Merge tag 'xfs-for-linus-4.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs

Pull xfs updates from Dave Chinner:
 "A pretty average collection of fixes, cleanups and improvements in
  this request.

  Summary:
   - fixes for mount line parsing, sparse warnings, read-only compat
     feature remount behaviour
   - allow fast path symlink lookups for inline symlinks.
   - attribute listing cleanups
   - writeback goes direct to bios rather than indirecting through
     bufferheads
   - transaction allocation cleanup
   - optimised kmem_realloc
   - added configurable error handling for metadata write errors,
     changed default error handling behaviour from "retry forever" to
     "retry until unmount then fail"
   - fixed several inode cluster writeback lookup vs reclaim race
     conditions
   - fixed inode cluster writeback checking wrong inode after lookup
   - fixed bugs where struct xfs_inode freeing wasn't actually RCU safe
   - cleaned up inode reclaim tagging"

* tag 'xfs-for-linus-4.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs: (39 commits)
  xfs: fix warning in xfs_finish_page_writeback for non-debug builds
  xfs: move reclaim tagging functions
  xfs: simplify inode reclaim tagging interfaces
  xfs: rename variables in xfs_iflush_cluster for clarity
  xfs: xfs_iflush_cluster has range issues
  xfs: mark reclaimed inodes invalid earlier
  xfs: xfs_inode_free() isn't RCU safe
  xfs: optimise xfs_iext_destroy
  xfs: skip stale inodes in xfs_iflush_cluster
  xfs: fix inode validity check in xfs_iflush_cluster
  xfs: xfs_iflush_cluster fails to abort on error
  xfs: remove xfs_fs_evict_inode()
  xfs: add "fail at unmount" error handling configuration
  xfs: add configuration handlers for specific errors
  xfs: add configuration of error failure speed
  xfs: introduce table-based init for error behaviors
  xfs: add configurable error support to metadata buffers
  xfs: introduce metadata IO error class
  xfs: configurable error behavior via sysfs
  xfs: buffer ->bi_end_io function requires irq-safe lock
  ...

fs/namei.c

@@ -4542,7 +4542,6 @@ int readlink_copy(char __user *buffer, int buflen, const char *link)
 out:
 	return len;
 }
-EXPORT_SYMBOL(readlink_copy);
 
 /*
  * A helper for ->readlink().  This should be used *ONLY* for symlinks that

fs/xfs/kmem.c

@@ -93,19 +93,23 @@ kmem_zalloc_large(size_t size, xfs_km_flags_t flags)
 }
 
 void *
-kmem_realloc(const void *ptr, size_t newsize, size_t oldsize,
-	     xfs_km_flags_t flags)
+kmem_realloc(const void *old, size_t newsize, xfs_km_flags_t flags)
 {
-	void	*new;
+	int	retries = 0;
+	gfp_t	lflags = kmem_flags_convert(flags);
+	void	*ptr;
 
-	new = kmem_alloc(newsize, flags);
-	if (ptr) {
-		if (new)
-			memcpy(new, ptr,
-				((oldsize < newsize) ? oldsize : newsize));
-		kmem_free(ptr);
-	}
-	return new;
+	do {
+		ptr = krealloc(old, newsize, lflags);
+		if (ptr || (flags & (KM_MAYFAIL|KM_NOSLEEP)))
+			return ptr;
+		if (!(++retries % 100))
+			xfs_err(NULL,
+	"%s(%u) possible memory allocation deadlock size %zu in %s (mode:0x%x)",
+				current->comm, current->pid,
+				newsize, __func__, lflags);
+		congestion_wait(BLK_RW_ASYNC, HZ/50);
+	} while (1);
 }
 
 void *

fs/xfs/kmem.h

@@ -62,7 +62,7 @@ kmem_flags_convert(xfs_km_flags_t flags)
 
 extern void *kmem_alloc(size_t, xfs_km_flags_t);
 extern void *kmem_zalloc_large(size_t size, xfs_km_flags_t);
-extern void *kmem_realloc(const void *, size_t, size_t, xfs_km_flags_t);
+extern void *kmem_realloc(const void *, size_t, xfs_km_flags_t);
 static inline void  kmem_free(const void *ptr)
 {
 	kvfree(ptr);

fs/xfs/libxfs/xfs_attr.c

@@ -242,37 +242,21 @@ xfs_attr_set(
 			return error;
 	}
 
-	/*
-	 * Start our first transaction of the day.
-	 *
-	 * All future transactions during this code must be "chained" off
-	 * this one via the trans_dup() call.  All transactions will contain
-	 * the inode, and the inode will always be marked with trans_ihold().
-	 * Since the inode will be locked in all transactions, we must log
-	 * the inode in every transaction to let it float upward through
-	 * the log.
-	 */
-	args.trans = xfs_trans_alloc(mp, XFS_TRANS_ATTR_SET);
+	tres.tr_logres = M_RES(mp)->tr_attrsetm.tr_logres +
+			 M_RES(mp)->tr_attrsetrt.tr_logres * args.total;
+	tres.tr_logcount = XFS_ATTRSET_LOG_COUNT;
+	tres.tr_logflags = XFS_TRANS_PERM_LOG_RES;
 
 	/*
 	 * Root fork attributes can use reserved data blocks for this
 	 * operation if necessary
 	 */
-
-	if (rsvd)
-		args.trans->t_flags |= XFS_TRANS_RESERVE;
-
-	tres.tr_logres = M_RES(mp)->tr_attrsetm.tr_logres +
-			 M_RES(mp)->tr_attrsetrt.tr_logres * args.total;
-	tres.tr_logcount = XFS_ATTRSET_LOG_COUNT;
-	tres.tr_logflags = XFS_TRANS_PERM_LOG_RES;
-	error = xfs_trans_reserve(args.trans, &tres, args.total, 0);
-	if (error) {
-		xfs_trans_cancel(args.trans);
+	error = xfs_trans_alloc(mp, &tres, args.total, 0,
+			rsvd ? XFS_TRANS_RESERVE : 0, &args.trans);
+	if (error)
 		return error;
-	}
-	xfs_ilock(dp, XFS_ILOCK_EXCL);
 
+	xfs_ilock(dp, XFS_ILOCK_EXCL);
 	error = xfs_trans_reserve_quota_nblks(args.trans, dp, args.total, 0,
 				rsvd ? XFS_QMOPT_RES_REGBLKS | XFS_QMOPT_FORCE_RES :
 				       XFS_QMOPT_RES_REGBLKS);
@@ -428,32 +412,16 @@ xfs_attr_remove(
 	if (error)
 		return error;
 
-	/*
-	 * Start our first transaction of the day.
-	 *
-	 * All future transactions during this code must be "chained" off
-	 * this one via the trans_dup() call.  All transactions will contain
-	 * the inode, and the inode will always be marked with trans_ihold().
-	 * Since the inode will be locked in all transactions, we must log
-	 * the inode in every transaction to let it float upward through
-	 * the log.
-	 */
-	args.trans = xfs_trans_alloc(mp, XFS_TRANS_ATTR_RM);
-
 	/*
 	 * Root fork attributes can use reserved data blocks for this
 	 * operation if necessary
 	 */
-
-	if (flags & ATTR_ROOT)
-		args.trans->t_flags |= XFS_TRANS_RESERVE;
-
-	error = xfs_trans_reserve(args.trans, &M_RES(mp)->tr_attrrm,
-				  XFS_ATTRRM_SPACE_RES(mp), 0);
-	if (error) {
-		xfs_trans_cancel(args.trans);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_attrrm,
+			XFS_ATTRRM_SPACE_RES(mp), 0,
+			(flags & ATTR_ROOT) ? XFS_TRANS_RESERVE : 0,
+			&args.trans);
+	if (error)
 		return error;
-	}
 
 	xfs_ilock(dp, XFS_ILOCK_EXCL);
 	/*

fs/xfs/libxfs/xfs_bmap.c

@@ -1121,15 +1121,14 @@ xfs_bmap_add_attrfork(
 
 	mp = ip->i_mount;
 	ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
-	tp = xfs_trans_alloc(mp, XFS_TRANS_ADDAFORK);
+
 	blks = XFS_ADDAFORK_SPACE_RES(mp);
-	if (rsvd)
-		tp->t_flags |= XFS_TRANS_RESERVE;
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_addafork, blks, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_addafork, blks, 0,
+			rsvd ? XFS_TRANS_RESERVE : 0, &tp);
+	if (error)
 		return error;
-	}
+
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	error = xfs_trans_reserve_quota_nblks(tp, ip, blks, 0, rsvd ?
 			XFS_QMOPT_RES_REGBLKS | XFS_QMOPT_FORCE_RES :
@@ -6026,13 +6025,10 @@ xfs_bmap_split_extent(
 	xfs_fsblock_t           firstfsb;
 	int                     error;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write,
-			XFS_DIOSTRAT_SPACE_RES(mp, 0), 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write,
+			XFS_DIOSTRAT_SPACE_RES(mp, 0), 0, 0, &tp);
+	if (error)
 		return error;
-	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);

fs/xfs/libxfs/xfs_dir2_sf.c

@@ -257,15 +257,12 @@ xfs_dir2_block_to_sf(
 	 *
 	 * Convert the inode to local format and copy the data in.
 	 */
-	dp->i_df.if_flags &= ~XFS_IFEXTENTS;
-	dp->i_df.if_flags |= XFS_IFINLINE;
-	dp->i_d.di_format = XFS_DINODE_FMT_LOCAL;
 	ASSERT(dp->i_df.if_bytes == 0);
-	xfs_idata_realloc(dp, size, XFS_DATA_FORK);
+	xfs_init_local_fork(dp, XFS_DATA_FORK, dst, size);
+	dp->i_d.di_format = XFS_DINODE_FMT_LOCAL;
+	dp->i_d.di_size = size;
 
 	logflags |= XFS_ILOG_DDATA;
-	memcpy(dp->i_df.if_u1.if_data, dst, size);
-	dp->i_d.di_size = size;
 	xfs_dir2_sf_check(args);
 out:
 	xfs_trans_log_inode(args->trans, dp, logflags);

fs/xfs/libxfs/xfs_inode_fork.c

@@ -231,6 +231,48 @@ xfs_iformat_fork(
 	return error;
 }
 
+void
+xfs_init_local_fork(
+	struct xfs_inode	*ip,
+	int			whichfork,
+	const void		*data,
+	int			size)
+{
+	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	int			mem_size = size, real_size = 0;
+	bool			zero_terminate;
+
+	/*
+	 * If we are using the local fork to store a symlink body we need to
+	 * zero-terminate it so that we can pass it back to the VFS directly.
+	 * Overallocate the in-memory fork by one for that and add a zero
+	 * to terminate it below.
+	 */
+	zero_terminate = S_ISLNK(VFS_I(ip)->i_mode);
+	if (zero_terminate)
+		mem_size++;
+
+	if (size == 0)
+		ifp->if_u1.if_data = NULL;
+	else if (mem_size <= sizeof(ifp->if_u2.if_inline_data))
+		ifp->if_u1.if_data = ifp->if_u2.if_inline_data;
+	else {
+		real_size = roundup(mem_size, 4);
+		ifp->if_u1.if_data = kmem_alloc(real_size, KM_SLEEP | KM_NOFS);
+	}
+
+	if (size) {
+		memcpy(ifp->if_u1.if_data, data, size);
+		if (zero_terminate)
+			ifp->if_u1.if_data[size] = '\0';
+	}
+
+	ifp->if_bytes = size;
+	ifp->if_real_bytes = real_size;
+	ifp->if_flags &= ~(XFS_IFEXTENTS | XFS_IFBROOT);
+	ifp->if_flags |= XFS_IFINLINE;
+}
+
 /*
  * The file is in-lined in the on-disk inode.
  * If it fits into if_inline_data, then copy
@@ -248,8 +290,6 @@ xfs_iformat_local(
 	int		whichfork,
 	int		size)
 {
-	xfs_ifork_t	*ifp;
-	int		real_size;
 
 	/*
 	 * If the size is unreasonable, then something
@@ -265,22 +305,8 @@ xfs_iformat_local(
 				     ip->i_mount, dip);
 		return -EFSCORRUPTED;
 	}
-	ifp = XFS_IFORK_PTR(ip, whichfork);
-	real_size = 0;
-	if (size == 0)
-		ifp->if_u1.if_data = NULL;
-	else if (size <= sizeof(ifp->if_u2.if_inline_data))
-		ifp->if_u1.if_data = ifp->if_u2.if_inline_data;
-	else {
-		real_size = roundup(size, 4);
-		ifp->if_u1.if_data = kmem_alloc(real_size, KM_SLEEP | KM_NOFS);
-	}
-	ifp->if_bytes = size;
-	ifp->if_real_bytes = real_size;
-	if (size)
-		memcpy(ifp->if_u1.if_data, XFS_DFORK_PTR(dip, whichfork), size);
-	ifp->if_flags &= ~XFS_IFEXTENTS;
-	ifp->if_flags |= XFS_IFINLINE;
+
+	xfs_init_local_fork(ip, whichfork, XFS_DFORK_PTR(dip, whichfork), size);
 	return 0;
 }
 
@@ -516,7 +542,6 @@ xfs_iroot_realloc(
 		new_max = cur_max + rec_diff;
 		new_size = XFS_BMAP_BROOT_SPACE_CALC(mp, new_max);
 		ifp->if_broot = kmem_realloc(ifp->if_broot, new_size,
-				XFS_BMAP_BROOT_SPACE_CALC(mp, cur_max),
 				KM_SLEEP | KM_NOFS);
 		op = (char *)XFS_BMAP_BROOT_PTR_ADDR(mp, ifp->if_broot, 1,
 						     ifp->if_broot_bytes);
@@ -660,7 +685,6 @@ xfs_idata_realloc(
 				ifp->if_u1.if_data =
 					kmem_realloc(ifp->if_u1.if_data,
 							real_size,
-							ifp->if_real_bytes,
 							KM_SLEEP | KM_NOFS);
 			}
 		} else {
@@ -1376,8 +1400,7 @@ xfs_iext_realloc_direct(
 		if (rnew_size != ifp->if_real_bytes) {
 			ifp->if_u1.if_extents =
 				kmem_realloc(ifp->if_u1.if_extents,
-						rnew_size,
-						ifp->if_real_bytes, KM_NOFS);
+						rnew_size, KM_NOFS);
 		}
 		if (rnew_size > ifp->if_real_bytes) {
 			memset(&ifp->if_u1.if_extents[ifp->if_bytes /
@@ -1461,9 +1484,8 @@ xfs_iext_realloc_indirect(
 	if (new_size == 0) {
 		xfs_iext_destroy(ifp);
 	} else {
-		ifp->if_u1.if_ext_irec = (xfs_ext_irec_t *)
-			kmem_realloc(ifp->if_u1.if_ext_irec,
-				new_size, size, KM_NOFS);
+		ifp->if_u1.if_ext_irec =
+			kmem_realloc(ifp->if_u1.if_ext_irec, new_size, KM_NOFS);
 	}
 }
 
@@ -1496,6 +1518,24 @@ xfs_iext_indirect_to_direct(
 	}
 }
 
+/*
+ * Remove all records from the indirection array.
+ */
+STATIC void
+xfs_iext_irec_remove_all(
+	struct xfs_ifork *ifp)
+{
+	int		nlists;
+	int		i;
+
+	ASSERT(ifp->if_flags & XFS_IFEXTIREC);
+	nlists = ifp->if_real_bytes / XFS_IEXT_BUFSZ;
+	for (i = 0; i < nlists; i++)
+		kmem_free(ifp->if_u1.if_ext_irec[i].er_extbuf);
+	kmem_free(ifp->if_u1.if_ext_irec);
+	ifp->if_flags &= ~XFS_IFEXTIREC;
+}
+
 /*
  * Free incore file extents.
  */
@@ -1504,14 +1544,7 @@ xfs_iext_destroy(
 	xfs_ifork_t	*ifp)		/* inode fork pointer */
 {
 	if (ifp->if_flags & XFS_IFEXTIREC) {
-		int	erp_idx;
-		int	nlists;
-
-		nlists = ifp->if_real_bytes / XFS_IEXT_BUFSZ;
-		for (erp_idx = nlists - 1; erp_idx >= 0 ; erp_idx--) {
-			xfs_iext_irec_remove(ifp, erp_idx);
-		}
-		ifp->if_flags &= ~XFS_IFEXTIREC;
+		xfs_iext_irec_remove_all(ifp);
 	} else if (ifp->if_real_bytes) {
 		kmem_free(ifp->if_u1.if_extents);
 	} else if (ifp->if_bytes) {

fs/xfs/libxfs/xfs_inode_fork.h

@@ -134,6 +134,7 @@ void		xfs_iroot_realloc(struct xfs_inode *, int, int);
 int		xfs_iread_extents(struct xfs_trans *, struct xfs_inode *, int);
 int		xfs_iextents_copy(struct xfs_inode *, struct xfs_bmbt_rec *,
 				  int);
+void		xfs_init_local_fork(struct xfs_inode *, int, const void *, int);
 
 struct xfs_bmbt_rec_host *
 		xfs_iext_get_ext(struct xfs_ifork *, xfs_extnum_t);

fs/xfs/libxfs/xfs_log_format.h

@@ -211,6 +211,11 @@ typedef struct xfs_trans_header {
 
 #define	XFS_TRANS_HEADER_MAGIC	0x5452414e	/* TRAN */
 
+/*
+ * The only type valid for th_type in CIL-enabled file system logs:
+ */
+#define XFS_TRANS_CHECKPOINT	40
+
 /*
  * Log item types.
  */

fs/xfs/libxfs/xfs_sb.c

@@ -838,12 +838,10 @@ xfs_sync_sb(
 	struct xfs_trans	*tp;
 	int			error;
 
-	tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_CHANGE, KM_SLEEP);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_sb, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_sb, 0, 0,
+			XFS_TRANS_NO_WRITECOUNT, &tp);
+	if (error)
 		return error;
-	}
 
 	xfs_log_sb(tp);
 	if (wait)

fs/xfs/libxfs/xfs_shared.h

@@ -55,103 +55,6 @@ extern const struct xfs_buf_ops xfs_sb_quiet_buf_ops;
 extern const struct xfs_buf_ops xfs_symlink_buf_ops;
 extern const struct xfs_buf_ops xfs_rtbuf_ops;
 
-/*
- * Transaction types.  Used to distinguish types of buffers. These never reach
- * the log.
- */
-#define XFS_TRANS_SETATTR_NOT_SIZE	1
-#define XFS_TRANS_SETATTR_SIZE		2
-#define XFS_TRANS_INACTIVE		3
-#define XFS_TRANS_CREATE		4
-#define XFS_TRANS_CREATE_TRUNC		5
-#define XFS_TRANS_TRUNCATE_FILE		6
-#define XFS_TRANS_REMOVE		7
-#define XFS_TRANS_LINK			8
-#define XFS_TRANS_RENAME		9
-#define XFS_TRANS_MKDIR			10
-#define XFS_TRANS_RMDIR			11
-#define XFS_TRANS_SYMLINK		12
-#define XFS_TRANS_SET_DMATTRS		13
-#define XFS_TRANS_GROWFS		14
-#define XFS_TRANS_STRAT_WRITE		15
-#define XFS_TRANS_DIOSTRAT		16
-/* 17 was XFS_TRANS_WRITE_SYNC */
-#define	XFS_TRANS_WRITEID		18
-#define	XFS_TRANS_ADDAFORK		19
-#define	XFS_TRANS_ATTRINVAL		20
-#define	XFS_TRANS_ATRUNCATE		21
-#define	XFS_TRANS_ATTR_SET		22
-#define	XFS_TRANS_ATTR_RM		23
-#define	XFS_TRANS_ATTR_FLAG		24
-#define	XFS_TRANS_CLEAR_AGI_BUCKET	25
-#define XFS_TRANS_SB_CHANGE		26
-/*
- * Dummy entries since we use the transaction type to index into the
- * trans_type[] in xlog_recover_print_trans_head()
- */
-#define XFS_TRANS_DUMMY1		27
-#define XFS_TRANS_DUMMY2		28
-#define XFS_TRANS_QM_QUOTAOFF		29
-#define XFS_TRANS_QM_DQALLOC		30
-#define XFS_TRANS_QM_SETQLIM		31
-#define XFS_TRANS_QM_DQCLUSTER		32
-#define XFS_TRANS_QM_QINOCREATE		33
-#define XFS_TRANS_QM_QUOTAOFF_END	34
-#define XFS_TRANS_FSYNC_TS		35
-#define	XFS_TRANS_GROWFSRT_ALLOC	36
-#define	XFS_TRANS_GROWFSRT_ZERO		37
-#define	XFS_TRANS_GROWFSRT_FREE		38
-#define	XFS_TRANS_SWAPEXT		39
-#define	XFS_TRANS_CHECKPOINT		40
-#define	XFS_TRANS_ICREATE		41
-#define	XFS_TRANS_CREATE_TMPFILE	42
-#define	XFS_TRANS_TYPE_MAX		43
-/* new transaction types need to be reflected in xfs_logprint(8) */
-
-#define XFS_TRANS_TYPES \
-	{ XFS_TRANS_SETATTR_NOT_SIZE,	"SETATTR_NOT_SIZE" }, \
-	{ XFS_TRANS_SETATTR_SIZE,	"SETATTR_SIZE" }, \
-	{ XFS_TRANS_INACTIVE,		"INACTIVE" }, \
-	{ XFS_TRANS_CREATE,		"CREATE" }, \
-	{ XFS_TRANS_CREATE_TRUNC,	"CREATE_TRUNC" }, \
-	{ XFS_TRANS_TRUNCATE_FILE,	"TRUNCATE_FILE" }, \
-	{ XFS_TRANS_REMOVE,		"REMOVE" }, \
-	{ XFS_TRANS_LINK,		"LINK" }, \
-	{ XFS_TRANS_RENAME,		"RENAME" }, \
-	{ XFS_TRANS_MKDIR,		"MKDIR" }, \
-	{ XFS_TRANS_RMDIR,		"RMDIR" }, \
-	{ XFS_TRANS_SYMLINK,		"SYMLINK" }, \
-	{ XFS_TRANS_SET_DMATTRS,	"SET_DMATTRS" }, \
-	{ XFS_TRANS_GROWFS,		"GROWFS" }, \
-	{ XFS_TRANS_STRAT_WRITE,	"STRAT_WRITE" }, \
-	{ XFS_TRANS_DIOSTRAT,		"DIOSTRAT" }, \
-	{ XFS_TRANS_WRITEID,		"WRITEID" }, \
-	{ XFS_TRANS_ADDAFORK,		"ADDAFORK" }, \
-	{ XFS_TRANS_ATTRINVAL,		"ATTRINVAL" }, \
-	{ XFS_TRANS_ATRUNCATE,		"ATRUNCATE" }, \
-	{ XFS_TRANS_ATTR_SET,		"ATTR_SET" }, \
-	{ XFS_TRANS_ATTR_RM,		"ATTR_RM" }, \
-	{ XFS_TRANS_ATTR_FLAG,		"ATTR_FLAG" }, \
-	{ XFS_TRANS_CLEAR_AGI_BUCKET,	"CLEAR_AGI_BUCKET" }, \
-	{ XFS_TRANS_SB_CHANGE,		"SBCHANGE" }, \
-	{ XFS_TRANS_DUMMY1,		"DUMMY1" }, \
-	{ XFS_TRANS_DUMMY2,		"DUMMY2" }, \
-	{ XFS_TRANS_QM_QUOTAOFF,	"QM_QUOTAOFF" }, \
-	{ XFS_TRANS_QM_DQALLOC,		"QM_DQALLOC" }, \
-	{ XFS_TRANS_QM_SETQLIM,		"QM_SETQLIM" }, \
-	{ XFS_TRANS_QM_DQCLUSTER,	"QM_DQCLUSTER" }, \
-	{ XFS_TRANS_QM_QINOCREATE,	"QM_QINOCREATE" }, \
-	{ XFS_TRANS_QM_QUOTAOFF_END,	"QM_QOFF_END" }, \
-	{ XFS_TRANS_FSYNC_TS,		"FSYNC_TS" }, \
-	{ XFS_TRANS_GROWFSRT_ALLOC,	"GROWFSRT_ALLOC" }, \
-	{ XFS_TRANS_GROWFSRT_ZERO,	"GROWFSRT_ZERO" }, \
-	{ XFS_TRANS_GROWFSRT_FREE,	"GROWFSRT_FREE" }, \
-	{ XFS_TRANS_SWAPEXT,		"SWAPEXT" }, \
-	{ XFS_TRANS_CHECKPOINT,		"CHECKPOINT" }, \
-	{ XFS_TRANS_ICREATE,		"ICREATE" }, \
-	{ XFS_TRANS_CREATE_TMPFILE,	"CREATE_TMPFILE" }, \
-	{ XLOG_UNMOUNT_REC_TYPE,	"UNMOUNT" }
-
 /*
  * This structure is used to track log items associated with
  * a transaction.  It points to the log item and keeps some
@@ -181,8 +84,9 @@ int	xfs_log_calc_minimum_size(struct xfs_mount *);
 #define	XFS_TRANS_SYNC		0x08	/* make commit synchronous */
 #define XFS_TRANS_DQ_DIRTY	0x10	/* at least one dquot in trx dirty */
 #define XFS_TRANS_RESERVE	0x20    /* OK to use reserved data blocks */
-#define XFS_TRANS_FREEZE_PROT	0x40	/* Transaction has elevated writer
-					   count in superblock */
+#define XFS_TRANS_NO_WRITECOUNT 0x40	/* do not elevate SB writecount */
+#define XFS_TRANS_NOFS		0x80	/* pass KM_NOFS to kmem_alloc */
+
 /*
  * Field values for xfs_trans_mod_sb.
  */

fs/xfs/xfs_aops.c

@@ -84,23 +84,71 @@ xfs_find_bdev_for_inode(
 }
 
 /*
- * We're now finished for good with this ioend structure.
- * Update the page state via the associated buffer_heads,
- * release holds on the inode and bio, and finally free
- * up memory.  Do not use the ioend after this.
+ * We're now finished for good with this page.  Update the page state via the
+ * associated buffer_heads, paying attention to the start and end offsets that
+ * we need to process on the page.
+ */
+static void
+xfs_finish_page_writeback(
+	struct inode		*inode,
+	struct bio_vec		*bvec,
+	int			error)
+{
+	unsigned int		end = bvec->bv_offset + bvec->bv_len - 1;
+	struct buffer_head	*head, *bh;
+	unsigned int		off = 0;
+
+	ASSERT(bvec->bv_offset < PAGE_SIZE);
+	ASSERT((bvec->bv_offset & ((1 << inode->i_blkbits) - 1)) == 0);
+	ASSERT(end < PAGE_SIZE);
+	ASSERT((bvec->bv_len & ((1 << inode->i_blkbits) - 1)) == 0);
+
+	bh = head = page_buffers(bvec->bv_page);
+
+	do {
+		if (off < bvec->bv_offset)
+			goto next_bh;
+		if (off > end)
+			break;
+		bh->b_end_io(bh, !error);
+next_bh:
+		off += bh->b_size;
+	} while ((bh = bh->b_this_page) != head);
+}
+
+/*
+ * We're now finished for good with this ioend structure.  Update the page
+ * state, release holds on bios, and finally free up memory.  Do not use the
+ * ioend after this.
  */
 STATIC void
 xfs_destroy_ioend(
-	xfs_ioend_t		*ioend)
+	struct xfs_ioend	*ioend,
+	int			error)
 {
-	struct buffer_head	*bh, *next;
+	struct inode		*inode = ioend->io_inode;
+	struct bio		*last = ioend->io_bio;
+	struct bio		*bio, *next;
 
-	for (bh = ioend->io_buffer_head; bh; bh = next) {
-		next = bh->b_private;
-		bh->b_end_io(bh, !ioend->io_error);
-	}
+	for (bio = &ioend->io_inline_bio; bio; bio = next) {
+		struct bio_vec	*bvec;
+		int		i;
+
+		/*
+		 * For the last bio, bi_private points to the ioend, so we
+		 * need to explicitly end the iteration here.
+		 */
+		if (bio == last)
+			next = NULL;
+		else
+			next = bio->bi_private;
 
-	mempool_free(ioend, xfs_ioend_pool);
+		/* walk each page on bio, ending page IO on them */
+		bio_for_each_segment_all(bvec, bio, i)
+			xfs_finish_page_writeback(inode, bvec, error);
+
+		bio_put(bio);
+	}
 }
 
 /*
@@ -120,13 +168,9 @@ xfs_setfilesize_trans_alloc(
 	struct xfs_trans	*tp;
 	int			error;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
-
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_fsyncts, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_fsyncts, 0, 0, 0, &tp);
+	if (error)
 		return error;
-	}
 
 	ioend->io_append_trans = tp;
 
@@ -174,7 +218,8 @@ xfs_setfilesize(
 
 STATIC int
 xfs_setfilesize_ioend(
-	struct xfs_ioend	*ioend)
+	struct xfs_ioend	*ioend,
+	int			error)
 {
 	struct xfs_inode	*ip = XFS_I(ioend->io_inode);
 	struct xfs_trans	*tp = ioend->io_append_trans;
@@ -188,36 +233,14 @@ xfs_setfilesize_ioend(
 	__sb_writers_acquired(VFS_I(ip)->i_sb, SB_FREEZE_FS);
 
 	/* we abort the update if there was an IO error */
-	if (ioend->io_error) {
+	if (error) {
 		xfs_trans_cancel(tp);
-		return ioend->io_error;
+		return error;
 	}
 
 	return xfs_setfilesize(ip, tp, ioend->io_offset, ioend->io_size);
 }
 
-/*
- * 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(
-	struct xfs_ioend	*ioend)
-{
-	if (atomic_dec_and_test(&ioend->io_remaining)) {
-		struct xfs_mount	*mp = XFS_I(ioend->io_inode)->i_mount;
-
-		if (ioend->io_type == XFS_IO_UNWRITTEN)
-			queue_work(mp->m_unwritten_workqueue, &ioend->io_work);
-		else if (ioend->io_append_trans)
-			queue_work(mp->m_data_workqueue, &ioend->io_work);
-		else
-			xfs_destroy_ioend(ioend);
-	}
-}
-
 /*
  * IO write completion.
  */
@@ -225,16 +248,17 @@ STATIC void
 xfs_end_io(
 	struct work_struct *work)
 {
-	xfs_ioend_t	*ioend = container_of(work, xfs_ioend_t, io_work);
-	struct xfs_inode *ip = XFS_I(ioend->io_inode);
-	int		error = 0;
+	struct xfs_ioend	*ioend =
+		container_of(work, struct xfs_ioend, io_work);
+	struct xfs_inode	*ip = XFS_I(ioend->io_inode);
+	int			error = ioend->io_bio->bi_error;
 
 	/*
 	 * Set an error if the mount has shut down and proceed with end I/O
 	 * processing so it can perform whatever cleanups are necessary.
 	 */
 	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
-		ioend->io_error = -EIO;
+		error = -EIO;
 
 	/*
 	 * For unwritten extents we need to issue transactions to convert a
@@ -244,55 +268,33 @@ xfs_end_io(
 	 * on error.
 	 */
 	if (ioend->io_type == XFS_IO_UNWRITTEN) {
-		if (ioend->io_error)
+		if (error)
 			goto done;
 		error = xfs_iomap_write_unwritten(ip, ioend->io_offset,
 						  ioend->io_size);
 	} else if (ioend->io_append_trans) {
-		error = xfs_setfilesize_ioend(ioend);
+		error = xfs_setfilesize_ioend(ioend, error);
 	} else {
 		ASSERT(!xfs_ioend_is_append(ioend));
 	}
 
 done:
-	if (error)
-		ioend->io_error = error;
-	xfs_destroy_ioend(ioend);
+	xfs_destroy_ioend(ioend, error);
 }
 
-/*
- * Allocate and initialise an IO completion structure.
- * We need to track unwritten extent write completion here initially.
- * We'll need to extend this for updating the ondisk inode size later
- * (vs. incore size).
- */
-STATIC xfs_ioend_t *
-xfs_alloc_ioend(
-	struct inode		*inode,
-	unsigned int		type)
+STATIC void
+xfs_end_bio(
+	struct bio		*bio)
 {
-	xfs_ioend_t		*ioend;
-
-	ioend = mempool_alloc(xfs_ioend_pool, GFP_NOFS);
-
-	/*
-	 * Set the count to 1 initially, which will prevent an I/O
-	 * completion callback from happening before we have started
-	 * all the I/O from calling the completion routine too early.
-	 */
-	atomic_set(&ioend->io_remaining, 1);
-	ioend->io_error = 0;
-	INIT_LIST_HEAD(&ioend->io_list);
-	ioend->io_type = type;
-	ioend->io_inode = inode;
-	ioend->io_buffer_head = NULL;
-	ioend->io_buffer_tail = NULL;
-	ioend->io_offset = 0;
-	ioend->io_size = 0;
-	ioend->io_append_trans = NULL;
+	struct xfs_ioend	*ioend = bio->bi_private;
+	struct xfs_mount	*mp = XFS_I(ioend->io_inode)->i_mount;
 
-	INIT_WORK(&ioend->io_work, xfs_end_io);
-	return ioend;
+	if (ioend->io_type == XFS_IO_UNWRITTEN)
+		queue_work(mp->m_unwritten_workqueue, &ioend->io_work);
+	else if (ioend->io_append_trans)
+		queue_work(mp->m_data_workqueue, &ioend->io_work);
+	else
+		xfs_destroy_ioend(ioend, bio->bi_error);
 }
 
 STATIC int
@@ -364,50 +366,6 @@ xfs_imap_valid(
 		offset < imap->br_startoff + imap->br_blockcount;
 }
 
-/*
- * BIO completion handler for buffered IO.
- */
-STATIC void
-xfs_end_bio(
-	struct bio		*bio)
-{
-	xfs_ioend_t		*ioend = bio->bi_private;
-
-	if (!ioend->io_error)
-		ioend->io_error = bio->bi_error;
-
-	/* Toss bio and pass work off to an xfsdatad thread */
-	bio->bi_private = NULL;
-	bio->bi_end_io = NULL;
-	bio_put(bio);
-
-	xfs_finish_ioend(ioend);
-}
-
-STATIC void
-xfs_submit_ioend_bio(
-	struct writeback_control *wbc,
-	xfs_ioend_t		*ioend,
-	struct bio		*bio)
-{
-	atomic_inc(&ioend->io_remaining);
-	bio->bi_private = ioend;
-	bio->bi_end_io = xfs_end_bio;
-	submit_bio(wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE, bio);
-}
-
-STATIC struct bio *
-xfs_alloc_ioend_bio(
-	struct buffer_head	*bh)
-{
-	struct bio		*bio = bio_alloc(GFP_NOIO, BIO_MAX_PAGES);
-
-	ASSERT(bio->bi_private == NULL);
-	bio->bi_iter.bi_sector = bh->b_blocknr * (bh->b_size >> 9);
-	bio->bi_bdev = bh->b_bdev;
-	return bio;
-}
-
 STATIC void
 xfs_start_buffer_writeback(
 	struct buffer_head	*bh)
@@ -452,28 +410,35 @@ static inline int xfs_bio_add_buffer(struct bio *bio, struct buffer_head *bh)
 }
 
 /*
- * Submit all of the bios for an ioend. We are only passed a single ioend at a
- * time; the caller is responsible for chaining prior to submission.
+ * Submit the bio for an ioend. We are passed an ioend with a bio attached to
+ * it, and we submit that bio. The ioend may be used for multiple bio
+ * submissions, so we only want to allocate an append transaction for the ioend
+ * once. In the case of multiple bio submission, each bio will take an IO
+ * reference to the ioend to ensure that the ioend completion is only done once
+ * all bios have been submitted and the ioend is really done.
  *
  * If @fail is non-zero, it means that we have a situation where some part of
  * the submission process has failed after we have marked paged for writeback
- * and unlocked them. In this situation, we need to fail the ioend chain rather
- * than submit it to IO. This typically only happens on a filesystem shutdown.
+ * and unlocked them. In this situation, we need to fail the bio and ioend
+ * rather than submit it to IO. This typically only happens on a filesystem
+ * shutdown.
  */
 STATIC int
 xfs_submit_ioend(
 	struct writeback_control *wbc,
-	xfs_ioend_t		*ioend,
+	struct xfs_ioend	*ioend,
 	int			status)
 {
-	struct buffer_head	*bh;
-	struct bio		*bio;
-	sector_t		lastblock = 0;
-
 	/* Reserve log space if we might write beyond the on-disk inode size. */
 	if (!status &&
-	     ioend->io_type != XFS_IO_UNWRITTEN && xfs_ioend_is_append(ioend))
+	    ioend->io_type != XFS_IO_UNWRITTEN &&
+	    xfs_ioend_is_append(ioend) &&
+	    !ioend->io_append_trans)
 		status = xfs_setfilesize_trans_alloc(ioend);
+
+	ioend->io_bio->bi_private = ioend;
+	ioend->io_bio->bi_end_io = xfs_end_bio;
+
 	/*
 	 * If we are failing the IO now, just mark the ioend with an
 	 * error and finish it. This will run IO completion immediately
@@ -481,33 +446,73 @@ xfs_submit_ioend(
 	 * time.
 	 */
 	if (status) {
-		ioend->io_error = status;
-		xfs_finish_ioend(ioend);
+		ioend->io_bio->bi_error = status;
+		bio_endio(ioend->io_bio);
 		return status;
 	}
 
-	bio = NULL;
-	for (bh = ioend->io_buffer_head; bh; bh = bh->b_private) {
+	submit_bio(wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE,
+		   ioend->io_bio);
+	return 0;
+}
 
-		if (!bio) {
-retry:
-			bio = xfs_alloc_ioend_bio(bh);
-		} else if (bh->b_blocknr != lastblock + 1) {
-			xfs_submit_ioend_bio(wbc, ioend, bio);
-			goto retry;
-		}
+static void
+xfs_init_bio_from_bh(
+	struct bio		*bio,
+	struct buffer_head	*bh)
+{
+	bio->bi_iter.bi_sector = bh->b_blocknr * (bh->b_size >> 9);
+	bio->bi_bdev = bh->b_bdev;
+}
 
-		if (xfs_bio_add_buffer(bio, bh) != bh->b_size) {
-			xfs_submit_ioend_bio(wbc, ioend, bio);
-			goto retry;
-		}
+static struct xfs_ioend *
+xfs_alloc_ioend(
+	struct inode		*inode,
+	unsigned int		type,
+	xfs_off_t		offset,
+	struct buffer_head	*bh)
+{
+	struct xfs_ioend	*ioend;
+	struct bio		*bio;
 
-		lastblock = bh->b_blocknr;
-	}
-	if (bio)
-		xfs_submit_ioend_bio(wbc, ioend, bio);
-	xfs_finish_ioend(ioend);
-	return 0;
+	bio = bio_alloc_bioset(GFP_NOFS, BIO_MAX_PAGES, xfs_ioend_bioset);
+	xfs_init_bio_from_bh(bio, bh);
+
+	ioend = container_of(bio, struct xfs_ioend, io_inline_bio);
+	INIT_LIST_HEAD(&ioend->io_list);
+	ioend->io_type = type;
+	ioend->io_inode = inode;
+	ioend->io_size = 0;
+	ioend->io_offset = offset;
+	INIT_WORK(&ioend->io_work, xfs_end_io);
+	ioend->io_append_trans = NULL;
+	ioend->io_bio = bio;
+	return ioend;
+}
+
+/*
+ * Allocate a new bio, and chain the old bio to the new one.
+ *
+ * Note that we have to do perform the chaining in this unintuitive order
+ * so that the bi_private linkage is set up in the right direction for the
+ * traversal in xfs_destroy_ioend().
+ */
+static void
+xfs_chain_bio(
+	struct xfs_ioend	*ioend,
+	struct writeback_control *wbc,
+	struct buffer_head	*bh)
+{
+	struct bio *new;
+
+	new = bio_alloc(GFP_NOFS, BIO_MAX_PAGES);
+	xfs_init_bio_from_bh(new, bh);
+
+	bio_chain(ioend->io_bio, new);
+	bio_get(ioend->io_bio);		/* for xfs_destroy_ioend */
+	submit_bio(wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE,
+		   ioend->io_bio);
+	ioend->io_bio = new;
 }
 
 /*
@@ -523,27 +528,24 @@ xfs_add_to_ioend(
 	struct buffer_head	*bh,
 	xfs_off_t		offset,
 	struct xfs_writepage_ctx *wpc,
+	struct writeback_control *wbc,
 	struct list_head	*iolist)
 {
 	if (!wpc->ioend || wpc->io_type != wpc->ioend->io_type ||
 	    bh->b_blocknr != wpc->last_block + 1 ||
 	    offset != wpc->ioend->io_offset + wpc->ioend->io_size) {
-		struct xfs_ioend	*new;
-
 		if (wpc->ioend)
 			list_add(&wpc->ioend->io_list, iolist);
-
-		new = xfs_alloc_ioend(inode, wpc->io_type);
-		new->io_offset = offset;
-		new->io_buffer_head = bh;
-		new->io_buffer_tail = bh;
-		wpc->ioend = new;
-	} else {
-		wpc->ioend->io_buffer_tail->b_private = bh;
-		wpc->ioend->io_buffer_tail = bh;
+		wpc->ioend = xfs_alloc_ioend(inode, wpc->io_type, offset, bh);
 	}
 
-	bh->b_private = NULL;
+	/*
+	 * If the buffer doesn't fit into the bio we need to allocate a new
+	 * one.  This shouldn't happen more than once for a given buffer.
+	 */
+	while (xfs_bio_add_buffer(wpc->ioend->io_bio, bh) != bh->b_size)
+		xfs_chain_bio(wpc->ioend, wbc, bh);
+
 	wpc->ioend->io_size += bh->b_size;
 	wpc->last_block = bh->b_blocknr;
 	xfs_start_buffer_writeback(bh);
@@ -803,7 +805,7 @@ xfs_writepage_map(
 			lock_buffer(bh);
 			if (wpc->io_type != XFS_IO_OVERWRITE)
 				xfs_map_at_offset(inode, bh, &wpc->imap, offset);
-			xfs_add_to_ioend(inode, bh, offset, wpc, &submit_list);
+			xfs_add_to_ioend(inode, bh, offset, wpc, wbc, &submit_list);
 			count++;
 		}
 
@@ -1391,13 +1393,10 @@ xfs_end_io_direct_write(
 
 		trace_xfs_end_io_direct_write_append(ip, offset, size);
 
-		tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_fsyncts, 0, 0);
-		if (error) {
-			xfs_trans_cancel(tp);
-			return error;
-		}
-		error = xfs_setfilesize(ip, tp, offset, size);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_fsyncts, 0, 0, 0,
+				&tp);
+		if (!error)
+			error = xfs_setfilesize(ip, tp, offset, size);
 	}
 
 	return error;

fs/xfs/xfs_aops.h

@@ -18,7 +18,7 @@
 #ifndef __XFS_AOPS_H__
 #define __XFS_AOPS_H__
 
-extern mempool_t *xfs_ioend_pool;
+extern struct bio_set *xfs_ioend_bioset;
 
 /*
  * Types of I/O for bmap clustering and I/O completion tracking.
@@ -37,22 +37,19 @@ enum {
 	{ XFS_IO_OVERWRITE,		"overwrite" }
 
 /*
- * xfs_ioend struct manages large extent writes for XFS.
- * It can manage several multi-page bio's at once.
+ * Structure for buffered I/O completions.
  */
-typedef struct xfs_ioend {
+struct xfs_ioend {
 	struct list_head	io_list;	/* next ioend in chain */
 	unsigned int		io_type;	/* delalloc / unwritten */
-	int			io_error;	/* I/O error code */
-	atomic_t		io_remaining;	/* hold count */
 	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 */
 	size_t			io_size;	/* size of the extent */
 	xfs_off_t		io_offset;	/* offset in the file */
 	struct work_struct	io_work;	/* xfsdatad work queue */
 	struct xfs_trans	*io_append_trans;/* xact. for size update */
-} xfs_ioend_t;
+	struct bio		*io_bio;	/* bio being built */
+	struct bio		io_inline_bio;	/* MUST BE LAST! */
+};
 
 extern const struct address_space_operations xfs_address_space_operations;
 

fs/xfs/xfs_attr.h

@@ -112,8 +112,9 @@ typedef struct attrlist_cursor_kern {
  *========================================================================*/
 
 
+/* Return 0 on success, or -errno; other state communicated via *context */
 typedef int (*put_listent_func_t)(struct xfs_attr_list_context *, int,
-			      unsigned char *, int, int, unsigned char *);
+			      unsigned char *, int, int);
 
 typedef struct xfs_attr_list_context {
 	struct xfs_inode		*dp;		/* inode */
@@ -126,7 +127,6 @@ typedef struct xfs_attr_list_context {
 	int				firstu;		/* first used byte in buffer */
 	int				flags;		/* from VOP call */
 	int				resynch;	/* T/F: resynch with cursor */
-	int				put_value;	/* T/F: need value for listent */
 	put_listent_func_t		put_listent;	/* list output fmt function */
 	int				index;		/* index into output buffer */
 } xfs_attr_list_context_t;

fs/xfs/xfs_attr_inactive.c

@@ -405,21 +405,11 @@ xfs_attr_inactive(
 		goto out_destroy_fork;
 	xfs_iunlock(dp, lock_mode);
 
-	/*
-	 * Start our first transaction of the day.
-	 *
-	 * All future transactions during this code must be "chained" off
-	 * this one via the trans_dup() call.  All transactions will contain
-	 * the inode, and the inode will always be marked with trans_ihold().
-	 * Since the inode will be locked in all transactions, we must log
-	 * the inode in every transaction to let it float upward through
-	 * the log.
-	 */
 	lock_mode = 0;
-	trans = xfs_trans_alloc(mp, XFS_TRANS_ATTRINVAL);
-	error = xfs_trans_reserve(trans, &M_RES(mp)->tr_attrinval, 0, 0);
+
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_attrinval, 0, 0, 0, &trans);
 	if (error)
-		goto out_cancel;
+		goto out_destroy_fork;
 
 	lock_mode = XFS_ILOCK_EXCL;
 	xfs_ilock(dp, lock_mode);

fs/xfs/xfs_attr_list.c

@@ -106,18 +106,15 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context)
 					   sfe->flags,
 					   sfe->nameval,
 					   (int)sfe->namelen,
-					   (int)sfe->valuelen,
-					   &sfe->nameval[sfe->namelen]);
-
+					   (int)sfe->valuelen);
+			if (error)
+				return error;
 			/*
 			 * Either search callback finished early or
 			 * didn't fit it all in the buffer after all.
 			 */
 			if (context->seen_enough)
 				break;
-
-			if (error)
-				return error;
 			sfe = XFS_ATTR_SF_NEXTENTRY(sfe);
 		}
 		trace_xfs_attr_list_sf_all(context);
@@ -200,8 +197,7 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context)
 					sbp->flags,
 					sbp->name,
 					sbp->namelen,
-					sbp->valuelen,
-					&sbp->name[sbp->namelen]);
+					sbp->valuelen);
 		if (error) {
 			kmem_free(sbuf);
 			return error;
@@ -416,6 +412,9 @@ xfs_attr3_leaf_list_int(
 	 */
 	retval = 0;
 	for (; i < ichdr.count; entry++, i++) {
+		char *name;
+		int namelen, valuelen;
+
 		if (be32_to_cpu(entry->hashval) != cursor->hashval) {
 			cursor->hashval = be32_to_cpu(entry->hashval);
 			cursor->offset = 0;
@@ -425,56 +424,25 @@ xfs_attr3_leaf_list_int(
 			continue;		/* skip incomplete entries */
 
 		if (entry->flags & XFS_ATTR_LOCAL) {
-			xfs_attr_leaf_name_local_t *name_loc =
-				xfs_attr3_leaf_name_local(leaf, i);
-
-			retval = context->put_listent(context,
-						entry->flags,
-						name_loc->nameval,
-						(int)name_loc->namelen,
-						be16_to_cpu(name_loc->valuelen),
-						&name_loc->nameval[name_loc->namelen]);
-			if (retval)
-				return retval;
+			xfs_attr_leaf_name_local_t *name_loc;
+
+			name_loc = xfs_attr3_leaf_name_local(leaf, i);
+			name = name_loc->nameval;
+			namelen = name_loc->namelen;
+			valuelen = be16_to_cpu(name_loc->valuelen);
 		} else {
-			xfs_attr_leaf_name_remote_t *name_rmt =
-				xfs_attr3_leaf_name_remote(leaf, i);
-
-			int valuelen = be32_to_cpu(name_rmt->valuelen);
-
-			if (context->put_value) {
-				xfs_da_args_t args;
-
-				memset((char *)&args, 0, sizeof(args));
-				args.geo = context->dp->i_mount->m_attr_geo;
-				args.dp = context->dp;
-				args.whichfork = XFS_ATTR_FORK;
-				args.valuelen = valuelen;
-				args.rmtvaluelen = valuelen;
-				args.value = kmem_alloc(valuelen, KM_SLEEP | KM_NOFS);
-				args.rmtblkno = be32_to_cpu(name_rmt->valueblk);
-				args.rmtblkcnt = xfs_attr3_rmt_blocks(
-							args.dp->i_mount, valuelen);
-				retval = xfs_attr_rmtval_get(&args);
-				if (!retval)
-					retval = context->put_listent(context,
-							entry->flags,
-							name_rmt->name,
-							(int)name_rmt->namelen,
-							valuelen,
-							args.value);
-				kmem_free(args.value);
-			} else {
-				retval = context->put_listent(context,
-						entry->flags,
-						name_rmt->name,
-						(int)name_rmt->namelen,
-						valuelen,
-						NULL);
-			}
-			if (retval)
-				return retval;
+			xfs_attr_leaf_name_remote_t *name_rmt;
+
+			name_rmt = xfs_attr3_leaf_name_remote(leaf, i);
+			name = name_rmt->name;
+			namelen = name_rmt->namelen;
+			valuelen = be32_to_cpu(name_rmt->valuelen);
 		}
+
+		retval = context->put_listent(context, entry->flags,
+					      name, namelen, valuelen);
+		if (retval)
+			break;
 		if (context->seen_enough)
 			break;
 		cursor->offset++;
@@ -551,8 +519,7 @@ xfs_attr_put_listent(
 	int		flags,
 	unsigned char	*name,
 	int		namelen,
-	int		valuelen,
-	unsigned char	*value)
+	int		valuelen)
 {
 	struct attrlist *alist = (struct attrlist *)context->alist;
 	attrlist_ent_t *aep;
@@ -581,7 +548,7 @@ xfs_attr_put_listent(
 		trace_xfs_attr_list_full(context);
 		alist->al_more = 1;
 		context->seen_enough = 1;
-		return 1;
+		return 0;
 	}
 
 	aep = (attrlist_ent_t *)&context->alist[context->firstu];

fs/xfs/xfs_bmap_util.c

@@ -900,19 +900,15 @@ xfs_free_eofblocks(
 		 * Free them up now by truncating the file to
 		 * its current size.
 		 */
-		tp = xfs_trans_alloc(mp, XFS_TRANS_INACTIVE);
-
 		if (need_iolock) {
-			if (!xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL)) {
-				xfs_trans_cancel(tp);
+			if (!xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL))
 				return -EAGAIN;
-			}
 		}
 
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0,
+				&tp);
 		if (error) {
 			ASSERT(XFS_FORCED_SHUTDOWN(mp));
-			xfs_trans_cancel(tp);
 			if (need_iolock)
 				xfs_iunlock(ip, XFS_IOLOCK_EXCL);
 			return error;
@@ -1037,9 +1033,9 @@ xfs_alloc_file_space(
 		/*
 		 * Allocate and setup the transaction.
 		 */
-		tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write,
-					  resblks, resrtextents);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks,
+				resrtextents, 0, &tp);
+
 		/*
 		 * Check for running out of space
 		 */
@@ -1048,7 +1044,6 @@ xfs_alloc_file_space(
 			 * Free the transaction structure.
 			 */
 			ASSERT(error == -ENOSPC || XFS_FORCED_SHUTDOWN(mp));
-			xfs_trans_cancel(tp);
 			break;
 		}
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
@@ -1311,18 +1306,10 @@ xfs_free_file_space(
 		 * transaction to dip into the reserve blocks to ensure
 		 * the freeing of the space succeeds at ENOSPC.
 		 */
-		tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write, resblks, 0);
-
-		/*
-		 * check for running out of space
-		 */
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks, 0, 0,
+				&tp);
 		if (error) {
-			/*
-			 * Free the transaction structure.
-			 */
 			ASSERT(error == -ENOSPC || XFS_FORCED_SHUTDOWN(mp));
-			xfs_trans_cancel(tp);
 			break;
 		}
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
@@ -1482,19 +1469,16 @@ xfs_shift_file_space(
 	}
 
 	while (!error && !done) {
-		tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
 		/*
 		 * We would need to reserve permanent block for transaction.
 		 * This will come into picture when after shifting extent into
 		 * hole we found that adjacent extents can be merged which
 		 * may lead to freeing of a block during record update.
 		 */
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write,
-				XFS_DIOSTRAT_SPACE_RES(mp, 0), 0);
-		if (error) {
-			xfs_trans_cancel(tp);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write,
+				XFS_DIOSTRAT_SPACE_RES(mp, 0), 0, 0, &tp);
+		if (error)
 			break;
-		}
 
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
 		error = xfs_trans_reserve_quota(tp, mp, ip->i_udquot,
@@ -1747,12 +1731,9 @@ xfs_swap_extents(
 	if (error)
 		goto out_unlock;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SWAPEXT);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0, 0, 0, &tp);
+	if (error)
 		goto out_unlock;
-	}
 
 	/*
 	 * Lock and join the inodes to the tansaction so that transaction commit

fs/xfs/xfs_buf.c

@@ -1100,22 +1100,18 @@ xfs_bwrite(
 	return error;
 }
 
-STATIC void
+static void
 xfs_buf_bio_end_io(
 	struct bio		*bio)
 {
-	xfs_buf_t		*bp = (xfs_buf_t *)bio->bi_private;
+	struct xfs_buf		*bp = (struct xfs_buf *)bio->bi_private;
 
 	/*
 	 * don't overwrite existing errors - otherwise we can lose errors on
 	 * buffers that require multiple bios to complete.
 	 */
-	if (bio->bi_error) {
-		spin_lock(&bp->b_lock);
-		if (!bp->b_io_error)
-			bp->b_io_error = bio->bi_error;
-		spin_unlock(&bp->b_lock);
-	}
+	if (bio->bi_error)
+		cmpxchg(&bp->b_io_error, 0, bio->bi_error);
 
 	if (!bp->b_error && xfs_buf_is_vmapped(bp) && (bp->b_flags & XBF_READ))
 		invalidate_kernel_vmap_range(bp->b_addr, xfs_buf_vmap_len(bp));

fs/xfs/xfs_buf.h

@@ -183,6 +183,26 @@ typedef struct xfs_buf {
 	unsigned int		b_page_count;	/* size of page array */
 	unsigned int		b_offset;	/* page offset in first page */
 	int			b_error;	/* error code on I/O */
+
+	/*
+	 * async write failure retry count. Initialised to zero on the first
+	 * failure, then when it exceeds the maximum configured without a
+	 * success the write is considered to be failed permanently and the
+	 * iodone handler will take appropriate action.
+	 *
+	 * For retry timeouts, we record the jiffie of the first failure. This
+	 * means that we can change the retry timeout for buffers already under
+	 * I/O and thus avoid getting stuck in a retry loop with a long timeout.
+	 *
+	 * last_error is used to ensure that we are getting repeated errors, not
+	 * different errors. e.g. a block device might change ENOSPC to EIO when
+	 * a failure timeout occurs, so we want to re-initialise the error
+	 * retry behaviour appropriately when that happens.
+	 */
+	int			b_retries;
+	unsigned long		b_first_retry_time; /* in jiffies */
+	int			b_last_error;
+
 	const struct xfs_buf_ops	*b_ops;
 
 #ifdef XFS_BUF_LOCK_TRACKING

fs/xfs/xfs_buf_item.c

@@ -1042,35 +1042,22 @@ xfs_buf_do_callbacks(
 	}
 }
 
-/*
- * This is the iodone() function for buffers which have had callbacks
- * attached to them by xfs_buf_attach_iodone().  It should remove each
- * log item from the buffer's list and call the callback of each in turn.
- * When done, the buffer's fsprivate field is set to NULL and the buffer
- * is unlocked with a call to iodone().
- */
-void
-xfs_buf_iodone_callbacks(
+static bool
+xfs_buf_iodone_callback_error(
 	struct xfs_buf		*bp)
 {
 	struct xfs_log_item	*lip = bp->b_fspriv;
 	struct xfs_mount	*mp = lip->li_mountp;
 	static ulong		lasttime;
 	static xfs_buftarg_t	*lasttarg;
-
-	if (likely(!bp->b_error))
-		goto do_callbacks;
+	struct xfs_error_cfg	*cfg;
 
 	/*
 	 * If we've already decided to shutdown the filesystem because of
 	 * I/O errors, there's no point in giving this a retry.
 	 */
-	if (XFS_FORCED_SHUTDOWN(mp)) {
-		xfs_buf_stale(bp);
-		bp->b_flags |= XBF_DONE;
-		trace_xfs_buf_item_iodone(bp, _RET_IP_);
-		goto do_callbacks;
-	}
+	if (XFS_FORCED_SHUTDOWN(mp))
+		goto out_stale;
 
 	if (bp->b_target != lasttarg ||
 	    time_after(jiffies, (lasttime + 5*HZ))) {
@@ -1079,45 +1066,93 @@ xfs_buf_iodone_callbacks(
 	}
 	lasttarg = bp->b_target;
 
+	/* synchronous writes will have callers process the error */
+	if (!(bp->b_flags & XBF_ASYNC))
+		goto out_stale;
+
+	trace_xfs_buf_item_iodone_async(bp, _RET_IP_);
+	ASSERT(bp->b_iodone != NULL);
+
 	/*
 	 * If the write was asynchronous then no one will be looking for the
-	 * error.  Clear the error state and write the buffer out again.
-	 *
-	 * XXX: This helps against transient write errors, but we need to find
-	 * a way to shut the filesystem down if the writes keep failing.
-	 *
-	 * In practice we'll shut the filesystem down soon as non-transient
-	 * errors tend to affect the whole device and a failing log write
-	 * will make us give up.  But we really ought to do better here.
+	 * error.  If this is the first failure of this type, clear the error
+	 * state and write the buffer out again. This means we always retry an
+	 * async write failure at least once, but we also need to set the buffer
+	 * up to behave correctly now for repeated failures.
 	 */
-	if (bp->b_flags & XBF_ASYNC) {
-		ASSERT(bp->b_iodone != NULL);
+	if (!(bp->b_flags & (XBF_STALE|XBF_WRITE_FAIL)) ||
+	     bp->b_last_error != bp->b_error) {
+		bp->b_flags |= (XBF_WRITE | XBF_ASYNC |
+			        XBF_DONE | XBF_WRITE_FAIL);
+		bp->b_last_error = bp->b_error;
+		bp->b_retries = 0;
+		bp->b_first_retry_time = jiffies;
+
+		xfs_buf_ioerror(bp, 0);
+		xfs_buf_submit(bp);
+		return true;
+	}
 
-		trace_xfs_buf_item_iodone_async(bp, _RET_IP_);
+	/*
+	 * Repeated failure on an async write. Take action according to the
+	 * error configuration we have been set up to use.
+	 */
+	cfg = xfs_error_get_cfg(mp, XFS_ERR_METADATA, bp->b_error);
 
-		xfs_buf_ioerror(bp, 0); /* errno of 0 unsets the flag */
+	if (cfg->max_retries != XFS_ERR_RETRY_FOREVER &&
+	    ++bp->b_retries > cfg->max_retries)
+			goto permanent_error;
+	if (cfg->retry_timeout &&
+	    time_after(jiffies, cfg->retry_timeout + bp->b_first_retry_time))
+			goto permanent_error;
 
-		if (!(bp->b_flags & (XBF_STALE|XBF_WRITE_FAIL))) {
-			bp->b_flags |= XBF_WRITE | XBF_ASYNC |
-				       XBF_DONE | XBF_WRITE_FAIL;
-			xfs_buf_submit(bp);
-		} else {
-			xfs_buf_relse(bp);
-		}
+	/* At unmount we may treat errors differently */
+	if ((mp->m_flags & XFS_MOUNT_UNMOUNTING) && mp->m_fail_unmount)
+		goto permanent_error;
 
-		return;
-	}
+	/* still a transient error, higher layers will retry */
+	xfs_buf_ioerror(bp, 0);
+	xfs_buf_relse(bp);
+	return true;
 
 	/*
-	 * If the write of the buffer was synchronous, we want to make
-	 * sure to return the error to the caller of xfs_bwrite().
+	 * Permanent error - we need to trigger a shutdown if we haven't already
+	 * to indicate that inconsistency will result from this action.
 	 */
+permanent_error:
+	xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR);
+out_stale:
 	xfs_buf_stale(bp);
 	bp->b_flags |= XBF_DONE;
-
 	trace_xfs_buf_error_relse(bp, _RET_IP_);
+	return false;
+}
+
+/*
+ * This is the iodone() function for buffers which have had callbacks attached
+ * to them by xfs_buf_attach_iodone(). We need to iterate the items on the
+ * callback list, mark the buffer as having no more callbacks and then push the
+ * buffer through IO completion processing.
+ */
+void
+xfs_buf_iodone_callbacks(
+	struct xfs_buf		*bp)
+{
+	/*
+	 * If there is an error, process it. Some errors require us
+	 * to run callbacks after failure processing is done so we
+	 * detect that and take appropriate action.
+	 */
+	if (bp->b_error && xfs_buf_iodone_callback_error(bp))
+		return;
+
+	/*
+	 * Successful IO or permanent error. Either way, we can clear the
+	 * retry state here in preparation for the next error that may occur.
+	 */
+	bp->b_last_error = 0;
+	bp->b_retries = 0;
 
-do_callbacks:
 	xfs_buf_do_callbacks(bp);
 	bp->b_fspriv = NULL;
 	bp->b_iodone = NULL;

fs/xfs/xfs_dquot.c

@@ -614,11 +614,10 @@ xfs_qm_dqread(
 	trace_xfs_dqread(dqp);
 
 	if (flags & XFS_QMOPT_DQALLOC) {
-		tp = xfs_trans_alloc(mp, XFS_TRANS_QM_DQALLOC);
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_dqalloc,
-					  XFS_QM_DQALLOC_SPACE_RES(mp), 0);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_qm_dqalloc,
+				XFS_QM_DQALLOC_SPACE_RES(mp), 0, 0, &tp);
 		if (error)
-			goto error1;
+			goto error0;
 	}
 
 	/*
@@ -692,7 +691,7 @@ error0:
  * end of the chunk, skip ahead to first id in next allocated chunk
  * using the SEEK_DATA interface.
  */
-int
+static int
 xfs_dq_get_next_id(
 	xfs_mount_t		*mp,
 	uint			type,

fs/xfs/xfs_file.c

@@ -145,12 +145,10 @@ xfs_update_prealloc_flags(
 	struct xfs_trans	*tp;
 	int			error;
 
-	tp = xfs_trans_alloc(ip->i_mount, XFS_TRANS_WRITEID);
-	error = xfs_trans_reserve(tp, &M_RES(ip->i_mount)->tr_writeid, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(ip->i_mount, &M_RES(ip->i_mount)->tr_writeid,
+			0, 0, 0, &tp);
+	if (error)
 		return error;
-	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);

fs/xfs/xfs_fsops.c

@@ -198,14 +198,10 @@ xfs_growfs_data_private(
 			return error;
 	}
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFS);
-	tp->t_flags |= XFS_TRANS_RESERVE;
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_growdata,
-				  XFS_GROWFS_SPACE_RES(mp), 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_growdata,
+			XFS_GROWFS_SPACE_RES(mp), 0, XFS_TRANS_RESERVE, &tp);
+	if (error)
 		return error;
-	}
 
 	/*
 	 * Write new AG headers to disk. Non-transactional, but written
@@ -243,8 +239,8 @@ xfs_growfs_data_private(
 		agf->agf_roots[XFS_BTNUM_CNTi] = cpu_to_be32(XFS_CNT_BLOCK(mp));
 		agf->agf_levels[XFS_BTNUM_BNOi] = cpu_to_be32(1);
 		agf->agf_levels[XFS_BTNUM_CNTi] = cpu_to_be32(1);
-		agf->agf_flfirst = 0;
-		agf->agf_fllast = cpu_to_be32(XFS_AGFL_SIZE(mp) - 1);
+		agf->agf_flfirst = cpu_to_be32(1);
+		agf->agf_fllast = 0;
 		agf->agf_flcount = 0;
 		tmpsize = agsize - XFS_PREALLOC_BLOCKS(mp);
 		agf->agf_freeblks = cpu_to_be32(tmpsize);

fs/xfs/xfs_icache.c

@@ -37,9 +37,6 @@
 #include <linux/kthread.h>
 #include <linux/freezer.h>
 
-STATIC void __xfs_inode_clear_reclaim_tag(struct xfs_mount *mp,
-				struct xfs_perag *pag, struct xfs_inode *ip);
-
 /*
  * Allocate and initialise an xfs_inode.
  */
@@ -94,13 +91,6 @@ xfs_inode_free_callback(
 	struct inode		*inode = container_of(head, struct inode, i_rcu);
 	struct xfs_inode	*ip = XFS_I(inode);
 
-	kmem_zone_free(xfs_inode_zone, ip);
-}
-
-void
-xfs_inode_free(
-	struct xfs_inode	*ip)
-{
 	switch (VFS_I(ip)->i_mode & S_IFMT) {
 	case S_IFREG:
 	case S_IFDIR:
@@ -118,6 +108,25 @@ xfs_inode_free(
 		ip->i_itemp = NULL;
 	}
 
+	kmem_zone_free(xfs_inode_zone, ip);
+}
+
+static void
+__xfs_inode_free(
+	struct xfs_inode	*ip)
+{
+	/* asserts to verify all state is correct here */
+	ASSERT(atomic_read(&ip->i_pincount) == 0);
+	ASSERT(!xfs_isiflocked(ip));
+	XFS_STATS_DEC(ip->i_mount, vn_active);
+
+	call_rcu(&VFS_I(ip)->i_rcu, xfs_inode_free_callback);
+}
+
+void
+xfs_inode_free(
+	struct xfs_inode	*ip)
+{
 	/*
 	 * Because we use RCU freeing we need to ensure the inode always
 	 * appears to be reclaimed with an invalid inode number when in the
@@ -129,12 +138,123 @@ xfs_inode_free(
 	ip->i_ino = 0;
 	spin_unlock(&ip->i_flags_lock);
 
-	/* asserts to verify all state is correct here */
-	ASSERT(atomic_read(&ip->i_pincount) == 0);
-	ASSERT(!xfs_isiflocked(ip));
-	XFS_STATS_DEC(ip->i_mount, vn_active);
+	__xfs_inode_free(ip);
+}
 
-	call_rcu(&VFS_I(ip)->i_rcu, xfs_inode_free_callback);
+/*
+ * Queue a new inode reclaim pass if there are reclaimable inodes and there
+ * isn't a reclaim pass already in progress. By default it runs every 5s based
+ * on the xfs periodic sync default of 30s. Perhaps this should have it's own
+ * tunable, but that can be done if this method proves to be ineffective or too
+ * aggressive.
+ */
+static void
+xfs_reclaim_work_queue(
+	struct xfs_mount        *mp)
+{
+
+	rcu_read_lock();
+	if (radix_tree_tagged(&mp->m_perag_tree, XFS_ICI_RECLAIM_TAG)) {
+		queue_delayed_work(mp->m_reclaim_workqueue, &mp->m_reclaim_work,
+			msecs_to_jiffies(xfs_syncd_centisecs / 6 * 10));
+	}
+	rcu_read_unlock();
+}
+
+/*
+ * This is a fast pass over the inode cache to try to get reclaim moving on as
+ * many inodes as possible in a short period of time. It kicks itself every few
+ * seconds, as well as being kicked by the inode cache shrinker when memory
+ * goes low. It scans as quickly as possible avoiding locked inodes or those
+ * already being flushed, and once done schedules a future pass.
+ */
+void
+xfs_reclaim_worker(
+	struct work_struct *work)
+{
+	struct xfs_mount *mp = container_of(to_delayed_work(work),
+					struct xfs_mount, m_reclaim_work);
+
+	xfs_reclaim_inodes(mp, SYNC_TRYLOCK);
+	xfs_reclaim_work_queue(mp);
+}
+
+static void
+xfs_perag_set_reclaim_tag(
+	struct xfs_perag	*pag)
+{
+	struct xfs_mount	*mp = pag->pag_mount;
+
+	ASSERT(spin_is_locked(&pag->pag_ici_lock));
+	if (pag->pag_ici_reclaimable++)
+		return;
+
+	/* propagate the reclaim tag up into the perag radix tree */
+	spin_lock(&mp->m_perag_lock);
+	radix_tree_tag_set(&mp->m_perag_tree, pag->pag_agno,
+			   XFS_ICI_RECLAIM_TAG);
+	spin_unlock(&mp->m_perag_lock);
+
+	/* schedule periodic background inode reclaim */
+	xfs_reclaim_work_queue(mp);
+
+	trace_xfs_perag_set_reclaim(mp, pag->pag_agno, -1, _RET_IP_);
+}
+
+static void
+xfs_perag_clear_reclaim_tag(
+	struct xfs_perag	*pag)
+{
+	struct xfs_mount	*mp = pag->pag_mount;
+
+	ASSERT(spin_is_locked(&pag->pag_ici_lock));
+	if (--pag->pag_ici_reclaimable)
+		return;
+
+	/* clear the reclaim tag from the perag radix tree */
+	spin_lock(&mp->m_perag_lock);
+	radix_tree_tag_clear(&mp->m_perag_tree, pag->pag_agno,
+			     XFS_ICI_RECLAIM_TAG);
+	spin_unlock(&mp->m_perag_lock);
+	trace_xfs_perag_clear_reclaim(mp, pag->pag_agno, -1, _RET_IP_);
+}
+
+
+/*
+ * We set the inode flag atomically with the radix tree tag.
+ * Once we get tag lookups on the radix tree, this inode flag
+ * can go away.
+ */
+void
+xfs_inode_set_reclaim_tag(
+	struct xfs_inode	*ip)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_perag	*pag;
+
+	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
+	spin_lock(&pag->pag_ici_lock);
+	spin_lock(&ip->i_flags_lock);
+
+	radix_tree_tag_set(&pag->pag_ici_root, XFS_INO_TO_AGINO(mp, ip->i_ino),
+			   XFS_ICI_RECLAIM_TAG);
+	xfs_perag_set_reclaim_tag(pag);
+	__xfs_iflags_set(ip, XFS_IRECLAIMABLE);
+
+	spin_unlock(&ip->i_flags_lock);
+	spin_unlock(&pag->pag_ici_lock);
+	xfs_perag_put(pag);
+}
+
+STATIC void
+xfs_inode_clear_reclaim_tag(
+	struct xfs_perag	*pag,
+	xfs_ino_t		ino)
+{
+	radix_tree_tag_clear(&pag->pag_ici_root,
+			     XFS_INO_TO_AGINO(pag->pag_mount, ino),
+			     XFS_ICI_RECLAIM_TAG);
+	xfs_perag_clear_reclaim_tag(pag);
 }
 
 /*
@@ -264,7 +384,7 @@ xfs_iget_cache_hit(
 		 */
 		ip->i_flags &= ~XFS_IRECLAIM_RESET_FLAGS;
 		ip->i_flags |= XFS_INEW;
-		__xfs_inode_clear_reclaim_tag(mp, pag, ip);
+		xfs_inode_clear_reclaim_tag(pag, ip->i_ino);
 		inode->i_state = I_NEW;
 
 		ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock));
@@ -722,121 +842,6 @@ xfs_inode_ag_iterator_tag(
 	return last_error;
 }
 
-/*
- * Queue a new inode reclaim pass if there are reclaimable inodes and there
- * isn't a reclaim pass already in progress. By default it runs every 5s based
- * on the xfs periodic sync default of 30s. Perhaps this should have it's own
- * tunable, but that can be done if this method proves to be ineffective or too
- * aggressive.
- */
-static void
-xfs_reclaim_work_queue(
-	struct xfs_mount        *mp)
-{
-
-	rcu_read_lock();
-	if (radix_tree_tagged(&mp->m_perag_tree, XFS_ICI_RECLAIM_TAG)) {
-		queue_delayed_work(mp->m_reclaim_workqueue, &mp->m_reclaim_work,
-			msecs_to_jiffies(xfs_syncd_centisecs / 6 * 10));
-	}
-	rcu_read_unlock();
-}
-
-/*
- * This is a fast pass over the inode cache to try to get reclaim moving on as
- * many inodes as possible in a short period of time. It kicks itself every few
- * seconds, as well as being kicked by the inode cache shrinker when memory
- * goes low. It scans as quickly as possible avoiding locked inodes or those
- * already being flushed, and once done schedules a future pass.
- */
-void
-xfs_reclaim_worker(
-	struct work_struct *work)
-{
-	struct xfs_mount *mp = container_of(to_delayed_work(work),
-					struct xfs_mount, m_reclaim_work);
-
-	xfs_reclaim_inodes(mp, SYNC_TRYLOCK);
-	xfs_reclaim_work_queue(mp);
-}
-
-static void
-__xfs_inode_set_reclaim_tag(
-	struct xfs_perag	*pag,
-	struct xfs_inode	*ip)
-{
-	radix_tree_tag_set(&pag->pag_ici_root,
-			   XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino),
-			   XFS_ICI_RECLAIM_TAG);
-
-	if (!pag->pag_ici_reclaimable) {
-		/* propagate the reclaim tag up into the perag radix tree */
-		spin_lock(&ip->i_mount->m_perag_lock);
-		radix_tree_tag_set(&ip->i_mount->m_perag_tree,
-				XFS_INO_TO_AGNO(ip->i_mount, ip->i_ino),
-				XFS_ICI_RECLAIM_TAG);
-		spin_unlock(&ip->i_mount->m_perag_lock);
-
-		/* schedule periodic background inode reclaim */
-		xfs_reclaim_work_queue(ip->i_mount);
-
-		trace_xfs_perag_set_reclaim(ip->i_mount, pag->pag_agno,
-							-1, _RET_IP_);
-	}
-	pag->pag_ici_reclaimable++;
-}
-
-/*
- * We set the inode flag atomically with the radix tree tag.
- * Once we get tag lookups on the radix tree, this inode flag
- * can go away.
- */
-void
-xfs_inode_set_reclaim_tag(
-	xfs_inode_t	*ip)
-{
-	struct xfs_mount *mp = ip->i_mount;
-	struct xfs_perag *pag;
-
-	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
-	spin_lock(&pag->pag_ici_lock);
-	spin_lock(&ip->i_flags_lock);
-	__xfs_inode_set_reclaim_tag(pag, ip);
-	__xfs_iflags_set(ip, XFS_IRECLAIMABLE);
-	spin_unlock(&ip->i_flags_lock);
-	spin_unlock(&pag->pag_ici_lock);
-	xfs_perag_put(pag);
-}
-
-STATIC void
-__xfs_inode_clear_reclaim(
-	xfs_perag_t	*pag,
-	xfs_inode_t	*ip)
-{
-	pag->pag_ici_reclaimable--;
-	if (!pag->pag_ici_reclaimable) {
-		/* clear the reclaim tag from the perag radix tree */
-		spin_lock(&ip->i_mount->m_perag_lock);
-		radix_tree_tag_clear(&ip->i_mount->m_perag_tree,
-				XFS_INO_TO_AGNO(ip->i_mount, ip->i_ino),
-				XFS_ICI_RECLAIM_TAG);
-		spin_unlock(&ip->i_mount->m_perag_lock);
-		trace_xfs_perag_clear_reclaim(ip->i_mount, pag->pag_agno,
-							-1, _RET_IP_);
-	}
-}
-
-STATIC void
-__xfs_inode_clear_reclaim_tag(
-	xfs_mount_t	*mp,
-	xfs_perag_t	*pag,
-	xfs_inode_t	*ip)
-{
-	radix_tree_tag_clear(&pag->pag_ici_root,
-			XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG);
-	__xfs_inode_clear_reclaim(pag, ip);
-}
-
 /*
  * Grab the inode for reclaim exclusively.
  * Return 0 if we grabbed it, non-zero otherwise.
@@ -929,6 +934,7 @@ xfs_reclaim_inode(
 	int			sync_mode)
 {
 	struct xfs_buf		*bp = NULL;
+	xfs_ino_t		ino = ip->i_ino; /* for radix_tree_delete */
 	int			error;
 
 restart:
@@ -993,6 +999,22 @@ restart:
 
 	xfs_iflock(ip);
 reclaim:
+	/*
+	 * Because we use RCU freeing we need to ensure the inode always appears
+	 * to be reclaimed with an invalid inode number when in the free state.
+	 * We do this as early as possible under the ILOCK and flush lock so
+	 * that xfs_iflush_cluster() can be guaranteed to detect races with us
+	 * here. By doing this, we guarantee that once xfs_iflush_cluster has
+	 * locked both the XFS_ILOCK and the flush lock that it will see either
+	 * a valid, flushable inode that will serialise correctly against the
+	 * locks below, or it will see a clean (and invalid) inode that it can
+	 * skip.
+	 */
+	spin_lock(&ip->i_flags_lock);
+	ip->i_flags = XFS_IRECLAIM;
+	ip->i_ino = 0;
+	spin_unlock(&ip->i_flags_lock);
+
 	xfs_ifunlock(ip);
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
 
@@ -1006,9 +1028,9 @@ reclaim:
 	 */
 	spin_lock(&pag->pag_ici_lock);
 	if (!radix_tree_delete(&pag->pag_ici_root,
-				XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino)))
+				XFS_INO_TO_AGINO(ip->i_mount, ino)))
 		ASSERT(0);
-	__xfs_inode_clear_reclaim(pag, ip);
+	xfs_perag_clear_reclaim_tag(pag);
 	spin_unlock(&pag->pag_ici_lock);
 
 	/*
@@ -1023,7 +1045,7 @@ reclaim:
 	xfs_qm_dqdetach(ip);
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
 
-	xfs_inode_free(ip);
+	__xfs_inode_free(ip);
 	return error;
 
 out_ifunlock:

fs/xfs/xfs_inode.c

@@ -1030,7 +1030,7 @@ xfs_dir_ialloc(
 			tp->t_flags &= ~(XFS_TRANS_DQ_DIRTY);
 		}
 
-		code = xfs_trans_roll(&tp, 0);
+		code = xfs_trans_roll(&tp, NULL);
 		if (committed != NULL)
 			*committed = 1;
 
@@ -1161,11 +1161,9 @@ xfs_create(
 		rdev = 0;
 		resblks = XFS_MKDIR_SPACE_RES(mp, name->len);
 		tres = &M_RES(mp)->tr_mkdir;
-		tp = xfs_trans_alloc(mp, XFS_TRANS_MKDIR);
 	} else {
 		resblks = XFS_CREATE_SPACE_RES(mp, name->len);
 		tres = &M_RES(mp)->tr_create;
-		tp = xfs_trans_alloc(mp, XFS_TRANS_CREATE);
 	}
 
 	/*
@@ -1174,20 +1172,19 @@ xfs_create(
 	 * the case we'll drop the one we have and get a more
 	 * appropriate transaction later.
 	 */
-	error = xfs_trans_reserve(tp, tres, resblks, 0);
+	error = xfs_trans_alloc(mp, tres, resblks, 0, 0, &tp);
 	if (error == -ENOSPC) {
 		/* flush outstanding delalloc blocks and retry */
 		xfs_flush_inodes(mp);
-		error = xfs_trans_reserve(tp, tres, resblks, 0);
+		error = xfs_trans_alloc(mp, tres, resblks, 0, 0, &tp);
 	}
 	if (error == -ENOSPC) {
 		/* No space at all so try a "no-allocation" reservation */
 		resblks = 0;
-		error = xfs_trans_reserve(tp, tres, 0, 0);
+		error = xfs_trans_alloc(mp, tres, 0, 0, 0, &tp);
 	}
 	if (error)
-		goto out_trans_cancel;
-
+		goto out_release_inode;
 
 	xfs_ilock(dp, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL |
 		      XFS_IOLOCK_PARENT | XFS_ILOCK_PARENT);
@@ -1337,17 +1334,16 @@ xfs_create_tmpfile(
 		return error;
 
 	resblks = XFS_IALLOC_SPACE_RES(mp);
-	tp = xfs_trans_alloc(mp, XFS_TRANS_CREATE_TMPFILE);
-
 	tres = &M_RES(mp)->tr_create_tmpfile;
-	error = xfs_trans_reserve(tp, tres, resblks, 0);
+
+	error = xfs_trans_alloc(mp, tres, resblks, 0, 0, &tp);
 	if (error == -ENOSPC) {
 		/* No space at all so try a "no-allocation" reservation */
 		resblks = 0;
-		error = xfs_trans_reserve(tp, tres, 0, 0);
+		error = xfs_trans_alloc(mp, tres, 0, 0, 0, &tp);
 	}
 	if (error)
-		goto out_trans_cancel;
+		goto out_release_inode;
 
 	error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp,
 						pdqp, resblks, 1, 0);
@@ -1432,15 +1428,14 @@ xfs_link(
 	if (error)
 		goto std_return;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_LINK);
 	resblks = XFS_LINK_SPACE_RES(mp, target_name->len);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_link, resblks, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_link, resblks, 0, 0, &tp);
 	if (error == -ENOSPC) {
 		resblks = 0;
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_link, 0, 0);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_link, 0, 0, 0, &tp);
 	}
 	if (error)
-		goto error_return;
+		goto std_return;
 
 	xfs_ilock(tdp, XFS_IOLOCK_EXCL | XFS_IOLOCK_PARENT);
 	xfs_lock_two_inodes(sip, tdp, XFS_ILOCK_EXCL);
@@ -1710,11 +1705,9 @@ xfs_inactive_truncate(
 	struct xfs_trans	*tp;
 	int			error;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_INACTIVE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp);
 	if (error) {
 		ASSERT(XFS_FORCED_SHUTDOWN(mp));
-		xfs_trans_cancel(tp);
 		return error;
 	}
 
@@ -1764,8 +1757,6 @@ xfs_inactive_ifree(
 	struct xfs_trans	*tp;
 	int			error;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_INACTIVE);
-
 	/*
 	 * The ifree transaction might need to allocate blocks for record
 	 * insertion to the finobt. We don't want to fail here at ENOSPC, so
@@ -1781,9 +1772,8 @@ xfs_inactive_ifree(
 	 * now remains allocated and sits on the unlinked list until the fs is
 	 * repaired.
 	 */
-	tp->t_flags |= XFS_TRANS_RESERVE;
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ifree,
-				  XFS_IFREE_SPACE_RES(mp), 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ifree,
+			XFS_IFREE_SPACE_RES(mp), 0, XFS_TRANS_RESERVE, &tp);
 	if (error) {
 		if (error == -ENOSPC) {
 			xfs_warn_ratelimited(mp,
@@ -1792,7 +1782,6 @@ xfs_inactive_ifree(
 		} else {
 			ASSERT(XFS_FORCED_SHUTDOWN(mp));
 		}
-		xfs_trans_cancel(tp);
 		return error;
 	}
 
@@ -2525,11 +2514,6 @@ xfs_remove(
 	if (error)
 		goto std_return;
 
-	if (is_dir)
-		tp = xfs_trans_alloc(mp, XFS_TRANS_RMDIR);
-	else
-		tp = xfs_trans_alloc(mp, XFS_TRANS_REMOVE);
-
 	/*
 	 * We try to get the real space reservation first,
 	 * allowing for directory btree deletion(s) implying
@@ -2540,14 +2524,15 @@ xfs_remove(
 	 * block from the directory.
 	 */
 	resblks = XFS_REMOVE_SPACE_RES(mp);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_remove, resblks, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_remove, resblks, 0, 0, &tp);
 	if (error == -ENOSPC) {
 		resblks = 0;
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_remove, 0, 0);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_remove, 0, 0, 0,
+				&tp);
 	}
 	if (error) {
 		ASSERT(error != -ENOSPC);
-		goto out_trans_cancel;
+		goto std_return;
 	}
 
 	xfs_ilock(dp, XFS_IOLOCK_EXCL | XFS_IOLOCK_PARENT);
@@ -2855,6 +2840,7 @@ xfs_rename_alloc_whiteout(
 	 * and flag it as linkable.
 	 */
 	drop_nlink(VFS_I(tmpfile));
+	xfs_setup_iops(tmpfile);
 	xfs_finish_inode_setup(tmpfile);
 	VFS_I(tmpfile)->i_state |= I_LINKABLE;
 
@@ -2910,15 +2896,15 @@ xfs_rename(
 	xfs_sort_for_rename(src_dp, target_dp, src_ip, target_ip, wip,
 				inodes, &num_inodes);
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_RENAME);
 	spaceres = XFS_RENAME_SPACE_RES(mp, target_name->len);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_rename, spaceres, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_rename, spaceres, 0, 0, &tp);
 	if (error == -ENOSPC) {
 		spaceres = 0;
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_rename, 0, 0);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_rename, 0, 0, 0,
+				&tp);
 	}
 	if (error)
-		goto out_trans_cancel;
+		goto out_release_wip;
 
 	/*
 	 * Attach the dquots to the inodes
@@ -3155,6 +3141,7 @@ out_bmap_cancel:
 	xfs_bmap_cancel(&free_list);
 out_trans_cancel:
 	xfs_trans_cancel(tp);
+out_release_wip:
 	if (wip)
 		IRELE(wip);
 	return error;
@@ -3162,16 +3149,16 @@ out_trans_cancel:
 
 STATIC int
 xfs_iflush_cluster(
-	xfs_inode_t	*ip,
-	xfs_buf_t	*bp)
+	struct xfs_inode	*ip,
+	struct xfs_buf		*bp)
 {
-	xfs_mount_t		*mp = ip->i_mount;
+	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_perag	*pag;
 	unsigned long		first_index, mask;
 	unsigned long		inodes_per_cluster;
-	int			ilist_size;
-	xfs_inode_t		**ilist;
-	xfs_inode_t		*iq;
+	int			cilist_size;
+	struct xfs_inode	**cilist;
+	struct xfs_inode	*cip;
 	int			nr_found;
 	int			clcount = 0;
 	int			bufwasdelwri;
@@ -3180,23 +3167,23 @@ xfs_iflush_cluster(
 	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
 
 	inodes_per_cluster = mp->m_inode_cluster_size >> mp->m_sb.sb_inodelog;
-	ilist_size = inodes_per_cluster * sizeof(xfs_inode_t *);
-	ilist = kmem_alloc(ilist_size, KM_MAYFAIL|KM_NOFS);
-	if (!ilist)
+	cilist_size = inodes_per_cluster * sizeof(xfs_inode_t *);
+	cilist = kmem_alloc(cilist_size, KM_MAYFAIL|KM_NOFS);
+	if (!cilist)
 		goto out_put;
 
 	mask = ~(((mp->m_inode_cluster_size >> mp->m_sb.sb_inodelog)) - 1);
 	first_index = XFS_INO_TO_AGINO(mp, ip->i_ino) & mask;
 	rcu_read_lock();
 	/* really need a gang lookup range call here */
-	nr_found = radix_tree_gang_lookup(&pag->pag_ici_root, (void**)ilist,
+	nr_found = radix_tree_gang_lookup(&pag->pag_ici_root, (void**)cilist,
 					first_index, inodes_per_cluster);
 	if (nr_found == 0)
 		goto out_free;
 
 	for (i = 0; i < nr_found; i++) {
-		iq = ilist[i];
-		if (iq == ip)
+		cip = cilist[i];
+		if (cip == ip)
 			continue;
 
 		/*
@@ -3205,20 +3192,30 @@ xfs_iflush_cluster(
 		 * We need to check under the i_flags_lock for a valid inode
 		 * here. Skip it if it is not valid or the wrong inode.
 		 */
-		spin_lock(&ip->i_flags_lock);
-		if (!ip->i_ino ||
-		    (XFS_INO_TO_AGINO(mp, iq->i_ino) & mask) != first_index) {
-			spin_unlock(&ip->i_flags_lock);
+		spin_lock(&cip->i_flags_lock);
+		if (!cip->i_ino ||
+		    __xfs_iflags_test(cip, XFS_ISTALE)) {
+			spin_unlock(&cip->i_flags_lock);
 			continue;
 		}
-		spin_unlock(&ip->i_flags_lock);
+
+		/*
+		 * Once we fall off the end of the cluster, no point checking
+		 * any more inodes in the list because they will also all be
+		 * outside the cluster.
+		 */
+		if ((XFS_INO_TO_AGINO(mp, cip->i_ino) & mask) != first_index) {
+			spin_unlock(&cip->i_flags_lock);
+			break;
+		}
+		spin_unlock(&cip->i_flags_lock);
 
 		/*
 		 * Do an un-protected check to see if the inode is dirty and
 		 * is a candidate for flushing.  These checks will be repeated
 		 * later after the appropriate locks are acquired.
 		 */
-		if (xfs_inode_clean(iq) && xfs_ipincount(iq) == 0)
+		if (xfs_inode_clean(cip) && xfs_ipincount(cip) == 0)
 			continue;
 
 		/*
@@ -3226,15 +3223,28 @@ xfs_iflush_cluster(
 		 * then this inode cannot be flushed and is skipped.
 		 */
 
-		if (!xfs_ilock_nowait(iq, XFS_ILOCK_SHARED))
+		if (!xfs_ilock_nowait(cip, XFS_ILOCK_SHARED))
+			continue;
+		if (!xfs_iflock_nowait(cip)) {
+			xfs_iunlock(cip, XFS_ILOCK_SHARED);
 			continue;
-		if (!xfs_iflock_nowait(iq)) {
-			xfs_iunlock(iq, XFS_ILOCK_SHARED);
+		}
+		if (xfs_ipincount(cip)) {
+			xfs_ifunlock(cip);
+			xfs_iunlock(cip, XFS_ILOCK_SHARED);
 			continue;
 		}
-		if (xfs_ipincount(iq)) {
-			xfs_ifunlock(iq);
-			xfs_iunlock(iq, XFS_ILOCK_SHARED);
+
+
+		/*
+		 * Check the inode number again, just to be certain we are not
+		 * racing with freeing in xfs_reclaim_inode(). See the comments
+		 * in that function for more information as to why the initial
+		 * check is not sufficient.
+		 */
+		if (!cip->i_ino) {
+			xfs_ifunlock(cip);
+			xfs_iunlock(cip, XFS_ILOCK_SHARED);
 			continue;
 		}
 
@@ -3242,18 +3252,18 @@ xfs_iflush_cluster(
 		 * arriving here means that this inode can be flushed.  First
 		 * re-check that it's dirty before flushing.
 		 */
-		if (!xfs_inode_clean(iq)) {
+		if (!xfs_inode_clean(cip)) {
 			int	error;
-			error = xfs_iflush_int(iq, bp);
+			error = xfs_iflush_int(cip, bp);
 			if (error) {
-				xfs_iunlock(iq, XFS_ILOCK_SHARED);
+				xfs_iunlock(cip, XFS_ILOCK_SHARED);
 				goto cluster_corrupt_out;
 			}
 			clcount++;
 		} else {
-			xfs_ifunlock(iq);
+			xfs_ifunlock(cip);
 		}
-		xfs_iunlock(iq, XFS_ILOCK_SHARED);
+		xfs_iunlock(cip, XFS_ILOCK_SHARED);
 	}
 
 	if (clcount) {
@@ -3263,7 +3273,7 @@ xfs_iflush_cluster(
 
 out_free:
 	rcu_read_unlock();
-	kmem_free(ilist);
+	kmem_free(cilist);
 out_put:
 	xfs_perag_put(pag);
 	return 0;
@@ -3306,8 +3316,8 @@ cluster_corrupt_out:
 	/*
 	 * Unlocks the flush lock
 	 */
-	xfs_iflush_abort(iq, false);
-	kmem_free(ilist);
+	xfs_iflush_abort(cip, false);
+	kmem_free(cilist);
 	xfs_perag_put(pag);
 	return -EFSCORRUPTED;
 }
@@ -3327,7 +3337,7 @@ xfs_iflush(
 	struct xfs_buf		**bpp)
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_buf		*bp;
+	struct xfs_buf		*bp = NULL;
 	struct xfs_dinode	*dip;
 	int			error;
 
@@ -3369,14 +3379,22 @@ xfs_iflush(
 	}
 
 	/*
-	 * Get the buffer containing the on-disk inode.
+	 * Get the buffer containing the on-disk inode. We are doing a try-lock
+	 * operation here, so we may get  an EAGAIN error. In that case, we
+	 * simply want to return with the inode still dirty.
+	 *
+	 * If we get any other error, we effectively have a corruption situation
+	 * and we cannot flush the inode, so we treat it the same as failing
+	 * xfs_iflush_int().
 	 */
 	error = xfs_imap_to_bp(mp, NULL, &ip->i_imap, &dip, &bp, XBF_TRYLOCK,
 			       0);
-	if (error || !bp) {
+	if (error == -EAGAIN) {
 		xfs_ifunlock(ip);
 		return error;
 	}
+	if (error)
+		goto corrupt_out;
 
 	/*
 	 * First flush out the inode that xfs_iflush was called with.
@@ -3404,7 +3422,8 @@ xfs_iflush(
 	return 0;
 
 corrupt_out:
-	xfs_buf_relse(bp);
+	if (bp)
+		xfs_buf_relse(bp);
 	xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
 cluster_corrupt_out:
 	error = -EFSCORRUPTED;

fs/xfs/xfs_inode.h

@@ -440,6 +440,9 @@ loff_t	__xfs_seek_hole_data(struct inode *inode, loff_t start,
 
 
 /* from xfs_iops.c */
+extern void xfs_setup_inode(struct xfs_inode *ip);
+extern void xfs_setup_iops(struct xfs_inode *ip);
+
 /*
  * When setting up a newly allocated inode, we need to call
  * xfs_finish_inode_setup() once the inode is fully instantiated at
@@ -447,7 +450,6 @@ loff_t	__xfs_seek_hole_data(struct inode *inode, loff_t start,
  * before we've completed instantiation. Otherwise we can do it
  * the moment the inode lookup is complete.
  */
-extern void xfs_setup_inode(struct xfs_inode *ip);
 static inline void xfs_finish_inode_setup(struct xfs_inode *ip)
 {
 	xfs_iflags_clear(ip, XFS_INEW);
@@ -458,6 +460,7 @@ static inline void xfs_finish_inode_setup(struct xfs_inode *ip)
 static inline void xfs_setup_existing_inode(struct xfs_inode *ip)
 {
 	xfs_setup_inode(ip);
+	xfs_setup_iops(ip);
 	xfs_finish_inode_setup(ip);
 }
 

fs/xfs/xfs_inode_item.c

@@ -210,7 +210,7 @@ xfs_inode_item_format_data_fork(
 			 */
 			data_bytes = roundup(ip->i_df.if_bytes, 4);
 			ASSERT(ip->i_df.if_real_bytes == 0 ||
-			       ip->i_df.if_real_bytes == data_bytes);
+			       ip->i_df.if_real_bytes >= data_bytes);
 			ASSERT(ip->i_df.if_u1.if_data != NULL);
 			ASSERT(ip->i_d.di_size > 0);
 			xlog_copy_iovec(lv, vecp, XLOG_REG_TYPE_ILOCAL,
@@ -305,7 +305,7 @@ xfs_inode_item_format_attr_fork(
 			 */
 			data_bytes = roundup(ip->i_afp->if_bytes, 4);
 			ASSERT(ip->i_afp->if_real_bytes == 0 ||
-			       ip->i_afp->if_real_bytes == data_bytes);
+			       ip->i_afp->if_real_bytes >= data_bytes);
 			ASSERT(ip->i_afp->if_u1.if_data != NULL);
 			xlog_copy_iovec(lv, vecp, XLOG_REG_TYPE_IATTR_LOCAL,
 					ip->i_afp->if_u1.if_data,
@@ -479,6 +479,8 @@ STATIC uint
 xfs_inode_item_push(
 	struct xfs_log_item	*lip,
 	struct list_head	*buffer_list)
+		__releases(&lip->li_ailp->xa_lock)
+		__acquires(&lip->li_ailp->xa_lock)
 {
 	struct xfs_inode_log_item *iip = INODE_ITEM(lip);
 	struct xfs_inode	*ip = iip->ili_inode;

fs/xfs/xfs_ioctl.c

@@ -277,7 +277,6 @@ xfs_readlink_by_handle(
 {
 	struct dentry		*dentry;
 	__u32			olen;
-	void			*link;
 	int			error;
 
 	if (!capable(CAP_SYS_ADMIN))
@@ -288,7 +287,7 @@ xfs_readlink_by_handle(
 		return PTR_ERR(dentry);
 
 	/* Restrict this handle operation to symlinks only. */
-	if (!d_is_symlink(dentry)) {
+	if (!d_inode(dentry)->i_op->readlink) {
 		error = -EINVAL;
 		goto out_dput;
 	}
@@ -298,21 +297,8 @@ xfs_readlink_by_handle(
 		goto out_dput;
 	}
 
-	link = kmalloc(MAXPATHLEN+1, GFP_KERNEL);
-	if (!link) {
-		error = -ENOMEM;
-		goto out_dput;
-	}
-
-	error = xfs_readlink(XFS_I(d_inode(dentry)), link);
-	if (error)
-		goto out_kfree;
-	error = readlink_copy(hreq->ohandle, olen, link);
-	if (error)
-		goto out_kfree;
+	error = d_inode(dentry)->i_op->readlink(dentry, hreq->ohandle, olen);
 
- out_kfree:
-	kfree(link);
  out_dput:
 	dput(dentry);
 	return error;
@@ -334,12 +320,10 @@ xfs_set_dmattrs(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -EIO;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SET_DMATTRS);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0, 0, 0, &tp);
+	if (error)
 		return error;
-	}
+
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 
@@ -1141,10 +1125,9 @@ xfs_ioctl_setattr_get_trans(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		goto out_unlock;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0, 0, 0, &tp);
 	if (error)
-		goto out_cancel;
+		return ERR_PTR(error);
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | join_flags);

fs/xfs/xfs_iomap.c

@@ -132,6 +132,7 @@ xfs_iomap_write_direct(
 	int		error;
 	int		lockmode;
 	int		bmapi_flags = XFS_BMAPI_PREALLOC;
+	uint		tflags = 0;
 
 	rt = XFS_IS_REALTIME_INODE(ip);
 	extsz = xfs_get_extsz_hint(ip);
@@ -191,11 +192,6 @@ xfs_iomap_write_direct(
 	if (error)
 		return error;
 
-	/*
-	 * Allocate and setup the transaction
-	 */
-	tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
-
 	/*
 	 * For DAX, we do not allocate unwritten extents, but instead we zero
 	 * the block before we commit the transaction.  Ideally we'd like to do
@@ -209,23 +205,17 @@ xfs_iomap_write_direct(
 	 * the reserve block pool for bmbt block allocation if there is no space
 	 * left but we need to do unwritten extent conversion.
 	 */
-
 	if (IS_DAX(VFS_I(ip))) {
 		bmapi_flags = XFS_BMAPI_CONVERT | XFS_BMAPI_ZERO;
 		if (ISUNWRITTEN(imap)) {
-			tp->t_flags |= XFS_TRANS_RESERVE;
+			tflags |= XFS_TRANS_RESERVE;
 			resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0) << 1;
 		}
 	}
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write,
-				  resblks, resrtextents);
-	/*
-	 * Check for running out of space, note: need lock to return
-	 */
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks, resrtextents,
+			tflags, &tp);
+	if (error)
 		return error;
-	}
 
 	lockmode = XFS_ILOCK_EXCL;
 	xfs_ilock(ip, lockmode);
@@ -726,15 +716,13 @@ xfs_iomap_write_allocate(
 
 		nimaps = 0;
 		while (nimaps == 0) {
-			tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE);
-			tp->t_flags |= XFS_TRANS_RESERVE;
 			nres = XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK);
-			error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write,
-						  nres, 0);
-			if (error) {
-				xfs_trans_cancel(tp);
+
+			error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, nres,
+					0, XFS_TRANS_RESERVE, &tp);
+			if (error)
 				return error;
-			}
+
 			xfs_ilock(ip, XFS_ILOCK_EXCL);
 			xfs_trans_ijoin(tp, ip, 0);
 
@@ -878,25 +866,18 @@ xfs_iomap_write_unwritten(
 
 	do {
 		/*
-		 * set up a transaction to convert the range of extents
+		 * Set up a transaction to convert the range of extents
 		 * from unwritten to real. Do allocations in a loop until
 		 * we have covered the range passed in.
 		 *
-		 * Note that we open code the transaction allocation here
-		 * to pass KM_NOFS--we can't risk to recursing back into
-		 * the filesystem here as we might be asked to write out
-		 * the same inode that we complete here and might deadlock
-		 * on the iolock.
+		 * Note that we can't risk to recursing back into the filesystem
+		 * here as we might be asked to write out the same inode that we
+		 * complete here and might deadlock on the iolock.
 		 */
-		sb_start_intwrite(mp->m_super);
-		tp = _xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE, KM_NOFS);
-		tp->t_flags |= XFS_TRANS_RESERVE | XFS_TRANS_FREEZE_PROT;
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write,
-					  resblks, 0);
-		if (error) {
-			xfs_trans_cancel(tp);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks, 0,
+				XFS_TRANS_RESERVE | XFS_TRANS_NOFS, &tp);
+		if (error)
 			return error;
-		}
 
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
 		xfs_trans_ijoin(tp, ip, 0);

fs/xfs/xfs_iops.c

@@ -181,6 +181,8 @@ xfs_generic_create(
 	}
 #endif
 
+	xfs_setup_iops(ip);
+
 	if (tmpfile)
 		d_tmpfile(dentry, inode);
 	else
@@ -368,6 +370,8 @@ xfs_vn_symlink(
 	if (unlikely(error))
 		goto out_cleanup_inode;
 
+	xfs_setup_iops(cip);
+
 	d_instantiate(dentry, inode);
 	xfs_finish_inode_setup(cip);
 	return 0;
@@ -442,6 +446,16 @@ xfs_vn_get_link(
 	return ERR_PTR(error);
 }
 
+STATIC const char *
+xfs_vn_get_link_inline(
+	struct dentry		*dentry,
+	struct inode		*inode,
+	struct delayed_call	*done)
+{
+	ASSERT(XFS_I(inode)->i_df.if_flags & XFS_IFINLINE);
+	return XFS_I(inode)->i_df.if_u1.if_data;
+}
+
 STATIC int
 xfs_vn_getattr(
 	struct vfsmount		*mnt,
@@ -599,12 +613,12 @@ xfs_setattr_nonsize(
 			return error;
 	}
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0, 0, 0, &tp);
 	if (error)
-		goto out_trans_cancel;
+		goto out_dqrele;
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, 0);
 
 	/*
 	 * Change file ownership.  Must be the owner or privileged.
@@ -633,12 +647,10 @@ xfs_setattr_nonsize(
 						NULL, capable(CAP_FOWNER) ?
 						XFS_QMOPT_FORCE_RES : 0);
 			if (error)	/* out of quota */
-				goto out_unlock;
+				goto out_cancel;
 		}
 	}
 
-	xfs_trans_ijoin(tp, ip, 0);
-
 	/*
 	 * Change file ownership.  Must be the owner or privileged.
 	 */
@@ -722,10 +734,9 @@ xfs_setattr_nonsize(
 
 	return 0;
 
-out_unlock:
-	xfs_iunlock(ip, XFS_ILOCK_EXCL);
-out_trans_cancel:
+out_cancel:
 	xfs_trans_cancel(tp);
+out_dqrele:
 	xfs_qm_dqrele(udqp);
 	xfs_qm_dqrele(gdqp);
 	return error;
@@ -834,7 +845,7 @@ xfs_setattr_size(
 	 * We have to do all the page cache truncate work outside the
 	 * transaction context as the "lock" order is page lock->log space
 	 * reservation as defined by extent allocation in the writeback path.
-	 * Hence a truncate can fail with ENOMEM from xfs_trans_reserve(), but
+	 * Hence a truncate can fail with ENOMEM from xfs_trans_alloc(), but
 	 * having already truncated the in-memory version of the file (i.e. made
 	 * user visible changes). There's not much we can do about this, except
 	 * to hope that the caller sees ENOMEM and retries the truncate
@@ -849,10 +860,9 @@ xfs_setattr_size(
 		return error;
 	truncate_setsize(inode, newsize);
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp);
 	if (error)
-		goto out_trans_cancel;
+		return error;
 
 	lock_flags |= XFS_ILOCK_EXCL;
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
@@ -971,12 +981,9 @@ xfs_vn_update_time(
 
 	trace_xfs_update_time(ip);
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_fsyncts, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_fsyncts, 0, 0, 0, &tp);
+	if (error)
 		return error;
-	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	if (flags & S_CTIME)
@@ -1167,6 +1174,18 @@ static const struct inode_operations xfs_symlink_inode_operations = {
 	.update_time		= xfs_vn_update_time,
 };
 
+static const struct inode_operations xfs_inline_symlink_inode_operations = {
+	.readlink		= generic_readlink,
+	.get_link		= xfs_vn_get_link_inline,
+	.getattr		= xfs_vn_getattr,
+	.setattr		= xfs_vn_setattr,
+	.setxattr		= generic_setxattr,
+	.getxattr		= generic_getxattr,
+	.removexattr		= generic_removexattr,
+	.listxattr		= xfs_vn_listxattr,
+	.update_time		= xfs_vn_update_time,
+};
+
 STATIC void
 xfs_diflags_to_iflags(
 	struct inode		*inode,
@@ -1193,7 +1212,7 @@ xfs_diflags_to_iflags(
 }
 
 /*
- * Initialize the Linux inode and set up the operation vectors.
+ * Initialize the Linux inode.
  *
  * When reading existing inodes from disk this is called directly from xfs_iget,
  * when creating a new inode it is called from xfs_ialloc after setting up the
@@ -1232,32 +1251,12 @@ xfs_setup_inode(
 	i_size_write(inode, ip->i_d.di_size);
 	xfs_diflags_to_iflags(inode, ip);
 
-	ip->d_ops = ip->i_mount->m_nondir_inode_ops;
-	lockdep_set_class(&ip->i_lock.mr_lock, &xfs_nondir_ilock_class);
-	switch (inode->i_mode & S_IFMT) {
-	case S_IFREG:
-		inode->i_op = &xfs_inode_operations;
-		inode->i_fop = &xfs_file_operations;
-		inode->i_mapping->a_ops = &xfs_address_space_operations;
-		break;
-	case S_IFDIR:
+	if (S_ISDIR(inode->i_mode)) {
 		lockdep_set_class(&ip->i_lock.mr_lock, &xfs_dir_ilock_class);
-		if (xfs_sb_version_hasasciici(&XFS_M(inode->i_sb)->m_sb))
-			inode->i_op = &xfs_dir_ci_inode_operations;
-		else
-			inode->i_op = &xfs_dir_inode_operations;
-		inode->i_fop = &xfs_dir_file_operations;
 		ip->d_ops = ip->i_mount->m_dir_inode_ops;
-		break;
-	case S_IFLNK:
-		inode->i_op = &xfs_symlink_inode_operations;
-		if (!(ip->i_df.if_flags & XFS_IFINLINE))
-			inode->i_mapping->a_ops = &xfs_address_space_operations;
-		break;
-	default:
-		inode->i_op = &xfs_inode_operations;
-		init_special_inode(inode, inode->i_mode, inode->i_rdev);
-		break;
+	} else {
+		ip->d_ops = ip->i_mount->m_nondir_inode_ops;
+		lockdep_set_class(&ip->i_lock.mr_lock, &xfs_nondir_ilock_class);
 	}
 
 	/*
@@ -1277,3 +1276,35 @@ xfs_setup_inode(
 		cache_no_acl(inode);
 	}
 }
+
+void
+xfs_setup_iops(
+	struct xfs_inode	*ip)
+{
+	struct inode		*inode = &ip->i_vnode;
+
+	switch (inode->i_mode & S_IFMT) {
+	case S_IFREG:
+		inode->i_op = &xfs_inode_operations;
+		inode->i_fop = &xfs_file_operations;
+		inode->i_mapping->a_ops = &xfs_address_space_operations;
+		break;
+	case S_IFDIR:
+		if (xfs_sb_version_hasasciici(&XFS_M(inode->i_sb)->m_sb))
+			inode->i_op = &xfs_dir_ci_inode_operations;
+		else
+			inode->i_op = &xfs_dir_inode_operations;
+		inode->i_fop = &xfs_dir_file_operations;
+		break;
+	case S_IFLNK:
+		if (ip->i_df.if_flags & XFS_IFINLINE)
+			inode->i_op = &xfs_inline_symlink_inode_operations;
+		else
+			inode->i_op = &xfs_symlink_inode_operations;
+		break;
+	default:
+		inode->i_op = &xfs_inode_operations;
+		init_special_inode(inode, inode->i_mode, inode->i_rdev);
+		break;
+	}
+}

fs/xfs/xfs_log.c

@@ -435,8 +435,7 @@ xfs_log_reserve(
 	int		 	cnt,
 	struct xlog_ticket	**ticp,
 	__uint8_t	 	client,
-	bool			permanent,
-	uint		 	t_type)
+	bool			permanent)
 {
 	struct xlog		*log = mp->m_log;
 	struct xlog_ticket	*tic;
@@ -456,7 +455,6 @@ xfs_log_reserve(
 	if (!tic)
 		return -ENOMEM;
 
-	tic->t_trans_type = t_type;
 	*ticp = tic;
 
 	xlog_grant_push_ail(log, tic->t_cnt ? tic->t_unit_res * tic->t_cnt
@@ -823,8 +821,7 @@ xfs_log_unmount_write(xfs_mount_t *mp)
 	} while (iclog != first_iclog);
 #endif
 	if (! (XLOG_FORCED_SHUTDOWN(log))) {
-		error = xfs_log_reserve(mp, 600, 1, &tic,
-					XFS_LOG, 0, XLOG_UNMOUNT_REC_TYPE);
+		error = xfs_log_reserve(mp, 600, 1, &tic, XFS_LOG, 0);
 		if (!error) {
 			/* the data section must be 32 bit size aligned */
 			struct {
@@ -2032,58 +2029,8 @@ xlog_print_tic_res(
 	    REG_TYPE_STR(ICREATE, "inode create")
 	};
 #undef REG_TYPE_STR
-#define TRANS_TYPE_STR(type)	[XFS_TRANS_##type] = #type
-	static char *trans_type_str[XFS_TRANS_TYPE_MAX] = {
-	    TRANS_TYPE_STR(SETATTR_NOT_SIZE),
-	    TRANS_TYPE_STR(SETATTR_SIZE),
-	    TRANS_TYPE_STR(INACTIVE),
-	    TRANS_TYPE_STR(CREATE),
-	    TRANS_TYPE_STR(CREATE_TRUNC),
-	    TRANS_TYPE_STR(TRUNCATE_FILE),
-	    TRANS_TYPE_STR(REMOVE),
-	    TRANS_TYPE_STR(LINK),
-	    TRANS_TYPE_STR(RENAME),
-	    TRANS_TYPE_STR(MKDIR),
-	    TRANS_TYPE_STR(RMDIR),
-	    TRANS_TYPE_STR(SYMLINK),
-	    TRANS_TYPE_STR(SET_DMATTRS),
-	    TRANS_TYPE_STR(GROWFS),
-	    TRANS_TYPE_STR(STRAT_WRITE),
-	    TRANS_TYPE_STR(DIOSTRAT),
-	    TRANS_TYPE_STR(WRITEID),
-	    TRANS_TYPE_STR(ADDAFORK),
-	    TRANS_TYPE_STR(ATTRINVAL),
-	    TRANS_TYPE_STR(ATRUNCATE),
-	    TRANS_TYPE_STR(ATTR_SET),
-	    TRANS_TYPE_STR(ATTR_RM),
-	    TRANS_TYPE_STR(ATTR_FLAG),
-	    TRANS_TYPE_STR(CLEAR_AGI_BUCKET),
-	    TRANS_TYPE_STR(SB_CHANGE),
-	    TRANS_TYPE_STR(DUMMY1),
-	    TRANS_TYPE_STR(DUMMY2),
-	    TRANS_TYPE_STR(QM_QUOTAOFF),
-	    TRANS_TYPE_STR(QM_DQALLOC),
-	    TRANS_TYPE_STR(QM_SETQLIM),
-	    TRANS_TYPE_STR(QM_DQCLUSTER),
-	    TRANS_TYPE_STR(QM_QINOCREATE),
-	    TRANS_TYPE_STR(QM_QUOTAOFF_END),
-	    TRANS_TYPE_STR(FSYNC_TS),
-	    TRANS_TYPE_STR(GROWFSRT_ALLOC),
-	    TRANS_TYPE_STR(GROWFSRT_ZERO),
-	    TRANS_TYPE_STR(GROWFSRT_FREE),
-	    TRANS_TYPE_STR(SWAPEXT),
-	    TRANS_TYPE_STR(CHECKPOINT),
-	    TRANS_TYPE_STR(ICREATE),
-	    TRANS_TYPE_STR(CREATE_TMPFILE)
-	};
-#undef TRANS_TYPE_STR
 
 	xfs_warn(mp, "xlog_write: reservation summary:");
-	xfs_warn(mp, "  trans type  = %s (%u)",
-		 ((ticket->t_trans_type <= 0 ||
-		   ticket->t_trans_type > XFS_TRANS_TYPE_MAX) ?
-		  "bad-trans-type" : trans_type_str[ticket->t_trans_type]),
-		 ticket->t_trans_type);
 	xfs_warn(mp, "  unit res    = %d bytes",
 		 ticket->t_unit_res);
 	xfs_warn(mp, "  current res = %d bytes",
@@ -3378,7 +3325,7 @@ xfs_log_force(
 {
 	int	error;
 
-	trace_xfs_log_force(mp, 0);
+	trace_xfs_log_force(mp, 0, _RET_IP_);
 	error = _xfs_log_force(mp, flags, NULL);
 	if (error)
 		xfs_warn(mp, "%s: error %d returned.", __func__, error);
@@ -3527,7 +3474,7 @@ xfs_log_force_lsn(
 {
 	int	error;
 
-	trace_xfs_log_force(mp, lsn);
+	trace_xfs_log_force(mp, lsn, _RET_IP_);
 	error = _xfs_log_force_lsn(mp, lsn, flags, NULL);
 	if (error)
 		xfs_warn(mp, "%s: error %d returned.", __func__, error);
@@ -3709,7 +3656,6 @@ xlog_ticket_alloc(
 	tic->t_tid		= prandom_u32();
 	tic->t_clientid		= client;
 	tic->t_flags		= XLOG_TIC_INITED;
-	tic->t_trans_type	= 0;
 	if (permanent)
 		tic->t_flags |= XLOG_TIC_PERM_RESERV;
 

fs/xfs/xfs_log.h

@@ -161,8 +161,7 @@ int	  xfs_log_reserve(struct xfs_mount *mp,
 			  int		   count,
 			  struct xlog_ticket **ticket,
 			  __uint8_t	   clientid,
-			  bool		   permanent,
-			  uint		   t_type);
+			  bool		   permanent);
 int	  xfs_log_regrant(struct xfs_mount *mp, struct xlog_ticket *tic);
 int	  xfs_log_unmount_write(struct xfs_mount *mp);
 void      xfs_log_unmount(struct xfs_mount *mp);

fs/xfs/xfs_log_cil.c

@@ -51,7 +51,6 @@ xlog_cil_ticket_alloc(
 
 	tic = xlog_ticket_alloc(log, 0, 1, XFS_TRANSACTION, 0,
 				KM_SLEEP|KM_NOFS);
-	tic->t_trans_type = XFS_TRANS_CHECKPOINT;
 
 	/*
 	 * set the current reservation to zero so we know to steal the basic

fs/xfs/xfs_log_priv.h

@@ -175,7 +175,6 @@ typedef struct xlog_ticket {
 	char		   t_cnt;	 /* current count		 : 1  */
 	char		   t_clientid;	 /* who does this belong to;	 : 1  */
 	char		   t_flags;	 /* properties of reservation	 : 1  */
-	uint		   t_trans_type; /* transaction type             : 4  */
 
         /* reservation array fields */
 	uint		   t_res_num;                    /* num in array : 4 */

fs/xfs/xfs_log_recover.c

@@ -3843,7 +3843,7 @@ xlog_recover_add_to_cont_trans(
 	old_ptr = item->ri_buf[item->ri_cnt-1].i_addr;
 	old_len = item->ri_buf[item->ri_cnt-1].i_len;
 
-	ptr = kmem_realloc(old_ptr, len+old_len, old_len, KM_SLEEP);
+	ptr = kmem_realloc(old_ptr, len + old_len, KM_SLEEP);
 	memcpy(&ptr[old_len], dp, len);
 	item->ri_buf[item->ri_cnt-1].i_len += len;
 	item->ri_buf[item->ri_cnt-1].i_addr = ptr;
@@ -4205,10 +4205,9 @@ xlog_recover_process_efi(
 		}
 	}
 
-	tp = xfs_trans_alloc(mp, 0);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp);
 	if (error)
-		goto abort_error;
+		return error;
 	efdp = xfs_trans_get_efd(tp, efip, efip->efi_format.efi_nextents);
 
 	for (i = 0; i < efip->efi_format.efi_nextents; i++) {
@@ -4355,10 +4354,9 @@ xlog_recover_clear_agi_bucket(
 	int		offset;
 	int		error;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_CLEAR_AGI_BUCKET);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_clearagi, 0, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_clearagi, 0, 0, 0, &tp);
 	if (error)
-		goto out_abort;
+		goto out_error;
 
 	error = xfs_read_agi(mp, tp, agno, &agibp);
 	if (error)

fs/xfs/xfs_mount.c

@@ -89,7 +89,6 @@ xfs_uuid_mount(
 	if (hole < 0) {
 		xfs_uuid_table = kmem_realloc(xfs_uuid_table,
 			(xfs_uuid_table_size + 1) * sizeof(*xfs_uuid_table),
-			xfs_uuid_table_size  * sizeof(*xfs_uuid_table),
 			KM_SLEEP);
 		hole = xfs_uuid_table_size++;
 	}
@@ -681,6 +680,9 @@ xfs_mountfs(
 
 	xfs_set_maxicount(mp);
 
+	/* enable fail_at_unmount as default */
+	mp->m_fail_unmount = 1;
+
 	error = xfs_sysfs_init(&mp->m_kobj, &xfs_mp_ktype, NULL, mp->m_fsname);
 	if (error)
 		goto out;
@@ -690,10 +692,15 @@ xfs_mountfs(
 	if (error)
 		goto out_remove_sysfs;
 
-	error = xfs_uuid_mount(mp);
+	error = xfs_error_sysfs_init(mp);
 	if (error)
 		goto out_del_stats;
 
+
+	error = xfs_uuid_mount(mp);
+	if (error)
+		goto out_remove_error_sysfs;
+
 	/*
 	 * Set the minimum read and write sizes
 	 */
@@ -957,6 +964,7 @@ xfs_mountfs(
 	cancel_delayed_work_sync(&mp->m_reclaim_work);
 	xfs_reclaim_inodes(mp, SYNC_WAIT);
  out_log_dealloc:
+	mp->m_flags |= XFS_MOUNT_UNMOUNTING;
 	xfs_log_mount_cancel(mp);
  out_fail_wait:
 	if (mp->m_logdev_targp && mp->m_logdev_targp != mp->m_ddev_targp)
@@ -968,6 +976,8 @@ xfs_mountfs(
 	xfs_da_unmount(mp);
  out_remove_uuid:
 	xfs_uuid_unmount(mp);
+ out_remove_error_sysfs:
+	xfs_error_sysfs_del(mp);
  out_del_stats:
 	xfs_sysfs_del(&mp->m_stats.xs_kobj);
  out_remove_sysfs:
@@ -1005,6 +1015,14 @@ xfs_unmountfs(
 	 */
 	xfs_log_force(mp, XFS_LOG_SYNC);
 
+	/*
+	 * We now need to tell the world we are unmounting. This will allow
+	 * us to detect that the filesystem is going away and we should error
+	 * out anything that we have been retrying in the background. This will
+	 * prevent neverending retries in AIL pushing from hanging the unmount.
+	 */
+	mp->m_flags |= XFS_MOUNT_UNMOUNTING;
+
 	/*
 	 * Flush all pending changes from the AIL.
 	 */
@@ -1056,6 +1074,7 @@ xfs_unmountfs(
 #endif
 	xfs_free_perag(mp);
 
+	xfs_error_sysfs_del(mp);
 	xfs_sysfs_del(&mp->m_stats.xs_kobj);
 	xfs_sysfs_del(&mp->m_kobj);
 }

fs/xfs/xfs_mount.h

@@ -37,6 +37,32 @@ enum {
 	XFS_LOWSP_MAX,
 };
 
+/*
+ * Error Configuration
+ *
+ * Error classes define the subsystem the configuration belongs to.
+ * Error numbers define the errors that are configurable.
+ */
+enum {
+	XFS_ERR_METADATA,
+	XFS_ERR_CLASS_MAX,
+};
+enum {
+	XFS_ERR_DEFAULT,
+	XFS_ERR_EIO,
+	XFS_ERR_ENOSPC,
+	XFS_ERR_ENODEV,
+	XFS_ERR_ERRNO_MAX,
+};
+
+#define XFS_ERR_RETRY_FOREVER	-1
+
+struct xfs_error_cfg {
+	struct xfs_kobj	kobj;
+	int		max_retries;
+	unsigned long	retry_timeout;	/* in jiffies, 0 = no timeout */
+};
+
 typedef struct xfs_mount {
 	struct super_block	*m_super;
 	xfs_tid_t		m_tid;		/* next unused tid for fs */
@@ -127,6 +153,9 @@ typedef struct xfs_mount {
 	int64_t			m_low_space[XFS_LOWSP_MAX];
 						/* low free space thresholds */
 	struct xfs_kobj		m_kobj;
+	struct xfs_kobj		m_error_kobj;
+	struct xfs_kobj		m_error_meta_kobj;
+	struct xfs_error_cfg	m_error_cfg[XFS_ERR_CLASS_MAX][XFS_ERR_ERRNO_MAX];
 	struct xstats		m_stats;	/* per-fs stats */
 
 	struct workqueue_struct *m_buf_workqueue;
@@ -148,6 +177,7 @@ typedef struct xfs_mount {
 	 */
 	__uint32_t		m_generation;
 
+	bool			m_fail_unmount;
 #ifdef DEBUG
 	/*
 	 * DEBUG mode instrumentation to test and/or trigger delayed allocation
@@ -166,6 +196,7 @@ typedef struct xfs_mount {
 #define XFS_MOUNT_WSYNC		(1ULL << 0)	/* for nfs - all metadata ops
 						   must be synchronous except
 						   for space allocations */
+#define XFS_MOUNT_UNMOUNTING	(1ULL << 1)	/* filesystem is unmounting */
 #define XFS_MOUNT_WAS_CLEAN	(1ULL << 3)
 #define XFS_MOUNT_FS_SHUTDOWN	(1ULL << 4)	/* atomic stop of all filesystem
 						   operations, typically for
@@ -364,4 +395,7 @@ extern void	xfs_set_low_space_thresholds(struct xfs_mount *);
 int	xfs_zero_extent(struct xfs_inode *ip, xfs_fsblock_t start_fsb,
 			xfs_off_t count_fsb);
 
+struct xfs_error_cfg * xfs_error_get_cfg(struct xfs_mount *mp,
+		int error_class, int error);
+
 #endif	/* __XFS_MOUNT_H__ */

fs/xfs/xfs_pnfs.c

@@ -308,12 +308,9 @@ xfs_fs_commit_blocks(
 			goto out_drop_iolock;
 	}
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0, 0, 0, &tp);
+	if (error)
 		goto out_drop_iolock;
-	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);

fs/xfs/xfs_qm.c

@@ -783,13 +783,10 @@ xfs_qm_qino_alloc(
 		}
 	}
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_QINOCREATE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_create,
-				  XFS_QM_QINOCREATE_SPACE_RES(mp), 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_create,
+			XFS_QM_QINOCREATE_SPACE_RES(mp), 0, 0, &tp);
+	if (error)
 		return error;
-	}
 
 	if (need_alloc) {
 		error = xfs_dir_ialloc(&tp, NULL, S_IFREG, 1, 0, 0, 1, ip,

fs/xfs/xfs_qm_syscalls.c

@@ -236,10 +236,8 @@ xfs_qm_scall_trunc_qfile(
 
 	xfs_ilock(ip, XFS_IOLOCK_EXCL);
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_TRUNCATE_FILE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp);
 	if (error) {
-		xfs_trans_cancel(tp);
 		xfs_iunlock(ip, XFS_IOLOCK_EXCL);
 		goto out_put;
 	}
@@ -436,12 +434,9 @@ xfs_qm_scall_setqlim(
 	defq = xfs_get_defquota(dqp, q);
 	xfs_dqunlock(dqp);
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SETQLIM);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_setqlim, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_qm_setqlim, 0, 0, 0, &tp);
+	if (error)
 		goto out_rele;
-	}
 
 	xfs_dqlock(dqp);
 	xfs_trans_dqjoin(tp, dqp);
@@ -569,13 +564,9 @@ xfs_qm_log_quotaoff_end(
 	int			error;
 	xfs_qoff_logitem_t	*qoffi;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_QUOTAOFF_END);
-
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_equotaoff, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_qm_equotaoff, 0, 0, 0, &tp);
+	if (error)
 		return error;
-	}
 
 	qoffi = xfs_trans_get_qoff_item(tp, startqoff,
 					flags & XFS_ALL_QUOTA_ACCT);
@@ -603,12 +594,9 @@ xfs_qm_log_quotaoff(
 
 	*qoffstartp = NULL;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_QUOTAOFF);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_quotaoff, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_qm_quotaoff, 0, 0, 0, &tp);
+	if (error)
 		goto out;
-	}
 
 	qoffi = xfs_trans_get_qoff_item(tp, NULL, flags & XFS_ALL_QUOTA_ACCT);
 	xfs_trans_log_quotaoff_item(tp, qoffi);

fs/xfs/xfs_rtalloc.c

@@ -780,15 +780,14 @@ xfs_growfs_rt_alloc(
 	 * Allocate space to the file, as necessary.
 	 */
 	while (oblocks < nblocks) {
-		tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFSRT_ALLOC);
 		resblks = XFS_GROWFSRT_SPACE_RES(mp, nblocks - oblocks);
 		/*
 		 * Reserve space & log for one extent added to the file.
 		 */
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_growrtalloc,
-					  resblks, 0);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_growrtalloc, resblks,
+				0, 0, &tp);
 		if (error)
-			goto out_trans_cancel;
+			return error;
 		/*
 		 * Lock the inode.
 		 */
@@ -823,14 +822,13 @@ xfs_growfs_rt_alloc(
 		for (bno = map.br_startoff, fsbno = map.br_startblock;
 		     bno < map.br_startoff + map.br_blockcount;
 		     bno++, fsbno++) {
-			tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFSRT_ZERO);
 			/*
 			 * Reserve log for one block zeroing.
 			 */
-			error = xfs_trans_reserve(tp, &M_RES(mp)->tr_growrtzero,
-						  0, 0);
+			error = xfs_trans_alloc(mp, &M_RES(mp)->tr_growrtzero,
+					0, 0, 0, &tp);
 			if (error)
-				goto out_trans_cancel;
+				return error;
 			/*
 			 * Lock the bitmap inode.
 			 */
@@ -994,11 +992,10 @@ xfs_growfs_rt(
 		/*
 		 * Start a transaction, get the log reservation.
 		 */
-		tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFSRT_FREE);
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_growrtfree,
-					  0, 0);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_growrtfree, 0, 0, 0,
+				&tp);
 		if (error)
-			goto error_cancel;
+			break;
 		/*
 		 * Lock out other callers by grabbing the bitmap inode lock.
 		 */

fs/xfs/xfs_super.c

@@ -58,8 +58,7 @@
 #include <linux/parser.h>
 
 static const struct super_operations xfs_super_operations;
-static kmem_zone_t *xfs_ioend_zone;
-mempool_t *xfs_ioend_pool;
+struct bio_set *xfs_ioend_bioset;
 
 static struct kset *xfs_kset;		/* top-level xfs sysfs dir */
 #ifdef DEBUG
@@ -350,6 +349,7 @@ xfs_parseargs(
 		case Opt_pqnoenforce:
 			mp->m_qflags |= (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE);
 			mp->m_qflags &= ~XFS_PQUOTA_ENFD;
+			break;
 		case Opt_gquota:
 		case Opt_grpquota:
 			mp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE |
@@ -928,7 +928,7 @@ xfs_fs_alloc_inode(
 
 /*
  * Now that the generic code is guaranteed not to be accessing
- * the linux inode, we can reclaim the inode.
+ * the linux inode, we can inactivate and reclaim the inode.
  */
 STATIC void
 xfs_fs_destroy_inode(
@@ -938,9 +938,14 @@ xfs_fs_destroy_inode(
 
 	trace_xfs_destroy_inode(ip);
 
-	XFS_STATS_INC(ip->i_mount, vn_reclaim);
+	ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock));
+	XFS_STATS_INC(ip->i_mount, vn_rele);
+	XFS_STATS_INC(ip->i_mount, vn_remove);
+
+	xfs_inactive(ip);
 
 	ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) || ip->i_delayed_blks == 0);
+	XFS_STATS_INC(ip->i_mount, vn_reclaim);
 
 	/*
 	 * We should never get here with one of the reclaim flags already set.
@@ -987,24 +992,6 @@ xfs_fs_inode_init_once(
 		     "xfsino", ip->i_ino);
 }
 
-STATIC void
-xfs_fs_evict_inode(
-	struct inode		*inode)
-{
-	xfs_inode_t		*ip = XFS_I(inode);
-
-	ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock));
-
-	trace_xfs_evict_inode(ip);
-
-	truncate_inode_pages_final(&inode->i_data);
-	clear_inode(inode);
-	XFS_STATS_INC(ip->i_mount, vn_rele);
-	XFS_STATS_INC(ip->i_mount, vn_remove);
-
-	xfs_inactive(ip);
-}
-
 /*
  * We do an unlocked check for XFS_IDONTCACHE here because we are already
  * serialised against cache hits here via the inode->i_lock and igrab() in
@@ -1276,6 +1263,16 @@ xfs_fs_remount(
 			return -EINVAL;
 		}
 
+		if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5 &&
+		    xfs_sb_has_ro_compat_feature(sbp,
+					XFS_SB_FEAT_RO_COMPAT_UNKNOWN)) {
+			xfs_warn(mp,
+"ro->rw transition prohibited on unknown (0x%x) ro-compat filesystem",
+				(sbp->sb_features_ro_compat &
+					XFS_SB_FEAT_RO_COMPAT_UNKNOWN));
+			return -EINVAL;
+		}
+
 		mp->m_flags &= ~XFS_MOUNT_RDONLY;
 
 		/*
@@ -1663,7 +1660,6 @@ xfs_fs_free_cached_objects(
 static const struct super_operations xfs_super_operations = {
 	.alloc_inode		= xfs_fs_alloc_inode,
 	.destroy_inode		= xfs_fs_destroy_inode,
-	.evict_inode		= xfs_fs_evict_inode,
 	.drop_inode		= xfs_fs_drop_inode,
 	.put_super		= xfs_fs_put_super,
 	.sync_fs		= xfs_fs_sync_fs,
@@ -1688,20 +1684,15 @@ MODULE_ALIAS_FS("xfs");
 STATIC int __init
 xfs_init_zones(void)
 {
-
-	xfs_ioend_zone = kmem_zone_init(sizeof(xfs_ioend_t), "xfs_ioend");
-	if (!xfs_ioend_zone)
+	xfs_ioend_bioset = bioset_create(4 * MAX_BUF_PER_PAGE,
+			offsetof(struct xfs_ioend, io_inline_bio));
+	if (!xfs_ioend_bioset)
 		goto out;
 
-	xfs_ioend_pool = mempool_create_slab_pool(4 * MAX_BUF_PER_PAGE,
-						  xfs_ioend_zone);
-	if (!xfs_ioend_pool)
-		goto out_destroy_ioend_zone;
-
 	xfs_log_ticket_zone = kmem_zone_init(sizeof(xlog_ticket_t),
 						"xfs_log_ticket");
 	if (!xfs_log_ticket_zone)
-		goto out_destroy_ioend_pool;
+		goto out_free_ioend_bioset;
 
 	xfs_bmap_free_item_zone = kmem_zone_init(sizeof(xfs_bmap_free_item_t),
 						"xfs_bmap_free_item");
@@ -1797,10 +1788,8 @@ xfs_init_zones(void)
 	kmem_zone_destroy(xfs_bmap_free_item_zone);
  out_destroy_log_ticket_zone:
 	kmem_zone_destroy(xfs_log_ticket_zone);
- out_destroy_ioend_pool:
-	mempool_destroy(xfs_ioend_pool);
- out_destroy_ioend_zone:
-	kmem_zone_destroy(xfs_ioend_zone);
+ out_free_ioend_bioset:
+	bioset_free(xfs_ioend_bioset);
  out:
 	return -ENOMEM;
 }
@@ -1826,9 +1815,7 @@ xfs_destroy_zones(void)
 	kmem_zone_destroy(xfs_btree_cur_zone);
 	kmem_zone_destroy(xfs_bmap_free_item_zone);
 	kmem_zone_destroy(xfs_log_ticket_zone);
-	mempool_destroy(xfs_ioend_pool);
-	kmem_zone_destroy(xfs_ioend_zone);
-
+	bioset_free(xfs_ioend_bioset);
 }
 
 STATIC int __init

fs/xfs/xfs_symlink.c

@@ -131,6 +131,8 @@ xfs_readlink(
 
 	trace_xfs_readlink(ip);
 
+	ASSERT(!(ip->i_df.if_flags & XFS_IFINLINE));
+
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -EIO;
 
@@ -150,12 +152,7 @@ xfs_readlink(
 	}
 
 
-	if (ip->i_df.if_flags & XFS_IFINLINE) {
-		memcpy(link, ip->i_df.if_u1.if_data, pathlen);
-		link[pathlen] = '\0';
-	} else {
-		error = xfs_readlink_bmap(ip, link);
-	}
+	error = xfs_readlink_bmap(ip, link);
 
  out:
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
@@ -221,7 +218,6 @@ xfs_symlink(
 	if (error)
 		return error;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SYMLINK);
 	/*
 	 * The symlink will fit into the inode data fork?
 	 * There can't be any attributes so we get the whole variable part.
@@ -231,13 +227,15 @@ xfs_symlink(
 	else
 		fs_blocks = xfs_symlink_blocks(mp, pathlen);
 	resblks = XFS_SYMLINK_SPACE_RES(mp, link_name->len, fs_blocks);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_symlink, resblks, 0);
+
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_symlink, resblks, 0, 0, &tp);
 	if (error == -ENOSPC && fs_blocks == 0) {
 		resblks = 0;
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_symlink, 0, 0);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_symlink, 0, 0, 0,
+				&tp);
 	}
 	if (error)
-		goto out_trans_cancel;
+		goto out_release_inode;
 
 	xfs_ilock(dp, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL |
 		      XFS_IOLOCK_PARENT | XFS_ILOCK_PARENT);
@@ -302,19 +300,11 @@ xfs_symlink(
 	 * If the symlink will fit into the inode, write it inline.
 	 */
 	if (pathlen <= XFS_IFORK_DSIZE(ip)) {
-		xfs_idata_realloc(ip, pathlen, XFS_DATA_FORK);
-		memcpy(ip->i_df.if_u1.if_data, target_path, pathlen);
-		ip->i_d.di_size = pathlen;
-
-		/*
-		 * The inode was initially created in extent format.
-		 */
-		ip->i_df.if_flags &= ~(XFS_IFEXTENTS | XFS_IFBROOT);
-		ip->i_df.if_flags |= XFS_IFINLINE;
+		xfs_init_local_fork(ip, XFS_DATA_FORK, target_path, pathlen);
 
+		ip->i_d.di_size = pathlen;
 		ip->i_d.di_format = XFS_DINODE_FMT_LOCAL;
 		xfs_trans_log_inode(tp, ip, XFS_ILOG_DDATA | XFS_ILOG_CORE);
-
 	} else {
 		int	offset;
 
@@ -455,12 +445,9 @@ xfs_inactive_symlink_rmt(
 	 */
 	ASSERT(ip->i_d.di_nextents > 0 && ip->i_d.di_nextents <= 2);
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_INACTIVE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp);
+	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp);
+	if (error)
 		return error;
-	}
 
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	xfs_trans_ijoin(tp, ip, 0);

fs/xfs/xfs_sysfs.c

@@ -17,10 +17,11 @@
  */
 
 #include "xfs.h"
-#include "xfs_sysfs.h"
+#include "xfs_shared.h"
 #include "xfs_format.h"
 #include "xfs_log_format.h"
 #include "xfs_trans_resv.h"
+#include "xfs_sysfs.h"
 #include "xfs_log.h"
 #include "xfs_log_priv.h"
 #include "xfs_stats.h"
@@ -362,3 +363,291 @@ struct kobj_type xfs_log_ktype = {
 	.sysfs_ops = &xfs_sysfs_ops,
 	.default_attrs = xfs_log_attrs,
 };
+
+/*
+ * Metadata IO error configuration
+ *
+ * The sysfs structure here is:
+ *	...xfs/<dev>/error/<class>/<errno>/<error_attrs>
+ *
+ * where <class> allows us to discriminate between data IO and metadata IO,
+ * and any other future type of IO (e.g. special inode or directory error
+ * handling) we care to support.
+ */
+static inline struct xfs_error_cfg *
+to_error_cfg(struct kobject *kobject)
+{
+	struct xfs_kobj *kobj = to_kobj(kobject);
+	return container_of(kobj, struct xfs_error_cfg, kobj);
+}
+
+static inline struct xfs_mount *
+err_to_mp(struct kobject *kobject)
+{
+	struct xfs_kobj *kobj = to_kobj(kobject);
+	return container_of(kobj, struct xfs_mount, m_error_kobj);
+}
+
+static ssize_t
+max_retries_show(
+	struct kobject	*kobject,
+	char		*buf)
+{
+	struct xfs_error_cfg *cfg = to_error_cfg(kobject);
+
+	return snprintf(buf, PAGE_SIZE, "%d\n", cfg->max_retries);
+}
+
+static ssize_t
+max_retries_store(
+	struct kobject	*kobject,
+	const char	*buf,
+	size_t		count)
+{
+	struct xfs_error_cfg *cfg = to_error_cfg(kobject);
+	int		ret;
+	int		val;
+
+	ret = kstrtoint(buf, 0, &val);
+	if (ret)
+		return ret;
+
+	if (val < -1)
+		return -EINVAL;
+
+	cfg->max_retries = val;
+	return count;
+}
+XFS_SYSFS_ATTR_RW(max_retries);
+
+static ssize_t
+retry_timeout_seconds_show(
+	struct kobject	*kobject,
+	char		*buf)
+{
+	struct xfs_error_cfg *cfg = to_error_cfg(kobject);
+
+	return snprintf(buf, PAGE_SIZE, "%ld\n",
+			jiffies_to_msecs(cfg->retry_timeout) / MSEC_PER_SEC);
+}
+
+static ssize_t
+retry_timeout_seconds_store(
+	struct kobject	*kobject,
+	const char	*buf,
+	size_t		count)
+{
+	struct xfs_error_cfg *cfg = to_error_cfg(kobject);
+	int		ret;
+	int		val;
+
+	ret = kstrtoint(buf, 0, &val);
+	if (ret)
+		return ret;
+
+	/* 1 day timeout maximum */
+	if (val < 0 || val > 86400)
+		return -EINVAL;
+
+	cfg->retry_timeout = msecs_to_jiffies(val * MSEC_PER_SEC);
+	return count;
+}
+XFS_SYSFS_ATTR_RW(retry_timeout_seconds);
+
+static ssize_t
+fail_at_unmount_show(
+	struct kobject	*kobject,
+	char		*buf)
+{
+	struct xfs_mount	*mp = err_to_mp(kobject);
+
+	return snprintf(buf, PAGE_SIZE, "%d\n", mp->m_fail_unmount);
+}
+
+static ssize_t
+fail_at_unmount_store(
+	struct kobject	*kobject,
+	const char	*buf,
+	size_t		count)
+{
+	struct xfs_mount	*mp = err_to_mp(kobject);
+	int		ret;
+	int		val;
+
+	ret = kstrtoint(buf, 0, &val);
+	if (ret)
+		return ret;
+
+	if (val < 0 || val > 1)
+		return -EINVAL;
+
+	mp->m_fail_unmount = val;
+	return count;
+}
+XFS_SYSFS_ATTR_RW(fail_at_unmount);
+
+static struct attribute *xfs_error_attrs[] = {
+	ATTR_LIST(max_retries),
+	ATTR_LIST(retry_timeout_seconds),
+	NULL,
+};
+
+
+struct kobj_type xfs_error_cfg_ktype = {
+	.release = xfs_sysfs_release,
+	.sysfs_ops = &xfs_sysfs_ops,
+	.default_attrs = xfs_error_attrs,
+};
+
+struct kobj_type xfs_error_ktype = {
+	.release = xfs_sysfs_release,
+	.sysfs_ops = &xfs_sysfs_ops,
+};
+
+/*
+ * Error initialization tables. These need to be ordered in the same
+ * order as the enums used to index the array. All class init tables need to
+ * define a "default" behaviour as the first entry, all other entries can be
+ * empty.
+ */
+struct xfs_error_init {
+	char		*name;
+	int		max_retries;
+	int		retry_timeout;	/* in seconds */
+};
+
+static const struct xfs_error_init xfs_error_meta_init[XFS_ERR_ERRNO_MAX] = {
+	{ .name = "default",
+	  .max_retries = XFS_ERR_RETRY_FOREVER,
+	  .retry_timeout = 0,
+	},
+	{ .name = "EIO",
+	  .max_retries = XFS_ERR_RETRY_FOREVER,
+	  .retry_timeout = 0,
+	},
+	{ .name = "ENOSPC",
+	  .max_retries = XFS_ERR_RETRY_FOREVER,
+	  .retry_timeout = 0,
+	},
+	{ .name = "ENODEV",
+	  .max_retries = 0,
+	},
+};
+
+static int
+xfs_error_sysfs_init_class(
+	struct xfs_mount	*mp,
+	int			class,
+	const char		*parent_name,
+	struct xfs_kobj		*parent_kobj,
+	const struct xfs_error_init init[])
+{
+	struct xfs_error_cfg	*cfg;
+	int			error;
+	int			i;
+
+	ASSERT(class < XFS_ERR_CLASS_MAX);
+
+	error = xfs_sysfs_init(parent_kobj, &xfs_error_ktype,
+				&mp->m_error_kobj, parent_name);
+	if (error)
+		return error;
+
+	for (i = 0; i < XFS_ERR_ERRNO_MAX; i++) {
+		cfg = &mp->m_error_cfg[class][i];
+		error = xfs_sysfs_init(&cfg->kobj, &xfs_error_cfg_ktype,
+					parent_kobj, init[i].name);
+		if (error)
+			goto out_error;
+
+		cfg->max_retries = init[i].max_retries;
+		cfg->retry_timeout = msecs_to_jiffies(
+					init[i].retry_timeout * MSEC_PER_SEC);
+	}
+	return 0;
+
+out_error:
+	/* unwind the entries that succeeded */
+	for (i--; i >= 0; i--) {
+		cfg = &mp->m_error_cfg[class][i];
+		xfs_sysfs_del(&cfg->kobj);
+	}
+	xfs_sysfs_del(parent_kobj);
+	return error;
+}
+
+int
+xfs_error_sysfs_init(
+	struct xfs_mount	*mp)
+{
+	int			error;
+
+	/* .../xfs/<dev>/error/ */
+	error = xfs_sysfs_init(&mp->m_error_kobj, &xfs_error_ktype,
+				&mp->m_kobj, "error");
+	if (error)
+		return error;
+
+	error = sysfs_create_file(&mp->m_error_kobj.kobject,
+				  ATTR_LIST(fail_at_unmount));
+
+	if (error)
+		goto out_error;
+
+	/* .../xfs/<dev>/error/metadata/ */
+	error = xfs_error_sysfs_init_class(mp, XFS_ERR_METADATA,
+				"metadata", &mp->m_error_meta_kobj,
+				xfs_error_meta_init);
+	if (error)
+		goto out_error;
+
+	return 0;
+
+out_error:
+	xfs_sysfs_del(&mp->m_error_kobj);
+	return error;
+}
+
+void
+xfs_error_sysfs_del(
+	struct xfs_mount	*mp)
+{
+	struct xfs_error_cfg	*cfg;
+	int			i, j;
+
+	for (i = 0; i < XFS_ERR_CLASS_MAX; i++) {
+		for (j = 0; j < XFS_ERR_ERRNO_MAX; j++) {
+			cfg = &mp->m_error_cfg[i][j];
+
+			xfs_sysfs_del(&cfg->kobj);
+		}
+	}
+	xfs_sysfs_del(&mp->m_error_meta_kobj);
+	xfs_sysfs_del(&mp->m_error_kobj);
+}
+
+struct xfs_error_cfg *
+xfs_error_get_cfg(
+	struct xfs_mount	*mp,
+	int			error_class,
+	int			error)
+{
+	struct xfs_error_cfg	*cfg;
+
+	switch (error) {
+	case EIO:
+		cfg = &mp->m_error_cfg[error_class][XFS_ERR_EIO];
+		break;
+	case ENOSPC:
+		cfg = &mp->m_error_cfg[error_class][XFS_ERR_ENOSPC];
+		break;
+	case ENODEV:
+		cfg = &mp->m_error_cfg[error_class][XFS_ERR_ENODEV];
+		break;
+	default:
+		cfg = &mp->m_error_cfg[error_class][XFS_ERR_DEFAULT];
+		break;
+	}
+
+	return cfg;
+}

fs/xfs/xfs_sysfs.h

@@ -58,4 +58,7 @@ xfs_sysfs_del(
 	wait_for_completion(&kobj->complete);
 }
 
+int	xfs_error_sysfs_init(struct xfs_mount *mp);
+void	xfs_error_sysfs_del(struct xfs_mount *mp);
+
 #endif	/* __XFS_SYSFS_H__ */

fs/xfs/xfs_trace.h

@@ -364,7 +364,6 @@ DEFINE_BUF_EVENT(xfs_buf_delwri_split);
 DEFINE_BUF_EVENT(xfs_buf_get_uncached);
 DEFINE_BUF_EVENT(xfs_bdstrat_shut);
 DEFINE_BUF_EVENT(xfs_buf_item_relse);
-DEFINE_BUF_EVENT(xfs_buf_item_iodone);
 DEFINE_BUF_EVENT(xfs_buf_item_iodone_async);
 DEFINE_BUF_EVENT(xfs_buf_error_relse);
 DEFINE_BUF_EVENT(xfs_buf_wait_buftarg);
@@ -944,7 +943,6 @@ DECLARE_EVENT_CLASS(xfs_loggrant_class,
 	TP_ARGS(log, tic),
 	TP_STRUCT__entry(
 		__field(dev_t, dev)
-		__field(unsigned, trans_type)
 		__field(char, ocnt)
 		__field(char, cnt)
 		__field(int, curr_res)
@@ -962,7 +960,6 @@ DECLARE_EVENT_CLASS(xfs_loggrant_class,
 	),
 	TP_fast_assign(
 		__entry->dev = log->l_mp->m_super->s_dev;
-		__entry->trans_type = tic->t_trans_type;
 		__entry->ocnt = tic->t_ocnt;
 		__entry->cnt = tic->t_cnt;
 		__entry->curr_res = tic->t_curr_res;
@@ -980,14 +977,13 @@ DECLARE_EVENT_CLASS(xfs_loggrant_class,
 		__entry->curr_block = log->l_curr_block;
 		__entry->tail_lsn = atomic64_read(&log->l_tail_lsn);
 	),
-	TP_printk("dev %d:%d type %s t_ocnt %u t_cnt %u t_curr_res %u "
+	TP_printk("dev %d:%d t_ocnt %u t_cnt %u t_curr_res %u "
 		  "t_unit_res %u t_flags %s reserveq %s "
 		  "writeq %s grant_reserve_cycle %d "
 		  "grant_reserve_bytes %d grant_write_cycle %d "
 		  "grant_write_bytes %d curr_cycle %d curr_block %d "
 		  "tail_cycle %d tail_block %d",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
-		  __print_symbolic(__entry->trans_type, XFS_TRANS_TYPES),
 		  __entry->ocnt,
 		  __entry->cnt,
 		  __entry->curr_res,
@@ -1053,19 +1049,21 @@ DECLARE_EVENT_CLASS(xfs_log_item_class,
 )
 
 TRACE_EVENT(xfs_log_force,
-	TP_PROTO(struct xfs_mount *mp, xfs_lsn_t lsn),
-	TP_ARGS(mp, lsn),
+	TP_PROTO(struct xfs_mount *mp, xfs_lsn_t lsn, unsigned long caller_ip),
+	TP_ARGS(mp, lsn, caller_ip),
 	TP_STRUCT__entry(
 		__field(dev_t, dev)
 		__field(xfs_lsn_t, lsn)
+		__field(unsigned long, caller_ip)
 	),
 	TP_fast_assign(
 		__entry->dev = mp->m_super->s_dev;
 		__entry->lsn = lsn;
+		__entry->caller_ip = caller_ip;
 	),
-	TP_printk("dev %d:%d lsn 0x%llx",
+	TP_printk("dev %d:%d lsn 0x%llx caller %ps",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
-		  __entry->lsn)
+		  __entry->lsn, (void *)__entry->caller_ip)
 )
 
 #define DEFINE_LOG_ITEM_EVENT(name) \

fs/xfs/xfs_trans.c

@@ -46,47 +46,6 @@ xfs_trans_init(
 	xfs_trans_resv_calc(mp, M_RES(mp));
 }
 
-/*
- * This routine is called to allocate a transaction structure.
- * The type parameter indicates the type of the transaction.  These
- * are enumerated in xfs_trans.h.
- *
- * Dynamically allocate the transaction structure from the transaction
- * zone, initialize it, and return it to the caller.
- */
-xfs_trans_t *
-xfs_trans_alloc(
-	xfs_mount_t	*mp,
-	uint		type)
-{
-	xfs_trans_t     *tp;
-
-	sb_start_intwrite(mp->m_super);
-	tp = _xfs_trans_alloc(mp, type, KM_SLEEP);
-	tp->t_flags |= XFS_TRANS_FREEZE_PROT;
-	return tp;
-}
-
-xfs_trans_t *
-_xfs_trans_alloc(
-	xfs_mount_t	*mp,
-	uint		type,
-	xfs_km_flags_t	memflags)
-{
-	xfs_trans_t	*tp;
-
-	WARN_ON(mp->m_super->s_writers.frozen == SB_FREEZE_COMPLETE);
-	atomic_inc(&mp->m_active_trans);
-
-	tp = kmem_zone_zalloc(xfs_trans_zone, memflags);
-	tp->t_magic = XFS_TRANS_HEADER_MAGIC;
-	tp->t_type = type;
-	tp->t_mountp = mp;
-	INIT_LIST_HEAD(&tp->t_items);
-	INIT_LIST_HEAD(&tp->t_busy);
-	return tp;
-}
-
 /*
  * Free the transaction structure.  If there is more clean up
  * to do when the structure is freed, add it here.
@@ -99,7 +58,7 @@ xfs_trans_free(
 	xfs_extent_busy_clear(tp->t_mountp, &tp->t_busy, false);
 
 	atomic_dec(&tp->t_mountp->m_active_trans);
-	if (tp->t_flags & XFS_TRANS_FREEZE_PROT)
+	if (!(tp->t_flags & XFS_TRANS_NO_WRITECOUNT))
 		sb_end_intwrite(tp->t_mountp->m_super);
 	xfs_trans_free_dqinfo(tp);
 	kmem_zone_free(xfs_trans_zone, tp);
@@ -125,7 +84,6 @@ xfs_trans_dup(
 	 * Initialize the new transaction structure.
 	 */
 	ntp->t_magic = XFS_TRANS_HEADER_MAGIC;
-	ntp->t_type = tp->t_type;
 	ntp->t_mountp = tp->t_mountp;
 	INIT_LIST_HEAD(&ntp->t_items);
 	INIT_LIST_HEAD(&ntp->t_busy);
@@ -135,9 +93,9 @@ xfs_trans_dup(
 
 	ntp->t_flags = XFS_TRANS_PERM_LOG_RES |
 		       (tp->t_flags & XFS_TRANS_RESERVE) |
-		       (tp->t_flags & XFS_TRANS_FREEZE_PROT);
+		       (tp->t_flags & XFS_TRANS_NO_WRITECOUNT);
 	/* We gave our writer reference to the new transaction */
-	tp->t_flags &= ~XFS_TRANS_FREEZE_PROT;
+	tp->t_flags |= XFS_TRANS_NO_WRITECOUNT;
 	ntp->t_ticket = xfs_log_ticket_get(tp->t_ticket);
 	ntp->t_blk_res = tp->t_blk_res - tp->t_blk_res_used;
 	tp->t_blk_res = tp->t_blk_res_used;
@@ -165,7 +123,7 @@ xfs_trans_dup(
  * This does not do quota reservations. That typically is done by the
  * caller afterwards.
  */
-int
+static int
 xfs_trans_reserve(
 	struct xfs_trans	*tp,
 	struct xfs_trans_res	*resp,
@@ -219,7 +177,7 @@ xfs_trans_reserve(
 						resp->tr_logres,
 						resp->tr_logcount,
 						&tp->t_ticket, XFS_TRANSACTION,
-						permanent, tp->t_type);
+						permanent);
 		}
 
 		if (error)
@@ -268,6 +226,42 @@ undo_blocks:
 	return error;
 }
 
+int
+xfs_trans_alloc(
+	struct xfs_mount	*mp,
+	struct xfs_trans_res	*resp,
+	uint			blocks,
+	uint			rtextents,
+	uint			flags,
+	struct xfs_trans	**tpp)
+{
+	struct xfs_trans	*tp;
+	int			error;
+
+	if (!(flags & XFS_TRANS_NO_WRITECOUNT))
+		sb_start_intwrite(mp->m_super);
+
+	WARN_ON(mp->m_super->s_writers.frozen == SB_FREEZE_COMPLETE);
+	atomic_inc(&mp->m_active_trans);
+
+	tp = kmem_zone_zalloc(xfs_trans_zone,
+		(flags & XFS_TRANS_NOFS) ? KM_NOFS : KM_SLEEP);
+	tp->t_magic = XFS_TRANS_HEADER_MAGIC;
+	tp->t_flags = flags;
+	tp->t_mountp = mp;
+	INIT_LIST_HEAD(&tp->t_items);
+	INIT_LIST_HEAD(&tp->t_busy);
+
+	error = xfs_trans_reserve(tp, resp, blocks, rtextents);
+	if (error) {
+		xfs_trans_cancel(tp);
+		return error;
+	}
+
+	*tpp = tp;
+	return 0;
+}
+
 /*
  * Record the indicated change to the given field for application
  * to the file system's superblock when the transaction commits.

fs/xfs/xfs_trans.h

@@ -90,7 +90,6 @@ void	xfs_log_item_init(struct xfs_mount *mp, struct xfs_log_item *item,
  */
 typedef struct xfs_trans {
 	unsigned int		t_magic;	/* magic number */
-	unsigned int		t_type;		/* transaction type */
 	unsigned int		t_log_res;	/* amt of log space resvd */
 	unsigned int		t_log_count;	/* count for perm log res */
 	unsigned int		t_blk_res;	/* # of blocks resvd */
@@ -148,10 +147,9 @@ typedef struct xfs_trans {
 /*
  * XFS transaction mechanism exported interfaces.
  */
-xfs_trans_t	*xfs_trans_alloc(struct xfs_mount *, uint);
-xfs_trans_t	*_xfs_trans_alloc(struct xfs_mount *, uint, xfs_km_flags_t);
-int		xfs_trans_reserve(struct xfs_trans *, struct xfs_trans_res *,
-				  uint, uint);
+int		xfs_trans_alloc(struct xfs_mount *mp, struct xfs_trans_res *resp,
+			uint blocks, uint rtextents, uint flags,
+			struct xfs_trans **tpp);
 void		xfs_trans_mod_sb(xfs_trans_t *, uint, int64_t);
 
 struct xfs_buf	*xfs_trans_get_buf_map(struct xfs_trans *tp,

fs/xfs/xfs_xattr.c

@@ -146,7 +146,7 @@ __xfs_xattr_put_listent(
 	arraytop = context->count + prefix_len + namelen + 1;
 	if (arraytop > context->firstu) {
 		context->count = -1;	/* insufficient space */
-		return 1;
+		return 0;
 	}
 	offset = (char *)context->alist + context->count;
 	strncpy(offset, prefix, prefix_len);
@@ -166,8 +166,7 @@ xfs_xattr_put_listent(
 	int		flags,
 	unsigned char	*name,
 	int		namelen,
-	int		valuelen,
-	unsigned char	*value)
+	int		valuelen)
 {
 	char *prefix;
 	int prefix_len;
@@ -221,11 +220,15 @@ xfs_xattr_put_listent(
 }
 
 ssize_t
-xfs_vn_listxattr(struct dentry *dentry, char *data, size_t size)
+xfs_vn_listxattr(
+	struct dentry	*dentry,
+	char		*data,
+	size_t		size)
 {
 	struct xfs_attr_list_context context;
 	struct attrlist_cursor_kern cursor = { 0 };
-	struct inode		*inode = d_inode(dentry);
+	struct inode	*inode = d_inode(dentry);
+	int		error;
 
 	/*
 	 * First read the regular on-disk attributes.
@@ -239,7 +242,9 @@ xfs_vn_listxattr(struct dentry *dentry, char *data, size_t size)
 	context.firstu = context.bufsize;
 	context.put_listent = xfs_xattr_put_listent;
 
-	xfs_attr_list_int(&context);
+	error = xfs_attr_list_int(&context);
+	if (error)
+		return error;
 	if (context.count < 0)
 		return -ERANGE;