Merge branch 'for-linus' of git://oss.sgi.com/xfs/xfs

* 'for-linus' of git://oss.sgi.com/xfs/xfs: (69 commits)
  xfs: add AIL pushing tracepoints
  xfs: put in missed fix for merge problem
  xfs: do not flush data workqueues in xfs_flush_buftarg
  xfs: remove XFS_bflush
  xfs: remove xfs_buf_target_name
  xfs: use xfs_ioerror_alert in xfs_buf_iodone_callbacks
  xfs: clean up xfs_ioerror_alert
  xfs: clean up buffer allocation
  xfs: remove buffers from the delwri list in xfs_buf_stale
  xfs: remove XFS_BUF_STALE and XFS_BUF_SUPER_STALE
  xfs: remove XFS_BUF_SET_VTYPE and XFS_BUF_SET_VTYPE_REF
  xfs: remove XFS_BUF_FINISH_IOWAIT
  xfs: remove xfs_get_buftarg_list
  xfs: fix buffer flushing during unmount
  xfs: optimize fsync on directories
  xfs: reduce the number of log forces from tail pushing
  xfs: Don't allocate new buffers on every call to _xfs_buf_find
  xfs: simplify xfs_trans_ijoin* again
  xfs: unlock the inode before log force in xfs_change_file_space
  xfs: unlock the inode before log force in xfs_fs_nfs_commit_metadata
  ...

fs/xfs/xfs_alloc.c

@@ -452,7 +452,7 @@ xfs_alloc_read_agfl(
 	if (error)
 		return error;
 	ASSERT(!xfs_buf_geterror(bp));
-	XFS_BUF_SET_VTYPE_REF(bp, B_FS_AGFL, XFS_AGFL_REF);
+	xfs_buf_set_ref(bp, XFS_AGFL_REF);
 	*bpp = bp;
 	return 0;
 }
@@ -2139,7 +2139,7 @@ xfs_read_agf(
 		xfs_trans_brelse(tp, *bpp);
 		return XFS_ERROR(EFSCORRUPTED);
 	}
-	XFS_BUF_SET_VTYPE_REF(*bpp, B_FS_AGF, XFS_AGF_REF);
+	xfs_buf_set_ref(*bpp, XFS_AGF_REF);
 	return 0;
 }
 

fs/xfs/xfs_aops.c

@@ -38,40 +38,6 @@
 #include <linux/pagevec.h>
 #include <linux/writeback.h>
 
-
-/*
- * Prime number of hash buckets since address is used as the key.
- */
-#define NVSYNC		37
-#define to_ioend_wq(v)	(&xfs_ioend_wq[((unsigned long)v) % NVSYNC])
-static wait_queue_head_t xfs_ioend_wq[NVSYNC];
-
-void __init
-xfs_ioend_init(void)
-{
-	int i;
-
-	for (i = 0; i < NVSYNC; i++)
-		init_waitqueue_head(&xfs_ioend_wq[i]);
-}
-
-void
-xfs_ioend_wait(
-	xfs_inode_t	*ip)
-{
-	wait_queue_head_t *wq = to_ioend_wq(ip);
-
-	wait_event(*wq, (atomic_read(&ip->i_iocount) == 0));
-}
-
-STATIC void
-xfs_ioend_wake(
-	xfs_inode_t	*ip)
-{
-	if (atomic_dec_and_test(&ip->i_iocount))
-		wake_up(to_ioend_wq(ip));
-}
-
 void
 xfs_count_page_state(
 	struct page		*page,
@@ -115,25 +81,20 @@ xfs_destroy_ioend(
 	xfs_ioend_t		*ioend)
 {
 	struct buffer_head	*bh, *next;
-	struct xfs_inode	*ip = XFS_I(ioend->io_inode);
 
 	for (bh = ioend->io_buffer_head; bh; bh = next) {
 		next = bh->b_private;
 		bh->b_end_io(bh, !ioend->io_error);
 	}
 
-	/*
-	 * Volume managers supporting multiple paths can send back ENODEV
-	 * when the final path disappears.  In this case continuing to fill
-	 * the page cache with dirty data which cannot be written out is
-	 * evil, so prevent that.
-	 */
-	if (unlikely(ioend->io_error == -ENODEV)) {
-		xfs_do_force_shutdown(ip->i_mount, SHUTDOWN_DEVICE_REQ,
-				      __FILE__, __LINE__);
+	if (ioend->io_iocb) {
+		if (ioend->io_isasync) {
+			aio_complete(ioend->io_iocb, ioend->io_error ?
+					ioend->io_error : ioend->io_result, 0);
+		}
+		inode_dio_done(ioend->io_inode);
 	}
 
-	xfs_ioend_wake(ip);
 	mempool_free(ioend, xfs_ioend_pool);
 }
 
@@ -155,6 +116,15 @@ xfs_ioend_new_eof(
 	return isize > ip->i_d.di_size ? isize : 0;
 }
 
+/*
+ * Fast and loose check if this write could update the on-disk inode size.
+ */
+static inline bool xfs_ioend_is_append(struct xfs_ioend *ioend)
+{
+	return ioend->io_offset + ioend->io_size >
+		XFS_I(ioend->io_inode)->i_d.di_size;
+}
+
 /*
  * Update on-disk file size now that data has been written to disk.  The
  * current in-memory file size is i_size.  If a write is beyond eof i_new_size
@@ -173,9 +143,6 @@ xfs_setfilesize(
 	xfs_inode_t		*ip = XFS_I(ioend->io_inode);
 	xfs_fsize_t		isize;
 
-	if (unlikely(ioend->io_error))
-		return 0;
-
 	if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL))
 		return EAGAIN;
 
@@ -192,6 +159,9 @@ xfs_setfilesize(
 
 /*
  * Schedule IO completion handling on the final put of an ioend.
+ *
+ * If there is no work to do we might as well call it a day and free the
+ * ioend right now.
  */
 STATIC void
 xfs_finish_ioend(
@@ -200,8 +170,10 @@ xfs_finish_ioend(
 	if (atomic_dec_and_test(&ioend->io_remaining)) {
 		if (ioend->io_type == IO_UNWRITTEN)
 			queue_work(xfsconvertd_workqueue, &ioend->io_work);
-		else
+		else if (xfs_ioend_is_append(ioend))
 			queue_work(xfsdatad_workqueue, &ioend->io_work);
+		else
+			xfs_destroy_ioend(ioend);
 	}
 }
 
@@ -216,17 +188,24 @@ xfs_end_io(
 	struct xfs_inode *ip = XFS_I(ioend->io_inode);
 	int		error = 0;
 
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
+		error = -EIO;
+		goto done;
+	}
+	if (ioend->io_error)
+		goto done;
+
 	/*
 	 * For unwritten extents we need to issue transactions to convert a
 	 * range to normal written extens after the data I/O has finished.
 	 */
-	if (ioend->io_type == IO_UNWRITTEN &&
-	    likely(!ioend->io_error && !XFS_FORCED_SHUTDOWN(ip->i_mount))) {
-
+	if (ioend->io_type == IO_UNWRITTEN) {
 		error = xfs_iomap_write_unwritten(ip, ioend->io_offset,
 						 ioend->io_size);
-		if (error)
-			ioend->io_error = error;
+		if (error) {
+			ioend->io_error = -error;
+			goto done;
+		}
 	}
 
 	/*
@@ -236,6 +215,7 @@ xfs_end_io(
 	error = xfs_setfilesize(ioend);
 	ASSERT(!error || error == EAGAIN);
 
+done:
 	/*
 	 * If we didn't complete processing of the ioend, requeue it to the
 	 * tail of the workqueue for another attempt later. Otherwise destroy
@@ -247,8 +227,6 @@ xfs_end_io(
 		/* ensure we don't spin on blocked ioends */
 		delay(1);
 	} else {
-		if (ioend->io_iocb)
-			aio_complete(ioend->io_iocb, ioend->io_result, 0);
 		xfs_destroy_ioend(ioend);
 	}
 }
@@ -285,13 +263,13 @@ xfs_alloc_ioend(
 	 * all the I/O from calling the completion routine too early.
 	 */
 	atomic_set(&ioend->io_remaining, 1);
+	ioend->io_isasync = 0;
 	ioend->io_error = 0;
 	ioend->io_list = NULL;
 	ioend->io_type = type;
 	ioend->io_inode = inode;
 	ioend->io_buffer_head = NULL;
 	ioend->io_buffer_tail = NULL;
-	atomic_inc(&XFS_I(ioend->io_inode)->i_iocount);
 	ioend->io_offset = 0;
 	ioend->io_size = 0;
 	ioend->io_iocb = NULL;
@@ -337,8 +315,8 @@ xfs_map_blocks(
 		count = mp->m_maxioffset - offset;
 	end_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + count);
 	offset_fsb = XFS_B_TO_FSBT(mp, offset);
-	error = xfs_bmapi(NULL, ip, offset_fsb, end_fsb - offset_fsb,
-			  bmapi_flags,  NULL, 0, imap, &nimaps, NULL);
+	error = xfs_bmapi_read(ip, offset_fsb, end_fsb - offset_fsb,
+				imap, &nimaps, bmapi_flags);
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
 
 	if (error)
@@ -551,7 +529,6 @@ xfs_cancel_ioend(
 			unlock_buffer(bh);
 		} while ((bh = next_bh) != NULL);
 
-		xfs_ioend_wake(XFS_I(ioend->io_inode));
 		mempool_free(ioend, xfs_ioend_pool);
 	} while ((ioend = next) != NULL);
 }
@@ -1161,8 +1138,8 @@ __xfs_get_blocks(
 	end_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + size);
 	offset_fsb = XFS_B_TO_FSBT(mp, offset);
 
-	error = xfs_bmapi(NULL, ip, offset_fsb, end_fsb - offset_fsb,
-			  XFS_BMAPI_ENTIRE,  NULL, 0, &imap, &nimaps, NULL);
+	error = xfs_bmapi_read(ip, offset_fsb, end_fsb - offset_fsb,
+				&imap, &nimaps, XFS_BMAPI_ENTIRE);
 	if (error)
 		goto out_unlock;
 
@@ -1300,7 +1277,6 @@ xfs_end_io_direct_write(
 	bool			is_async)
 {
 	struct xfs_ioend	*ioend = iocb->private;
-	struct inode		*inode = ioend->io_inode;
 
 	/*
 	 * blockdev_direct_IO can return an error even after the I/O
@@ -1311,28 +1287,17 @@ xfs_end_io_direct_write(
 
 	ioend->io_offset = offset;
 	ioend->io_size = size;
+	ioend->io_iocb = iocb;
+	ioend->io_result = ret;
 	if (private && size > 0)
 		ioend->io_type = IO_UNWRITTEN;
 
 	if (is_async) {
-		/*
-		 * If we are converting an unwritten extent we need to delay
-		 * the AIO completion until after the unwrittent extent
-		 * conversion has completed, otherwise do it ASAP.
-		 */
-		if (ioend->io_type == IO_UNWRITTEN) {
-			ioend->io_iocb = iocb;
-			ioend->io_result = ret;
-		} else {
-			aio_complete(iocb, ret, 0);
-		}
+		ioend->io_isasync = 1;
 		xfs_finish_ioend(ioend);
 	} else {
 		xfs_finish_ioend_sync(ioend);
 	}
-
-	/* XXX: probably should move into the real I/O completion handler */
-	inode_dio_done(inode);
 }
 
 STATIC ssize_t

fs/xfs/xfs_aops.h

@@ -47,6 +47,7 @@ typedef struct xfs_ioend {
 	unsigned int		io_type;	/* delalloc / unwritten */
 	int			io_error;	/* I/O error code */
 	atomic_t		io_remaining;	/* hold count */
+	unsigned int		io_isasync : 1;	/* needs aio_complete */
 	struct inode		*io_inode;	/* file being written to */
 	struct buffer_head	*io_buffer_head;/* buffer linked list head */
 	struct buffer_head	*io_buffer_tail;/* buffer linked list tail */
@@ -60,9 +61,6 @@ typedef struct xfs_ioend {
 extern const struct address_space_operations xfs_address_space_operations;
 extern int xfs_get_blocks(struct inode *, sector_t, struct buffer_head *, int);
 
-extern void xfs_ioend_init(void);
-extern void xfs_ioend_wait(struct xfs_inode *);
-
 extern void xfs_count_page_state(struct page *, int *, int *);
 
 #endif /* __XFS_AOPS_H__ */

fs/xfs/xfs_attr.c

@@ -319,7 +319,7 @@ xfs_attr_set_int(
 		return (error);
 	}
 
-	xfs_trans_ijoin(args.trans, dp);
+	xfs_trans_ijoin(args.trans, dp, 0);
 
 	/*
 	 * If the attribute list is non-existent or a shortform list,
@@ -389,7 +389,7 @@ xfs_attr_set_int(
 		 * a new one.  We need the inode to be in all transactions.
 		 */
 		if (committed)
-			xfs_trans_ijoin(args.trans, dp);
+			xfs_trans_ijoin(args.trans, dp, 0);
 
 		/*
 		 * Commit the leaf transformation.  We'll need another (linked)
@@ -537,7 +537,7 @@ xfs_attr_remove_int(xfs_inode_t *dp, struct xfs_name *name, int flags)
 	 * No need to make quota reservations here. We expect to release some
 	 * blocks not allocate in the common case.
 	 */
-	xfs_trans_ijoin(args.trans, dp);
+	xfs_trans_ijoin(args.trans, dp, 0);
 
 	/*
 	 * Decide on what work routines to call based on the inode size.
@@ -809,7 +809,7 @@ xfs_attr_inactive(xfs_inode_t *dp)
 	 * No need to make quota reservations here. We expect to release some
 	 * blocks, not allocate, in the common case.
 	 */
-	xfs_trans_ijoin(trans, dp);
+	xfs_trans_ijoin(trans, dp, 0);
 
 	/*
 	 * Decide on what work routines to call based on the inode size.
@@ -823,18 +823,6 @@ xfs_attr_inactive(xfs_inode_t *dp)
 	if (error)
 		goto out;
 
-	/*
-	 * Signal synchronous inactive transactions unless this is a
-	 * synchronous mount filesystem in which case we know that we're here
-	 * because we've been called out of xfs_inactive which means that the
-	 * last reference is gone and the unlink transaction has already hit
-	 * the disk so async inactive transactions are safe.
-	 */
-	if (!(mp->m_flags & XFS_MOUNT_WSYNC)) {
-		if (dp->i_d.di_anextents > 0)
-			xfs_trans_set_sync(trans);
-	}
-
 	error = xfs_itruncate_extents(&trans, dp, XFS_ATTR_FORK, 0);
 	if (error)
 		goto out;
@@ -973,7 +961,7 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
 		 * a new one.  We need the inode to be in all transactions.
 		 */
 		if (committed)
-			xfs_trans_ijoin(args->trans, dp);
+			xfs_trans_ijoin(args->trans, dp, 0);
 
 		/*
 		 * Commit the current trans (including the inode) and start
@@ -1075,7 +1063,7 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
 			 * in all transactions.
 			 */
 			if (committed)
-				xfs_trans_ijoin(args->trans, dp);
+				xfs_trans_ijoin(args->trans, dp, 0);
 		} else
 			xfs_da_buf_done(bp);
 
@@ -1149,7 +1137,7 @@ xfs_attr_leaf_removename(xfs_da_args_t *args)
 		 * a new one.  We need the inode to be in all transactions.
 		 */
 		if (committed)
-			xfs_trans_ijoin(args->trans, dp);
+			xfs_trans_ijoin(args->trans, dp, 0);
 	} else
 		xfs_da_buf_done(bp);
 	return(0);
@@ -1303,7 +1291,7 @@ restart:
 			 * in all transactions.
 			 */
 			if (committed)
-				xfs_trans_ijoin(args->trans, dp);
+				xfs_trans_ijoin(args->trans, dp, 0);
 
 			/*
 			 * Commit the node conversion and start the next
@@ -1340,7 +1328,7 @@ restart:
 		 * a new one.  We need the inode to be in all transactions.
 		 */
 		if (committed)
-			xfs_trans_ijoin(args->trans, dp);
+			xfs_trans_ijoin(args->trans, dp, 0);
 	} else {
 		/*
 		 * Addition succeeded, update Btree hashvals.
@@ -1452,7 +1440,7 @@ restart:
 			 * in all transactions.
 			 */
 			if (committed)
-				xfs_trans_ijoin(args->trans, dp);
+				xfs_trans_ijoin(args->trans, dp, 0);
 		}
 
 		/*
@@ -1584,7 +1572,7 @@ xfs_attr_node_removename(xfs_da_args_t *args)
 		 * a new one.  We need the inode to be in all transactions.
 		 */
 		if (committed)
-			xfs_trans_ijoin(args->trans, dp);
+			xfs_trans_ijoin(args->trans, dp, 0);
 
 		/*
 		 * Commit the Btree join operation and start a new trans.
@@ -1635,7 +1623,7 @@ xfs_attr_node_removename(xfs_da_args_t *args)
 			 * in all transactions.
 			 */
 			if (committed)
-				xfs_trans_ijoin(args->trans, dp);
+				xfs_trans_ijoin(args->trans, dp, 0);
 		} else
 			xfs_da_brelse(args->trans, bp);
 	}
@@ -1975,10 +1963,9 @@ xfs_attr_rmtval_get(xfs_da_args_t *args)
 	lblkno = args->rmtblkno;
 	while (valuelen > 0) {
 		nmap = ATTR_RMTVALUE_MAPSIZE;
-		error = xfs_bmapi(args->trans, args->dp, (xfs_fileoff_t)lblkno,
-				  args->rmtblkcnt,
-				  XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA,
-				  NULL, 0, map, &nmap, NULL);
+		error = xfs_bmapi_read(args->dp, (xfs_fileoff_t)lblkno,
+				       args->rmtblkcnt, map, &nmap,
+				       XFS_BMAPI_ATTRFORK);
 		if (error)
 			return(error);
 		ASSERT(nmap >= 1);
@@ -2052,10 +2039,9 @@ xfs_attr_rmtval_set(xfs_da_args_t *args)
 		 */
 		xfs_bmap_init(args->flist, args->firstblock);
 		nmap = 1;
-		error = xfs_bmapi(args->trans, dp, (xfs_fileoff_t)lblkno,
+		error = xfs_bmapi_write(args->trans, dp, (xfs_fileoff_t)lblkno,
 				  blkcnt,
-				  XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA |
-							XFS_BMAPI_WRITE,
+				  XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA,
 				  args->firstblock, args->total, &map, &nmap,
 				  args->flist);
 		if (!error) {
@@ -2074,7 +2060,7 @@ xfs_attr_rmtval_set(xfs_da_args_t *args)
 		 * a new one.  We need the inode to be in all transactions.
 		 */
 		if (committed)
-			xfs_trans_ijoin(args->trans, dp);
+			xfs_trans_ijoin(args->trans, dp, 0);
 
 		ASSERT(nmap == 1);
 		ASSERT((map.br_startblock != DELAYSTARTBLOCK) &&
@@ -2104,14 +2090,11 @@ xfs_attr_rmtval_set(xfs_da_args_t *args)
 		 */
 		xfs_bmap_init(args->flist, args->firstblock);
 		nmap = 1;
-		error = xfs_bmapi(NULL, dp, (xfs_fileoff_t)lblkno,
-				  args->rmtblkcnt,
-				  XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA,
-				  args->firstblock, 0, &map, &nmap,
-				  NULL);
-		if (error) {
+		error = xfs_bmapi_read(dp, (xfs_fileoff_t)lblkno,
+				       args->rmtblkcnt, &map, &nmap,
+				       XFS_BMAPI_ATTRFORK);
+		if (error)
 			return(error);
-		}
 		ASSERT(nmap == 1);
 		ASSERT((map.br_startblock != DELAYSTARTBLOCK) &&
 		       (map.br_startblock != HOLESTARTBLOCK));
@@ -2121,16 +2104,17 @@ xfs_attr_rmtval_set(xfs_da_args_t *args)
 
 		bp = xfs_buf_get(mp->m_ddev_targp, dblkno, blkcnt,
 				 XBF_LOCK | XBF_DONT_BLOCK);
-		ASSERT(!xfs_buf_geterror(bp));
-
+		if (!bp)
+			return ENOMEM;
 		tmp = (valuelen < XFS_BUF_SIZE(bp)) ? valuelen :
 							XFS_BUF_SIZE(bp);
 		xfs_buf_iomove(bp, 0, tmp, src, XBRW_WRITE);
 		if (tmp < XFS_BUF_SIZE(bp))
 			xfs_buf_zero(bp, tmp, XFS_BUF_SIZE(bp) - tmp);
-		if ((error = xfs_bwrite(mp, bp))) {/* GROT: NOTE: synchronous write */
-			return (error);
-		}
+		error = xfs_bwrite(bp);	/* GROT: NOTE: synchronous write */
+		xfs_buf_relse(bp);
+		if (error)
+			return error;
 		src += tmp;
 		valuelen -= tmp;
 
@@ -2166,16 +2150,12 @@ xfs_attr_rmtval_remove(xfs_da_args_t *args)
 		/*
 		 * Try to remember where we decided to put the value.
 		 */
-		xfs_bmap_init(args->flist, args->firstblock);
 		nmap = 1;
-		error = xfs_bmapi(NULL, args->dp, (xfs_fileoff_t)lblkno,
-					args->rmtblkcnt,
-					XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA,
-					args->firstblock, 0, &map, &nmap,
-					args->flist);
-		if (error) {
+		error = xfs_bmapi_read(args->dp, (xfs_fileoff_t)lblkno,
+				       args->rmtblkcnt, &map, &nmap,
+				       XFS_BMAPI_ATTRFORK);
+		if (error)
 			return(error);
-		}
 		ASSERT(nmap == 1);
 		ASSERT((map.br_startblock != DELAYSTARTBLOCK) &&
 		       (map.br_startblock != HOLESTARTBLOCK));
@@ -2188,8 +2168,7 @@ xfs_attr_rmtval_remove(xfs_da_args_t *args)
 		 */
 		bp = xfs_incore(mp->m_ddev_targp, dblkno, blkcnt, XBF_TRYLOCK);
 		if (bp) {
-			XFS_BUF_STALE(bp);
-			XFS_BUF_UNDELAYWRITE(bp);
+			xfs_buf_stale(bp);
 			xfs_buf_relse(bp);
 			bp = NULL;
 		}
@@ -2227,7 +2206,7 @@ xfs_attr_rmtval_remove(xfs_da_args_t *args)
 		 * a new one.  We need the inode to be in all transactions.
 		 */
 		if (committed)
-			xfs_trans_ijoin(args->trans, args->dp);
+			xfs_trans_ijoin(args->trans, args->dp, 0);
 
 		/*
 		 * Close out trans and start the next one in the chain.

fs/xfs/xfs_attr_leaf.c

@@ -2926,9 +2926,8 @@ xfs_attr_leaf_freextent(xfs_trans_t **trans, xfs_inode_t *dp,
 		 * Try to remember where we decided to put the value.
 		 */
 		nmap = 1;
-		error = xfs_bmapi(*trans, dp, (xfs_fileoff_t)tblkno, tblkcnt,
-					XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA,
-					NULL, 0, &map, &nmap, NULL);
+		error = xfs_bmapi_read(dp, (xfs_fileoff_t)tblkno, tblkcnt,
+				       &map, &nmap, XFS_BMAPI_ATTRFORK);
 		if (error) {
 			return(error);
 		}
@@ -2948,6 +2947,8 @@ xfs_attr_leaf_freextent(xfs_trans_t **trans, xfs_inode_t *dp,
 			bp = xfs_trans_get_buf(*trans,
 					dp->i_mount->m_ddev_targp,
 					dblkno, dblkcnt, XBF_LOCK);
+			if (!bp)
+				return ENOMEM;
 			xfs_trans_binval(*trans, bp);
 			/*
 			 * Roll to next transaction.

fs/xfs/xfs_bmap.c

@@ -50,17 +50,22 @@
 #include "xfs_trace.h"
 
 
-#ifdef DEBUG
-STATIC void
-xfs_bmap_check_leaf_extents(xfs_btree_cur_t *cur, xfs_inode_t *ip, int whichfork);
-#endif
-
 kmem_zone_t		*xfs_bmap_free_item_zone;
 
 /*
  * Prototypes for internal bmap routines.
  */
 
+#ifdef DEBUG
+STATIC void
+xfs_bmap_check_leaf_extents(
+	struct xfs_btree_cur	*cur,
+	struct xfs_inode	*ip,
+	int			whichfork);
+#else
+#define xfs_bmap_check_leaf_extents(cur, ip, whichfork)		do { } while (0)
+#endif
+
 
 /*
  * Called from xfs_bmap_add_attrfork to handle extents format files.
@@ -84,58 +89,6 @@ xfs_bmap_add_attrfork_local(
 	xfs_bmap_free_t		*flist,		/* blocks to free at commit */
 	int			*flags);	/* inode logging flags */
 
-/*
- * Called by xfs_bmap_add_extent to handle cases converting a delayed
- * allocation to a real allocation.
- */
-STATIC int				/* error */
-xfs_bmap_add_extent_delay_real(
-	struct xfs_trans	*tp,	/* transaction pointer */
-	xfs_inode_t		*ip,	/* incore inode pointer */
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
-	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_filblks_t		*dnew,	/* new delayed-alloc indirect blocks */
-	xfs_fsblock_t		*first,	/* pointer to firstblock variable */
-	xfs_bmap_free_t		*flist,	/* list of extents to be freed */
-	int			*logflagsp); /* inode logging flags */
-
-/*
- * Called by xfs_bmap_add_extent to handle cases converting a hole
- * to a delayed allocation.
- */
-STATIC int				/* error */
-xfs_bmap_add_extent_hole_delay(
-	xfs_inode_t		*ip,	/* incore inode pointer */
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
-	xfs_bmbt_irec_t		*new,	/* new data to add to file extents */
-	int			*logflagsp); /* inode logging flags */
-
-/*
- * Called by xfs_bmap_add_extent to handle cases converting a hole
- * to a real allocation.
- */
-STATIC int				/* error */
-xfs_bmap_add_extent_hole_real(
-	xfs_inode_t		*ip,	/* incore inode pointer */
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
-	xfs_btree_cur_t		*cur,	/* if null, not a btree */
-	xfs_bmbt_irec_t		*new,	/* new data to add to file extents */
-	int			*logflagsp, /* inode logging flags */
-	int			whichfork); /* data or attr fork */
-
-/*
- * Called by xfs_bmap_add_extent to handle cases converting an unwritten
- * allocation to a real allocation or vice versa.
- */
-STATIC int				/* error */
-xfs_bmap_add_extent_unwritten_real(
-	xfs_inode_t		*ip,	/* incore inode pointer */
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
-	xfs_btree_cur_t		**curp,	/* if *curp is null, not a btree */
-	xfs_bmbt_irec_t		*new,	/* new data to add to file extents */
-	int			*logflagsp); /* inode logging flags */
-
 /*
  * xfs_bmap_alloc is called by xfs_bmapi to allocate an extent for a file.
  * It figures out where to ask the underlying allocator to put the new extent.
@@ -214,19 +167,6 @@ xfs_bmap_search_extents(
 	xfs_bmbt_irec_t	*gotp,		/* out: extent entry found */
 	xfs_bmbt_irec_t	*prevp);	/* out: previous extent entry found */
 
-/*
- * Check the last inode extent to determine whether this allocation will result
- * in blocks being allocated at the end of the file. When we allocate new data
- * blocks at the end of the file which do not start at the previous data block,
- * we will try to align the new blocks at stripe unit boundaries.
- */
-STATIC int				/* error */
-xfs_bmap_isaeof(
-	xfs_inode_t	*ip,		/* incore inode pointer */
-	xfs_fileoff_t   off,		/* file offset in fsblocks */
-	int             whichfork,	/* data or attribute fork */
-	char		*aeof);		/* return value */
-
 /*
  * Compute the worst-case number of indirect blocks that will be used
  * for ip's delayed extent of length "len".
@@ -431,188 +371,13 @@ xfs_bmap_add_attrfork_local(
 }
 
 /*
- * Called by xfs_bmapi to update file extent records and the btree
- * after allocating space (or doing a delayed allocation).
- */
-STATIC int				/* error */
-xfs_bmap_add_extent(
-	struct xfs_trans	*tp,	/* transaction pointer */
-	xfs_inode_t		*ip,	/* incore inode pointer */
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
-	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 */
-	xfs_bmap_free_t		*flist,	/* list of extents to be freed */
-	int			*logflagsp, /* inode logging flags */
-	int			whichfork) /* data or attr fork */
-{
-	xfs_btree_cur_t		*cur;	/* btree cursor or null */
-	xfs_filblks_t		da_new; /* new count del alloc blocks used */
-	xfs_filblks_t		da_old; /* old count del alloc blocks used */
-	int			error;	/* error return value */
-	xfs_ifork_t		*ifp;	/* inode fork ptr */
-	int			logflags; /* returned value */
-	xfs_extnum_t		nextents; /* number of extents in file now */
-
-	XFS_STATS_INC(xs_add_exlist);
-
-	cur = *curp;
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
-	da_old = da_new = 0;
-	error = 0;
-
-	ASSERT(*idx >= 0);
-	ASSERT(*idx <= nextents);
-
-	/*
-	 * This is the first extent added to a new/empty file.
-	 * Special case this one, so other routines get to assume there are
-	 * already extents in the list.
-	 */
-	if (nextents == 0) {
-		xfs_iext_insert(ip, *idx, 1, new,
-				whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0);
-
-		ASSERT(cur == NULL);
-
-		if (!isnullstartblock(new->br_startblock)) {
-			XFS_IFORK_NEXT_SET(ip, whichfork, 1);
-			logflags = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
-		} else
-			logflags = 0;
-	}
-	/*
-	 * Any kind of new delayed allocation goes here.
-	 */
-	else if (isnullstartblock(new->br_startblock)) {
-		if (cur)
-			ASSERT((cur->bc_private.b.flags &
-				XFS_BTCUR_BPRV_WASDEL) == 0);
-		error = xfs_bmap_add_extent_hole_delay(ip, idx, new,
-						       &logflags);
-	}
-	/*
-	 * Real allocation off the end of the file.
-	 */
-	else if (*idx == nextents) {
-		if (cur)
-			ASSERT((cur->bc_private.b.flags &
-				XFS_BTCUR_BPRV_WASDEL) == 0);
-		error = xfs_bmap_add_extent_hole_real(ip, idx, cur, new,
-				&logflags, whichfork);
-	} else {
-		xfs_bmbt_irec_t	prev;	/* old extent at offset idx */
-
-		/*
-		 * Get the record referred to by idx.
-		 */
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &prev);
-		/*
-		 * If it's a real allocation record, and the new allocation ends
-		 * after the start of the referred to record, then we're filling
-		 * in a delayed or unwritten allocation with a real one, or
-		 * converting real back to unwritten.
-		 */
-		if (!isnullstartblock(new->br_startblock) &&
-		    new->br_startoff + new->br_blockcount > prev.br_startoff) {
-			if (prev.br_state != XFS_EXT_UNWRITTEN &&
-			    isnullstartblock(prev.br_startblock)) {
-				da_old = startblockval(prev.br_startblock);
-				if (cur)
-					ASSERT(cur->bc_private.b.flags &
-						XFS_BTCUR_BPRV_WASDEL);
-				error = xfs_bmap_add_extent_delay_real(tp, ip,
-						idx, &cur, new, &da_new,
-						first, flist, &logflags);
-			} else {
-				ASSERT(new->br_state == XFS_EXT_NORM ||
-				       new->br_state == XFS_EXT_UNWRITTEN);
-
-				error = xfs_bmap_add_extent_unwritten_real(ip,
-						idx, &cur, new, &logflags);
-				if (error)
-					goto done;
-			}
-		}
-		/*
-		 * Otherwise we're filling in a hole with an allocation.
-		 */
-		else {
-			if (cur)
-				ASSERT((cur->bc_private.b.flags &
-					XFS_BTCUR_BPRV_WASDEL) == 0);
-			error = xfs_bmap_add_extent_hole_real(ip, idx, cur,
-					new, &logflags, whichfork);
-		}
-	}
-
-	if (error)
-		goto done;
-	ASSERT(*curp == cur || *curp == NULL);
-
-	/*
-	 * Convert to a btree if necessary.
-	 */
-	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS &&
-	    XFS_IFORK_NEXTENTS(ip, whichfork) > ifp->if_ext_max) {
-		int	tmp_logflags;	/* partial log flag return val */
-
-		ASSERT(cur == NULL);
-		error = xfs_bmap_extents_to_btree(tp, ip, first,
-			flist, &cur, da_old > 0, &tmp_logflags, whichfork);
-		logflags |= tmp_logflags;
-		if (error)
-			goto done;
-	}
-	/*
-	 * Adjust for changes in reserved delayed indirect blocks.
-	 * Nothing to do for disk quotas here.
-	 */
-	if (da_old || da_new) {
-		xfs_filblks_t	nblks;
-
-		nblks = da_new;
-		if (cur)
-			nblks += cur->bc_private.b.allocated;
-		ASSERT(nblks <= da_old);
-		if (nblks < da_old)
-			xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS,
-				(int64_t)(da_old - nblks), 0);
-	}
-	/*
-	 * Clear out the allocated field, done with it now in any case.
-	 */
-	if (cur) {
-		cur->bc_private.b.allocated = 0;
-		*curp = cur;
-	}
-done:
-#ifdef DEBUG
-	if (!error)
-		xfs_bmap_check_leaf_extents(*curp, ip, whichfork);
-#endif
-	*logflagsp = logflags;
-	return error;
-}
-
-/*
- * Called by xfs_bmap_add_extent to handle cases converting a delayed
- * allocation to a real allocation.
+ * Convert a delayed allocation to a real allocation.
  */
 STATIC int				/* error */
 xfs_bmap_add_extent_delay_real(
-	struct xfs_trans	*tp,	/* transaction pointer */
-	xfs_inode_t		*ip,	/* incore inode pointer */
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
-	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_filblks_t		*dnew,	/* new delayed-alloc indirect blocks */
-	xfs_fsblock_t		*first,	/* pointer to firstblock variable */
-	xfs_bmap_free_t		*flist,	/* list of extents to be freed */
-	int			*logflagsp) /* inode logging flags */
+	struct xfs_bmalloca	*bma)
 {
-	xfs_btree_cur_t		*cur;	/* btree cursor */
+	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 */
@@ -623,10 +388,22 @@ xfs_bmap_add_extent_delay_real(
 					/* left is 0, right is 1, prev is 2 */
 	int			rval=0;	/* return value (logging flags) */
 	int			state = 0;/* state bits, accessed thru macros */
-	xfs_filblks_t		temp=0;	/* value for dnew calculations */
-	xfs_filblks_t		temp2=0;/* value for dnew calculations */
+	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 */
 
+	ifp = XFS_IFORK_PTR(bma->ip, XFS_DATA_FORK);
+
+	ASSERT(bma->idx >= 0);
+	ASSERT(bma->idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
+	ASSERT(!isnullstartblock(new->br_startblock));
+	ASSERT(!bma->cur ||
+	       (bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
+
+	XFS_STATS_INC(xs_add_exlist);
+
 #define	LEFT		r[0]
 #define	RIGHT		r[1]
 #define	PREV		r[2]
@@ -634,14 +411,15 @@ xfs_bmap_add_extent_delay_real(
 	/*
 	 * Set up a bunch of variables to make the tests simpler.
 	 */
-	cur = *curp;
-	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
-	ep = xfs_iext_get_ext(ifp, *idx);
+	ep = xfs_iext_get_ext(ifp, bma->idx);
 	xfs_bmbt_get_all(ep, &PREV);
 	new_endoff = new->br_startoff + new->br_blockcount;
 	ASSERT(PREV.br_startoff <= new->br_startoff);
 	ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);
 
+	da_old = startblockval(PREV.br_startblock);
+	da_new = 0;
+
 	/*
 	 * Set flags determining what part of the previous delayed allocation
 	 * extent is being replaced by a real allocation.
@@ -655,9 +433,9 @@ 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 (*idx > 0) {
+	if (bma->idx > 0) {
 		state |= BMAP_LEFT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &LEFT);
+		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &LEFT);
 
 		if (isnullstartblock(LEFT.br_startblock))
 			state |= BMAP_LEFT_DELAY;
@@ -675,9 +453,9 @@ 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 (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) {
+	if (bma->idx < bma->ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) {
 		state |= BMAP_RIGHT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx + 1), &RIGHT);
+		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx + 1), &RIGHT);
 
 		if (isnullstartblock(RIGHT.br_startblock))
 			state |= BMAP_RIGHT_DELAY;
@@ -708,38 +486,41 @@ 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.
 		 */
-		--*idx;
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
+		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(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		xfs_iext_remove(ip, *idx + 1, 2, state);
-		ip->i_d.di_nextents--;
-		if (cur == NULL)
+		xfs_iext_remove(bma->ip, bma->idx + 1, 2, state);
+		bma->ip->i_d.di_nextents--;
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
 					RIGHT.br_startblock,
-					RIGHT.br_blockcount, &i)))
+					RIGHT.br_blockcount, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_btree_delete(cur, &i)))
+			error = xfs_btree_delete(bma->cur, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_btree_decrement(cur, 0, &i)))
+			error = xfs_btree_decrement(bma->cur, 0, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
+			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
 					LEFT.br_startblock,
 					LEFT.br_blockcount +
 					PREV.br_blockcount +
-					RIGHT.br_blockcount, LEFT.br_state)))
+					RIGHT.br_blockcount, LEFT.br_state);
+			if (error)
 				goto done;
 		}
-		*dnew = 0;
 		break;
 
 	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
@@ -747,30 +528,31 @@ xfs_bmap_add_extent_delay_real(
 		 * Filling in all of a previously delayed allocation extent.
 		 * The left neighbor is contiguous, the right is not.
 		 */
-		--*idx;
+		bma->idx--;
 
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *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);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		xfs_iext_remove(ip, *idx + 1, 1, state);
-		if (cur == NULL)
+		xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur, LEFT.br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
 					LEFT.br_startblock, LEFT.br_blockcount,
-					&i)))
+					&i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
+			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
 					LEFT.br_startblock,
 					LEFT.br_blockcount +
-					PREV.br_blockcount, LEFT.br_state)))
+					PREV.br_blockcount, LEFT.br_state);
+			if (error)
 				goto done;
 		}
-		*dnew = 0;
 		break;
 
 	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
@@ -778,30 +560,30 @@ 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(ip, *idx, state, _THIS_IP_);
+		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(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		xfs_iext_remove(ip, *idx + 1, 1, state);
-		if (cur == NULL)
+		xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
 					RIGHT.br_startblock,
-					RIGHT.br_blockcount, &i)))
+					RIGHT.br_blockcount, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
+			error = xfs_bmbt_update(bma->cur, PREV.br_startoff,
 					new->br_startblock,
 					PREV.br_blockcount +
-					RIGHT.br_blockcount, PREV.br_state)))
+					RIGHT.br_blockcount, PREV.br_state);
+			if (error)
 				goto done;
 		}
-
-		*dnew = 0;
 		break;
 
 	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
@@ -810,27 +592,27 @@ xfs_bmap_add_extent_delay_real(
 		 * Neither the left nor right neighbors are contiguous with
 		 * the new one.
 		 */
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
 		xfs_bmbt_set_startblock(ep, new->br_startblock);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		ip->i_d.di_nextents++;
-		if (cur == NULL)
+		bma->ip->i_d.di_nextents++;
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
 					new->br_startblock, new->br_blockcount,
-					&i)))
+					&i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 0, done);
-			cur->bc_rec.b.br_state = XFS_EXT_NORM;
-			if ((error = xfs_btree_insert(cur, &i)))
+			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
+			error = xfs_btree_insert(bma->cur, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
 		}
-
-		*dnew = 0;
 		break;
 
 	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG:
@@ -838,39 +620,40 @@ 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(ip, *idx - 1, state, _THIS_IP_);
-		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx - 1),
+		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(ip, *idx - 1, state, _THIS_IP_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
 
 		temp = PREV.br_blockcount - new->br_blockcount;
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
 		xfs_bmbt_set_blockcount(ep, temp);
-		if (cur == NULL)
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur, LEFT.br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
 					LEFT.br_startblock, LEFT.br_blockcount,
-					&i)))
+					&i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
+			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
 					LEFT.br_startblock,
 					LEFT.br_blockcount +
 					new->br_blockcount,
-					LEFT.br_state)))
+					LEFT.br_state);
+			if (error)
 				goto done;
 		}
-		temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
 			startblockval(PREV.br_startblock));
-		xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
+		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		--*idx;
-		*dnew = temp;
+		bma->idx--;
 		break;
 
 	case BMAP_LEFT_FILLING:
@@ -878,43 +661,43 @@ 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(ip, *idx, state, _THIS_IP_);
+		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(ip, *idx, 1, new, state);
-		ip->i_d.di_nextents++;
-		if (cur == NULL)
+		xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
+		bma->ip->i_d.di_nextents++;
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
 					new->br_startblock, new->br_blockcount,
-					&i)))
+					&i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 0, done);
-			cur->bc_rec.b.br_state = XFS_EXT_NORM;
-			if ((error = xfs_btree_insert(cur, &i)))
+			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
+			error = xfs_btree_insert(bma->cur, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
 		}
-		if (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
-		    ip->i_d.di_nextents > ip->i_df.if_ext_max) {
-			error = xfs_bmap_extents_to_btree(tp, ip,
-					first, flist, &cur, 1, &tmp_rval,
-					XFS_DATA_FORK);
+		if (bma->ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
+		    bma->ip->i_d.di_nextents > bma->ip->i_df.if_ext_max) {
+			error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+					bma->firstblock, bma->flist,
+					&bma->cur, 1, &tmp_rval, XFS_DATA_FORK);
 			rval |= tmp_rval;
 			if (error)
 				goto done;
 		}
-		temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
 			startblockval(PREV.br_startblock) -
-			(cur ? cur->bc_private.b.allocated : 0));
-		ep = xfs_iext_get_ext(ifp, *idx + 1);
-		xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-		trace_xfs_bmap_post_update(ip, *idx + 1, state, _THIS_IP_);
-
-		*dnew = temp;
+			(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_);
 		break;
 
 	case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
@@ -923,38 +706,39 @@ xfs_bmap_add_extent_delay_real(
 		 * The right neighbor is contiguous with the new allocation.
 		 */
 		temp = PREV.br_blockcount - new->br_blockcount;
-		trace_xfs_bmap_pre_update(ip, *idx + 1, state, _THIS_IP_);
+		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, *idx + 1),
+		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(ip, *idx + 1, state, _THIS_IP_);
-		if (cur == NULL)
+		trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_DEXT;
 		else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
 					RIGHT.br_startblock,
-					RIGHT.br_blockcount, &i)))
+					RIGHT.br_blockcount, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_bmbt_update(cur, new->br_startoff,
+			error = xfs_bmbt_update(bma->cur, new->br_startoff,
 					new->br_startblock,
 					new->br_blockcount +
 					RIGHT.br_blockcount,
-					RIGHT.br_state)))
+					RIGHT.br_state);
+			if (error)
 				goto done;
 		}
 
-		temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
 			startblockval(PREV.br_startblock));
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
-		xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		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_);
 
-		++*idx;
-		*dnew = temp;
+		bma->idx++;
 		break;
 
 	case BMAP_RIGHT_FILLING:
@@ -963,42 +747,43 @@ xfs_bmap_add_extent_delay_real(
 		 * The right neighbor is not contiguous.
 		 */
 		temp = PREV.br_blockcount - new->br_blockcount;
-		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
 		xfs_bmbt_set_blockcount(ep, temp);
-		xfs_iext_insert(ip, *idx + 1, 1, new, state);
-		ip->i_d.di_nextents++;
-		if (cur == NULL)
+		xfs_iext_insert(bma->ip, bma->idx + 1, 1, new, state);
+		bma->ip->i_d.di_nextents++;
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
 					new->br_startblock, new->br_blockcount,
-					&i)))
+					&i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 0, done);
-			cur->bc_rec.b.br_state = XFS_EXT_NORM;
-			if ((error = xfs_btree_insert(cur, &i)))
+			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
+			error = xfs_btree_insert(bma->cur, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
 		}
-		if (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
-		    ip->i_d.di_nextents > ip->i_df.if_ext_max) {
-			error = xfs_bmap_extents_to_btree(tp, ip,
-				first, flist, &cur, 1, &tmp_rval,
-				XFS_DATA_FORK);
+		if (bma->ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
+		    bma->ip->i_d.di_nextents > bma->ip->i_df.if_ext_max) {
+			error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+				bma->firstblock, bma->flist, &bma->cur, 1,
+				&tmp_rval, XFS_DATA_FORK);
 			rval |= tmp_rval;
 			if (error)
 				goto done;
 		}
-		temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
 			startblockval(PREV.br_startblock) -
-			(cur ? cur->bc_private.b.allocated : 0));
-		ep = xfs_iext_get_ext(ifp, *idx);
-		xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+			(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_);
 
-		++*idx;
-		*dnew = temp;
+		bma->idx++;
 		break;
 
 	case 0:
@@ -1024,82 +809,65 @@ xfs_bmap_add_extent_delay_real(
 		 */
 		temp = new->br_startoff - PREV.br_startoff;
 		temp2 = PREV.br_startoff + PREV.br_blockcount - new_endoff;
-		trace_xfs_bmap_pre_update(ip, *idx, 0, _THIS_IP_);
+		trace_xfs_bmap_pre_update(bma->ip, bma->idx, 0, _THIS_IP_);
 		xfs_bmbt_set_blockcount(ep, temp);	/* truncate PREV */
 		LEFT = *new;
 		RIGHT.br_state = PREV.br_state;
 		RIGHT.br_startblock = nullstartblock(
-				(int)xfs_bmap_worst_indlen(ip, temp2));
+				(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(ip, *idx + 1, 2, &LEFT, state);
-		ip->i_d.di_nextents++;
-		if (cur == NULL)
+		xfs_iext_insert(bma->ip, bma->idx + 1, 2, &LEFT, state);
+		bma->ip->i_d.di_nextents++;
+		if (bma->cur == NULL)
 			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
 		else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
+			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
 					new->br_startblock, new->br_blockcount,
-					&i)))
+					&i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 0, done);
-			cur->bc_rec.b.br_state = XFS_EXT_NORM;
-			if ((error = xfs_btree_insert(cur, &i)))
+			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
+			error = xfs_btree_insert(bma->cur, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
 		}
-		if (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
-		    ip->i_d.di_nextents > ip->i_df.if_ext_max) {
-			error = xfs_bmap_extents_to_btree(tp, ip,
-					first, flist, &cur, 1, &tmp_rval,
-					XFS_DATA_FORK);
+		if (bma->ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
+		    bma->ip->i_d.di_nextents > bma->ip->i_df.if_ext_max) {
+			error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+					bma->firstblock, bma->flist, &bma->cur,
+					1, &tmp_rval, XFS_DATA_FORK);
 			rval |= tmp_rval;
 			if (error)
 				goto done;
 		}
-		temp = xfs_bmap_worst_indlen(ip, temp);
-		temp2 = xfs_bmap_worst_indlen(ip, temp2);
+		temp = xfs_bmap_worst_indlen(bma->ip, temp);
+		temp2 = xfs_bmap_worst_indlen(bma->ip, temp2);
 		diff = (int)(temp + temp2 - startblockval(PREV.br_startblock) -
-			(cur ? cur->bc_private.b.allocated : 0));
-		if (diff > 0 &&
-		    xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS,
-					     -((int64_t)diff), 0)) {
-			/*
-			 * Ick gross gag me with a spoon.
-			 */
-			ASSERT(0);	/* want to see if this ever happens! */
-			while (diff > 0) {
-				if (temp) {
-					temp--;
-					diff--;
-					if (!diff ||
-					    !xfs_icsb_modify_counters(ip->i_mount,
-						    XFS_SBS_FDBLOCKS,
-						    -((int64_t)diff), 0))
-						break;
-				}
-				if (temp2) {
-					temp2--;
-					diff--;
-					if (!diff ||
-					    !xfs_icsb_modify_counters(ip->i_mount,
-						    XFS_SBS_FDBLOCKS,
-						    -((int64_t)diff), 0))
-						break;
-				}
-			}
+			(bma->cur ? bma->cur->bc_private.b.allocated : 0));
+		if (diff > 0) {
+			error = xfs_icsb_modify_counters(bma->ip->i_mount,
+					XFS_SBS_FDBLOCKS,
+					-((int64_t)diff), 0);
+			ASSERT(!error);
+			if (error)
+				goto done;
 		}
-		ep = xfs_iext_get_ext(ifp, *idx);
+
+		ep = xfs_iext_get_ext(ifp, bma->idx);
 		xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-		trace_xfs_bmap_pre_update(ip, *idx + 2, state, _THIS_IP_);
-		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx + 2),
+		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(ip, *idx + 2, state, _THIS_IP_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx + 2, state, _THIS_IP_);
 
-		++*idx;
-		*dnew = temp + temp2;
+		bma->idx++;
+		da_new = temp + temp2;
 		break;
 
 	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
@@ -1114,9 +882,40 @@ xfs_bmap_add_extent_delay_real(
 		 */
 		ASSERT(0);
 	}
-	*curp = cur;
+
+	/* convert to a btree if necessary */
+	if (XFS_IFORK_FORMAT(bma->ip, XFS_DATA_FORK) == XFS_DINODE_FMT_EXTENTS &&
+	    XFS_IFORK_NEXTENTS(bma->ip, XFS_DATA_FORK) > ifp->if_ext_max) {
+		int	tmp_logflags;	/* partial log flag return val */
+
+		ASSERT(bma->cur == NULL);
+		error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+				bma->firstblock, bma->flist, &bma->cur,
+				da_old > 0, &tmp_logflags, XFS_DATA_FORK);
+		bma->logflags |= tmp_logflags;
+		if (error)
+			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;
+		ASSERT(temp <= da_old);
+		if (temp < da_old)
+			xfs_icsb_modify_counters(bma->ip->i_mount,
+					XFS_SBS_FDBLOCKS,
+					(int64_t)(da_old - temp), 0);
+	}
+
+	/* clear out the allocated field, done with it now in any case. */
+	if (bma->cur)
+		bma->cur->bc_private.b.allocated = 0;
+
+	xfs_bmap_check_leaf_extents(bma->cur, bma->ip, XFS_DATA_FORK);
 done:
-	*logflagsp = rval;
+	bma->logflags |= rval;
 	return error;
 #undef	LEFT
 #undef	RIGHT
@@ -1124,15 +923,17 @@ done:
 }
 
 /*
- * Called by xfs_bmap_add_extent to handle cases converting an unwritten
- * allocation to a real allocation or vice versa.
+ * Convert an unwritten allocation to a real allocation or vice versa.
  */
 STATIC int				/* error */
 xfs_bmap_add_extent_unwritten_real(
+	struct xfs_trans	*tp,
 	xfs_inode_t		*ip,	/* incore inode pointer */
 	xfs_extnum_t		*idx,	/* extent number to update/insert */
 	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 */
+	xfs_bmap_free_t		*flist,	/* list of extents to be freed */
 	int			*logflagsp) /* inode logging flags */
 {
 	xfs_btree_cur_t		*cur;	/* btree cursor */
@@ -1148,15 +949,25 @@ xfs_bmap_add_extent_unwritten_real(
 	int			rval=0;	/* return value (logging flags) */
 	int			state = 0;/* state bits, accessed thru macros */
 
+	*logflagsp = 0;
+
+	cur = *curp;
+	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+
+	ASSERT(*idx >= 0);
+	ASSERT(*idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
+	ASSERT(!isnullstartblock(new->br_startblock));
+
+	XFS_STATS_INC(xs_add_exlist);
+
 #define	LEFT		r[0]
 #define	RIGHT		r[1]
 #define	PREV		r[2]
+
 	/*
 	 * Set up a bunch of variables to make the tests simpler.
 	 */
 	error = 0;
-	cur = *curp;
-	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
 	ep = xfs_iext_get_ext(ifp, *idx);
 	xfs_bmbt_get_all(ep, &PREV);
 	newext = new->br_state;
@@ -1406,10 +1217,11 @@ xfs_bmap_add_extent_unwritten_real(
 				goto done;
 			if ((error = xfs_btree_decrement(cur, 0, &i)))
 				goto done;
-			if (xfs_bmbt_update(cur, LEFT.br_startoff,
+			error = xfs_bmbt_update(cur, LEFT.br_startoff,
 				LEFT.br_startblock,
 				LEFT.br_blockcount + new->br_blockcount,
-				LEFT.br_state))
+				LEFT.br_state);
+			if (error)
 				goto done;
 		}
 		break;
@@ -1607,9 +1419,29 @@ xfs_bmap_add_extent_unwritten_real(
 		 */
 		ASSERT(0);
 	}
-	*curp = cur;
+
+	/* convert to a btree if necessary */
+	if (XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) == XFS_DINODE_FMT_EXTENTS &&
+	    XFS_IFORK_NEXTENTS(ip, XFS_DATA_FORK) > ifp->if_ext_max) {
+		int	tmp_logflags;	/* partial log flag return val */
+
+		ASSERT(cur == NULL);
+		error = xfs_bmap_extents_to_btree(tp, ip, first, flist, &cur,
+				0, &tmp_logflags, XFS_DATA_FORK);
+		*logflagsp |= tmp_logflags;
+		if (error)
+			goto done;
+	}
+
+	/* clear out the allocated field, done with it now in any case. */
+	if (cur) {
+		cur->bc_private.b.allocated = 0;
+		*curp = cur;
+	}
+
+	xfs_bmap_check_leaf_extents(*curp, ip, XFS_DATA_FORK);
 done:
-	*logflagsp = rval;
+	*logflagsp |= rval;
 	return error;
 #undef	LEFT
 #undef	RIGHT
@@ -1617,16 +1449,13 @@ done:
 }
 
 /*
- * Called by xfs_bmap_add_extent to handle cases converting a hole
- * to a delayed allocation.
+ * Convert a hole to a delayed allocation.
  */
-/*ARGSUSED*/
-STATIC int				/* error */
+STATIC void
 xfs_bmap_add_extent_hole_delay(
 	xfs_inode_t		*ip,	/* incore inode pointer */
 	xfs_extnum_t		*idx,	/* extent number to update/insert */
-	xfs_bmbt_irec_t		*new,	/* new data to add to file extents */
-	int			*logflagsp) /* inode logging flags */
+	xfs_bmbt_irec_t		*new)	/* new data to add to file extents */
 {
 	xfs_ifork_t		*ifp;	/* inode fork pointer */
 	xfs_bmbt_irec_t		left;	/* left neighbor extent entry */
@@ -1761,23 +1590,17 @@ xfs_bmap_add_extent_hole_delay(
 		 * Nothing to do for disk quota accounting here.
 		 */
 	}
-	*logflagsp = 0;
-	return 0;
 }
 
 /*
- * Called by xfs_bmap_add_extent to handle cases converting a hole
- * to a real allocation.
+ * Convert a hole to a real allocation.
  */
 STATIC int				/* error */
 xfs_bmap_add_extent_hole_real(
-	xfs_inode_t		*ip,	/* incore inode pointer */
-	xfs_extnum_t		*idx,	/* extent number to update/insert */
-	xfs_btree_cur_t		*cur,	/* if null, not a btree */
-	xfs_bmbt_irec_t		*new,	/* new data to add to file extents */
-	int			*logflagsp, /* inode logging flags */
-	int			whichfork) /* data or attr fork */
+	struct xfs_bmalloca	*bma,
+	int			whichfork)
 {
+	struct xfs_bmbt_irec	*new = &bma->got;
 	int			error;	/* error return value */
 	int			i;	/* temp state */
 	xfs_ifork_t		*ifp;	/* inode fork pointer */
@@ -1786,19 +1609,26 @@ xfs_bmap_add_extent_hole_real(
 	int			rval=0;	/* return value (logging flags) */
 	int			state;	/* state bits, accessed thru macros */
 
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	ASSERT(*idx <= ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t));
-	state = 0;
+	ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+
+	ASSERT(bma->idx >= 0);
+	ASSERT(bma->idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
+	ASSERT(!isnullstartblock(new->br_startblock));
+	ASSERT(!bma->cur ||
+	       !(bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
 
+	XFS_STATS_INC(xs_add_exlist);
+
+	state = 0;
 	if (whichfork == XFS_ATTR_FORK)
 		state |= BMAP_ATTRFORK;
 
 	/*
 	 * Check and set flags if this segment has a left neighbor.
 	 */
-	if (*idx > 0) {
+	if (bma->idx > 0) {
 		state |= BMAP_LEFT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &left);
+		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &left);
 		if (isnullstartblock(left.br_startblock))
 			state |= BMAP_LEFT_DELAY;
 	}
@@ -1807,9 +1637,9 @@ 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 < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) {
+	if (bma->idx < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) {
 		state |= BMAP_RIGHT_VALID;
-		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &right);
+		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &right);
 		if (isnullstartblock(right.br_startblock))
 			state |= BMAP_RIGHT_DELAY;
 	}
@@ -1846,39 +1676,42 @@ 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),
+		--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 + new->br_blockcount +
 			right.br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		xfs_iext_remove(ip, *idx + 1, 1, state);
+		xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
 
-		XFS_IFORK_NEXT_SET(ip, whichfork,
-			XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
-		if (cur == NULL) {
+		XFS_IFORK_NEXT_SET(bma->ip, whichfork,
+			XFS_IFORK_NEXTENTS(bma->ip, whichfork) - 1);
+		if (bma->cur == NULL) {
 			rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
 		} 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(bma->cur, right.br_startoff,
+					right.br_startblock, right.br_blockcount,
+					&i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_btree_delete(cur, &i)))
+			error = xfs_btree_delete(bma->cur, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_btree_decrement(cur, 0, &i)))
+			error = xfs_btree_decrement(bma->cur, 0, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_bmbt_update(cur, left.br_startoff,
+			error = xfs_bmbt_update(bma->cur, left.br_startoff,
 					left.br_startblock,
 					left.br_blockcount +
 						new->br_blockcount +
 						right.br_blockcount,
-					left.br_state)))
+					left.br_state);
+			if (error)
 				goto done;
 		}
 		break;
@@ -1889,27 +1722,28 @@ 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),
+		--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 + new->br_blockcount);
-		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		if (cur == NULL) {
+		if (bma->cur == NULL) {
 			rval = xfs_ilog_fext(whichfork);
 		} else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur,
-					left.br_startoff,
-					left.br_startblock,
-					left.br_blockcount, &i)))
+			error = xfs_bmbt_lookup_eq(bma->cur, left.br_startoff,
+					left.br_startblock, left.br_blockcount,
+					&i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_bmbt_update(cur, left.br_startoff,
+			error = xfs_bmbt_update(bma->cur, left.br_startoff,
 					left.br_startblock,
 					left.br_blockcount +
 						new->br_blockcount,
-					left.br_state)))
+					left.br_state);
+			if (error)
 				goto done;
 		}
 		break;
@@ -1920,28 +1754,30 @@ 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),
+		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, bma->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_);
+		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
 
-		if (cur == NULL) {
+		if (bma->cur == NULL) {
 			rval = xfs_ilog_fext(whichfork);
 		} else {
 			rval = 0;
-			if ((error = xfs_bmbt_lookup_eq(cur,
+			error = xfs_bmbt_lookup_eq(bma->cur,
 					right.br_startoff,
 					right.br_startblock,
-					right.br_blockcount, &i)))
+					right.br_blockcount, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
-			if ((error = xfs_bmbt_update(cur, new->br_startoff,
+			error = xfs_bmbt_update(bma->cur, new->br_startoff,
 					new->br_startblock,
 					new->br_blockcount +
 						right.br_blockcount,
-					right.br_state)))
+					right.br_state);
+			if (error)
 				goto done;
 		}
 		break;
@@ -1952,28 +1788,50 @@ xfs_bmap_add_extent_hole_real(
 		 * real allocation.
 		 * Insert a new entry.
 		 */
-		xfs_iext_insert(ip, *idx, 1, new, state);
-		XFS_IFORK_NEXT_SET(ip, whichfork,
-			XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
-		if (cur == NULL) {
+		xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
+		XFS_IFORK_NEXT_SET(bma->ip, whichfork,
+			XFS_IFORK_NEXTENTS(bma->ip, whichfork) + 1);
+		if (bma->cur == NULL) {
 			rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
 		} else {
 			rval = XFS_ILOG_CORE;
-			if ((error = xfs_bmbt_lookup_eq(cur,
+			error = xfs_bmbt_lookup_eq(bma->cur,
 					new->br_startoff,
 					new->br_startblock,
-					new->br_blockcount, &i)))
+					new->br_blockcount, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 0, done);
-			cur->bc_rec.b.br_state = new->br_state;
-			if ((error = xfs_btree_insert(cur, &i)))
+			bma->cur->bc_rec.b.br_state = new->br_state;
+			error = xfs_btree_insert(bma->cur, &i);
+			if (error)
 				goto done;
 			XFS_WANT_CORRUPTED_GOTO(i == 1, done);
 		}
 		break;
 	}
+
+	/* convert to a btree if necessary */
+	if (XFS_IFORK_FORMAT(bma->ip, whichfork) == XFS_DINODE_FMT_EXTENTS &&
+	    XFS_IFORK_NEXTENTS(bma->ip, whichfork) > ifp->if_ext_max) {
+		int	tmp_logflags;	/* partial log flag return val */
+
+		ASSERT(bma->cur == NULL);
+		error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+				bma->firstblock, bma->flist, &bma->cur,
+				0, &tmp_logflags, whichfork);
+		bma->logflags |= tmp_logflags;
+		if (error)
+			goto done;
+	}
+
+	/* clear out the allocated field, done with it now in any case. */
+	if (bma->cur)
+		bma->cur->bc_private.b.allocated = 0;
+
+	xfs_bmap_check_leaf_extents(bma->cur, bma->ip, whichfork);
 done:
-	*logflagsp = rval;
+	bma->logflags |= rval;
 	return error;
 }
 
@@ -2160,26 +2018,26 @@ xfs_bmap_adjacent(
 		XFS_FSB_TO_AGBNO(mp, x) < mp->m_sb.sb_agblocks)
 
 	mp = ap->ip->i_mount;
-	nullfb = ap->firstblock == NULLFSBLOCK;
+	nullfb = *ap->firstblock == NULLFSBLOCK;
 	rt = XFS_IS_REALTIME_INODE(ap->ip) && ap->userdata;
-	fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, ap->firstblock);
+	fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
 	/*
 	 * If allocating at eof, and there's a previous real block,
 	 * try to use its last block as our starting point.
 	 */
-	if (ap->eof && ap->prevp->br_startoff != NULLFILEOFF &&
-	    !isnullstartblock(ap->prevp->br_startblock) &&
-	    ISVALID(ap->prevp->br_startblock + ap->prevp->br_blockcount,
-		    ap->prevp->br_startblock)) {
-		ap->rval = ap->prevp->br_startblock + ap->prevp->br_blockcount;
+	if (ap->eof && ap->prev.br_startoff != NULLFILEOFF &&
+	    !isnullstartblock(ap->prev.br_startblock) &&
+	    ISVALID(ap->prev.br_startblock + ap->prev.br_blockcount,
+		    ap->prev.br_startblock)) {
+		ap->blkno = ap->prev.br_startblock + ap->prev.br_blockcount;
 		/*
 		 * Adjust for the gap between prevp and us.
 		 */
-		adjust = ap->off -
-			(ap->prevp->br_startoff + ap->prevp->br_blockcount);
+		adjust = ap->offset -
+			(ap->prev.br_startoff + ap->prev.br_blockcount);
 		if (adjust &&
-		    ISVALID(ap->rval + adjust, ap->prevp->br_startblock))
-			ap->rval += adjust;
+		    ISVALID(ap->blkno + adjust, ap->prev.br_startblock))
+			ap->blkno += adjust;
 	}
 	/*
 	 * If not at eof, then compare the two neighbor blocks.
@@ -2196,17 +2054,17 @@ xfs_bmap_adjacent(
 		 * If there's a previous (left) block, select a requested
 		 * start block based on it.
 		 */
-		if (ap->prevp->br_startoff != NULLFILEOFF &&
-		    !isnullstartblock(ap->prevp->br_startblock) &&
-		    (prevbno = ap->prevp->br_startblock +
-			       ap->prevp->br_blockcount) &&
-		    ISVALID(prevbno, ap->prevp->br_startblock)) {
+		if (ap->prev.br_startoff != NULLFILEOFF &&
+		    !isnullstartblock(ap->prev.br_startblock) &&
+		    (prevbno = ap->prev.br_startblock +
+			       ap->prev.br_blockcount) &&
+		    ISVALID(prevbno, ap->prev.br_startblock)) {
 			/*
 			 * Calculate gap to end of previous block.
 			 */
-			adjust = prevdiff = ap->off -
-				(ap->prevp->br_startoff +
-				 ap->prevp->br_blockcount);
+			adjust = prevdiff = ap->offset -
+				(ap->prev.br_startoff +
+				 ap->prev.br_blockcount);
 			/*
 			 * Figure the startblock based on the previous block's
 			 * end and the gap size.
@@ -2215,9 +2073,9 @@ xfs_bmap_adjacent(
 			 * allocating, or using it gives us an invalid block
 			 * number, then just use the end of the previous block.
 			 */
-			if (prevdiff <= XFS_ALLOC_GAP_UNITS * ap->alen &&
+			if (prevdiff <= XFS_ALLOC_GAP_UNITS * ap->length &&
 			    ISVALID(prevbno + prevdiff,
-				    ap->prevp->br_startblock))
+				    ap->prev.br_startblock))
 				prevbno += adjust;
 			else
 				prevdiff += adjust;
@@ -2238,16 +2096,16 @@ xfs_bmap_adjacent(
 		 * If there's a following (right) block, select a requested
 		 * start block based on it.
 		 */
-		if (!isnullstartblock(ap->gotp->br_startblock)) {
+		if (!isnullstartblock(ap->got.br_startblock)) {
 			/*
 			 * Calculate gap to start of next block.
 			 */
-			adjust = gotdiff = ap->gotp->br_startoff - ap->off;
+			adjust = gotdiff = ap->got.br_startoff - ap->offset;
 			/*
 			 * Figure the startblock based on the next block's
 			 * start and the gap size.
 			 */
-			gotbno = ap->gotp->br_startblock;
+			gotbno = ap->got.br_startblock;
 			/*
 			 * Heuristic!
 			 * If the gap is large relative to the piece we're
@@ -2255,12 +2113,12 @@ xfs_bmap_adjacent(
 			 * number, then just use the start of the next block
 			 * offset by our length.
 			 */
-			if (gotdiff <= XFS_ALLOC_GAP_UNITS * ap->alen &&
+			if (gotdiff <= XFS_ALLOC_GAP_UNITS * ap->length &&
 			    ISVALID(gotbno - gotdiff, gotbno))
 				gotbno -= adjust;
-			else if (ISVALID(gotbno - ap->alen, gotbno)) {
-				gotbno -= ap->alen;
-				gotdiff += adjust - ap->alen;
+			else if (ISVALID(gotbno - ap->length, gotbno)) {
+				gotbno -= ap->length;
+				gotdiff += adjust - ap->length;
 			} else
 				gotdiff += adjust;
 			/*
@@ -2278,14 +2136,14 @@ xfs_bmap_adjacent(
 			gotbno = NULLFSBLOCK;
 		/*
 		 * If both valid, pick the better one, else the only good
-		 * one, else ap->rval is already set (to 0 or the inode block).
+		 * one, else ap->blkno is already set (to 0 or the inode block).
 		 */
 		if (prevbno != NULLFSBLOCK && gotbno != NULLFSBLOCK)
-			ap->rval = prevdiff <= gotdiff ? prevbno : gotbno;
+			ap->blkno = prevdiff <= gotdiff ? prevbno : gotbno;
 		else if (prevbno != NULLFSBLOCK)
-			ap->rval = prevbno;
+			ap->blkno = prevbno;
 		else if (gotbno != NULLFSBLOCK)
-			ap->rval = gotbno;
+			ap->blkno = gotbno;
 	}
 #undef ISVALID
 }
@@ -2305,24 +2163,24 @@ xfs_bmap_rtalloc(
 	mp = ap->ip->i_mount;
 	align = xfs_get_extsz_hint(ap->ip);
 	prod = align / mp->m_sb.sb_rextsize;
-	error = xfs_bmap_extsize_align(mp, ap->gotp, ap->prevp,
+	error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev,
 					align, 1, ap->eof, 0,
-					ap->conv, &ap->off, &ap->alen);
+					ap->conv, &ap->offset, &ap->length);
 	if (error)
 		return error;
-	ASSERT(ap->alen);
-	ASSERT(ap->alen % mp->m_sb.sb_rextsize == 0);
+	ASSERT(ap->length);
+	ASSERT(ap->length % mp->m_sb.sb_rextsize == 0);
 
 	/*
 	 * If the offset & length are not perfectly aligned
 	 * then kill prod, it will just get us in trouble.
 	 */
-	if (do_mod(ap->off, align) || ap->alen % align)
+	if (do_mod(ap->offset, align) || ap->length % align)
 		prod = 1;
 	/*
 	 * Set ralen to be the actual requested length in rtextents.
 	 */
-	ralen = ap->alen / mp->m_sb.sb_rextsize;
+	ralen = ap->length / mp->m_sb.sb_rextsize;
 	/*
 	 * If the old value was close enough to MAXEXTLEN that
 	 * we rounded up to it, cut it back so it's valid again.
@@ -2337,21 +2195,21 @@ xfs_bmap_rtalloc(
 	 * Lock out other modifications to the RT bitmap inode.
 	 */
 	xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin_ref(ap->tp, mp->m_rbmip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(ap->tp, mp->m_rbmip, XFS_ILOCK_EXCL);
 
 	/*
 	 * If it's an allocation to an empty file at offset 0,
 	 * pick an extent that will space things out in the rt area.
 	 */
-	if (ap->eof && ap->off == 0) {
+	if (ap->eof && ap->offset == 0) {
 		xfs_rtblock_t uninitialized_var(rtx); /* realtime extent no */
 
 		error = xfs_rtpick_extent(mp, ap->tp, ralen, &rtx);
 		if (error)
 			return error;
-		ap->rval = rtx * mp->m_sb.sb_rextsize;
+		ap->blkno = rtx * mp->m_sb.sb_rextsize;
 	} else {
-		ap->rval = 0;
+		ap->blkno = 0;
 	}
 
 	xfs_bmap_adjacent(ap);
@@ -2359,23 +2217,23 @@ xfs_bmap_rtalloc(
 	/*
 	 * Realtime allocation, done through xfs_rtallocate_extent.
 	 */
-	atype = ap->rval == 0 ?  XFS_ALLOCTYPE_ANY_AG : XFS_ALLOCTYPE_NEAR_BNO;
-	do_div(ap->rval, mp->m_sb.sb_rextsize);
-	rtb = ap->rval;
-	ap->alen = ralen;
-	if ((error = xfs_rtallocate_extent(ap->tp, ap->rval, 1, ap->alen,
+	atype = ap->blkno == 0 ?  XFS_ALLOCTYPE_ANY_AG : XFS_ALLOCTYPE_NEAR_BNO;
+	do_div(ap->blkno, mp->m_sb.sb_rextsize);
+	rtb = ap->blkno;
+	ap->length = ralen;
+	if ((error = xfs_rtallocate_extent(ap->tp, ap->blkno, 1, ap->length,
 				&ralen, atype, ap->wasdel, prod, &rtb)))
 		return error;
 	if (rtb == NULLFSBLOCK && prod > 1 &&
-	    (error = xfs_rtallocate_extent(ap->tp, ap->rval, 1,
-					   ap->alen, &ralen, atype,
+	    (error = xfs_rtallocate_extent(ap->tp, ap->blkno, 1,
+					   ap->length, &ralen, atype,
 					   ap->wasdel, 1, &rtb)))
 		return error;
-	ap->rval = rtb;
-	if (ap->rval != NULLFSBLOCK) {
-		ap->rval *= mp->m_sb.sb_rextsize;
+	ap->blkno = rtb;
+	if (ap->blkno != NULLFSBLOCK) {
+		ap->blkno *= mp->m_sb.sb_rextsize;
 		ralen *= mp->m_sb.sb_rextsize;
-		ap->alen = ralen;
+		ap->length = ralen;
 		ap->ip->i_d.di_nblocks += ralen;
 		xfs_trans_log_inode(ap->tp, ap->ip, XFS_ILOG_CORE);
 		if (ap->wasdel)
@@ -2388,7 +2246,7 @@ xfs_bmap_rtalloc(
 			ap->wasdel ? XFS_TRANS_DQ_DELRTBCOUNT :
 					XFS_TRANS_DQ_RTBCOUNT, (long) ralen);
 	} else {
-		ap->alen = 0;
+		ap->length = 0;
 	}
 	return 0;
 }
@@ -2503,7 +2361,7 @@ xfs_bmap_btalloc_nullfb(
 	 * AG as the stream may have moved.
 	 */
 	if (xfs_inode_is_filestream(ap->ip))
-		ap->rval = args->fsbno = XFS_AGB_TO_FSB(mp, ag, 0);
+		ap->blkno = args->fsbno = XFS_AGB_TO_FSB(mp, ag, 0);
 
 	return 0;
 }
@@ -2528,52 +2386,52 @@ xfs_bmap_btalloc(
 	mp = ap->ip->i_mount;
 	align = ap->userdata ? xfs_get_extsz_hint(ap->ip) : 0;
 	if (unlikely(align)) {
-		error = xfs_bmap_extsize_align(mp, ap->gotp, ap->prevp,
+		error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev,
 						align, 0, ap->eof, 0, ap->conv,
-						&ap->off, &ap->alen);
+						&ap->offset, &ap->length);
 		ASSERT(!error);
-		ASSERT(ap->alen);
+		ASSERT(ap->length);
 	}
-	nullfb = ap->firstblock == NULLFSBLOCK;
-	fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, ap->firstblock);
+	nullfb = *ap->firstblock == NULLFSBLOCK;
+	fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
 	if (nullfb) {
 		if (ap->userdata && xfs_inode_is_filestream(ap->ip)) {
 			ag = xfs_filestream_lookup_ag(ap->ip);
 			ag = (ag != NULLAGNUMBER) ? ag : 0;
-			ap->rval = XFS_AGB_TO_FSB(mp, ag, 0);
+			ap->blkno = XFS_AGB_TO_FSB(mp, ag, 0);
 		} else {
-			ap->rval = XFS_INO_TO_FSB(mp, ap->ip->i_ino);
+			ap->blkno = XFS_INO_TO_FSB(mp, ap->ip->i_ino);
 		}
 	} else
-		ap->rval = ap->firstblock;
+		ap->blkno = *ap->firstblock;
 
 	xfs_bmap_adjacent(ap);
 
 	/*
-	 * If allowed, use ap->rval; otherwise must use firstblock since
+	 * If allowed, use ap->blkno; otherwise must use firstblock since
 	 * it's in the right allocation group.
 	 */
-	if (nullfb || XFS_FSB_TO_AGNO(mp, ap->rval) == fb_agno)
+	if (nullfb || XFS_FSB_TO_AGNO(mp, ap->blkno) == fb_agno)
 		;
 	else
-		ap->rval = ap->firstblock;
+		ap->blkno = *ap->firstblock;
 	/*
 	 * Normal allocation, done through xfs_alloc_vextent.
 	 */
 	tryagain = isaligned = 0;
 	args.tp = ap->tp;
 	args.mp = mp;
-	args.fsbno = ap->rval;
+	args.fsbno = ap->blkno;
 
 	/* Trim the allocation back to the maximum an AG can fit. */
-	args.maxlen = MIN(ap->alen, XFS_ALLOC_AG_MAX_USABLE(mp));
-	args.firstblock = ap->firstblock;
+	args.maxlen = MIN(ap->length, XFS_ALLOC_AG_MAX_USABLE(mp));
+	args.firstblock = *ap->firstblock;
 	blen = 0;
 	if (nullfb) {
 		error = xfs_bmap_btalloc_nullfb(ap, &args, &blen);
 		if (error)
 			return error;
-	} else if (ap->low) {
+	} else if (ap->flist->xbf_low) {
 		if (xfs_inode_is_filestream(ap->ip))
 			args.type = XFS_ALLOCTYPE_FIRST_AG;
 		else
@@ -2587,14 +2445,14 @@ xfs_bmap_btalloc(
 	/* apply extent size hints if obtained earlier */
 	if (unlikely(align)) {
 		args.prod = align;
-		if ((args.mod = (xfs_extlen_t)do_mod(ap->off, args.prod)))
+		if ((args.mod = (xfs_extlen_t)do_mod(ap->offset, args.prod)))
 			args.mod = (xfs_extlen_t)(args.prod - args.mod);
 	} else if (mp->m_sb.sb_blocksize >= PAGE_CACHE_SIZE) {
 		args.prod = 1;
 		args.mod = 0;
 	} else {
 		args.prod = PAGE_CACHE_SIZE >> mp->m_sb.sb_blocklog;
-		if ((args.mod = (xfs_extlen_t)(do_mod(ap->off, args.prod))))
+		if ((args.mod = (xfs_extlen_t)(do_mod(ap->offset, args.prod))))
 			args.mod = (xfs_extlen_t)(args.prod - args.mod);
 	}
 	/*
@@ -2606,8 +2464,8 @@ xfs_bmap_btalloc(
 	 * is >= the stripe unit and the allocation offset is
 	 * at the end of file.
 	 */
-	if (!ap->low && ap->aeof) {
-		if (!ap->off) {
+	if (!ap->flist->xbf_low && ap->aeof) {
+		if (!ap->offset) {
 			args.alignment = mp->m_dalign;
 			atype = args.type;
 			isaligned = 1;
@@ -2660,7 +2518,7 @@ xfs_bmap_btalloc(
 		 * turned on.
 		 */
 		args.type = atype;
-		args.fsbno = ap->rval;
+		args.fsbno = ap->blkno;
 		args.alignment = mp->m_dalign;
 		args.minlen = nextminlen;
 		args.minalignslop = 0;
@@ -2674,7 +2532,7 @@ xfs_bmap_btalloc(
 		 * try again.
 		 */
 		args.type = atype;
-		args.fsbno = ap->rval;
+		args.fsbno = ap->blkno;
 		args.alignment = 0;
 		if ((error = xfs_alloc_vextent(&args)))
 			return error;
@@ -2683,7 +2541,7 @@ xfs_bmap_btalloc(
 	    args.minlen > ap->minlen) {
 		args.minlen = ap->minlen;
 		args.type = XFS_ALLOCTYPE_START_BNO;
-		args.fsbno = ap->rval;
+		args.fsbno = ap->blkno;
 		if ((error = xfs_alloc_vextent(&args)))
 			return error;
 	}
@@ -2694,13 +2552,26 @@ xfs_bmap_btalloc(
 		args.minleft = 0;
 		if ((error = xfs_alloc_vextent(&args)))
 			return error;
-		ap->low = 1;
+		ap->flist->xbf_low = 1;
 	}
 	if (args.fsbno != NULLFSBLOCK) {
-		ap->firstblock = ap->rval = args.fsbno;
+		/*
+		 * check the allocation happened at the same or higher AG than
+		 * the first block that was allocated.
+		 */
+		ASSERT(*ap->firstblock == NULLFSBLOCK ||
+		       XFS_FSB_TO_AGNO(mp, *ap->firstblock) ==
+		       XFS_FSB_TO_AGNO(mp, args.fsbno) ||
+		       (ap->flist->xbf_low &&
+			XFS_FSB_TO_AGNO(mp, *ap->firstblock) <
+			XFS_FSB_TO_AGNO(mp, args.fsbno)));
+
+		ap->blkno = args.fsbno;
+		if (*ap->firstblock == NULLFSBLOCK)
+			*ap->firstblock = args.fsbno;
 		ASSERT(nullfb || fb_agno == args.agno ||
-		       (ap->low && fb_agno < args.agno));
-		ap->alen = args.len;
+		       (ap->flist->xbf_low && fb_agno < args.agno));
+		ap->length = args.len;
 		ap->ip->i_d.di_nblocks += args.len;
 		xfs_trans_log_inode(ap->tp, ap->ip, XFS_ILOG_CORE);
 		if (ap->wasdel)
@@ -2714,8 +2585,8 @@ xfs_bmap_btalloc(
 					XFS_TRANS_DQ_BCOUNT,
 			(long) args.len);
 	} else {
-		ap->rval = NULLFSBLOCK;
-		ap->alen = 0;
+		ap->blkno = NULLFSBLOCK;
+		ap->length = 0;
 	}
 	return 0;
 }
@@ -3589,7 +3460,7 @@ xfs_bmap_add_attrfork(
 	}
 	ASSERT(ip->i_d.di_anextents == 0);
 
-	xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 
 	switch (ip->i_d.di_format) {
@@ -3782,19 +3653,11 @@ xfs_bmap_compute_maxlevels(
  * Routine to be called at transaction's end by xfs_bmapi, xfs_bunmapi
  * caller.  Frees all the extents that need freeing, which must be done
  * last due to locking considerations.  We never free any extents in
- * the first transaction.  This is to allow the caller to make the first
- * transaction a synchronous one so that the pointers to the data being
- * broken in this transaction will be permanent before the data is actually
- * freed.  This is necessary to prevent blocks from being reallocated
- * and written to before the free and reallocation are actually permanent.
- * We do not just make the first transaction synchronous here, because
- * there are more efficient ways to gain the same protection in some cases
- * (see the file truncation code).
+ * the first transaction.
  *
  * Return 1 if the given transaction was committed and a new one
  * started, and 0 otherwise in the committed parameter.
  */
-/*ARGSUSED*/
 int						/* error */
 xfs_bmap_finish(
 	xfs_trans_t		**tp,		/* transaction pointer addr */
@@ -3994,42 +3857,122 @@ xfs_bmap_last_before(
 	return 0;
 }
 
+STATIC int
+xfs_bmap_last_extent(
+	struct xfs_trans	*tp,
+	struct xfs_inode	*ip,
+	int			whichfork,
+	struct xfs_bmbt_irec	*rec,
+	int			*is_empty)
+{
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	int			error;
+	int			nextents;
+
+	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+		error = xfs_iread_extents(tp, ip, whichfork);
+		if (error)
+			return error;
+	}
+
+	nextents = ifp->if_bytes / sizeof(xfs_bmbt_rec_t);
+	if (nextents == 0) {
+		*is_empty = 1;
+		return 0;
+	}
+
+	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, nextents - 1), rec);
+	*is_empty = 0;
+	return 0;
+}
+
+/*
+ * Check the last inode extent to determine whether this allocation will result
+ * in blocks being allocated at the end of the file. When we allocate new data
+ * blocks at the end of the file which do not start at the previous data block,
+ * we will try to align the new blocks at stripe unit boundaries.
+ *
+ * Returns 0 in bma->aeof if the file (fork) is empty as any new write will be
+ * at, or past the EOF.
+ */
+STATIC int
+xfs_bmap_isaeof(
+	struct xfs_bmalloca	*bma,
+	int			whichfork)
+{
+	struct xfs_bmbt_irec	rec;
+	int			is_empty;
+	int			error;
+
+	bma->aeof = 0;
+	error = xfs_bmap_last_extent(NULL, bma->ip, whichfork, &rec,
+				     &is_empty);
+	if (error || is_empty)
+		return error;
+
+	/*
+	 * Check if we are allocation or past the last extent, or at least into
+	 * the last delayed allocated extent.
+	 */
+	bma->aeof = bma->offset >= rec.br_startoff + rec.br_blockcount ||
+		(bma->offset >= rec.br_startoff &&
+		 isnullstartblock(rec.br_startblock));
+	return 0;
+}
+
+/*
+ * Check if the endoff is outside the last extent. If so the caller will grow
+ * the allocation to a stripe unit boundary.  All offsets are considered outside
+ * the end of file for an empty fork, so 1 is returned in *eof in that case.
+ */
+int
+xfs_bmap_eof(
+	struct xfs_inode	*ip,
+	xfs_fileoff_t		endoff,
+	int			whichfork,
+	int			*eof)
+{
+	struct xfs_bmbt_irec	rec;
+	int			error;
+
+	error = xfs_bmap_last_extent(NULL, ip, whichfork, &rec, eof);
+	if (error || *eof)
+		return error;
+
+	*eof = endoff >= rec.br_startoff + rec.br_blockcount;
+	return 0;
+}
+
 /*
  * Returns the file-relative block number of the first block past eof in
  * the file.  This is not based on i_size, it is based on the extent records.
  * Returns 0 for local files, as they do not have extent records.
  */
-int						/* error */
+int
 xfs_bmap_last_offset(
-	xfs_trans_t	*tp,			/* transaction pointer */
-	xfs_inode_t	*ip,			/* incore inode */
-	xfs_fileoff_t	*last_block,		/* last block */
-	int		whichfork)		/* data or attr fork */
+	struct xfs_trans	*tp,
+	struct xfs_inode	*ip,
+	xfs_fileoff_t		*last_block,
+	int			whichfork)
 {
-	xfs_bmbt_rec_host_t *ep;		/* pointer to last extent */
-	int		error;			/* error return value */
-	xfs_ifork_t	*ifp;			/* inode fork pointer */
-	xfs_extnum_t	nextents;		/* number of extent entries */
+	struct xfs_bmbt_irec	rec;
+	int			is_empty;
+	int			error;
+
+	*last_block = 0;
+
+	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL)
+		return 0;
 
 	if (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)
+	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
 	       return XFS_ERROR(EIO);
-	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
-		*last_block = 0;
-		return 0;
-	}
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
-	    (error = xfs_iread_extents(tp, ip, whichfork)))
+
+	error = xfs_bmap_last_extent(NULL, ip, whichfork, &rec, &is_empty);
+	if (error || is_empty)
 		return error;
-	nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
-	if (!nextents) {
-		*last_block = 0;
-		return 0;
-	}
-	ep = xfs_iext_get_ext(ifp, nextents - 1);
-	*last_block = xfs_bmbt_get_startoff(ep) + xfs_bmbt_get_blockcount(ep);
+
+	*last_block = rec.br_startoff + rec.br_blockcount;
 	return 0;
 }
 
@@ -4159,7 +4102,6 @@ xfs_bmap_read_extents(
 		xfs_extnum_t	num_recs;
 		xfs_extnum_t	start;
 
-
 		num_recs = xfs_btree_get_numrecs(block);
 		if (unlikely(i + num_recs > room)) {
 			ASSERT(i + num_recs <= room);
@@ -4282,9 +4224,8 @@ xfs_bmap_validate_ret(
 		ASSERT(i == 0 ||
 		       mval[i - 1].br_startoff + mval[i - 1].br_blockcount ==
 		       mval[i].br_startoff);
-		if ((flags & XFS_BMAPI_WRITE) && !(flags & XFS_BMAPI_DELAY))
-			ASSERT(mval[i].br_startblock != DELAYSTARTBLOCK &&
-			       mval[i].br_startblock != HOLESTARTBLOCK);
+		ASSERT(mval[i].br_startblock != DELAYSTARTBLOCK &&
+		       mval[i].br_startblock != HOLESTARTBLOCK);
 		ASSERT(mval[i].br_state == XFS_EXT_NORM ||
 		       mval[i].br_state == XFS_EXT_UNWRITTEN);
 	}
@@ -4293,66 +4234,609 @@ xfs_bmap_validate_ret(
 
 
 /*
- * Map file blocks to filesystem blocks.
- * File range is given by the bno/len pair.
- * Adds blocks to file if a write ("flags & XFS_BMAPI_WRITE" set)
- * into a hole or past eof.
- * Only allocates blocks from a single allocation group,
- * to avoid locking problems.
+ * Trim the returned map to the required bounds
+ */
+STATIC void
+xfs_bmapi_trim_map(
+	struct xfs_bmbt_irec	*mval,
+	struct xfs_bmbt_irec	*got,
+	xfs_fileoff_t		*bno,
+	xfs_filblks_t		len,
+	xfs_fileoff_t		obno,
+	xfs_fileoff_t		end,
+	int			n,
+	int			flags)
+{
+	if ((flags & XFS_BMAPI_ENTIRE) ||
+	    got->br_startoff + got->br_blockcount <= obno) {
+		*mval = *got;
+		if (isnullstartblock(got->br_startblock))
+			mval->br_startblock = DELAYSTARTBLOCK;
+		return;
+	}
+
+	if (obno > *bno)
+		*bno = obno;
+	ASSERT((*bno >= obno) || (n == 0));
+	ASSERT(*bno < end);
+	mval->br_startoff = *bno;
+	if (isnullstartblock(got->br_startblock))
+		mval->br_startblock = DELAYSTARTBLOCK;
+	else
+		mval->br_startblock = got->br_startblock +
+					(*bno - got->br_startoff);
+	/*
+	 * Return the minimum of what we got and what we asked for for
+	 * the length.  We can use the len variable here because it is
+	 * modified below and we could have been there before coming
+	 * here if the first part of the allocation didn't overlap what
+	 * was asked for.
+	 */
+	mval->br_blockcount = XFS_FILBLKS_MIN(end - *bno,
+			got->br_blockcount - (*bno - got->br_startoff));
+	mval->br_state = got->br_state;
+	ASSERT(mval->br_blockcount <= len);
+	return;
+}
+
+/*
+ * Update and validate the extent map to return
+ */
+STATIC void
+xfs_bmapi_update_map(
+	struct xfs_bmbt_irec	**map,
+	xfs_fileoff_t		*bno,
+	xfs_filblks_t		*len,
+	xfs_fileoff_t		obno,
+	xfs_fileoff_t		end,
+	int			*n,
+	int			flags)
+{
+	xfs_bmbt_irec_t	*mval = *map;
+
+	ASSERT((flags & XFS_BMAPI_ENTIRE) ||
+	       ((mval->br_startoff + mval->br_blockcount) <= end));
+	ASSERT((flags & XFS_BMAPI_ENTIRE) || (mval->br_blockcount <= *len) ||
+	       (mval->br_startoff < obno));
+
+	*bno = mval->br_startoff + mval->br_blockcount;
+	*len = end - *bno;
+	if (*n > 0 && mval->br_startoff == mval[-1].br_startoff) {
+		/* update previous map with new information */
+		ASSERT(mval->br_startblock == mval[-1].br_startblock);
+		ASSERT(mval->br_blockcount > mval[-1].br_blockcount);
+		ASSERT(mval->br_state == mval[-1].br_state);
+		mval[-1].br_blockcount = mval->br_blockcount;
+		mval[-1].br_state = mval->br_state;
+	} else if (*n > 0 && mval->br_startblock != DELAYSTARTBLOCK &&
+		   mval[-1].br_startblock != DELAYSTARTBLOCK &&
+		   mval[-1].br_startblock != HOLESTARTBLOCK &&
+		   mval->br_startblock == mval[-1].br_startblock +
+					  mval[-1].br_blockcount &&
+		   ((flags & XFS_BMAPI_IGSTATE) ||
+			mval[-1].br_state == mval->br_state)) {
+		ASSERT(mval->br_startoff ==
+		       mval[-1].br_startoff + mval[-1].br_blockcount);
+		mval[-1].br_blockcount += mval->br_blockcount;
+	} else if (*n > 0 &&
+		   mval->br_startblock == DELAYSTARTBLOCK &&
+		   mval[-1].br_startblock == DELAYSTARTBLOCK &&
+		   mval->br_startoff ==
+		   mval[-1].br_startoff + mval[-1].br_blockcount) {
+		mval[-1].br_blockcount += mval->br_blockcount;
+		mval[-1].br_state = mval->br_state;
+	} else if (!((*n == 0) &&
+		     ((mval->br_startoff + mval->br_blockcount) <=
+		      obno))) {
+		mval++;
+		(*n)++;
+	}
+	*map = mval;
+}
+
+/*
+ * Map file blocks to filesystem blocks without allocation.
+ */
+int
+xfs_bmapi_read(
+	struct xfs_inode	*ip,
+	xfs_fileoff_t		bno,
+	xfs_filblks_t		len,
+	struct xfs_bmbt_irec	*mval,
+	int			*nmap,
+	int			flags)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_ifork	*ifp;
+	struct xfs_bmbt_irec	got;
+	struct xfs_bmbt_irec	prev;
+	xfs_fileoff_t		obno;
+	xfs_fileoff_t		end;
+	xfs_extnum_t		lastx;
+	int			error;
+	int			eof;
+	int			n = 0;
+	int			whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
+						XFS_ATTR_FORK : XFS_DATA_FORK;
+
+	ASSERT(*nmap >= 1);
+	ASSERT(!(flags & ~(XFS_BMAPI_ATTRFORK|XFS_BMAPI_ENTIRE|
+			   XFS_BMAPI_IGSTATE)));
+
+	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_RANDOM_BMAPIFORMAT))) {
+		XFS_ERROR_REPORT("xfs_bmapi_read", XFS_ERRLEVEL_LOW, mp);
+		return XFS_ERROR(EFSCORRUPTED);
+	}
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return XFS_ERROR(EIO);
+
+	XFS_STATS_INC(xs_blk_mapr);
+
+	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ASSERT(ifp->if_ext_max ==
+	       XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t));
+
+	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+		error = xfs_iread_extents(NULL, ip, whichfork);
+		if (error)
+			return error;
+	}
+
+	xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got, &prev);
+	end = bno + len;
+	obno = bno;
+
+	while (bno < end && n < *nmap) {
+		/* Reading past eof, act as though there's a hole up to end. */
+		if (eof)
+			got.br_startoff = end;
+		if (got.br_startoff > bno) {
+			/* Reading in a hole.  */
+			mval->br_startoff = bno;
+			mval->br_startblock = HOLESTARTBLOCK;
+			mval->br_blockcount =
+				XFS_FILBLKS_MIN(len, got.br_startoff - bno);
+			mval->br_state = XFS_EXT_NORM;
+			bno += mval->br_blockcount;
+			len -= mval->br_blockcount;
+			mval++;
+			n++;
+			continue;
+		}
+
+		/* set up the extent map to return. */
+		xfs_bmapi_trim_map(mval, &got, &bno, len, obno, end, n, flags);
+		xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);
+
+		/* If we're done, stop now. */
+		if (bno >= end || n >= *nmap)
+			break;
+
+		/* Else go on to the next record. */
+		if (++lastx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t))
+			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx), &got);
+		else
+			eof = 1;
+	}
+	*nmap = n;
+	return 0;
+}
+
+STATIC int
+xfs_bmapi_reserve_delalloc(
+	struct xfs_inode	*ip,
+	xfs_fileoff_t		aoff,
+	xfs_filblks_t		len,
+	struct xfs_bmbt_irec	*got,
+	struct xfs_bmbt_irec	*prev,
+	xfs_extnum_t		*lastx,
+	int			eof)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	xfs_extlen_t		alen;
+	xfs_extlen_t		indlen;
+	char			rt = XFS_IS_REALTIME_INODE(ip);
+	xfs_extlen_t		extsz;
+	int			error;
+
+	alen = XFS_FILBLKS_MIN(len, MAXEXTLEN);
+	if (!eof)
+		alen = XFS_FILBLKS_MIN(alen, got->br_startoff - aoff);
+
+	/* Figure out the extent size, adjust alen */
+	extsz = xfs_get_extsz_hint(ip);
+	if (extsz) {
+		/*
+		 * Make sure we don't exceed a single extent length when we
+		 * align the extent by reducing length we are going to
+		 * allocate by the maximum amount extent size aligment may
+		 * require.
+		 */
+		alen = XFS_FILBLKS_MIN(len, MAXEXTLEN - (2 * extsz - 1));
+		error = xfs_bmap_extsize_align(mp, got, prev, extsz, rt, eof,
+					       1, 0, &aoff, &alen);
+		ASSERT(!error);
+	}
+
+	if (rt)
+		extsz = alen / mp->m_sb.sb_rextsize;
+
+	/*
+	 * Make a transaction-less quota reservation for delayed allocation
+	 * blocks.  This number gets adjusted later.  We return if we haven't
+	 * allocated blocks already inside this loop.
+	 */
+	error = xfs_trans_reserve_quota_nblks(NULL, ip, (long)alen, 0,
+			rt ? XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS);
+	if (error)
+		return error;
+
+	/*
+	 * Split changing sb for alen and indlen since they could be coming
+	 * from different places.
+	 */
+	indlen = (xfs_extlen_t)xfs_bmap_worst_indlen(ip, alen);
+	ASSERT(indlen > 0);
+
+	if (rt) {
+		error = xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS,
+					  -((int64_t)extsz), 0);
+	} else {
+		error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
+						 -((int64_t)alen), 0);
+	}
+
+	if (error)
+		goto out_unreserve_quota;
+
+	error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
+					 -((int64_t)indlen), 0);
+	if (error)
+		goto out_unreserve_blocks;
+
+
+	ip->i_delayed_blks += alen;
+
+	got->br_startoff = aoff;
+	got->br_startblock = nullstartblock(indlen);
+	got->br_blockcount = alen;
+	got->br_state = XFS_EXT_NORM;
+	xfs_bmap_add_extent_hole_delay(ip, lastx, got);
+
+	/*
+	 * Update our extent pointer, given that xfs_bmap_add_extent_hole_delay
+	 * might have merged it into one of the neighbouring ones.
+	 */
+	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *lastx), got);
+
+	ASSERT(got->br_startoff <= aoff);
+	ASSERT(got->br_startoff + got->br_blockcount >= aoff + alen);
+	ASSERT(isnullstartblock(got->br_startblock));
+	ASSERT(got->br_state == XFS_EXT_NORM);
+	return 0;
+
+out_unreserve_blocks:
+	if (rt)
+		xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS, extsz, 0);
+	else
+		xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, alen, 0);
+out_unreserve_quota:
+	if (XFS_IS_QUOTA_ON(mp))
+		xfs_trans_unreserve_quota_nblks(NULL, ip, alen, 0, rt ?
+				XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS);
+	return error;
+}
+
+/*
+ * Map file blocks to filesystem blocks, adding delayed allocations as needed.
+ */
+int
+xfs_bmapi_delay(
+	struct xfs_inode	*ip,	/* incore inode */
+	xfs_fileoff_t		bno,	/* starting file offs. mapped */
+	xfs_filblks_t		len,	/* length to map in file */
+	struct xfs_bmbt_irec	*mval,	/* output: map values */
+	int			*nmap,	/* i/o: mval size/count */
+	int			flags)	/* XFS_BMAPI_... */
+{
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	struct xfs_bmbt_irec	got;	/* current file extent record */
+	struct xfs_bmbt_irec	prev;	/* previous file extent record */
+	xfs_fileoff_t		obno;	/* old block number (offset) */
+	xfs_fileoff_t		end;	/* end of mapped file region */
+	xfs_extnum_t		lastx;	/* last useful extent number */
+	int			eof;	/* we've hit the end of extents */
+	int			n = 0;	/* current extent index */
+	int			error = 0;
+
+	ASSERT(*nmap >= 1);
+	ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
+	ASSERT(!(flags & ~XFS_BMAPI_ENTIRE));
+
+	if (unlikely(XFS_TEST_ERROR(
+	    (XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_EXTENTS &&
+	     XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_BTREE),
+	     mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
+		XFS_ERROR_REPORT("xfs_bmapi_delay", XFS_ERRLEVEL_LOW, mp);
+		return XFS_ERROR(EFSCORRUPTED);
+	}
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return XFS_ERROR(EIO);
+
+	XFS_STATS_INC(xs_blk_mapw);
+
+	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+		error = xfs_iread_extents(NULL, ip, XFS_DATA_FORK);
+		if (error)
+			return error;
+	}
+
+	xfs_bmap_search_extents(ip, bno, XFS_DATA_FORK, &eof, &lastx, &got, &prev);
+	end = bno + len;
+	obno = bno;
+
+	while (bno < end && n < *nmap) {
+		if (eof || got.br_startoff > bno) {
+			error = xfs_bmapi_reserve_delalloc(ip, bno, len, &got,
+							   &prev, &lastx, eof);
+			if (error) {
+				if (n == 0) {
+					*nmap = 0;
+					return error;
+				}
+				break;
+			}
+		}
+
+		/* set up the extent map to return. */
+		xfs_bmapi_trim_map(mval, &got, &bno, len, obno, end, n, flags);
+		xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);
+
+		/* If we're done, stop now. */
+		if (bno >= end || n >= *nmap)
+			break;
+
+		/* Else go on to the next record. */
+		prev = got;
+		if (++lastx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t))
+			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx), &got);
+		else
+			eof = 1;
+	}
+
+	*nmap = n;
+	return 0;
+}
+
+
+STATIC int
+xfs_bmapi_allocate(
+	struct xfs_bmalloca	*bma,
+	int			flags)
+{
+	struct xfs_mount	*mp = bma->ip->i_mount;
+	int			whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
+						XFS_ATTR_FORK : XFS_DATA_FORK;
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+	int			tmp_logflags = 0;
+	int			error;
+	int			rt;
+
+	rt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(bma->ip);
+
+	/*
+	 * For the wasdelay case, we could also just allocate the stuff asked
+	 * for in this bmap call but that wouldn't be as good.
+	 */
+	if (bma->wasdel) {
+		bma->length = (xfs_extlen_t)bma->got.br_blockcount;
+		bma->offset = bma->got.br_startoff;
+		if (bma->idx != NULLEXTNUM && bma->idx) {
+			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1),
+					 &bma->prev);
+		}
+	} else {
+		bma->length = XFS_FILBLKS_MIN(bma->length, MAXEXTLEN);
+		if (!bma->eof)
+			bma->length = XFS_FILBLKS_MIN(bma->length,
+					bma->got.br_startoff - bma->offset);
+	}
+
+	/*
+	 * Indicate if this is the first user data in the file, or just any
+	 * user data.
+	 */
+	if (!(flags & XFS_BMAPI_METADATA)) {
+		bma->userdata = (bma->offset == 0) ?
+			XFS_ALLOC_INITIAL_USER_DATA : XFS_ALLOC_USERDATA;
+	}
+
+	bma->minlen = (flags & XFS_BMAPI_CONTIG) ? bma->length : 1;
+
+	/*
+	 * Only want to do the alignment at the eof if it is userdata and
+	 * allocation length is larger than a stripe unit.
+	 */
+	if (mp->m_dalign && bma->length >= mp->m_dalign &&
+	    !(flags & XFS_BMAPI_METADATA) && whichfork == XFS_DATA_FORK) {
+		error = xfs_bmap_isaeof(bma, whichfork);
+		if (error)
+			return error;
+	}
+
+	error = xfs_bmap_alloc(bma);
+	if (error)
+		return error;
+
+	if (bma->flist->xbf_low)
+		bma->minleft = 0;
+	if (bma->cur)
+		bma->cur->bc_private.b.firstblock = *bma->firstblock;
+	if (bma->blkno == NULLFSBLOCK)
+		return 0;
+	if ((ifp->if_flags & XFS_IFBROOT) && !bma->cur) {
+		bma->cur = xfs_bmbt_init_cursor(mp, bma->tp, bma->ip, whichfork);
+		bma->cur->bc_private.b.firstblock = *bma->firstblock;
+		bma->cur->bc_private.b.flist = bma->flist;
+	}
+	/*
+	 * Bump the number of extents we've allocated
+	 * in this call.
+	 */
+	bma->nallocs++;
+
+	if (bma->cur)
+		bma->cur->bc_private.b.flags =
+			bma->wasdel ? XFS_BTCUR_BPRV_WASDEL : 0;
+
+	bma->got.br_startoff = bma->offset;
+	bma->got.br_startblock = bma->blkno;
+	bma->got.br_blockcount = bma->length;
+	bma->got.br_state = XFS_EXT_NORM;
+
+	/*
+	 * A wasdelay extent has been initialized, so shouldn't be flagged
+	 * as unwritten.
+	 */
+	if (!bma->wasdel && (flags & XFS_BMAPI_PREALLOC) &&
+	    xfs_sb_version_hasextflgbit(&mp->m_sb))
+		bma->got.br_state = XFS_EXT_UNWRITTEN;
+
+	if (bma->wasdel)
+		error = xfs_bmap_add_extent_delay_real(bma);
+	else
+		error = xfs_bmap_add_extent_hole_real(bma, whichfork);
+
+	bma->logflags |= tmp_logflags;
+	if (error)
+		return error;
+
+	/*
+	 * Update our extent pointer, given that xfs_bmap_add_extent_delay_real
+	 * 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);
+
+	ASSERT(bma->got.br_startoff <= bma->offset);
+	ASSERT(bma->got.br_startoff + bma->got.br_blockcount >=
+	       bma->offset + bma->length);
+	ASSERT(bma->got.br_state == XFS_EXT_NORM ||
+	       bma->got.br_state == XFS_EXT_UNWRITTEN);
+	return 0;
+}
+
+STATIC int
+xfs_bmapi_convert_unwritten(
+	struct xfs_bmalloca	*bma,
+	struct xfs_bmbt_irec	*mval,
+	xfs_filblks_t		len,
+	int			flags)
+{
+	int			whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
+						XFS_ATTR_FORK : XFS_DATA_FORK;
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+	int			tmp_logflags = 0;
+	int			error;
+
+	/* check if we need to do unwritten->real conversion */
+	if (mval->br_state == XFS_EXT_UNWRITTEN &&
+	    (flags & XFS_BMAPI_PREALLOC))
+		return 0;
+
+	/* check if we need to do real->unwritten conversion */
+	if (mval->br_state == XFS_EXT_NORM &&
+	    (flags & (XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT)) !=
+			(XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT))
+		return 0;
+
+	/*
+	 * Modify (by adding) the state flag, if writing.
+	 */
+	ASSERT(mval->br_blockcount <= len);
+	if ((ifp->if_flags & XFS_IFBROOT) && !bma->cur) {
+		bma->cur = xfs_bmbt_init_cursor(bma->ip->i_mount, bma->tp,
+					bma->ip, whichfork);
+		bma->cur->bc_private.b.firstblock = *bma->firstblock;
+		bma->cur->bc_private.b.flist = bma->flist;
+	}
+	mval->br_state = (mval->br_state == XFS_EXT_UNWRITTEN)
+				? XFS_EXT_NORM : XFS_EXT_UNWRITTEN;
+
+	error = xfs_bmap_add_extent_unwritten_real(bma->tp, bma->ip, &bma->idx,
+			&bma->cur, mval, bma->firstblock, bma->flist,
+			&tmp_logflags);
+	bma->logflags |= tmp_logflags;
+	if (error)
+		return error;
+
+	/*
+	 * Update our extent pointer, given that
+	 * 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);
+
+	/*
+	 * We may have combined previously unwritten space with written space,
+	 * so generate another request.
+	 */
+	if (mval->br_blockcount < len)
+		return EAGAIN;
+	return 0;
+}
+
+/*
+ * Map file blocks to filesystem blocks, and allocate blocks or convert the
+ * extent state if necessary.  Details behaviour is controlled by the flags
+ * parameter.  Only allocates blocks from a single allocation group, to avoid
+ * locking problems.
+ *
  * The returned value in "firstblock" from the first call in a transaction
  * must be remembered and presented to subsequent calls in "firstblock".
  * An upper bound for the number of blocks to be allocated is supplied to
  * the first call in "total"; if no allocation group has that many free
  * blocks then the call will fail (return NULLFSBLOCK in "firstblock").
  */
-int					/* error */
-xfs_bmapi(
-	xfs_trans_t	*tp,		/* transaction pointer */
-	xfs_inode_t	*ip,		/* incore inode */
-	xfs_fileoff_t	bno,		/* starting file offs. mapped */
-	xfs_filblks_t	len,		/* length to map in file */
-	int		flags,		/* XFS_BMAPI_... */
-	xfs_fsblock_t	*firstblock,	/* first allocated block
-					   controls a.g. for allocs */
-	xfs_extlen_t	total,		/* total blocks needed */
-	xfs_bmbt_irec_t	*mval,		/* output: map values */
-	int		*nmap,		/* i/o: mval size/count */
-	xfs_bmap_free_t	*flist)		/* i/o: list extents to free */
+int
+xfs_bmapi_write(
+	struct xfs_trans	*tp,		/* transaction pointer */
+	struct xfs_inode	*ip,		/* incore inode */
+	xfs_fileoff_t		bno,		/* starting file offs. mapped */
+	xfs_filblks_t		len,		/* length to map in file */
+	int			flags,		/* XFS_BMAPI_... */
+	xfs_fsblock_t		*firstblock,	/* first allocated block
+						   controls a.g. for allocs */
+	xfs_extlen_t		total,		/* total blocks needed */
+	struct xfs_bmbt_irec	*mval,		/* output: map values */
+	int			*nmap,		/* i/o: mval size/count */
+	struct xfs_bmap_free	*flist)		/* i/o: list extents to free */
 {
-	xfs_fsblock_t	abno;		/* allocated block number */
-	xfs_extlen_t	alen;		/* allocated extent length */
-	xfs_fileoff_t	aoff;		/* allocated file offset */
-	xfs_bmalloca_t	bma = { 0 };	/* args for xfs_bmap_alloc */
-	xfs_btree_cur_t	*cur;		/* bmap btree cursor */
-	xfs_fileoff_t	end;		/* end of mapped file region */
-	int		eof;		/* we've hit the end of extents */
-	xfs_bmbt_rec_host_t *ep;	/* extent record pointer */
-	int		error;		/* error return */
-	xfs_bmbt_irec_t	got;		/* current file extent record */
-	xfs_ifork_t	*ifp;		/* inode fork pointer */
-	xfs_extlen_t	indlen;		/* indirect blocks length */
-	xfs_extnum_t	lastx;		/* last useful extent number */
-	int		logflags;	/* flags for transaction logging */
-	xfs_extlen_t	minleft;	/* min blocks left after allocation */
-	xfs_extlen_t	minlen;		/* min allocation size */
-	xfs_mount_t	*mp;		/* xfs mount structure */
-	int		n;		/* current extent index */
-	int		nallocs;	/* number of extents alloc'd */
-	xfs_extnum_t	nextents;	/* number of extents in file */
-	xfs_fileoff_t	obno;		/* old block number (offset) */
-	xfs_bmbt_irec_t	prev;		/* previous file extent record */
-	int		tmp_logflags;	/* temp flags holder */
-	int		whichfork;	/* data or attr fork */
-	char		inhole;		/* current location is hole in file */
-	char		wasdelay;	/* old extent was delayed */
-	char		wr;		/* this is a write request */
-	char		rt;		/* this is a realtime file */
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_ifork	*ifp;
+	struct xfs_bmalloca	bma = { 0 };	/* args for xfs_bmap_alloc */
+	xfs_fileoff_t		end;		/* end of mapped file region */
+	int			eof;		/* after the end of extents */
+	int			error;		/* error return */
+	int			n;		/* current extent index */
+	xfs_fileoff_t		obno;		/* old block number (offset) */
+	int			whichfork;	/* data or attr fork */
+	char			inhole;		/* current location is hole in file */
+	char			wasdelay;	/* old extent was delayed */
+
 #ifdef DEBUG
-	xfs_fileoff_t	orig_bno;	/* original block number value */
-	int		orig_flags;	/* original flags arg value */
-	xfs_filblks_t	orig_len;	/* original value of len arg */
-	xfs_bmbt_irec_t	*orig_mval;	/* original value of mval */
-	int		orig_nmap;	/* original value of *nmap */
+	xfs_fileoff_t		orig_bno;	/* original block number value */
+	int			orig_flags;	/* original flags arg value */
+	xfs_filblks_t		orig_len;	/* original value of len arg */
+	struct xfs_bmbt_irec	*orig_mval;	/* original value of mval */
+	int			orig_nmap;	/* original value of *nmap */
 
 	orig_bno = bno;
 	orig_len = len;
@@ -4360,488 +4844,133 @@ xfs_bmapi(
 	orig_mval = mval;
 	orig_nmap = *nmap;
 #endif
+
 	ASSERT(*nmap >= 1);
-	ASSERT(*nmap <= XFS_BMAP_MAX_NMAP || !(flags & XFS_BMAPI_WRITE));
+	ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
+	ASSERT(!(flags & XFS_BMAPI_IGSTATE));
+	ASSERT(tp != NULL);
+
 	whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
 		XFS_ATTR_FORK : XFS_DATA_FORK;
-	mp = ip->i_mount;
+
 	if (unlikely(XFS_TEST_ERROR(
 	    (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
 	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
 	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL),
 	     mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
-		XFS_ERROR_REPORT("xfs_bmapi", XFS_ERRLEVEL_LOW, mp);
+		XFS_ERROR_REPORT("xfs_bmapi_write", XFS_ERRLEVEL_LOW, mp);
 		return XFS_ERROR(EFSCORRUPTED);
 	}
+
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return XFS_ERROR(EIO);
-	rt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(ip);
+
 	ifp = XFS_IFORK_PTR(ip, whichfork);
 	ASSERT(ifp->if_ext_max ==
 	       XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t));
-	if ((wr = (flags & XFS_BMAPI_WRITE)) != 0)
-		XFS_STATS_INC(xs_blk_mapw);
-	else
-		XFS_STATS_INC(xs_blk_mapr);
-	/*
-	 * IGSTATE flag is used to combine extents which
-	 * differ only due to the state of the extents.
-	 * This technique is used from xfs_getbmap()
-	 * when the caller does not wish to see the
-	 * separation (which is the default).
-	 *
-	 * This technique is also used when writing a
-	 * buffer which has been partially written,
-	 * (usually by being flushed during a chunkread),
-	 * to ensure one write takes place. This also
-	 * prevents a change in the xfs inode extents at
-	 * this time, intentionally. This change occurs
-	 * on completion of the write operation, in
-	 * xfs_strat_comp(), where the xfs_bmapi() call
-	 * is transactioned, and the extents combined.
-	 */
-	if ((flags & XFS_BMAPI_IGSTATE) && wr)	/* if writing unwritten space */
-		wr = 0;				/* no allocations are allowed */
-	ASSERT(wr || !(flags & XFS_BMAPI_DELAY));
-	logflags = 0;
-	nallocs = 0;
-	cur = NULL;
+
+	XFS_STATS_INC(xs_blk_mapw);
+
 	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
-		ASSERT(wr && tp);
-		if ((error = xfs_bmap_local_to_extents(tp, ip,
-				firstblock, total, &logflags, whichfork)))
+		error = xfs_bmap_local_to_extents(tp, ip, firstblock, total,
+						  &bma.logflags, whichfork);
+		if (error)
 			goto error0;
 	}
-	if (wr && *firstblock == NULLFSBLOCK) {
+
+	if (*firstblock == NULLFSBLOCK) {
 		if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE)
-			minleft = be16_to_cpu(ifp->if_broot->bb_level) + 1;
+			bma.minleft = be16_to_cpu(ifp->if_broot->bb_level) + 1;
 		else
-			minleft = 1;
-	} else
-		minleft = 0;
-	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
-	    (error = xfs_iread_extents(tp, ip, whichfork)))
-		goto error0;
-	ep = xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got,
-		&prev);
-	nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
+			bma.minleft = 1;
+	} else {
+		bma.minleft = 0;
+	}
+
+	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+		error = xfs_iread_extents(tp, ip, whichfork);
+		if (error)
+			goto error0;
+	}
+
+	xfs_bmap_search_extents(ip, bno, whichfork, &eof, &bma.idx, &bma.got,
+				&bma.prev);
 	n = 0;
 	end = bno + len;
 	obno = bno;
-	bma.ip = NULL;
+
+	bma.tp = tp;
+	bma.ip = ip;
+	bma.total = total;
+	bma.userdata = 0;
+	bma.flist = flist;
+	bma.firstblock = firstblock;
 
 	while (bno < end && n < *nmap) {
-		/*
-		 * Reading past eof, act as though there's a hole
-		 * up to end.
-		 */
-		if (eof && !wr)
-			got.br_startoff = end;
-		inhole = eof || got.br_startoff > bno;
-		wasdelay = wr && !inhole && !(flags & XFS_BMAPI_DELAY) &&
-			isnullstartblock(got.br_startblock);
+		inhole = eof || bma.got.br_startoff > bno;
+		wasdelay = !inhole && isnullstartblock(bma.got.br_startblock);
+
 		/*
 		 * First, deal with the hole before the allocated space
 		 * that we found, if any.
 		 */
-		if (wr && (inhole || wasdelay)) {
-			/*
-			 * For the wasdelay case, we could also just
-			 * allocate the stuff asked for in this bmap call
-			 * but that wouldn't be as good.
-			 */
-			if (wasdelay) {
-				alen = (xfs_extlen_t)got.br_blockcount;
-				aoff = got.br_startoff;
-				if (lastx != NULLEXTNUM && lastx) {
-					ep = xfs_iext_get_ext(ifp, lastx - 1);
-					xfs_bmbt_get_all(ep, &prev);
-				}
-			} else {
-				alen = (xfs_extlen_t)
-					XFS_FILBLKS_MIN(len, MAXEXTLEN);
-				if (!eof)
-					alen = (xfs_extlen_t)
-						XFS_FILBLKS_MIN(alen,
-							got.br_startoff - bno);
-				aoff = bno;
-			}
-			minlen = (flags & XFS_BMAPI_CONTIG) ? alen : 1;
-			if (flags & XFS_BMAPI_DELAY) {
-				xfs_extlen_t	extsz;
-
-				/* Figure out the extent size, adjust alen */
-				extsz = xfs_get_extsz_hint(ip);
-				if (extsz) {
-					/*
-					 * make sure we don't exceed a single
-					 * extent length when we align the
-					 * extent by reducing length we are
-					 * going to allocate by the maximum
-					 * amount extent size aligment may
-					 * require.
-					 */
-					alen = XFS_FILBLKS_MIN(len,
-						   MAXEXTLEN - (2 * extsz - 1));
-					error = xfs_bmap_extsize_align(mp,
-							&got, &prev, extsz,
-							rt, eof,
-							flags&XFS_BMAPI_DELAY,
-							flags&XFS_BMAPI_CONVERT,
-							&aoff, &alen);
-					ASSERT(!error);
-				}
-
-				if (rt)
-					extsz = alen / mp->m_sb.sb_rextsize;
-
-				/*
-				 * Make a transaction-less quota reservation for
-				 * delayed allocation blocks. This number gets
-				 * adjusted later.  We return if we haven't
-				 * allocated blocks already inside this loop.
-				 */
-				error = xfs_trans_reserve_quota_nblks(
-						NULL, ip, (long)alen, 0,
-						rt ? XFS_QMOPT_RES_RTBLKS :
-						     XFS_QMOPT_RES_REGBLKS);
-				if (error) {
-					if (n == 0) {
-						*nmap = 0;
-						ASSERT(cur == NULL);
-						return error;
-					}
-					break;
-				}
-
-				/*
-				 * Split changing sb for alen and indlen since
-				 * they could be coming from different places.
-				 */
-				indlen = (xfs_extlen_t)
-					xfs_bmap_worst_indlen(ip, alen);
-				ASSERT(indlen > 0);
-
-				if (rt) {
-					error = xfs_mod_incore_sb(mp,
-							XFS_SBS_FREXTENTS,
-							-((int64_t)extsz), 0);
-				} else {
-					error = xfs_icsb_modify_counters(mp,
-							XFS_SBS_FDBLOCKS,
-							-((int64_t)alen), 0);
-				}
-				if (!error) {
-					error = xfs_icsb_modify_counters(mp,
-							XFS_SBS_FDBLOCKS,
-							-((int64_t)indlen), 0);
-					if (error && rt)
-						xfs_mod_incore_sb(mp,
-							XFS_SBS_FREXTENTS,
-							(int64_t)extsz, 0);
-					else if (error)
-						xfs_icsb_modify_counters(mp,
-							XFS_SBS_FDBLOCKS,
-							(int64_t)alen, 0);
-				}
-
-				if (error) {
-					if (XFS_IS_QUOTA_ON(mp))
-						/* unreserve the blocks now */
-						(void)
-						xfs_trans_unreserve_quota_nblks(
-							NULL, ip,
-							(long)alen, 0, rt ?
-							XFS_QMOPT_RES_RTBLKS :
-							XFS_QMOPT_RES_REGBLKS);
-					break;
-				}
-
-				ip->i_delayed_blks += alen;
-				abno = nullstartblock(indlen);
-			} else {
-				/*
-				 * If first time, allocate and fill in
-				 * once-only bma fields.
-				 */
-				if (bma.ip == NULL) {
-					bma.tp = tp;
-					bma.ip = ip;
-					bma.prevp = &prev;
-					bma.gotp = &got;
-					bma.total = total;
-					bma.userdata = 0;
-				}
-				/* Indicate if this is the first user data
-				 * in the file, or just any user data.
-				 */
-				if (!(flags & XFS_BMAPI_METADATA)) {
-					bma.userdata = (aoff == 0) ?
-						XFS_ALLOC_INITIAL_USER_DATA :
-						XFS_ALLOC_USERDATA;
-				}
-				/*
-				 * Fill in changeable bma fields.
-				 */
-				bma.eof = eof;
-				bma.firstblock = *firstblock;
-				bma.alen = alen;
-				bma.off = aoff;
-				bma.conv = !!(flags & XFS_BMAPI_CONVERT);
-				bma.wasdel = wasdelay;
-				bma.minlen = minlen;
-				bma.low = flist->xbf_low;
-				bma.minleft = minleft;
-				/*
-				 * Only want to do the alignment at the
-				 * eof if it is userdata and allocation length
-				 * is larger than a stripe unit.
-				 */
-				if (mp->m_dalign && alen >= mp->m_dalign &&
-				    (!(flags & XFS_BMAPI_METADATA)) &&
-				    (whichfork == XFS_DATA_FORK)) {
-					if ((error = xfs_bmap_isaeof(ip, aoff,
-							whichfork, &bma.aeof)))
-						goto error0;
-				} else
-					bma.aeof = 0;
-				/*
-				 * Call allocator.
-				 */
-				if ((error = xfs_bmap_alloc(&bma)))
-					goto error0;
-				/*
-				 * Copy out result fields.
-				 */
-				abno = bma.rval;
-				if ((flist->xbf_low = bma.low))
-					minleft = 0;
-				alen = bma.alen;
-				aoff = bma.off;
-				ASSERT(*firstblock == NULLFSBLOCK ||
-				       XFS_FSB_TO_AGNO(mp, *firstblock) ==
-				       XFS_FSB_TO_AGNO(mp, bma.firstblock) ||
-				       (flist->xbf_low &&
-					XFS_FSB_TO_AGNO(mp, *firstblock) <
-					XFS_FSB_TO_AGNO(mp, bma.firstblock)));
-				*firstblock = bma.firstblock;
-				if (cur)
-					cur->bc_private.b.firstblock =
-						*firstblock;
-				if (abno == NULLFSBLOCK)
-					break;
-				if ((ifp->if_flags & XFS_IFBROOT) && !cur) {
-					cur = xfs_bmbt_init_cursor(mp, tp,
-						ip, whichfork);
-					cur->bc_private.b.firstblock =
-						*firstblock;
-					cur->bc_private.b.flist = flist;
-				}
-				/*
-				 * Bump the number of extents we've allocated
-				 * in this call.
-				 */
-				nallocs++;
-			}
-			if (cur)
-				cur->bc_private.b.flags =
-					wasdelay ? XFS_BTCUR_BPRV_WASDEL : 0;
-			got.br_startoff = aoff;
-			got.br_startblock = abno;
-			got.br_blockcount = alen;
-			got.br_state = XFS_EXT_NORM;	/* assume normal */
-			/*
-			 * Determine state of extent, and the filesystem.
-			 * A wasdelay extent has been initialized, so
-			 * shouldn't be flagged as unwritten.
-			 */
-			if (wr && xfs_sb_version_hasextflgbit(&mp->m_sb)) {
-				if (!wasdelay && (flags & XFS_BMAPI_PREALLOC))
-					got.br_state = XFS_EXT_UNWRITTEN;
-			}
-			error = xfs_bmap_add_extent(tp, ip, &lastx, &cur, &got,
-				firstblock, flist, &tmp_logflags,
-				whichfork);
-			logflags |= tmp_logflags;
+		if (inhole || wasdelay) {
+			bma.eof = eof;
+			bma.conv = !!(flags & XFS_BMAPI_CONVERT);
+			bma.wasdel = wasdelay;
+			bma.length = len;
+			bma.offset = bno;
+
+			error = xfs_bmapi_allocate(&bma, flags);
 			if (error)
 				goto error0;
-			ep = xfs_iext_get_ext(ifp, lastx);
-			nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
-			xfs_bmbt_get_all(ep, &got);
-			ASSERT(got.br_startoff <= aoff);
-			ASSERT(got.br_startoff + got.br_blockcount >=
-				aoff + alen);
-#ifdef DEBUG
-			if (flags & XFS_BMAPI_DELAY) {
-				ASSERT(isnullstartblock(got.br_startblock));
-				ASSERT(startblockval(got.br_startblock) > 0);
-			}
-			ASSERT(got.br_state == XFS_EXT_NORM ||
-			       got.br_state == XFS_EXT_UNWRITTEN);
-#endif
-			/*
-			 * Fall down into the found allocated space case.
-			 */
-		} else if (inhole) {
-			/*
-			 * Reading in a hole.
-			 */
-			mval->br_startoff = bno;
-			mval->br_startblock = HOLESTARTBLOCK;
-			mval->br_blockcount =
-				XFS_FILBLKS_MIN(len, got.br_startoff - bno);
-			mval->br_state = XFS_EXT_NORM;
-			bno += mval->br_blockcount;
-			len -= mval->br_blockcount;
-			mval++;
-			n++;
-			continue;
-		}
-		/*
-		 * Then deal with the allocated space we found.
-		 */
-		ASSERT(ep != NULL);
-		if (!(flags & XFS_BMAPI_ENTIRE) &&
-		    (got.br_startoff + got.br_blockcount > obno)) {
-			if (obno > bno)
-				bno = obno;
-			ASSERT((bno >= obno) || (n == 0));
-			ASSERT(bno < end);
-			mval->br_startoff = bno;
-			if (isnullstartblock(got.br_startblock)) {
-				ASSERT(!wr || (flags & XFS_BMAPI_DELAY));
-				mval->br_startblock = DELAYSTARTBLOCK;
-			} else
-				mval->br_startblock =
-					got.br_startblock +
-					(bno - got.br_startoff);
-			/*
-			 * Return the minimum of what we got and what we
-			 * asked for for the length.  We can use the len
-			 * variable here because it is modified below
-			 * and we could have been there before coming
-			 * here if the first part of the allocation
-			 * didn't overlap what was asked for.
-			 */
-			mval->br_blockcount =
-				XFS_FILBLKS_MIN(end - bno, got.br_blockcount -
-					(bno - got.br_startoff));
-			mval->br_state = got.br_state;
-			ASSERT(mval->br_blockcount <= len);
-		} else {
-			*mval = got;
-			if (isnullstartblock(mval->br_startblock)) {
-				ASSERT(!wr || (flags & XFS_BMAPI_DELAY));
-				mval->br_startblock = DELAYSTARTBLOCK;
-			}
+			if (bma.blkno == NULLFSBLOCK)
+				break;
 		}
 
-		/*
-		 * Check if writing previously allocated but
-		 * unwritten extents.
-		 */
-		if (wr &&
-		    ((mval->br_state == XFS_EXT_UNWRITTEN &&
-		      ((flags & (XFS_BMAPI_PREALLOC|XFS_BMAPI_DELAY)) == 0)) ||
-		     (mval->br_state == XFS_EXT_NORM &&
-		      ((flags & (XFS_BMAPI_PREALLOC|XFS_BMAPI_CONVERT)) ==
-				(XFS_BMAPI_PREALLOC|XFS_BMAPI_CONVERT))))) {
-			/*
-			 * Modify (by adding) the state flag, if writing.
-			 */
-			ASSERT(mval->br_blockcount <= len);
-			if ((ifp->if_flags & XFS_IFBROOT) && !cur) {
-				cur = xfs_bmbt_init_cursor(mp,
-					tp, ip, whichfork);
-				cur->bc_private.b.firstblock =
-					*firstblock;
-				cur->bc_private.b.flist = flist;
-			}
-			mval->br_state = (mval->br_state == XFS_EXT_UNWRITTEN)
-						? XFS_EXT_NORM
-						: XFS_EXT_UNWRITTEN;
-			error = xfs_bmap_add_extent(tp, ip, &lastx, &cur, mval,
-				firstblock, flist, &tmp_logflags,
-				whichfork);
-			logflags |= tmp_logflags;
-			if (error)
-				goto error0;
-			ep = xfs_iext_get_ext(ifp, lastx);
-			nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
-			xfs_bmbt_get_all(ep, &got);
-			/*
-			 * We may have combined previously unwritten
-			 * space with written space, so generate
-			 * another request.
-			 */
-			if (mval->br_blockcount < len)
-				continue;
-		}
+		/* Deal with the allocated space we found.  */
+		xfs_bmapi_trim_map(mval, &bma.got, &bno, len, obno,
+							end, n, flags);
+
+		/* Execute unwritten extent conversion if necessary */
+		error = xfs_bmapi_convert_unwritten(&bma, mval, len, flags);
+		if (error == EAGAIN)
+			continue;
+		if (error)
+			goto error0;
+
+		/* update the extent map to return */
+		xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);
 
-		ASSERT((flags & XFS_BMAPI_ENTIRE) ||
-		       ((mval->br_startoff + mval->br_blockcount) <= end));
-		ASSERT((flags & XFS_BMAPI_ENTIRE) ||
-		       (mval->br_blockcount <= len) ||
-		       (mval->br_startoff < obno));
-		bno = mval->br_startoff + mval->br_blockcount;
-		len = end - bno;
-		if (n > 0 && mval->br_startoff == mval[-1].br_startoff) {
-			ASSERT(mval->br_startblock == mval[-1].br_startblock);
-			ASSERT(mval->br_blockcount > mval[-1].br_blockcount);
-			ASSERT(mval->br_state == mval[-1].br_state);
-			mval[-1].br_blockcount = mval->br_blockcount;
-			mval[-1].br_state = mval->br_state;
-		} else if (n > 0 && mval->br_startblock != DELAYSTARTBLOCK &&
-			   mval[-1].br_startblock != DELAYSTARTBLOCK &&
-			   mval[-1].br_startblock != HOLESTARTBLOCK &&
-			   mval->br_startblock ==
-			   mval[-1].br_startblock + mval[-1].br_blockcount &&
-			   ((flags & XFS_BMAPI_IGSTATE) ||
-				mval[-1].br_state == mval->br_state)) {
-			ASSERT(mval->br_startoff ==
-			       mval[-1].br_startoff + mval[-1].br_blockcount);
-			mval[-1].br_blockcount += mval->br_blockcount;
-		} else if (n > 0 &&
-			   mval->br_startblock == DELAYSTARTBLOCK &&
-			   mval[-1].br_startblock == DELAYSTARTBLOCK &&
-			   mval->br_startoff ==
-			   mval[-1].br_startoff + mval[-1].br_blockcount) {
-			mval[-1].br_blockcount += mval->br_blockcount;
-			mval[-1].br_state = mval->br_state;
-		} else if (!((n == 0) &&
-			     ((mval->br_startoff + mval->br_blockcount) <=
-			      obno))) {
-			mval++;
-			n++;
-		}
 		/*
 		 * If we're done, stop now.  Stop when we've allocated
 		 * XFS_BMAP_MAX_NMAP extents no matter what.  Otherwise
 		 * the transaction may get too big.
 		 */
-		if (bno >= end || n >= *nmap || nallocs >= *nmap)
+		if (bno >= end || n >= *nmap || bma.nallocs >= *nmap)
 			break;
-		/*
-		 * Else go on to the next record.
-		 */
-		prev = got;
-		if (++lastx < nextents) {
-			ep = xfs_iext_get_ext(ifp, lastx);
-			xfs_bmbt_get_all(ep, &got);
-		} else {
+
+		/* Else go on to the next record. */
+		bma.prev = bma.got;
+		if (++bma.idx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t)) {
+			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma.idx),
+					 &bma.got);
+		} else
 			eof = 1;
-		}
 	}
 	*nmap = n;
+
 	/*
 	 * Transform from btree to extents, give it cur.
 	 */
-	if (tp && XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE &&
+	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE &&
 	    XFS_IFORK_NEXTENTS(ip, whichfork) <= ifp->if_ext_max) {
-		ASSERT(wr && cur);
-		error = xfs_bmap_btree_to_extents(tp, ip, cur,
+		int		tmp_logflags = 0;
+
+		ASSERT(bma.cur);
+		error = xfs_bmap_btree_to_extents(tp, ip, bma.cur,
 			&tmp_logflags, whichfork);
-		logflags |= tmp_logflags;
+		bma.logflags |= tmp_logflags;
 		if (error)
 			goto error0;
 	}
@@ -4855,34 +4984,33 @@ error0:
 	 * Log everything.  Do this after conversion, there's no point in
 	 * logging the extent records if we've converted to btree format.
 	 */
-	if ((logflags & xfs_ilog_fext(whichfork)) &&
+	if ((bma.logflags & xfs_ilog_fext(whichfork)) &&
 	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
-		logflags &= ~xfs_ilog_fext(whichfork);
-	else if ((logflags & xfs_ilog_fbroot(whichfork)) &&
+		bma.logflags &= ~xfs_ilog_fext(whichfork);
+	else if ((bma.logflags & xfs_ilog_fbroot(whichfork)) &&
 		 XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)
-		logflags &= ~xfs_ilog_fbroot(whichfork);
+		bma.logflags &= ~xfs_ilog_fbroot(whichfork);
 	/*
 	 * Log whatever the flags say, even if error.  Otherwise we might miss
 	 * detecting a case where the data is changed, there's an error,
 	 * and it's not logged so we don't shutdown when we should.
 	 */
-	if (logflags) {
-		ASSERT(tp && wr);
-		xfs_trans_log_inode(tp, ip, logflags);
-	}
-	if (cur) {
+	if (bma.logflags)
+		xfs_trans_log_inode(tp, ip, bma.logflags);
+
+	if (bma.cur) {
 		if (!error) {
 			ASSERT(*firstblock == NULLFSBLOCK ||
 			       XFS_FSB_TO_AGNO(mp, *firstblock) ==
 			       XFS_FSB_TO_AGNO(mp,
-				       cur->bc_private.b.firstblock) ||
+				       bma.cur->bc_private.b.firstblock) ||
 			       (flist->xbf_low &&
 				XFS_FSB_TO_AGNO(mp, *firstblock) <
 				XFS_FSB_TO_AGNO(mp,
-					cur->bc_private.b.firstblock)));
-			*firstblock = cur->bc_private.b.firstblock;
+					bma.cur->bc_private.b.firstblock)));
+			*firstblock = bma.cur->bc_private.b.firstblock;
 		}
-		xfs_btree_del_cursor(cur,
+		xfs_btree_del_cursor(bma.cur,
 			error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
 	}
 	if (!error)
@@ -4891,58 +5019,6 @@ error0:
 	return error;
 }
 
-/*
- * Map file blocks to filesystem blocks, simple version.
- * One block (extent) only, read-only.
- * For flags, only the XFS_BMAPI_ATTRFORK flag is examined.
- * For the other flag values, the effect is as if XFS_BMAPI_METADATA
- * was set and all the others were clear.
- */
-int						/* error */
-xfs_bmapi_single(
-	xfs_trans_t	*tp,		/* transaction pointer */
-	xfs_inode_t	*ip,		/* incore inode */
-	int		whichfork,	/* data or attr fork */
-	xfs_fsblock_t	*fsb,		/* output: mapped block */
-	xfs_fileoff_t	bno)		/* starting file offs. mapped */
-{
-	int		eof;		/* we've hit the end of extents */
-	int		error;		/* error return */
-	xfs_bmbt_irec_t	got;		/* current file extent record */
-	xfs_ifork_t	*ifp;		/* inode fork pointer */
-	xfs_extnum_t	lastx;		/* last useful extent number */
-	xfs_bmbt_irec_t	prev;		/* previous file extent record */
-
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	if (unlikely(
-	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
-	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)) {
-	       XFS_ERROR_REPORT("xfs_bmapi_single", XFS_ERRLEVEL_LOW,
-				ip->i_mount);
-	       return XFS_ERROR(EFSCORRUPTED);
-	}
-	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
-		return XFS_ERROR(EIO);
-	XFS_STATS_INC(xs_blk_mapr);
-	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
-	    (error = xfs_iread_extents(tp, ip, whichfork)))
-		return error;
-	(void)xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got,
-		&prev);
-	/*
-	 * Reading past eof, act as though there's a hole
-	 * up to end.
-	 */
-	if (eof || got.br_startoff > bno) {
-		*fsb = NULLFSBLOCK;
-		return 0;
-	}
-	ASSERT(!isnullstartblock(got.br_startblock));
-	ASSERT(bno < got.br_startoff + got.br_blockcount);
-	*fsb = got.br_startblock + (bno - got.br_startoff);
-	return 0;
-}
-
 /*
  * Unmap (remove) blocks from a file.
  * If nexts is nonzero then the number of extents to remove is limited to
@@ -5114,9 +5190,9 @@ xfs_bunmapi(
 				del.br_blockcount = mod;
 			}
 			del.br_state = XFS_EXT_UNWRITTEN;
-			error = xfs_bmap_add_extent(tp, ip, &lastx, &cur, &del,
-				firstblock, flist, &logflags,
-				XFS_DATA_FORK);
+			error = xfs_bmap_add_extent_unwritten_real(tp, ip,
+					&lastx, &cur, &del, firstblock, flist,
+					&logflags);
 			if (error)
 				goto error0;
 			goto nodelete;
@@ -5172,18 +5248,18 @@ xfs_bunmapi(
 				}
 				prev.br_state = XFS_EXT_UNWRITTEN;
 				lastx--;
-				error = xfs_bmap_add_extent(tp, ip, &lastx,
-						&cur, &prev, firstblock, flist,
-						&logflags, XFS_DATA_FORK);
+				error = xfs_bmap_add_extent_unwritten_real(tp,
+						ip, &lastx, &cur, &prev,
+						firstblock, flist, &logflags);
 				if (error)
 					goto error0;
 				goto nodelete;
 			} else {
 				ASSERT(del.br_state == XFS_EXT_NORM);
 				del.br_state = XFS_EXT_UNWRITTEN;
-				error = xfs_bmap_add_extent(tp, ip, &lastx,
-						&cur, &del, firstblock, flist,
-						&logflags, XFS_DATA_FORK);
+				error = xfs_bmap_add_extent_unwritten_real(tp,
+						ip, &lastx, &cur, &del,
+						firstblock, flist, &logflags);
 				if (error)
 					goto error0;
 				goto nodelete;
@@ -5505,10 +5581,9 @@ xfs_getbmap(
 
 	do {
 		nmap = (nexleft > subnex) ? subnex : nexleft;
-		error = xfs_bmapi(NULL, ip, XFS_BB_TO_FSBT(mp, bmv->bmv_offset),
-				  XFS_BB_TO_FSB(mp, bmv->bmv_length),
-				  bmapi_flags, NULL, 0, map, &nmap,
-				  NULL);
+		error = xfs_bmapi_read(ip, XFS_BB_TO_FSBT(mp, bmv->bmv_offset),
+				       XFS_BB_TO_FSB(mp, bmv->bmv_length),
+				       map, &nmap, bmapi_flags);
 		if (error)
 			goto out_free_map;
 		ASSERT(nmap <= subnex);
@@ -5582,89 +5657,6 @@ xfs_getbmap(
 	return error;
 }
 
-/*
- * Check the last inode extent to determine whether this allocation will result
- * in blocks being allocated at the end of the file. When we allocate new data
- * blocks at the end of the file which do not start at the previous data block,
- * we will try to align the new blocks at stripe unit boundaries.
- */
-STATIC int				/* error */
-xfs_bmap_isaeof(
-	xfs_inode_t	*ip,		/* incore inode pointer */
-	xfs_fileoff_t   off,		/* file offset in fsblocks */
-	int             whichfork,	/* data or attribute fork */
-	char		*aeof)		/* return value */
-{
-	int		error;		/* error return value */
-	xfs_ifork_t	*ifp;		/* inode fork pointer */
-	xfs_bmbt_rec_host_t *lastrec;	/* extent record pointer */
-	xfs_extnum_t	nextents;	/* number of file extents */
-	xfs_bmbt_irec_t	s;		/* expanded extent record */
-
-	ASSERT(whichfork == XFS_DATA_FORK);
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
-	    (error = xfs_iread_extents(NULL, ip, whichfork)))
-		return error;
-	nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
-	if (nextents == 0) {
-		*aeof = 1;
-		return 0;
-	}
-	/*
-	 * Go to the last extent
-	 */
-	lastrec = xfs_iext_get_ext(ifp, nextents - 1);
-	xfs_bmbt_get_all(lastrec, &s);
-	/*
-	 * Check we are allocating in the last extent (for delayed allocations)
-	 * or past the last extent for non-delayed allocations.
-	 */
-	*aeof = (off >= s.br_startoff &&
-		 off < s.br_startoff + s.br_blockcount &&
-		 isnullstartblock(s.br_startblock)) ||
-		off >= s.br_startoff + s.br_blockcount;
-	return 0;
-}
-
-/*
- * Check if the endoff is outside the last extent. If so the caller will grow
- * the allocation to a stripe unit boundary.
- */
-int					/* error */
-xfs_bmap_eof(
-	xfs_inode_t	*ip,		/* incore inode pointer */
-	xfs_fileoff_t	endoff,		/* file offset in fsblocks */
-	int		whichfork,	/* data or attribute fork */
-	int		*eof)		/* result value */
-{
-	xfs_fsblock_t	blockcount;	/* extent block count */
-	int		error;		/* error return value */
-	xfs_ifork_t	*ifp;		/* inode fork pointer */
-	xfs_bmbt_rec_host_t *lastrec;	/* extent record pointer */
-	xfs_extnum_t	nextents;	/* number of file extents */
-	xfs_fileoff_t	startoff;	/* extent starting file offset */
-
-	ASSERT(whichfork == XFS_DATA_FORK);
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
-	    (error = xfs_iread_extents(NULL, ip, whichfork)))
-		return error;
-	nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
-	if (nextents == 0) {
-		*eof = 1;
-		return 0;
-	}
-	/*
-	 * Go to the last extent
-	 */
-	lastrec = xfs_iext_get_ext(ifp, nextents - 1);
-	startoff = xfs_bmbt_get_startoff(lastrec);
-	blockcount = xfs_bmbt_get_blockcount(lastrec);
-	*eof = endoff >= startoff + blockcount;
-	return 0;
-}
-
 #ifdef DEBUG
 STATIC struct xfs_buf *
 xfs_bmap_get_bp(
@@ -6099,9 +6091,8 @@ xfs_bmap_punch_delalloc_range(
 		 * trying to remove a real extent (which requires a
 		 * transaction) or a hole, which is probably a bad idea...
 		 */
-		error = xfs_bmapi(NULL, ip, start_fsb, 1,
-				XFS_BMAPI_ENTIRE,  NULL, 0, &imap,
-				&nimaps, NULL);
+		error = xfs_bmapi_read(ip, start_fsb, 1, &imap, &nimaps,
+				       XFS_BMAPI_ENTIRE);
 
 		if (error) {
 			/* something screwed, just bail */

fs/xfs/xfs_bmap.h

@@ -62,27 +62,23 @@ typedef	struct xfs_bmap_free
 #define	XFS_BMAP_MAX_NMAP	4
 
 /*
- * Flags for xfs_bmapi
+ * Flags for xfs_bmapi_*
  */
-#define	XFS_BMAPI_WRITE		0x001	/* write operation: allocate space */
-#define XFS_BMAPI_DELAY		0x002	/* delayed write operation */
-#define XFS_BMAPI_ENTIRE	0x004	/* return entire extent, not trimmed */
-#define XFS_BMAPI_METADATA	0x008	/* mapping metadata not user data */
-#define XFS_BMAPI_ATTRFORK	0x010	/* use attribute fork not data */
-#define	XFS_BMAPI_PREALLOC	0x040	/* preallocation op: unwritten space */
-#define	XFS_BMAPI_IGSTATE	0x080	/* Ignore state - */
+#define XFS_BMAPI_ENTIRE	0x001	/* return entire extent, not trimmed */
+#define XFS_BMAPI_METADATA	0x002	/* mapping metadata not user data */
+#define XFS_BMAPI_ATTRFORK	0x004	/* use attribute fork not data */
+#define XFS_BMAPI_PREALLOC	0x008	/* preallocation op: unwritten space */
+#define XFS_BMAPI_IGSTATE	0x010	/* Ignore state - */
 					/* combine contig. space */
-#define	XFS_BMAPI_CONTIG	0x100	/* must allocate only one extent */
+#define XFS_BMAPI_CONTIG	0x020	/* must allocate only one extent */
 /*
  * unwritten extent conversion - this needs write cache flushing and no additional
  * allocation alignments. When specified with XFS_BMAPI_PREALLOC it converts
  * from written to unwritten, otherwise convert from unwritten to written.
  */
-#define XFS_BMAPI_CONVERT	0x200
+#define XFS_BMAPI_CONVERT	0x040
 
 #define XFS_BMAPI_FLAGS \
-	{ XFS_BMAPI_WRITE,	"WRITE" }, \
-	{ XFS_BMAPI_DELAY,	"DELAY" }, \
 	{ XFS_BMAPI_ENTIRE,	"ENTIRE" }, \
 	{ XFS_BMAPI_METADATA,	"METADATA" }, \
 	{ XFS_BMAPI_ATTRFORK,	"ATTRFORK" }, \
@@ -113,21 +109,28 @@ static inline void xfs_bmap_init(xfs_bmap_free_t *flp, xfs_fsblock_t *fbp)
  * Argument structure for xfs_bmap_alloc.
  */
 typedef struct xfs_bmalloca {
-	xfs_fsblock_t		firstblock; /* i/o first block allocated */
-	xfs_fsblock_t		rval;	/* starting block of new extent */
-	xfs_fileoff_t		off;	/* offset in file filling in */
+	xfs_fsblock_t		*firstblock; /* i/o first block allocated */
+	struct xfs_bmap_free	*flist;	/* bmap freelist */
 	struct xfs_trans	*tp;	/* transaction pointer */
 	struct xfs_inode	*ip;	/* incore inode pointer */
-	struct xfs_bmbt_irec	*prevp;	/* extent before the new one */
-	struct xfs_bmbt_irec	*gotp;	/* extent after, or delayed */
-	xfs_extlen_t		alen;	/* i/o length asked/allocated */
+	struct xfs_bmbt_irec	prev;	/* extent before the new one */
+	struct xfs_bmbt_irec	got;	/* extent after, or delayed */
+
+	xfs_fileoff_t		offset;	/* offset in file filling in */
+	xfs_extlen_t		length;	/* i/o length asked/allocated */
+	xfs_fsblock_t		blkno;	/* starting block of new extent */
+
+	struct xfs_btree_cur	*cur;	/* btree cursor */
+	xfs_extnum_t		idx;	/* current extent index */
+	int			nallocs;/* number of extents alloc'd */
+	int			logflags;/* flags for transaction logging */
+
 	xfs_extlen_t		total;	/* total blocks needed for xaction */
 	xfs_extlen_t		minlen;	/* minimum allocation size (blocks) */
 	xfs_extlen_t		minleft; /* amount must be left after alloc */
 	char			eof;	/* set if allocating past last extent */
 	char			wasdel;	/* replacing a delayed allocation */
 	char			userdata;/* set if is user data */
-	char			low;	/* low on space, using seq'l ags */
 	char			aeof;	/* allocated space at eof */
 	char			conv;	/* overwriting unwritten extents */
 } xfs_bmalloca_t;
@@ -152,251 +155,62 @@ typedef struct xfs_bmalloca {
 	{ BMAP_RIGHT_FILLING,	"RF" }, \
 	{ BMAP_ATTRFORK,	"ATTR" }
 
-/*
- * Add bmap trace insert entries for all the contents of the extent list.
- *
- * Quite excessive tracing.  Only do this for debug builds.
- */
 #if defined(__KERNEL) && defined(DEBUG)
-void
-xfs_bmap_trace_exlist(
-	struct xfs_inode	*ip,		/* incore inode pointer */
-	xfs_extnum_t		cnt,		/* count of entries in list */
-	int			whichfork,
-	unsigned long		caller_ip);	/* data or attr fork */
+void	xfs_bmap_trace_exlist(struct xfs_inode *ip, xfs_extnum_t cnt,
+		int whichfork, unsigned long caller_ip);
 #define	XFS_BMAP_TRACE_EXLIST(ip,c,w)	\
 	xfs_bmap_trace_exlist(ip,c,w, _THIS_IP_)
 #else
 #define	XFS_BMAP_TRACE_EXLIST(ip,c,w)
 #endif
 
-/*
- * Convert inode from non-attributed to attributed.
- * Must not be in a transaction, ip must not be locked.
- */
-int					/* error code */
-xfs_bmap_add_attrfork(
-	struct xfs_inode	*ip,	/* incore inode pointer */
-	int			size,	/* space needed for new attribute */
-	int			rsvd);	/* flag for reserved block allocation */
-
-/*
- * Add the extent to the list of extents to be free at transaction end.
- * The list is maintained sorted (by block number).
- */
-void
-xfs_bmap_add_free(
-	xfs_fsblock_t		bno,		/* fs block number of extent */
-	xfs_filblks_t		len,		/* length of extent */
-	xfs_bmap_free_t		*flist,		/* list of extents */
-	struct xfs_mount	*mp);		/* mount point structure */
-
-/*
- * Routine to clean up the free list data structure when
- * an error occurs during a transaction.
- */
-void
-xfs_bmap_cancel(
-	xfs_bmap_free_t		*flist);	/* free list to clean up */
-
-/*
- * Compute and fill in the value of the maximum depth of a bmap btree
- * in this filesystem.  Done once, during mount.
- */
-void
-xfs_bmap_compute_maxlevels(
-	struct xfs_mount	*mp,	/* file system mount structure */
-	int			whichfork);	/* data or attr fork */
-
-/*
- * Returns the file-relative block number of the first unused block in the file.
- * This is the lowest-address hole if the file has holes, else the first block
- * past the end of file.
- */
-int						/* error */
-xfs_bmap_first_unused(
-	struct xfs_trans	*tp,		/* transaction pointer */
-	struct xfs_inode	*ip,		/* incore inode */
-	xfs_extlen_t		len,		/* size of hole to find */
-	xfs_fileoff_t		*unused,	/* unused block num */
-	int			whichfork);	/* data or attr fork */
-
-/*
- * Returns the file-relative block number of the last block + 1 before
- * last_block (input value) in the file.
- * This is not based on i_size, it is based on the extent list.
- * Returns 0 for local files, as they do not have an extent list.
- */
-int						/* error */
-xfs_bmap_last_before(
-	struct xfs_trans	*tp,		/* transaction pointer */
-	struct xfs_inode	*ip,		/* incore inode */
-	xfs_fileoff_t		*last_block,	/* last block */
-	int			whichfork);	/* data or attr fork */
-
-/*
- * Returns the file-relative block number of the first block past eof in
- * the file.  This is not based on i_size, it is based on the extent list.
- * Returns 0 for local files, as they do not have an extent list.
- */
-int						/* error */
-xfs_bmap_last_offset(
-	struct xfs_trans	*tp,		/* transaction pointer */
-	struct xfs_inode	*ip,		/* incore inode */
-	xfs_fileoff_t		*unused,	/* last block num */
-	int			whichfork);	/* data or attr fork */
-
-/*
- * Returns whether the selected fork of the inode has exactly one
- * block or not.  For the data fork we check this matches di_size,
- * implying the file's range is 0..bsize-1.
- */
-int
-xfs_bmap_one_block(
-	struct xfs_inode	*ip,		/* incore inode */
-	int			whichfork);	/* data or attr fork */
-
-/*
- * Read in the extents to iu_extents.
- * All inode fields are set up by caller, we just traverse the btree
- * and copy the records in.
- */
-int						/* error */
-xfs_bmap_read_extents(
-	struct xfs_trans	*tp,		/* transaction pointer */
-	struct xfs_inode	*ip,		/* incore inode */
-	int			whichfork);	/* data or attr fork */
-
-/*
- * Map file blocks to filesystem blocks.
- * File range is given by the bno/len pair.
- * Adds blocks to file if a write ("flags & XFS_BMAPI_WRITE" set)
- * into a hole or past eof.
- * Only allocates blocks from a single allocation group,
- * to avoid locking problems.
- * The returned value in "firstblock" from the first call in a transaction
- * must be remembered and presented to subsequent calls in "firstblock".
- * An upper bound for the number of blocks to be allocated is supplied to
- * the first call in "total"; if no allocation group has that many free
- * blocks then the call will fail (return NULLFSBLOCK in "firstblock").
- */
-int						/* error */
-xfs_bmapi(
-	struct xfs_trans	*tp,		/* transaction pointer */
-	struct xfs_inode	*ip,		/* incore inode */
-	xfs_fileoff_t		bno,		/* starting file offs. mapped */
-	xfs_filblks_t		len,		/* length to map in file */
-	int			flags,		/* XFS_BMAPI_... */
-	xfs_fsblock_t		*firstblock,	/* first allocated block
-						   controls a.g. for allocs */
-	xfs_extlen_t		total,		/* total blocks needed */
-	struct xfs_bmbt_irec	*mval,		/* output: map values */
-	int			*nmap,		/* i/o: mval size/count */
-	xfs_bmap_free_t		*flist);	/* i/o: list extents to free */
-
-/*
- * Map file blocks to filesystem blocks, simple version.
- * One block only, read-only.
- * For flags, only the XFS_BMAPI_ATTRFORK flag is examined.
- * For the other flag values, the effect is as if XFS_BMAPI_METADATA
- * was set and all the others were clear.
- */
-int						/* error */
-xfs_bmapi_single(
-	struct xfs_trans	*tp,		/* transaction pointer */
-	struct xfs_inode	*ip,		/* incore inode */
-	int			whichfork,	/* data or attr fork */
-	xfs_fsblock_t		*fsb,		/* output: mapped block */
-	xfs_fileoff_t		bno);		/* starting file offs. mapped */
-
-/*
- * Unmap (remove) blocks from a file.
- * If nexts is nonzero then the number of extents to remove is limited to
- * that value.  If not all extents in the block range can be removed then
- * *done is set.
- */
-int						/* error */
-xfs_bunmapi(
-	struct xfs_trans	*tp,		/* transaction pointer */
-	struct xfs_inode	*ip,		/* incore inode */
-	xfs_fileoff_t		bno,		/* starting offset to unmap */
-	xfs_filblks_t		len,		/* length to unmap in file */
-	int			flags,		/* XFS_BMAPI_... */
-	xfs_extnum_t		nexts,		/* number of extents max */
-	xfs_fsblock_t		*firstblock,	/* first allocated block
-						   controls a.g. for allocs */
-	xfs_bmap_free_t		*flist,		/* i/o: list extents to free */
-	int			*done);		/* set if not done yet */
-
-/*
- * Check an extent list, which has just been read, for
- * any bit in the extent flag field.
- */
-int
-xfs_check_nostate_extents(
-	struct xfs_ifork	*ifp,
-	xfs_extnum_t		idx,
-	xfs_extnum_t		num);
-
-uint
-xfs_default_attroffset(
-	struct xfs_inode	*ip);
+int	xfs_bmap_add_attrfork(struct xfs_inode *ip, int size, int rsvd);
+void	xfs_bmap_add_free(xfs_fsblock_t bno, xfs_filblks_t len,
+		struct xfs_bmap_free *flist, struct xfs_mount *mp);
+void	xfs_bmap_cancel(struct xfs_bmap_free *flist);
+void	xfs_bmap_compute_maxlevels(struct xfs_mount *mp, int whichfork);
+int	xfs_bmap_first_unused(struct xfs_trans *tp, struct xfs_inode *ip,
+		xfs_extlen_t len, xfs_fileoff_t *unused, int whichfork);
+int	xfs_bmap_last_before(struct xfs_trans *tp, struct xfs_inode *ip,
+		xfs_fileoff_t *last_block, int whichfork);
+int	xfs_bmap_last_offset(struct xfs_trans *tp, struct xfs_inode *ip,
+		xfs_fileoff_t *unused, int whichfork);
+int	xfs_bmap_one_block(struct xfs_inode *ip, int whichfork);
+int	xfs_bmap_read_extents(struct xfs_trans *tp, struct xfs_inode *ip,
+		int whichfork);
+int	xfs_bmapi_read(struct xfs_inode *ip, xfs_fileoff_t bno,
+		xfs_filblks_t len, struct xfs_bmbt_irec *mval,
+		int *nmap, int flags);
+int	xfs_bmapi_delay(struct xfs_inode *ip, xfs_fileoff_t bno,
+		xfs_filblks_t len, struct xfs_bmbt_irec *mval,
+		int *nmap, int flags);
+int	xfs_bmapi_write(struct xfs_trans *tp, struct xfs_inode *ip,
+		xfs_fileoff_t bno, xfs_filblks_t len, int flags,
+		xfs_fsblock_t *firstblock, xfs_extlen_t total,
+		struct xfs_bmbt_irec *mval, int *nmap,
+		struct xfs_bmap_free *flist);
+int	xfs_bunmapi(struct xfs_trans *tp, struct xfs_inode *ip,
+		xfs_fileoff_t bno, xfs_filblks_t len, int flags,
+		xfs_extnum_t nexts, xfs_fsblock_t *firstblock,
+		struct xfs_bmap_free *flist, int *done);
+int	xfs_check_nostate_extents(struct xfs_ifork *ifp, xfs_extnum_t idx,
+		xfs_extnum_t num);
+uint	xfs_default_attroffset(struct xfs_inode *ip);
 
 #ifdef __KERNEL__
-
-/*
- * Routine to be called at transaction's end by xfs_bmapi, xfs_bunmapi
- * caller.  Frees all the extents that need freeing, which must be done
- * last due to locking considerations.
- *
- * Return 1 if the given transaction was committed and a new one allocated,
- * and 0 otherwise.
- */
-int						/* error */
-xfs_bmap_finish(
-	struct xfs_trans	**tp,		/* transaction pointer addr */
-	xfs_bmap_free_t		*flist,		/* i/o: list extents to free */
-	int			*committed);	/* xact committed or not */
-
 /* bmap to userspace formatter - copy to user & advance pointer */
 typedef int (*xfs_bmap_format_t)(void **, struct getbmapx *, int *);
 
-/*
- * Get inode's extents as described in bmv, and format for output.
- */
-int						/* error code */
-xfs_getbmap(
-	xfs_inode_t		*ip,
-	struct getbmapx		*bmv,		/* user bmap structure */
-	xfs_bmap_format_t	formatter,	/* format to user */
-	void			*arg);		/* formatter arg */
-
-/*
- * Check if the endoff is outside the last extent. If so the caller will grow
- * the allocation to a stripe unit boundary
- */
-int
-xfs_bmap_eof(
-	struct xfs_inode        *ip,
-	xfs_fileoff_t           endoff,
-	int                     whichfork,
-	int                     *eof);
-
-/*
- * Count fsblocks of the given fork.
- */
-int
-xfs_bmap_count_blocks(
-	xfs_trans_t		*tp,
-	struct xfs_inode	*ip,
-	int			whichfork,
-	int			*count);
-
-int
-xfs_bmap_punch_delalloc_range(
-	struct xfs_inode	*ip,
-	xfs_fileoff_t		start_fsb,
-	xfs_fileoff_t		length);
+int	xfs_bmap_finish(struct xfs_trans **tp, struct xfs_bmap_free *flist,
+		int *committed);
+int	xfs_getbmap(struct xfs_inode *ip, struct getbmapx *bmv,
+		xfs_bmap_format_t formatter, void *arg);
+int	xfs_bmap_eof(struct xfs_inode *ip, xfs_fileoff_t endoff,
+		int whichfork, int *eof);
+int	xfs_bmap_count_blocks(struct xfs_trans *tp, struct xfs_inode *ip,
+		int whichfork, int *count);
+int	xfs_bmap_punch_delalloc_range(struct xfs_inode *ip,
+		xfs_fileoff_t start_fsb, xfs_fileoff_t length);
 #endif	/* __KERNEL__ */
 
 #endif	/* __XFS_BMAP_H__ */

fs/xfs/xfs_btree.c

@@ -631,7 +631,7 @@ xfs_btree_read_bufl(
 	}
 	ASSERT(!xfs_buf_geterror(bp));
 	if (bp)
-		XFS_BUF_SET_VTYPE_REF(bp, B_FS_MAP, refval);
+		xfs_buf_set_ref(bp, refval);
 	*bpp = bp;
 	return 0;
 }
@@ -939,13 +939,13 @@ xfs_btree_set_refs(
 	switch (cur->bc_btnum) {
 	case XFS_BTNUM_BNO:
 	case XFS_BTNUM_CNT:
-		XFS_BUF_SET_VTYPE_REF(bp, B_FS_MAP, XFS_ALLOC_BTREE_REF);
+		xfs_buf_set_ref(bp, XFS_ALLOC_BTREE_REF);
 		break;
 	case XFS_BTNUM_INO:
-		XFS_BUF_SET_VTYPE_REF(bp, B_FS_INOMAP, XFS_INO_BTREE_REF);
+		xfs_buf_set_ref(bp, XFS_INO_BTREE_REF);
 		break;
 	case XFS_BTNUM_BMAP:
-		XFS_BUF_SET_VTYPE_REF(bp, B_FS_MAP, XFS_BMAP_BTREE_REF);
+		xfs_buf_set_ref(bp, XFS_BMAP_BTREE_REF);
 		break;
 	default:
 		ASSERT(0);
@@ -970,7 +970,8 @@ xfs_btree_get_buf_block(
 	*bpp = xfs_trans_get_buf(cur->bc_tp, mp->m_ddev_targp, d,
 				 mp->m_bsize, flags);
 
-	ASSERT(!xfs_buf_geterror(*bpp));
+	if (!*bpp)
+		return ENOMEM;
 
 	*block = XFS_BUF_TO_BLOCK(*bpp);
 	return 0;

fs/xfs/xfs_buf.c

@@ -43,7 +43,6 @@
 
 static kmem_zone_t *xfs_buf_zone;
 STATIC int xfsbufd(void *);
-STATIC void xfs_buf_delwri_queue(xfs_buf_t *, int);
 
 static struct workqueue_struct *xfslogd_workqueue;
 struct workqueue_struct *xfsdatad_workqueue;
@@ -66,10 +65,6 @@ struct workqueue_struct *xfsconvertd_workqueue;
 #define xb_to_km(flags) \
 	 (((flags) & XBF_DONT_BLOCK) ? KM_NOFS : KM_SLEEP)
 
-#define xfs_buf_allocate(flags) \
-	kmem_zone_alloc(xfs_buf_zone, xb_to_km(flags))
-#define xfs_buf_deallocate(bp) \
-	kmem_zone_free(xfs_buf_zone, (bp));
 
 static inline int
 xfs_buf_is_vmapped(
@@ -152,6 +147,7 @@ xfs_buf_stale(
 	struct xfs_buf	*bp)
 {
 	bp->b_flags |= XBF_STALE;
+	xfs_buf_delwri_dequeue(bp);
 	atomic_set(&(bp)->b_lru_ref, 0);
 	if (!list_empty(&bp->b_lru)) {
 		struct xfs_buftarg *btp = bp->b_target;
@@ -167,14 +163,19 @@ xfs_buf_stale(
 	ASSERT(atomic_read(&bp->b_hold) >= 1);
 }
 
-STATIC void
-_xfs_buf_initialize(
-	xfs_buf_t		*bp,
-	xfs_buftarg_t		*target,
+struct xfs_buf *
+xfs_buf_alloc(
+	struct xfs_buftarg	*target,
 	xfs_off_t		range_base,
 	size_t			range_length,
 	xfs_buf_flags_t		flags)
 {
+	struct xfs_buf		*bp;
+
+	bp = kmem_zone_alloc(xfs_buf_zone, xb_to_km(flags));
+	if (unlikely(!bp))
+		return NULL;
+
 	/*
 	 * We don't want certain flags to appear in b_flags.
 	 */
@@ -203,8 +204,9 @@ _xfs_buf_initialize(
 	init_waitqueue_head(&bp->b_waiters);
 
 	XFS_STATS_INC(xb_create);
-
 	trace_xfs_buf_init(bp, _RET_IP_);
+
+	return bp;
 }
 
 /*
@@ -277,7 +279,7 @@ xfs_buf_free(
 	} else if (bp->b_flags & _XBF_KMEM)
 		kmem_free(bp->b_addr);
 	_xfs_buf_free_pages(bp);
-	xfs_buf_deallocate(bp);
+	kmem_zone_free(xfs_buf_zone, bp);
 }
 
 /*
@@ -416,10 +418,7 @@ _xfs_buf_map_pages(
 /*
  *	Look up, and creates if absent, a lockable buffer for
  *	a given range of an inode.  The buffer is returned
- *	locked.	 If other overlapping buffers exist, they are
- *	released before the new buffer is created and locked,
- *	which may imply that this call will block until those buffers
- *	are unlocked.  No I/O is implied by this call.
+ *	locked.	No I/O is implied by this call.
  */
 xfs_buf_t *
 _xfs_buf_find(
@@ -481,8 +480,6 @@ _xfs_buf_find(
 
 	/* No match found */
 	if (new_bp) {
-		_xfs_buf_initialize(new_bp, btp, range_base,
-				range_length, flags);
 		rb_link_node(&new_bp->b_rbnode, parent, rbp);
 		rb_insert_color(&new_bp->b_rbnode, &pag->pag_buf_tree);
 		/* the buffer keeps the perag reference until it is freed */
@@ -525,35 +522,51 @@ found:
 }
 
 /*
- *	Assembles a buffer covering the specified range.
- *	Storage in memory for all portions of the buffer will be allocated,
- *	although backing storage may not be.
+ * Assembles a buffer covering the specified range. The code is optimised for
+ * cache hits, as metadata intensive workloads will see 3 orders of magnitude
+ * more hits than misses.
  */
-xfs_buf_t *
+struct xfs_buf *
 xfs_buf_get(
 	xfs_buftarg_t		*target,/* target for buffer		*/
 	xfs_off_t		ioff,	/* starting offset of range	*/
 	size_t			isize,	/* length of range		*/
 	xfs_buf_flags_t		flags)
 {
-	xfs_buf_t		*bp, *new_bp;
+	struct xfs_buf		*bp;
+	struct xfs_buf		*new_bp;
 	int			error = 0;
 
-	new_bp = xfs_buf_allocate(flags);
+	bp = _xfs_buf_find(target, ioff, isize, flags, NULL);
+	if (likely(bp))
+		goto found;
+
+	new_bp = xfs_buf_alloc(target, ioff << BBSHIFT, isize << BBSHIFT,
+			       flags);
 	if (unlikely(!new_bp))
 		return NULL;
 
 	bp = _xfs_buf_find(target, ioff, isize, flags, new_bp);
+	if (!bp) {
+		kmem_zone_free(xfs_buf_zone, new_bp);
+		return NULL;
+	}
+
 	if (bp == new_bp) {
 		error = xfs_buf_allocate_memory(bp, flags);
 		if (error)
 			goto no_buffer;
-	} else {
-		xfs_buf_deallocate(new_bp);
-		if (unlikely(bp == NULL))
-			return NULL;
-	}
+	} else
+		kmem_zone_free(xfs_buf_zone, new_bp);
 
+	/*
+	 * Now we have a workable buffer, fill in the block number so
+	 * that we can do IO on it.
+	 */
+	bp->b_bn = ioff;
+	bp->b_count_desired = bp->b_buffer_length;
+
+found:
 	if (!(bp->b_flags & XBF_MAPPED)) {
 		error = _xfs_buf_map_pages(bp, flags);
 		if (unlikely(error)) {
@@ -564,18 +577,10 @@ xfs_buf_get(
 	}
 
 	XFS_STATS_INC(xb_get);
-
-	/*
-	 * Always fill in the block number now, the mapped cases can do
-	 * their own overlay of this later.
-	 */
-	bp->b_bn = ioff;
-	bp->b_count_desired = bp->b_buffer_length;
-
 	trace_xfs_buf_get(bp, flags, _RET_IP_);
 	return bp;
 
- no_buffer:
+no_buffer:
 	if (flags & (XBF_LOCK | XBF_TRYLOCK))
 		xfs_buf_unlock(bp);
 	xfs_buf_rele(bp);
@@ -689,19 +694,6 @@ xfs_buf_read_uncached(
 	return bp;
 }
 
-xfs_buf_t *
-xfs_buf_get_empty(
-	size_t			len,
-	xfs_buftarg_t		*target)
-{
-	xfs_buf_t		*bp;
-
-	bp = xfs_buf_allocate(0);
-	if (bp)
-		_xfs_buf_initialize(bp, target, 0, len, 0);
-	return bp;
-}
-
 /*
  * Return a buffer allocated as an empty buffer and associated to external
  * memory via xfs_buf_associate_memory() back to it's empty state.
@@ -787,10 +779,9 @@ xfs_buf_get_uncached(
 	int			error, i;
 	xfs_buf_t		*bp;
 
-	bp = xfs_buf_allocate(0);
+	bp = xfs_buf_alloc(target, 0, len, 0);
 	if (unlikely(bp == NULL))
 		goto fail;
-	_xfs_buf_initialize(bp, target, 0, len, 0);
 
 	error = _xfs_buf_get_pages(bp, page_count, 0);
 	if (error)
@@ -818,7 +809,7 @@ xfs_buf_get_uncached(
 		__free_page(bp->b_pages[i]);
 	_xfs_buf_free_pages(bp);
  fail_free_buf:
-	xfs_buf_deallocate(bp);
+	kmem_zone_free(xfs_buf_zone, bp);
  fail:
 	return NULL;
 }
@@ -937,12 +928,6 @@ void
 xfs_buf_unlock(
 	struct xfs_buf		*bp)
 {
-	if ((bp->b_flags & (XBF_DELWRI|_XBF_DELWRI_Q)) == XBF_DELWRI) {
-		atomic_inc(&bp->b_hold);
-		bp->b_flags |= XBF_ASYNC;
-		xfs_buf_delwri_queue(bp, 0);
-	}
-
 	XB_CLEAR_OWNER(bp);
 	up(&bp->b_sema);
 
@@ -1019,9 +1004,19 @@ xfs_buf_ioerror(
 	trace_xfs_buf_ioerror(bp, error, _RET_IP_);
 }
 
+void
+xfs_buf_ioerror_alert(
+	struct xfs_buf		*bp,
+	const char		*func)
+{
+	xfs_alert(bp->b_target->bt_mount,
+"metadata I/O error: block 0x%llx (\"%s\") error %d buf count %zd",
+		(__uint64_t)XFS_BUF_ADDR(bp), func,
+		bp->b_error, XFS_BUF_COUNT(bp));
+}
+
 int
 xfs_bwrite(
-	struct xfs_mount	*mp,
 	struct xfs_buf		*bp)
 {
 	int			error;
@@ -1033,25 +1028,13 @@ xfs_bwrite(
 	xfs_bdstrat_cb(bp);
 
 	error = xfs_buf_iowait(bp);
-	if (error)
-		xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR);
-	xfs_buf_relse(bp);
+	if (error) {
+		xfs_force_shutdown(bp->b_target->bt_mount,
+				   SHUTDOWN_META_IO_ERROR);
+	}
 	return error;
 }
 
-void
-xfs_bdwrite(
-	void			*mp,
-	struct xfs_buf		*bp)
-{
-	trace_xfs_buf_bdwrite(bp, _RET_IP_);
-
-	bp->b_flags &= ~XBF_READ;
-	bp->b_flags |= (XBF_DELWRI | XBF_ASYNC);
-
-	xfs_buf_delwri_queue(bp, 1);
-}
-
 /*
  * Called when we want to stop a buffer from getting written or read.
  * We attach the EIO error, muck with its flags, and call xfs_buf_ioend
@@ -1074,9 +1057,8 @@ xfs_bioerror(
 	 * We're calling xfs_buf_ioend, so delete XBF_DONE flag.
 	 */
 	XFS_BUF_UNREAD(bp);
-	XFS_BUF_UNDELAYWRITE(bp);
 	XFS_BUF_UNDONE(bp);
-	XFS_BUF_STALE(bp);
+	xfs_buf_stale(bp);
 
 	xfs_buf_ioend(bp, 0);
 
@@ -1103,9 +1085,8 @@ xfs_bioerror_relse(
 	 * change that interface.
 	 */
 	XFS_BUF_UNREAD(bp);
-	XFS_BUF_UNDELAYWRITE(bp);
 	XFS_BUF_DONE(bp);
-	XFS_BUF_STALE(bp);
+	xfs_buf_stale(bp);
 	bp->b_iodone = NULL;
 	if (!(fl & XBF_ASYNC)) {
 		/*
@@ -1115,7 +1096,7 @@ xfs_bioerror_relse(
 		 * ASYNC buffers.
 		 */
 		xfs_buf_ioerror(bp, EIO);
-		XFS_BUF_FINISH_IOWAIT(bp);
+		complete(&bp->b_iowait);
 	} else {
 		xfs_buf_relse(bp);
 	}
@@ -1275,15 +1256,10 @@ xfs_buf_iorequest(
 {
 	trace_xfs_buf_iorequest(bp, _RET_IP_);
 
-	if (bp->b_flags & XBF_DELWRI) {
-		xfs_buf_delwri_queue(bp, 1);
-		return 0;
-	}
+	ASSERT(!(bp->b_flags & XBF_DELWRI));
 
-	if (bp->b_flags & XBF_WRITE) {
+	if (bp->b_flags & XBF_WRITE)
 		xfs_buf_wait_unpin(bp);
-	}
-
 	xfs_buf_hold(bp);
 
 	/* Set the count to 1 initially, this will stop an I/O
@@ -1481,9 +1457,13 @@ xfs_setsize_buftarg_flags(
 	btp->bt_smask = sectorsize - 1;
 
 	if (set_blocksize(btp->bt_bdev, sectorsize)) {
+		char name[BDEVNAME_SIZE];
+
+		bdevname(btp->bt_bdev, name);
+
 		xfs_warn(btp->bt_mount,
 			"Cannot set_blocksize to %u on device %s\n",
-			sectorsize, xfs_buf_target_name(btp));
+			sectorsize, name);
 		return EINVAL;
 	}
 
@@ -1514,12 +1494,12 @@ xfs_setsize_buftarg(
 }
 
 STATIC int
-xfs_alloc_delwrite_queue(
+xfs_alloc_delwri_queue(
 	xfs_buftarg_t		*btp,
 	const char		*fsname)
 {
-	INIT_LIST_HEAD(&btp->bt_delwrite_queue);
-	spin_lock_init(&btp->bt_delwrite_lock);
+	INIT_LIST_HEAD(&btp->bt_delwri_queue);
+	spin_lock_init(&btp->bt_delwri_lock);
 	btp->bt_flags = 0;
 	btp->bt_task = kthread_run(xfsbufd, btp, "xfsbufd/%s", fsname);
 	if (IS_ERR(btp->bt_task))
@@ -1549,7 +1529,7 @@ xfs_alloc_buftarg(
 	spin_lock_init(&btp->bt_lru_lock);
 	if (xfs_setsize_buftarg_early(btp, bdev))
 		goto error;
-	if (xfs_alloc_delwrite_queue(btp, fsname))
+	if (xfs_alloc_delwri_queue(btp, fsname))
 		goto error;
 	btp->bt_shrinker.shrink = xfs_buftarg_shrink;
 	btp->bt_shrinker.seeks = DEFAULT_SEEKS;
@@ -1565,56 +1545,48 @@ error:
 /*
  *	Delayed write buffer handling
  */
-STATIC void
+void
 xfs_buf_delwri_queue(
-	xfs_buf_t		*bp,
-	int			unlock)
+	xfs_buf_t		*bp)
 {
-	struct list_head	*dwq = &bp->b_target->bt_delwrite_queue;
-	spinlock_t		*dwlk = &bp->b_target->bt_delwrite_lock;
+	struct xfs_buftarg	*btp = bp->b_target;
 
 	trace_xfs_buf_delwri_queue(bp, _RET_IP_);
 
-	ASSERT((bp->b_flags&(XBF_DELWRI|XBF_ASYNC)) == (XBF_DELWRI|XBF_ASYNC));
+	ASSERT(!(bp->b_flags & XBF_READ));
 
-	spin_lock(dwlk);
-	/* If already in the queue, dequeue and place at tail */
+	spin_lock(&btp->bt_delwri_lock);
 	if (!list_empty(&bp->b_list)) {
+		/* if already in the queue, move it to the tail */
 		ASSERT(bp->b_flags & _XBF_DELWRI_Q);
-		if (unlock)
-			atomic_dec(&bp->b_hold);
-		list_del(&bp->b_list);
-	}
-
-	if (list_empty(dwq)) {
+		list_move_tail(&bp->b_list, &btp->bt_delwri_queue);
+	} else {
 		/* start xfsbufd as it is about to have something to do */
-		wake_up_process(bp->b_target->bt_task);
-	}
+		if (list_empty(&btp->bt_delwri_queue))
+			wake_up_process(bp->b_target->bt_task);
 
-	bp->b_flags |= _XBF_DELWRI_Q;
-	list_add_tail(&bp->b_list, dwq);
+		atomic_inc(&bp->b_hold);
+		bp->b_flags |= XBF_DELWRI | _XBF_DELWRI_Q | XBF_ASYNC;
+		list_add_tail(&bp->b_list, &btp->bt_delwri_queue);
+	}
 	bp->b_queuetime = jiffies;
-	spin_unlock(dwlk);
-
-	if (unlock)
-		xfs_buf_unlock(bp);
+	spin_unlock(&btp->bt_delwri_lock);
 }
 
 void
 xfs_buf_delwri_dequeue(
 	xfs_buf_t		*bp)
 {
-	spinlock_t		*dwlk = &bp->b_target->bt_delwrite_lock;
 	int			dequeued = 0;
 
-	spin_lock(dwlk);
+	spin_lock(&bp->b_target->bt_delwri_lock);
 	if ((bp->b_flags & XBF_DELWRI) && !list_empty(&bp->b_list)) {
 		ASSERT(bp->b_flags & _XBF_DELWRI_Q);
 		list_del_init(&bp->b_list);
 		dequeued = 1;
 	}
 	bp->b_flags &= ~(XBF_DELWRI|_XBF_DELWRI_Q);
-	spin_unlock(dwlk);
+	spin_unlock(&bp->b_target->bt_delwri_lock);
 
 	if (dequeued)
 		xfs_buf_rele(bp);
@@ -1646,16 +1618,9 @@ xfs_buf_delwri_promote(
 	if (bp->b_queuetime < jiffies - age)
 		return;
 	bp->b_queuetime = jiffies - age;
-	spin_lock(&btp->bt_delwrite_lock);
-	list_move(&bp->b_list, &btp->bt_delwrite_queue);
-	spin_unlock(&btp->bt_delwrite_lock);
-}
-
-STATIC void
-xfs_buf_runall_queues(
-	struct workqueue_struct	*queue)
-{
-	flush_workqueue(queue);
+	spin_lock(&btp->bt_delwri_lock);
+	list_move(&bp->b_list, &btp->bt_delwri_queue);
+	spin_unlock(&btp->bt_delwri_lock);
 }
 
 /*
@@ -1669,15 +1634,13 @@ xfs_buf_delwri_split(
 	unsigned long	age)
 {
 	xfs_buf_t	*bp, *n;
-	struct list_head *dwq = &target->bt_delwrite_queue;
-	spinlock_t	*dwlk = &target->bt_delwrite_lock;
 	int		skipped = 0;
 	int		force;
 
 	force = test_and_clear_bit(XBT_FORCE_FLUSH, &target->bt_flags);
 	INIT_LIST_HEAD(list);
-	spin_lock(dwlk);
-	list_for_each_entry_safe(bp, n, dwq, b_list) {
+	spin_lock(&target->bt_delwri_lock);
+	list_for_each_entry_safe(bp, n, &target->bt_delwri_queue, b_list) {
 		ASSERT(bp->b_flags & XBF_DELWRI);
 
 		if (!xfs_buf_ispinned(bp) && xfs_buf_trylock(bp)) {
@@ -1694,10 +1657,9 @@ xfs_buf_delwri_split(
 		} else
 			skipped++;
 	}
-	spin_unlock(dwlk);
 
+	spin_unlock(&target->bt_delwri_lock);
 	return skipped;
-
 }
 
 /*
@@ -1747,7 +1709,7 @@ xfsbufd(
 		}
 
 		/* sleep for a long time if there is nothing to do. */
-		if (list_empty(&target->bt_delwrite_queue))
+		if (list_empty(&target->bt_delwri_queue))
 			tout = MAX_SCHEDULE_TIMEOUT;
 		schedule_timeout_interruptible(tout);
 
@@ -1783,9 +1745,7 @@ xfs_flush_buftarg(
 	LIST_HEAD(wait_list);
 	struct blk_plug plug;
 
-	xfs_buf_runall_queues(xfsconvertd_workqueue);
-	xfs_buf_runall_queues(xfsdatad_workqueue);
-	xfs_buf_runall_queues(xfslogd_workqueue);
+	flush_workqueue(xfslogd_workqueue);
 
 	set_bit(XBT_FORCE_FLUSH, &target->bt_flags);
 	pincount = xfs_buf_delwri_split(target, &tmp_list, 0);
@@ -1866,11 +1826,3 @@ xfs_buf_terminate(void)
 	destroy_workqueue(xfslogd_workqueue);
 	kmem_zone_destroy(xfs_buf_zone);
 }
-
-#ifdef CONFIG_KDB_MODULES
-struct list_head *
-xfs_get_buftarg_list(void)
-{
-	return &xfs_buftarg_list;
-}
-#endif

fs/xfs/xfs_buf.h

@@ -105,8 +105,8 @@ typedef struct xfs_buftarg {
 
 	/* per device delwri queue */
 	struct task_struct	*bt_task;
-	struct list_head	bt_delwrite_queue;
-	spinlock_t		bt_delwrite_lock;
+	struct list_head	bt_delwri_queue;
+	spinlock_t		bt_delwri_lock;
 	unsigned long		bt_flags;
 
 	/* LRU control structures */
@@ -175,7 +175,8 @@ extern xfs_buf_t *xfs_buf_get(xfs_buftarg_t *, xfs_off_t, size_t,
 extern xfs_buf_t *xfs_buf_read(xfs_buftarg_t *, xfs_off_t, size_t,
 				xfs_buf_flags_t);
 
-extern xfs_buf_t *xfs_buf_get_empty(size_t, xfs_buftarg_t *);
+struct xfs_buf *xfs_buf_alloc(struct xfs_buftarg *, xfs_off_t, size_t,
+			      xfs_buf_flags_t);
 extern void xfs_buf_set_empty(struct xfs_buf *bp, size_t len);
 extern xfs_buf_t *xfs_buf_get_uncached(struct xfs_buftarg *, size_t, int);
 extern int xfs_buf_associate_memory(xfs_buf_t *, void *, size_t);
@@ -197,14 +198,14 @@ extern void xfs_buf_unlock(xfs_buf_t *);
 	((bp)->b_sema.count <= 0)
 
 /* Buffer Read and Write Routines */
-extern int xfs_bwrite(struct xfs_mount *mp, struct xfs_buf *bp);
-extern void xfs_bdwrite(void *mp, xfs_buf_t *bp);
+extern int xfs_bwrite(struct xfs_buf *bp);
 
 extern void xfsbdstrat(struct xfs_mount *, struct xfs_buf *);
 extern int xfs_bdstrat_cb(struct xfs_buf *);
 
 extern void xfs_buf_ioend(xfs_buf_t *,	int);
 extern void xfs_buf_ioerror(xfs_buf_t *, int);
+extern void xfs_buf_ioerror_alert(struct xfs_buf *, const char *func);
 extern int xfs_buf_iorequest(xfs_buf_t *);
 extern int xfs_buf_iowait(xfs_buf_t *);
 extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, void *,
@@ -221,38 +222,22 @@ static inline int xfs_buf_geterror(xfs_buf_t *bp)
 extern xfs_caddr_t xfs_buf_offset(xfs_buf_t *, size_t);
 
 /* Delayed Write Buffer Routines */
-extern void xfs_buf_delwri_dequeue(xfs_buf_t *);
-extern void xfs_buf_delwri_promote(xfs_buf_t *);
+extern void xfs_buf_delwri_queue(struct xfs_buf *);
+extern void xfs_buf_delwri_dequeue(struct xfs_buf *);
+extern void xfs_buf_delwri_promote(struct xfs_buf *);
 
 /* Buffer Daemon Setup Routines */
 extern int xfs_buf_init(void);
 extern void xfs_buf_terminate(void);
 
-static inline const char *
-xfs_buf_target_name(struct xfs_buftarg *target)
-{
-	static char __b[BDEVNAME_SIZE];
-
-	return bdevname(target->bt_bdev, __b);
-}
-
-
 #define XFS_BUF_ZEROFLAGS(bp) \
 	((bp)->b_flags &= ~(XBF_READ|XBF_WRITE|XBF_ASYNC|XBF_DELWRI| \
 			    XBF_SYNCIO|XBF_FUA|XBF_FLUSH))
 
 void xfs_buf_stale(struct xfs_buf *bp);
-#define XFS_BUF_STALE(bp)	xfs_buf_stale(bp);
 #define XFS_BUF_UNSTALE(bp)	((bp)->b_flags &= ~XBF_STALE)
 #define XFS_BUF_ISSTALE(bp)	((bp)->b_flags & XBF_STALE)
-#define XFS_BUF_SUPER_STALE(bp)	do {				\
-					XFS_BUF_STALE(bp);	\
-					xfs_buf_delwri_dequeue(bp);	\
-					XFS_BUF_DONE(bp);	\
-				} while (0)
-
-#define XFS_BUF_DELAYWRITE(bp)		((bp)->b_flags |= XBF_DELWRI)
-#define XFS_BUF_UNDELAYWRITE(bp)	xfs_buf_delwri_dequeue(bp)
+
 #define XFS_BUF_ISDELAYWRITE(bp)	((bp)->b_flags & XBF_DELWRI)
 
 #define XFS_BUF_DONE(bp)	((bp)->b_flags |= XBF_DONE)
@@ -280,23 +265,16 @@ void xfs_buf_stale(struct xfs_buf *bp);
 #define XFS_BUF_SIZE(bp)		((bp)->b_buffer_length)
 #define XFS_BUF_SET_SIZE(bp, cnt)	((bp)->b_buffer_length = (cnt))
 
-static inline void
-xfs_buf_set_ref(
-	struct xfs_buf	*bp,
-	int		lru_ref)
+static inline void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref)
 {
 	atomic_set(&bp->b_lru_ref, lru_ref);
 }
-#define XFS_BUF_SET_VTYPE_REF(bp, type, ref)	xfs_buf_set_ref(bp, ref)
-#define XFS_BUF_SET_VTYPE(bp, type)		do { } while (0)
 
 static inline int xfs_buf_ispinned(struct xfs_buf *bp)
 {
 	return atomic_read(&bp->b_pin_count);
 }
 
-#define XFS_BUF_FINISH_IOWAIT(bp)	complete(&bp->b_iowait);
-
 static inline void xfs_buf_relse(xfs_buf_t *bp)
 {
 	xfs_buf_unlock(bp);
@@ -313,14 +291,7 @@ extern void xfs_wait_buftarg(xfs_buftarg_t *);
 extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int, unsigned int);
 extern int xfs_flush_buftarg(xfs_buftarg_t *, int);
 
-#ifdef CONFIG_KDB_MODULES
-extern struct list_head *xfs_get_buftarg_list(void);
-#endif
-
 #define xfs_getsize_buftarg(buftarg)	block_size((buftarg)->bt_bdev)
 #define xfs_readonly_buftarg(buftarg)	bdev_read_only((buftarg)->bt_bdev)
 
-#define xfs_binval(buftarg)		xfs_flush_buftarg(buftarg, 1)
-#define XFS_bflush(buftarg)		xfs_flush_buftarg(buftarg, 1)
-
 #endif	/* __XFS_BUF_H__ */

fs/xfs/xfs_buf_item.c

@@ -967,7 +967,8 @@ xfs_buf_iodone_callbacks(
 	 * I/O errors, there's no point in giving this a retry.
 	 */
 	if (XFS_FORCED_SHUTDOWN(mp)) {
-		XFS_BUF_SUPER_STALE(bp);
+		xfs_buf_stale(bp);
+		XFS_BUF_DONE(bp);
 		trace_xfs_buf_item_iodone(bp, _RET_IP_);
 		goto do_callbacks;
 	}
@@ -975,9 +976,7 @@ xfs_buf_iodone_callbacks(
 	if (bp->b_target != lasttarg ||
 	    time_after(jiffies, (lasttime + 5*HZ))) {
 		lasttime = jiffies;
-		xfs_alert(mp, "Device %s: metadata write error block 0x%llx",
-			xfs_buf_target_name(bp->b_target),
-		      (__uint64_t)XFS_BUF_ADDR(bp));
+		xfs_buf_ioerror_alert(bp, __func__);
 	}
 	lasttarg = bp->b_target;
 
@@ -993,7 +992,7 @@ xfs_buf_iodone_callbacks(
 		xfs_buf_ioerror(bp, 0); /* errno of 0 unsets the flag */
 
 		if (!XFS_BUF_ISSTALE(bp)) {
-			XFS_BUF_DELAYWRITE(bp);
+			xfs_buf_delwri_queue(bp);
 			XFS_BUF_DONE(bp);
 		}
 		ASSERT(bp->b_iodone != NULL);
@@ -1006,9 +1005,8 @@ xfs_buf_iodone_callbacks(
 	 * If the write of the buffer was synchronous, we want to make
 	 * sure to return the error to the caller of xfs_bwrite().
 	 */
-	XFS_BUF_STALE(bp);
+	xfs_buf_stale(bp);
 	XFS_BUF_DONE(bp);
-	XFS_BUF_UNDELAYWRITE(bp);
 
 	trace_xfs_buf_error_relse(bp, _RET_IP_);
 

fs/xfs/xfs_da_btree.c

@@ -1578,9 +1578,8 @@ xfs_da_grow_inode_int(
 	 */
 	nmap = 1;
 	ASSERT(args->firstblock != NULL);
-	error = xfs_bmapi(tp, dp, *bno, count,
-			xfs_bmapi_aflag(w)|XFS_BMAPI_WRITE|XFS_BMAPI_METADATA|
-			XFS_BMAPI_CONTIG,
+	error = xfs_bmapi_write(tp, dp, *bno, count,
+			xfs_bmapi_aflag(w)|XFS_BMAPI_METADATA|XFS_BMAPI_CONTIG,
 			args->firstblock, args->total, &map, &nmap,
 			args->flist);
 	if (error)
@@ -1602,9 +1601,8 @@ xfs_da_grow_inode_int(
 		for (b = *bno, mapi = 0; b < *bno + count; ) {
 			nmap = MIN(XFS_BMAP_MAX_NMAP, count);
 			c = (int)(*bno + count - b);
-			error = xfs_bmapi(tp, dp, b, c,
-					xfs_bmapi_aflag(w)|XFS_BMAPI_WRITE|
-					XFS_BMAPI_METADATA,
+			error = xfs_bmapi_write(tp, dp, b, c,
+					xfs_bmapi_aflag(w)|XFS_BMAPI_METADATA,
 					args->firstblock, args->total,
 					&mapp[mapi], &nmap, args->flist);
 			if (error)
@@ -1975,33 +1973,16 @@ xfs_da_do_buf(
 		/*
 		 * Optimize the one-block case.
 		 */
-		if (nfsb == 1) {
-			xfs_fsblock_t	fsb;
-
-			if ((error =
-			    xfs_bmapi_single(trans, dp, whichfork, &fsb,
-				    (xfs_fileoff_t)bno))) {
-				return error;
-			}
+		if (nfsb == 1)
 			mapp = &map;
-			if (fsb == NULLFSBLOCK) {
-				nmap = 0;
-			} else {
-				map.br_startblock = fsb;
-				map.br_startoff = (xfs_fileoff_t)bno;
-				map.br_blockcount = 1;
-				nmap = 1;
-			}
-		} else {
+		else
 			mapp = kmem_alloc(sizeof(*mapp) * nfsb, KM_SLEEP);
-			nmap = nfsb;
-			if ((error = xfs_bmapi(trans, dp, (xfs_fileoff_t)bno,
-					nfsb,
-					XFS_BMAPI_METADATA |
-						xfs_bmapi_aflag(whichfork),
-					NULL, 0, mapp, &nmap, NULL)))
-				goto exit0;
-		}
+
+		nmap = nfsb;
+		error = xfs_bmapi_read(dp, (xfs_fileoff_t)bno, nfsb, mapp,
+				       &nmap, xfs_bmapi_aflag(whichfork));
+		if (error)
+			goto exit0;
 	} else {
 		map.br_startblock = XFS_DADDR_TO_FSB(mp, mappedbno);
 		map.br_startoff = (xfs_fileoff_t)bno;
@@ -2072,13 +2053,10 @@ xfs_da_do_buf(
 		if (!bp)
 			continue;
 		if (caller == 1) {
-			if (whichfork == XFS_ATTR_FORK) {
-				XFS_BUF_SET_VTYPE_REF(bp, B_FS_ATTR_BTREE,
-						XFS_ATTR_BTREE_REF);
-			} else {
-				XFS_BUF_SET_VTYPE_REF(bp, B_FS_DIR_BTREE,
-						XFS_DIR_BTREE_REF);
-			}
+			if (whichfork == XFS_ATTR_FORK)
+				xfs_buf_set_ref(bp, XFS_ATTR_BTREE_REF);
+			else
+				xfs_buf_set_ref(bp, XFS_DIR_BTREE_REF);
 		}
 		if (bplist) {
 			bplist[nbplist++] = bp;

fs/xfs/xfs_dfrag.c

@@ -425,8 +425,8 @@ xfs_swap_extents(
 	}
 
 
-	xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
-	xfs_trans_ijoin_ref(tp, tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
+	xfs_trans_ijoin(tp, tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
 
 	xfs_trans_log_inode(tp, ip,  ilf_fields);
 	xfs_trans_log_inode(tp, tip, tilf_fields);
@@ -438,7 +438,7 @@ xfs_swap_extents(
 	if (mp->m_flags & XFS_MOUNT_WSYNC)
 		xfs_trans_set_sync(tp);
 
-	error = xfs_trans_commit(tp, XFS_TRANS_SWAPEXT);
+	error = xfs_trans_commit(tp, 0);
 
 	trace_xfs_swap_extent_after(ip, 0);
 	trace_xfs_swap_extent_after(tip, 1);

fs/xfs/xfs_dir2_leaf.c

@@ -888,12 +888,10 @@ xfs_dir2_leaf_getdents(
 				 * we already have in the table.
 				 */
 				nmap = map_size - map_valid;
-				error = xfs_bmapi(NULL, dp,
-					map_off,
+				error = xfs_bmapi_read(dp, map_off,
 					xfs_dir2_byte_to_da(mp,
 						XFS_DIR2_LEAF_OFFSET) - map_off,
-					XFS_BMAPI_METADATA, NULL, 0,
-					&map[map_valid], &nmap, NULL);
+					&map[map_valid], &nmap, 0);
 				/*
 				 * Don't know if we should ignore this or
 				 * try to return an error.

fs/xfs/xfs_discard.c

@@ -38,7 +38,7 @@ xfs_trim_extents(
 	struct xfs_mount	*mp,
 	xfs_agnumber_t		agno,
 	xfs_fsblock_t		start,
-	xfs_fsblock_t		len,
+	xfs_fsblock_t		end,
 	xfs_fsblock_t		minlen,
 	__uint64_t		*blocks_trimmed)
 {
@@ -100,7 +100,7 @@ xfs_trim_extents(
 		 * down partially overlapping ranges for now.
 		 */
 		if (XFS_AGB_TO_FSB(mp, agno, fbno) + flen < start ||
-		    XFS_AGB_TO_FSB(mp, agno, fbno) >= start + len) {
+		    XFS_AGB_TO_FSB(mp, agno, fbno) > end) {
 			trace_xfs_discard_exclude(mp, agno, fbno, flen);
 			goto next_extent;
 		}
@@ -145,7 +145,7 @@ xfs_ioc_trim(
 	struct request_queue	*q = mp->m_ddev_targp->bt_bdev->bd_disk->queue;
 	unsigned int		granularity = q->limits.discard_granularity;
 	struct fstrim_range	range;
-	xfs_fsblock_t		start, len, minlen;
+	xfs_fsblock_t		start, end, minlen;
 	xfs_agnumber_t		start_agno, end_agno, agno;
 	__uint64_t		blocks_trimmed = 0;
 	int			error, last_error = 0;
@@ -165,19 +165,19 @@ xfs_ioc_trim(
 	 * matter as trimming blocks is an advisory interface.
 	 */
 	start = XFS_B_TO_FSBT(mp, range.start);
-	len = XFS_B_TO_FSBT(mp, range.len);
+	end = start + XFS_B_TO_FSBT(mp, range.len) - 1;
 	minlen = XFS_B_TO_FSB(mp, max_t(u64, granularity, range.minlen));
 
-	start_agno = XFS_FSB_TO_AGNO(mp, start);
-	if (start_agno >= mp->m_sb.sb_agcount)
+	if (start >= mp->m_sb.sb_dblocks)
 		return -XFS_ERROR(EINVAL);
+	if (end > mp->m_sb.sb_dblocks - 1)
+		end = mp->m_sb.sb_dblocks - 1;
 
-	end_agno = XFS_FSB_TO_AGNO(mp, start + len);
-	if (end_agno >= mp->m_sb.sb_agcount)
-		end_agno = mp->m_sb.sb_agcount - 1;
+	start_agno = XFS_FSB_TO_AGNO(mp, start);
+	end_agno = XFS_FSB_TO_AGNO(mp, end);
 
 	for (agno = start_agno; agno <= end_agno; agno++) {
-		error = -xfs_trim_extents(mp, agno, start, len, minlen,
+		error = -xfs_trim_extents(mp, agno, start, end, minlen,
 					  &blocks_trimmed);
 		if (error)
 			last_error = error;

fs/xfs/xfs_dquot.c

@@ -377,16 +377,14 @@ xfs_qm_dqalloc(
 		return (ESRCH);
 	}
 
-	xfs_trans_ijoin_ref(tp, quotip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, quotip, XFS_ILOCK_EXCL);
 	nmaps = 1;
-	if ((error = xfs_bmapi(tp, quotip,
-			      offset_fsb, XFS_DQUOT_CLUSTER_SIZE_FSB,
-			      XFS_BMAPI_METADATA | XFS_BMAPI_WRITE,
-			      &firstblock,
-			      XFS_QM_DQALLOC_SPACE_RES(mp),
-			      &map, &nmaps, &flist))) {
+	error = xfs_bmapi_write(tp, quotip, offset_fsb,
+				XFS_DQUOT_CLUSTER_SIZE_FSB, XFS_BMAPI_METADATA,
+				&firstblock, XFS_QM_DQALLOC_SPACE_RES(mp),
+				&map, &nmaps, &flist);
+	if (error)
 		goto error0;
-	}
 	ASSERT(map.br_blockcount == XFS_DQUOT_CLUSTER_SIZE_FSB);
 	ASSERT(nmaps == 1);
 	ASSERT((map.br_startblock != DELAYSTARTBLOCK) &&
@@ -402,8 +400,11 @@ xfs_qm_dqalloc(
 			       dqp->q_blkno,
 			       mp->m_quotainfo->qi_dqchunklen,
 			       0);
-	if (!bp || (error = xfs_buf_geterror(bp)))
+
+	error = xfs_buf_geterror(bp);
+	if (error)
 		goto error1;
+
 	/*
 	 * Make a chunk of dquots out of this buffer and log
 	 * the entire thing.
@@ -485,9 +486,8 @@ xfs_qm_dqtobp(
 	/*
 	 * Find the block map; no allocations yet
 	 */
-	error = xfs_bmapi(NULL, quotip, dqp->q_fileoffset,
-			  XFS_DQUOT_CLUSTER_SIZE_FSB, XFS_BMAPI_METADATA,
-			  NULL, 0, &map, &nmaps, NULL);
+	error = xfs_bmapi_read(quotip, dqp->q_fileoffset,
+			       XFS_DQUOT_CLUSTER_SIZE_FSB, &map, &nmaps, 0);
 
 	xfs_iunlock(quotip, XFS_ILOCK_SHARED);
 	if (error)
@@ -605,7 +605,7 @@ xfs_qm_dqread(
 	dqp->q_res_rtbcount = be64_to_cpu(ddqp->d_rtbcount);
 
 	/* Mark the buf so that this will stay incore a little longer */
-	XFS_BUF_SET_VTYPE_REF(bp, B_FS_DQUOT, XFS_DQUOT_REF);
+	xfs_buf_set_ref(bp, XFS_DQUOT_REF);
 
 	/*
 	 * We got the buffer with a xfs_trans_read_buf() (in dqtobp())
@@ -1242,9 +1242,11 @@ xfs_qm_dqflush(
 	}
 
 	if (flags & SYNC_WAIT)
-		error = xfs_bwrite(mp, bp);
+		error = xfs_bwrite(bp);
 	else
-		xfs_bdwrite(mp, bp);
+		xfs_buf_delwri_queue(bp);
+
+	xfs_buf_relse(bp);
 
 	trace_xfs_dqflush_done(dqp);
 

fs/xfs/xfs_export.c

@@ -229,16 +229,16 @@ xfs_fs_nfs_commit_metadata(
 {
 	struct xfs_inode	*ip = XFS_I(inode);
 	struct xfs_mount	*mp = ip->i_mount;
-	int			error = 0;
+	xfs_lsn_t		lsn = 0;
 
 	xfs_ilock(ip, XFS_ILOCK_SHARED);
-	if (xfs_ipincount(ip)) {
-		error = _xfs_log_force_lsn(mp, ip->i_itemp->ili_last_lsn,
-				XFS_LOG_SYNC, NULL);
-	}
+	if (xfs_ipincount(ip))
+		lsn = ip->i_itemp->ili_last_lsn;
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
 
-	return error;
+	if (!lsn)
+		return 0;
+	return _xfs_log_force_lsn(mp, lsn, XFS_LOG_SYNC, NULL);
 }
 
 const struct export_operations xfs_export_operations = {

fs/xfs/xfs_file.c

@@ -124,6 +124,35 @@ xfs_iozero(
 	return (-status);
 }
 
+/*
+ * Fsync operations on directories are much simpler than on regular files,
+ * as there is no file data to flush, and thus also no need for explicit
+ * cache flush operations, and there are no non-transaction metadata updates
+ * on directories either.
+ */
+STATIC int
+xfs_dir_fsync(
+	struct file		*file,
+	loff_t			start,
+	loff_t			end,
+	int			datasync)
+{
+	struct xfs_inode	*ip = XFS_I(file->f_mapping->host);
+	struct xfs_mount	*mp = ip->i_mount;
+	xfs_lsn_t		lsn = 0;
+
+	trace_xfs_dir_fsync(ip);
+
+	xfs_ilock(ip, XFS_ILOCK_SHARED);
+	if (xfs_ipincount(ip))
+		lsn = ip->i_itemp->ili_last_lsn;
+	xfs_iunlock(ip, XFS_ILOCK_SHARED);
+
+	if (!lsn)
+		return 0;
+	return _xfs_log_force_lsn(mp, lsn, XFS_LOG_SYNC, NULL);
+}
+
 STATIC int
 xfs_file_fsync(
 	struct file		*file,
@@ -137,6 +166,7 @@ xfs_file_fsync(
 	struct xfs_trans	*tp;
 	int			error = 0;
 	int			log_flushed = 0;
+	xfs_lsn_t		lsn = 0;
 
 	trace_xfs_file_fsync(ip);
 
@@ -149,10 +179,6 @@ xfs_file_fsync(
 
 	xfs_iflags_clear(ip, XFS_ITRUNCATED);
 
-	xfs_ilock(ip, XFS_IOLOCK_SHARED);
-	xfs_ioend_wait(ip);
-	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
-
 	if (mp->m_flags & XFS_MOUNT_BARRIER) {
 		/*
 		 * If we have an RT and/or log subvolume we need to make sure
@@ -216,11 +242,11 @@ xfs_file_fsync(
 		 * transaction.	 So we play it safe and fire off the
 		 * transaction anyway.
 		 */
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-		xfs_trans_set_sync(tp);
-		error = _xfs_trans_commit(tp, 0, &log_flushed);
+		error = xfs_trans_commit(tp, 0);
 
+		lsn = ip->i_itemp->ili_last_lsn;
 		xfs_iunlock(ip, XFS_ILOCK_EXCL);
 	} else {
 		/*
@@ -231,14 +257,14 @@ xfs_file_fsync(
 		 * disk yet, the inode will be still be pinned.  If it is,
 		 * force the log.
 		 */
-		if (xfs_ipincount(ip)) {
-			error = _xfs_log_force_lsn(mp,
-					ip->i_itemp->ili_last_lsn,
-					XFS_LOG_SYNC, &log_flushed);
-		}
+		if (xfs_ipincount(ip))
+			lsn = ip->i_itemp->ili_last_lsn;
 		xfs_iunlock(ip, XFS_ILOCK_SHARED);
 	}
 
+	if (!error && lsn)
+		error = _xfs_log_force_lsn(mp, lsn, XFS_LOG_SYNC, &log_flushed);
+
 	/*
 	 * If we only have a single device, and the log force about was
 	 * a no-op we might have to flush the data device cache here.
@@ -317,7 +343,19 @@ xfs_file_aio_read(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -EIO;
 
-	if (unlikely(ioflags & IO_ISDIRECT)) {
+	/*
+	 * Locking is a bit tricky here. If we take an exclusive lock
+	 * for direct IO, we effectively serialise all new concurrent
+	 * read IO to this file and block it behind IO that is currently in
+	 * progress because IO in progress holds the IO lock shared. We only
+	 * need to hold the lock exclusive to blow away the page cache, so
+	 * only take lock exclusively if the page cache needs invalidation.
+	 * This allows the normal direct IO case of no page cache pages to
+	 * proceeed concurrently without serialisation.
+	 */
+	xfs_rw_ilock(ip, XFS_IOLOCK_SHARED);
+	if ((ioflags & IO_ISDIRECT) && inode->i_mapping->nrpages) {
+		xfs_rw_iunlock(ip, XFS_IOLOCK_SHARED);
 		xfs_rw_ilock(ip, XFS_IOLOCK_EXCL);
 
 		if (inode->i_mapping->nrpages) {
@@ -330,8 +368,7 @@ xfs_file_aio_read(
 			}
 		}
 		xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL);
-	} else
-		xfs_rw_ilock(ip, XFS_IOLOCK_SHARED);
+	}
 
 	trace_xfs_file_read(ip, size, iocb->ki_pos, ioflags);
 
@@ -407,11 +444,13 @@ xfs_aio_write_isize_update(
  */
 STATIC void
 xfs_aio_write_newsize_update(
-	struct xfs_inode	*ip)
+	struct xfs_inode	*ip,
+	xfs_fsize_t		new_size)
 {
-	if (ip->i_new_size) {
+	if (new_size == ip->i_new_size) {
 		xfs_rw_ilock(ip, XFS_ILOCK_EXCL);
-		ip->i_new_size = 0;
+		if (new_size == ip->i_new_size)
+			ip->i_new_size = 0;
 		if (ip->i_d.di_size > ip->i_size)
 			ip->i_d.di_size = ip->i_size;
 		xfs_rw_iunlock(ip, XFS_ILOCK_EXCL);
@@ -462,7 +501,7 @@ xfs_file_splice_write(
 	ret = generic_file_splice_write(pipe, outfilp, ppos, count, flags);
 
 	xfs_aio_write_isize_update(inode, ppos, ret);
-	xfs_aio_write_newsize_update(ip);
+	xfs_aio_write_newsize_update(ip, new_size);
 	xfs_iunlock(ip, XFS_IOLOCK_EXCL);
 	return ret;
 }
@@ -500,11 +539,9 @@ xfs_zero_last_block(
 
 	last_fsb = XFS_B_TO_FSBT(mp, isize);
 	nimaps = 1;
-	error = xfs_bmapi(NULL, ip, last_fsb, 1, 0, NULL, 0, &imap,
-			  &nimaps, NULL);
-	if (error) {
+	error = xfs_bmapi_read(ip, last_fsb, 1, &imap, &nimaps, 0);
+	if (error)
 		return error;
-	}
 	ASSERT(nimaps > 0);
 	/*
 	 * If the block underlying isize is just a hole, then there
@@ -595,8 +632,8 @@ xfs_zero_eof(
 	while (start_zero_fsb <= end_zero_fsb) {
 		nimaps = 1;
 		zero_count_fsb = end_zero_fsb - start_zero_fsb + 1;
-		error = xfs_bmapi(NULL, ip, start_zero_fsb, zero_count_fsb,
-				  0, NULL, 0, &imap, &nimaps, NULL);
+		error = xfs_bmapi_read(ip, start_zero_fsb, zero_count_fsb,
+					  &imap, &nimaps, 0);
 		if (error) {
 			ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL));
 			return error;
@@ -659,6 +696,7 @@ xfs_file_aio_write_checks(
 	struct file		*file,
 	loff_t			*pos,
 	size_t			*count,
+	xfs_fsize_t		*new_sizep,
 	int			*iolock)
 {
 	struct inode		*inode = file->f_mapping->host;
@@ -666,6 +704,9 @@ xfs_file_aio_write_checks(
 	xfs_fsize_t		new_size;
 	int			error = 0;
 
+	xfs_rw_ilock(ip, XFS_ILOCK_EXCL);
+	*new_sizep = 0;
+restart:
 	error = generic_write_checks(file, pos, count, S_ISBLK(inode->i_mode));
 	if (error) {
 		xfs_rw_iunlock(ip, XFS_ILOCK_EXCL | *iolock);
@@ -673,20 +714,41 @@ xfs_file_aio_write_checks(
 		return error;
 	}
 
-	new_size = *pos + *count;
-	if (new_size > ip->i_size)
-		ip->i_new_size = new_size;
-
 	if (likely(!(file->f_mode & FMODE_NOCMTIME)))
 		file_update_time(file);
 
 	/*
 	 * If the offset is beyond the size of the file, we need to zero any
 	 * blocks that fall between the existing EOF and the start of this
-	 * write.
+	 * write. There is no need to issue zeroing if another in-flght IO ends
+	 * at or before this one If zeronig is needed and we are currently
+	 * holding the iolock shared, we need to update it to exclusive which
+	 * involves dropping all locks and relocking to maintain correct locking
+	 * order. If we do this, restart the function to ensure all checks and
+	 * values are still valid.
 	 */
-	if (*pos > ip->i_size)
+	if ((ip->i_new_size && *pos > ip->i_new_size) ||
+	    (!ip->i_new_size && *pos > ip->i_size)) {
+		if (*iolock == XFS_IOLOCK_SHARED) {
+			xfs_rw_iunlock(ip, XFS_ILOCK_EXCL | *iolock);
+			*iolock = XFS_IOLOCK_EXCL;
+			xfs_rw_ilock(ip, XFS_ILOCK_EXCL | *iolock);
+			goto restart;
+		}
 		error = -xfs_zero_eof(ip, *pos, ip->i_size);
+	}
+
+	/*
+	 * If this IO extends beyond EOF, we may need to update ip->i_new_size.
+	 * We have already zeroed space beyond EOF (if necessary).  Only update
+	 * ip->i_new_size if this IO ends beyond any other in-flight writes.
+	 */
+	new_size = *pos + *count;
+	if (new_size > ip->i_size) {
+		if (new_size > ip->i_new_size)
+			ip->i_new_size = new_size;
+		*new_sizep = new_size;
+	}
 
 	xfs_rw_iunlock(ip, XFS_ILOCK_EXCL);
 	if (error)
@@ -721,7 +783,7 @@ xfs_file_aio_write_checks(
  * the dio layer.  To avoid the problem with aio, we also need to wait for
  * outstanding IOs to complete so that unwritten extent conversion is completed
  * before we try to map the overlapping block. This is currently implemented by
- * hitting it with a big hammer (i.e. xfs_ioend_wait()).
+ * hitting it with a big hammer (i.e. inode_dio_wait()).
  *
  * Returns with locks held indicated by @iolock and errors indicated by
  * negative return values.
@@ -733,6 +795,7 @@ xfs_file_dio_aio_write(
 	unsigned long		nr_segs,
 	loff_t			pos,
 	size_t			ocount,
+	xfs_fsize_t		*new_size,
 	int			*iolock)
 {
 	struct file		*file = iocb->ki_filp;
@@ -753,18 +816,35 @@ xfs_file_dio_aio_write(
 	if ((pos & mp->m_blockmask) || ((pos + count) & mp->m_blockmask))
 		unaligned_io = 1;
 
-	if (unaligned_io || mapping->nrpages || pos > ip->i_size)
+	/*
+	 * We don't need to take an exclusive lock unless there page cache needs
+	 * to be invalidated or unaligned IO is being executed. We don't need to
+	 * consider the EOF extension case here because
+	 * xfs_file_aio_write_checks() will relock the inode as necessary for
+	 * EOF zeroing cases and fill out the new inode size as appropriate.
+	 */
+	if (unaligned_io || mapping->nrpages)
 		*iolock = XFS_IOLOCK_EXCL;
 	else
 		*iolock = XFS_IOLOCK_SHARED;
-	xfs_rw_ilock(ip, XFS_ILOCK_EXCL | *iolock);
+	xfs_rw_ilock(ip, *iolock);
 
-	ret = xfs_file_aio_write_checks(file, &pos, &count, iolock);
+	/*
+	 * Recheck if there are cached pages that need invalidate after we got
+	 * the iolock to protect against other threads adding new pages while
+	 * we were waiting for the iolock.
+	 */
+	if (mapping->nrpages && *iolock == XFS_IOLOCK_SHARED) {
+		xfs_rw_iunlock(ip, *iolock);
+		*iolock = XFS_IOLOCK_EXCL;
+		xfs_rw_ilock(ip, *iolock);
+	}
+
+	ret = xfs_file_aio_write_checks(file, &pos, &count, new_size, iolock);
 	if (ret)
 		return ret;
 
 	if (mapping->nrpages) {
-		WARN_ON(*iolock != XFS_IOLOCK_EXCL);
 		ret = -xfs_flushinval_pages(ip, (pos & PAGE_CACHE_MASK), -1,
 							FI_REMAPF_LOCKED);
 		if (ret)
@@ -776,7 +856,7 @@ xfs_file_dio_aio_write(
 	 * otherwise demote the lock if we had to flush cached pages
 	 */
 	if (unaligned_io)
-		xfs_ioend_wait(ip);
+		inode_dio_wait(inode);
 	else if (*iolock == XFS_IOLOCK_EXCL) {
 		xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL);
 		*iolock = XFS_IOLOCK_SHARED;
@@ -798,6 +878,7 @@ xfs_file_buffered_aio_write(
 	unsigned long		nr_segs,
 	loff_t			pos,
 	size_t			ocount,
+	xfs_fsize_t		*new_size,
 	int			*iolock)
 {
 	struct file		*file = iocb->ki_filp;
@@ -809,9 +890,9 @@ xfs_file_buffered_aio_write(
 	size_t			count = ocount;
 
 	*iolock = XFS_IOLOCK_EXCL;
-	xfs_rw_ilock(ip, XFS_ILOCK_EXCL | *iolock);
+	xfs_rw_ilock(ip, *iolock);
 
-	ret = xfs_file_aio_write_checks(file, &pos, &count, iolock);
+	ret = xfs_file_aio_write_checks(file, &pos, &count, new_size, iolock);
 	if (ret)
 		return ret;
 
@@ -851,6 +932,7 @@ xfs_file_aio_write(
 	ssize_t			ret;
 	int			iolock;
 	size_t			ocount = 0;
+	xfs_fsize_t		new_size = 0;
 
 	XFS_STATS_INC(xs_write_calls);
 
@@ -870,10 +952,10 @@ xfs_file_aio_write(
 
 	if (unlikely(file->f_flags & O_DIRECT))
 		ret = xfs_file_dio_aio_write(iocb, iovp, nr_segs, pos,
-						ocount, &iolock);
+						ocount, &new_size, &iolock);
 	else
 		ret = xfs_file_buffered_aio_write(iocb, iovp, nr_segs, pos,
-						ocount, &iolock);
+						ocount, &new_size, &iolock);
 
 	xfs_aio_write_isize_update(inode, &iocb->ki_pos, ret);
 
@@ -894,7 +976,7 @@ xfs_file_aio_write(
 	}
 
 out_unlock:
-	xfs_aio_write_newsize_update(ip);
+	xfs_aio_write_newsize_update(ip, new_size);
 	xfs_rw_iunlock(ip, iolock);
 	return ret;
 }
@@ -1087,7 +1169,7 @@ const struct file_operations xfs_dir_file_operations = {
 #ifdef CONFIG_COMPAT
 	.compat_ioctl	= xfs_file_compat_ioctl,
 #endif
-	.fsync		= xfs_file_fsync,
+	.fsync		= xfs_dir_fsync,
 };
 
 static const struct vm_operations_struct xfs_file_vm_ops = {

fs/xfs/xfs_filestream.c

@@ -682,7 +682,7 @@ xfs_filestream_new_ag(
 	ip = ap->ip;
 	mp = ip->i_mount;
 	cache = mp->m_filestream;
-	minlen = ap->alen;
+	minlen = ap->length;
 	*agp = NULLAGNUMBER;
 
 	/*
@@ -761,7 +761,7 @@ xfs_filestream_new_ag(
 	 */
 	ag = (ag == NULLAGNUMBER) ? 0 : (ag + 1) % mp->m_sb.sb_agcount;
 	flags = (ap->userdata ? XFS_PICK_USERDATA : 0) |
-	        (ap->low ? XFS_PICK_LOWSPACE : 0);
+	        (ap->flist->xbf_low ? XFS_PICK_LOWSPACE : 0);
 
 	err = _xfs_filestream_pick_ag(mp, ag, agp, flags, minlen);
 	if (err || *agp == NULLAGNUMBER)

fs/xfs/xfs_fsops.c

@@ -194,6 +194,10 @@ xfs_growfs_data_private(
 		bp = xfs_buf_get(mp->m_ddev_targp,
 				 XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)),
 				 XFS_FSS_TO_BB(mp, 1), XBF_LOCK | XBF_MAPPED);
+		if (!bp) {
+			error = ENOMEM;
+			goto error0;
+		}
 		agf = XFS_BUF_TO_AGF(bp);
 		memset(agf, 0, mp->m_sb.sb_sectsize);
 		agf->agf_magicnum = cpu_to_be32(XFS_AGF_MAGIC);
@@ -216,16 +220,21 @@ xfs_growfs_data_private(
 		tmpsize = agsize - XFS_PREALLOC_BLOCKS(mp);
 		agf->agf_freeblks = cpu_to_be32(tmpsize);
 		agf->agf_longest = cpu_to_be32(tmpsize);
-		error = xfs_bwrite(mp, bp);
-		if (error) {
+		error = xfs_bwrite(bp);
+		xfs_buf_relse(bp);
+		if (error)
 			goto error0;
-		}
+
 		/*
 		 * AG inode header block
 		 */
 		bp = xfs_buf_get(mp->m_ddev_targp,
 				 XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)),
 				 XFS_FSS_TO_BB(mp, 1), XBF_LOCK | XBF_MAPPED);
+		if (!bp) {
+			error = ENOMEM;
+			goto error0;
+		}
 		agi = XFS_BUF_TO_AGI(bp);
 		memset(agi, 0, mp->m_sb.sb_sectsize);
 		agi->agi_magicnum = cpu_to_be32(XFS_AGI_MAGIC);
@@ -240,10 +249,11 @@ xfs_growfs_data_private(
 		agi->agi_dirino = cpu_to_be32(NULLAGINO);
 		for (bucket = 0; bucket < XFS_AGI_UNLINKED_BUCKETS; bucket++)
 			agi->agi_unlinked[bucket] = cpu_to_be32(NULLAGINO);
-		error = xfs_bwrite(mp, bp);
-		if (error) {
+		error = xfs_bwrite(bp);
+		xfs_buf_relse(bp);
+		if (error)
 			goto error0;
-		}
+
 		/*
 		 * BNO btree root block
 		 */
@@ -251,6 +261,10 @@ xfs_growfs_data_private(
 				 XFS_AGB_TO_DADDR(mp, agno, XFS_BNO_BLOCK(mp)),
 				 BTOBB(mp->m_sb.sb_blocksize),
 				 XBF_LOCK | XBF_MAPPED);
+		if (!bp) {
+			error = ENOMEM;
+			goto error0;
+		}
 		block = XFS_BUF_TO_BLOCK(bp);
 		memset(block, 0, mp->m_sb.sb_blocksize);
 		block->bb_magic = cpu_to_be32(XFS_ABTB_MAGIC);
@@ -262,10 +276,11 @@ xfs_growfs_data_private(
 		arec->ar_startblock = cpu_to_be32(XFS_PREALLOC_BLOCKS(mp));
 		arec->ar_blockcount = cpu_to_be32(
 			agsize - be32_to_cpu(arec->ar_startblock));
-		error = xfs_bwrite(mp, bp);
-		if (error) {
+		error = xfs_bwrite(bp);
+		xfs_buf_relse(bp);
+		if (error)
 			goto error0;
-		}
+
 		/*
 		 * CNT btree root block
 		 */
@@ -273,6 +288,10 @@ xfs_growfs_data_private(
 				 XFS_AGB_TO_DADDR(mp, agno, XFS_CNT_BLOCK(mp)),
 				 BTOBB(mp->m_sb.sb_blocksize),
 				 XBF_LOCK | XBF_MAPPED);
+		if (!bp) {
+			error = ENOMEM;
+			goto error0;
+		}
 		block = XFS_BUF_TO_BLOCK(bp);
 		memset(block, 0, mp->m_sb.sb_blocksize);
 		block->bb_magic = cpu_to_be32(XFS_ABTC_MAGIC);
@@ -285,10 +304,11 @@ xfs_growfs_data_private(
 		arec->ar_blockcount = cpu_to_be32(
 			agsize - be32_to_cpu(arec->ar_startblock));
 		nfree += be32_to_cpu(arec->ar_blockcount);
-		error = xfs_bwrite(mp, bp);
-		if (error) {
+		error = xfs_bwrite(bp);
+		xfs_buf_relse(bp);
+		if (error)
 			goto error0;
-		}
+
 		/*
 		 * INO btree root block
 		 */
@@ -296,6 +316,10 @@ xfs_growfs_data_private(
 				 XFS_AGB_TO_DADDR(mp, agno, XFS_IBT_BLOCK(mp)),
 				 BTOBB(mp->m_sb.sb_blocksize),
 				 XBF_LOCK | XBF_MAPPED);
+		if (!bp) {
+			error = ENOMEM;
+			goto error0;
+		}
 		block = XFS_BUF_TO_BLOCK(bp);
 		memset(block, 0, mp->m_sb.sb_blocksize);
 		block->bb_magic = cpu_to_be32(XFS_IBT_MAGIC);
@@ -303,10 +327,10 @@ xfs_growfs_data_private(
 		block->bb_numrecs = 0;
 		block->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK);
 		block->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK);
-		error = xfs_bwrite(mp, bp);
-		if (error) {
+		error = xfs_bwrite(bp);
+		xfs_buf_relse(bp);
+		if (error)
 			goto error0;
-		}
 	}
 	xfs_trans_agblocks_delta(tp, nfree);
 	/*
@@ -396,9 +420,9 @@ xfs_growfs_data_private(
 		 * just issue a warning and continue.  The real work is
 		 * already done and committed.
 		 */
-		if (!(error = xfs_bwrite(mp, bp))) {
-			continue;
-		} else {
+		error = xfs_bwrite(bp);
+		xfs_buf_relse(bp);
+		if (error) {
 			xfs_warn(mp,
 		"write error %d updating secondary superblock for ag %d",
 				error, agno);

fs/xfs/xfs_ialloc.c

@@ -150,7 +150,7 @@ xfs_check_agi_freecount(
 /*
  * Initialise a new set of inodes.
  */
-STATIC void
+STATIC int
 xfs_ialloc_inode_init(
 	struct xfs_mount	*mp,
 	struct xfs_trans	*tp,
@@ -202,8 +202,8 @@ xfs_ialloc_inode_init(
 		fbuf = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
 					 mp->m_bsize * blks_per_cluster,
 					 XBF_LOCK);
-		ASSERT(!xfs_buf_geterror(fbuf));
-
+		if (!fbuf)
+			return ENOMEM;
 		/*
 		 * Initialize all inodes in this buffer and then log them.
 		 *
@@ -225,6 +225,7 @@ xfs_ialloc_inode_init(
 		}
 		xfs_trans_inode_alloc_buf(tp, fbuf);
 	}
+	return 0;
 }
 
 /*
@@ -369,9 +370,11 @@ xfs_ialloc_ag_alloc(
 	 * rather than a linear progression to prevent the next generation
 	 * number from being easily guessable.
 	 */
-	xfs_ialloc_inode_init(args.mp, tp, agno, args.agbno, args.len,
-			      random32());
+	error = xfs_ialloc_inode_init(args.mp, tp, agno, args.agbno,
+			args.len, random32());
 
+	if (error)
+		return error;
 	/*
 	 * Convert the results.
 	 */
@@ -1502,7 +1505,7 @@ xfs_read_agi(
 		return XFS_ERROR(EFSCORRUPTED);
 	}
 
-	XFS_BUF_SET_VTYPE_REF(*bpp, B_FS_AGI, XFS_AGI_REF);
+	xfs_buf_set_ref(*bpp, XFS_AGI_REF);
 
 	xfs_check_agi_unlinked(agi);
 	return 0;

fs/xfs/xfs_iget.c

@@ -75,7 +75,6 @@ xfs_inode_alloc(
 		return NULL;
 	}
 
-	ASSERT(atomic_read(&ip->i_iocount) == 0);
 	ASSERT(atomic_read(&ip->i_pincount) == 0);
 	ASSERT(!spin_is_locked(&ip->i_flags_lock));
 	ASSERT(completion_done(&ip->i_flush));
@@ -150,7 +149,6 @@ xfs_inode_free(
 	}
 
 	/* asserts to verify all state is correct here */
-	ASSERT(atomic_read(&ip->i_iocount) == 0);
 	ASSERT(atomic_read(&ip->i_pincount) == 0);
 	ASSERT(!spin_is_locked(&ip->i_flags_lock));
 	ASSERT(completion_done(&ip->i_flush));

fs/xfs/xfs_inode.c

@@ -190,12 +190,6 @@ xfs_imap_to_bp(
 	}
 
 	xfs_inobp_check(mp, bp);
-
-	/*
-	 * Mark the buffer as an inode buffer now that it looks good
-	 */
-	XFS_BUF_SET_VTYPE(bp, B_FS_INO);
-
 	*bpp = bp;
 	return 0;
 }
@@ -1152,7 +1146,7 @@ xfs_ialloc(
 	/*
 	 * Log the new values stuffed into the inode.
 	 */
-	xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 	xfs_trans_log_inode(tp, ip, flags);
 
 	/* now that we have an i_mode we can setup inode ops and unlock */
@@ -1187,6 +1181,7 @@ xfs_isize_check(
 	xfs_fileoff_t		map_first;
 	int			nimaps;
 	xfs_bmbt_irec_t		imaps[2];
+	int			error;
 
 	if (!S_ISREG(ip->i_d.di_mode))
 		return;
@@ -1203,13 +1198,12 @@ xfs_isize_check(
 	 * The filesystem could be shutting down, so bmapi may return
 	 * an error.
 	 */
-	if (xfs_bmapi(NULL, ip, map_first,
+	error = xfs_bmapi_read(ip, map_first,
 			 (XFS_B_TO_FSB(mp,
-				       (xfs_ufsize_t)XFS_MAXIOFFSET(mp)) -
-			  map_first),
-			 XFS_BMAPI_ENTIRE, NULL, 0, imaps, &nimaps,
-			 NULL))
-	    return;
+			       (xfs_ufsize_t)XFS_MAXIOFFSET(mp)) - map_first),
+			 imaps, &nimaps, XFS_BMAPI_ENTIRE);
+	if (error)
+		return;
 	ASSERT(nimaps == 1);
 	ASSERT(imaps[0].br_startblock == HOLESTARTBLOCK);
 }
@@ -1297,7 +1291,7 @@ xfs_itruncate_extents(
 		 */
 		error = xfs_bmap_finish(&tp, &free_list, &committed);
 		if (committed)
-			xfs_trans_ijoin(tp, ip);
+			xfs_trans_ijoin(tp, ip, 0);
 		if (error)
 			goto out_bmap_cancel;
 
@@ -1313,7 +1307,7 @@ xfs_itruncate_extents(
 		error = xfs_trans_commit(tp, 0);
 		tp = ntp;
 
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 
 		if (error)
 			goto out;
@@ -1644,7 +1638,7 @@ xfs_iunlink_remove(
  * inodes that are in memory - they all must be marked stale and attached to
  * the cluster buffer.
  */
-STATIC void
+STATIC int
 xfs_ifree_cluster(
 	xfs_inode_t	*free_ip,
 	xfs_trans_t	*tp,
@@ -1690,6 +1684,8 @@ xfs_ifree_cluster(
 					mp->m_bsize * blks_per_cluster,
 					XBF_LOCK);
 
+		if (!bp)
+			return ENOMEM;
 		/*
 		 * Walk the inodes already attached to the buffer and mark them
 		 * stale. These will all have the flush locks held, so an
@@ -1799,6 +1795,7 @@ retry:
 	}
 
 	xfs_perag_put(pag);
+	return 0;
 }
 
 /*
@@ -1878,10 +1875,10 @@ xfs_ifree(
 	dip->di_mode = 0;
 
 	if (delete) {
-		xfs_ifree_cluster(ip, tp, first_ino);
+		error = xfs_ifree_cluster(ip, tp, first_ino);
 	}
 
-	return 0;
+	return error;
 }
 
 /*
@@ -2472,11 +2469,11 @@ cluster_corrupt_out:
 		 */
 		if (bp->b_iodone) {
 			XFS_BUF_UNDONE(bp);
-			XFS_BUF_STALE(bp);
+			xfs_buf_stale(bp);
 			xfs_buf_ioerror(bp, EIO);
 			xfs_buf_ioend(bp, 0);
 		} else {
-			XFS_BUF_STALE(bp);
+			xfs_buf_stale(bp);
 			xfs_buf_relse(bp);
 		}
 	}
@@ -2597,9 +2594,11 @@ xfs_iflush(
 		goto cluster_corrupt_out;
 
 	if (flags & SYNC_WAIT)
-		error = xfs_bwrite(mp, bp);
+		error = xfs_bwrite(bp);
 	else
-		xfs_bdwrite(mp, bp);
+		xfs_buf_delwri_queue(bp);
+
+	xfs_buf_relse(bp);
 	return error;
 
 corrupt_out:

fs/xfs/xfs_inode.h

@@ -257,7 +257,6 @@ typedef struct xfs_inode {
 
 	xfs_fsize_t		i_size;		/* in-memory size */
 	xfs_fsize_t		i_new_size;	/* size when write completes */
-	atomic_t		i_iocount;	/* outstanding I/O count */
 
 	/* VFS inode */
 	struct inode		i_vnode;	/* embedded VFS inode */

fs/xfs/xfs_inode_item.c

@@ -658,10 +658,8 @@ xfs_inode_item_unlock(
 
 	lock_flags = iip->ili_lock_flags;
 	iip->ili_lock_flags = 0;
-	if (lock_flags) {
+	if (lock_flags)
 		xfs_iunlock(ip, lock_flags);
-		IRELE(ip);
-	}
 }
 
 /*

fs/xfs/xfs_ioctl.c

@@ -1069,7 +1069,7 @@ xfs_ioctl_setattr(
 		}
 	}
 
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, 0);
 
 	/*
 	 * Change file ownership.  Must be the owner or privileged.

fs/xfs/xfs_iomap.c

@@ -208,22 +208,20 @@ xfs_iomap_write_direct(
 	if (error)
 		goto error1;
 
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, 0);
 
-	bmapi_flag = XFS_BMAPI_WRITE;
+	bmapi_flag = 0;
 	if (offset < ip->i_size || extsz)
 		bmapi_flag |= XFS_BMAPI_PREALLOC;
 
 	/*
-	 * Issue the xfs_bmapi() call to allocate the blocks.
-	 *
 	 * From this point onwards we overwrite the imap pointer that the
 	 * caller gave to us.
 	 */
 	xfs_bmap_init(&free_list, &firstfsb);
 	nimaps = 1;
-	error = xfs_bmapi(tp, ip, offset_fsb, count_fsb, bmapi_flag,
-		&firstfsb, 0, imap, &nimaps, &free_list);
+	error = xfs_bmapi_write(tp, ip, offset_fsb, count_fsb, bmapi_flag,
+				&firstfsb, 0, imap, &nimaps, &free_list);
 	if (error)
 		goto error0;
 
@@ -300,8 +298,8 @@ xfs_iomap_eof_want_preallocate(
 	while (count_fsb > 0) {
 		imaps = nimaps;
 		firstblock = NULLFSBLOCK;
-		error = xfs_bmapi(NULL, ip, start_fsb, count_fsb, 0,
-				  &firstblock, 0, imap, &imaps, NULL);
+		error = xfs_bmapi_read(ip, start_fsb, count_fsb, imap, &imaps,
+				       0);
 		if (error)
 			return error;
 		for (n = 0; n < imaps; n++) {
@@ -381,7 +379,6 @@ xfs_iomap_write_delay(
 	xfs_fileoff_t	last_fsb;
 	xfs_off_t	aligned_offset;
 	xfs_fileoff_t	ioalign;
-	xfs_fsblock_t	firstblock;
 	xfs_extlen_t	extsz;
 	int		nimaps;
 	xfs_bmbt_irec_t imap[XFS_WRITE_IMAPS];
@@ -425,12 +422,8 @@ retry:
 	}
 
 	nimaps = XFS_WRITE_IMAPS;
-	firstblock = NULLFSBLOCK;
-	error = xfs_bmapi(NULL, ip, offset_fsb,
-			  (xfs_filblks_t)(last_fsb - offset_fsb),
-			  XFS_BMAPI_DELAY | XFS_BMAPI_WRITE |
-			  XFS_BMAPI_ENTIRE, &firstblock, 1, imap,
-			  &nimaps, NULL);
+	error = xfs_bmapi_delay(ip, offset_fsb, last_fsb - offset_fsb,
+				imap, &nimaps, XFS_BMAPI_ENTIRE);
 	switch (error) {
 	case 0:
 	case ENOSPC:
@@ -535,7 +528,7 @@ xfs_iomap_write_allocate(
 				return XFS_ERROR(error);
 			}
 			xfs_ilock(ip, XFS_ILOCK_EXCL);
-			xfs_trans_ijoin(tp, ip);
+			xfs_trans_ijoin(tp, ip, 0);
 
 			xfs_bmap_init(&free_list, &first_block);
 
@@ -587,14 +580,12 @@ xfs_iomap_write_allocate(
 			}
 
 			/*
-			 * Go get the actual blocks.
-	 	 	 *
 			 * From this point onwards we overwrite the imap
 			 * pointer that the caller gave to us.
 			 */
-			error = xfs_bmapi(tp, ip, map_start_fsb, count_fsb,
-					XFS_BMAPI_WRITE, &first_block, 1,
-					imap, &nimaps, &free_list);
+			error = xfs_bmapi_write(tp, ip, map_start_fsb,
+						count_fsb, 0, &first_block, 1,
+						imap, &nimaps, &free_list);
 			if (error)
 				goto trans_cancel;
 
@@ -701,15 +692,15 @@ xfs_iomap_write_unwritten(
 		}
 
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 
 		/*
 		 * Modify the unwritten extent state of the buffer.
 		 */
 		xfs_bmap_init(&free_list, &firstfsb);
 		nimaps = 1;
-		error = xfs_bmapi(tp, ip, offset_fsb, count_fsb,
-				  XFS_BMAPI_WRITE|XFS_BMAPI_CONVERT, &firstfsb,
+		error = xfs_bmapi_write(tp, ip, offset_fsb, count_fsb,
+				  XFS_BMAPI_CONVERT, &firstfsb,
 				  1, &imap, &nimaps, &free_list);
 		if (error)
 			goto error_on_bmapi_transaction;

fs/xfs/xfs_iops.c

@@ -466,7 +466,7 @@ xfs_vn_getattr(
 	trace_xfs_getattr(ip);
 
 	if (XFS_FORCED_SHUTDOWN(mp))
-		return XFS_ERROR(EIO);
+		return -XFS_ERROR(EIO);
 
 	stat->size = XFS_ISIZE(ip);
 	stat->dev = inode->i_sb->s_dev;
@@ -612,7 +612,7 @@ xfs_setattr_nonsize(
 		}
 	}
 
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, 0);
 
 	/*
 	 * Change file ownership.  Must be the owner or privileged.
@@ -834,16 +834,16 @@ xfs_setattr_size(
 	 * care about here.
 	 */
 	if (ip->i_size != ip->i_d.di_size && iattr->ia_size > ip->i_d.di_size) {
-		error = xfs_flush_pages(ip, ip->i_d.di_size, iattr->ia_size,
-					XBF_ASYNC, FI_NONE);
+		error = xfs_flush_pages(ip, ip->i_d.di_size, iattr->ia_size, 0,
+					FI_NONE);
 		if (error)
 			goto out_unlock;
 	}
 
 	/*
-	 * Wait for all I/O to complete.
+	 * Wait for all direct I/O to complete.
 	 */
-	xfs_ioend_wait(ip);
+	inode_dio_wait(inode);
 
 	error = -block_truncate_page(inode->i_mapping, iattr->ia_size,
 				     xfs_get_blocks);
@@ -864,7 +864,7 @@ xfs_setattr_size(
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, 0);
 
 	/*
 	 * Only change the c/mtime if we are changing the size or we are

fs/xfs/xfs_log.c

@@ -880,8 +880,8 @@ xlog_iodone(xfs_buf_t *bp)
 	 */
 	if (XFS_TEST_ERROR((xfs_buf_geterror(bp)), l->l_mp,
 			XFS_ERRTAG_IODONE_IOERR, XFS_RANDOM_IODONE_IOERR)) {
-		xfs_ioerror_alert("xlog_iodone", l->l_mp, bp, XFS_BUF_ADDR(bp));
-		XFS_BUF_STALE(bp);
+		xfs_buf_ioerror_alert(bp, __func__);
+		xfs_buf_stale(bp);
 		xfs_force_shutdown(l->l_mp, SHUTDOWN_LOG_IO_ERROR);
 		/*
 		 * This flag will be propagated to the trans-committed
@@ -1047,7 +1047,7 @@ xlog_alloc_log(xfs_mount_t	*mp,
 	xlog_get_iclog_buffer_size(mp, log);
 
 	error = ENOMEM;
-	bp = xfs_buf_get_empty(log->l_iclog_size, mp->m_logdev_targp);
+	bp = xfs_buf_alloc(mp->m_logdev_targp, 0, log->l_iclog_size, 0);
 	if (!bp)
 		goto out_free_log;
 	bp->b_iodone = xlog_iodone;
@@ -1247,7 +1247,7 @@ xlog_bdstrat(
 
 	if (iclog->ic_state & XLOG_STATE_IOERROR) {
 		xfs_buf_ioerror(bp, EIO);
-		XFS_BUF_STALE(bp);
+		xfs_buf_stale(bp);
 		xfs_buf_ioend(bp, 0);
 		/*
 		 * It would seem logical to return EIO here, but we rely on
@@ -1387,9 +1387,9 @@ xlog_sync(xlog_t		*log,
 	 */
 	XFS_BUF_WRITE(bp);
 
-	if ((error = xlog_bdstrat(bp))) {
-		xfs_ioerror_alert("xlog_sync", log->l_mp, bp,
-				  XFS_BUF_ADDR(bp));
+	error = xlog_bdstrat(bp);
+	if (error) {
+		xfs_buf_ioerror_alert(bp, "xlog_sync");
 		return error;
 	}
 	if (split) {
@@ -1423,9 +1423,9 @@ xlog_sync(xlog_t		*log,
 		/* account for internal log which doesn't start at block #0 */
 		XFS_BUF_SET_ADDR(bp, XFS_BUF_ADDR(bp) + log->l_logBBstart);
 		XFS_BUF_WRITE(bp);
-		if ((error = xlog_bdstrat(bp))) {
-			xfs_ioerror_alert("xlog_sync (split)", log->l_mp,
-					  bp, XFS_BUF_ADDR(bp));
+		error = xlog_bdstrat(bp);
+		if (error) {
+			xfs_buf_ioerror_alert(bp, "xlog_sync (split)");
 			return error;
 		}
 	}

fs/xfs/xfs_log_recover.c

@@ -183,8 +183,7 @@ xlog_bread_noalign(
 	xfsbdstrat(log->l_mp, bp);
 	error = xfs_buf_iowait(bp);
 	if (error)
-		xfs_ioerror_alert("xlog_bread", log->l_mp,
-				  bp, XFS_BUF_ADDR(bp));
+		xfs_buf_ioerror_alert(bp, __func__);
 	return error;
 }
 
@@ -268,9 +267,10 @@ xlog_bwrite(
 	xfs_buf_lock(bp);
 	XFS_BUF_SET_COUNT(bp, BBTOB(nbblks));
 
-	if ((error = xfs_bwrite(log->l_mp, bp)))
-		xfs_ioerror_alert("xlog_bwrite", log->l_mp,
-				  bp, XFS_BUF_ADDR(bp));
+	error = xfs_bwrite(bp);
+	if (error)
+		xfs_buf_ioerror_alert(bp, __func__);
+	xfs_buf_relse(bp);
 	return error;
 }
 
@@ -361,9 +361,7 @@ xlog_recover_iodone(
 		 * We're not going to bother about retrying
 		 * this during recovery. One strike!
 		 */
-		xfs_ioerror_alert("xlog_recover_iodone",
-					bp->b_target->bt_mount, bp,
-					XFS_BUF_ADDR(bp));
+		xfs_buf_ioerror_alert(bp, __func__);
 		xfs_force_shutdown(bp->b_target->bt_mount,
 					SHUTDOWN_META_IO_ERROR);
 	}
@@ -2135,8 +2133,7 @@ xlog_recover_buffer_pass2(
 		return XFS_ERROR(ENOMEM);
 	error = bp->b_error;
 	if (error) {
-		xfs_ioerror_alert("xlog_recover_do..(read#1)", mp,
-				  bp, buf_f->blf_blkno);
+		xfs_buf_ioerror_alert(bp, "xlog_recover_do..(read#1)");
 		xfs_buf_relse(bp);
 		return error;
 	}
@@ -2171,15 +2168,16 @@ xlog_recover_buffer_pass2(
 	    be16_to_cpu(*((__be16 *)xfs_buf_offset(bp, 0))) &&
 	    (XFS_BUF_COUNT(bp) != MAX(log->l_mp->m_sb.sb_blocksize,
 			(__uint32_t)XFS_INODE_CLUSTER_SIZE(log->l_mp)))) {
-		XFS_BUF_STALE(bp);
-		error = xfs_bwrite(mp, bp);
+		xfs_buf_stale(bp);
+		error = xfs_bwrite(bp);
 	} else {
 		ASSERT(bp->b_target->bt_mount == mp);
 		bp->b_iodone = xlog_recover_iodone;
-		xfs_bdwrite(mp, bp);
+		xfs_buf_delwri_queue(bp);
 	}
 
-	return (error);
+	xfs_buf_relse(bp);
+	return error;
 }
 
 STATIC int
@@ -2230,8 +2228,7 @@ xlog_recover_inode_pass2(
 	}
 	error = bp->b_error;
 	if (error) {
-		xfs_ioerror_alert("xlog_recover_do..(read#2)", mp,
-				  bp, in_f->ilf_blkno);
+		xfs_buf_ioerror_alert(bp, "xlog_recover_do..(read#2)");
 		xfs_buf_relse(bp);
 		goto error;
 	}
@@ -2439,7 +2436,8 @@ xlog_recover_inode_pass2(
 write_inode_buffer:
 	ASSERT(bp->b_target->bt_mount == mp);
 	bp->b_iodone = xlog_recover_iodone;
-	xfs_bdwrite(mp, bp);
+	xfs_buf_delwri_queue(bp);
+	xfs_buf_relse(bp);
 error:
 	if (need_free)
 		kmem_free(in_f);
@@ -2537,8 +2535,7 @@ xlog_recover_dquot_pass2(
 			     XFS_FSB_TO_BB(mp, dq_f->qlf_len),
 			     0, &bp);
 	if (error) {
-		xfs_ioerror_alert("xlog_recover_do..(read#3)", mp,
-				  bp, dq_f->qlf_blkno);
+		xfs_buf_ioerror_alert(bp, "xlog_recover_do..(read#3)");
 		return error;
 	}
 	ASSERT(bp);
@@ -2561,7 +2558,8 @@ xlog_recover_dquot_pass2(
 	ASSERT(dq_f->qlf_size == 2);
 	ASSERT(bp->b_target->bt_mount == mp);
 	bp->b_iodone = xlog_recover_iodone;
-	xfs_bdwrite(mp, bp);
+	xfs_buf_delwri_queue(bp);
+	xfs_buf_relse(bp);
 
 	return (0);
 }
@@ -3656,7 +3654,7 @@ xlog_do_recover(
 		return error;
 	}
 
-	XFS_bflush(log->l_mp->m_ddev_targp);
+	xfs_flush_buftarg(log->l_mp->m_ddev_targp, 1);
 
 	/*
 	 * If IO errors happened during recovery, bail out.
@@ -3689,8 +3687,7 @@ xlog_do_recover(
 	xfsbdstrat(log->l_mp, bp);
 	error = xfs_buf_iowait(bp);
 	if (error) {
-		xfs_ioerror_alert("xlog_do_recover",
-				  log->l_mp, bp, XFS_BUF_ADDR(bp));
+		xfs_buf_ioerror_alert(bp, __func__);
 		ASSERT(0);
 		xfs_buf_relse(bp);
 		return error;

fs/xfs/xfs_mount.c

@@ -44,9 +44,6 @@
 #include "xfs_trace.h"
 
 
-STATIC void	xfs_unmountfs_wait(xfs_mount_t *);
-
-
 #ifdef HAVE_PERCPU_SB
 STATIC void	xfs_icsb_balance_counter(xfs_mount_t *, xfs_sb_field_t,
 						int);
@@ -1484,7 +1481,7 @@ xfs_unmountfs(
 	 * state as much as possible.
 	 */
 	xfs_reclaim_inodes(mp, 0);
-	XFS_bflush(mp->m_ddev_targp);
+	xfs_flush_buftarg(mp->m_ddev_targp, 1);
 	xfs_reclaim_inodes(mp, SYNC_WAIT);
 
 	xfs_qm_unmount(mp);
@@ -1496,11 +1493,6 @@ xfs_unmountfs(
 	 */
 	xfs_log_force(mp, XFS_LOG_SYNC);
 
-	xfs_binval(mp->m_ddev_targp);
-	if (mp->m_rtdev_targp) {
-		xfs_binval(mp->m_rtdev_targp);
-	}
-
 	/*
 	 * Unreserve any blocks we have so that when we unmount we don't account
 	 * the reserved free space as used. This is really only necessary for
@@ -1526,7 +1518,16 @@ xfs_unmountfs(
 		xfs_warn(mp, "Unable to update superblock counters. "
 				"Freespace may not be correct on next mount.");
 	xfs_unmountfs_writesb(mp);
-	xfs_unmountfs_wait(mp); 		/* wait for async bufs */
+
+	/*
+	 * Make sure all buffers have been flushed and completed before
+	 * unmounting the log.
+	 */
+	error = xfs_flush_buftarg(mp->m_ddev_targp, 1);
+	if (error)
+		xfs_warn(mp, "%d busy buffers during unmount.", error);
+	xfs_wait_buftarg(mp->m_ddev_targp);
+
 	xfs_log_unmount_write(mp);
 	xfs_log_unmount(mp);
 	xfs_uuid_unmount(mp);
@@ -1537,16 +1538,6 @@ xfs_unmountfs(
 	xfs_free_perag(mp);
 }
 
-STATIC void
-xfs_unmountfs_wait(xfs_mount_t *mp)
-{
-	if (mp->m_logdev_targp != mp->m_ddev_targp)
-		xfs_wait_buftarg(mp->m_logdev_targp);
-	if (mp->m_rtdev_targp)
-		xfs_wait_buftarg(mp->m_rtdev_targp);
-	xfs_wait_buftarg(mp->m_ddev_targp);
-}
-
 int
 xfs_fs_writable(xfs_mount_t *mp)
 {
@@ -1612,15 +1603,14 @@ xfs_unmountfs_writesb(xfs_mount_t *mp)
 
 		XFS_BUF_UNDONE(sbp);
 		XFS_BUF_UNREAD(sbp);
-		XFS_BUF_UNDELAYWRITE(sbp);
+		xfs_buf_delwri_dequeue(sbp);
 		XFS_BUF_WRITE(sbp);
 		XFS_BUF_UNASYNC(sbp);
 		ASSERT(sbp->b_target == mp->m_ddev_targp);
 		xfsbdstrat(mp, sbp);
 		error = xfs_buf_iowait(sbp);
 		if (error)
-			xfs_ioerror_alert("xfs_unmountfs_writesb",
-					  mp, sbp, XFS_BUF_ADDR(sbp));
+			xfs_buf_ioerror_alert(sbp, __func__);
 		xfs_buf_relse(sbp);
 	}
 	return error;

fs/xfs/xfs_qm.c

@@ -1296,7 +1296,8 @@ xfs_qm_dqiter_bufs(
 			break;
 
 		xfs_qm_reset_dqcounts(mp, bp, firstid, type);
-		xfs_bdwrite(mp, bp);
+		xfs_buf_delwri_queue(bp);
+		xfs_buf_relse(bp);
 		/*
 		 * goto the next block.
 		 */
@@ -1346,11 +1347,8 @@ xfs_qm_dqiterate(
 		 * the inode is never added to the transaction.
 		 */
 		xfs_ilock(qip, XFS_ILOCK_SHARED);
-		error = xfs_bmapi(NULL, qip, lblkno,
-				  maxlblkcnt - lblkno,
-				  XFS_BMAPI_METADATA,
-				  NULL,
-				  0, map, &nmaps, NULL);
+		error = xfs_bmapi_read(qip, lblkno, maxlblkcnt - lblkno,
+				       map, &nmaps, 0);
 		xfs_iunlock(qip, XFS_ILOCK_SHARED);
 		if (error)
 			break;
@@ -1683,7 +1681,7 @@ xfs_qm_quotacheck(
 	 * quotacheck'd stamp on the superblock. So, here we do a synchronous
 	 * flush.
 	 */
-	XFS_bflush(mp->m_ddev_targp);
+	xfs_flush_buftarg(mp->m_ddev_targp, 1);
 
 	/*
 	 * If one type of quotas is off, then it will lose its

fs/xfs/xfs_qm_syscalls.c

@@ -261,7 +261,7 @@ xfs_qm_scall_trunc_qfile(
 	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, 0);
 
 	error = xfs_itruncate_data(&tp, ip, 0);
 	if (error) {

fs/xfs/xfs_rename.c

@@ -170,12 +170,12 @@ xfs_rename(
 	 * we can rely on either trans_commit or trans_cancel to unlock
 	 * them.
 	 */
-	xfs_trans_ijoin_ref(tp, src_dp, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, src_dp, XFS_ILOCK_EXCL);
 	if (new_parent)
-		xfs_trans_ijoin_ref(tp, target_dp, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin_ref(tp, src_ip, XFS_ILOCK_EXCL);
+		xfs_trans_ijoin(tp, target_dp, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, src_ip, XFS_ILOCK_EXCL);
 	if (target_ip)
-		xfs_trans_ijoin_ref(tp, target_ip, XFS_ILOCK_EXCL);
+		xfs_trans_ijoin(tp, target_ip, XFS_ILOCK_EXCL);
 
 	/*
 	 * If we are using project inheritance, we only allow renames

fs/xfs/xfs_rtalloc.c

@@ -112,7 +112,7 @@ xfs_growfs_rt_alloc(
 		 * Lock the inode.
 		 */
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
-		xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL);
+		xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 
 		xfs_bmap_init(&flist, &firstblock);
 		/*
@@ -120,9 +120,9 @@ xfs_growfs_rt_alloc(
 		 */
 		nmap = 1;
 		cancelflags |= XFS_TRANS_ABORT;
-		error = xfs_bmapi(tp, ip, oblocks, nblocks - oblocks,
-			XFS_BMAPI_WRITE | XFS_BMAPI_METADATA, &firstblock,
-			resblks, &map, &nmap, &flist);
+		error = xfs_bmapi_write(tp, ip, oblocks, nblocks - oblocks,
+					XFS_BMAPI_METADATA, &firstblock,
+					resblks, &map, &nmap, &flist);
 		if (!error && nmap < 1)
 			error = XFS_ERROR(ENOSPC);
 		if (error)
@@ -155,7 +155,7 @@ xfs_growfs_rt_alloc(
 			 * Lock the bitmap inode.
 			 */
 			xfs_ilock(ip, XFS_ILOCK_EXCL);
-			xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL);
+			xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 			/*
 			 * Get a buffer for the block.
 			 */
@@ -856,33 +856,23 @@ xfs_rtbuf_get(
 	xfs_buf_t	**bpp)		/* output: buffer for the block */
 {
 	xfs_buf_t	*bp;		/* block buffer, result */
-	xfs_daddr_t	d;		/* disk addr of block */
-	int		error;		/* error value */
-	xfs_fsblock_t	fsb;		/* fs block number for block */
 	xfs_inode_t	*ip;		/* bitmap or summary inode */
+	xfs_bmbt_irec_t	map;
+	int		nmap;
+	int		error;		/* error value */
 
 	ip = issum ? mp->m_rsumip : mp->m_rbmip;
-	/*
-	 * Map from the file offset (block) and inode number to the
-	 * file system block.
-	 */
-	error = xfs_bmapi_single(tp, ip, XFS_DATA_FORK, &fsb, block);
-	if (error) {
+
+	error = xfs_bmapi_read(ip, block, 1, &map, &nmap, XFS_DATA_FORK);
+	if (error)
 		return error;
-	}
-	ASSERT(fsb != NULLFSBLOCK);
-	/*
-	 * Convert to disk address for buffer cache.
-	 */
-	d = XFS_FSB_TO_DADDR(mp, fsb);
-	/*
-	 * Read the buffer.
-	 */
-	error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, d,
+
+	ASSERT(map.br_startblock != NULLFSBLOCK);
+	error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp,
+				   XFS_FSB_TO_DADDR(mp, map.br_startblock),
 				   mp->m_bsize, 0, &bp);
-	if (error) {
+	if (error)
 		return error;
-	}
 	ASSERT(!xfs_buf_geterror(bp));
 	*bpp = bp;
 	return 0;
@@ -1970,7 +1960,7 @@ xfs_growfs_rt(
 		 * Lock out other callers by grabbing the bitmap inode lock.
 		 */
 		xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
-		xfs_trans_ijoin_ref(tp, mp->m_rbmip, XFS_ILOCK_EXCL);
+		xfs_trans_ijoin(tp, mp->m_rbmip, XFS_ILOCK_EXCL);
 		/*
 		 * Update the bitmap inode's size.
 		 */
@@ -1982,7 +1972,7 @@ xfs_growfs_rt(
 		 * Get the summary inode into the transaction.
 		 */
 		xfs_ilock(mp->m_rsumip, XFS_ILOCK_EXCL);
-		xfs_trans_ijoin_ref(tp, mp->m_rsumip, XFS_ILOCK_EXCL);
+		xfs_trans_ijoin(tp, mp->m_rsumip, XFS_ILOCK_EXCL);
 		/*
 		 * Update the summary inode's size.
 		 */
@@ -2153,7 +2143,7 @@ xfs_rtfree_extent(
 	 * Synchronize by locking the bitmap inode.
 	 */
 	xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin_ref(tp, mp->m_rbmip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, mp->m_rbmip, XFS_ILOCK_EXCL);
 
 #if defined(__KERNEL__) && defined(DEBUG)
 	/*

fs/xfs/xfs_rw.c

@@ -91,24 +91,6 @@ xfs_do_force_shutdown(
 	}
 }
 
-/*
- * Prints out an ALERT message about I/O error.
- */
-void
-xfs_ioerror_alert(
-	char			*func,
-	struct xfs_mount	*mp,
-	xfs_buf_t		*bp,
-	xfs_daddr_t		blkno)
-{
-	xfs_alert(mp,
-		 "I/O error occurred: meta-data dev %s block 0x%llx"
-		 "       (\"%s\") error %d buf count %zd",
-		xfs_buf_target_name(bp->b_target),
-		(__uint64_t)blkno, func,
-		bp->b_error, XFS_BUF_COUNT(bp));
-}
-
 /*
  * This isn't an absolute requirement, but it is
  * just a good idea to call xfs_read_buf instead of
@@ -143,14 +125,13 @@ xfs_read_buf(
 	} else {
 		*bpp = NULL;
 		if (error) {
-			xfs_ioerror_alert("xfs_read_buf", mp, bp, XFS_BUF_ADDR(bp));
+			xfs_buf_ioerror_alert(bp, __func__);
 		} else {
 			error = XFS_ERROR(EIO);
 		}
 		if (bp) {
 			XFS_BUF_UNDONE(bp);
-			XFS_BUF_UNDELAYWRITE(bp);
-			XFS_BUF_STALE(bp);
+			xfs_buf_stale(bp);
 			/*
 			 * brelse clears B_ERROR and b_error
 			 */

fs/xfs/xfs_rw.h

@@ -42,8 +42,6 @@ xfs_fsb_to_db(struct xfs_inode *ip, xfs_fsblock_t fsb)
 extern int xfs_read_buf(struct xfs_mount *mp, xfs_buftarg_t *btp,
 			xfs_daddr_t blkno, int len, uint flags,
 			struct xfs_buf **bpp);
-extern void xfs_ioerror_alert(char *func, struct xfs_mount *mp,
-				xfs_buf_t *bp, xfs_daddr_t blkno);
 extern xfs_extlen_t xfs_get_extsz_hint(struct xfs_inode *ip);
 
 #endif /* __XFS_RW_H__ */

fs/xfs/xfs_super.c

@@ -796,8 +796,6 @@ xfs_fs_destroy_inode(
 	if (is_bad_inode(inode))
 		goto out_reclaim;
 
-	xfs_ioend_wait(ip);
-
 	ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) || ip->i_delayed_blks == 0);
 
 	/*
@@ -837,7 +835,6 @@ xfs_fs_inode_init_once(
 	inode_init_once(VFS_I(ip));
 
 	/* xfs inode */
-	atomic_set(&ip->i_iocount, 0);
 	atomic_set(&ip->i_pincount, 0);
 	spin_lock_init(&ip->i_flags_lock);
 	init_waitqueue_head(&ip->i_ipin_wait);
@@ -887,7 +884,7 @@ xfs_log_inode(
 	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 	return xfs_trans_commit(tp, 0);
 }
@@ -914,9 +911,8 @@ xfs_fs_write_inode(
 		 * of forcing it all the way to stable storage using a
 		 * synchronous transaction we let the log force inside the
 		 * ->sync_fs call do that for thus, which reduces the number
-		 * of synchronous log foces dramatically.
+		 * of synchronous log forces dramatically.
 		 */
-		xfs_ioend_wait(ip);
 		error = xfs_log_inode(ip);
 		if (error)
 			goto out;
@@ -1019,7 +1015,7 @@ xfs_fs_put_super(
 	 */
 	xfs_filestream_unmount(mp);
 
-	XFS_bflush(mp->m_ddev_targp);
+	xfs_flush_buftarg(mp->m_ddev_targp, 1);
 
 	xfs_unmountfs(mp);
 	xfs_freesb(mp);
@@ -1443,7 +1439,7 @@ xfs_fs_fill_super(
 	 */
 	xfs_filestream_unmount(mp);
 
-	XFS_bflush(mp->m_ddev_targp);
+	xfs_flush_buftarg(mp->m_ddev_targp, 1);
 
 	xfs_unmountfs(mp);
 	goto out_free_sb;
@@ -1670,7 +1666,6 @@ init_xfs_fs(void)
 	printk(KERN_INFO XFS_VERSION_STRING " with "
 			 XFS_BUILD_OPTIONS " enabled\n");
 
-	xfs_ioend_init();
 	xfs_dir_startup();
 
 	error = xfs_init_zones();

fs/xfs/xfs_sync.c

@@ -227,21 +227,17 @@ xfs_sync_inode_data(
 	int			error = 0;
 
 	if (!mapping_tagged(mapping, PAGECACHE_TAG_DIRTY))
-		goto out_wait;
+		return 0;
 
 	if (!xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED)) {
 		if (flags & SYNC_TRYLOCK)
-			goto out_wait;
+			return 0;
 		xfs_ilock(ip, XFS_IOLOCK_SHARED);
 	}
 
 	error = xfs_flush_pages(ip, 0, -1, (flags & SYNC_WAIT) ?
 				0 : XBF_ASYNC, FI_NONE);
 	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
-
- out_wait:
-	if (flags & SYNC_WAIT)
-		xfs_ioend_wait(ip);
 	return error;
 }
 
@@ -322,6 +318,7 @@ xfs_sync_fsdata(
 	struct xfs_mount	*mp)
 {
 	struct xfs_buf		*bp;
+	int			error;
 
 	/*
 	 * If the buffer is pinned then push on the log so we won't get stuck
@@ -334,8 +331,9 @@ xfs_sync_fsdata(
 	bp = xfs_getsb(mp, 0);
 	if (xfs_buf_ispinned(bp))
 		xfs_log_force(mp, 0);
-
-	return xfs_bwrite(mp, bp);
+	error = xfs_bwrite(bp);
+	xfs_buf_relse(bp);
+	return error;
 }
 
 /*
@@ -379,7 +377,7 @@ xfs_quiesce_data(
 
 	/* flush data-only devices */
 	if (mp->m_rtdev_targp)
-		XFS_bflush(mp->m_rtdev_targp);
+		xfs_flush_buftarg(mp->m_rtdev_targp, 1);
 
 	return error ? error : error2;
 }

fs/xfs/xfs_trace.h

@@ -30,6 +30,7 @@ struct xfs_buf_log_item;
 struct xfs_da_args;
 struct xfs_da_node_entry;
 struct xfs_dquot;
+struct xfs_log_item;
 struct xlog_ticket;
 struct log;
 struct xlog_recover;
@@ -320,7 +321,6 @@ DEFINE_BUF_EVENT(xfs_buf_rele);
 DEFINE_BUF_EVENT(xfs_buf_iodone);
 DEFINE_BUF_EVENT(xfs_buf_iorequest);
 DEFINE_BUF_EVENT(xfs_buf_bawrite);
-DEFINE_BUF_EVENT(xfs_buf_bdwrite);
 DEFINE_BUF_EVENT(xfs_buf_lock);
 DEFINE_BUF_EVENT(xfs_buf_lock_done);
 DEFINE_BUF_EVENT(xfs_buf_trylock);
@@ -577,6 +577,7 @@ DEFINE_INODE_EVENT(xfs_vm_bmap);
 DEFINE_INODE_EVENT(xfs_file_ioctl);
 DEFINE_INODE_EVENT(xfs_file_compat_ioctl);
 DEFINE_INODE_EVENT(xfs_ioctl_setattr);
+DEFINE_INODE_EVENT(xfs_dir_fsync);
 DEFINE_INODE_EVENT(xfs_file_fsync);
 DEFINE_INODE_EVENT(xfs_destroy_inode);
 DEFINE_INODE_EVENT(xfs_write_inode);
@@ -853,6 +854,42 @@ DEFINE_LOGGRANT_EVENT(xfs_log_ungrant_enter);
 DEFINE_LOGGRANT_EVENT(xfs_log_ungrant_exit);
 DEFINE_LOGGRANT_EVENT(xfs_log_ungrant_sub);
 
+DECLARE_EVENT_CLASS(xfs_log_item_class,
+	TP_PROTO(struct xfs_log_item *lip),
+	TP_ARGS(lip),
+	TP_STRUCT__entry(
+		__field(dev_t, dev)
+		__field(void *, lip)
+		__field(uint, type)
+		__field(uint, flags)
+		__field(xfs_lsn_t, lsn)
+	),
+	TP_fast_assign(
+		__entry->dev = lip->li_mountp->m_super->s_dev;
+		__entry->lip = lip;
+		__entry->type = lip->li_type;
+		__entry->flags = lip->li_flags;
+		__entry->lsn = lip->li_lsn;
+	),
+	TP_printk("dev %d:%d lip 0x%p lsn %d/%d type %s flags %s",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __entry->lip,
+		  CYCLE_LSN(__entry->lsn), BLOCK_LSN(__entry->lsn),
+		  __print_symbolic(__entry->type, XFS_LI_TYPE_DESC),
+		  __print_flags(__entry->flags, "|", XFS_LI_FLAGS))
+)
+
+#define DEFINE_LOG_ITEM_EVENT(name) \
+DEFINE_EVENT(xfs_log_item_class, name, \
+	TP_PROTO(struct xfs_log_item *lip), \
+	TP_ARGS(lip))
+DEFINE_LOG_ITEM_EVENT(xfs_ail_push);
+DEFINE_LOG_ITEM_EVENT(xfs_ail_pushbuf);
+DEFINE_LOG_ITEM_EVENT(xfs_ail_pushbuf_pinned);
+DEFINE_LOG_ITEM_EVENT(xfs_ail_pinned);
+DEFINE_LOG_ITEM_EVENT(xfs_ail_locked);
+
+
 DECLARE_EVENT_CLASS(xfs_file_class,
 	TP_PROTO(struct xfs_inode *ip, size_t count, loff_t offset, int flags),
 	TP_ARGS(ip, count, offset, flags),

fs/xfs/xfs_trans.c

@@ -1790,9 +1790,7 @@ xfs_trans_commit_cil(
 }
 
 /*
- * xfs_trans_commit
- *
- * Commit the given transaction to the log a/synchronously.
+ * Commit the given transaction to the log.
  *
  * XFS disk error handling mechanism is not based on a typical
  * transaction abort mechanism. Logically after the filesystem
@@ -1804,10 +1802,9 @@ xfs_trans_commit_cil(
  * Do not reference the transaction structure after this call.
  */
 int
-_xfs_trans_commit(
+xfs_trans_commit(
 	struct xfs_trans	*tp,
-	uint			flags,
-	int			*log_flushed)
+	uint			flags)
 {
 	struct xfs_mount	*mp = tp->t_mountp;
 	xfs_lsn_t		commit_lsn = -1;
@@ -1866,7 +1863,7 @@ _xfs_trans_commit(
 	if (sync) {
 		if (!error) {
 			error = _xfs_log_force_lsn(mp, commit_lsn,
-				      XFS_LOG_SYNC, log_flushed);
+				      XFS_LOG_SYNC, NULL);
 		}
 		XFS_STATS_INC(xs_trans_sync);
 	} else {
@@ -2021,6 +2018,6 @@ xfs_trans_roll(
 	if (error)
 		return error;
 
-	xfs_trans_ijoin(trans, dp);
+	xfs_trans_ijoin(trans, dp, 0);
 	return 0;
 }

fs/xfs/xfs_trans.h

@@ -470,8 +470,7 @@ void		xfs_trans_stale_inode_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_dquot_buf(xfs_trans_t *, struct xfs_buf *, uint);
 void		xfs_trans_inode_alloc_buf(xfs_trans_t *, struct xfs_buf *);
 void		xfs_trans_ichgtime(struct xfs_trans *, struct xfs_inode *, int);
-void		xfs_trans_ijoin_ref(struct xfs_trans *, struct xfs_inode *, uint);
-void		xfs_trans_ijoin(struct xfs_trans *, struct xfs_inode *);
+void		xfs_trans_ijoin(struct xfs_trans *, struct xfs_inode *, uint);
 void		xfs_trans_log_buf(xfs_trans_t *, struct xfs_buf *, uint, uint);
 void		xfs_trans_log_inode(xfs_trans_t *, struct xfs_inode *, uint);
 struct xfs_efi_log_item	*xfs_trans_get_efi(xfs_trans_t *, uint);
@@ -487,10 +486,7 @@ void		xfs_trans_log_efd_extent(xfs_trans_t *,
 					 struct xfs_efd_log_item *,
 					 xfs_fsblock_t,
 					 xfs_extlen_t);
-int		_xfs_trans_commit(xfs_trans_t *,
-				  uint flags,
-				  int *);
-#define xfs_trans_commit(tp, flags)	_xfs_trans_commit(tp, flags, NULL)
+int		xfs_trans_commit(xfs_trans_t *, uint flags);
 void		xfs_trans_cancel(xfs_trans_t *, int);
 int		xfs_trans_ail_init(struct xfs_mount *);
 void		xfs_trans_ail_destroy(struct xfs_mount *);

fs/xfs/xfs_trans_ail.c

@@ -26,6 +26,7 @@
 #include "xfs_ag.h"
 #include "xfs_mount.h"
 #include "xfs_trans_priv.h"
+#include "xfs_trace.h"
 #include "xfs_error.h"
 
 #ifdef DEBUG
@@ -364,12 +365,24 @@ xfsaild_push(
 	xfs_lsn_t		lsn;
 	xfs_lsn_t		target;
 	long			tout = 10;
-	int			flush_log = 0;
 	int			stuck = 0;
 	int			count = 0;
 	int			push_xfsbufd = 0;
 
+	/*
+	 * If last time we ran we encountered pinned items, force the log first
+	 * and wait for it before pushing again.
+	 */
 	spin_lock(&ailp->xa_lock);
+	if (ailp->xa_last_pushed_lsn == 0 && ailp->xa_log_flush &&
+	    !list_empty(&ailp->xa_ail)) {
+		ailp->xa_log_flush = 0;
+		spin_unlock(&ailp->xa_lock);
+		XFS_STATS_INC(xs_push_ail_flush);
+		xfs_log_force(mp, XFS_LOG_SYNC);
+		spin_lock(&ailp->xa_lock);
+	}
+
 	target = ailp->xa_target;
 	lip = xfs_trans_ail_cursor_first(ailp, &cur, ailp->xa_last_pushed_lsn);
 	if (!lip || XFS_FORCED_SHUTDOWN(mp)) {
@@ -413,16 +426,20 @@ xfsaild_push(
 		switch (lock_result) {
 		case XFS_ITEM_SUCCESS:
 			XFS_STATS_INC(xs_push_ail_success);
+			trace_xfs_ail_push(lip);
+
 			IOP_PUSH(lip);
 			ailp->xa_last_pushed_lsn = lsn;
 			break;
 
 		case XFS_ITEM_PUSHBUF:
 			XFS_STATS_INC(xs_push_ail_pushbuf);
+			trace_xfs_ail_pushbuf(lip);
 
 			if (!IOP_PUSHBUF(lip)) {
+				trace_xfs_ail_pushbuf_pinned(lip);
 				stuck++;
-				flush_log = 1;
+				ailp->xa_log_flush++;
 			} else {
 				ailp->xa_last_pushed_lsn = lsn;
 			}
@@ -431,12 +448,15 @@ xfsaild_push(
 
 		case XFS_ITEM_PINNED:
 			XFS_STATS_INC(xs_push_ail_pinned);
+			trace_xfs_ail_pinned(lip);
+
 			stuck++;
-			flush_log = 1;
+			ailp->xa_log_flush++;
 			break;
 
 		case XFS_ITEM_LOCKED:
 			XFS_STATS_INC(xs_push_ail_locked);
+			trace_xfs_ail_locked(lip);
 			stuck++;
 			break;
 
@@ -476,16 +496,6 @@ xfsaild_push(
 	xfs_trans_ail_cursor_done(ailp, &cur);
 	spin_unlock(&ailp->xa_lock);
 
-	if (flush_log) {
-		/*
-		 * If something we need to push out was pinned, then
-		 * push out the log so it will become unpinned and
-		 * move forward in the AIL.
-		 */
-		XFS_STATS_INC(xs_push_ail_flush);
-		xfs_log_force(mp, 0);
-	}
-
 	if (push_xfsbufd) {
 		/* we've got delayed write buffers to flush */
 		wake_up_process(mp->m_ddev_targp->bt_task);
@@ -496,6 +506,7 @@ out_done:
 	if (!count) {
 		/* We're past our target or empty, so idle */
 		ailp->xa_last_pushed_lsn = 0;
+		ailp->xa_log_flush = 0;
 
 		tout = 50;
 	} else if (XFS_LSN_CMP(lsn, target) >= 0) {
@@ -514,9 +525,13 @@ out_done:
 		 * were stuck.
 		 *
 		 * Backoff a bit more to allow some I/O to complete before
-		 * continuing from where we were.
+		 * restarting from the start of the AIL. This prevents us
+		 * from spinning on the same items, and if they are pinned will
+		 * all the restart to issue a log force to unpin the stuck
+		 * items.
 		 */
 		tout = 20;
+		ailp->xa_last_pushed_lsn = 0;
 	}
 
 	return tout;

fs/xfs/xfs_trans_buf.c

@@ -160,8 +160,10 @@ xfs_trans_get_buf(xfs_trans_t	*tp,
 	bp = xfs_trans_buf_item_match(tp, target_dev, blkno, len);
 	if (bp != NULL) {
 		ASSERT(xfs_buf_islocked(bp));
-		if (XFS_FORCED_SHUTDOWN(tp->t_mountp))
-			XFS_BUF_SUPER_STALE(bp);
+		if (XFS_FORCED_SHUTDOWN(tp->t_mountp)) {
+			xfs_buf_stale(bp);
+			XFS_BUF_DONE(bp);
+		}
 
 		/*
 		 * If the buffer is stale then it was binval'ed
@@ -294,8 +296,7 @@ xfs_trans_read_buf(
 
 		if (bp->b_error) {
 			error = bp->b_error;
-			xfs_ioerror_alert("xfs_trans_read_buf", mp,
-					  bp, blkno);
+			xfs_buf_ioerror_alert(bp, __func__);
 			xfs_buf_relse(bp);
 			return error;
 		}
@@ -337,8 +338,7 @@ xfs_trans_read_buf(
 			xfsbdstrat(tp->t_mountp, bp);
 			error = xfs_buf_iowait(bp);
 			if (error) {
-				xfs_ioerror_alert("xfs_trans_read_buf", mp,
-						  bp, blkno);
+				xfs_buf_ioerror_alert(bp, __func__);
 				xfs_buf_relse(bp);
 				/*
 				 * We can gracefully recover from most read
@@ -387,9 +387,9 @@ xfs_trans_read_buf(
 	}
 	if (bp->b_error) {
 		error = bp->b_error;
-		XFS_BUF_SUPER_STALE(bp);
-		xfs_ioerror_alert("xfs_trans_read_buf", mp,
-				  bp, blkno);
+		xfs_buf_stale(bp);
+		XFS_BUF_DONE(bp);
+		xfs_buf_ioerror_alert(bp, __func__);
 		if (tp->t_flags & XFS_TRANS_DIRTY)
 			xfs_force_shutdown(tp->t_mountp, SHUTDOWN_META_IO_ERROR);
 		xfs_buf_relse(bp);
@@ -643,13 +643,14 @@ xfs_trans_log_buf(xfs_trans_t	*tp,
 	 * inside the b_bdstrat callback so that this won't get written to
 	 * disk.
 	 */
-	XFS_BUF_DELAYWRITE(bp);
 	XFS_BUF_DONE(bp);
 
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 	bp->b_iodone = xfs_buf_iodone_callbacks;
 	bip->bli_item.li_cb = xfs_buf_iodone;
 
+	xfs_buf_delwri_queue(bp);
+
 	trace_xfs_trans_log_buf(bip);
 
 	/*
@@ -738,8 +739,7 @@ xfs_trans_binval(
 	 * We set the stale bit in the buffer as well since we're getting
 	 * rid of it.
 	 */
-	XFS_BUF_UNDELAYWRITE(bp);
-	XFS_BUF_STALE(bp);
+	xfs_buf_stale(bp);
 	bip->bli_flags |= XFS_BLI_STALE;
 	bip->bli_flags &= ~(XFS_BLI_INODE_BUF | XFS_BLI_LOGGED | XFS_BLI_DIRTY);
 	bip->bli_format.blf_flags &= ~XFS_BLF_INODE_BUF;

fs/xfs/xfs_trans_inode.c

@@ -47,11 +47,13 @@ xfs_trans_inode_broot_debug(
  * Add a locked inode to the transaction.
  *
  * The inode must be locked, and it cannot be associated with any transaction.
+ * If lock_flags is non-zero the inode will be unlocked on transaction commit.
  */
 void
 xfs_trans_ijoin(
 	struct xfs_trans	*tp,
-	struct xfs_inode	*ip)
+	struct xfs_inode	*ip,
+	uint			lock_flags)
 {
 	xfs_inode_log_item_t	*iip;
 
@@ -59,7 +61,9 @@ xfs_trans_ijoin(
 	if (ip->i_itemp == NULL)
 		xfs_inode_item_init(ip, ip->i_mount);
 	iip = ip->i_itemp;
+
 	ASSERT(iip->ili_lock_flags == 0);
+	iip->ili_lock_flags = lock_flags;
 
 	/*
 	 * Get a log_item_desc to point at the new item.
@@ -69,25 +73,6 @@ xfs_trans_ijoin(
 	xfs_trans_inode_broot_debug(ip);
 }
 
-/*
- * Add a locked inode to the transaction.
- *
- *
- * Grabs a reference to the inode which will be dropped when the transaction
- * is committed.  The inode will also be unlocked at that point.  The inode
- * must be locked, and it cannot be associated with any transaction.
- */
-void
-xfs_trans_ijoin_ref(
-	struct xfs_trans	*tp,
-	struct xfs_inode	*ip,
-	uint			lock_flags)
-{
-	xfs_trans_ijoin(tp, ip);
-	IHOLD(ip);
-	ip->i_itemp->ili_lock_flags = lock_flags;
-}
-
 /*
  * Transactional inode timestamp update. Requires the inode to be locked and
  * joined to the transaction supplied. Relies on the transaction subsystem to

fs/xfs/xfs_trans_priv.h

@@ -70,6 +70,7 @@ struct xfs_ail {
 	struct list_head	xa_cursors;
 	spinlock_t		xa_lock;
 	xfs_lsn_t		xa_last_pushed_lsn;
+	int			xa_log_flush;
 };
 
 /*

fs/xfs/xfs_vnodeops.c

@@ -72,8 +72,8 @@ xfs_readlink_bmap(
 	xfs_buf_t	*bp;
 	int		error = 0;
 
-	error = xfs_bmapi(NULL, ip, 0, XFS_B_TO_FSB(mp, pathlen), 0, NULL, 0,
-			mval, &nmaps, NULL);
+	error = xfs_bmapi_read(ip, 0, XFS_B_TO_FSB(mp, pathlen), mval, &nmaps,
+			       0);
 	if (error)
 		goto out;
 
@@ -87,8 +87,7 @@ xfs_readlink_bmap(
 			return XFS_ERROR(ENOMEM);
 		error = bp->b_error;
 		if (error) {
-			xfs_ioerror_alert("xfs_readlink",
-				  ip->i_mount, bp, XFS_BUF_ADDR(bp));
+			xfs_buf_ioerror_alert(bp, __func__);
 			xfs_buf_relse(bp);
 			goto out;
 		}
@@ -178,8 +177,7 @@ xfs_free_eofblocks(
 
 	nimaps = 1;
 	xfs_ilock(ip, XFS_ILOCK_SHARED);
-	error = xfs_bmapi(NULL, ip, end_fsb, map_len, 0,
-			  NULL, 0, &imap, &nimaps, NULL);
+	error = xfs_bmapi_read(ip, end_fsb, map_len, &imap, &nimaps, 0);
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
 
 	if (!error && (nimaps != 0) &&
@@ -220,7 +218,7 @@ xfs_free_eofblocks(
 		}
 
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 
 		error = xfs_itruncate_data(&tp, ip, ip->i_size);
 		if (error) {
@@ -289,7 +287,7 @@ xfs_inactive_symlink_rmt(
 	xfs_ilock(ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL);
 	size = (int)ip->i_d.di_size;
 	ip->i_d.di_size = 0;
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, 0);
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 	/*
 	 * Find the block(s) so we can inval and unmap them.
@@ -297,9 +295,9 @@ xfs_inactive_symlink_rmt(
 	done = 0;
 	xfs_bmap_init(&free_list, &first_block);
 	nmaps = ARRAY_SIZE(mval);
-	if ((error = xfs_bmapi(tp, ip, 0, XFS_B_TO_FSB(mp, size),
-			XFS_BMAPI_METADATA, &first_block, 0, mval, &nmaps,
-			&free_list)))
+	error = xfs_bmapi_read(ip, 0, XFS_B_TO_FSB(mp, size),
+				mval, &nmaps, 0);
+	if (error)
 		goto error0;
 	/*
 	 * Invalidate the block(s).
@@ -308,6 +306,10 @@ xfs_inactive_symlink_rmt(
 		bp = xfs_trans_get_buf(tp, mp->m_ddev_targp,
 			XFS_FSB_TO_DADDR(mp, mval[i].br_startblock),
 			XFS_FSB_TO_BB(mp, mval[i].br_blockcount), 0);
+		if (!bp) {
+			error = ENOMEM;
+			goto error1;
+		}
 		xfs_trans_binval(tp, bp);
 	}
 	/*
@@ -333,7 +335,7 @@ xfs_inactive_symlink_rmt(
 	 * Mark it dirty so it will be logged and moved forward in the log as
 	 * part of every commit.
 	 */
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, 0);
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 	/*
 	 * Get a new, empty transaction to return to our caller.
@@ -466,7 +468,7 @@ xfs_inactive_attrs(
 		goto error_cancel;
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, 0);
 	xfs_idestroy_fork(ip, XFS_ATTR_FORK);
 
 	ASSERT(ip->i_d.di_anextents == 0);
@@ -647,8 +649,6 @@ xfs_inactive(
 	if (truncate) {
 		xfs_ilock(ip, XFS_IOLOCK_EXCL);
 
-		xfs_ioend_wait(ip);
-
 		error = xfs_trans_reserve(tp, 0,
 					  XFS_ITRUNCATE_LOG_RES(mp),
 					  0, XFS_TRANS_PERM_LOG_RES,
@@ -662,7 +662,7 @@ xfs_inactive(
 		}
 
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 
 		error = xfs_itruncate_data(&tp, ip, 0);
 		if (error) {
@@ -686,7 +686,7 @@ xfs_inactive(
 			return VN_INACTIVE_CACHE;
 		}
 
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 	} else {
 		error = xfs_trans_reserve(tp, 0,
 					  XFS_IFREE_LOG_RES(mp),
@@ -699,7 +699,7 @@ xfs_inactive(
 		}
 
 		xfs_ilock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 	}
 
 	/*
@@ -939,7 +939,7 @@ xfs_create(
 	 * the transaction cancel unlocking dp so don't do it explicitly in the
 	 * error path.
 	 */
-	xfs_trans_ijoin_ref(tp, dp, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
 	unlock_dp_on_error = B_FALSE;
 
 	error = xfs_dir_createname(tp, dp, name, ip->i_ino,
@@ -1260,8 +1260,8 @@ xfs_remove(
 
 	xfs_lock_two_inodes(dp, ip, XFS_ILOCK_EXCL);
 
-	xfs_trans_ijoin_ref(tp, dp, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 
 	/*
 	 * If we're removing a directory perform some additional validation.
@@ -1406,8 +1406,8 @@ xfs_link(
 
 	xfs_lock_two_inodes(sip, tdp, XFS_ILOCK_EXCL);
 
-	xfs_trans_ijoin_ref(tp, sip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin_ref(tp, tdp, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, sip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, tdp, XFS_ILOCK_EXCL);
 
 	/*
 	 * If the source has too many links, we can't make any more to it.
@@ -1601,7 +1601,7 @@ xfs_symlink(
 	 * transaction cancel unlocking dp so don't do it explicitly in the
 	 * error path.
 	 */
-	xfs_trans_ijoin_ref(tp, dp, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
 	unlock_dp_on_error = B_FALSE;
 
 	/*
@@ -1632,10 +1632,9 @@ xfs_symlink(
 		first_fsb = 0;
 		nmaps = SYMLINK_MAPS;
 
-		error = xfs_bmapi(tp, ip, first_fsb, fs_blocks,
-				  XFS_BMAPI_WRITE | XFS_BMAPI_METADATA,
-				  &first_block, resblks, mval, &nmaps,
-				  &free_list);
+		error = xfs_bmapi_write(tp, ip, first_fsb, fs_blocks,
+				  XFS_BMAPI_METADATA, &first_block, resblks,
+				  mval, &nmaps, &free_list);
 		if (error)
 			goto error2;
 
@@ -1650,7 +1649,10 @@ xfs_symlink(
 			byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
 			bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
 					       BTOBB(byte_cnt), 0);
-			ASSERT(!xfs_buf_geterror(bp));
+			if (!bp) {
+				error = ENOMEM;
+				goto error2;
+			}
 			if (pathlen < byte_cnt) {
 				byte_cnt = pathlen;
 			}
@@ -1732,7 +1734,7 @@ xfs_set_dmattrs(
 		return error;
 	}
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 
 	ip->i_d.di_dmevmask = evmask;
 	ip->i_d.di_dmstate  = state;
@@ -1778,7 +1780,6 @@ xfs_alloc_file_space(
 	xfs_fileoff_t		startoffset_fsb;
 	xfs_fsblock_t		firstfsb;
 	int			nimaps;
-	int			bmapi_flag;
 	int			quota_flag;
 	int			rt;
 	xfs_trans_t		*tp;
@@ -1806,7 +1807,6 @@ xfs_alloc_file_space(
 	count = len;
 	imapp = &imaps[0];
 	nimaps = 1;
-	bmapi_flag = XFS_BMAPI_WRITE | alloc_type;
 	startoffset_fsb	= XFS_B_TO_FSBT(mp, offset);
 	allocatesize_fsb = XFS_B_TO_FSB(mp, count);
 
@@ -1877,16 +1877,12 @@ xfs_alloc_file_space(
 		if (error)
 			goto error1;
 
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 
-		/*
-		 * Issue the xfs_bmapi() call to allocate the blocks
-		 */
 		xfs_bmap_init(&free_list, &firstfsb);
-		error = xfs_bmapi(tp, ip, startoffset_fsb,
-				  allocatesize_fsb, bmapi_flag,
-				  &firstfsb, 0, imapp, &nimaps,
-				  &free_list);
+		error = xfs_bmapi_write(tp, ip, startoffset_fsb,
+					allocatesize_fsb, alloc_type, &firstfsb,
+					0, imapp, &nimaps, &free_list);
 		if (error) {
 			goto error0;
 		}
@@ -1976,8 +1972,7 @@ xfs_zero_remaining_bytes(
 	for (offset = startoff; offset <= endoff; offset = lastoffset + 1) {
 		offset_fsb = XFS_B_TO_FSBT(mp, offset);
 		nimap = 1;
-		error = xfs_bmapi(NULL, ip, offset_fsb, 1, 0,
-			NULL, 0, &imap, &nimap, NULL);
+		error = xfs_bmapi_read(ip, offset_fsb, 1, &imap, &nimap, 0);
 		if (error || nimap < 1)
 			break;
 		ASSERT(imap.br_blockcount >= 1);
@@ -1997,8 +1992,8 @@ xfs_zero_remaining_bytes(
 		xfsbdstrat(mp, bp);
 		error = xfs_buf_iowait(bp);
 		if (error) {
-			xfs_ioerror_alert("xfs_zero_remaining_bytes(read)",
-					  mp, bp, XFS_BUF_ADDR(bp));
+			xfs_buf_ioerror_alert(bp,
+					"xfs_zero_remaining_bytes(read)");
 			break;
 		}
 		memset(bp->b_addr +
@@ -2010,8 +2005,8 @@ xfs_zero_remaining_bytes(
 		xfsbdstrat(mp, bp);
 		error = xfs_buf_iowait(bp);
 		if (error) {
-			xfs_ioerror_alert("xfs_zero_remaining_bytes(write)",
-					  mp, bp, XFS_BUF_ADDR(bp));
+			xfs_buf_ioerror_alert(bp,
+					"xfs_zero_remaining_bytes(write)");
 			break;
 		}
 	}
@@ -2076,7 +2071,7 @@ xfs_free_file_space(
 	if (need_iolock) {
 		xfs_ilock(ip, XFS_IOLOCK_EXCL);
 		/* wait for the completion of any pending DIOs */
-		xfs_ioend_wait(ip);
+		inode_dio_wait(VFS_I(ip));
 	}
 
 	rounding = max_t(uint, 1 << mp->m_sb.sb_blocklog, PAGE_CACHE_SIZE);
@@ -2096,8 +2091,8 @@ xfs_free_file_space(
 	 */
 	if (rt && !xfs_sb_version_hasextflgbit(&mp->m_sb)) {
 		nimap = 1;
-		error = xfs_bmapi(NULL, ip, startoffset_fsb,
-			1, 0, NULL, 0, &imap, &nimap, NULL);
+		error = xfs_bmapi_read(ip, startoffset_fsb, 1,
+					&imap, &nimap, 0);
 		if (error)
 			goto out_unlock_iolock;
 		ASSERT(nimap == 0 || nimap == 1);
@@ -2111,8 +2106,8 @@ xfs_free_file_space(
 				startoffset_fsb += mp->m_sb.sb_rextsize - mod;
 		}
 		nimap = 1;
-		error = xfs_bmapi(NULL, ip, endoffset_fsb - 1,
-			1, 0, NULL, 0, &imap, &nimap, NULL);
+		error = xfs_bmapi_read(ip, endoffset_fsb - 1, 1,
+					&imap, &nimap, 0);
 		if (error)
 			goto out_unlock_iolock;
 		ASSERT(nimap == 0 || nimap == 1);
@@ -2180,7 +2175,7 @@ xfs_free_file_space(
 		if (error)
 			goto error1;
 
-		xfs_trans_ijoin(tp, ip);
+		xfs_trans_ijoin(tp, ip, 0);
 
 		/*
 		 * issue the bunmapi() call to free the blocks
@@ -2353,8 +2348,7 @@ xfs_change_file_space(
 	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-
-	xfs_trans_ijoin(tp, ip);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 
 	if ((attr_flags & XFS_ATTR_DMI) == 0) {
 		ip->i_d.di_mode &= ~S_ISUID;
@@ -2379,10 +2373,5 @@ xfs_change_file_space(
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 	if (attr_flags & XFS_ATTR_SYNC)
 		xfs_trans_set_sync(tp);
-
-	error = xfs_trans_commit(tp, 0);
-
-	xfs_iunlock(ip, XFS_ILOCK_EXCL);
-
-	return error;
+	return xfs_trans_commit(tp, 0);
 }