Merge tag 'for-linus-v3.11-rc1' of git://oss.sgi.com/xfs/xfs

Pull xfs update from Ben Myers:
 "This includes several bugfixes, part of the work for project quotas
  and group quotas to be used together, performance improvements for
  inode creation/deletion, buffer readahead, and bulkstat,
  implementation of the inode change count, an inode create transaction,
  and the removal of a bunch of dead code.

  There are also some duplicate commits that you already have from the
  3.10-rc series.

   - part of the work to allow project quotas and group quotas to be
     used together
   - inode change count
   - inode create transaction
   - block queue plugging in buffer readahead and bulkstat
   - ordered log vector support
   - removal of dead code in and around xfs_sync_inode_grab,
     xfs_ialloc_get_rec, XFS_MOUNT_RETERR, XFS_ALLOCFREE_LOG_RES,
     XFS_DIROP_LOG_RES, xfs_chash, ctl_table, and
     xfs_growfs_data_private
   - don't keep silent if sunit/swidth can not be changed via mount
   - fix a leak of remote symlink blocks into the filesystem when xattrs
     are used on symlinks
   - fix for fiemap to return FIEMAP_EXTENT_UNKOWN flag on delay extents
   - part of a fix for xfs_fsr
   - disable speculative preallocation with small files
   - performance improvements for inode creates and deletes"

* tag 'for-linus-v3.11-rc1' of git://oss.sgi.com/xfs/xfs: (61 commits)
  xfs: Remove incore use of XFS_OQUOTA_ENFD and XFS_OQUOTA_CHKD
  xfs: Change xfs_dquot_acct to be a 2-dimensional array
  xfs: Code cleanup and removal of some typedef usage
  xfs: Replace macro XFS_DQ_TO_QIP with a function
  xfs: Replace macro XFS_DQUOT_TREE with a function
  xfs: Define a new function xfs_is_quota_inode()
  xfs: implement inode change count
  xfs: Use inode create transaction
  xfs: Inode create item recovery
  xfs: Inode create transaction reservations
  xfs: Inode create log items
  xfs: Introduce an ordered buffer item
  xfs: Introduce ordered log vector support
  xfs: xfs_ifree doesn't need to modify the inode buffer
  xfs: don't do IO when creating an new inode
  xfs: don't use speculative prealloc for small files
  xfs: plug directory buffer readahead
  xfs: add pluging for bulkstat readahead
  xfs: Remove dead function prototype xfs_sync_inode_grab()
  xfs: Remove the left function variable from xfs_ialloc_get_rec()
  ...

43 files changed, 1180 insertions, 477 deletions

diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 6313b69b6644..4a4508023a3c 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -71,6 +71,7 @@ xfs-y				+= xfs_alloc.o \
 				   xfs_dir2_sf.o \
 				   xfs_ialloc.o \
 				   xfs_ialloc_btree.o \
+				   xfs_icreate_item.o \
 				   xfs_inode.o \
 				   xfs_log_recover.o \
 				   xfs_mount.o \
diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c
index 5673bcfda2f0..71596e57283a 100644
--- a/fs/xfs/xfs_alloc.c
+++ b/fs/xfs/xfs_alloc.c
@@ -175,6 +175,7 @@ xfs_alloc_compute_diff(
 	xfs_agblock_t	wantbno,	/* target starting block */
 	xfs_extlen_t	wantlen,	/* target length */
 	xfs_extlen_t	alignment,	/* target alignment */
+	char		userdata,	/* are we allocating data? */
 	xfs_agblock_t	freebno,	/* freespace's starting block */
 	xfs_extlen_t	freelen,	/* freespace's length */
 	xfs_agblock_t	*newbnop)	/* result: best start block from free */
@@ -189,7 +190,14 @@ xfs_alloc_compute_diff(
 	ASSERT(freelen >= wantlen);
 	freeend = freebno + freelen;
 	wantend = wantbno + wantlen;
-	if (freebno >= wantbno) {
+	/*
+	 * We want to allocate from the start of a free extent if it is past
+	 * the desired block or if we are allocating user data and the free
+	 * extent is before desired block. The second case is there to allow
+	 * for contiguous allocation from the remaining free space if the file
+	 * grows in the short term.
+	 */
+	if (freebno >= wantbno || (userdata && freeend < wantend)) {
 		if ((newbno1 = roundup(freebno, alignment)) >= freeend)
 			newbno1 = NULLAGBLOCK;
 	} else if (freeend >= wantend && alignment > 1) {
@@ -805,7 +813,8 @@ xfs_alloc_find_best_extent(
 			xfs_alloc_fix_len(args);
 
 			sdiff = xfs_alloc_compute_diff(args->agbno, args->len,
-						       args->alignment, *sbnoa,
+						       args->alignment,
+						       args->userdata, *sbnoa,
 						       *slena, &new);
 
 			/*
@@ -976,7 +985,8 @@ restart:
 			if (args->len < blen)
 				continue;
 			ltdiff = xfs_alloc_compute_diff(args->agbno, args->len,
-				args->alignment, ltbnoa, ltlena, &ltnew);
+				args->alignment, args->userdata, ltbnoa,
+				ltlena, &ltnew);
 			if (ltnew != NULLAGBLOCK &&
 			    (args->len > blen || ltdiff < bdiff)) {
 				bdiff = ltdiff;
@@ -1128,7 +1138,8 @@ restart:
 			args->len = XFS_EXTLEN_MIN(ltlena, args->maxlen);
 			xfs_alloc_fix_len(args);
 			ltdiff = xfs_alloc_compute_diff(args->agbno, args->len,
-				args->alignment, ltbnoa, ltlena, &ltnew);
+				args->alignment, args->userdata, ltbnoa,
+				ltlena, &ltnew);
 
 			error = xfs_alloc_find_best_extent(args,
 						&bno_cur_lt, &bno_cur_gt,
@@ -1144,7 +1155,8 @@ restart:
 			args->len = XFS_EXTLEN_MIN(gtlena, args->maxlen);
 			xfs_alloc_fix_len(args);
 			gtdiff = xfs_alloc_compute_diff(args->agbno, args->len,
-				args->alignment, gtbnoa, gtlena, &gtnew);
+				args->alignment, args->userdata, gtbnoa,
+				gtlena, &gtnew);
 
 			error = xfs_alloc_find_best_extent(args,
 						&bno_cur_gt, &bno_cur_lt,
@@ -1203,7 +1215,7 @@ restart:
 	}
 	rlen = args->len;
 	(void)xfs_alloc_compute_diff(args->agbno, rlen, args->alignment,
-				     ltbnoa, ltlena, &ltnew);
+				     args->userdata, ltbnoa, ltlena, &ltnew);
 	ASSERT(ltnew >= ltbno);
 	ASSERT(ltnew + rlen <= ltbnoa + ltlena);
 	ASSERT(ltnew + rlen <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length));
diff --git a/fs/xfs/xfs_bmap_btree.h b/fs/xfs/xfs_bmap_btree.h
index 70c43d9f72c1..1b726d626941 100644
--- a/fs/xfs/xfs_bmap_btree.h
+++ b/fs/xfs/xfs_bmap_btree.h
@@ -196,6 +196,8 @@ typedef __be64 xfs_bmbt_ptr_t, xfs_bmdr_ptr_t;
 #define XFS_BMDR_SPACE_CALC(nrecs) \
 	(int)(sizeof(xfs_bmdr_block_t) + \
 	       ((nrecs) * (sizeof(xfs_bmbt_key_t) + sizeof(xfs_bmbt_ptr_t))))
+#define XFS_BMAP_BMDR_SPACE(bb) \
+	(XFS_BMDR_SPACE_CALC(be16_to_cpu((bb)->bb_numrecs)))
 
 /*
  * Maximum number of bmap btree levels.
diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index 4ec431777048..bfc4e0c26fd3 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -140,6 +140,16 @@ xfs_buf_item_size(
 
 	ASSERT(bip->bli_flags & XFS_BLI_LOGGED);
 
+	if (bip->bli_flags & XFS_BLI_ORDERED) {
+		/*
+		 * The buffer has been logged just to order it.
+		 * It is not being included in the transaction
+		 * commit, so no vectors are used at all.
+		 */
+		trace_xfs_buf_item_size_ordered(bip);
+		return XFS_LOG_VEC_ORDERED;
+	}
+
 	/*
 	 * the vector count is based on the number of buffer vectors we have
 	 * dirty bits in. This will only be greater than one when we have a
@@ -212,6 +222,7 @@ xfs_buf_item_format_segment(
 		goto out;
 	}
 
+
 	/*
 	 * Fill in an iovec for each set of contiguous chunks.
 	 */
@@ -299,18 +310,36 @@ xfs_buf_item_format(
 
 	/*
 	 * If it is an inode buffer, transfer the in-memory state to the
-	 * format flags and clear the in-memory state. We do not transfer
+	 * format flags and clear the in-memory state.
+	 *
+	 * For buffer based inode allocation, we do not transfer
 	 * this state if the inode buffer allocation has not yet been committed
 	 * to the log as setting the XFS_BLI_INODE_BUF flag will prevent
 	 * correct replay of the inode allocation.
+	 *
+	 * For icreate item based inode allocation, the buffers aren't written
+	 * to the journal during allocation, and hence we should always tag the
+	 * buffer as an inode buffer so that the correct unlinked list replay
+	 * occurs during recovery.
 	 */
 	if (bip->bli_flags & XFS_BLI_INODE_BUF) {
-		if (!((bip->bli_flags & XFS_BLI_INODE_ALLOC_BUF) &&
+		if (xfs_sb_version_hascrc(&lip->li_mountp->m_sb) ||
+		    !((bip->bli_flags & XFS_BLI_INODE_ALLOC_BUF) &&
 		      xfs_log_item_in_current_chkpt(lip)))
 			bip->__bli_format.blf_flags |= XFS_BLF_INODE_BUF;
 		bip->bli_flags &= ~XFS_BLI_INODE_BUF;
 	}
 
+	if ((bip->bli_flags & (XFS_BLI_ORDERED|XFS_BLI_STALE)) ==
+							XFS_BLI_ORDERED) {
+		/*
+		 * The buffer has been logged just to order it.  It is not being
+		 * included in the transaction commit, so don't format it.
+		 */
+		trace_xfs_buf_item_format_ordered(bip);
+		return;
+	}
+
 	for (i = 0; i < bip->bli_format_count; i++) {
 		vecp = xfs_buf_item_format_segment(bip, vecp, offset,
 						&bip->bli_formats[i]);
@@ -340,6 +369,7 @@ xfs_buf_item_pin(
 
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 	ASSERT((bip->bli_flags & XFS_BLI_LOGGED) ||
+	       (bip->bli_flags & XFS_BLI_ORDERED) ||
 	       (bip->bli_flags & XFS_BLI_STALE));
 
 	trace_xfs_buf_item_pin(bip);
@@ -512,8 +542,9 @@ xfs_buf_item_unlock(
 {
 	struct xfs_buf_log_item	*bip = BUF_ITEM(lip);
 	struct xfs_buf		*bp = bip->bli_buf;
-	int			aborted, clean, i;
-	uint			hold;
+	bool			clean;
+	bool			aborted;
+	int			flags;
 
 	/* Clear the buffer's association with this transaction. */
 	bp->b_transp = NULL;
@@ -524,23 +555,21 @@ xfs_buf_item_unlock(
 	 * (cancelled) buffers at unpin time, but we'll never go through the
 	 * pin/unpin cycle if we abort inside commit.
 	 */
-	aborted = (lip->li_flags & XFS_LI_ABORTED) != 0;
-
+	aborted = (lip->li_flags & XFS_LI_ABORTED) ? true : false;
 	/*
-	 * Before possibly freeing the buf item, determine if we should
-	 * release the buffer at the end of this routine.
+	 * Before possibly freeing the buf item, copy the per-transaction state
+	 * so we can reference it safely later after clearing it from the
+	 * buffer log item.
 	 */
-	hold = bip->bli_flags & XFS_BLI_HOLD;
-
-	/* Clear the per transaction state. */
-	bip->bli_flags &= ~(XFS_BLI_LOGGED | XFS_BLI_HOLD);
+	flags = bip->bli_flags;
+	bip->bli_flags &= ~(XFS_BLI_LOGGED | XFS_BLI_HOLD | XFS_BLI_ORDERED);
 
 	/*
 	 * If the buf item is marked stale, then don't do anything.  We'll
 	 * unlock the buffer and free the buf item when the buffer is unpinned
 	 * for the last time.
 	 */
-	if (bip->bli_flags & XFS_BLI_STALE) {
+	if (flags & XFS_BLI_STALE) {
 		trace_xfs_buf_item_unlock_stale(bip);
 		ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL);
 		if (!aborted) {
@@ -557,13 +586,19 @@ xfs_buf_item_unlock(
 	 * be the only reference to the buf item, so we free it anyway
 	 * regardless of whether it is dirty or not. A dirty abort implies a
 	 * shutdown, anyway.
+	 *
+	 * Ordered buffers are dirty but may have no recorded changes, so ensure
+	 * we only release clean items here.
 	 */
-	clean = 1;
-	for (i = 0; i < bip->bli_format_count; i++) {
-		if (!xfs_bitmap_empty(bip->bli_formats[i].blf_data_map,
-			     bip->bli_formats[i].blf_map_size)) {
-			clean = 0;
-			break;
+	clean = (flags & XFS_BLI_DIRTY) ? false : true;
+	if (clean) {
+		int i;
+		for (i = 0; i < bip->bli_format_count; i++) {
+			if (!xfs_bitmap_empty(bip->bli_formats[i].blf_data_map,
+				     bip->bli_formats[i].blf_map_size)) {
+				clean = false;
+				break;
+			}
 		}
 	}
 	if (clean)
@@ -576,7 +611,7 @@ xfs_buf_item_unlock(
 	} else
 		atomic_dec(&bip->bli_refcount);
 
-	if (!hold)
+	if (!(flags & XFS_BLI_HOLD))
 		xfs_buf_relse(bp);
 }
 
@@ -842,12 +877,6 @@ xfs_buf_item_log(
 	struct xfs_buf		*bp = bip->bli_buf;
 
 	/*
-	 * Mark the item as having some dirty data for
-	 * quick reference in xfs_buf_item_dirty.
-	 */
-	bip->bli_flags |= XFS_BLI_DIRTY;
-
-	/*
 	 * walk each buffer segment and mark them dirty appropriately.
 	 */
 	start = 0;
@@ -873,7 +902,7 @@ xfs_buf_item_log(
 
 
 /*
- * Return 1 if the buffer has some data that has been logged (at any
+ * Return 1 if the buffer has been logged or ordered in a transaction (at any
  * point, not just the current transaction) and 0 if not.
  */
 uint
@@ -907,11 +936,11 @@ void
 xfs_buf_item_relse(
 	xfs_buf_t	*bp)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	trace_xfs_buf_item_relse(bp, _RET_IP_);
+	ASSERT(!(bip->bli_item.li_flags & XFS_LI_IN_AIL));
 
-	bip = bp->b_fspriv;
 	bp->b_fspriv = bip->bli_item.li_bio_list;
 	if (bp->b_fspriv == NULL)
 		bp->b_iodone = NULL;
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index 2573d2a75fc8..0f1c247dc680 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -120,6 +120,7 @@ xfs_blft_from_flags(struct xfs_buf_log_format *blf)
 #define	XFS_BLI_INODE_ALLOC_BUF	0x10
 #define XFS_BLI_STALE_INODE	0x20
 #define	XFS_BLI_INODE_BUF	0x40
+#define	XFS_BLI_ORDERED		0x80
 
 #define XFS_BLI_FLAGS \
 	{ XFS_BLI_HOLD,		"HOLD" }, \
@@ -128,7 +129,8 @@ xfs_blft_from_flags(struct xfs_buf_log_format *blf)
 	{ XFS_BLI_LOGGED,	"LOGGED" }, \
 	{ XFS_BLI_INODE_ALLOC_BUF, "INODE_ALLOC" }, \
 	{ XFS_BLI_STALE_INODE,	"STALE_INODE" }, \
-	{ XFS_BLI_INODE_BUF,	"INODE_BUF" }
+	{ XFS_BLI_INODE_BUF,	"INODE_BUF" }, \
+	{ XFS_BLI_ORDERED,	"ORDERED" }
 
 
 #ifdef __KERNEL__
diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c
index c407e1ccff43..e36445ceaf80 100644
--- a/fs/xfs/xfs_dfrag.c
+++ b/fs/xfs/xfs_dfrag.c
@@ -24,6 +24,9 @@
 #include "xfs_ag.h"
 #include "xfs_mount.h"
 #include "xfs_bmap_btree.h"
+#include "xfs_alloc_btree.h"
+#include "xfs_ialloc_btree.h"
+#include "xfs_btree.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_inode_item.h"
@@ -182,7 +185,7 @@ xfs_swap_extents_check_format(
 	 */
 	if (tip->i_d.di_format == XFS_DINODE_FMT_BTREE) {
 		if (XFS_IFORK_BOFF(ip) &&
-		    tip->i_df.if_broot_bytes > XFS_IFORK_BOFF(ip))
+		    XFS_BMAP_BMDR_SPACE(tip->i_df.if_broot) > XFS_IFORK_BOFF(ip))
 			return EINVAL;
 		if (XFS_IFORK_NEXTENTS(tip, XFS_DATA_FORK) <=
 		    XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK))
@@ -192,9 +195,8 @@ xfs_swap_extents_check_format(
 	/* Reciprocal target->temp btree format checks */
 	if (ip->i_d.di_format == XFS_DINODE_FMT_BTREE) {
 		if (XFS_IFORK_BOFF(tip) &&
-		    ip->i_df.if_broot_bytes > XFS_IFORK_BOFF(tip))
+		    XFS_BMAP_BMDR_SPACE(ip->i_df.if_broot) > XFS_IFORK_BOFF(tip))
 			return EINVAL;
-
 		if (XFS_IFORK_NEXTENTS(ip, XFS_DATA_FORK) <=
 		    XFS_IFORK_MAXEXT(tip, XFS_DATA_FORK))
 			return EINVAL;
diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c
index e0cc1243a8aa..2aed25cae04d 100644
--- a/fs/xfs/xfs_dir2_leaf.c
+++ b/fs/xfs/xfs_dir2_leaf.c
@@ -1108,6 +1108,7 @@ xfs_dir2_leaf_readbuf(
 	struct xfs_mount	*mp = dp->i_mount;
 	struct xfs_buf		*bp = *bpp;
 	struct xfs_bmbt_irec	*map = mip->map;
+	struct blk_plug		plug;
 	int			error = 0;
 	int			length;
 	int			i;
@@ -1236,6 +1237,7 @@ xfs_dir2_leaf_readbuf(
 	/*
 	 * Do we need more readahead?
 	 */
+	blk_start_plug(&plug);
 	for (mip->ra_index = mip->ra_offset = i = 0;
 	     mip->ra_want > mip->ra_current && i < mip->map_blocks;
 	     i += mp->m_dirblkfsbs) {
@@ -1287,6 +1289,7 @@ xfs_dir2_leaf_readbuf(
 			}
 		}
 	}
+	blk_finish_plug(&plug);
 
 out:
 	*bpp = bp;
diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c
index 044e97a33c8d..f01012de06d0 100644
--- a/fs/xfs/xfs_dquot.c
+++ b/fs/xfs/xfs_dquot.c
@@ -570,13 +570,13 @@ xfs_qm_dqtobp(
 	xfs_buf_t		**O_bpp,
 	uint			flags)
 {
-	xfs_bmbt_irec_t map;
-	int		nmaps = 1, error;
-	xfs_buf_t	*bp;
-	xfs_inode_t	*quotip = XFS_DQ_TO_QIP(dqp);
-	xfs_mount_t	*mp = dqp->q_mount;
-	xfs_dqid_t	id = be32_to_cpu(dqp->q_core.d_id);
-	xfs_trans_t	*tp = (tpp ? *tpp : NULL);
+	struct xfs_bmbt_irec	map;
+	int			nmaps = 1, error;
+	struct xfs_buf		*bp;
+	struct xfs_inode	*quotip = xfs_dq_to_quota_inode(dqp);
+	struct xfs_mount	*mp = dqp->q_mount;
+	xfs_dqid_t		id = be32_to_cpu(dqp->q_core.d_id);
+	struct xfs_trans	*tp = (tpp ? *tpp : NULL);
 
 	dqp->q_fileoffset = (xfs_fileoff_t)id / mp->m_quotainfo->qi_dqperchunk;
 
@@ -804,7 +804,7 @@ xfs_qm_dqget(
 	xfs_dquot_t	**O_dqpp) /* OUT : locked incore dquot */
 {
 	struct xfs_quotainfo	*qi = mp->m_quotainfo;
-	struct radix_tree_root *tree = XFS_DQUOT_TREE(qi, type);
+	struct radix_tree_root *tree = xfs_dquot_tree(qi, type);
 	struct xfs_dquot	*dqp;
 	int			error;
 
diff --git a/fs/xfs/xfs_dquot.h b/fs/xfs/xfs_dquot.h
index 4f0ebfc43cc9..b596626249b8 100644
--- a/fs/xfs/xfs_dquot.h
+++ b/fs/xfs/xfs_dquot.h
@@ -143,10 +143,6 @@ static inline xfs_dquot_t *xfs_inode_dquot(struct xfs_inode *ip, int type)
 #define XFS_QM_ISUDQ(dqp)	((dqp)->dq_flags & XFS_DQ_USER)
 #define XFS_QM_ISPDQ(dqp)	((dqp)->dq_flags & XFS_DQ_PROJ)
 #define XFS_QM_ISGDQ(dqp)	((dqp)->dq_flags & XFS_DQ_GROUP)
-#define XFS_DQ_TO_QINF(dqp)	((dqp)->q_mount->m_quotainfo)
-#define XFS_DQ_TO_QIP(dqp)	(XFS_QM_ISUDQ(dqp) ? \
-				 XFS_DQ_TO_QINF(dqp)->qi_uquotaip : \
-				 XFS_DQ_TO_QINF(dqp)->qi_gquotaip)
 
 extern int		xfs_qm_dqread(struct xfs_mount *, xfs_dqid_t, uint,
 					uint, struct xfs_dquot	**);
diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index 3c3644ea825b..614eb0cc3608 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -176,7 +176,7 @@ xfs_growfs_data_private(
 	if (!bp)
 		return EIO;
 	if (bp->b_error) {
-		int	error = bp->b_error;
+		error = bp->b_error;
 		xfs_buf_relse(bp);
 		return error;
 	}
diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c
index c8f5ae1debf2..7a0c17d7ec09 100644
--- a/fs/xfs/xfs_ialloc.c
+++ b/fs/xfs/xfs_ialloc.c
@@ -38,6 +38,7 @@
 #include "xfs_bmap.h"
 #include "xfs_cksum.h"
 #include "xfs_buf_item.h"
+#include "xfs_icreate_item.h"
 
 
 /*
@@ -150,12 +151,16 @@ xfs_check_agi_freecount(
 #endif
 
 /*
- * Initialise a new set of inodes.
+ * Initialise a new set of inodes. When called without a transaction context
+ * (e.g. from recovery) we initiate a delayed write of the inode buffers rather
+ * than logging them (which in a transaction context puts them into the AIL
+ * for writeback rather than the xfsbufd queue).
  */
-STATIC int
+int
 xfs_ialloc_inode_init(
 	struct xfs_mount	*mp,
 	struct xfs_trans	*tp,
+	struct list_head	*buffer_list,
 	xfs_agnumber_t		agno,
 	xfs_agblock_t		agbno,
 	xfs_agblock_t		length,
@@ -208,6 +213,18 @@ xfs_ialloc_inode_init(
 		version = 3;
 		ino = XFS_AGINO_TO_INO(mp, agno,
 				       XFS_OFFBNO_TO_AGINO(mp, agbno, 0));
+
+		/*
+		 * log the initialisation that is about to take place as an
+		 * logical operation. This means the transaction does not
+		 * need to log the physical changes to the inode buffers as log
+		 * recovery will know what initialisation is actually needed.
+		 * Hence we only need to log the buffers as "ordered" buffers so
+		 * they track in the AIL as if they were physically logged.
+		 */
+		if (tp)
+			xfs_icreate_log(tp, agno, agbno, XFS_IALLOC_INODES(mp),
+					mp->m_sb.sb_inodesize, length, gen);
 	} else if (xfs_sb_version_hasnlink(&mp->m_sb))
 		version = 2;
 	else
@@ -223,13 +240,8 @@ xfs_ialloc_inode_init(
 					 XBF_UNMAPPED);
 		if (!fbuf)
 			return ENOMEM;
-		/*
-		 * Initialize all inodes in this buffer and then log them.
-		 *
-		 * XXX: It would be much better if we had just one transaction
-		 *	to log a whole cluster of inodes instead of all the
-		 *	individual transactions causing a lot of log traffic.
-		 */
+
+		/* Initialize the inode buffers and log them appropriately. */
 		fbuf->b_ops = &xfs_inode_buf_ops;
 		xfs_buf_zero(fbuf, 0, BBTOB(fbuf->b_length));
 		for (i = 0; i < ninodes; i++) {
@@ -247,18 +259,39 @@ xfs_ialloc_inode_init(
 				ino++;
 				uuid_copy(&free->di_uuid, &mp->m_sb.sb_uuid);
 				xfs_dinode_calc_crc(mp, free);
-			} else {
+			} else if (tp) {
 				/* just log the inode core */
 				xfs_trans_log_buf(tp, fbuf, ioffset,
 						  ioffset + isize - 1);
 			}
 		}
-		if (version == 3) {
-			/* need to log the entire buffer */
-			xfs_trans_log_buf(tp, fbuf, 0,
-					  BBTOB(fbuf->b_length) - 1);
+
+		if (tp) {
+			/*
+			 * Mark the buffer as an inode allocation buffer so it
+			 * sticks in AIL at the point of this allocation
+			 * transaction. This ensures the they are on disk before
+			 * the tail of the log can be moved past this
+			 * transaction (i.e. by preventing relogging from moving
+			 * it forward in the log).
+			 */
+			xfs_trans_inode_alloc_buf(tp, fbuf);
+			if (version == 3) {
+				/*
+				 * Mark the buffer as ordered so that they are
+				 * not physically logged in the transaction but
+				 * still tracked in the AIL as part of the
+				 * transaction and pin the log appropriately.
+				 */
+				xfs_trans_ordered_buf(tp, fbuf);
+				xfs_trans_log_buf(tp, fbuf, 0,
+						  BBTOB(fbuf->b_length) - 1);
+			}
+		} else {
+			fbuf->b_flags |= XBF_DONE;
+			xfs_buf_delwri_queue(fbuf, buffer_list);
+			xfs_buf_relse(fbuf);
 		}
-		xfs_trans_inode_alloc_buf(tp, fbuf);
 	}
 	return 0;
 }
@@ -303,7 +336,7 @@ xfs_ialloc_ag_alloc(
 	 * First try to allocate inodes contiguous with the last-allocated
 	 * chunk of inodes.  If the filesystem is striped, this will fill
 	 * an entire stripe unit with inodes.
- 	 */
+	 */
 	agi = XFS_BUF_TO_AGI(agbp);
 	newino = be32_to_cpu(agi->agi_newino);
 	agno = be32_to_cpu(agi->agi_seqno);
@@ -402,7 +435,7 @@ xfs_ialloc_ag_alloc(
 	 * rather than a linear progression to prevent the next generation
 	 * number from being easily guessable.
 	 */
-	error = xfs_ialloc_inode_init(args.mp, tp, agno, args.agbno,
+	error = xfs_ialloc_inode_init(args.mp, tp, NULL, agno, args.agbno,
 			args.len, prandom_u32());
 
 	if (error)
@@ -615,8 +648,7 @@ xfs_ialloc_get_rec(
 	struct xfs_btree_cur	*cur,
 	xfs_agino_t		agino,
 	xfs_inobt_rec_incore_t	*rec,
-	int			*done,
-	int			left)
+	int			*done)
 {
 	int                     error;
 	int			i;
@@ -724,12 +756,12 @@ xfs_dialloc_ag(
 		    pag->pagl_leftrec != NULLAGINO &&
 		    pag->pagl_rightrec != NULLAGINO) {
 			error = xfs_ialloc_get_rec(tcur, pag->pagl_leftrec,
-						   &trec, &doneleft, 1);
+						   &trec, &doneleft);
 			if (error)
 				goto error1;
 
 			error = xfs_ialloc_get_rec(cur, pag->pagl_rightrec,
-						   &rec, &doneright, 0);
+						   &rec, &doneright);
 			if (error)
 				goto error1;
 		} else {
diff --git a/fs/xfs/xfs_ialloc.h b/fs/xfs/xfs_ialloc.h
index c8da3df271e6..68c07320f096 100644
--- a/fs/xfs/xfs_ialloc.h
+++ b/fs/xfs/xfs_ialloc.h
@@ -150,6 +150,14 @@ int xfs_inobt_lookup(struct xfs_btree_cur *cur, xfs_agino_t ino,
 int xfs_inobt_get_rec(struct xfs_btree_cur *cur,
 		xfs_inobt_rec_incore_t *rec, int *stat);
 
+/*
+ * Inode chunk initialisation routine
+ */
+int xfs_ialloc_inode_init(struct xfs_mount *mp, struct xfs_trans *tp,
+			  struct list_head *buffer_list,
+			  xfs_agnumber_t agno, xfs_agblock_t agbno,
+			  xfs_agblock_t length, unsigned int gen);
+
 extern const struct xfs_buf_ops xfs_agi_buf_ops;
 
 #endif	/* __XFS_IALLOC_H__ */
diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index 96e344e3e927..9560dc1f15a9 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -335,7 +335,8 @@ xfs_iget_cache_miss(
 	iflags = XFS_INEW;
 	if (flags & XFS_IGET_DONTCACHE)
 		iflags |= XFS_IDONTCACHE;
-	ip->i_udquot = ip->i_gdquot = NULL;
+	ip->i_udquot = NULL;
+	ip->i_gdquot = NULL;
 	xfs_iflags_set(ip, iflags);
 
 	/* insert the new inode */
diff --git a/fs/xfs/xfs_icache.h b/fs/xfs/xfs_icache.h
index e0f138c70a2f..a01afbb3909a 100644
--- a/fs/xfs/xfs_icache.h
+++ b/fs/xfs/xfs_icache.h
@@ -40,7 +40,6 @@ void xfs_inode_clear_eofblocks_tag(struct xfs_inode *ip);
 int xfs_icache_free_eofblocks(struct xfs_mount *, struct xfs_eofblocks *);
 void xfs_eofblocks_worker(struct work_struct *);
 
-int xfs_sync_inode_grab(struct xfs_inode *ip);
 int xfs_inode_ag_iterator(struct xfs_mount *mp,
 	int (*execute)(struct xfs_inode *ip, struct xfs_perag *pag,
 		int flags, void *args),
diff --git a/fs/xfs/xfs_icreate_item.c b/fs/xfs/xfs_icreate_item.c
new file mode 100644
index 000000000000..7716a4e7375e
--- /dev/null
+++ b/fs/xfs/xfs_icreate_item.c
@@ -0,0 +1,195 @@
+/*
+ * Copyright (c) 2008-2010, 2013 Dave Chinner