Merge tag 'xfs-4.15-merge-1' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux

Pull xfs updates from Darrick Wong:
 "xfs: great scads of new stuff for 4.15.

  This merge cycle, we're making some substantive changes to XFS. The
  in-core extent mappings have been refactored to use proper iterators
  and a btree to handle heavily fragmented files without needing
  high-order memory allocations; some important log recovery bug fixes;
  and the first part of the online fsck functionality.

  (The online fsck feature is disabled by default and more pieces of it
  will be coming in future release cycles.)

  This giant pile of patches has been run through a full xfstests run
  over the weekend and through a quick xfstests run against this
  morning's master, with no major failures reported.

  New in this version:

   - Refactor the incore extent map manipulations to use a cursor
     instead of directly modifying extent data.

   - Refactor the incore extent map cursor to use an in-memory btree
     instead of a single high-order allocation. This eliminates a major
     source of complaints about insufficient memory when opening a
     heavily fragmented file into a system whose memory is also heavily
     fragmented.

   - Fix a longstanding bug where deleting a file with a complex
     extended attribute btree incorrectly handled memory pointers, which
     could lead to memory corruption.

   - Improve metadata validation to eliminate crashing problems found
     while fuzzing xfs.

   - Move the error injection tag definitions into libxfs to be shared
     with userspace components.

   - Fix some log recovery bugs where we'd underflow log block position
     vector and incorrectly fail log recovery.

   - Drain the buffer lru after log recovery to force recovered buffers
     back through the verifiers after mount. On a v4 filesystem the log
     never attaches verifiers during log replay (v5 does), so we could
     end up with buffers marked verified but without having ever been
     verified.

   - Fix various other bugs.

   - Introduce the first part of a new online fsck tool. The new fsck
     tool will be able to iterate every piece of metadata in the
     filesystem to look for obvious errors and corruptions. In the next
     release cycle the checking will be extended to cross-reference with
     the other fs metadata, so this feature should only be used by the
     developers in the mean time"

* tag 'xfs-4.15-merge-1' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (131 commits)
  xfs: on failed mount, force-reclaim inodes after unmounting quota controls
  xfs: check the uniqueness of the AGFL entries
  xfs: remove u_int* type usage
  xfs: handle zero entries case in xfs_iext_rebalance_leaf
  xfs: add comments documenting the rebalance algorithm
  xfs: trivial indentation fixup for xfs_iext_remove_node
  xfs: remove a superflous assignment in xfs_iext_remove_node
  xfs: add some comments to xfs_iext_insert/xfs_iext_insert_node
  xfs: fix number of records handling in xfs_iext_split_leaf
  fs/xfs: Remove NULL check before kmem_cache_destroy
  xfs: only check da node header padding on v5 filesystems
  xfs: fix btree scrub deref check
  xfs: fix uninitialized return values in scrub code
  xfs: pass inode number to xfs_scrub_ino_set_{preen,warning}
  xfs: refactor the directory data block bestfree checks
  xfs: mark xlog_verify_dest_ptr STATIC
  xfs: mark xlog_recover_check_summary STATIC
  xfs: mark xfs_btree_check_lblock and xfs_btree_check_ptr static
  xfs: remove unreachable error injection code in xfs_qm_dqget
  xfs: remove unused debug counts for xfs_lock_inodes
  ...

1 files changed, 815 insertions, 1246 deletions

diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index 89263797cf32..08df809e2315 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -38,6 +38,7 @@
 #include "xfs_bmap_util.h"
 #include "xfs_bmap_btree.h"
 #include "xfs_rtalloc.h"
+#include "xfs_errortag.h"
 #include "xfs_error.h"
 #include "xfs_quota.h"
 #include "xfs_trans_space.h"
@@ -112,28 +113,21 @@ xfs_bmap_compute_maxlevels(
 STATIC int				/* error */
 xfs_bmbt_lookup_eq(
 	struct xfs_btree_cur	*cur,
-	xfs_fileoff_t		off,
-	xfs_fsblock_t		bno,
-	xfs_filblks_t		len,
+	struct xfs_bmbt_irec	*irec,
 	int			*stat)	/* success/failure */
 {
-	cur->bc_rec.b.br_startoff = off;
-	cur->bc_rec.b.br_startblock = bno;
-	cur->bc_rec.b.br_blockcount = len;
+	cur->bc_rec.b = *irec;
 	return xfs_btree_lookup(cur, XFS_LOOKUP_EQ, stat);
 }
 
 STATIC int				/* error */
-xfs_bmbt_lookup_ge(
+xfs_bmbt_lookup_first(
 	struct xfs_btree_cur	*cur,
-	xfs_fileoff_t		off,
-	xfs_fsblock_t		bno,
-	xfs_filblks_t		len,
 	int			*stat)	/* success/failure */
 {
-	cur->bc_rec.b.br_startoff = off;
-	cur->bc_rec.b.br_startblock = bno;
-	cur->bc_rec.b.br_blockcount = len;
+	cur->bc_rec.b.br_startoff = 0;
+	cur->bc_rec.b.br_startblock = 0;
+	cur->bc_rec.b.br_blockcount = 0;
 	return xfs_btree_lookup(cur, XFS_LOOKUP_GE, stat);
 }
 
@@ -160,21 +154,17 @@ static inline bool xfs_bmap_wants_extents(struct xfs_inode *ip, int whichfork)
 }
 
 /*
- * Update the record referred to by cur to the value given
- * by [off, bno, len, state].
+ * Update the record referred to by cur to the value given by irec
  * This either works (return 0) or gets an EFSCORRUPTED error.
  */
 STATIC int
 xfs_bmbt_update(
 	struct xfs_btree_cur	*cur,
-	xfs_fileoff_t		off,
-	xfs_fsblock_t		bno,
-	xfs_filblks_t		len,
-	xfs_exntst_t		state)
+	struct xfs_bmbt_irec	*irec)
 {
 	union xfs_btree_rec	rec;
 
-	xfs_bmbt_disk_set_allf(&rec.bmbt, off, bno, len, state);
+	xfs_bmbt_disk_set_all(&rec.bmbt, irec);
 	return xfs_btree_update(cur, &rec);
 }
 
@@ -242,7 +232,6 @@ xfs_bmap_forkoff_reset(
 {
 	if (whichfork == XFS_ATTR_FORK &&
 	    ip->i_d.di_format != XFS_DINODE_FMT_DEV &&
-	    ip->i_d.di_format != XFS_DINODE_FMT_UUID &&
 	    ip->i_d.di_format != XFS_DINODE_FMT_BTREE) {
 		uint	dfl_forkoff = xfs_default_attroffset(ip) >> 3;
 
@@ -499,31 +488,6 @@ error_norelse:
 }
 
 /*
- * Add bmap trace insert entries for all the contents of the extent records.
- */
-void
-xfs_bmap_trace_exlist(
-	xfs_inode_t	*ip,		/* incore inode pointer */
-	xfs_extnum_t	cnt,		/* count of entries in the list */
-	int		whichfork,	/* data or attr or cow fork */
-	unsigned long	caller_ip)
-{
-	xfs_extnum_t	idx;		/* extent record index */
-	xfs_ifork_t	*ifp;		/* inode fork pointer */
-	int		state = 0;
-
-	if (whichfork == XFS_ATTR_FORK)
-		state |= BMAP_ATTRFORK;
-	else if (whichfork == XFS_COW_FORK)
-		state |= BMAP_COWFORK;
-
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	ASSERT(cnt == xfs_iext_count(ifp));
-	for (idx = 0; idx < cnt; idx++)
-		trace_xfs_extlist(ip, idx, state, caller_ip);
-}
-
-/*
  * Validate that the bmbt_irecs being returned from bmapi are valid
  * given the caller's original parameters.  Specifically check the
  * ranges of the returned irecs to ensure that they only extend beyond
@@ -657,8 +621,8 @@ xfs_bmap_btree_to_extents(
 	cbno = be64_to_cpu(*pp);
 	*logflagsp = 0;
 #ifdef DEBUG
-	if ((error = xfs_btree_check_lptr(cur, cbno, 1)))
-		return error;
+	XFS_WANT_CORRUPTED_RETURN(cur->bc_mp,
+			xfs_btree_check_lptr(cur, cbno, 1));
 #endif
 	error = xfs_btree_read_bufl(mp, tp, cbno, 0, &cbp, XFS_BMAP_BTREE_REF,
 				&xfs_bmbt_buf_ops);
@@ -703,14 +667,14 @@ xfs_bmap_extents_to_btree(
 	xfs_bmbt_rec_t		*arp;		/* child record pointer */
 	struct xfs_btree_block	*block;		/* btree root block */
 	xfs_btree_cur_t		*cur;		/* bmap btree cursor */
-	xfs_bmbt_rec_host_t	*ep;		/* extent record pointer */
 	int			error;		/* error return value */
-	xfs_extnum_t		i, cnt;		/* extent record index */
 	xfs_ifork_t		*ifp;		/* inode fork pointer */
 	xfs_bmbt_key_t		*kp;		/* root block key pointer */
 	xfs_mount_t		*mp;		/* mount structure */
-	xfs_extnum_t		nextents;	/* number of file extents */
 	xfs_bmbt_ptr_t		*pp;		/* root block address pointer */
+	struct xfs_iext_cursor	icur;
+	struct xfs_bmbt_irec	rec;
+	xfs_extnum_t		cnt = 0;
 
 	mp = ip->i_mount;
 	ASSERT(whichfork != XFS_COW_FORK);
@@ -789,15 +753,12 @@ xfs_bmap_extents_to_btree(
 				XFS_BTNUM_BMAP, 0, 0, ip->i_ino,
 				XFS_BTREE_LONG_PTRS);
 
-	arp = XFS_BMBT_REC_ADDR(mp, ablock, 1);
-	nextents =  xfs_iext_count(ifp);
-	for (cnt = i = 0; i < nextents; i++) {
-		ep = xfs_iext_get_ext(ifp, i);
-		if (!isnullstartblock(xfs_bmbt_get_startblock(ep))) {
-			arp->l0 = cpu_to_be64(ep->l0);
-			arp->l1 = cpu_to_be64(ep->l1);
-			arp++; cnt++;
-		}
+	for_each_xfs_iext(ifp, &icur, &rec) {
+		if (isnullstartblock(rec.br_startblock))
+			continue;
+		arp = XFS_BMBT_REC_ADDR(mp, ablock, 1 + cnt);
+		xfs_bmbt_disk_set_all(arp, &rec);
+		cnt++;
 	}
 	ASSERT(cnt == XFS_IFORK_NEXTENTS(ip, whichfork));
 	xfs_btree_set_numrecs(ablock, cnt);
@@ -845,6 +806,8 @@ xfs_bmap_local_to_extents_empty(
 	xfs_bmap_forkoff_reset(ip, whichfork);
 	ifp->if_flags &= ~XFS_IFINLINE;
 	ifp->if_flags |= XFS_IFEXTENTS;
+	ifp->if_u1.if_root = NULL;
+	ifp->if_height = 0;
 	XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
 }
 
@@ -868,6 +831,7 @@ xfs_bmap_local_to_extents(
 	xfs_alloc_arg_t	args;		/* allocation arguments */
 	xfs_buf_t	*bp;		/* buffer for extent block */
 	struct xfs_bmbt_irec rec;
+	struct xfs_iext_cursor icur;
 
 	/*
 	 * We don't want to deal with the case of keeping inode data inline yet.
@@ -885,8 +849,7 @@ xfs_bmap_local_to_extents(
 
 	flags = 0;
 	error = 0;
-	ASSERT((ifp->if_flags & (XFS_IFINLINE|XFS_IFEXTENTS|XFS_IFEXTIREC)) ==
-								XFS_IFINLINE);
+	ASSERT((ifp->if_flags & (XFS_IFINLINE|XFS_IFEXTENTS)) == XFS_IFINLINE);
 	memset(&args, 0, sizeof(args));
 	args.tp = tp;
 	args.mp = ip->i_mount;
@@ -930,15 +893,16 @@ xfs_bmap_local_to_extents(
 	xfs_bmap_local_to_extents_empty(ip, whichfork);
 	flags |= XFS_ILOG_CORE;
 
+	ifp->if_u1.if_root = NULL;
+	ifp->if_height = 0;
+
 	rec.br_startoff = 0;
 	rec.br_startblock = args.fsbno;
 	rec.br_blockcount = 1;
 	rec.br_state = XFS_EXT_NORM;
-	xfs_iext_insert(ip, 0, 1, &rec, 0);
+	xfs_iext_first(ifp, &icur);
+	xfs_iext_insert(ip, &icur, &rec, 0);
 
-	trace_xfs_bmap_post_update(ip, 0,
-			whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0,
-			_THIS_IP_);
 	XFS_IFORK_NEXT_SET(ip, whichfork, 1);
 	ip->i_d.di_nblocks = 1;
 	xfs_trans_mod_dquot_byino(tp, ip,
@@ -973,7 +937,8 @@ xfs_bmap_add_attrfork_btree(
 		cur = xfs_bmbt_init_cursor(mp, tp, ip, XFS_DATA_FORK);
 		cur->bc_private.b.dfops = dfops;
 		cur->bc_private.b.firstblock = *firstblock;
-		if ((error = xfs_bmbt_lookup_ge(cur, 0, 0, 0, &stat)))
+		error = xfs_bmbt_lookup_first(cur, &stat);
+		if (error)
 			goto error0;
 		/* must be at least one entry */
 		XFS_WANT_CORRUPTED_GOTO(mp, stat == 1, error0);
@@ -1124,9 +1089,6 @@ xfs_bmap_add_attrfork(
 	case XFS_DINODE_FMT_DEV:
 		ip->i_d.di_forkoff = roundup(sizeof(xfs_dev_t), 8) >> 3;
 		break;
-	case XFS_DINODE_FMT_UUID:
-		ip->i_d.di_forkoff = roundup(sizeof(uuid_t), 8) >> 3;
-		break;
 	case XFS_DINODE_FMT_LOCAL:
 	case XFS_DINODE_FMT_EXTENTS:
 	case XFS_DINODE_FMT_BTREE:
@@ -1206,32 +1168,35 @@ trans_cancel:
  */
 
 /*
- * Read in the extents to if_extents.
- * All inode fields are set up by caller, we just traverse the btree
- * and copy the records in. If the file system cannot contain unwritten
- * extents, the records are checked for no "state" flags.
+ * Read in extents from a btree-format inode.
  */
-int					/* error */
-xfs_bmap_read_extents(
-	xfs_trans_t		*tp,	/* transaction pointer */
-	xfs_inode_t		*ip,	/* incore inode */
-	int			whichfork) /* data or attr fork */
+int
+xfs_iread_extents(
+	struct xfs_trans	*tp,
+	struct xfs_inode	*ip,
+	int			whichfork)
 {
-	struct xfs_btree_block	*block;	/* current btree block */
-	xfs_fsblock_t		bno;	/* block # of "block" */
-	xfs_buf_t		*bp;	/* buffer for "block" */
-	int			error;	/* error return value */
-	xfs_extnum_t		i, j;	/* index into the extents list */
-	xfs_ifork_t		*ifp;	/* fork structure */
-	int			level;	/* btree level, for checking */
-	xfs_mount_t		*mp;	/* file system mount structure */
-	__be64			*pp;	/* pointer to block address */
-	/* REFERENCED */
-	xfs_extnum_t		room;	/* number of entries there's room for */
+	struct xfs_mount	*mp = ip->i_mount;
+	int			state = xfs_bmap_fork_to_state(whichfork);
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	xfs_extnum_t		nextents = XFS_IFORK_NEXTENTS(ip, whichfork);
+	struct xfs_btree_block	*block = ifp->if_broot;
+	struct xfs_iext_cursor	icur;
+	struct xfs_bmbt_irec	new;
+	xfs_fsblock_t		bno;
+	struct xfs_buf		*bp;
+	xfs_extnum_t		i, j;
+	int			level;
+	__be64			*pp;
+	int			error;
+
+	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
+
+	if (unlikely(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)) {
+		XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_LOW, mp);
+		return -EFSCORRUPTED;
+	}
 
-	mp = ip->i_mount;
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	block = ifp->if_broot;
 	/*
 	 * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
 	 */
@@ -1248,21 +1213,23 @@ xfs_bmap_read_extents(
 		error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
 				XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops);
 		if (error)
-			return error;
+			goto out;
 		block = XFS_BUF_TO_BLOCK(bp);
 		if (level == 0)
 			break;
 		pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
 		bno = be64_to_cpu(*pp);
 		XFS_WANT_CORRUPTED_GOTO(mp,
-			XFS_FSB_SANITY_CHECK(mp, bno), error0);
+			XFS_FSB_SANITY_CHECK(mp, bno), out_brelse);
 		xfs_trans_brelse(tp, bp);
 	}
+
 	/*
 	 * Here with bp and block set to the leftmost leaf node in the tree.
 	 */
-	room = xfs_iext_count(ifp);
 	i = 0;
+	xfs_iext_first(ifp, &icur);
+
 	/*
 	 * Loop over all leaf nodes.  Copy information to the extent records.
 	 */
@@ -1272,14 +1239,15 @@ xfs_bmap_read_extents(
 		xfs_extnum_t	num_recs;
 
 		num_recs = xfs_btree_get_numrecs(block);
-		if (unlikely(i + num_recs > room)) {
-			ASSERT(i + num_recs <= room);
+		if (unlikely(i + num_recs > nextents)) {
+			ASSERT(i + num_recs <= nextents);
 			xfs_warn(ip->i_mount,
 				"corrupt dinode %Lu, (btree extents).",
 				(unsigned long long) ip->i_ino);
-			XFS_CORRUPTION_ERROR("xfs_bmap_read_extents(1)",
+			XFS_CORRUPTION_ERROR(__func__,
 				XFS_ERRLEVEL_LOW, ip->i_mount, block);
-			goto error0;
+			error = -EFSCORRUPTED;
+			goto out_brelse;
 		}
 		/*
 		 * Read-ahead the next leaf block, if any.
@@ -1292,15 +1260,17 @@ xfs_bmap_read_extents(
 		 * Copy records into the extent records.
 		 */
 		frp = XFS_BMBT_REC_ADDR(mp, block, 1);
-		for (j = 0; j < num_recs; j++, i++, frp++) {
-			xfs_bmbt_rec_host_t *trp = xfs_iext_get_ext(ifp, i);
-			trp->l0 = be64_to_cpu(frp->l0);
-			trp->l1 = be64_to_cpu(frp->l1);
-			if (!xfs_bmbt_validate_extent(mp, whichfork, trp)) {
+		for (j = 0; j < num_recs; j++, frp++, i++) {
+			xfs_bmbt_disk_get_all(frp, &new);
+			if (!xfs_bmbt_validate_extent(mp, whichfork, &new)) {
 				XFS_ERROR_REPORT("xfs_bmap_read_extents(2)",
 						 XFS_ERRLEVEL_LOW, mp);
-				goto error0;
+				error = -EFSCORRUPTED;
+				goto out_brelse;
 			}
+			xfs_iext_insert(ip, &icur, &new, state);
+			trace_xfs_read_extent(ip, &icur, state, _THIS_IP_);
+			xfs_iext_next(ifp, &icur);
 		}
 		xfs_trans_brelse(tp, bp);
 		bno = nextbno;
@@ -1312,71 +1282,74 @@ xfs_bmap_read_extents(
 		error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
 				XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops);
 		if (error)
-			return error;
+			goto out;
 		block = XFS_BUF_TO_BLOCK(bp);
 	}
-	if (i != XFS_IFORK_NEXTENTS(ip, whichfork))
-		return -EFSCORRUPTED;
+
+	if (i != XFS_IFORK_NEXTENTS(ip, whichfork)) {
+		error = -EFSCORRUPTED;
+		goto out;
+	}
 	ASSERT(i == xfs_iext_count(ifp));
-	XFS_BMAP_TRACE_EXLIST(ip, i, whichfork);
+
+	ifp->if_flags |= XFS_IFEXTENTS;
 	return 0;
-error0:
+
+out_brelse:
 	xfs_trans_brelse(tp, bp);
-	return -EFSCORRUPTED;
+out:
+	xfs_iext_destroy(ifp);
+	return error;
 }
 
 /*
- * Returns the file-relative block number of the first unused block(s)
- * in the file with at least "len" logically contiguous blocks free.
- * This is the lowest-address hole if the file has holes, else the first block
- * past the end of file.
- * Return 0 if the file is currently local (in-inode).
+ * Returns the relative block number of the first unused block(s) in the given
+ * fork with at least "len" logically contiguous blocks free.  This is the
+ * lowest-address hole if the fork has holes, else the first block past the end
+ * of fork.  Return 0 if the fork is currently local (in-inode).
  */
 int						/* error */
 xfs_bmap_first_unused(
-	xfs_trans_t	*tp,			/* transaction pointer */
-	xfs_inode_t	*ip,			/* incore inode */
-	xfs_extlen_t	len,			/* size of hole to find */
-	xfs_fileoff_t	*first_unused,		/* unused block */
-	int		whichfork)		/* data or attr fork */
+	struct xfs_trans	*tp,		/* transaction pointer */
+	struct xfs_inode	*ip,		/* incore inode */
+	xfs_extlen_t		len,		/* size of hole to find */
+	xfs_fileoff_t		*first_unused,	/* unused block */
+	int			whichfork)	/* data or attr fork */
 {
-	int		error;			/* error return value */
-	int		idx;			/* extent record index */
-	xfs_ifork_t	*ifp;			/* inode fork pointer */
-	xfs_fileoff_t	lastaddr;		/* last block number seen */
-	xfs_fileoff_t	lowest;			/* lowest useful block */
-	xfs_fileoff_t	max;			/* starting useful block */
-	xfs_extnum_t	nextents;		/* number of extent entries */
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_bmbt_irec	got;
+	struct xfs_iext_cursor	icur;
+	xfs_fileoff_t		lastaddr = 0;
+	xfs_fileoff_t		lowest, max;
+	int			error;
 
 	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE ||
 	       XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS ||
 	       XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL);
+
 	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
 		*first_unused = 0;
 		return 0;
 	}
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
-	    (error = xfs_iread_extents(tp, ip, whichfork)))
-		return error;
-	lowest = *first_unused;
-	nextents = xfs_iext_count(ifp);
-	for (idx = 0, lastaddr = 0, max = lowest; idx < nextents; idx++) {
-		struct xfs_bmbt_irec got;
 
-		xfs_iext_get_extent(ifp, idx, &got);
+	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+		error = xfs_iread_extents(tp, ip, whichfork);
+		if (error)
+			return error;
+	}
 
+	lowest = max = *first_unused;
+	for_each_xfs_iext(ifp, &icur, &got) {
 		/*
 		 * See if the hole before this extent will work.
 		 */
 		if (got.br_startoff >= lowest + len &&
-		    got.br_startoff - max >= len) {
-			*first_unused = max;
-			return 0;
-		}
+		    got.br_startoff - max >= len)
+			break;
 		lastaddr = got.br_startoff + got.br_blockcount;
 		max = XFS_FILEOFF_MAX(lastaddr, lowest);
 	}
+
 	*first_unused = max;
 	return 0;
 }
@@ -1396,7 +1369,7 @@ xfs_bmap_last_before(
 {
 	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
 	struct xfs_bmbt_irec	got;
-	xfs_extnum_t		idx;
+	struct xfs_iext_cursor	icur;
 	int			error;
 
 	switch (XFS_IFORK_FORMAT(ip, whichfork)) {
@@ -1416,17 +1389,8 @@ xfs_bmap_last_before(
 			return error;
 	}
 
-	if (xfs_iext_lookup_extent(ip, ifp, *last_block - 1, &idx, &got)) {
-		if (got.br_startoff <= *last_block - 1)
-			return 0;
-	}
-
-	if (xfs_iext_get_extent(ifp, idx - 1, &got)) {
-		*last_block = got.br_startoff + got.br_blockcount;
-		return 0;
-	}
-
-	*last_block = 0;
+	if (!xfs_iext_lookup_extent_before(ip, ifp, last_block, &icur, &got))
+		*last_block = 0;
 	return 0;
 }
 
@@ -1439,8 +1403,8 @@ xfs_bmap_last_extent(
 	int			*is_empty)
 {
 	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_iext_cursor	icur;
 	int			error;
-	int			nextents;
 
 	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
 		error = xfs_iread_extents(tp, ip, whichfork);
@@ -1448,14 +1412,11 @@ xfs_bmap_last_extent(
 			return error;
 	}
 
-	nextents = xfs_iext_count(ifp);
-	if (nextents == 0) {
+	xfs_iext_last(ifp, &icur);
+	if (!xfs_iext_get_extent(ifp, &icur, rec))
 		*is_empty = 1;
-		return 0;
-	}
-
-	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, nextents - 1), rec);
-	*is_empty = 0;
+	else
+		*is_empty = 0;
 	return 0;
 }
 
@@ -1540,10 +1501,10 @@ xfs_bmap_one_block(
 	xfs_inode_t	*ip,		/* incore inode */
 	int		whichfork)	/* data or attr fork */
 {
-	xfs_bmbt_rec_host_t *ep;	/* ptr to fork's extent */
 	xfs_ifork_t	*ifp;		/* inode fork pointer */
 	int		rval;		/* return value */
 	xfs_bmbt_irec_t	s;		/* internal version of extent */
+	struct xfs_iext_cursor icur;
 
 #ifndef DEBUG
 	if (whichfork == XFS_DATA_FORK)
@@ -1555,8 +1516,8 @@ xfs_bmap_one_block(
 		return 0;
 	ifp = XFS_IFORK_PTR(ip, whichfork);
 	ASSERT(ifp->if_flags & XFS_IFEXTENTS);
-	ep = xfs_iext_get_ext(ifp, 0);
-	xfs_bmbt_get_all(ep, &s);
+	xfs_iext_first(ifp, &icur);
+	xfs_iext_get_extent(ifp, &icur, &s);
 	rval = s.br_startoff == 0 && s.br_blockcount == 1;
 	if (rval && whichfork == XFS_DATA_FORK)
 		ASSERT(XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize);
@@ -1576,8 +1537,6 @@ xfs_bmap_add_extent_delay_real(
 	int			whichfork)
 {
 	struct xfs_bmbt_irec	*new = &bma->got;
-	int			diff;	/* temp value */
-	xfs_bmbt_rec_host_t	*ep;	/* extent entry for idx */
 	int			error;	/* error return value */
 	int			i;	/* temp state */
 	xfs_ifork_t		*ifp;	/* inode fork pointer */
@@ -1585,14 +1544,14 @@ xfs_bmap_add_extent_delay_real(
 	xfs_bmbt_irec_t		r[3];	/* neighbor extent entries */
 					/* left is 0, right is 1, prev is 2 */
 	int			rval=0;	/* return value (logging flags) */
-	int			state = 0;/* state bits, accessed thru macros */
+	int			state = xfs_bmap_fork_to_state(whichfork);
 	xfs_filblks_t		da_new; /* new count del alloc blocks used */
 	xfs_filblks_t		da_old; /* old count del alloc blocks used */
 	xfs_filblks_t		temp=0;	/* value for da_new calculations */
-	xfs_filblks_t		temp2=0;/* value for da_new calculations */
 	int			tmp_rval;	/* partial logging flags */
 	struct xfs_mount	*mp;
 	xfs_extnum_t		*nextents;
+	struct xfs_bmbt_irec	old;
 
 	mp = bma->ip->i_mount;
 	ifp = XFS_IFORK_PTR(bma->ip, whichfork);
@@ -1600,8 +1559,6 @@ xfs_bmap_add_extent_delay_real(
 	nextents = (whichfork == XFS_COW_FORK ? &bma->ip->i_cnextents :
 						&bma->ip->i_d.di_nextents);
 
-	ASSERT(bma->idx >= 0);
-	ASSERT(bma->idx <= xfs_iext_count(ifp));
 	ASSERT(!isnullstartblock(new->br_startblock));
 	ASSERT(!bma->cur ||
 	       (bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
@@ -1612,15 +1569,12 @@ xfs_bmap_add_extent_delay_real(
 #define	RIGHT		r[1]
 #define	PREV		r[2]
 
-	if (whichfork == XFS_COW_FORK)
-		state |= BMAP_COWFORK;
-
 	/*
 	 * Set up a bunch of variables to make the tests simpler.
 	 */
-	ep = xfs_iext_get_ext(ifp, bma->idx);
-	xfs_bmbt_get_all(ep, &PREV);
+	xfs_iext_get_extent(ifp, &bma->icur, &PREV);
 	new_endoff = new->br_startoff + new->br_blockcount;
+	ASSERT(isnullstartblock(PREV.br_startblock));
 	ASSERT(PREV.br_startoff <= new->br_startoff);
 	ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);
 
@@ -1640,10 +1594,8 @@ xfs_bmap_add_extent_delay_real(
 	 * Check and set flags if this segment has a left neighbor.
 	 * Don't set contiguous if the combined extent would be too large.
 	 */
-	if (bma->idx > 0) {
+	if (xfs_iext_peek_prev_extent(ifp, &bma->icur, &LEFT)) {
 		state |= BMAP_LEFT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &LEFT);
-
 		if (isnullstartblock(LEFT.br_startblock))
 			state |= BMAP_LEFT_DELAY;
 	}
@@ -1660,10 +1612,8 @@ xfs_bmap_add_extent_delay_real(
 	 * Don't set contiguous if the combined extent would be too large.
 	 * Also check for all-three-contiguous being too large.
 	 */
-	if (bma->idx < xfs_iext_count(ifp) - 1) {
+	if (xfs_iext_peek_next_extent(ifp, &bma->icur, &RIGHT)) {
 		state |= BMAP_RIGHT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx + 1), &RIGHT);
-
 		if (isnullstartblock(RIGHT.br_startblock))
 			state |= BMAP_RIGHT_DELAY;
 	}
@@ -1693,22 +1643,19 @@ xfs_bmap_add_extent_delay_real(
 		 * Filling in all of a previously delayed allocation extent.
 		 * The left and right neighbors are both contiguous with new.
 		 */
-		bma->idx--;
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
-			LEFT.br_blockcount + PREV.br_blockcount +
-			RIGHT.br_blockcount);
-		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-
-		xfs_iext_remove(bma->ip, bma->idx + 1, 2, state);
+		LEFT.br_blockcount += PREV.br_blockcount + RIGHT.br_blockcount;
+
+		xfs_iext_remove(bma->ip, &bma->icur, state);
+		xfs_iext_remove(bma->ip, &bma->icur, state);
+		xfs_iext_prev(ifp, &bma->icur);
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &LEFT);
 		(*nextents)--;
+
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
-					RIGHT.br_startblock,
-					RIGHT.br_blockcount, &i);
+			error = xfs_bmbt_lookup_eq(bma->cur, &RIGHT, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
@@ -1720,11 +1667,7 @@ xfs_bmap_add_extent_delay_real(
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
-					LEFT.br_startblock,
-					LEFT.br_blockcount +
-					PREV.br_blockcount +
-					RIGHT.br_blockcount, LEFT.br_state);
+			error = xfs_bmbt_update(bma->cur, &LEFT);
 			if (error)
 				goto done;
 		}
@@ -1735,28 +1678,22 @@ xfs_bmap_add_extent_delay_real(
 		 * Filling in all of a previously delayed allocation extent.
 		 * The left neighbor is contiguous, the right is not.
 		 */
-		bma->idx--;
+		old = LEFT;
+		LEFT.br_blockcount += PREV.br_blockcount;
 
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
-			LEFT.br_blockcount + PREV.br_blockcount);
-		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+		xfs_iext_remove(bma->ip, &bma->icur, state);
+		xfs_iext_prev(ifp, &bma->icur);
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &LEFT);
 
-		xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
-					LEFT.br_startblock, LEFT.br_blockcount,
-					&i);
+			error = xfs_bmbt_lookup_eq(bma->cur, &old, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
-					LEFT.br_startblock,
-					LEFT.br_blockcount +
-					PREV.br_blockcount, LEFT.br_state);
+			error = xfs_bmbt_update(bma->cur, &LEFT);
 			if (error)
 				goto done;
 		}
@@ -1767,27 +1704,23 @@ xfs_bmap_add_extent_delay_real(
 		 * Filling in all of a previously delayed allocation extent.
 		 * The right neighbor is contiguous, the left is not.
 		 */
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
-		xfs_bmbt_set_startblock(ep, new->br_startblock);
-		xfs_bmbt_set_blockcount(ep,
-			PREV.br_blockcount + RIGHT.br_blockcount);
-		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+		PREV.br_startblock = new->br_startblock;
+		PREV.br_blockcount += RIGHT.br_blockcount;
+
+		xfs_iext_next(ifp, &bma->icur);
+		xfs_iext_remove(bma->ip, &bma->icur, state);
+		xfs_iext_prev(ifp, &bma->icur);
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &PREV);
 
-		xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
-					RIGHT.br_startblock,
-					RIGHT.br_blockcount, &i);
+			error = xfs_bmbt_lookup_eq(bma->cur, &RIGHT, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			error = xfs_bmbt_update(bma->cur, PREV.br_startoff,
-					new->br_startblock,
-					PREV.br_blockcount +
-					RIGHT.br_blockcount, PREV.br_state);
+			error = xfs_bmbt_update(bma->cur, &PREV);
 			if (error)
 				goto done;
 		}
@@ -1799,23 +1732,19 @@ xfs_bmap_add_extent_delay_real(
 		 * Neither the left nor right neighbors are contiguous with
 		 * the new one.
 		 */
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
-		xfs_bmbt_set_startblock(ep, new->br_startblock);
-		xfs_bmbt_set_state(ep, new->br_state);
-		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+		PREV.br_startblock = new->br_startblock;
+		PREV.br_state = new->br_state;
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &PREV);
 
 		(*nextents)++;
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
-					new->br_startblock, new->br_blockcount,
-					&i);
+			error = xfs_bmbt_lookup_eq(bma->cur, new, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
-			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
 			error = xfs_btree_insert(bma->cur, &i);
 			if (error)
 				goto done;
@@ -1828,40 +1757,33 @@ xfs_bmap_add_extent_delay_real(
 		 * Filling in the first part of a previous delayed allocation.
 		 * The left neighbor is contiguous.
 		 */
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx - 1),
-			LEFT.br_blockcount + new->br_blockcount);
-		xfs_bmbt_set_startoff(ep,
-			PREV.br_startoff + new->br_blockcount);
-		trace_xfs_bmap_post_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
-
+		old = LEFT;
 		temp = PREV.br_blockcount - new->br_blockcount;
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep, temp);
+		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
+				startblockval(PREV.br_startblock));
+
+		LEFT.br_blockcount += new->br_blockcount;
+
+		PREV.br_blockcount = temp;
+		PREV.br_startoff += new->br_blockcount;
+		PREV.br_startblock = nullstartblock(da_new);
+
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &PREV);
+		xfs_iext_prev(ifp, &bma->icur);
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &LEFT);
+
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
-					LEFT.br_startblock, LEFT.br_blockcount,
-					&i);
+			error = xfs_bmbt_lookup_eq(bma->cur, &old, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
-					LEFT.br_startblock,
-					LEFT.br_blockcount +
-					new->br_blockcount,
-					LEFT.br_state);
+			error = xfs_bmbt_update(bma->cur, &LEFT);
 			if (error)
 				goto done;
 		}
-		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
-			startblockval(PREV.br_startblock));
-		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
-		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-
-		bma->idx--;
 		break;
 
 	case BMAP_LEFT_FILLING:
@@ -1869,23 +1791,16 @@ xfs_bmap_add_extent_delay_real(
 		 * Filling in the first part of a previous delayed allocation.
 		 * The left neighbor is not contiguous.
 		 */
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
-		xfs_bmbt_set_startoff(ep, new_endoff);
-		temp = PREV.br_blockcount - new->br_blockcount;
-		xfs_bmbt_set_blockcount(ep, temp);
-		xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, new);
 		(*nextents)++;
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
-					new->br_startblock, new->br_blockcount,
-					&i);
+			error = xfs_bmbt_lookup_eq(bma->cur, new, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
-			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
 			error = xfs_btree_insert(bma->cur, &i);
 			if (error)
 				goto done;
@@ -1900,12 +1815,18 @@ xfs_bmap_add_extent_delay_real(
 			if (error)
 				goto done;
 		}
+
+		temp = PREV.br_blockcount - new->br_blockcount;
 		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
 			startblockval(PREV.br_startblock) -
 			(bma->cur ? bma->cur->bc_private.b.allocated : 0));
-		ep = xfs_iext_get_ext(ifp, bma->idx + 1);
-		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
-		trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
+
+		PREV.br_startoff = new_endoff;
+		PREV.br_blockcount = temp;
+		PREV.br_startblock = nullstartblock(da_new);
+		xfs_iext_next(ifp, &bma->icur);
+		xfs_iext_insert(bma->ip, &bma->icur, &PREV, state);
+		xfs_iext_prev(ifp, &bma->icur);
 		break;
 
 	case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
@@ -1913,40 +1834,34 @@ xfs_bmap_add_extent_delay_real(
 		 * Filling in the last part of a previous delayed allocation.
 		 * The right neighbor is contiguous with the new allocation.
 		 */
-		temp = PREV.br_blockcount - new->br_blockcount;
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep, temp);
-		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, bma->idx + 1),
-			new->br_startoff, new->br_startblock,
-			new->br_blockcount + RIGHT.br_blockcount,
-			RIGHT.br_state);
-		trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
+		old = RIGHT;
+		RIGHT.br_startoff = new->br_startoff;
+		RIGHT.br_startblock = new->br_startblock;
+		RIGHT.br_blockcount += new->br_blockcount;
+
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
-					RIGHT.br_startblock,
-					RIGHT.br_blockcount, &i);
+			error = xfs_bmbt_lookup_eq(bma->cur, &old, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			error = xfs_bmbt_update(bma->cur, new->br_startoff,
-					new->br_startblock,
-					new->br_blockcount +
-					RIGHT.br_blockcount,
-					RIGHT.br_state);
+			error = xfs_bmbt_update(bma->cur, &RIGHT);
 			if (error)
 				goto done;
 		}
 
+		temp = PREV.br_blockcount - new->br_blockcount;
 		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
 			startblockval(PREV.br_startblock));
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
-		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
-		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		bma->idx++;
+		PREV.br_blockcount = temp;
+		PREV.br_startblock = nullstartblock(da_new);
+
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &PREV);
+		xfs_iext_next(ifp, &bma->icur);
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &RIGHT);
 		break;
 
 	case BMAP_RIGHT_FILLING:
@@ -1954,22 +1869,16 @@ xfs_bmap_add_extent_delay_real(
 		 * Filling in the last part of a previous delayed allocation.
 		 * The right neighbor is not contiguous.
 		 */
-		temp = PREV.br_blockcount - new->br_blockcount;
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep, temp);
-		xfs_iext_insert(bma->ip, bma->idx + 1, 1, new, state);
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, new);
 		(*nextents)++;
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
-					new->br_startblock, new->br_blockcount,
-					&i);
+			error = xfs_bmbt_lookup_eq(bma->cur, new, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
-			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
 			error = xfs_btree_insert(bma->cur, &i);
 			if (error)
 				goto done;
@@ -1984,14 +1893,16 @@ xfs_bmap_add_extent_delay_real(
 			if (error)
 				goto done;
 		}
+
+		temp = PREV.br_blockcount - new->br_blockcount;
 		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
 			startblockval(PREV.br_startblock) -
 			(bma->cur ? bma->cur->bc_private.b.allocated : 0));
-		ep = xfs_iext_get_ext(ifp, bma->idx);
-		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
-		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		bma->idx++;
+		PREV.br_startblock = nullstartblock(da_new);
+		PREV.br_blockcount = temp;
+		xfs_iext_insert(bma->ip, &bma->icur, &PREV, state);
+		xfs_iext_next(ifp, &bma->icur);
 		break;
 
 	case 0:
@@ -2015,30 +1926,40 @@ xfs_bmap_add_extent_delay_real(
 		 *  PREV @ idx          LEFT              RIGHT
 		 *                      inserted at idx + 1
 		 */
-		temp = new->br_startoff - PREV.br_startoff;
-		temp2 = PREV.br_startoff + PREV.br_blockcount - new_endoff;
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx, 0, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep, temp);	/* truncate PREV */
+		old = PREV;
+
+		/* LEFT is the new middle */
 		LEFT = *new;
+
+		/* RIGHT is the new right */
 		RIGHT.br_state = PREV.br_state;
-		RIGHT.br_startblock = nullstartblock(
-				(int)xfs_bmap_worst_indlen(bma->ip, temp2));
 		RIGHT.br_startoff = new_endoff;
-		RIGHT.br_blockcount = temp2;
-		/* insert LEFT (r[0]) and RIGHT (r[1]) at the same time */
-		xfs_iext_insert(bma->ip, bma->idx + 1, 2, &LEFT, state);
+		RIGHT.br_blockcount =
+			PREV.br_startoff + PREV.br_blockcount - new_endoff;
+		RIGHT.br_startblock =
+			nullstartblock(xfs_bmap_worst_indlen(bma->ip,
+					RIGHT.br_blockcount));
+
+		/* truncate PREV */
+		PREV.br_blockcount = new->br_startoff - PREV.br_startoff;
+		PREV.br_startblock =
+			nullstartblock(xfs_bmap_worst_indlen(bma->ip,
+					PREV.br_blockcount));
+		xfs_iext_update_extent(bma->ip, state, &bma->icur, &PREV);
+
+		xfs_iext_next(ifp, &bma->icur);
+		xfs_iext_insert(bma->ip, &bma->icur, &RIGHT, state);
+		xfs_iext_insert(bma->ip, &bma->icur, &LEFT, state);
 		(*nextents)++;
+
 		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
-					new->br_startblock, new->br_blockcount,
-					&i);
+			error = xfs_bmbt_lookup_eq(bma->cur, new, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
-			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
 			error = xfs_btree_insert(bma->cur, &i);
 			if (error)
 				goto done;
@@ -2053,30 +1974,9 @@ xfs_bmap_add_extent_delay_real(
 			if (error)
 				goto done;
 		}
-		temp = xfs_bmap_worst_indlen(bma->ip, temp);
-		temp2 = xfs_bmap_worst_indlen(bma->ip, temp2);
-		diff = (int)(temp + temp2 -
-			     (startblockval(PREV.br_startblock) -
-			      (bma->cur ?
-			       bma->cur->bc_private.b.allocated : 0)));
-		if (diff > 0) {
-			error = xfs_mod_fdblocks(bma->ip->i_mount,
-						 -((int64_t)diff), false);
-			ASSERT(!error);
-			if (error)
-				goto done;
-		}
-
-		ep = xfs_iext_get_ext(ifp, bma->idx);
-		xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-		trace_xfs_bmap_pre_update(bma->ip, bma->idx + 2, state, _THIS_IP_);
-		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, bma->idx + 2),
-			nullstartblock((int)temp2));
-		trace_xfs_bmap_post_update(bma->ip, bma->idx + 2, state, _THIS_IP_);
 
-		bma->idx++;
-		da_new = temp + temp2;
+		da_new = startblockval(PREV.br_startblock) +
+			 startblockval(RIGHT.br_startblock);
 		break;
 
 	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
@@ -2110,19 +2010,17 @@ xfs_bmap_add_extent_delay_real(
 			goto done;
 	}
 
-	/* adjust for changes in reserved delayed indirect blocks */
-	if (da_old || da_new) {
-		temp = da_new;
-		if (bma->cur)
-			temp += bma->cur->bc_private.b.allocated;
-		if (temp < da_old)
-			xfs_mod_fdblocks(bma->ip->i_mount,
-					(int64_t)(da_old - temp), false);
+	if (bma->cur) {
+		da_new += bma->cur->bc_private.b.allocated;
+		bma->cur->bc_private.b.allocated = 0;
 	}
 
-	/* clear out the allocated field, done with it now in any case. */
-	if (bma->cur)
-		bma->cur->bc_private.b.allocated = 0;
+	/* adjust for changes in reserved delayed indirect blocks */
+	if (da_new != da_old) {
+		ASSERT(state == 0 || da_new < da_old);
+		error = xfs_mod_fdblocks(mp, (int64_t)(da_old - da_new),
+				false);
+	}
 
 	xfs_bmap_check_leaf_extents(bma->cur, bma->ip, whichfork);
 done:
@@ -2142,7 +2040,7 @@ xfs_bmap_add_extent_unwritten_real(
 	struct xfs_trans	*tp,
 	xfs_inode_t		*ip,	/* incore inode pointer */
 	int			whichfork,
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
+	struct xfs_iext_cursor	*icur,
 	xfs_btree_cur_t		**curp,	/* if *curp is null, not a btree */
 	xfs_bmbt_irec_t		*new,	/* new data to add to file extents */
 	xfs_fsblock_t		*first,	/* pointer to firstblock variable */
@@ -2150,28 +2048,22 @@ xfs_bmap_add_extent_unwritten_real(
 	int			*logflagsp) /* inode logging flags */
 {
 	xfs_btree_cur_t		*cur;	/* btree cursor */
-	xfs_bmbt_rec_host_t	*ep;	/* extent entry for idx */
 	int			error;	/* error return value */
 	int			i;	/* temp state */
 	xfs_ifork_t		*ifp;	/* inode fork pointer */
 	xfs_fileoff_t		new_endoff;	/* end offset of new entry */
-	xfs_exntst_t		newext;	/* new extent state */
-	xfs_exntst_t		oldext;	/* old extent state */
 	xfs_bmbt_irec_t		r[3];	/* neighbor extent entries */
 					/* left is 0, right is 1, prev is 2 */
 	int			rval=0;	/* return value (logging flags) */
-	int			state = 0;/* state bits, accessed thru macros */
+	int			state = xfs_bmap_fork_to_state(whichfork);
 	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_bmbt_irec	old;
 
 	*logflagsp = 0;
 
 	cur = *curp;
 	ifp = XFS_IFORK_PTR(ip, whichfork);
-	if (whichfork == XFS_COW_FORK)
-		state |= BMAP_COWFORK;
 
-	ASSERT(*idx >= 0);
-	ASSERT(*idx <= xfs_iext_count(ifp));
 	ASSERT(!isnullstartblock(new->br_startblock));
 
 	XFS_STATS_INC(mp, xs_add_exlist);
@@ -2184,12 +2076,8 @@ xfs_bmap_add_extent_unwritten_real(
 	 * Set up a bunch of variables to make the tests simpler.
 	 */
 	error = 0;
-	ep = xfs_iext_get_ext(ifp, *idx);
-	xfs_bmbt_get_all(ep, &PREV);
-	newext = new->br_state;
-	oldext = (newext == XFS_EXT_UNWRITTEN) ?
-		XFS_EXT_NORM : XFS_EXT_UNWRITTEN;
-	ASSERT(PREV.br_state == oldext);
+	xfs_iext_get_extent(ifp, icur, &PREV);
+	ASSERT(new->br_state != PREV.br_state);
 	new_endoff = new->br_startoff + new->br_blockcount;
 	ASSERT(PREV.br_startoff <= new->br_startoff);
 	ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);
@@ -2207,10 +2095,8 @@ xfs_bmap_add_extent_unwritten_real(
 	 * Check and set flags if this segment has a left neighbor.
 	 * Don't set contiguous if the combined extent would be too large.
 	 */
-	if (*idx > 0) {
+	if (xfs_iext_peek_prev_extent(ifp, icur, &LEFT)) {
 		state |= BMAP_LEFT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &LEFT);
-
 		if (isnullstartblock(LEFT.br_startblock))
 			state |= BMAP_LEFT_DELAY;
 	}
@@ -2218,7 +2104,7 @@ xfs_bmap_add_extent_unwritten_real(
 	if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
 	    LEFT.br_startoff + LEFT.br_blockcount == new->br_startoff &&
 	    LEFT.br_startblock + LEFT.br_blockcount == new->br_startblock &&
-	    LEFT.br_state == newext &&
+	    LEFT.br_state == new->br_state &&
 	    LEFT.br_blockcount + new->br_blockcount <= MAXEXTLEN)
 		state |= BMAP_LEFT_CONTIG;
 
@@ -2227,9 +2113,8 @@ xfs_bmap_add_extent_unwritten_real(
 	 * Don't set contiguous if the combined extent would be too large.
 	 * Also check for all-three-contiguous being too large.
 	 */
-	if (*idx < xfs_iext_count(ifp) - 1) {
+	if (xfs_iext_peek_next_extent(ifp, icur, &RIGHT)) {
 		state |= BMAP_RIGHT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx + 1), &RIGHT);
 		if (isnullstartblock(RIGHT.br_startblock))
 			state |= BMAP_RIGHT_DELAY;
 	}
@@ -2237,7 +2122,7 @@ xfs_bmap_add_extent_unwritten_real(
 	if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
 	    new_endoff == RIGHT.br_startoff &&
 	    new->br_startblock + new->br_blockcount == RIGHT.br_startblock &&
-	    newext == RIGHT.br_state &&
+	    new->br_state == RIGHT.br_state &&
 	    new->br_blockcount + RIGHT.br_blockcount <= MAXEXTLEN &&
 	    ((state & (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
 		       BMAP_RIGHT_FILLING)) !=
@@ -2258,24 +2143,20 @@ xfs_bmap_add_extent_unwritten_real(
 		 * Setting all of a previous oldext extent to newext.
 		 * The left and right neighbors are both contiguous with new.
 		 */
-		--*idx;
-
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
-			LEFT.br_blockcount + PREV.br_blockcount +
-			RIGHT.br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		LEFT.br_blockcount += PREV.br_blockcount + RIGHT.br_blockcount;
 
-		xfs_iext_remove(ip, *idx + 1, 2, state);
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_prev(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &LEFT);
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 				XFS_IFORK_NEXTENTS(ip, whichfork) - 2);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
-					RIGHT.br_startblock,
-					RIGHT.br_blockcount, &i)))
+			error = xfs_bmbt_lookup_eq(cur, &RIGHT, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
 			if ((error = xfs_btree_delete(cur, &i)))
@@ -2290,10 +2171,8 @@ xfs_bmap_add_extent_unwritten_real(
 			if ((error = xfs_btree_decrement(cur, 0, &i)))
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
-				LEFT.br_startblock,
-				LEFT.br_blockcount + PREV.br_blockcount +
-				RIGHT.br_blockcount, LEFT.br_state)))
+			error = xfs_bmbt_update(cur, &LEFT);
+			if (error)
 				goto done;
 		}
 		break;
@@ -2303,23 +2182,19 @@ xfs_bmap_add_extent_unwritten_real(
 		 * Setting all of a previous oldext extent to newext.
 		 * The left neighbor is contiguous, the right is not.
 		 */
-		--*idx;
+		LEFT.br_blockcount += PREV.br_blockcount;
 
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
-			LEFT.br_blockcount + PREV.br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-
-		xfs_iext_remove(ip, *idx + 1, 1, state);
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_prev(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &LEFT);
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 				XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
-					PREV.br_startblock, PREV.br_blockcount,
-					&i)))
+			error = xfs_bmbt_lookup_eq(cur, &PREV, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
 			if ((error = xfs_btree_delete(cur, &i)))
@@ -2328,10 +2203,8 @@ xfs_bmap_add_extent_unwritten_real(
 			if ((error = xfs_btree_decrement(cur, 0, &i)))
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
-				LEFT.br_startblock,
-				LEFT.br_blockcount + PREV.br_blockcount,
-				LEFT.br_state)))
+			error = xfs_bmbt_update(cur, &LEFT);
+			if (error)
 				goto done;
 		}
 		break;
@@ -2341,21 +2214,22 @@ xfs_bmap_add_extent_unwritten_real(
 		 * Setting all of a previous oldext extent to newext.
 		 * The right neighbor is contiguous, the left is not.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep,
-			PREV.br_blockcount + RIGHT.br_blockcount);
-		xfs_bmbt_set_state(ep, newext);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-		xfs_iext_remove(ip, *idx + 1, 1, state);
+		PREV.br_blockcount += RIGHT.br_blockcount;
+		PREV.br_state = new->br_state;
+
+		xfs_iext_next(ifp, icur);
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_prev(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &PREV);
+
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 				XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
-					RIGHT.br_startblock,
-					RIGHT.br_blockcount, &i)))
+			error = xfs_bmbt_lookup_eq(cur, &RIGHT, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
 			if ((error = xfs_btree_delete(cur, &i)))
@@ -2364,10 +2238,8 @@ xfs_bmap_add_extent_unwritten_real(
 			if ((error = xfs_btree_decrement(cur, 0, &i)))
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			if ((error = xfs_bmbt_update(cur, new->br_startoff,
-				new->br_startblock,
-				new->br_blockcount + RIGHT.br_blockcount,
-				newext)))
+			error = xfs_bmbt_update(cur, &PREV);
+			if (error)
 				goto done;
 		}
 		break;
@@ -2378,22 +2250,19 @@ xfs_bmap_add_extent_unwritten_real(
 		 * Neither the left nor right neighbors are contiguous with
 		 * the new one.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_state(ep, newext);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		PREV.br_state = new->br_state;
+		xfs_iext_update_extent(ip, state, icur, &PREV);
 
 		if (cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
-					new->br_startblock, new->br_blockcount,
-					&i)))
+			error = xfs_bmbt_lookup_eq(cur, new, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			if ((error = xfs_bmbt_update(cur, new->br_startoff,
-				new->br_startblock, new->br_blockcount,
-				newext)))
+			error = xfs_bmbt_update(cur, &PREV);
+			if (error)
 				goto done;
 		}
 		break;
@@ -2403,43 +2272,32 @@ xfs_bmap_add_extent_unwritten_real(
 		 * Setting the first part of a previous oldext extent to newext.
 		 * The left neighbor is contiguous.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx - 1, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx - 1),
-			LEFT.br_blockcount + new->br_blockcount);
-		xfs_bmbt_set_startoff(ep,
-			PREV.br_startoff + new->br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx - 1, state, _THIS_IP_);
-
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_startblock(ep,
-			new->br_startblock + new->br_blockcount);
-		xfs_bmbt_set_blockcount(ep,
-			PREV.br_blockcount - new->br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-
-		--*idx;
+		LEFT.br_blockcount += new->br_blockcount;
+
+		old = PREV;
+		PREV.br_startoff += new->br_blockcount;
+		PREV.br_startblock += new->br_blockcount;
+		PREV.br_blockcount -= new->br_blockcount;
+
+		xfs_iext_update_extent(ip, state, icur, &PREV);
+		xfs_iext_prev(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &LEFT);
 
 		if (cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
-					PREV.br_startblock, PREV.br_blockcount,
-					&i)))
+			error = xfs_bmbt_lookup_eq(cur, &old, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			if ((error = xfs_bmbt_update(cur,
-				PREV.br_startoff + new->br_blockcount,
-				PREV.br_startblock + new->br_blockcount,
-				PREV.br_blockcount - new->br_blockcount,
-				oldext)))
+			error = xfs_bmbt_update(cur, &PREV);
+			if (error)
 				goto done;
-			if ((error = xfs_btree_decrement(cur, 0, &i)))
+			error = xfs_btree_decrement(cur, 0, &i);
+			if (error)
 				goto done;
-			error = xfs_bmbt_update(cur, LEFT.br_startoff,
-				LEFT.br_startblock,
-				LEFT.br_blockcount + new->br_blockcount,
-				LEFT.br_state);
+			error = xfs_bmbt_update(cur, &LEFT);
 			if (error)
 				goto done;
 		}
@@ -2450,32 +2308,25 @@ xfs_bmap_add_extent_unwritten_real(
 		 * Setting the first part of a previous oldext extent to newext.
 		 * The left neighbor is not contiguous.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		ASSERT(ep && xfs_bmbt_get_state(ep) == oldext);
-		xfs_bmbt_set_startoff(ep, new_endoff);
-		xfs_bmbt_set_blockcount(ep,
-			PREV.br_blockcount - new->br_blockcount);
-		xfs_bmbt_set_startblock(ep,
-			new->br_startblock + new->br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-
-		xfs_iext_insert(ip, *idx, 1, new, state);
+		old = PREV;
+		PREV.br_startoff += new->br_blockcount;
+		PREV.br_startblock += new->br_blockcount;
+		PREV.br_blockcount -= new->br_blockcount;
+
+		xfs_iext_update_extent(ip, state, icur, &PREV);
+		xfs_iext_insert(ip, icur, new, state);
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 				XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
 		if (cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
-					PREV.br_startblock, PREV.br_blockcount,
-					&i)))
+			error = xfs_bmbt_lookup_eq(cur, &old, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			if ((error = xfs_bmbt_update(cur,
-				PREV.br_startoff + new->br_blockcount,
-				PREV.br_startblock + new->br_blockcount,
-				PREV.br_blockcount - new->br_blockcount,
-				oldext)))
+			error = xfs_bmbt_update(cur, &PREV);
+			if (error)
 				goto done;
 			cur->bc_rec.b = *new;
 			if ((error = xfs_btree_insert(cur, &i)))
@@ -2489,39 +2340,33 @@ xfs_bmap_add_extent_unwritten_real(
 		 * Setting the last part of a previous oldext extent to newext.
 		 * The right neighbor is contiguous with the new allocation.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep,
-			PREV.br_blockcount - new->br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		old = PREV;
+		PREV.br_blockcount -= new->br_blockcount;
 
-		++*idx;
+		RIGHT.br_startoff = new->br_startoff;
+		RIGHT.br_startblock = new->br_startblock;
+		RIGHT.br_blockcount += new->br_blockcount;
 
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
-			new->br_startoff, new->br_startblock,
-			new->br_blockcount + RIGHT.br_blockcount, newext);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		xfs_iext_update_extent(ip, state, icur, &PREV);
+		xfs_iext_next(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &RIGHT);
 
 		if (cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
-					PREV.br_startblock,
-					PREV.br_blockcount, &i)))
+			error = xfs_bmbt_lookup_eq(cur, &old, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
-				PREV.br_startblock,
-				PREV.br_blockcount - new->br_blockcount,
-				oldext)))
+			error = xfs_bmbt_update(cur, &PREV);
+			if (error)
 				goto done;
-			if ((error = xfs_btree_increment(cur, 0, &i)))
+			error = xfs_btree_increment(cur, 0, &i);
+			if (error)
 				goto done;
-			if ((error = xfs_bmbt_update(cur, new->br_startoff,
-				new->br_startblock,
-				new->br_blockcount + RIGHT.br_blockcount,
-				newext)))
+			error = xfs_bmbt_update(cur, &RIGHT);
+			if (error)
 				goto done;
 		}
 		break;
@@ -2531,13 +2376,12 @@ xfs_bmap_add_extent_unwritten_real(
 		 * Setting the last part of a previous oldext extent to newext.
 		 * The right neighbor is not contiguous.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep,
-			PREV.br_blockcount - new->br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		old = PREV;
+		PREV.br_blockcount -= new->br_blockcount;
 
-		++*idx;
-		xfs_iext_insert(ip, *idx, 1, new, state);
+		xfs_iext_update_extent(ip, state, icur, &PREV);
+		xfs_iext_next(ifp, icur);
+		xfs_iext_insert(ip, icur, new, state);
 
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 				XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
@@ -2545,22 +2389,17 @@ xfs_bmap_add_extent_unwritten_real(
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
-					PREV.br_startblock, PREV.br_blockcount,
-					&i)))
+			error = xfs_bmbt_lookup_eq(cur, &old, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
-				PREV.br_startblock,
-				PREV.br_blockcount - new->br_blockcount,
-				oldext)))
+			error = xfs_bmbt_update(cur, &PREV);
+			if (error)
 				goto done;
-			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
-					new->br_startblock, new->br_blockcount,
-					&i)))
+			error = xfs_bmbt_lookup_eq(cur, new, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
-			cur->bc_rec.b.br_state = XFS_EXT_NORM;
 			if ((error = xfs_btree_insert(cur, &i)))
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
@@ -2573,20 +2412,20 @@ xfs_bmap_add_extent_unwritten_real(
 		 * newext.  Contiguity is impossible here.
 		 * One extent becomes three extents.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep,
-			new->br_startoff - PREV.br_startoff);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		old = PREV;
+		PREV.br_blockcount = new->br_startoff - PREV.br_startoff;
 
 		r[0] = *new;
 		r[1].br_startoff = new_endoff;
 		r[1].br_blockcount =
-			PREV.br_startoff + PREV.br_blockcount - new_endoff;
+			old.br_startoff + old.br_blockcount - new_endoff;
 		r[1].br_startblock = new->br_startblock + new->br_blockcount;
-		r[1].br_state = oldext;
+		r[1].br_state = PREV.br_state;
 
-		++*idx;
-		xfs_iext_insert(ip, *idx, 2, &r[0], state);
+		xfs_iext_update_extent(ip, state, icur, &PREV);
+		xfs_iext_next(ifp, icur);
+		xfs_iext_insert(ip, icur, &r[1], state);
+		xfs_iext_insert(ip, icur, &r[0], state);
 
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 				XFS_IFORK_NEXTENTS(ip, whichfork) + 2);
@@ -2594,20 +2433,16 @@ xfs_bmap_add_extent_unwritten_real(
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
-					PREV.br_startblock, PREV.br_blockcount,
-					&i)))
+			error = xfs_bmbt_lookup_eq(cur, &old, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
 			/* new right extent - oldext */
-			if ((error = xfs_bmbt_update(cur, r[1].br_startoff,
-				r[1].br_startblock, r[1].br_blockcount,
-				r[1].br_state)))
+			error = xfs_bmbt_update(cur, &r[1]);
+			if (error)
 				goto done;
 			/* new left extent - oldext */
 			cur->bc_rec.b = PREV;
-			cur->bc_rec.b.br_blockcount =
-				new->br_startoff - PREV.br_startoff;
 			if ((error = xfs_btree_insert(cur, &i)))
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
@@ -2616,13 +2451,11 @@ xfs_bmap_add_extent_unwritten_real(
 			 * we are about to insert as we can't trust it after
 			 * the previous insert.
 			 */
-			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
-					new->br_startblock, new->br_blockcount,
-					&i)))
+			error = xfs_bmbt_lookup_eq(cur, new, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
 			/* new middle extent - newext */
-			cur->bc_rec.b.br_state = new->br_state;
 			if ((error = xfs_btree_insert(cur, &i)))
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
@@ -2681,7 +2514,7 @@ STATIC void
 xfs_bmap_add_extent_hole_delay(
 	xfs_inode_t		*ip,	/* incore inode pointer */
 	int			whichfork,
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
+	struct xfs_iext_cursor	*icur,
 	xfs_bmbt_irec_t		*new)	/* new data to add to file extents */
 {
 	xfs_ifork_t		*ifp;	/* inode fork pointer */
@@ -2689,22 +2522,17 @@ xfs_bmap_add_extent_hole_delay(
 	xfs_filblks_t		newlen=0;	/* new indirect size */
 	xfs_filblks_t		oldlen=0;	/* old indirect size */
 	xfs_bmbt_irec_t		right;	/* right neighbor extent entry */
-	int			state;  /* state bits, accessed thru macros */
-	xfs_filblks_t		temp=0;	/* temp for indirect calculations */
+	int			state = xfs_bmap_fork_to_state(whichfork);
+	xfs_filblks_t		temp;	 /* temp for indirect calculations */
 
 	ifp = XFS_IFORK_PTR(ip, whichfork);
-	state = 0;
-	if (whichfork == XFS_COW_FORK)
-		state |= BMAP_COWFORK;
 	ASSERT(isnullstartblock(new->br_startblock));
 
 	/*
 	 * Check and set flags if this segment has a left neighbor
 	 */
-	if (*idx > 0) {
+	if (xfs_iext_peek_prev_extent(ifp, icur, &left)) {
 		state |= BMAP_LEFT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &left);
-
 		if (isnullstartblock(left.br_startblock))
 			state |= BMAP_LEFT_DELAY;
 	}
@@ -2713,10 +2541,8 @@ xfs_bmap_add_extent_hole_delay(
 	 * Check and set flags if the current (right) segment exists.
 	 * If it doesn't exist, we're converting the hole at end-of-file.
 	 */
-	if (*idx < xfs_iext_count(ifp)) {
+	if (xfs_iext_get_extent(ifp, icur, &right)) {
 		state |= BMAP_RIGHT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &right);
-
 		if (isnullstartblock(right.br_startblock))
 			state |= BMAP_RIGHT_DELAY;
 	}
@@ -2748,22 +2574,20 @@ xfs_bmap_add_extent_hole_delay(
 		 * on the left and on the right.
 		 * Merge all three into a single extent record.
 		 */
-		--*idx;
 		temp = left.br_blockcount + new->br_blockcount +
 			right.br_blockcount;
 
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
 		oldlen = startblockval(left.br_startblock) +
 			startblockval(new->br_startblock) +
 			startblockval(right.br_startblock);
 		newlen = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
 					 oldlen);
-		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
-			nullstartblock((int)newlen));
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		left.br_startblock = nullstartblock(newlen);
+		left.br_blockcount = temp;
 
-		xfs_iext_remove(ip, *idx + 1, 1, state);
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_prev(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &left);
 		break;
 
 	case BMAP_LEFT_CONTIG:
@@ -2772,18 +2596,17 @@ xfs_bmap_add_extent_hole_delay(
 		 * on the left.
 		 * Merge the new allocation with the left neighbor.
 		 */
-		--*idx;
 		temp = left.br_blockcount + new->br_blockcount;
 
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
 		oldlen = startblockval(left.br_startblock) +
 			startblockval(new->br_startblock);
 		newlen = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
 					 oldlen);
-		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
-			nullstartblock((int)newlen));
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		left.br_blockcount = temp;
+		left.br_startblock = nullstartblock(newlen);
+
+		xfs_iext_prev(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &left);
 		break;
 
 	case BMAP_RIGHT_CONTIG:
@@ -2792,16 +2615,15 @@ xfs_bmap_add_extent_hole_delay(
 		 * on the right.
 		 * Merge the new allocation with the right neighbor.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
 		temp = new->br_blockcount + right.br_blockcount;
 		oldlen = startblockval(new->br_startblock) +
 			startblockval(right.br_startblock);
 		newlen = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
 					 oldlen);
-		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
-			new->br_startoff,
-			nullstartblock((int)newlen), temp, right.br_state);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		right.br_startoff = new->br_startoff;
+		right.br_startblock = nullstartblock(newlen);
+		right.br_blockcount = temp;
+		xfs_iext_update_extent(ip, state, icur, &right);
 		break;
 
 	case 0:
@@ -2811,7 +2633,7 @@ xfs_bmap_add_extent_hole_delay(
 		 * Insert a new entry.
 		 */
 		oldlen = newlen = 0;
-		xfs_iext_insert(ip, *idx, 1, new, state);
+		xfs_iext_insert(ip, icur, new, state);
 		break;
 	}
 	if (oldlen != newlen) {
@@ -2832,7 +2654,7 @@ xfs_bmap_add_extent_hole_real(
 	struct xfs_trans	*tp,
 	struct xfs_inode	*ip,
 	int			whichfork,
-	xfs_extnum_t		*idx,
+	struct xfs_iext_cursor	*icur,
 	struct xfs_btree_cur	**curp,
 	struct xfs_bmbt_irec	*new,
 	xfs_fsblock_t		*first,
@@ -2847,27 +2669,19 @@ xfs_bmap_add_extent_hole_real(
 	xfs_bmbt_irec_t		left;	/* left neighbor extent entry */
 	xfs_bmbt_irec_t		right;	/* right neighbor extent entry */
 	int			rval=0;	/* return value (logging flags) */
-	int			state;	/* state bits, accessed thru macros */
+	int			state = xfs_bmap_fork_to_state(whichfork);
+	struct xfs_bmbt_irec	old;
 
-	ASSERT(*idx >= 0);
-	ASSERT(*idx <= xfs_iext_count(ifp));
 	ASSERT(!isnullstartblock(new->br_startblock));
 	ASSERT(!cur || !(cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
 
 	XFS_STATS_INC(mp, xs_add_exlist);
 
-	state = 0;
-	if (whichfork == XFS_ATTR_FORK)
-		state |= BMAP_ATTRFORK;
-	if (whichfork == XFS_COW_FORK)
-		state |= BMAP_COWFORK;
-
 	/*
 	 * Check and set flags if this segment has a left neighbor.
 	 */
-	if (*idx > 0) {
+	if (xfs_iext_peek_prev_extent(ifp, icur, &left)) {
 		state |= BMAP_LEFT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &left);
 		if (isnullstartblock(left.br_startblock))
 			state |= BMAP_LEFT_DELAY;
 	}
@@ -2876,9 +2690,8 @@ xfs_bmap_add_extent_hole_real(
 	 * Check and set flags if this segment has a current value.
 	 * Not true if we're inserting into the "hole" at eof.
 	 */
-	if (*idx < xfs_iext_count(ifp)) {
+	if (xfs_iext_get_extent(ifp, icur, &right)) {
 		state |= BMAP_RIGHT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &right);
 		if (isnullstartblock(right.br_startblock))
 			state |= BMAP_RIGHT_DELAY;
 	}
@@ -2915,14 +2728,11 @@ xfs_bmap_add_extent_hole_real(
 		 * left and on the right.
 		 * Merge all three into a single extent record.
 		 */
-		--*idx;
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
-			left.br_blockcount + new->br_blockcount +
-			right.br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		left.br_blockcount += new->br_blockcount + right.br_blockcount;
 
-		xfs_iext_remove(ip, *idx + 1, 1, state);
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_prev(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &left);
 
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 			XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
@@ -2930,9 +2740,7 @@ xfs_bmap_add_extent_hole_real(
 			rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
 		} else {
 			rval = XFS_ILOG_CORE;
-			error = xfs_bmbt_lookup_eq(cur, right.br_startoff,
-					right.br_startblock, right.br_blockcount,
-					&i);
+			error = xfs_bmbt_lookup_eq(cur, &right, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
@@ -2944,12 +2752,7 @@ xfs_bmap_add_extent_hole_real(
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			error = xfs_bmbt_update(cur, left.br_startoff,
-					left.br_startblock,
-					left.br_blockcount +
-						new->br_blockcount +
-						right.br_blockcount,
-					left.br_state);
+			error = xfs_bmbt_update(cur, &left);
 			if (error)
 				goto done;
 		}
@@ -2961,27 +2764,21 @@ xfs_bmap_add_extent_hole_real(
 		 * on the left.
 		 * Merge the new allocation with the left neighbor.
 		 */
-		--*idx;
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
-			left.br_blockcount + new->br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		old = left;
+		left.br_blockcount += new->br_blockcount;
+
+		xfs_iext_prev(ifp, icur);
+		xfs_iext_update_extent(ip, state, icur, &left);
 
 		if (cur == NULL) {
 			rval = xfs_ilog_fext(whichfork);
 		} else {
 			rval = 0;
-			error = xfs_bmbt_lookup_eq(cur, left.br_startoff,
-					left.br_startblock, left.br_blockcount,
-					&i);
+			error = xfs_bmbt_lookup_eq(cur, &old, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			error = xfs_bmbt_update(cur, left.br_startoff,
-					left.br_startblock,
-					left.br_blockcount +
-						new->br_blockcount,
-					left.br_state);
+			error = xfs_bmbt_update(cur, &left);
 			if (error)
 				goto done;
 		}
@@ -2993,29 +2790,22 @@ xfs_bmap_add_extent_hole_real(
 		 * on the right.
 		 * Merge the new allocation with the right neighbor.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
-			new->br_startoff, new->br_startblock,
-			new->br_blockcount + right.br_blockcount,
-			right.br_state);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		old = right;
+
+		right.br_startoff = new->br_startoff;
+		right.br_startblock = new->br_startblock;
+		right.br_blockcount += new->br_blockcount;
+		xfs_iext_update_extent(ip, state, icur, &right);
 
 		if (cur == NULL) {
 			rval = xfs_ilog_fext(whichfork);
 		} else {
 			rval = 0;
-			error = xfs_bmbt_lookup_eq(cur,
-					right.br_startoff,
-					right.br_startblock,
-					right.br_blockcount, &i);
+			error = xfs_bmbt_lookup_eq(cur, &old, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			error = xfs_bmbt_update(cur, new->br_startoff,
-					new->br_startblock,
-					new->br_blockcount +
-						right.br_blockcount,
-					right.br_state);
+			error = xfs_bmbt_update(cur, &right);
 			if (error)
 				goto done;
 		}
@@ -3027,21 +2817,17 @@ xfs_bmap_add_extent_hole_real(
 		 * real allocation.
 		 * Insert a new entry.
 		 */
-		xfs_iext_insert(ip, *idx, 1, new, state);
+		xfs_iext_insert(ip, icur, new, state);
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 			XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
 		if (cur == NULL) {
 			rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
 		} else {
 			rval = XFS_ILOG_CORE;
-			error = xfs_bmbt_lookup_eq(cur,
-					new->br_startoff,
-					new->br_startblock,
-					new->br_blockcount, &i);
+			error = xfs_bmbt_lookup_eq(cur, new, &i);
 			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
-			cur->bc_rec.b.br_state = new->br_state;
 			error = xfs_btree_insert(cur, &i);
 			if (error)
 				goto done;
@@ -3981,7 +3767,7 @@ xfs_bmapi_read(
 	struct xfs_bmbt_irec	got;
 	xfs_fileoff_t		obno;
 	xfs_fileoff_t		end;
-	xfs_extnum_t		idx;
+	struct xfs_iext_cursor	icur;
 	int			error;
 	bool			eof = false;
 	int			n = 0;
@@ -4023,7 +3809,7 @@ xfs_bmapi_read(
 			return error;
 	}
 
-	if (!xfs_iext_lookup_extent(ip, ifp, bno, &idx, &got))
+	if (!xfs_iext_lookup_extent(ip, ifp, bno, &icur, &got))
 		eof = true;
 	end = bno + len;
 	obno = bno;
@@ -4055,7 +3841,7 @@ xfs_bmapi_read(
 			break;
 
 		/* Else go on to the next record. */
-		if (!xfs_iext_get_extent(ifp, ++idx, &got))
+		if (!xfs_iext_next_extent(ifp, &icur, &got))
 			eof = true;
 	}
 	*nmap = n;
@@ -4083,7 +3869,7 @@ xfs_bmapi_reserve_delalloc(
 	xfs_filblks_t		len,
 	xfs_filblks_t		prealloc,
 	struct xfs_bmbt_irec	*got,
-	xfs_extnum_t		*lastx,
+	struct xfs_iext_cursor	*icur,
 	int			eof)
 {
 	struct xfs_mount	*mp = ip->i_mount;
@@ -4113,7 +3899,7 @@ xfs_bmapi_reserve_delalloc(
 	if (extsz) {
 		struct xfs_bmbt_irec	prev;
 
-		if (!xfs_iext_get_extent(ifp, *lastx - 1, &prev))
+		if (!xfs_iext_peek_prev_extent(ifp, icur, &prev))
 			prev.br_startoff = NULLFILEOFF;
 
 		error = xfs_bmap_extsize_align(mp, got, &prev, extsz, rt, eof,
@@ -4162,7 +3948,7 @@ xfs_bmapi_reserve_delalloc(
 	got->br_blockcount = alen;
 	got->br_state = XFS_EXT_NORM;
 
-	xfs_bmap_add_extent_hole_delay(ip, whichfork, lastx, got);
+	xfs_bmap_add_extent_hole_delay(ip, whichfork, icur, got);
 
 	/*
 	 * Tag the inode if blocks were preallocated. Note that COW fork
@@ -4207,10 +3993,7 @@ xfs_bmapi_allocate(
 	if (bma->wasdel) {
 		bma->length = (xfs_extlen_t)bma->got.br_blockcount;
 		bma->offset = bma->got.br_startoff;
-		if (bma->idx) {
-			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1),
-					 &bma->prev);
-		}
+		xfs_iext_peek_prev_extent(ifp, &bma->icur, &bma->prev);
 	} else {
 		bma->length = XFS_FILBLKS_MIN(bma->length, MAXEXTLEN);
 		if (!bma->eof)
@@ -4295,7 +4078,7 @@ xfs_bmapi_allocate(
 		error = xfs_bmap_add_extent_delay_real(bma, whichfork);
 	else
 		error = xfs_bmap_add_extent_hole_real(bma->tp, bma->ip,
-				whichfork, &bma->idx, &bma->cur, &bma->got,
+				whichfork, &bma->icur, &bma->cur, &bma->got,
 				bma->firstblock, bma->dfops, &bma->logflags);
 
 	bma->logflags |= tmp_logflags;
@@ -4307,7 +4090,7 @@ xfs_bmapi_allocate(
 	 * or xfs_bmap_add_extent_hole_real might have merged it into one of
 	 * the neighbouring ones.
 	 */
-	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &bma->got);
+	xfs_iext_get_extent(ifp, &bma->icur, &bma->got);
 
 	ASSERT(bma->got.br_startoff <= bma->offset);
 	ASSERT(bma->got.br_startoff + bma->got.br_blockcount >=
@@ -4365,8 +4148,8 @@ xfs_bmapi_convert_unwritten(
 	}
 
 	error = xfs_bmap_add_extent_unwritten_real(bma->tp, bma->ip, whichfork,
-			&bma->idx, &bma->cur, mval, bma->firstblock, bma->dfops,
-			&tmp_logflags);
+			&bma->icur, &bma->cur, mval, bma->firstblock,
+			bma->dfops, &tmp_logflags);
 	/*
 	 * Log the inode core unconditionally in the unwritten extent conversion
 	 * path because the conversion might not have done so (e.g., if the
@@ -4388,7 +4171,7 @@ xfs_bmapi_convert_unwritten(
 	 * xfs_bmap_add_extent_unwritten_real might have merged it into one
 	 * of the neighbouring ones.
 	 */
-	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &bma->got);
+	xfs_iext_get_extent(ifp, &bma->icur, &bma->got);
 
 	/*
 	 * We may have combined previously unwritten space with written space,
@@ -4507,9 +4290,9 @@ xfs_bmapi_write(
 	end = bno + len;
 	obno = bno;
 
-	if (!xfs_iext_lookup_extent(ip, ifp, bno, &bma.idx, &bma.got))
+	if (!xfs_iext_lookup_extent(ip, ifp, bno, &bma.icur, &bma.got))
 		eof = true;
-	if (!xfs_iext_get_extent(ifp, bma.idx - 1, &bma.prev))
+	if (!xfs_iext_peek_prev_extent(ifp, &bma.icur, &bma.prev))
 		bma.prev.br_startoff = NULLFILEOFF;
 	bma.tp = tp;
 	bma.ip = ip;
@@ -4551,7 +4334,8 @@ xfs_bmapi_write(
 		 * First, deal with the hole before the allocated space
 		 * that we found, if any.
 		 */
-		if (need_alloc || wasdelay) {
+		if ((need_alloc || wasdelay) &&
+		    !(flags & XFS_BMAPI_CONVERT_ONLY)) {
 			bma.eof = eof;
 			bma.conv = !!(flags & XFS_BMAPI_CONVERT);
 			bma.wasdel = wasdelay;
@@ -4614,7 +4398,7 @@ xfs_bmapi_write(
 
 		/* Else go on to the next record. */
 		bma.prev = bma.got;
-		if (!xfs_iext_get_extent(ifp, ++bma.idx, &bma.got))
+		if (!xfs_iext_next_extent(ifp, &bma.icur, &bma.got))
 			eof = true;
 	}
 	*nmap = n;
@@ -4687,7 +4471,7 @@ xfs_bmapi_remap(
 	struct xfs_btree_cur	*cur = NULL;
 	xfs_fsblock_t		firstblock = NULLFSBLOCK;
 	struct xfs_bmbt_irec	got;
-	xfs_extnum_t		idx;
+	struct xfs_iext_cursor	icur;
 	int			logflags = 0, error;
 
 	ASSERT(len > 0);
@@ -4711,7 +4495,7 @@ xfs_bmapi_remap(
 			return error;
 	}
 
-	if (xfs_iext_lookup_extent(ip, ifp, bno, &idx, &got)) {
+	if (xfs_iext_lookup_extent(ip, ifp, bno, &icur, &got)) {
 		/* make sure we only reflink into a hole. */
 		ASSERT(got.br_startoff > bno);
 		ASSERT(got.br_startoff - bno >= len);
@@ -4732,8 +4516,8 @@ xfs_bmapi_remap(
 	got.br_blockcount = len;
 	got.br_state = XFS_EXT_NORM;
 
-	error = xfs_bmap_add_extent_hole_real(tp, ip, XFS_DATA_FORK, &idx, &cur,
-			&got, &firstblock, dfops, &logflags);
+	error = xfs_bmap_add_extent_hole_real(tp, ip, XFS_DATA_FORK, &icur,
+			&cur, &got, &firstblock, dfops, &logflags);
 	if (error)
 		goto error0;
 
@@ -4849,7 +4633,7 @@ int
 xfs_bmap_del_extent_delay(
 	struct xfs_inode	*ip,
 	int			whichfork,
-	xfs_extnum_t		*idx,
+	struct xfs_iext_cursor	*icur,
 	struct xfs_bmbt_irec	*got,
 	struct xfs_bmbt_irec	*del)
 {
@@ -4859,7 +4643,8 @@ xfs_bmap_del_extent_delay(
 	int64_t			da_old, da_new, da_diff = 0;
 	xfs_fileoff_t		del_endoff, got_endoff;
 	xfs_filblks_t		got_indlen, new_indlen, stolen;
-	int			error = 0, state = 0;
+	int			state = xfs_bmap_fork_to_state(whichfork);
+	int			error = 0;
 	bool			isrt;
 
 	XFS_STATS_INC(mp, xs_del_exlist);
@@ -4870,8 +4655,6 @@ xfs_bmap_del_extent_delay(
 	da_old = startblockval(got->br_startblock);
 	da_new = 0;
 
-	ASSERT(*idx >= 0);
-	ASSERT(*idx <= xfs_iext_count(ifp));
 	ASSERT(del->br_blockcount > 0);
 	ASSERT(got->br_startoff <= del->br_startoff);
 	ASSERT(got_endoff >= del_endoff);
@@ -4895,46 +4678,39 @@ xfs_bmap_del_extent_delay(
 		return error;
 	ip->i_delayed_blks -= del->br_blockcount;
 
-	if (whichfork == XFS_COW_FORK)
-		state |= BMAP_COWFORK;
-
 	if (got->br_startoff == del->br_startoff)
-		state |= BMAP_LEFT_CONTIG;
+		state |= BMAP_LEFT_FILLING;
 	if (got_endoff == del_endoff)
-		state |= BMAP_RIGHT_CONTIG;
+		state |= BMAP_RIGHT_FILLING;
 
-	switch (state & (BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG)) {
-	case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+	switch (state & (BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING)) {
+	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
 		/*
 		 * Matches the whole extent.  Delete the entry.
 		 */
-		xfs_iext_remove(ip, *idx, 1, state);
-		--*idx;
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_prev(ifp, icur);
 		break;
-	case BMAP_LEFT_CONTIG:
+	case BMAP_LEFT_FILLING:
 		/*
 		 * Deleting the first part of the extent.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
 		got->br_startoff = del_endoff;
 		got->br_blockcount -= del->br_blockcount;
 		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip,
 				got->br_blockcount), da_old);
 		got->br_startblock = nullstartblock((int)da_new);
-		xfs_iext_update_extent(ifp, *idx, got);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		xfs_iext_update_extent(ip, state, icur, got);
 		break;
-	case BMAP_RIGHT_CONTIG:
+	case BMAP_RIGHT_FILLING:
 		/*
 		 * Deleting the last part of the extent.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
 		got->br_blockcount = got->br_blockcount - del->br_blockcount;
 		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip,
 				got->br_blockcount), da_old);
 		got->br_startblock = nullstartblock((int)da_new);
-		xfs_iext_update_extent(ifp, *idx, got);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		xfs_iext_update_extent(ip, state, icur, got);
 		break;
 	case 0:
 		/*
@@ -4946,8 +4722,6 @@ xfs_bmap_del_extent_delay(
 		 * Warn if either of the new indlen reservations is zero as this
 		 * can lead to delalloc problems.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-
 		got->br_blockcount = del->br_startoff - got->br_startoff;
 		got_indlen = xfs_bmap_worst_indlen(ip, got->br_blockcount);
 
@@ -4959,15 +4733,14 @@ xfs_bmap_del_extent_delay(
 						       del->br_blockcount);
 
 		got->br_startblock = nullstartblock((int)got_indlen);
-		xfs_iext_update_extent(ifp, *idx, got);
-		trace_xfs_bmap_post_update(ip, *idx, 0, _THIS_IP_);
 
 		new.br_startoff = del_endoff;
 		new.br_state = got->br_state;
 		new.br_startblock = nullstartblock((int)new_indlen);
 
-		++*idx;
-		xfs_iext_insert(ip, *idx, 1, &new, state);
+		xfs_iext_update_extent(ip, state, icur, got);
+		xfs_iext_next(ifp, icur);
+		xfs_iext_insert(ip, icur, &new, state);
 
 		da_new = got_indlen + new_indlen - stolen;
 		del->br_blockcount -= stolen;
@@ -4986,7 +4759,7 @@ xfs_bmap_del_extent_delay(
 void
 xfs_bmap_del_extent_cow(
 	struct xfs_inode	*ip,
-	xfs_extnum_t		*idx,
+	struct xfs_iext_cursor	*icur,
 	struct xfs_bmbt_irec	*got,
 	struct xfs_bmbt_irec	*del)
 {
@@ -5001,75 +4774,67 @@ xfs_bmap_del_extent_cow(
 	del_endoff = del->br_startoff + del->br_blockcount;
 	got_endoff = got->br_startoff + got->br_blockcount;
 
-	ASSERT(*idx >= 0);
-	ASSERT(*idx <= xfs_iext_count(ifp));
 	ASSERT(del->br_blockcount > 0);
 	ASSERT(got->br_startoff <= del->br_startoff);
 	ASSERT(got_endoff >= del_endoff);
 	ASSERT(!isnullstartblock(got->br_startblock));
 
 	if (got->br_startoff == del->br_startoff)
-		state |= BMAP_LEFT_CONTIG;
+		state |= BMAP_LEFT_FILLING;
 	if (got_endoff == del_endoff)
-		state |= BMAP_RIGHT_CONTIG;
+		state |= BMAP_RIGHT_FILLING;
 
-	switch (state & (BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG)) {
-	case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+	switch (state & (BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING)) {
+	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
 		/*
 		 * Matches the whole extent.  Delete the entry.
 		 */
-		xfs_iext_remove(ip, *idx, 1, state);
-		--*idx;
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_prev(ifp, icur);
 		break;
-	case BMAP_LEFT_CONTIG:
+	case BMAP_LEFT_FILLING:
 		/*
 		 * Deleting the first part of the extent.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
 		got->br_startoff = del_endoff;
 		got->br_blockcount -= del->br_blockcount;
 		got->br_startblock = del->br_startblock + del->br_blockcount;
-		xfs_iext_update_extent(ifp, *idx, got);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		xfs_iext_update_extent(ip, state, icur, got);
 		break;
-	case BMAP_RIGHT_CONTIG:
+	case BMAP_RIGHT_FILLING:
 		/*
 		 * Deleting the last part of the extent.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
 		got->br_blockcount -= del->br_blockcount;
-		xfs_iext_update_extent(ifp, *idx, got);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		xfs_iext_update_extent(ip, state, icur, got);
 		break;
 	case 0:
 		/*
 		 * Deleting the middle of the extent.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
 		got->br_blockcount = del->br_startoff - got->br_startoff;
-		xfs_iext_update_extent(ifp, *idx, got);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
 
 		new.br_startoff = del_endoff;
 		new.br_blockcount = got_endoff - del_endoff;
 		new.br_state = got->br_state;
 		new.br_startblock = del->br_startblock + del->br_blockcount;
 
-		++*idx;
-		xfs_iext_insert(ip, *idx, 1, &new, state);
+		xfs_iext_update_extent(ip, state, icur, got);
+		xfs_iext_next(ifp, icur);
+		xfs_iext_insert(ip, icur, &new, state);
 		break;
 	}
 }
 
 /*
  * Called by xfs_bmapi to update file extent records and the btree
- * after removing space (or undoing a delayed allocation).
+ * after removing space.
  */
 STATIC int				/* error */
-xfs_bmap_del_extent(
+xfs_bmap_del_extent_real(
 	xfs_inode_t		*ip,	/* incore inode pointer */
 	xfs_trans_t		*tp,	/* current transaction pointer */
-	xfs_extnum_t		*idx,	/* extent number to update/delete */
+	struct xfs_iext_cursor	*icur,
 	struct xfs_defer_ops	*dfops,	/* list of extents to be freed */
 	xfs_btree_cur_t		*cur,	/* if null, not a btree */
 	xfs_bmbt_irec_t		*del,	/* data to remove from extents */
@@ -5077,16 +4842,12 @@ xfs_bmap_del_extent(
 	int			whichfork, /* data or attr fork */
 	int			bflags)	/* bmapi flags */
 {
-	xfs_filblks_t		da_new;	/* new delay-alloc indirect blocks */
-	xfs_filblks_t		da_old;	/* old delay-alloc indirect blocks */
 	xfs_fsblock_t		del_endblock=0;	/* first block past del */
 	xfs_fileoff_t		del_endoff;	/* first offset past del */
-	int			delay;	/* current block is delayed allocated */
 	int			do_fx;	/* free extent at end of routine */
-	xfs_bmbt_rec_host_t	*ep;	/* current extent entry pointer */
 	int			error;	/* error return value */
-	int			flags;	/* inode logging flags */
-	xfs_bmbt_irec_t		got;	/* current extent entry */
+	int			flags = 0;/* inode logging flags */
+	struct xfs_bmbt_irec	got;	/* current extent entry */
 	xfs_fileoff_t		got_endoff;	/* first offset past got */
 	int			i;	/* temp state */
 	xfs_ifork_t		*ifp;	/* inode fork pointer */
@@ -5095,103 +4856,81 @@ xfs_bmap_del_extent(
 	xfs_bmbt_irec_t		new;	/* new record to be inserted */
 	/* REFERENCED */
 	uint			qfield;	/* quota field to update */
-	xfs_filblks_t		temp;	/* for indirect length calculations */
-	xfs_filblks_t		temp2;	/* for indirect length calculations */
-	int			state = 0;
+	int			state = xfs_bmap_fork_to_state(whichfork);
+	struct xfs_bmbt_irec	old;
 
 	mp = ip->i_mount;
 	XFS_STATS_INC(mp, xs_del_exlist);
 
-	if (whichfork == XFS_ATTR_FORK)
-		state |= BMAP_ATTRFORK;
-	else if (whichfork == XFS_COW_FORK)
-		state |= BMAP_COWFORK;
-
 	ifp = XFS_IFORK_PTR(ip, whichfork);
-	ASSERT((*idx >= 0) && (*idx < xfs_iext_count(ifp)));
 	ASSERT(del->br_blockcount > 0);
-	ep = xfs_iext_get_ext(ifp, *idx);
-	xfs_bmbt_get_all(ep, &got);
+	xfs_iext_get_extent(ifp, icur, &got);
 	ASSERT(got.br_startoff <= del->br_startoff);
 	del_endoff = del->br_startoff + del->br_blockcount;
 	got_endoff = got.br_startoff + got.br_blockcount;
 	ASSERT(got_endoff >= del_endoff);
-	delay = isnullstartblock(got.br_startblock);
-	ASSERT(isnullstartblock(del->br_startblock) == delay);
-	flags = 0;
+	ASSERT(!isnullstartblock(got.br_startblock));
 	qfield = 0;
 	error = 0;
+
 	/*
-	 * If deleting a real allocation, must free up the disk space.
+	 * If it's the case where the directory code is running with no block
+	 * reservation, and the deleted block is in the middle of its extent,
+	 * and the resulting insert of an extent would cause transformation to
+	 * btree format, then reject it.  The calling code will then swap blocks
+	 * around instead.  We have to do this now, rather than waiting for the
+	 * conversion to btree format, since the transaction will be dirty then.
 	 */
-	if (!delay) {
-		flags = XFS_ILOG_CORE;
-		/*
-		 * Realtime allocation.  Free it and record di_nblocks update.
-		 */
-		if (whichfork == XFS_DATA_FORK && XFS_IS_REALTIME_INODE(ip)) {
-			xfs_fsblock_t	bno;
-			xfs_filblks_t	len;
-
-			ASSERT(do_mod(del->br_blockcount,
-				      mp->m_sb.sb_rextsize) == 0);
-			ASSERT(do_mod(del->br_startblock,
-				      mp->m_sb.sb_rextsize) == 0);
-			bno = del->br_startblock;
-			len = del->br_blockcount;
-			do_div(bno, mp->m_sb.sb_rextsize);
-			do_div(len, mp->m_sb.sb_rextsize);
-			error = xfs_rtfree_extent(tp, bno, (xfs_extlen_t)len);
-			if (error)
-				goto done;
-			do_fx = 0;
-			nblks = len * mp->m_sb.sb_rextsize;
-			qfield = XFS_TRANS_DQ_RTBCOUNT;
-		}
-		/*
-		 * Ordinary allocation.
-		 */
-		else {
-			do_fx = 1;
-			nblks = del->br_blockcount;
-			qfield = XFS_TRANS_DQ_BCOUNT;
-		}
-		/*
-		 * Set up del_endblock and cur for later.
-		 */
-		del_endblock = del->br_startblock + del->br_blockcount;
-		if (cur) {
-			if ((error = xfs_bmbt_lookup_eq(cur, got.br_startoff,
-					got.br_startblock, got.br_blockcount,
-					&i)))
-				goto done;
-			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-		}
-		da_old = da_new = 0;
-	} else {
-		da_old = startblockval(got.br_startblock);
-		da_new = 0;
-		nblks = 0;
+	if (tp->t_blk_res == 0 &&
+	    XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS &&
+	    XFS_IFORK_NEXTENTS(ip, whichfork) >=
+			XFS_IFORK_MAXEXT(ip, whichfork) &&
+	    del->br_startoff > got.br_startoff && del_endoff < got_endoff)
+		return -ENOSPC;
+
+	flags = XFS_ILOG_CORE;
+	if (whichfork == XFS_DATA_FORK && XFS_IS_REALTIME_INODE(ip)) {
+		xfs_fsblock_t	bno;
+		xfs_filblks_t	len;
+
+		ASSERT(do_mod(del->br_blockcount, mp->m_sb.sb_rextsize) == 0);
+		ASSERT(do_mod(del->br_startblock, mp->m_sb.sb_rextsize) == 0);
+		bno = del->br_startblock;
+		len = del->br_blockcount;
+		do_div(bno, mp->m_sb.sb_rextsize);
+		do_div(len, mp->m_sb.sb_rextsize);
+		error = xfs_rtfree_extent(tp, bno, (xfs_extlen_t)len);
+		if (error)
+			goto done;
 		do_fx = 0;
+		nblks = len * mp->m_sb.sb_rextsize;
+		qfield = XFS_TRANS_DQ_RTBCOUNT;
+	} else {
+		do_fx = 1;
+		nblks = del->br_blockcount;
+		qfield = XFS_TRANS_DQ_BCOUNT;
 	}
 
-	/*
-	 * Set flag value to use in switch statement.
-	 * Left-contig is 2, right-contig is 1.
-	 */
-	switch (((got.br_startoff == del->br_startoff) << 1) |
-		(got_endoff == del_endoff)) {
-	case 3:
+	del_endblock = del->br_startblock + del->br_blockcount;
+	if (cur) {
+		error = xfs_bmbt_lookup_eq(cur, &got, &i);
+		if (error)
+			goto done;
+		XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
+	}
+
+	if (got.br_startoff == del->br_startoff)
+		state |= BMAP_LEFT_FILLING;
+	if (got_endoff == del_endoff)
+		state |= BMAP_RIGHT_FILLING;
+
+	switch (state & (BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING)) {
+	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
 		/*
 		 * Matches the whole extent.  Delete the entry.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_iext_remove(ip, *idx, 1,
-				whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0);
-		--*idx;
-		if (delay)
-			break;
-
+		xfs_iext_remove(ip, icur, state);
+		xfs_iext_prev(ifp, icur);
 		XFS_IFORK_NEXT_SET(ip, whichfork,
 			XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
 		flags |= XFS_ILOG_CORE;
@@ -5203,168 +4942,106 @@ xfs_bmap_del_extent(
 			goto done;
 		XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
 		break;
-
-	case 2:
+	case BMAP_LEFT_FILLING:
 		/*
 		 * Deleting the first part of the extent.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_startoff(ep, del_endoff);
-		temp = got.br_blockcount - del->br_blockcount;
-		xfs_bmbt_set_blockcount(ep, temp);
-		if (delay) {
-			temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
-				da_old);
-			xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-			trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-			da_new = temp;
-			break;
-		}
-		xfs_bmbt_set_startblock(ep, del_endblock);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		got.br_startoff = del_endoff;
+		got.br_startblock = del_endblock;
+		got.br_blockcount -= del->br_blockcount;
+		xfs_iext_update_extent(ip, state, icur, &got);
 		if (!cur) {
 			flags |= xfs_ilog_fext(whichfork);
 			break;
 		}
-		if ((error = xfs_bmbt_update(cur, del_endoff, del_endblock,
-				got.br_blockcount - del->br_blockcount,
-				got.br_state)))
+		error = xfs_bmbt_update(cur, &got);
+		if (error)
 			goto done;
 		break;
-
-	case 1:
+	case BMAP_RIGHT_FILLING:
 		/*
 		 * Deleting the last part of the extent.
 		 */
-		temp = got.br_blockcount - del->br_blockcount;
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep, temp);
-		if (delay) {
-			temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
-				da_old);
-			xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-			trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-			da_new = temp;
-			break;
-		}
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		got.br_blockcount -= del->br_blockcount;
+		xfs_iext_update_extent(ip, state, icur, &got);
 		if (!cur) {
 			flags |= xfs_ilog_fext(whichfork);
 			break;
 		}
-		if ((error = xfs_bmbt_update(cur, got.br_startoff,
-				got.br_startblock,
-				got.br_blockcount - del->br_blockcount,
-				got.br_state)))
+		error = xfs_bmbt_update(cur, &got);
+		if (error)
 			goto done;
 		break;
-
 	case 0:
 		/*
 		 * Deleting the middle of the extent.
 		 */
-		temp = del->br_startoff - got.br_startoff;
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(ep, temp);
+		old = got;
+
+		got.br_blockcount = del->br_startoff - got.br_startoff;
+		xfs_iext_update_extent(ip, state, icur, &got);
+
 		new.br_startoff = del_endoff;
-		temp2 = got_endoff - del_endoff;
-		new.br_blockcount = temp2;
+		new.br_blockcount = got_endoff - del_endoff;
 		new.br_state = got.br_state;
-		if (!delay) {
-			new.br_startblock = del_endblock;
-			flags |= XFS_ILOG_CORE;
-			if (cur) {
-				if ((error = xfs_bmbt_update(cur,
-						got.br_startoff,
-						got.br_startblock, temp,
-						got.br_state)))
-					goto done;
-				if ((error = xfs_btree_increment(cur, 0, &i)))
-					goto done;
-				cur->bc_rec.b = new;
-				error = xfs_btree_insert(cur, &i);
-				if (error && error != -ENOSPC)
-					goto done;
+		new.br_startblock = del_endblock;
+
+		flags |= XFS_ILOG_CORE;
+		if (cur) {
+			error = xfs_bmbt_update(cur, &got);
+			if (error)
+				goto done;
+			error = xfs_btree_increment(cur, 0, &i);
+			if (error)
+				goto done;
+			cur->bc_rec.b = new;
+			error = xfs_btree_insert(cur, &i);
+			if (error && error != -ENOSPC)
+				goto done;
+			/*
+			 * If get no-space back from btree insert, it tried a
+			 * split, and we have a zero block reservation.  Fix up
+			 * our state and return the error.
+			 */
+			if (error == -ENOSPC) {
 				/*
-				 * If get no-space back from btree insert,
-				 * it tried a split, and we have a zero
-				 * block reservation.
-				 * Fix up our state and return the error.
+				 * Reset the cursor, don't trust it after any
+				 * insert operation.
 				 */
-				if (error == -ENOSPC) {
-					/*
-					 * Reset the cursor, don't trust
-					 * it after any insert operation.
-					 */
-					if ((error = xfs_bmbt_lookup_eq(cur,
-							got.br_startoff,
-							got.br_startblock,
-							temp, &i)))
-						goto done;
-					XFS_WANT_CORRUPTED_GOTO(mp,
-								i == 1, done);
-					/*
-					 * Update the btree record back
-					 * to the original value.
-					 */
-					if ((error = xfs_bmbt_update(cur,
-							got.br_startoff,
-							got.br_startblock,
-							got.br_blockcount,
-							got.br_state)))
-						goto done;
-					/*
-					 * Reset the extent record back
-					 * to the original value.
-					 */
-					xfs_bmbt_set_blockcount(ep,
-						got.br_blockcount);
-					flags = 0;
-					error = -ENOSPC;
+				error = xfs_bmbt_lookup_eq(cur, &got, &i);
+				if (error)
 					goto done;
-				}
 				XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
-			} else
-				flags |= xfs_ilog_fext(whichfork);
-			XFS_IFORK_NEXT_SET(ip, whichfork,
-				XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
-		} else {
-			xfs_filblks_t	stolen;
-			ASSERT(whichfork == XFS_DATA_FORK);
-
-			/*
-			 * Distribute the original indlen reservation across the
-			 * two new extents. Steal blocks from the deleted extent
-			 * if necessary. Stealing blocks simply fudges the
-			 * fdblocks accounting in xfs_bunmapi().
-			 */
-			temp = xfs_bmap_worst_indlen(ip, got.br_blockcount);
-			temp2 = xfs_bmap_worst_indlen(ip, new.br_blockcount);
-			stolen = xfs_bmap_split_indlen(da_old, &temp, &temp2,
-						       del->br_blockcount);
-			da_new = temp + temp2 - stolen;
-			del->br_blockcount -= stolen;
-
-			/*
-			 * Set the reservation for each extent. Warn if either
-			 * is zero as this can lead to delalloc problems.
-			 */
-			WARN_ON_ONCE(!temp || !temp2);
-			xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-			new.br_startblock = nullstartblock((int)temp2);
-		}
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-		xfs_iext_insert(ip, *idx + 1, 1, &new, state);
-		++*idx;
+				/*
+				 * Update the btree record back
+				 * to the original value.
+				 */
+				error = xfs_bmbt_update(cur, &old);
+				if (error)
+					goto done;
+				/*
+				 * Reset the extent record back
+				 * to the original value.
+				 */
+				xfs_iext_update_extent(ip, state, icur, &old);
+				flags = 0;
+				error = -ENOSPC;
+				goto done;
+			}
+			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
+		} else
+			flags |= xfs_ilog_fext(whichfork);
+		XFS_IFORK_NEXT_SET(ip, whichfork,
+			XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
+		xfs_iext_next(ifp, icur);
+		xfs_iext_insert(ip, icur, &new, state);
 		break;
 	}
 
 	/* remove reverse mapping */
-	if (!delay) {
-		error = xfs_rmap_unmap_extent(mp, dfops, ip, whichfork, del);
-		if (error)
-			goto done;
-	}
+	error = xfs_rmap_unmap_extent(mp, dfops, ip, whichfork, del);
+	if (error)
+		goto done;
 
 	/*
 	 * If we need to, add to list of extents to delete.
@@ -5390,13 +5067,6 @@ xfs_bmap_del_extent(
 	if (qfield && !(bflags & XFS_BMAPI_REMAP))
 		xfs_trans_mod_dquot_byino(tp, ip, qfield, (long)-nblks);
 
-	/*
-	 * Account for change in delayed indirect blocks.
-	 * Nothing to do for disk quota accounting here.
-	 */
-	ASSERT(da_old >= da_new);
-	if (da_old > da_new)
-		xfs_mod_fdblocks(mp, (int64_t)(da_old - da_new), false);
 done:
 	*logflagsp = flags;
 	return error;
@@ -5412,7 +5082,7 @@ int						/* error */
 __xfs_bunmapi(
 	xfs_trans_t		*tp,		/* transaction pointer */
 	struct xfs_inode	*ip,		/* incore inode */
-	xfs_fileoff_t		bno,		/* starting offset to unmap */
+	xfs_fileoff_t		start,		/* first file offset deleted */
 	xfs_filblks_t		*rlen,		/* i/o: amount remaining */
 	int			flags,		/* misc flags */
 	xfs_extnum_t		nexts,		/* number of extents max */
@@ -5427,11 +5097,9 @@ __xfs_bunmapi(
 	xfs_bmbt_irec_t		got;		/* current extent record */
 	xfs_ifork_t		*ifp;		/* inode fork pointer */
 	int			isrt;		/* freeing in rt area */
-	xfs_extnum_t		lastx;		/* last extent index used */
 	int			logflags;	/* transaction logging flags */
 	xfs_extlen_t		mod;		/* rt extent offset */
 	xfs_mount_t		*mp;		/* mount structure */
-	xfs_fileoff_t		start;		/* first file offset deleted */
 	int			tmp_logflags;	/* partial logging flags */
 	int			wasdel;		/* was a delayed alloc extent */
 	int			whichfork;	/* data or attribute fork */
@@ -5439,8 +5107,11 @@ __xfs_bunmapi(
 	xfs_filblks_t		len = *rlen;	/* length to unmap in file */
 	xfs_fileoff_t		max_len;
 	xfs_agnumber_t		prev_agno = NULLAGNUMBER, agno;
+	xfs_fileoff_t		end;
+	struct xfs_iext_cursor	icur;
+	bool			done = false;
 
-	trace_xfs_bunmap(ip, bno, len, flags, _RET_IP_);
+	trace_xfs_bunmap(ip, start, len, flags, _RET_IP_);
 
 	whichfork = xfs_bmapi_whichfork(flags);
 	ASSERT(whichfork != XFS_COW_FORK);
@@ -5479,18 +5150,13 @@ __xfs_bunmapi(
 	}
 	XFS_STATS_INC(mp, xs_blk_unmap);
 	isrt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(ip);
-	start = bno;
-	bno = start + len - 1;
+	end = start + len;
 
-	/*
-	 * Check to see if the given block number is past the end of the
-	 * file, back up to the last block if so...
-	 */
-	if (!xfs_iext_lookup_extent(ip, ifp, bno, &lastx, &got)) {
-		ASSERT(lastx > 0);
-		xfs_iext_get_extent(ifp, --lastx, &got);
-		bno = got.br_startoff + got.br_blockcount - 1;
+	if (!xfs_iext_lookup_extent_before(ip, ifp, &end, &icur, &got)) {
+		*rlen = 0;
+		return 0;
 	}
+	end--;
 
 	logflags = 0;
 	if (ifp->if_flags & XFS_IFBROOT) {
@@ -5513,24 +5179,24 @@ __xfs_bunmapi(
 	}
 
 	extno = 0;
-	while (bno != (xfs_fileoff_t)-1 && bno >= start && lastx >= 0 &&
+	while (end != (xfs_fileoff_t)-1 && end >= start &&
 	       (nexts == 0 || extno < nexts) && max_len > 0) {
 		/*
-		 * Is the found extent after a hole in which bno lives?
+		 * Is the found extent after a hole in which end lives?
 		 * Just back up to the previous extent, if so.
 		 */
-		if (got.br_startoff > bno) {
-			if (--lastx < 0)
-				break;
-			xfs_iext_get_extent(ifp, lastx, &got);
+		if (got.br_startoff > end &&
+		    !xfs_iext_prev_extent(ifp, &icur, &got)) {
+			done = true;
+			break;
 		}
 		/*
 		 * Is the last block of this extent before the range
 		 * we're supposed to delete?  If so, we're done.
 		 */
-		bno = XFS_FILEOFF_MIN(bno,
+		end = XFS_FILEOFF_MIN(end,
 			got.br_startoff + got.br_blockcount - 1);
-		if (bno < start)
+		if (end < start)
 			break;
 		/*
 		 * Then deal with the (possibly delayed) allocated space
@@ -5555,8 +5221,8 @@ __xfs_bunmapi(
 			if (!wasdel)
 				del.br_startblock += start - got.br_startoff;
 		}
-		if (del.br_startoff + del.br_blockcount > bno + 1)
-			del.br_blockcount = bno + 1 - del.br_startoff;
+		if (del.br_startoff + del.br_blockcount > end + 1)
+			del.br_blockcount = end + 1 - del.br_startoff;
 
 		/* How much can we safely unmap? */
 		if (max_len < del.br_blockcount) {
@@ -5582,13 +5248,13 @@ __xfs_bunmapi(
 				 * This piece is unwritten, or we're not
 				 * using unwritten extents.  Skip over it.
 				 */
-				ASSERT(bno >= mod);
-				bno -= mod > del.br_blockcount ?
+				ASSERT(end >= mod);
+				end -= mod > del.br_blockcount ?
 					del.br_blockcount : mod;
-				if (bno < got.br_startoff) {
-					if (--lastx >= 0)
-						xfs_bmbt_get_all(xfs_iext_get_ext(
-							ifp, lastx), &got);
+				if (end < got.br_startoff &&
+				    !xfs_iext_prev_extent(ifp, &icur, &got)) {
+					done = true;
+					break;
 				}
 				continue;
 			}
@@ -5609,7 +5275,7 @@ __xfs_bunmapi(
 			}
 			del.br_state = XFS_EXT_UNWRITTEN;
 			error = xfs_bmap_add_extent_unwritten_real(tp, ip,
-					whichfork, &lastx, &cur, &del,
+					whichfork, &icur, &cur, &del,
 					firstblock, dfops, &logflags);
 			if (error)
 				goto error0;
@@ -5634,10 +5300,13 @@ __xfs_bunmapi(
 				 * Can't make it unwritten.  There isn't
 				 * a full extent here so just skip it.
 				 */
-				ASSERT(bno >= del.br_blockcount);
-				bno -= del.br_blockcount;
-				if (got.br_startoff > bno && --lastx >= 0)
-					xfs_iext_get_extent(ifp, lastx, &got);
+				ASSERT(end >= del.br_blockcount);
+				end -= del.br_blockcount;
+				if (got.br_startoff > end &&
+				    !xfs_iext_prev_extent(ifp, &icur, &got)) {
+					done = true;
+					break;
+				}
 				continue;
 			} else if (del.br_state == XFS_EXT_UNWRITTEN) {
 				struct xfs_bmbt_irec	prev;
@@ -5648,8 +5317,8 @@ __xfs_bunmapi(
 				 * Unwrite the killed part of that one and
 				 * try again.
 				 */
-				ASSERT(lastx > 0);
-				xfs_iext_get_extent(ifp, lastx - 1, &prev);
+				if (!xfs_iext_prev_extent(ifp, &icur, &prev))
+					ASSERT(0);
 				ASSERT(prev.br_state == XFS_EXT_NORM);
 				ASSERT(!isnullstartblock(prev.br_startblock));
 				ASSERT(del.br_startblock ==
@@ -5661,9 +5330,8 @@ __xfs_bunmapi(
 					prev.br_startoff = start;
 				}
 				prev.br_state = XFS_EXT_UNWRITTEN;
-				lastx--;
 				error = xfs_bmap_add_extent_unwritten_real(tp,
-						ip, whichfork, &lastx, &cur,
+						ip, whichfork, &icur, &cur,
 						&prev, firstblock, dfops,
 						&logflags);
 				if (error)
@@ -5673,7 +5341,7 @@ __xfs_bunmapi(
 				ASSERT(del.br_state == XFS_EXT_NORM);
 				del.br_state = XFS_EXT_UNWRITTEN;
 				error = xfs_bmap_add_extent_unwritten_real(tp,
-						ip, whichfork, &lastx, &cur,
+						ip, whichfork, &icur, &cur,
 						&del, firstblock, dfops,
 						&logflags);
 				if (error)
@@ -5682,85 +5350,39 @@ __xfs_bunmapi(
 			}
 		}
 
-		/*
-		 * If it's the case where the directory code is running
-		 * with no block reservation, and the deleted block is in
-		 * the middle of its extent, and the resulting insert
-		 * of an extent would cause transformation to btree format,
-		 * then reject it.  The calling code will then swap
-		 * blocks around instead.
-		 * We have to do this now, rather than waiting for the
-		 * conversion to btree format, since the transaction
-		 * will be dirty.
-		 */
-		if (!wasdel && tp->t_blk_res == 0 &&
-		    XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS &&
-		    XFS_IFORK_NEXTENTS(ip, whichfork) >= /* Note the >= */
-			XFS_IFORK_MAXEXT(ip, whichfork) &&
-		    del.br_startoff > got.br_startoff &&
-		    del.br_startoff + del.br_blockcount <
-		    got.br_startoff + got.br_blockcount) {
-			error = -ENOSPC;
-			goto error0;
+		if (wasdel) {
+			error = xfs_bmap_del_extent_delay(ip, whichfork, &icur,
+					&got, &del);
+		} else {
+			error = xfs_bmap_del_extent_real(ip, tp, &icur, dfops,
+					cur, &del, &tmp_logflags, whichfork,
+					flags);
+			logflags |= tmp_logflags;
 		}
 
-		/*
-		 * Unreserve quota and update realtime free space, if
-		 * appropriate. If delayed allocation, update the inode delalloc
-		 * counter now and wait to update the sb counters as
-		 * xfs_bmap_del_extent() might need to borrow some blocks.
-		 */
-		if (wasdel) {
-			ASSERT(startblockval(del.br_startblock) > 0);
-			if (isrt) {
-				xfs_filblks_t rtexts;
-
-				rtexts = XFS_FSB_TO_B(mp, del.br_blockcount);
-				do_div(rtexts, mp->m_sb.sb_rextsize);
-				xfs_mod_frextents(mp, (int64_t)rtexts);
-				(void)xfs_trans_reserve_quota_nblks(NULL,
-					ip, -((long)del.br_blockcount), 0,
-					XFS_QMOPT_RES_RTBLKS);
-			} else {
-				(void)xfs_trans_reserve_quota_nblks(NULL,
-					ip, -((long)del.br_blockcount), 0,
-					XFS_QMOPT_RES_REGBLKS);
-			}
-			ip->i_delayed_blks -= del.br_blockcount;
-			if (cur)
-				cur->bc_private.b.flags |=
-					XFS_BTCUR_BPRV_WASDEL;
-		} else if (cur)
-			cur->bc_private.b.flags &= ~XFS_BTCUR_BPRV_WASDEL;
-
-		error = xfs_bmap_del_extent(ip, tp, &lastx, dfops, cur, &del,
-				&tmp_logflags, whichfork, flags);
-		logflags |= tmp_logflags;
 		if (error)
 			goto error0;
 
-		if (!isrt && wasdel)
-			xfs_mod_fdblocks(mp, (int64_t)del.br_blockcount, false);
-
 		max_len -= del.br_blockcount;
-		bno = del.br_startoff - 1;
+		end = del.br_startoff - 1;
 nodelete:
 		/*
 		 * If not done go on to the next (previous) record.
 		 */
-		if (bno != (xfs_fileoff_t)-1 && bno >= start) {
-			if (lastx >= 0) {
-				xfs_iext_get_extent(ifp, lastx, &got);
-				if (got.br_startoff > bno && --lastx >= 0)
-					xfs_iext_get_extent(ifp, lastx, &got);
+		if (end != (xfs_fileoff_t)-1 && end >= start) {
+			if (!xfs_iext_get_extent(ifp, &icur, &got) ||
+			    (got.br_startoff > end &&
+			     !xfs_iext_prev_extent(ifp, &icur, &got))) {
+				done = true;
+				break;
 			}
 			extno++;
 		}
 	}
-	if (bno == (xfs_fileoff_t)-1 || bno < start || lastx < 0)
+	if (done || end == (xfs_fileoff_t)-1 || end < start)
 		*rlen = 0;
 	else
-		*rlen = bno - start + 1;
+		*rlen = end - start + 1;
 
 	/*
 	 * Convert to a btree if necessary.
@@ -5878,14 +5500,13 @@ xfs_bmse_merge(
 	struct xfs_inode		*ip,
 	int				whichfork,
 	xfs_fileoff_t			shift,		/* shift fsb */
-	int				current_ext,	/* idx of gotp */
+	struct xfs_iext_cursor		*icur,
 	struct xfs_bmbt_irec		*got,		/* extent to shift */
 	struct xfs_bmbt_irec		*left,		/* preceding extent */
 	struct xfs_btree_cur		*cur,
 	int				*logflags,	/* output */
 	struct xfs_defer_ops		*dfops)
 {
-	struct xfs_ifork		*ifp = XFS_IFORK_PTR(ip, whichfork);
 	struct xfs_bmbt_irec		new;
 	xfs_filblks_t			blockcount;
 	int				error, i;
@@ -5913,8 +5534,7 @@ xfs_bmse_merge(
 	}
 
 	/* lookup and remove the extent to merge */
-	error = xfs_bmbt_lookup_eq(cur, got->br_startoff, got->br_startblock,
-				   got->br_blockcount, &i);
+	error = xfs_bmbt_lookup_eq(cur, got, &i);
 	if (error)
 		return error;
 	XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
@@ -5925,20 +5545,20 @@ xfs_bmse_merge(
 	XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
 
 	/* lookup and update size of the previous extent */
-	error = xfs_bmbt_lookup_eq(cur, left->br_startoff, left->br_startblock,
-				   left->br_blockcount, &i);
+	error = xfs_bmbt_lookup_eq(cur, left, &i);
 	if (error)
 		return error;
 	XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
 
-	error = xfs_bmbt_update(cur, new.br_startoff, new.br_startblock,
-			        new.br_blockcount, new.br_state);
+	error = xfs_bmbt_update(cur, &new);
 	if (error)
 		return error;
 
 done:
-	xfs_iext_update_extent(ifp, current_ext - 1, &new);
-	xfs_iext_remove(ip, current_ext, 1, 0);
+	xfs_iext_remove(ip, icur, 0);
+	xfs_iext_prev(XFS_IFORK_PTR(ip, whichfork), icur);
+	xfs_iext_update_extent(ip, xfs_bmap_fork_to_state(whichfork), icur,
+			&new);
 
 	/* update reverse mapping. rmap functions merge the rmaps for us */
 	error = xfs_rmap_unmap_extent(mp, dfops, ip, whichfork, got);
@@ -5949,183 +5569,83 @@ done:
 	return xfs_rmap_map_extent(mp, dfops, ip, whichfork, &new);
 }
 
-/*
- * Shift a single extent.
- */
-STATIC int
-xfs_bmse_shift_one(
-	struct xfs_inode		*ip,
-	int				whichfork,
-	xfs_fileoff_t			offset_shift_fsb,
-	int				*current_ext,
-	struct xfs_bmbt_irec		*got,
-	struct xfs_btree_cur		*cur,
-	int				*logflags,
-	enum shift_direction		direction,
-	struct xfs_defer_ops		*dfops)
+static int
+xfs_bmap_shift_update_extent(
+	struct xfs_inode	*ip,
+	int			whichfork,
+	struct xfs_iext_cursor	*icur,
+	struct xfs_bmbt_irec	*got,
+	struct xfs_btree_cur	*cur,
+	int			*logflags,
+	struct xfs_defer_ops	*dfops,
+	xfs_fileoff_t		startoff)
 {
-	struct xfs_ifork		*ifp;
-	struct xfs_mount		*mp;
-	xfs_fileoff_t			startoff;
-	struct xfs_bmbt_irec		adj_irec, new;
-	int				error;
-	int				i;
-	int				total_extents;
-
-	mp = ip->i_mount;
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	total_extents = xfs_iext_count(ifp);
-
-	/* delalloc extents should be prevented by caller */
-	XFS_WANT_CORRUPTED_RETURN(mp, !isnullstartblock(got->br_startblock));
-
-	if (direction == SHIFT_LEFT) {
-		startoff = got->br_startoff - offset_shift_fsb;
-
-		/*
-		 * Check for merge if we've got an extent to the left,
-		 * otherwise make sure there's enough room at the start
-		 * of the file for the shift.
-		 */
-		if (!*current_ext) {
-			if (got->br_startoff < offset_shift_fsb)
-				return -EINVAL;
-			goto update_current_ext;
-		}
-
-		/*
-		 * grab the left extent and check for a large enough hole.
-		 */
-		xfs_iext_get_extent(ifp, *current_ext - 1, &adj_irec);
-		if (startoff < adj_irec.br_startoff + adj_irec.br_blockcount)
-			return -EINVAL;
-
-		/* check whether to merge the extent or shift it down */
-		if (xfs_bmse_can_merge(&adj_irec, got, offset_shift_fsb)) {
-			return xfs_bmse_merge(ip, whichfork, offset_shift_fsb,
-					      *current_ext, got, &adj_irec,
-					      cur, logflags, dfops);
-		}
-	} else {
-		startoff = got->br_startoff + offset_shift_fsb;
-		/* nothing to move if this is the last extent */
-		if (*current_ext >= (total_extents - 1))
-			goto update_current_ext;
-
-		/*
-		 * If this is not the last extent in the file, make sure there
-		 * is enough room between current extent and next extent for
-		 * accommodating the shift.
-		 */
-		xfs_iext_get_extent(ifp, *current_ext + 1, &adj_irec);
-		if (startoff + got->br_blockcount > adj_irec.br_startoff)
-			return -EINVAL;
-
-		/*
-		 * Unlike a left shift (which involves a hole punch),
-		 * a right shift does not modify extent neighbors
-		 * in any way. We should never find mergeable extents
-		 * in this scenario. Check anyways and warn if we
-		 * encounter two extents that could be one.
-		 */
-		if (xfs_bmse_can_merge(got, &adj_irec, offset_shift_fsb))
-			WARN_ON_ONCE(1);
-	}
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_bmbt_irec	prev = *got;
+	int			error, i;
 
-	/*
-	 * Increment the extent index for the next iteration, update the start
-	 * offset of the in-core extent and update the btree if applicable.
-	 */
-update_current_ext:
 	*logflags |= XFS_ILOG_CORE;
 
-	new = *got;
-	new.br_startoff = startoff;
+	got->br_startoff = startoff;
 
 	if (cur) {
-		error = xfs_bmbt_lookup_eq(cur, got->br_startoff,
-				got->br_startblock, got->br_blockcount, &i);
+		error = xfs_bmbt_lookup_eq(cur, &prev, &i);
 		if (error)
 			return error;
 		XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
 
-		error = xfs_bmbt_update(cur, new.br_startoff,
-				new.br_startblock, new.br_blockcount,
-				new.br_state);
+		error = xfs_bmbt_update(cur, got);
 		if (error)
 			return error;
 	} else {
 		*logflags |= XFS_ILOG_DEXT;
 	}
 
-	xfs_iext_update_extent(ifp, *current_ext, &new);
-
-	if (direction == SHIFT_LEFT)
-		(*current_ext)++;
-	else
-		(*current_ext)--;
+	xfs_iext_update_extent(ip, xfs_bmap_fork_to_state(whichfork), icur,
+			got);
 
 	/* update reverse mapping */
-	error = xfs_rmap_unmap_extent(mp, dfops, ip, whichfork, got);
+	error = xfs_rmap_unmap_extent(mp, dfops, ip, whichfork, &prev);
 	if (error)
 		return error;
-	return xfs_rmap_map_extent(mp, dfops, ip, whichfork, &new);
+	return xfs_rmap_map_extent(mp, dfops, ip, whichfork, got);
 }
 
-/*
- * Shift extent records to the left/right to cover/create a hole.
- *
- * The maximum number of extents to be shifted in a single operation is
- * @num_exts. @stop_fsb specifies the file offset at which to stop shift and the
- * file offset where we've left off is returned in @next_fsb. @offset_shift_fsb
- * is the length by which each extent is shifted. If there is no hole to shift
- * the extents into, this will be considered invalid operation and we abort
- * immediately.
- */
 int
-xfs_bmap_shift_extents(
+xfs_bmap_collapse_extents(
 	struct xfs_trans	*tp,
 	struct xfs_inode	*ip,
 	xfs_fileoff_t		*next_fsb,
 	xfs_fileoff_t		offset_shift_fsb,
-	int			*done,
+	bool			*done,
 	xfs_fileoff_t		stop_fsb,
 	xfs_fsblock_t		*firstblock,
-	struct xfs_defer_ops	*dfops,
-	enum shift_direction	direction,
-	int			num_exts)
+	struct xfs_defer_ops	*dfops)
 {
-	struct xfs_btree_cur		*cur = NULL;
-	struct xfs_bmbt_irec            got;
-	struct xfs_mount		*mp = ip->i_mount;
-	struct xfs_ifork		*ifp;
-	xfs_extnum_t			nexts = 0;
-	xfs_extnum_t			current_ext;
-	xfs_extnum_t			total_extents;
-	xfs_extnum_t			stop_extent;
-	int				error = 0;
-	int				whichfork = XFS_DATA_FORK;
-	int				logflags = 0;
+	int			whichfork = XFS_DATA_FORK;
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_btree_cur	*cur = NULL;
+	struct xfs_bmbt_irec	got, prev;
+	struct xfs_iext_cursor	icur;
+	xfs_fileoff_t		new_startoff;
+	int			error = 0;
+	int			logflags = 0;
 
 	if (unlikely(XFS_TEST_ERROR(
 	    (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
 	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
 	     mp, XFS_ERRTAG_BMAPIFORMAT))) {
-		XFS_ERROR_REPORT("xfs_bmap_shift_extents",
-				 XFS_ERRLEVEL_LOW, mp);
+		XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_LOW, mp);
 		return -EFSCORRUPTED;
 	}
 
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -EIO;
 
-	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
-	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
-	ASSERT(direction == SHIFT_LEFT || direction == SHIFT_RIGHT);
+	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL));
 
-	ifp = XFS_IFORK_PTR(ip, whichfork);
 	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
-		/* Read in all the extents */
 		error = xfs_iread_extents(tp, ip, whichfork);
 		if (error)
 			return error;
@@ -6138,107 +5658,165 @@ xfs_bmap_shift_extents(
 		cur->bc_private.b.flags = 0;
 	}
 
-	/*
-	 * There may be delalloc extents in the data fork before the range we
-	 * are collapsing out, so we cannot use the count of real extents here.
-	 * Instead we have to calculate it from the incore fork.
-	 */
-	total_extents = xfs_iext_count(ifp);
-	if (total_extents == 0) {
-		*done = 1;
+	if (!xfs_iext_lookup_extent(ip, ifp, *next_fsb, &icur, &got)) {
+		*done = true;
 		goto del_cursor;
 	}
+	XFS_WANT_CORRUPTED_RETURN(mp, !isnullstartblock(got.br_startblock));
 
-	/*
-	 * In case of first right shift, we need to initialize next_fsb
-	 */
-	if (*next_fsb == NULLFSBLOCK) {
-		ASSERT(direction == SHIFT_RIGHT);
-
-		current_ext = total_extents - 1;
-		xfs_iext_get_extent(ifp, current_ext, &got);
-		if (stop_fsb > got.br_startoff) {
-			*done = 1;
+	new_startoff = got.br_startoff - offset_shift_fsb;
+	if (xfs_iext_peek_prev_extent(ifp, &icur, &prev)) {
+		if (new_startoff < prev.br_startoff + prev.br_blockcount) {
+			error = -EINVAL;
 			goto del_cursor;
 		}
-		*next_fsb = got.br_startoff;
+
+		if (xfs_bmse_can_merge(&prev, &got, offset_shift_fsb)) {
+			error = xfs_bmse_merge(ip, whichfork, offset_shift_fsb,
+					&icur, &got, &prev, cur, &logflags,
+					dfops);
+			if (error)
+				goto del_cursor;
+			goto done;
+		}
 	} else {
-		/*
-		 * Look up the extent index for the fsb where we start shifting. We can
-		 * henceforth iterate with current_ext as extent list changes are locked
-		 * out via ilock.
-		 *
-		 * If next_fsb lies in a hole beyond which there are no extents we are
-		 * done.
-		 */
-		if (!xfs_iext_lookup_extent(ip, ifp, *next_fsb, &current_ext,
-				&got)) {
-			*done = 1;
+		if (got.br_startoff < offset_shift_fsb) {
+			error = -EINVAL;
 			goto del_cursor;
 		}
 	}
 
-	/* Lookup the extent index at which we have to stop */
-	if (direction == SHIFT_RIGHT) {
-		struct xfs_bmbt_irec s;
+	error = xfs_bmap_shift_update_extent(ip, whichfork, &icur, &got, cur,
+			&logflags, dfops, new_startoff);
+	if (error)
+		goto del_cursor;
+
+done:
+	if (!xfs_iext_next_extent(ifp, &icur, &got)) {
+		*done = true;
+		goto del_cursor;
+	}
+
+	*next_fsb = got.br_startoff;
+del_cursor:
+	if (cur)
+		xfs_btree_del_cursor(cur,
+			error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
+	if (logflags)
+		xfs_trans_log_inode(tp, ip, logflags);
+	return error;
+}
+
+int
+xfs_bmap_insert_extents(
+	struct xfs_trans	*tp,
+	struct xfs_inode	*ip,
+	xfs_fileoff_t		*next_fsb,
+	xfs_fileoff_t		offset_shift_fsb,
+	bool			*done,
+	xfs_fileoff_t		stop_fsb,
+	xfs_fsblock_t		*firstblock,
+	struct xfs_defer_ops	*dfops)
+{
+	int			whichfork = XFS_DATA_FORK;
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_btree_cur	*cur = NULL;
+	struct xfs_bmbt_irec	got, next;
+	struct xfs_iext_cursor	icur;
+	xfs_fileoff_t		new_startoff;
+	int			error = 0;
+	int			logflags = 0;
+
+	if (unlikely(XFS_TEST_ERROR(
+	    (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
+	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
+	     mp, XFS_ERRTAG_BMAPIFORMAT))) {
+		XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_LOW, mp);
+		return -EFSCORRUPTED;
+	}
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return -EIO;
+
+	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL));
+
+	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+		error = xfs_iread_extents(tp, ip, whichfork);
+		if (error)
+			return error;
+	}
+
+	if (ifp->if_flags & XFS_IFBROOT) {
+		cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
+		cur->bc_private.b.firstblock = *firstblock;
+		cur->bc_private.b.dfops = dfops;
+		cur->bc_private.b.flags = 0;
+	}
 
-		xfs_iext_lookup_extent(ip, ifp, stop_fsb, &stop_extent, &s);
-		/* Make stop_extent exclusive of shift range */
-		stop_extent--;
-		if (current_ext <= stop_extent) {
-			error = -EIO;
+	if (*next_fsb == NULLFSBLOCK) {
+		xfs_iext_last(ifp, &icur);
+		if (!xfs_iext_get_extent(ifp, &icur, &got) ||
+		    stop_fsb > got.br_startoff) {
+			*done = true;
 			goto del_cursor;
 		}
 	} else {
-		stop_extent = total_extents;
-		if (current_ext >= stop_extent) {
-			error = -EIO;
+		if (!xfs_iext_lookup_extent(ip, ifp, *next_fsb, &icur, &got)) {
+			*done = true;
 			goto del_cursor;
 		}
 	}
+	XFS_WANT_CORRUPTED_RETURN(mp, !isnullstartblock(got.br_startblock));
 
-	while (nexts++ < num_exts) {
-		error = xfs_bmse_shift_one(ip, whichfork, offset_shift_fsb,
-					   &current_ext, &got, cur, &logflags,
-					   direction, dfops);
-		if (error)
+	if (stop_fsb >= got.br_startoff + got.br_blockcount) {
+		error = -EIO;
+		goto del_cursor;
+	}
+
+	new_startoff = got.br_startoff + offset_shift_fsb;
+	if (xfs_iext_peek_next_extent(ifp, &icur, &next)) {
+		if (new_startoff + got.br_blockcount > next.br_startoff) {
+			error = -EINVAL;
 			goto del_cursor;
-		/*
-		 * If there was an extent merge during the shift, the extent
-		 * count can change. Update the total and grade the next record.
-		 */
-		if (direction == SHIFT_LEFT) {
-			total_extents = xfs_iext_count(ifp);
-			stop_extent = total_extents;
 		}
 
-		if (current_ext == stop_extent) {
-			*done = 1;
-			*next_fsb = NULLFSBLOCK;
-			break;
-		}
-		xfs_iext_get_extent(ifp, current_ext, &got);
+		/*
+		 * Unlike a left shift (which involves a hole punch), a right
+		 * shift does not modify extent neighbors in any way.  We should
+		 * never find mergeable extents in this scenario.  Check anyways
+		 * and warn if we encounter two extents that could be one.
+		 */
+		if (xfs_bmse_can_merge(&got, &next, offset_shift_fsb))
+			WARN_ON_ONCE(1);
 	}
 
-	if (!*done)
-		*next_fsb = got.br_startoff;
+	error = xfs_bmap_shift_update_extent(ip, whichfork, &icur, &got, cur,
+			&logflags, dfops, new_startoff);
+	if (error)
+		goto del_cursor;
+
+	if (!xfs_iext_prev_extent(ifp, &icur, &got) ||
+	    stop_fsb >= got.br_startoff + got.br_blockcount) {
+		*done = true;
+		goto del_cursor;
+	}
 
+	*next_fsb = got.br_startoff;
 del_cursor:
 	if (cur)
 		xfs_btree_del_cursor(cur,
 			error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
-
 	if (logflags)
 		xfs_trans_log_inode(tp, ip, logflags);
-
 	return error;
 }
 
 /*
- * Splits an extent into two extents at split_fsb block such that it is
- * the first block of the current_ext. @current_ext is a target extent
- * to be split. @split_fsb is a block where the extents is split.
- * If split_fsb lies in a hole or the first block of extents, just return 0.
+ * Splits an extent into two extents at split_fsb block such that it is the
+ * first block of the current_ext. @ext is a target extent to be split.
+ * @split_fsb is a block where the extents is split.  If split_fsb lies in a
+ * hole or the first block of extents, just return 0.
  */
 STATIC int
 xfs_bmap_split_extent_at(
@@ -6255,7 +5833,7 @@ xfs_bmap_split_extent_at(
 	struct xfs_mount		*mp = ip->i_mount;
 	struct xfs_ifork		*ifp;
 	xfs_fsblock_t			gotblkcnt; /* new block count for got */
-	xfs_extnum_t			current_ext;
+	struct xfs_iext_cursor		icur;
 	int				error = 0;
 	int				logflags = 0;
 	int				i = 0;
@@ -6283,7 +5861,7 @@ xfs_bmap_split_extent_at(
 	/*
 	 * If there are not extents, or split_fsb lies in a hole we are done.
 	 */
-	if (!xfs_iext_lookup_extent(ip, ifp, split_fsb, &current_ext, &got) ||
+	if (!xfs_iext_lookup_extent(ip, ifp, split_fsb, &icur, &got) ||
 	    got.br_startoff >= split_fsb)
 		return 0;
 
@@ -6298,44 +5876,35 @@ xfs_bmap_split_extent_at(
 		cur->bc_private.b.firstblock = *firstfsb;
 		cur->bc_private.b.dfops = dfops;
 		cur->bc_private.b.flags = 0;
-		error = xfs_bmbt_lookup_eq(cur, got.br_startoff,
-				got.br_startblock,
-				got.br_blockcount,
-				&i);
+		error = xfs_bmbt_lookup_eq(cur, &got, &i);
 		if (error)
 			goto del_cursor;
 		XFS_WANT_CORRUPTED_GOTO(mp, i == 1, del_cursor);
 	}
 
 	got.br_blockcount = gotblkcnt;
-	xfs_iext_update_extent(ifp, current_ext, &got);
+	xfs_iext_update_extent(ip, xfs_bmap_fork_to_state(whichfork), &icur,
+			&got);
 
 	logflags = XFS_ILOG_CORE;
 	if (cur) {
-		error = xfs_bmbt_update(cur, got.br_startoff,
-				got.br_startblock,
-				got.br_blockcount,
-				got.br_state);
+		error = xfs_bmbt_update(cur, &got);
 		if (error)
 			goto del_cursor;
 	} else
 		logflags |= XFS_ILOG_DEXT;
 
 	/* Add new extent */
-	current_ext++;
-	xfs_iext_insert(ip, current_ext, 1, &new, 0);
+	xfs_iext_next(ifp, &icur);
+	xfs_iext_insert(ip, &icur, &new, 0);
 	XFS_IFORK_NEXT_SET(ip, whichfork,
 			   XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
 
 	if (cur) {
-		error = xfs_bmbt_lookup_eq(cur, new.br_startoff,
-				new.br_startblock, new.br_blockcount,
-				&i);
+		error = xfs_bmbt_lookup_eq(cur, &new, &i);
 		if (error)
 			goto del_cursor;
 		XFS_WANT_CORRUPTED_GOTO(mp, i == 0, del_cursor);
-		cur->bc_rec.b.br_state = new.br_state;
-
 		error = xfs_btree_insert(cur, &i);
 		if (error)
 			goto del_cursor;