: Fix up i_blocks calculation to know about holes
  ocfs2: Fix extent lookup to return true size of holes
  ocfs2: Read from an unwritten extent returns zeros
  ocfs2: make room for unwritten extents flag
  ocfs2: Use own splice write actor
  ocfs2: Use do_sync_mapping_range() in ocfs2_zero_tail_for_truncate()
  [PATCH] Turn do_sync_file_range() into do_sync_mapping_range()
  ocfs2: zero tail of sparse files on truncate
  ocfs2: Teach ocfs2_get_block() about holes
  ocfs2: remove ocfs2_prepare_write() and ocfs2_commit_write()
  ocfs2: teach ocfs2_file_aio_write() about sparse files
  ocfs2: Turn off shared writeable mmap for local files systems with holes.
  ocfs2: abstract out allocation locking
  ocfs2: teach extend/truncate about sparse files
  ocfs2: temporarily remove extent map caching
  ocfs2: sparse b-tree support
  ocfs2: small cleanup of ocfs2_request_delete()
  ocfs2: remove unused code
  ...

fs/ocfs2/alloc.h

@@ -31,7 +31,8 @@ int ocfs2_insert_extent(struct ocfs2_super *osb,
 			handle_t *handle,
 			struct inode *inode,
 			struct buffer_head *fe_bh,
-			u64 blkno,
+			u32 cpos,
+			u64 start_blk,
 			u32 new_clusters,
 			struct ocfs2_alloc_context *meta_ac);
 int ocfs2_num_free_extents(struct ocfs2_super *osb,
@@ -70,6 +71,8 @@ struct ocfs2_truncate_context {
 	struct buffer_head *tc_last_eb_bh;
 };
 
+int ocfs2_zero_tail_for_truncate(struct inode *inode, handle_t *handle,
+				 u64 new_i_size);
 int ocfs2_prepare_truncate(struct ocfs2_super *osb,
 			   struct inode *inode,
 			   struct buffer_head *fe_bh,
@@ -79,4 +82,26 @@ int ocfs2_commit_truncate(struct ocfs2_super *osb,
 			  struct buffer_head *fe_bh,
 			  struct ocfs2_truncate_context *tc);
 
+int ocfs2_find_leaf(struct inode *inode, struct ocfs2_extent_list *root_el,
+		    u32 cpos, struct buffer_head **leaf_bh);
+
+/*
+ * Helper function to look at the # of clusters in an extent record.
+ */
+static inline unsigned int ocfs2_rec_clusters(struct ocfs2_extent_list *el,
+					      struct ocfs2_extent_rec *rec)
+{
+	/*
+	 * Cluster count in extent records is slightly different
+	 * between interior nodes and leaf nodes. This is to support
+	 * unwritten extents which need a flags field in leaf node
+	 * records, thus shrinking the available space for a clusters
+	 * field.
+	 */
+	if (el->l_tree_depth)
+		return le32_to_cpu(rec->e_int_clusters);
+	else
+		return le16_to_cpu(rec->e_leaf_clusters);
+}
+
 #endif /* OCFS2_ALLOC_H */

fs/ocfs2/aops.h

@@ -30,12 +30,83 @@ handle_t *ocfs2_start_walk_page_trans(struct inode *inode,
 							 unsigned from,
 							 unsigned to);
 
+int ocfs2_map_page_blocks(struct page *page, u64 *p_blkno,
+			  struct inode *inode, unsigned int from,
+			  unsigned int to, int new);
+
+int walk_page_buffers(	handle_t *handle,
+			struct buffer_head *head,
+			unsigned from,
+			unsigned to,
+			int *partial,
+			int (*fn)(	handle_t *handle,
+					struct buffer_head *bh));
+
+struct ocfs2_write_ctxt;
+typedef int (ocfs2_page_writer)(struct inode *, struct ocfs2_write_ctxt *,
+				u64 *, unsigned int *, unsigned int *);
+
+ssize_t ocfs2_buffered_write_cluster(struct file *file, loff_t pos,
+				     size_t count, ocfs2_page_writer *actor,
+				     void *priv);
+
+struct ocfs2_write_ctxt {
+	size_t				w_count;
+	loff_t				w_pos;
+	u32				w_cpos;
+	unsigned int			w_finished_copy;
+
+	/* This is true if page_size > cluster_size */
+	unsigned int			w_large_pages;
+
+	/* Filler callback and private data */
+	ocfs2_page_writer		*w_write_data_page;
+	void				*w_private;
+
+	/* Only valid for the filler callback */
+	struct page			*w_this_page;
+	unsigned int			w_this_page_new;
+};
+
+struct ocfs2_buffered_write_priv {
+	char				*b_src_buf;
+	const struct iovec		*b_cur_iov; /* Current iovec */
+	size_t				b_cur_off; /* Offset in the
+						    * current iovec */
+};
+int ocfs2_map_and_write_user_data(struct inode *inode,
+				  struct ocfs2_write_ctxt *wc,
+				  u64 *p_blkno,
+				  unsigned int *ret_from,
+				  unsigned int *ret_to);
+
+struct ocfs2_splice_write_priv {
+	struct splice_desc		*s_sd;
+	struct pipe_buffer		*s_buf;
+	struct pipe_inode_info		*s_pipe;
+	/* Neither offset value is ever larger than one page */
+	unsigned int			s_offset;
+	unsigned int			s_buf_offset;
+};
+int ocfs2_map_and_write_splice_data(struct inode *inode,
+				    struct ocfs2_write_ctxt *wc,
+				    u64 *p_blkno,
+				    unsigned int *ret_from,
+				    unsigned int *ret_to);
+
 /* all ocfs2_dio_end_io()'s fault */
 #define ocfs2_iocb_is_rw_locked(iocb) \
 	test_bit(0, (unsigned long *)&iocb->private)
-#define ocfs2_iocb_set_rw_locked(iocb) \
-	set_bit(0, (unsigned long *)&iocb->private)
+static inline void ocfs2_iocb_set_rw_locked(struct kiocb *iocb, int level)
+{
+	set_bit(0, (unsigned long *)&iocb->private);
+	if (level)
+		set_bit(1, (unsigned long *)&iocb->private);
+	else
+		clear_bit(1, (unsigned long *)&iocb->private);
+}
 #define ocfs2_iocb_clear_rw_locked(iocb) \
 	clear_bit(0, (unsigned long *)&iocb->private)
-
+#define ocfs2_iocb_rw_locked_level(iocb) \
+	test_bit(1, (unsigned long *)&iocb->private)
 #endif /* OCFS2_FILE_H */

fs/ocfs2/cluster/quorum.c

@@ -46,6 +46,7 @@
 #include <linux/kernel.h>
 #include <linux/slab.h>
 #include <linux/workqueue.h>
+#include <linux/reboot.h>
 
 #include "heartbeat.h"
 #include "nodemanager.h"
@@ -72,7 +73,9 @@ static void o2quo_fence_self(void)
 	/* panic spins with interrupts enabled.  with preempt
 	 * threads can still schedule, etc, etc */
 	o2hb_stop_all_regions();
-	panic("ocfs2 is very sorry to be fencing this system by panicing\n");
+
+	printk("ocfs2 is very sorry to be fencing this system by restarting\n");
+	emergency_restart();
 }
 
 /* Indicate that a timeout occured on a hearbeat region write. The

fs/ocfs2/cluster/tcp_internal.h

@@ -38,6 +38,9 @@
  * locking semantics of the file system using the protocol.  It should 
  * be somewhere else, I'm sure, but right now it isn't.
  *
+ * New in version 8:
+ * 	- Replace delete inode votes with a cluster lock
+ *
  * New in version 7:
  * 	- DLM join domain includes the live nodemap
  *
@@ -57,7 +60,7 @@
  * 	- full 64 bit i_size in the metadata lock lvbs
  * 	- introduction of "rw" lock and pushing meta/data locking down
  */
-#define O2NET_PROTOCOL_VERSION 7ULL
+#define O2NET_PROTOCOL_VERSION 8ULL
 struct o2net_handshake {
 	__be64	protocol_version;
 	__be64	connector_id;

fs/ocfs2/dir.c

@@ -358,15 +358,17 @@ int ocfs2_do_extend_dir(struct super_block *sb,
 {
 	int status;
 	int extend;
-	u64 p_blkno;
+	u64 p_blkno, v_blkno;
 
 	spin_lock(&OCFS2_I(dir)->ip_lock);
 	extend = (i_size_read(dir) == ocfs2_clusters_to_bytes(sb, OCFS2_I(dir)->ip_clusters));
 	spin_unlock(&OCFS2_I(dir)->ip_lock);
 
 	if (extend) {
-		status = ocfs2_do_extend_allocation(OCFS2_SB(sb), dir, 1,
-						    parent_fe_bh, handle,
+		u32 offset = OCFS2_I(dir)->ip_clusters;
+
+		status = ocfs2_do_extend_allocation(OCFS2_SB(sb), dir, &offset,
+						    1, parent_fe_bh, handle,
 						    data_ac, meta_ac, NULL);
 		BUG_ON(status == -EAGAIN);
 		if (status < 0) {
@@ -375,9 +377,8 @@ int ocfs2_do_extend_dir(struct super_block *sb,
 		}
 	}
 
-	status = ocfs2_extent_map_get_blocks(dir, (dir->i_blocks >>
-						   (sb->s_blocksize_bits - 9)),
-					     1, &p_blkno, NULL);
+	v_blkno = ocfs2_blocks_for_bytes(sb, i_size_read(dir));
+	status = ocfs2_extent_map_get_blocks(dir, v_blkno, &p_blkno, NULL, NULL);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -486,7 +487,7 @@ static int ocfs2_extend_dir(struct ocfs2_super *osb,
 
 	dir_i_size += dir->i_sb->s_blocksize;
 	i_size_write(dir, dir_i_size);
-	dir->i_blocks = ocfs2_align_bytes_to_sectors(dir_i_size);
+	dir->i_blocks = ocfs2_inode_sector_count(dir);
 	status = ocfs2_mark_inode_dirty(handle, dir, parent_fe_bh);
 	if (status < 0) {
 		mlog_errno(status);

fs/ocfs2/dlm/dlmdomain.c

@@ -430,11 +430,10 @@ redo_bucket:
 
 			dlm_lockres_put(res);
 
-			cond_resched_lock(&dlm->spinlock);
-
 			if (dropped)
 				goto redo_bucket;
 		}
+		cond_resched_lock(&dlm->spinlock);
 		num += n;
 		mlog(0, "%s: touched %d lockreses in bucket %d "
 		     "(tot=%d)\n", dlm->name, n, i, num);
@@ -1035,7 +1034,7 @@ static int dlm_try_to_join_domain(struct dlm_ctxt *dlm)
 {
 	int status = 0, tmpstat, node;
 	struct domain_join_ctxt *ctxt;
-	enum dlm_query_join_response response;
+	enum dlm_query_join_response response = JOIN_DISALLOW;
 
 	mlog_entry("%p", dlm);
 

fs/ocfs2/dlm/dlmrecovery.c

@@ -611,6 +611,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
 			}
 		} while (status != 0);
 
+		spin_lock(&dlm_reco_state_lock);
 		switch (ndata->state) {
 			case DLM_RECO_NODE_DATA_INIT:
 			case DLM_RECO_NODE_DATA_FINALIZE_SENT:
@@ -641,6 +642,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
 				     ndata->node_num, dead_node);
 				break;
 		}
+		spin_unlock(&dlm_reco_state_lock);
 	}
 
 	mlog(0, "done requesting all lock info\n");

fs/ocfs2/dlmglue.c

@@ -225,11 +225,17 @@ static struct ocfs2_lock_res_ops ocfs2_dentry_lops = {
 	.flags		= 0,
 };
 
+static struct ocfs2_lock_res_ops ocfs2_inode_open_lops = {
+	.get_osb	= ocfs2_get_inode_osb,
+	.flags		= 0,
+};
+
 static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres)
 {
 	return lockres->l_type == OCFS2_LOCK_TYPE_META ||
 		lockres->l_type == OCFS2_LOCK_TYPE_DATA ||
-		lockres->l_type == OCFS2_LOCK_TYPE_RW;
+		lockres->l_type == OCFS2_LOCK_TYPE_RW ||
+		lockres->l_type == OCFS2_LOCK_TYPE_OPEN;
 }
 
 static inline struct inode *ocfs2_lock_res_inode(struct ocfs2_lock_res *lockres)
@@ -373,6 +379,9 @@ void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res,
 		case OCFS2_LOCK_TYPE_DATA:
 			ops = &ocfs2_inode_data_lops;
 			break;
+		case OCFS2_LOCK_TYPE_OPEN:
+			ops = &ocfs2_inode_open_lops;
+			break;
 		default:
 			mlog_bug_on_msg(1, "type: %d\n", type);
 			ops = NULL; /* thanks, gcc */
@@ -1129,6 +1138,12 @@ int ocfs2_create_new_inode_locks(struct inode *inode)
 		goto bail;
 	}
 
+	ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_open_lockres, 0, 0);
+	if (ret) {
+		mlog_errno(ret);
+		goto bail;
+	}
+
 bail:
 	mlog_exit(ret);
 	return ret;
@@ -1182,6 +1197,99 @@ void ocfs2_rw_unlock(struct inode *inode, int write)
 	mlog_exit_void();
 }
 
+/*
+ * ocfs2_open_lock always get PR mode lock.
+ */
+int ocfs2_open_lock(struct inode *inode)
+{
+	int status = 0;
+	struct ocfs2_lock_res *lockres;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+
+	BUG_ON(!inode);
+
+	mlog_entry_void();
+
+	mlog(0, "inode %llu take PRMODE open lock\n",
+	     (unsigned long long)OCFS2_I(inode)->ip_blkno);
+
+	if (ocfs2_mount_local(osb))
+		goto out;
+
+	lockres = &OCFS2_I(inode)->ip_open_lockres;
+
+	status = ocfs2_cluster_lock(OCFS2_SB(inode->i_sb), lockres,
+				    LKM_PRMODE, 0, 0);
+	if (status < 0)
+		mlog_errno(status);
+
+out:
+	mlog_exit(status);
+	return status;
+}
+
+int ocfs2_try_open_lock(struct inode *inode, int write)
+{
+	int status = 0, level;
+	struct ocfs2_lock_res *lockres;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+
+	BUG_ON(!inode);
+
+	mlog_entry_void();
+
+	mlog(0, "inode %llu try to take %s open lock\n",
+	     (unsigned long long)OCFS2_I(inode)->ip_blkno,
+	     write ? "EXMODE" : "PRMODE");
+
+	if (ocfs2_mount_local(osb))
+		goto out;
+
+	lockres = &OCFS2_I(inode)->ip_open_lockres;
+
+	level = write ? LKM_EXMODE : LKM_PRMODE;
+
+	/*
+	 * The file system may already holding a PRMODE/EXMODE open lock.
+	 * Since we pass LKM_NOQUEUE, the request won't block waiting on
+	 * other nodes and the -EAGAIN will indicate to the caller that
+	 * this inode is still in use.
+	 */
+	status = ocfs2_cluster_lock(OCFS2_SB(inode->i_sb), lockres,
+				    level, LKM_NOQUEUE, 0);
+
+out:
+	mlog_exit(status);
+	return status;
+}
+
+/*
+ * ocfs2_open_unlock unlock PR and EX mode open locks.
+ */
+void ocfs2_open_unlock(struct inode *inode)
+{
+	struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_open_lockres;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+
+	mlog_entry_void();
+
+	mlog(0, "inode %llu drop open lock\n",
+	     (unsigned long long)OCFS2_I(inode)->ip_blkno);
+
+	if (ocfs2_mount_local(osb))
+		goto out;
+
+	if(lockres->l_ro_holders)
+		ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres,
+				     LKM_PRMODE);
+	if(lockres->l_ex_holders)
+		ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres,
+				     LKM_EXMODE);
+
+out:
+	mlog_exit_void();
+}
+
 int ocfs2_data_lock_full(struct inode *inode,
 			 int write,
 			 int arg_flags)
@@ -1387,8 +1495,7 @@ static void ocfs2_refresh_inode_from_lvb(struct inode *inode)
 	if (S_ISLNK(inode->i_mode) && !oi->ip_clusters)
 		inode->i_blocks = 0;
 	else
-		inode->i_blocks =
-			ocfs2_align_bytes_to_sectors(i_size_read(inode));
+		inode->i_blocks = ocfs2_inode_sector_count(inode);
 
 	inode->i_uid     = be32_to_cpu(lvb->lvb_iuid);
 	inode->i_gid     = be32_to_cpu(lvb->lvb_igid);
@@ -1479,12 +1586,15 @@ static int ocfs2_meta_lock_update(struct inode *inode,
 {
 	int status = 0;
 	struct ocfs2_inode_info *oi = OCFS2_I(inode);
-	struct ocfs2_lock_res *lockres = NULL;
+	struct ocfs2_lock_res *lockres = &oi->ip_meta_lockres;
 	struct ocfs2_dinode *fe;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 
 	mlog_entry_void();
 
+	if (ocfs2_mount_local(osb))
+		goto bail;
+
 	spin_lock(&oi->ip_lock);
 	if (oi->ip_flags & OCFS2_INODE_DELETED) {
 		mlog(0, "Orphaned inode %llu was deleted while we "
@@ -1496,22 +1606,16 @@ static int ocfs2_meta_lock_update(struct inode *inode,
 	}
 	spin_unlock(&oi->ip_lock);
 
-	if (!ocfs2_mount_local(osb)) {
-		lockres = &oi->ip_meta_lockres;
-
-		if (!ocfs2_should_refresh_lock_res(lockres))
-			goto bail;
-	}
+	if (!ocfs2_should_refresh_lock_res(lockres))
+		goto bail;
 
 	/* This will discard any caching information we might have had
 	 * for the inode metadata. */
 	ocfs2_metadata_cache_purge(inode);
 
-	/* will do nothing for inode types that don't use the extent
-	 * map (directories, bitmap files, etc) */
 	ocfs2_extent_map_trunc(inode, 0);
 
-	if (lockres && ocfs2_meta_lvb_is_trustable(inode, lockres)) {
+	if (ocfs2_meta_lvb_is_trustable(inode, lockres)) {
 		mlog(0, "Trusting LVB on inode %llu\n",
 		     (unsigned long long)oi->ip_blkno);
 		ocfs2_refresh_inode_from_lvb(inode);
@@ -1558,8 +1662,7 @@ static int ocfs2_meta_lock_update(struct inode *inode,
 
 	status = 0;
 bail_refresh:
-	if (lockres)
-		ocfs2_complete_lock_res_refresh(lockres, status);
+	ocfs2_complete_lock_res_refresh(lockres, status);
 bail:
 	mlog_exit(status);
 	return status;
@@ -1630,7 +1733,6 @@ int ocfs2_meta_lock_full(struct inode *inode,
 		wait_event(osb->recovery_event,
 			   ocfs2_node_map_is_empty(osb, &osb->recovery_map));
 
-	acquired = 0;
 	lockres = &OCFS2_I(inode)->ip_meta_lockres;
 	level = ex ? LKM_EXMODE : LKM_PRMODE;
 	dlm_flags = 0;
@@ -2458,12 +2560,19 @@ int ocfs2_drop_inode_locks(struct inode *inode)
 	 * ocfs2_clear_inode has done it for us. */
 
 	err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
-			      &OCFS2_I(inode)->ip_data_lockres);
+			      &OCFS2_I(inode)->ip_open_lockres);
 	if (err < 0)
 		mlog_errno(err);
 
 	status = err;
 
+	err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
+			      &OCFS2_I(inode)->ip_data_lockres);
+	if (err < 0)
+		mlog_errno(err);
+	if (err < 0 && !status)
+		status = err;
+
 	err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
 			      &OCFS2_I(inode)->ip_meta_lockres);
 	if (err < 0)

fs/ocfs2/dlmglue.h

@@ -80,6 +80,9 @@ void ocfs2_data_unlock(struct inode *inode,
 		       int write);
 int ocfs2_rw_lock(struct inode *inode, int write);
 void ocfs2_rw_unlock(struct inode *inode, int write);
+int ocfs2_open_lock(struct inode *inode);
+int ocfs2_try_open_lock(struct inode *inode, int write);
+void ocfs2_open_unlock(struct inode *inode);
 int ocfs2_meta_lock_atime(struct inode *inode,
 			  struct vfsmount *vfsmnt,
 			  int *level);

fs/ocfs2/extent_map.h

@@ -25,22 +25,29 @@
 #ifndef _EXTENT_MAP_H
 #define _EXTENT_MAP_H
 
-int init_ocfs2_extent_maps(void);
-void exit_ocfs2_extent_maps(void);
+struct ocfs2_extent_map_item {
+	unsigned int			ei_cpos;
+	unsigned int			ei_phys;
+	unsigned int			ei_clusters;
+	unsigned int			ei_flags;
 
-/*
- * EVERY CALL here except _init, _trunc, and _drop expects alloc_sem
- * to be held.  The allocation cannot change at all while the map is
- * in the process of being updated.
- */
-int ocfs2_extent_map_init(struct inode *inode);
-int ocfs2_extent_map_append(struct inode *inode,
-			    struct ocfs2_extent_rec *rec,
-			    u32 new_clusters);
-int ocfs2_extent_map_get_blocks(struct inode *inode,
-				u64 v_blkno, int count,
-				u64 *p_blkno, int *ret_count);
-int ocfs2_extent_map_drop(struct inode *inode, u32 new_clusters);
-int ocfs2_extent_map_trunc(struct inode *inode, u32 new_clusters);
+	struct list_head		ei_list;
+};
+
+#define OCFS2_MAX_EXTENT_MAP_ITEMS			3
+struct ocfs2_extent_map {
+	unsigned int			em_num_items;
+	struct list_head		em_list;
+};
+
+void ocfs2_extent_map_init(struct inode *inode);
+void ocfs2_extent_map_trunc(struct inode *inode, unsigned int cluster);
+void ocfs2_extent_map_insert_rec(struct inode *inode,
+				 struct ocfs2_extent_rec *rec);
+
+int ocfs2_get_clusters(struct inode *inode, u32 v_cluster, u32 *p_cluster,
+		       u32 *num_clusters, unsigned int *extent_flags);
+int ocfs2_extent_map_get_blocks(struct inode *inode, u64 v_blkno, u64 *p_blkno,
+				u64 *ret_count, unsigned int *extent_flags);
 
 #endif  /* _EXTENT_MAP_H */