Merge tag 'for-linus-v3.9-rc1' of git://oss.sgi.com/xfs/xfs

Pull xfs update from Ben Myers:
 "Primarily bugfixes and a few cleanups:

   - fix(es) for compound buffers

   - remove unused XFS_TRANS_DEBUG routines

   - fix for dquot soft timer asserts due to overflow of d_blk_softlimit

   - don't zero allocation args structure members after they are memset(0)

   - fix for regression in dir v2 code introduced in commit 20f7e9f3

   - remove obsolete simple_strto<foo>

   - fix return value when filesystem probe finds no XFS magic, a
     regression introduced in 9802182.

   - remove boolean_t typedef completely

   - fix stack switch in __xfs_bmapi_allocate by moving the check for
     stack switch up into xfs_bmapi_write.

   - fix build error due to incomplete boolean_t removal

   - fix oops in _xfs_buf_find by validating that the requested block is
     within the filesystem bounds.

   - limit speculative preallocation near ENOSPC.

   - fix an unmount hang in xfs_wait_buftarg by freeing the
     xfs_buf_log_item in xfs_buf_item_unlock.

   - fix a possible use after free with AIO.

   - fix xfs_swap_extents after removal of xfs_flushinval_pages, a
     regression introduced in fb59581404a.

   - replace hardcoded 128 with log header size

   - add memory barrier before wake_up_bit in xfs_ifunlock

   - limit speculative preallocation on sparse files

   - fix xa_lock recursion bug introduced in 90810b9e82a3

   - fix write verifier for symlinks"

Fixed up conflicts in fs/xfs/xfs_buf_item.c (due to bli_format rename in
commit 0f22f9d0cd8a affecting the removed XFS_TRANS_DEBUG routines in
commit ec47eb6b0b45).

* tag 'for-linus-v3.9-rc1' of git://oss.sgi.com/xfs/xfs: (36 commits)
  xfs: xfs_bmap_add_attrfork_local is too generic
  xfs: remove log force from xfs_buf_trylock()
  xfs: recheck buffer pinned status after push trylock failure
  xfs: limit speculative prealloc size on sparse files
  xfs: memory barrier before wake_up_bit()
  xfs: refactor space log reservation for XFS_TRANS_ATTR_SET
  xfs: make use of XFS_SB_LOG_RES() at xfs_fs_log_dummy()
  xfs: make use of XFS_SB_LOG_RES() at xfs_mount_log_sb()
  xfs: make use of XFS_SB_LOG_RES() at xfs_log_sbcount()
  xfs: introduce XFS_SB_LOG_RES() for transactions that modify sb on disk
  xfs: calculate XFS_TRANS_QM_QUOTAOFF_END space log reservation at mount time
  xfs: calculate XFS_TRANS_QM_QUOTAOFF space log reservation at mount time
  xfs: calculate XFS_TRANS_QM_DQALLOC space log reservation at mount time
  xfs: calcuate XFS_TRANS_QM_SETQLIM space log reservation at mount time
  xfs: calculate xfs_qm_write_sb_changes() space log reservation at mount time
  xfs: calculate XFS_TRANS_QM_SBCHANGE space log reservation at mount time
  xfs: make use of xfs_calc_buf_res() in xfs_trans.c
  xfs: add a helper to figure out the space log reservation per item
  xfs: Fix xfs_swap_extents() after removal of xfs_flushinval_pages()
  xfs: Fix possible use-after-free with AIO
  ...

28 files changed, 503 insertions, 453 deletions

diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c
index 393055fe3ae..0ad23253e8b 100644
--- a/fs/xfs/xfs_alloc.c
+++ b/fs/xfs/xfs_alloc.c
@@ -1925,8 +1925,6 @@ xfs_alloc_fix_freelist(
 	targs.mp = mp;
 	targs.agbp = agbp;
 	targs.agno = args->agno;
-	targs.mod = targs.minleft = targs.wasdel = targs.userdata =
-		targs.minalignslop = 0;
 	targs.alignment = targs.minlen = targs.prod = targs.isfl = 1;
 	targs.type = XFS_ALLOCTYPE_THIS_AG;
 	targs.pag = pag;
diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c
index aaf472532b3..888683844d9 100644
--- a/fs/xfs/xfs_attr.c
+++ b/fs/xfs/xfs_attr.c
@@ -300,9 +300,12 @@ xfs_attr_set_int(
 	if (rsvd)
 		args.trans->t_flags |= XFS_TRANS_RESERVE;
 
-	if ((error = xfs_trans_reserve(args.trans, args.total,
-			XFS_ATTRSET_LOG_RES(mp, args.total), 0,
-			XFS_TRANS_PERM_LOG_RES, XFS_ATTRSET_LOG_COUNT))) {
+	error = xfs_trans_reserve(args.trans, args.total,
+				  XFS_ATTRSETM_LOG_RES(mp) +
+				  XFS_ATTRSETRT_LOG_RES(mp) * args.total,
+				  0, XFS_TRANS_PERM_LOG_RES,
+				  XFS_ATTRSET_LOG_COUNT);
+	if (error) {
 		xfs_trans_cancel(args.trans, 0);
 		return(error);
 	}
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index cdb2d334858..b44af9211bd 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -147,7 +147,10 @@ xfs_bmap_local_to_extents(
 	xfs_fsblock_t	*firstblock,	/* first block allocated in xaction */
 	xfs_extlen_t	total,		/* total blocks needed by transaction */
 	int		*logflagsp,	/* inode logging flags */
-	int		whichfork);	/* data or attr fork */
+	int		whichfork,	/* data or attr fork */
+	void		(*init_fn)(struct xfs_buf *bp,
+				   struct xfs_inode *ip,
+				   struct xfs_ifork *ifp));
 
 /*
  * Search the extents list for the inode, for the extent containing bno.
@@ -357,7 +360,42 @@ xfs_bmap_add_attrfork_extents(
 }
 
 /*
- * Called from xfs_bmap_add_attrfork to handle local format files.
+ * Block initialisation functions for local to extent format conversion.
+ * As these get more complex, they will be moved to the relevant files,
+ * but for now they are too simple to worry about.
+ */
+STATIC void
+xfs_bmap_local_to_extents_init_fn(
+	struct xfs_buf		*bp,
+	struct xfs_inode	*ip,
+	struct xfs_ifork	*ifp)
+{
+	bp->b_ops = &xfs_bmbt_buf_ops;
+	memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+}
+
+STATIC void
+xfs_symlink_local_to_remote(
+	struct xfs_buf		*bp,
+	struct xfs_inode	*ip,
+	struct xfs_ifork	*ifp)
+{
+	/* remote symlink blocks are not verifiable until CRCs come along */
+	bp->b_ops = NULL;
+	memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+}
+
+/*
+ * Called from xfs_bmap_add_attrfork to handle local format files. Each
+ * different data fork content type needs a different callout to do the
+ * conversion. Some are basic and only require special block initialisation
+ * callouts for the data formating, others (directories) are so specialised they
+ * handle everything themselves.
+ *
+ * XXX (dgc): investigate whether directory conversion can use the generic
+ * formatting callout. It should be possible - it's just a very complex
+ * formatter. it would also require passing the transaction through to the init
+ * function.
  */
 STATIC int					/* error */
 xfs_bmap_add_attrfork_local(
@@ -368,25 +406,29 @@ xfs_bmap_add_attrfork_local(
 	int			*flags)		/* inode logging flags */
 {
 	xfs_da_args_t		dargs;		/* args for dir/attr code */
-	int			error;		/* error return value */
-	xfs_mount_t		*mp;		/* mount structure pointer */
 
 	if (ip->i_df.if_bytes <= XFS_IFORK_DSIZE(ip))
 		return 0;
+
 	if (S_ISDIR(ip->i_d.di_mode)) {
-		mp = ip->i_mount;
 		memset(&dargs, 0, sizeof(dargs));
 		dargs.dp = ip;
 		dargs.firstblock = firstblock;
 		dargs.flist = flist;
-		dargs.total = mp->m_dirblkfsbs;
+		dargs.total = ip->i_mount->m_dirblkfsbs;
 		dargs.whichfork = XFS_DATA_FORK;
 		dargs.trans = tp;
-		error = xfs_dir2_sf_to_block(&dargs);
-	} else
-		error = xfs_bmap_local_to_extents(tp, ip, firstblock, 1, flags,
-			XFS_DATA_FORK);
-	return error;
+		return xfs_dir2_sf_to_block(&dargs);
+	}
+
+	if (S_ISLNK(ip->i_d.di_mode))
+		return xfs_bmap_local_to_extents(tp, ip, firstblock, 1,
+						 flags, XFS_DATA_FORK,
+						 xfs_symlink_local_to_remote);
+
+	return xfs_bmap_local_to_extents(tp, ip, firstblock, 1, flags,
+					 XFS_DATA_FORK,
+					 xfs_bmap_local_to_extents_init_fn);
 }
 
 /*
@@ -3099,8 +3141,6 @@ xfs_bmap_extents_to_btree(
 		args.fsbno = *firstblock;
 	}
 	args.minlen = args.maxlen = args.prod = 1;
-	args.total = args.minleft = args.alignment = args.mod = args.isfl =
-		args.minalignslop = 0;
 	args.wasdel = wasdel;
 	*logflagsp = 0;
 	if ((error = xfs_alloc_vextent(&args))) {
@@ -3221,7 +3261,10 @@ xfs_bmap_local_to_extents(
 	xfs_fsblock_t	*firstblock,	/* first block allocated in xaction */
 	xfs_extlen_t	total,		/* total blocks needed by transaction */
 	int		*logflagsp,	/* inode logging flags */
-	int		whichfork)	/* data or attr fork */
+	int		whichfork,
+	void		(*init_fn)(struct xfs_buf *bp,
+				   struct xfs_inode *ip,
+				   struct xfs_ifork *ifp))
 {
 	int		error;		/* error return value */
 	int		flags;		/* logging flags returned */
@@ -3241,12 +3284,12 @@ xfs_bmap_local_to_extents(
 		xfs_buf_t	*bp;	/* buffer for extent block */
 		xfs_bmbt_rec_host_t *ep;/* extent record pointer */
 
+		ASSERT((ifp->if_flags &
+			(XFS_IFINLINE|XFS_IFEXTENTS|XFS_IFEXTIREC)) == XFS_IFINLINE);
 		memset(&args, 0, sizeof(args));
 		args.tp = tp;
 		args.mp = ip->i_mount;
 		args.firstblock = *firstblock;
-		ASSERT((ifp->if_flags &
-			(XFS_IFINLINE|XFS_IFEXTENTS|XFS_IFEXTIREC)) == XFS_IFINLINE);
 		/*
 		 * Allocate a block.  We know we need only one, since the
 		 * file currently fits in an inode.
@@ -3259,20 +3302,21 @@ xfs_bmap_local_to_extents(
 			args.type = XFS_ALLOCTYPE_NEAR_BNO;
 		}
 		args.total = total;
-		args.mod = args.minleft = args.alignment = args.wasdel =
-			args.isfl = args.minalignslop = 0;
 		args.minlen = args.maxlen = args.prod = 1;
-		if ((error = xfs_alloc_vextent(&args)))
+		error = xfs_alloc_vextent(&args);
+		if (error)
 			goto done;
-		/*
-		 * Can't fail, the space was reserved.
-		 */
+
+		/* Can't fail, the space was reserved. */
 		ASSERT(args.fsbno != NULLFSBLOCK);
 		ASSERT(args.len == 1);
 		*firstblock = args.fsbno;
 		bp = xfs_btree_get_bufl(args.mp, tp, args.fsbno, 0);
-		bp->b_ops = &xfs_bmbt_buf_ops;
-		memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+
+		/* initialise the block and copy the data */
+		init_fn(bp, ip, ifp);
+
+		/* account for the change in fork size and log everything */
 		xfs_trans_log_buf(tp, bp, 0, ifp->if_bytes - 1);
 		xfs_bmap_forkoff_reset(args.mp, ip, whichfork);
 		xfs_idata_realloc(ip, -ifp->if_bytes, whichfork);
@@ -4919,8 +4963,32 @@ xfs_bmapi_write(
 	XFS_STATS_INC(xs_blk_mapw);
 
 	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
+		/*
+		 * XXX (dgc): This assumes we are only called for inodes that
+		 * contain content neutral data in local format. Anything that
+		 * contains caller-specific data in local format that needs
+		 * transformation to move to a block format needs to do the
+		 * conversion to extent format itself.
+		 *
+		 * Directory data forks and attribute forks handle this
+		 * themselves, but with the addition of metadata verifiers every
+		 * data fork in local format now contains caller specific data
+		 * and as such conversion through this function is likely to be
+		 * broken.
+		 *
+		 * The only likely user of this branch is for remote symlinks,
+		 * but we cannot overwrite the data fork contents of the symlink
+		 * (EEXIST occurs higher up the stack) and so it will never go
+		 * from local format to extent format here. Hence I don't think
+		 * this branch is ever executed intentionally and we should
+		 * consider removing it and asserting that xfs_bmapi_write()
+		 * cannot be called directly on local format forks. i.e. callers
+		 * are completely responsible for local to extent format
+		 * conversion, not xfs_bmapi_write().
+		 */
 		error = xfs_bmap_local_to_extents(tp, ip, firstblock, total,
-						  &bma.logflags, whichfork);
+					&bma.logflags, whichfork,
+					xfs_bmap_local_to_extents_init_fn);
 		if (error)
 			goto error0;
 	}
diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c
index fbbb9eb92e3..4e8f0df82d0 100644
--- a/fs/xfs/xfs_buf.c
+++ b/fs/xfs/xfs_buf.c
@@ -951,8 +951,6 @@ xfs_buf_trylock(
 	locked = down_trylock(&bp->b_sema) == 0;
 	if (locked)
 		XB_SET_OWNER(bp);
-	else if (atomic_read(&bp->b_pin_count) && (bp->b_flags & XBF_STALE))
-		xfs_log_force(bp->b_target->bt_mount, 0);
 
 	trace_xfs_buf_trylock(bp, _RET_IP_);
 	return locked;
diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index 3f9949fee39..cf263476d6b 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -37,109 +37,6 @@ static inline struct xfs_buf_log_item *BUF_ITEM(struct xfs_log_item *lip)
 	return container_of(lip, struct xfs_buf_log_item, bli_item);
 }
 
-
-#ifdef XFS_TRANS_DEBUG
-/*
- * This function uses an alternate strategy for tracking the bytes
- * that the user requests to be logged.  This can then be used
- * in conjunction with the bli_orig array in the buf log item to
- * catch bugs in our callers' code.
- *
- * We also double check the bits set in xfs_buf_item_log using a
- * simple algorithm to check that every byte is accounted for.
- */
-STATIC void
-xfs_buf_item_log_debug(
-	xfs_buf_log_item_t	*bip,
-	uint			first,
-	uint			last)
-{
-	uint	x;
-	uint	byte;
-	uint	nbytes;
-	uint	chunk_num;
-	uint	word_num;
-	uint	bit_num;
-	uint	bit_set;
-	uint	*wordp;
-
-	ASSERT(bip->bli_logged != NULL);
-	byte = first;
-	nbytes = last - first + 1;
-	bfset(bip->bli_logged, first, nbytes);
-	for (x = 0; x < nbytes; x++) {
-		chunk_num = byte >> XFS_BLF_SHIFT;
-		word_num = chunk_num >> BIT_TO_WORD_SHIFT;
-		bit_num = chunk_num & (NBWORD - 1);
-		wordp = &(bip->__bli_format.blf_data_map[word_num]);
-		bit_set = *wordp & (1 << bit_num);
-		ASSERT(bit_set);
-		byte++;
-	}
-}
-
-/*
- * This function is called when we flush something into a buffer without
- * logging it.  This happens for things like inodes which are logged
- * separately from the buffer.
- */
-void
-xfs_buf_item_flush_log_debug(
-	xfs_buf_t	*bp,
-	uint		first,
-	uint		last)
-{
-	xfs_buf_log_item_t	*bip = bp->b_fspriv;
-	uint			nbytes;
-
-	if (bip == NULL || (bip->bli_item.li_type != XFS_LI_BUF))
-		return;
-
-	ASSERT(bip->bli_logged != NULL);
-	nbytes = last - first + 1;
-	bfset(bip->bli_logged, first, nbytes);
-}
-
-/*
- * This function is called to verify that our callers have logged
- * all the bytes that they changed.
- *
- * It does this by comparing the original copy of the buffer stored in
- * the buf log item's bli_orig array to the current copy of the buffer
- * and ensuring that all bytes which mismatch are set in the bli_logged
- * array of the buf log item.
- */
-STATIC void
-xfs_buf_item_log_check(
-	xfs_buf_log_item_t	*bip)
-{
-	char		*orig;
-	char		*buffer;
-	int		x;
-	xfs_buf_t	*bp;
-
-	ASSERT(bip->bli_orig != NULL);
-	ASSERT(bip->bli_logged != NULL);
-
-	bp = bip->bli_buf;
-	ASSERT(bp->b_length > 0);
-	ASSERT(bp->b_addr != NULL);
-	orig = bip->bli_orig;
-	buffer = bp->b_addr;
-	for (x = 0; x < BBTOB(bp->b_length); x++) {
-		if (orig[x] != buffer[x] && !btst(bip->bli_logged, x)) {
-			xfs_emerg(bp->b_mount,
-				"%s: bip %x buffer %x orig %x index %d",
-				__func__, bip, bp, orig, x);
-			ASSERT(0);
-		}
-	}
-}
-#else
-#define		xfs_buf_item_log_debug(x,y,z)
-#define		xfs_buf_item_log_check(x)
-#endif
-
 STATIC void	xfs_buf_do_callbacks(struct xfs_buf *bp);
 
 /*
@@ -429,7 +326,6 @@ xfs_buf_item_format(
 	 * Check to make sure everything is consistent.
 	 */
 	trace_xfs_buf_item_format(bip);
-	xfs_buf_item_log_check(bip);
 }
 
 /*
@@ -573,8 +469,18 @@ xfs_buf_item_push(
 
 	if (xfs_buf_ispinned(bp))
 		return XFS_ITEM_PINNED;
-	if (!xfs_buf_trylock(bp))
+	if (!xfs_buf_trylock(bp)) {
+		/*
+		 * If we have just raced with a buffer being pinned and it has
+		 * been marked stale, we could end up stalling until someone else
+		 * issues a log force to unpin the stale buffer. Check for the
+		 * race condition here so xfsaild recognizes the buffer is pinned
+		 * and queues a log force to move it along.
+		 */
+		if (xfs_buf_ispinned(bp))
+			return XFS_ITEM_PINNED;
 		return XFS_ITEM_LOCKED;
+	}
 
 	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
 
@@ -923,8 +829,6 @@ xfs_buf_item_log_segment(
 		mask = (1 << end_bit) - 1;
 		*wordp |= mask;
 	}
-
-	xfs_buf_item_log_debug(bip, first, last);
 }
 
 /*
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index 16def435944..ee36c88ecfd 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -98,10 +98,6 @@ typedef struct xfs_buf_log_item {
 	unsigned int		bli_flags;	/* misc flags */
 	unsigned int		bli_recur;	/* lock recursion count */
 	atomic_t		bli_refcount;	/* cnt of tp refs */
-#ifdef XFS_TRANS_DEBUG
-	char			*bli_orig;	/* original buffer copy */
-	char			*bli_logged;	/* bytes logged (bitmap) */
-#endif
 	int			bli_format_count;	/* count of headers */
 	struct xfs_buf_log_format *bli_formats;	/* array of in-log header ptrs */
 	struct xfs_buf_log_format __bli_format;	/* embedded in-log header */
@@ -117,16 +113,6 @@ void	xfs_buf_attach_iodone(struct xfs_buf *,
 void	xfs_buf_iodone_callbacks(struct xfs_buf *);
 void	xfs_buf_iodone(struct xfs_buf *, struct xfs_log_item *);
 
-#ifdef XFS_TRANS_DEBUG
-void
-xfs_buf_item_flush_log_debug(
-	struct xfs_buf *bp,
-	uint	first,
-	uint	last);
-#else
-#define	xfs_buf_item_flush_log_debug(bp, first, last)
-#endif
-
 #endif	/* __KERNEL__ */
 
 #endif	/* __XFS_BUF_ITEM_H__ */
diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c
index 9e1bf5294c9..8025eb23ad7 100644
--- a/fs/xfs/xfs_dquot.c
+++ b/fs/xfs/xfs_dquot.c
@@ -612,15 +612,9 @@ xfs_qm_dqread(
 	if (flags & XFS_QMOPT_DQALLOC) {
 		tp = xfs_trans_alloc(mp, XFS_TRANS_QM_DQALLOC);
 		error = xfs_trans_reserve(tp, XFS_QM_DQALLOC_SPACE_RES(mp),
-				XFS_WRITE_LOG_RES(mp) +
-				/*
-				 * Round the chunklen up to the next multiple
-				 * of 128 (buf log item chunk size)).
-				 */
-				BBTOB(mp->m_quotainfo->qi_dqchunklen) - 1 + 128,
-				0,
-				XFS_TRANS_PERM_LOG_RES,
-				XFS_WRITE_LOG_COUNT);
+					  XFS_QM_DQALLOC_LOG_RES(mp), 0,
+					  XFS_TRANS_PERM_LOG_RES,
+					  XFS_WRITE_LOG_COUNT);
 		if (error)
 			goto error1;
 		cancelflags = XFS_TRANS_RELEASE_LOG_RES;
diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index 94eaeedc549..2866b8c78b7 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -709,8 +709,8 @@ xfs_fs_log_dummy(
 	int		error;
 
 	tp = _xfs_trans_alloc(mp, XFS_TRANS_DUMMY1, KM_SLEEP);
-	error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0,
-					XFS_DEFAULT_LOG_COUNT);
+	error = xfs_trans_reserve(tp, 0, XFS_SB_LOG_RES(mp), 0, 0,
+				  XFS_DEFAULT_LOG_COUNT);
 	if (error) {
 		xfs_trans_cancel(tp, 0);
 		return error;
diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c
index a815412eab8..515bf71ce01 100644
--- a/fs/xfs/xfs_ialloc.c
+++ b/fs/xfs/xfs_ialloc.c
@@ -279,8 +279,6 @@ xfs_ialloc_ag_alloc(
 		  (args.agbno < be32_to_cpu(agi->agi_length)))) {
 		args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno);
 		args.type = XFS_ALLOCTYPE_THIS_BNO;
-		args.mod = args.total = args.wasdel = args.isfl =
-			args.userdata = args.minalignslop = 0;
 		args.prod = 1;
 
 		/*
@@ -333,8 +331,6 @@ xfs_ialloc_ag_alloc(
 		 * Allocate a fixed-size extent of inodes.
 		 */
 		args.type = XFS_ALLOCTYPE_NEAR_BNO;
-		args.mod = args.total = args.wasdel = args.isfl =
-			args.userdata = args.minalignslop = 0;
 		args.prod = 1;
 		/*
 		 * Allow space for the inode btree to split.
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 66282dcb821..4f201656d2d 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -2379,9 +2379,6 @@ xfs_iflush_fork(
 	char			*cp;
 	xfs_ifork_t		*ifp;
 	xfs_mount_t		*mp;
-#ifdef XFS_TRANS_DEBUG
-	int			first;
-#endif
 	static const short	brootflag[2] =
 		{ XFS_ILOG_DBROOT, XFS_ILOG_ABROOT };
 	static const short	dataflag[2] =
@@ -2724,9 +2721,6 @@ xfs_iflush_int(
 	xfs_inode_log_item_t	*iip;
 	xfs_dinode_t		*dip;
 	xfs_mount_t		*mp;
-#ifdef XFS_TRANS_DEBUG
-	int			first;
-#endif
 
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED));
 	ASSERT(xfs_isiflocked(ip));
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 22baf6ea4fa..237e7f6f2ab 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -419,6 +419,7 @@ static inline void xfs_iflock(struct xfs_inode *ip)
 static inline void xfs_ifunlock(struct xfs_inode *ip)
 {
 	xfs_iflags_clear(ip, XFS_IFLOCK);
+	smp_mb();
 	wake_up_bit(&ip->i_flags, __XFS_IFLOCK_BIT);
 }
 
diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index d041d47d9d8..f034bd1652f 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -269,17 +269,6 @@ xfs_inode_item_format(
 		} else {
 			ASSERT(!(iip->ili_fields &
 				 XFS_ILOG_DBROOT));
-#ifdef XFS_TRANS_DEBUG
-			if (iip->ili_root_size > 0) {
-				ASSERT(iip->ili_root_size ==
-				       ip->i_df.if_broot_bytes);
-				ASSERT(memcmp(iip->ili_orig_root,
-					    ip->i_df.if_broot,
-					    iip->ili_root_size) == 0);
-			} else {
-				ASSERT(ip->i_df.if_broot_bytes == 0);
-			}
-#endif
 			iip->ili_fields &= ~XFS_ILOG_DBROOT;
 		}
 		break;
@@ -678,11 +667,6 @@ void
 xfs_inode_item_destroy(
 	xfs_inode_t	*ip)
 {
-#ifdef XFS_TRANS_DEBUG
-	if (ip->i_itemp->ili_root_size != 0) {
-		kmem_free(ip->i_itemp->ili_orig_root);
-	}
-#endif
 	kmem_zone_free(xfs_ili_zone, ip->i_itemp);
 }
 
diff --git a/fs/xfs/xfs_inode_item.h b/fs/xfs/xfs_inode_item.h
index 376d4d0b263..779812fb3d8 100644
--- a/fs/xfs/xfs_inode_item.h
+++ b/fs/xfs/xfs_inode_item.h
@@ -148,10 +148,6 @@ typedef struct xfs_inode_log_item {
 						      data exts */
 	struct xfs_bmbt_rec	*ili_aextents_buf; /* array of logged
 						      attr exts */
-#ifdef XFS_TRANS_DEBUG
-	int			ili_root_size;
-	char			*ili_orig_root;
-#endif
 	xfs_inode_log_format_t	ili_format;	   /* logged structure */
 } xfs_inode_log_item_t;
 
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 364818eef40..912d83d8860 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -311,6 +311,62 @@ xfs_iomap_eof_want_preallocate(
 }
 
 /*
+ * Determine the initial size of the preallocation. We are beyond the current
+ * EOF here, but we need to take into account whether this is a sparse write or
+ * an extending write when determining the preallocation size.  Hence we need to
+ * look up the extent that ends at the current write offset and use the result
+ * to determine the preallocation size.
+ *
+ * If the extent is a hole, then preallocation is essentially disabled.
+ * Otherwise we take the size of the preceeding data extent as the basis for the
+ * preallocation size. If the size of the extent is greater than half the
+ * maximum extent length, then use the current offset as the basis. This ensures
+ * that for large files the preallocation size always extends to MAXEXTLEN
+ * rather than falling short due to things like stripe unit/width alignment of
+ * real extents.
+ */
+STATIC int
+xfs_iomap_eof_prealloc_initial_size(
+	struct xfs_mount	*mp,
+	struct xfs_inode	*ip,
+	xfs_off_t		offset,
+	xfs_bmbt_irec_t		*imap,
+	int			nimaps)
+{
+	xfs_fileoff_t   start_fsb;
+	int		imaps = 1;
+	int		error;
+
+	ASSERT(nimaps >= imaps);
+
+	/* if we are using a specific prealloc size, return now */
+	if (mp->m_flags & XFS_MOUNT_DFLT_IOSIZE)
+		return 0;
+
+	/*
+	 * As we write multiple pages, the offset will always align to the
+	 * start of a page and hence point to a hole at EOF. i.e. if the size is
+	 * 4096 bytes, we only have one block at FSB 0, but XFS_B_TO_FSB(4096)
+	 * will return FSB 1. Hence if there are blocks in the file, we want to
+	 * point to the block prior to the EOF block and not the hole that maps
+	 * directly at @offset.
+	 */
+	start_fsb = XFS_B_TO_FSB(mp, offset);
+	if (start_fsb)
+		start_fsb--;
+	error = xfs_bmapi_read(ip, start_fsb, 1, imap, &imaps, XFS_BMAPI_ENTIRE);
+	if (error)
+		return 0;
+
+	ASSERT(imaps == 1);
+	if (imap[0].br_startblock == HOLESTARTBLOCK)
+		return 0;
+	if (imap[0].br_blockcount <= (MAXEXTLEN >> 1))
+		return imap[0].br_blockcount;
+	return XFS_B_TO_FSB(mp, offset);
+}
+
+/*
  * If we don't have a user specified preallocation size, dynamically increase
  * the preallocation size as the size of the file grows. Cap the maximum size
  * at a single extent or less if the filesystem is near full. The closer the
@@ -319,20 +375,19 @@ xfs_iomap_eof_want_preallocate(
 STATIC xfs_fsblock_t
 xfs_iomap_prealloc_size(
 	struct xfs_mount	*mp,
-	struct xfs_inode	*ip)
+	struct xfs_inode	*ip,
+	xfs_off_t		offset,
+	struct xfs_bmbt_irec	*imap,
+	int			nimaps)
 {
 	xfs_fsblock_t		alloc_blocks = 0;
 
-	if (!(mp->m_flags & XFS_MOUNT_DFLT_IOSIZE)) {
+	alloc_blocks = xfs_iomap_eof_prealloc_initial_size(mp, ip, offset,
+							   imap, nimaps);
+	if (alloc_blocks > 0) {
 		int shift = 0;
 		int64_t freesp;
 
-		/*
-		 * rounddown_pow_of_two() returns an undefined result
-		 * if we pass in alloc_blocks = 0. Hence the "+ 1" to
-		 * ensure we always pass in a non-zero value.
-		 */
-		alloc_blocks = XFS_B_TO_FSB(mp, XFS_ISIZE(ip)) + 1;
 		alloc_blocks = XFS_FILEOFF_MIN(MAXEXTLEN,
 					rounddown_pow_of_two(alloc_blocks));
 
@@ -399,7 +454,6 @@ xfs_iomap_write_delay(
 	extsz = xfs_get_extsz_hint(ip);
 	offset_fsb = XFS_B_TO_FSBT(mp, offset);
 
-
 	error = xfs_iomap_eof_want_preallocate(mp, ip, offset, count,
 				imap, XFS_WRITE_IMAPS, &prealloc);
 	if (error)
@@ -407,7 +461,10 @@ xfs_iomap_write_delay(
 
 retry:
 	if (prealloc) {
-		xfs_fsblock_t	alloc_blocks = xfs_iomap_prealloc_size(mp, ip);
+		xfs_fsblock_t	alloc_blocks;
+
+		alloc_blocks = xfs_iomap_prealloc_size(mp, ip, offset, imap,
+						       XFS_WRITE_IMAPS);
 
 		aligned_offset = XFS_WRITEIO_ALIGN(mp, (offset + count - 1));
 		ioalign = XFS_B_TO_FSBT(mp, aligned_offset);
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 46bd9d52ab5..eec226f78a4 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -120,7 +120,7 @@ xlog_verify_iclog(
 	struct xlog		*log,
 	struct xlog_in_core	*iclog,
 	int			count,
-	boolean_t		syncing);
+	bool                    syncing);
 STATIC void
 xlog_verify_tail_lsn(
 	struct xlog		*log,
@@ -1737,7 +1737,7 @@ xlog_sync(
 	ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1);
 	ASSERT(XFS_BUF_ADDR(bp) + BTOBB(count) <= log->l_logBBsize);
 
-	xlog_verify_iclog(log, iclog, count, B_TRUE);
+	xlog_verify_iclog(log, iclog, count, true);
 
 	/* account for log which doesn't start at block #0 */
 	XFS_BUF_SET_ADDR(bp, XFS_BUF_ADDR(bp) + log->l_logBBstart);
@@ -3611,7 +3611,7 @@ xlog_verify_iclog(
 	struct xlog		*log,
 	struct xlog_in_core	*iclog,
 	int			count,
-	boolean_t		syncing)
+	bool                    syncing)
 {
 	xlog_op_header_t	*ophead;
 	xlog_in_core_t		*icptr;
@@ -3659,7 +3659,7 @@ xlog_verify_iclog(
 		/* clientid is only 1 byte */
 		field_offset = (__psint_t)
 			       ((xfs_caddr_t)&(ophead->oh_clientid) - base_ptr);
-		if (syncing == B_FALSE || (field_offset & 0x1ff)) {
+		if (!syncing || (field_offset & 0x1ff)) {
 			clientid = ophead->oh_clientid;
 		} else {
 			idx = BTOBBT((xfs_caddr_t)&(ophead->oh_clientid) - iclog->ic_datap);
@@ -3682,7 +3682,7 @@ xlog_verify_iclog(
 		/* check length */
 		field_offset = (__psint_t)
 			       ((xfs_caddr_t)&(ophead->oh_len) - base_ptr);
-		if (syncing == B_FALSE || (field_offset & 0x1ff)) {
+		if (!syncing || (field_offset & 0x1ff)) {
 			op_len = be32_to_cpu(ophead->oh_len);
 		} else {
 			idx = BTOBBT((__psint_t)&ophead->oh_len -
diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index 7d6df7c00c3..3806088a8f7 100644
--- a/fs/xfs/xfs_mount.c
+++ b/fs/xfs/xfs_mount.c
@@ -1109,8 +1109,8 @@ xfs_mount_reset_sbqflags(
 		return 0;
 
 	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SBCHANGE);
-	error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0,
-				      XFS_DEFAULT_LOG_COUNT);
+	error = xfs_trans_reserve(tp, 0, XFS_QM_SBCHANGE_LOG_RES(mp),
+				  0, 0, XFS_DEFAULT_LOG_COUNT);
 	if (error) {
 		xfs_trans_cancel(tp, 0);
 		xfs_alert(mp, "%s: Superblock update failed!", __func__);
@@ -1583,8 +1583,8 @@ xfs_log_sbcount(xfs_mount_t *mp)
 		return 0;
 
 	tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_COUNT, KM_SLEEP);
-	error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0,
-					XFS_DEFAULT_LOG_COUNT);
+	error = xfs_trans_reserve(tp, 0, XFS_SB_LOG_RES(mp), 0, 0,
+				  XFS_DEFAULT_LOG_COUNT);
 	if (error) {
 		xfs_trans_cancel(tp, 0);
 		return error;
@@ -1945,8 +1945,8 @@ xfs_mount_log_sb(
 			 XFS_SB_VERSIONNUM));
 
 	tp = xfs_trans_alloc(mp, XFS_TRANS_SB_UNIT);
-	error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0,
-				XFS_DEFAULT_LOG_COUNT);
+	error = xfs_trans_reserve(tp, 0, XFS_SB_LOG_RES(mp), 0, 0,
+				  XFS_DEFAULT_LOG_COUNT);
 	if (error) {
 		xfs_trans_cancel(tp, 0);
 		return error;
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index bab8314507e..bc907061d39 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -34,12 +34,19 @@ type