Merge tag 'xfs-for-linus-3.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs

Pull xfs update from Dave Chinner:
 "This update contains:

   - RENAME_EXCHANGE support

   - Rework of the superblock logging infrastructure

   - Rework of the XFS_IOCTL_SETXATTR implementation
       * enables use inside user namespaces
       * fixes inconsistencies setting extent size hints

   - fixes for missing buffer type annotations used in log recovery

   - more consolidation of libxfs headers

   - preparation patches for block based PNFS support

   - miscellaneous bug fixes and cleanups"

* tag 'xfs-for-linus-3.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs: (37 commits)
  xfs: only trace buffer items if they exist
  xfs: report proper f_files in statfs if we overshoot imaxpct
  xfs: fix panic_mask documentation
  xfs: xfs_ioctl_setattr_check_projid can be static
  xfs: growfs should use synchronous transactions
  xfs: fix behaviour of XFS_IOC_FSSETXATTR on directories
  xfs: factor projid hint checking out of xfs_ioctl_setattr
  xfs: factor extsize hint checking out of xfs_ioctl_setattr
  xfs: XFS_IOCTL_SETXATTR can run in user namespaces
  xfs: kill xfs_ioctl_setattr behaviour mask
  xfs: disaggregate xfs_ioctl_setattr
  xfs: factor out xfs_ioctl_setattr transaciton preamble
  xfs: separate xflags from xfs_ioctl_setattr
  xfs: FSX_NONBLOCK is not used
  xfs: don't allocate an ioend for direct I/O completions
  xfs: change kmem_free to use generic kvfree()
  xfs: factor out a xfs_update_prealloc_flags() helper
  xfs: remove incorrect error negation in attr_multi ioctl
  xfs: set superblock buffer type correctly
  xfs: set buf types when converting extent formats
  ...

Documentation/filesystems/xfs.txt

@@ -287,9 +287,9 @@ The following sysctls are available for the XFS filesystem:
 		XFS_ERRLEVEL_LOW:       1
 		XFS_ERRLEVEL_HIGH:      5
 
-  fs.xfs.panic_mask		(Min: 0  Default: 0  Max: 127)
+  fs.xfs.panic_mask		(Min: 0  Default: 0  Max: 255)
 	Causes certain error conditions to call BUG(). Value is a bitmask;
-	AND together the tags which represent errors which should cause panics:
+	OR together the tags which represent errors which should cause panics:
 
 		XFS_NO_PTAG                     0
 		XFS_PTAG_IFLUSH                 0x00000001
@@ -299,6 +299,7 @@ The following sysctls are available for the XFS filesystem:
 		XFS_PTAG_SHUTDOWN_CORRUPT       0x00000010
 		XFS_PTAG_SHUTDOWN_IOERROR       0x00000020
 		XFS_PTAG_SHUTDOWN_LOGERROR      0x00000040
+		XFS_PTAG_FSBLOCK_ZERO           0x00000080
 
 	This option is intended for debugging only.
 
@@ -348,16 +349,13 @@ The following sysctls are available for the XFS filesystem:
 Deprecated Sysctls
 ==================
 
-  fs.xfs.xfsbufd_centisecs	(Min: 50  Default: 100	Max: 3000)
-	Dirty metadata is now tracked by the log subsystem and
-	flushing is driven by log space and idling demands. The
-	xfsbufd no longer exists, so this syctl does nothing.
+None at present.
 
-	Due for removal in 3.14.
 
-  fs.xfs.age_buffer_centisecs	(Min: 100  Default: 1500  Max: 720000)
-	Dirty metadata is now tracked by the log subsystem and
-	flushing is driven by log space and idling demands. The
-	xfsbufd no longer exists, so this syctl does nothing.
+Removed Sysctls
+===============
 
-	Due for removal in 3.14.
+  Name				Removed
+  ----				-------
+  fs.xfs.xfsbufd_centisec	v3.20
+  fs.xfs.age_buffer_centisecs	v3.20

fs/xfs/kmem.c

@@ -91,16 +91,6 @@ kmem_zalloc_large(size_t size, xfs_km_flags_t flags)
 	return ptr;
 }
 
-void
-kmem_free(const void *ptr)
-{
-	if (!is_vmalloc_addr(ptr)) {
-		kfree(ptr);
-	} else {
-		vfree(ptr);
-	}
-}
-
 void *
 kmem_realloc(const void *ptr, size_t newsize, size_t oldsize,
 	     xfs_km_flags_t flags)

fs/xfs/kmem.h

@@ -63,7 +63,10 @@ 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_free(const void *);
+static inline void  kmem_free(const void *ptr)
+{
+	kvfree(ptr);
+}
 
 
 extern void *kmem_zalloc_greedy(size_t *, size_t, size_t);

fs/xfs/libxfs/xfs_attr_leaf.c

@@ -403,7 +403,7 @@ xfs_sbversion_add_attr2(xfs_mount_t *mp, xfs_trans_t *tp)
 		if (!xfs_sb_version_hasattr2(&mp->m_sb)) {
 			xfs_sb_version_addattr2(&mp->m_sb);
 			spin_unlock(&mp->m_sb_lock);
-			xfs_mod_sb(tp, XFS_SB_VERSIONNUM | XFS_SB_FEATURES2);
+			xfs_log_sb(tp);
 		} else
 			spin_unlock(&mp->m_sb_lock);
 	}

fs/xfs/libxfs/xfs_bmap.c

@@ -973,7 +973,11 @@ xfs_bmap_local_to_extents(
 	*firstblock = args.fsbno;
 	bp = xfs_btree_get_bufl(args.mp, tp, args.fsbno, 0);
 
-	/* initialise the block and copy the data */
+	/*
+	 * Initialise the block and copy the data
+	 *
+	 * Note: init_fn must set the buffer log item type correctly!
+	 */
 	init_fn(tp, bp, ip, ifp);
 
 	/* account for the change in fork size and log everything */
@@ -1221,22 +1225,20 @@ xfs_bmap_add_attrfork(
 		goto bmap_cancel;
 	if (!xfs_sb_version_hasattr(&mp->m_sb) ||
 	   (!xfs_sb_version_hasattr2(&mp->m_sb) && version == 2)) {
-		__int64_t sbfields = 0;
+		bool log_sb = false;
 
 		spin_lock(&mp->m_sb_lock);
 		if (!xfs_sb_version_hasattr(&mp->m_sb)) {
 			xfs_sb_version_addattr(&mp->m_sb);
-			sbfields |= XFS_SB_VERSIONNUM;
+			log_sb = true;
 		}
 		if (!xfs_sb_version_hasattr2(&mp->m_sb) && version == 2) {
 			xfs_sb_version_addattr2(&mp->m_sb);
-			sbfields |= (XFS_SB_VERSIONNUM | XFS_SB_FEATURES2);
+			log_sb = true;
 		}
-		if (sbfields) {
-			spin_unlock(&mp->m_sb_lock);
-			xfs_mod_sb(tp, sbfields);
-		} else
-			spin_unlock(&mp->m_sb_lock);
+		spin_unlock(&mp->m_sb_lock);
+		if (log_sb)
+			xfs_log_sb(tp);
 	}
 
 	error = xfs_bmap_finish(&tp, &flist, &committed);

fs/xfs/libxfs/xfs_bmap.h

@@ -27,6 +27,37 @@ struct xfs_trans;
 
 extern kmem_zone_t	*xfs_bmap_free_item_zone;
 
+/*
+ * Argument structure for xfs_bmap_alloc.
+ */
+struct xfs_bmalloca {
+	xfs_fsblock_t		*firstblock; /* i/o first block allocated */
+	struct xfs_bmap_free	*flist;	/* bmap freelist */
+	struct xfs_trans	*tp;	/* transaction pointer */
+	struct xfs_inode	*ip;	/* incore inode pointer */
+	struct xfs_bmbt_irec	prev;	/* extent before the new one */
+	struct xfs_bmbt_irec	got;	/* extent after, or delayed */
+
+	xfs_fileoff_t		offset;	/* offset in file filling in */
+	xfs_extlen_t		length;	/* i/o length asked/allocated */
+	xfs_fsblock_t		blkno;	/* starting block of new extent */
+
+	struct xfs_btree_cur	*cur;	/* btree cursor */
+	xfs_extnum_t		idx;	/* current extent index */
+	int			nallocs;/* number of extents alloc'd */
+	int			logflags;/* flags for transaction logging */
+
+	xfs_extlen_t		total;	/* total blocks needed for xaction */
+	xfs_extlen_t		minlen;	/* minimum allocation size (blocks) */
+	xfs_extlen_t		minleft; /* amount must be left after alloc */
+	bool			eof;	/* set if allocating past last extent */
+	bool			wasdel;	/* replacing a delayed allocation */
+	bool			userdata;/* set if is user data */
+	bool			aeof;	/* allocated space at eof */
+	bool			conv;	/* overwriting unwritten extents */
+	int			flags;
+};
+
 /*
  * List of extents to be free "later".
  * The list is kept sorted on xbf_startblock.
@@ -149,6 +180,8 @@ void	xfs_bmap_local_to_extents_empty(struct xfs_inode *ip, int whichfork);
 void	xfs_bmap_add_free(xfs_fsblock_t bno, xfs_filblks_t len,
 		struct xfs_bmap_free *flist, struct xfs_mount *mp);
 void	xfs_bmap_cancel(struct xfs_bmap_free *flist);
+int	xfs_bmap_finish(struct xfs_trans **tp, struct xfs_bmap_free *flist,
+			int *committed);
 void	xfs_bmap_compute_maxlevels(struct xfs_mount *mp, int whichfork);
 int	xfs_bmap_first_unused(struct xfs_trans *tp, struct xfs_inode *ip,
 		xfs_extlen_t len, xfs_fileoff_t *unused, int whichfork);

fs/xfs/libxfs/xfs_format.h

@@ -151,10 +151,13 @@ typedef struct xfs_sb {
 	__uint32_t	sb_features2;	/* additional feature bits */
 
 	/*
-	 * bad features2 field as a result of failing to pad the sb
-	 * structure to 64 bits. Some machines will be using this field
-	 * for features2 bits. Easiest just to mark it bad and not use
-	 * it for anything else.
+	 * bad features2 field as a result of failing to pad the sb structure to
+	 * 64 bits. Some machines will be using this field for features2 bits.
+	 * Easiest just to mark it bad and not use it for anything else.
+	 *
+	 * This is not kept up to date in memory; it is always overwritten by
+	 * the value in sb_features2 when formatting the incore superblock to
+	 * the disk buffer.
 	 */
 	__uint32_t	sb_bad_features2;
 
@@ -304,8 +307,8 @@ typedef enum {
 #define XFS_SB_ICOUNT		XFS_SB_MVAL(ICOUNT)
 #define XFS_SB_IFREE		XFS_SB_MVAL(IFREE)
 #define XFS_SB_FDBLOCKS		XFS_SB_MVAL(FDBLOCKS)
-#define XFS_SB_FEATURES2	XFS_SB_MVAL(FEATURES2)
-#define XFS_SB_BAD_FEATURES2	XFS_SB_MVAL(BAD_FEATURES2)
+#define XFS_SB_FEATURES2	(XFS_SB_MVAL(FEATURES2) | \
+				 XFS_SB_MVAL(BAD_FEATURES2))
 #define XFS_SB_FEATURES_COMPAT	XFS_SB_MVAL(FEATURES_COMPAT)
 #define XFS_SB_FEATURES_RO_COMPAT XFS_SB_MVAL(FEATURES_RO_COMPAT)
 #define XFS_SB_FEATURES_INCOMPAT XFS_SB_MVAL(FEATURES_INCOMPAT)
@@ -319,9 +322,9 @@ typedef enum {
 	 XFS_SB_VERSIONNUM | XFS_SB_UQUOTINO | XFS_SB_GQUOTINO | \
 	 XFS_SB_QFLAGS | XFS_SB_SHARED_VN | XFS_SB_UNIT | XFS_SB_WIDTH | \
 	 XFS_SB_ICOUNT | XFS_SB_IFREE | XFS_SB_FDBLOCKS | XFS_SB_FEATURES2 | \
-	 XFS_SB_BAD_FEATURES2 | XFS_SB_FEATURES_COMPAT | \
-	 XFS_SB_FEATURES_RO_COMPAT | XFS_SB_FEATURES_INCOMPAT | \
-	 XFS_SB_FEATURES_LOG_INCOMPAT | XFS_SB_PQUOTINO)
+	 XFS_SB_FEATURES_COMPAT | XFS_SB_FEATURES_RO_COMPAT | \
+	 XFS_SB_FEATURES_INCOMPAT | XFS_SB_FEATURES_LOG_INCOMPAT | \
+	 XFS_SB_PQUOTINO)
 
 
 /*
@@ -453,13 +456,11 @@ static inline void xfs_sb_version_addattr2(struct xfs_sb *sbp)
 {
 	sbp->sb_versionnum |= XFS_SB_VERSION_MOREBITSBIT;
 	sbp->sb_features2 |= XFS_SB_VERSION2_ATTR2BIT;
-	sbp->sb_bad_features2 |= XFS_SB_VERSION2_ATTR2BIT;
 }
 
 static inline void xfs_sb_version_removeattr2(struct xfs_sb *sbp)
 {
 	sbp->sb_features2 &= ~XFS_SB_VERSION2_ATTR2BIT;
-	sbp->sb_bad_features2 &= ~XFS_SB_VERSION2_ATTR2BIT;
 	if (!sbp->sb_features2)
 		sbp->sb_versionnum &= ~XFS_SB_VERSION_MOREBITSBIT;
 }
@@ -475,7 +476,6 @@ static inline void xfs_sb_version_addprojid32bit(struct xfs_sb *sbp)
 {
 	sbp->sb_versionnum |= XFS_SB_VERSION_MOREBITSBIT;
 	sbp->sb_features2 |= XFS_SB_VERSION2_PROJID32BIT;
-	sbp->sb_bad_features2 |= XFS_SB_VERSION2_PROJID32BIT;
 }
 
 /*

fs/xfs/libxfs/xfs_sb.c

@@ -40,69 +40,6 @@
  * Physical superblock buffer manipulations. Shared with libxfs in userspace.
  */
 
-static const struct {
-	short offset;
-	short type;	/* 0 = integer
-			 * 1 = binary / string (no translation)
-			 */
-} xfs_sb_info[] = {
-	{ offsetof(xfs_sb_t, sb_magicnum),	0 },
-	{ offsetof(xfs_sb_t, sb_blocksize),	0 },
-	{ offsetof(xfs_sb_t, sb_dblocks),	0 },
-	{ offsetof(xfs_sb_t, sb_rblocks),	0 },
-	{ offsetof(xfs_sb_t, sb_rextents),	0 },
-	{ offsetof(xfs_sb_t, sb_uuid),		1 },
-	{ offsetof(xfs_sb_t, sb_logstart),	0 },
-	{ offsetof(xfs_sb_t, sb_rootino),	0 },
-	{ offsetof(xfs_sb_t, sb_rbmino),	0 },
-	{ offsetof(xfs_sb_t, sb_rsumino),	0 },
-	{ offsetof(xfs_sb_t, sb_rextsize),	0 },
-	{ offsetof(xfs_sb_t, sb_agblocks),	0 },
-	{ offsetof(xfs_sb_t, sb_agcount),	0 },
-	{ offsetof(xfs_sb_t, sb_rbmblocks),	0 },
-	{ offsetof(xfs_sb_t, sb_logblocks),	0 },
-	{ offsetof(xfs_sb_t, sb_versionnum),	0 },
-	{ offsetof(xfs_sb_t, sb_sectsize),	0 },
-	{ offsetof(xfs_sb_t, sb_inodesize),	0 },
-	{ offsetof(xfs_sb_t, sb_inopblock),	0 },
-	{ offsetof(xfs_sb_t, sb_fname[0]),	1 },
-	{ offsetof(xfs_sb_t, sb_blocklog),	0 },
-	{ offsetof(xfs_sb_t, sb_sectlog),	0 },
-	{ offsetof(xfs_sb_t, sb_inodelog),	0 },
-	{ offsetof(xfs_sb_t, sb_inopblog),	0 },
-	{ offsetof(xfs_sb_t, sb_agblklog),	0 },
-	{ offsetof(xfs_sb_t, sb_rextslog),	0 },
-	{ offsetof(xfs_sb_t, sb_inprogress),	0 },
-	{ offsetof(xfs_sb_t, sb_imax_pct),	0 },
-	{ offsetof(xfs_sb_t, sb_icount),	0 },
-	{ offsetof(xfs_sb_t, sb_ifree),		0 },
-	{ offsetof(xfs_sb_t, sb_fdblocks),	0 },
-	{ offsetof(xfs_sb_t, sb_frextents),	0 },
-	{ offsetof(xfs_sb_t, sb_uquotino),	0 },
-	{ offsetof(xfs_sb_t, sb_gquotino),	0 },
-	{ offsetof(xfs_sb_t, sb_qflags),	0 },
-	{ offsetof(xfs_sb_t, sb_flags),		0 },
-	{ offsetof(xfs_sb_t, sb_shared_vn),	0 },
-	{ offsetof(xfs_sb_t, sb_inoalignmt),	0 },
-	{ offsetof(xfs_sb_t, sb_unit),		0 },
-	{ offsetof(xfs_sb_t, sb_width),		0 },
-	{ offsetof(xfs_sb_t, sb_dirblklog),	0 },
-	{ offsetof(xfs_sb_t, sb_logsectlog),	0 },
-	{ offsetof(xfs_sb_t, sb_logsectsize),	0 },
-	{ offsetof(xfs_sb_t, sb_logsunit),	0 },
-	{ offsetof(xfs_sb_t, sb_features2),	0 },
-	{ offsetof(xfs_sb_t, sb_bad_features2),	0 },
-	{ offsetof(xfs_sb_t, sb_features_compat),	0 },
-	{ offsetof(xfs_sb_t, sb_features_ro_compat),	0 },
-	{ offsetof(xfs_sb_t, sb_features_incompat),	0 },
-	{ offsetof(xfs_sb_t, sb_features_log_incompat),	0 },
-	{ offsetof(xfs_sb_t, sb_crc),		0 },
-	{ offsetof(xfs_sb_t, sb_pad),		0 },
-	{ offsetof(xfs_sb_t, sb_pquotino),	0 },
-	{ offsetof(xfs_sb_t, sb_lsn),		0 },
-	{ sizeof(xfs_sb_t),			0 }
-};
-
 /*
  * Reference counting access wrappers to the perag structures.
  * Because we never free per-ag structures, the only thing we
@@ -461,58 +398,49 @@ xfs_sb_from_disk(
 	__xfs_sb_from_disk(to, from, true);
 }
 
-static inline void
+static void
 xfs_sb_quota_to_disk(
-	xfs_dsb_t	*to,
-	xfs_sb_t	*from,
-	__int64_t	*fields)
+	struct xfs_dsb	*to,
+	struct xfs_sb	*from)
 {
 	__uint16_t	qflags = from->sb_qflags;
 
+	to->sb_uquotino = cpu_to_be64(from->sb_uquotino);
+	if (xfs_sb_version_has_pquotino(from)) {
+		to->sb_qflags = cpu_to_be16(from->sb_qflags);
+		to->sb_gquotino = cpu_to_be64(from->sb_gquotino);
+		to->sb_pquotino = cpu_to_be64(from->sb_pquotino);
+		return;
+	}
+
 	/*
-	 * We need to do these manipilations only if we are working
-	 * with an older version of on-disk superblock.
+	 * The in-core version of sb_qflags do not have XFS_OQUOTA_*
+	 * flags, whereas the on-disk version does.  So, convert incore
+	 * XFS_{PG}QUOTA_* flags to on-disk XFS_OQUOTA_* flags.
 	 */
-	if (xfs_sb_version_has_pquotino(from))
-		return;
+	qflags &= ~(XFS_PQUOTA_ENFD | XFS_PQUOTA_CHKD |
+			XFS_GQUOTA_ENFD | XFS_GQUOTA_CHKD);
 
-	if (*fields & XFS_SB_QFLAGS) {
-		/*
-		 * The in-core version of sb_qflags do not have
-		 * XFS_OQUOTA_* flags, whereas the on-disk version
-		 * does.  So, convert incore XFS_{PG}QUOTA_* flags
-		 * to on-disk XFS_OQUOTA_* flags.
-		 */
-		qflags &= ~(XFS_PQUOTA_ENFD | XFS_PQUOTA_CHKD |
-				XFS_GQUOTA_ENFD | XFS_GQUOTA_CHKD);
-
-		if (from->sb_qflags &
-				(XFS_PQUOTA_ENFD | XFS_GQUOTA_ENFD))
-			qflags |= XFS_OQUOTA_ENFD;
-		if (from->sb_qflags &
-				(XFS_PQUOTA_CHKD | XFS_GQUOTA_CHKD))
-			qflags |= XFS_OQUOTA_CHKD;
-		to->sb_qflags = cpu_to_be16(qflags);
-		*fields &= ~XFS_SB_QFLAGS;
-	}
+	if (from->sb_qflags &
+			(XFS_PQUOTA_ENFD | XFS_GQUOTA_ENFD))
+		qflags |= XFS_OQUOTA_ENFD;
+	if (from->sb_qflags &
+			(XFS_PQUOTA_CHKD | XFS_GQUOTA_CHKD))
+		qflags |= XFS_OQUOTA_CHKD;
+	to->sb_qflags = cpu_to_be16(qflags);
 
 	/*
-	 * GQUOTINO and PQUOTINO cannot be used together in versions of
-	 * superblock that do not have pquotino. from->sb_flags tells us which
-	 * quota is active and should be copied to disk. If neither are active,
-	 * make sure we write NULLFSINO to the sb_gquotino field as a quota
-	 * inode value of "0" is invalid when the XFS_SB_VERSION_QUOTA feature
-	 * bit is set.
+	 * GQUOTINO and PQUOTINO cannot be used together in versions
+	 * of superblock that do not have pquotino. from->sb_flags
+	 * tells us which quota is active and should be copied to
+	 * disk. If neither are active, we should NULL the inode.
 	 *
-	 * Note that we don't need to handle the sb_uquotino or sb_pquotino here
-	 * as they do not require any translation. Hence the main sb field loop
-	 * will write them appropriately from the in-core superblock.
+	 * In all cases, the separate pquotino must remain 0 because it
+	 * it beyond the "end" of the valid non-pquotino superblock.
 	 */
-	if ((*fields & XFS_SB_GQUOTINO) &&
-				(from->sb_qflags & XFS_GQUOTA_ACCT))
+	if (from->sb_qflags & XFS_GQUOTA_ACCT)
 		to->sb_gquotino = cpu_to_be64(from->sb_gquotino);
-	else if ((*fields & XFS_SB_PQUOTINO) &&
-				(from->sb_qflags & XFS_PQUOTA_ACCT))
+	else if (from->sb_qflags & XFS_PQUOTA_ACCT)
 		to->sb_gquotino = cpu_to_be64(from->sb_pquotino);
 	else {
 		/*
@@ -526,63 +454,78 @@ xfs_sb_quota_to_disk(
 			to->sb_gquotino = cpu_to_be64(NULLFSINO);
 	}
 
-	*fields &= ~(XFS_SB_PQUOTINO | XFS_SB_GQUOTINO);
+	to->sb_pquotino = 0;
 }
 
-/*
- * Copy in core superblock to ondisk one.
- *
- * The fields argument is mask of superblock fields to copy.
- */
 void
 xfs_sb_to_disk(
-	xfs_dsb_t	*to,
-	xfs_sb_t	*from,
-	__int64_t	fields)
+	struct xfs_dsb	*to,
+	struct xfs_sb	*from)
 {
-	xfs_caddr_t	to_ptr = (xfs_caddr_t)to;
-	xfs_caddr_t	from_ptr = (xfs_caddr_t)from;
-	xfs_sb_field_t	f;
-	int		first;
-	int		size;
-
-	ASSERT(fields);
-	if (!fields)
-		return;
+	xfs_sb_quota_to_disk(to, from);
 
-	/* We should never write the crc here, it's updated in the IO path */
-	fields &= ~XFS_SB_CRC;
-
-	xfs_sb_quota_to_disk(to, from, &fields);
-	while (fields) {
-		f = (xfs_sb_field_t)xfs_lowbit64((__uint64_t)fields);
-		first = xfs_sb_info[f].offset;
-		size = xfs_sb_info[f + 1].offset - first;
-
-		ASSERT(xfs_sb_info[f].type == 0 || xfs_sb_info[f].type == 1);
-
-		if (size == 1 || xfs_sb_info[f].type == 1) {
-			memcpy(to_ptr + first, from_ptr + first, size);
-		} else {
-			switch (size) {
-			case 2:
-				*(__be16 *)(to_ptr + first) =
-				      cpu_to_be16(*(__u16 *)(from_ptr + first));
-				break;
-			case 4:
-				*(__be32 *)(to_ptr + first) =
-				      cpu_to_be32(*(__u32 *)(from_ptr + first));
-				break;
-			case 8:
-				*(__be64 *)(to_ptr + first) =
-				      cpu_to_be64(*(__u64 *)(from_ptr + first));
-				break;
-			default:
-				ASSERT(0);
-			}
-		}
+	to->sb_magicnum = cpu_to_be32(from->sb_magicnum);
+	to->sb_blocksize = cpu_to_be32(from->sb_blocksize);
+	to->sb_dblocks = cpu_to_be64(from->sb_dblocks);
+	to->sb_rblocks = cpu_to_be64(from->sb_rblocks);
+	to->sb_rextents = cpu_to_be64(from->sb_rextents);
+	memcpy(&to->sb_uuid, &from->sb_uuid, sizeof(to->sb_uuid));
+	to->sb_logstart = cpu_to_be64(from->sb_logstart);
+	to->sb_rootino = cpu_to_be64(from->sb_rootino);
+	to->sb_rbmino = cpu_to_be64(from->sb_rbmino);
+	to->sb_rsumino = cpu_to_be64(from->sb_rsumino);
+	to->sb_rextsize = cpu_to_be32(from->sb_rextsize);
+	to->sb_agblocks = cpu_to_be32(from->sb_agblocks);
+	to->sb_agcount = cpu_to_be32(from->sb_agcount);
+	to->sb_rbmblocks = cpu_to_be32(from->sb_rbmblocks);
+	to->sb_logblocks = cpu_to_be32(from->sb_logblocks);
+	to->sb_versionnum = cpu_to_be16(from->sb_versionnum);
+	to->sb_sectsize = cpu_to_be16(from->sb_sectsize);
+	to->sb_inodesize = cpu_to_be16(from->sb_inodesize);
+	to->sb_inopblock = cpu_to_be16(from->sb_inopblock);
+	memcpy(&to->sb_fname, &from->sb_fname, sizeof(to->sb_fname));
+	to->sb_blocklog = from->sb_blocklog;
+	to->sb_sectlog = from->sb_sectlog;
+	to->sb_inodelog = from->sb_inodelog;
+	to->sb_inopblog = from->sb_inopblog;
+	to->sb_agblklog = from->sb_agblklog;
+	to->sb_rextslog = from->sb_rextslog;
+	to->sb_inprogress = from->sb_inprogress;
+	to->sb_imax_pct = from->sb_imax_pct;
+	to->sb_icount = cpu_to_be64(from->sb_icount);
+	to->sb_ifree = cpu_to_be64(from->sb_ifree);
+	to->sb_fdblocks = cpu_to_be64(from->sb_fdblocks);
+	to->sb_frextents = cpu_to_be64(from->sb_frextents);
 
-		fields &= ~(1LL << f);
+	to->sb_flags = from->sb_flags;
+	to->sb_shared_vn = from->sb_shared_vn;
+	to->sb_inoalignmt = cpu_to_be32(from->sb_inoalignmt);
+	to->sb_unit = cpu_to_be32(from->sb_unit);
+	to->sb_width = cpu_to_be32(from->sb_width);
+	to->sb_dirblklog = from->sb_dirblklog;
+	to->sb_logsectlog = from->sb_logsectlog;
+	to->sb_logsectsize = cpu_to_be16(from->sb_logsectsize);
+	to->sb_logsunit = cpu_to_be32(from->sb_logsunit);
+
+	/*
+	 * We need to ensure that bad_features2 always matches features2.
+	 * Hence we enforce that here rather than having to remember to do it
+	 * everywhere else that updates features2.
+	 */
+	from->sb_bad_features2 = from->sb_features2;
+	to->sb_features2 = cpu_to_be32(from->sb_features2);
+	to->sb_bad_features2 = cpu_to_be32(from->sb_bad_features2);
+
+	if (xfs_sb_version_hascrc(from)) {
+		to->sb_features_compat = cpu_to_be32(from->sb_features_compat);
+		to->sb_features_ro_compat =
+				cpu_to_be32(from->sb_features_ro_compat);
+		to->sb_features_incompat =
+				cpu_to_be32(from->sb_features_incompat);
+		to->sb_features_log_incompat =
+				cpu_to_be32(from->sb_features_log_incompat);
+		to->sb_pad = 0;
+		to->sb_lsn = cpu_to_be64(from->sb_lsn);
 	}
 }
 
@@ -816,42 +759,51 @@ xfs_initialize_perag_data(
 }
 
 /*
- * xfs_mod_sb() can be used to copy arbitrary changes to the
- * in-core superblock into the superblock buffer to be logged.
- * It does not provide the higher level of locking that is
- * needed to protect the in-core superblock from concurrent
- * access.
+ * xfs_log_sb() can be used to copy arbitrary changes to the in-core superblock
+ * into the superblock buffer to be logged.  It does not provide the higher
+ * level of locking that is needed to protect the in-core superblock from
+ * concurrent access.
  */
 void
-xfs_mod_sb(xfs_trans_t *tp, __int64_t fields)
+xfs_log_sb(
+	struct xfs_trans	*tp)
 {
-	xfs_buf_t	*bp;
-	int		first;
-	int		last;
-	xfs_mount_t	*mp;
-	xfs_sb_field_t	f;
-
-	ASSERT(fields);
-	if (!fields)
-		return;
-	mp = tp->t_mountp;
-	bp = xfs_trans_getsb(tp, mp, 0);
-	first = sizeof(xfs_sb_t);
-	last = 0;
-
-	/* translate/copy */
+	struct xfs_mount	*mp = tp->t_mountp;
+	struct xfs_buf		*bp = xfs_trans_getsb(tp, mp, 0);
 
-	xfs_sb_to_disk(XFS_BUF_TO_SBP(bp), &mp->m_sb, fields);
+	xfs_sb_to_disk(XFS_BUF_TO_SBP(bp), &mp->m_sb);
+	xfs_trans_buf_set_type(tp, bp, XFS_BLFT_SB_BUF);
+	xfs_trans_log_buf(tp, bp, 0, sizeof(struct xfs_dsb));
+}
 
-	/* find modified range */
-	f = (xfs_sb_field_t)xfs_highbit64((__uint64_t)fields);
-	ASSERT((1LL << f) & XFS_SB_MOD_BITS);
-	last = xfs_sb_info[f + 1].offset - 1;
+/*
+ * xfs_sync_sb
+ *
+ * Sync the superblock to disk.
+ *
+ * Note that the caller is responsible for checking the frozen state of the
+ * filesystem. This procedure uses the non-blocking transaction allocator and
+ * thus will allow modifications to a frozen fs. This is required because this
+ * code can be called during the process of freezing where use of the high-level
+ * allocator would deadlock.
+ */
+int
+xfs_sync_sb(
+	struct xfs_mount	*mp,
+	bool			wait)
+{
+	struct xfs_trans	*tp;
+	int			error;
 
-	f = (xfs_sb_field_t)xfs_lowbit64((__uint64_t)fields);
-	ASSERT((1LL << f) & XFS_SB_MOD_BITS);
-	first = xfs_sb_info[f].offset;
+	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, 0);
+		return error;
+	}
 
-	xfs_trans_buf_set_type(tp, bp, XFS_BLFT_SB_BUF);
-	xfs_trans_log_buf(tp, bp, first, last);
+	xfs_log_sb(tp);
+	if (wait)
+		xfs_trans_set_sync(tp);
+	return xfs_trans_commit(tp, 0);
 }

fs/xfs/libxfs/xfs_sb.h

@@ -27,11 +27,12 @@ extern struct xfs_perag *xfs_perag_get_tag(struct xfs_mount *, xfs_agnumber_t,
 extern void	xfs_perag_put(struct xfs_perag *pag);
 extern int	xfs_initialize_perag_data(struct xfs_mount *, xfs_agnumber_t);
 
-extern void	xfs_sb_calc_crc(struct xfs_buf	*);
-extern void	xfs_mod_sb(struct xfs_trans *, __int64_t);
-extern void	xfs_sb_mount_common(struct xfs_mount *, struct xfs_sb *);
-extern void	xfs_sb_from_disk(struct xfs_sb *, struct xfs_dsb *);
-extern void	xfs_sb_to_disk(struct xfs_dsb *, struct xfs_sb *, __int64_t);
+extern void	xfs_sb_calc_crc(struct xfs_buf *bp);
+extern void	xfs_log_sb(struct xfs_trans *tp);
+extern int	xfs_sync_sb(struct xfs_mount *mp, bool wait);
+extern void	xfs_sb_mount_common(struct xfs_mount *mp, struct xfs_sb *sbp);
+extern void	xfs_sb_from_disk(struct xfs_sb *to, struct xfs_dsb *from);
+extern void	xfs_sb_to_disk(struct xfs_dsb *to, struct xfs_sb *from);
 extern void	xfs_sb_quota_from_disk(struct xfs_sb *sbp);
 
 #endif	/* __XFS_SB_H__ */

fs/xfs/libxfs/xfs_shared.h

@@ -82,7 +82,7 @@ extern const struct xfs_buf_ops xfs_symlink_buf_ops;
 #define	XFS_TRANS_ATTR_RM		23
 #define	XFS_TRANS_ATTR_FLAG		24
 #define	XFS_TRANS_CLEAR_AGI_BUCKET	25
-#define XFS_TRANS_QM_SBCHANGE		26
+#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()
@@ -95,17 +95,15 @@ extern const struct xfs_buf_ops xfs_symlink_buf_ops;
 #define XFS_TRANS_QM_DQCLUSTER		32
 #define XFS_TRANS_QM_QINOCREATE		33
 #define XFS_TRANS_QM_QUOTAOFF_END	34
-#define XFS_TRANS_SB_UNIT		35
-#define XFS_TRANS_FSYNC_TS		36
-#define	XFS_TRANS_GROWFSRT_ALLOC	37
-#define	XFS_TRANS_GROWFSRT_ZERO		38
-#define	XFS_TRANS_GROWFSRT_FREE		39
-#define	XFS_TRANS_SWAPEXT		40
-#define	XFS_TRANS_SB_COUNT		41
-#define	XFS_TRANS_CHECKPOINT		42
-#define	XFS_TRANS_ICREATE		43
-#define	XFS_TRANS_CREATE_TMPFILE	44
-#define	XFS_TRANS_TYPE_MAX		44
+#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 \
@@ -113,7 +111,6 @@ extern const struct xfs_buf_ops xfs_symlink_buf_ops;
 	{ XFS_TRANS_SETATTR_SIZE,	"SETATTR_SIZE" }, \
 	{ XFS_TRANS_INACTIVE,		"INACTIVE" }, \
 	{ XFS_TRANS_CREATE,		"CREATE" }, \
-	{ XFS_TRANS_CREATE_TMPFILE,	"CREATE_TMPFILE" }, \
 	{ XFS_TRANS_CREATE_TRUNC,	"CREATE_TRUNC" }, \
 	{ XFS_TRANS_TRUNCATE_FILE,	"TRUNCATE_FILE" }, \
 	{ XFS_TRANS_REMOVE,		"REMOVE" }, \
@@ -134,23 +131,23 @@ extern const struct xfs_buf_ops xfs_symlink_buf_ops;
 	{ XFS_TRANS_ATTR_RM,		"ATTR_RM" }, \
 	{ XFS_TRANS_ATTR_FLAG,		"ATTR_FLAG" }, \
 	{ XFS_TRANS_CLEAR_AGI_BUCKET,	"CLEAR_AGI_BUCKET" }, \
-	{ XFS_TRANS_QM_SBCHANGE,	"QM_SBCHANGE" }, \
+	{ 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_SB_UNIT,		"SB_UNIT" }, \
 	{ 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_SB_COUNT,		"SB_COUNT" }, \
 	{ XFS_TRANS_CHECKPOINT,		"CHECKPOINT" }, \
-	{ XFS_TRANS_DUMMY1,		"DUMMY1" }, \
-	{ XFS_TRANS_DUMMY2,		"DUMMY2" }, \
+	{ XFS_TRANS_ICREATE,		"ICREATE" }, \
+	{ XFS_TRANS_CREATE_TMPFILE,	"CREATE_TMPFILE" }, \
 	{ XLOG_UNMOUNT_REC_TYPE,	"UNMOUNT" }
 
 /*

fs/xfs/libxfs/xfs_symlink_remote.c

@@ -178,6 +178,8 @@ xfs_symlink_local_to_remote(
 	struct xfs_mount	*mp = ip->i_mount;
 	char			*buf;
 
+	xfs_trans_buf_set_type(tp, bp, XFS_BLFT_SYMLINK_BUF);
+
 	if (!xfs_sb_version_hascrc(&mp->m_sb)) {
 		bp->b_ops = NULL;
 		memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);

fs/xfs/libxfs/xfs_trans_resv.c

@@ -715,17 +715,6 @@ xfs_calc_clear_agi_bucket_reservation(
 	return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
 }
 
-/*
- * Clearing the quotaflags in the superblock.
- *	the super block for changing quota flags: sector size
- */
-STATIC uint
-xfs_calc_qm_sbchange_reservation(
-	struct xfs_mount	*mp)
-{
-	return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
-}
-
 /*
  * Adjusting quota limits.
  *    the xfs_disk_dquot_t: sizeof(struct xfs_disk_dquot)
@@ -864,9 +853,6 @@ xfs_trans_resv_calc(
 	 * The following transactions are logged in logical format with
 	 * a default log count.
 	 */
-	resp->tr_qm_sbchange.tr_logres = xfs_calc_qm_sbchange_reservation(mp);
-	resp->tr_qm_sbchange.tr_logcount = XFS_DEFAULT_LOG_COUNT;
-
 	resp->tr_qm_setqlim.tr_logres = xfs_calc_qm_setqlim_reservation(mp);
 	resp->tr_qm_setqlim.tr_logcount = XFS_DEFAULT_LOG_COUNT;
 

fs/xfs/libxfs/xfs_trans_resv.h

@@ -56,7 +56,6 @@ struct xfs_trans_resv {
 	struct xfs_trans_res	tr_growrtalloc;	/* grow realtime allocations */
 	struct xfs_trans_res	tr_growrtzero;	/* grow realtime zeroing */
 	struct xfs_trans_res	tr_growrtfree;	/* grow realtime freeing */
-	struct xfs_trans_res	tr_qm_sbchange;	/* change quota flags */
 	struct xfs_trans_res	tr_qm_setqlim;	/* adjust quota limits */
 	struct xfs_trans_res	tr_qm_dqalloc;	/* allocate quota on disk */
 	struct xfs_trans_res	tr_qm_quotaoff;	/* turn quota off */

fs/xfs/xfs_aops.c

@@ -135,30 +135,22 @@ xfs_setfilesize_trans_alloc(
  */
 STATIC int
 xfs_setfilesize(
-	struct xfs_ioend	*ioend)
+	struct xfs_inode	*ip,
+	struct xfs_trans	*tp,
+	xfs_off_t		offset,
+	size_t			size)
 {
-	struct xfs_inode	*ip = XFS_I(ioend->io_inode);
-	struct xfs_trans	*tp = ioend->io_append_trans;
 	xfs_fsize_t		isize;
 
-	/*
-	 * The transaction may have been allocated in the I/O submission thread,
-	 * thus we need to mark ourselves as beeing in a transaction manually.
-	 * Similarly for freeze protection.
-	 */
-	current_set_flags_nested(&tp->t_pflags, PF_FSTRANS);
-	rwsem_acquire_read(&VFS_I(ip)->i_sb->s_writers.lock_map[SB_FREEZE_FS-1],
-			   0, 1, _THIS_IP_);
-
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	isize = xfs_new_eof(ip, ioend->io_offset + ioend->io_size);
+	isize = xfs_new_eof(ip, offset + size);
 	if (!isize) {
 		xfs_iunlock(ip, XFS_ILOCK_EXCL);
 		xfs_trans_cancel(tp, 0);
 		return 0;
 	}
 
-	trace_xfs_setfilesize(ip, ioend->io_offset, ioend->io_size);
+	trace_xfs_setfilesize(ip, offset, size);
 
 	ip->i_d.di_size = isize;
 	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
@@ -167,6 +159,25 @@ xfs_setfilesize(
 	return xfs_trans_commit(tp, 0);
 }
 
+STATIC int
+xfs_setfilesize_ioend(
+	struct xfs_ioend	*ioend)
+{
+	struct xfs_inode	*ip = XFS_I(ioend->io_inode);
+	struct xfs_trans	*tp = ioend->io_append_trans;
+
+	/*
+	 * The transaction may have been allocated in the I/O submission thread,
+	 * thus we need to mark ourselves as being in a transaction manually.
+	 * Similarly for freeze protection.
+	 */
+	current_set_flags_nested(&tp->t_pflags, PF_FSTRANS);
+	rwsem_acquire_read(&VFS_I(ip)->i_sb->s_writers.lock_map[SB_FREEZE_FS-1],
+			   0, 1, _THIS_IP_);
+
+	return xfs_setfilesize(ip, tp, ioend->io_offset, ioend->io_size);
+}
+
 /*
  * Schedule IO completion handling on the final put of an ioend.
  *
@@ -182,8 +193,7 @@ xfs_finish_ioend(
 
 		if (ioend->io_type == XFS_IO_UNWRITTEN)
 			queue_work(mp->m_unwritten_workqueue, &ioend->io_work);
-		else if (ioend->io_append_trans ||
-			 (ioend->io_isdirect && xfs_ioend_is_append(ioend)))
+		else if (ioend->io_append_trans)
 			queue_work(mp->m_data_workqueue, &ioend->io_work);
 		else
 			xfs_destroy_ioend(ioend);
@@ -215,22 +225,8 @@ xfs_end_io(
 	if (ioend->io_type == XFS_IO_UNWRITTEN) {
 		error = xfs_iomap_write_unwritten(ip, ioend->io_offset,
 						  ioend->io_size);
-	} else if (ioend->io_isdirect && xfs_ioend_is_append(ioend)) {
-		/*
-		 * For direct I/O we do not know if we need to allocate blocks
-		 * or not so we can't preallocate an append transaction as that
-		 * results in nested reservations and log space deadlocks. Hence
-		 * allocate the transaction here. While this is sub-optimal and
-		 * can block IO completion for some time, we're stuck with doing
-		 * it this way until we can pass the ioend to the direct IO
-		 * allocation callbacks and avoid nesting that way.
-		 */
-		error = xfs_setfilesize_trans_alloc(ioend);
-		if (error)
-			goto done;
-		error = xfs_setfilesize(ioend);
 	} else if (ioend->io_append_trans) {
-		error = xfs_setfilesize(ioend);
+		error = xfs_setfilesize_ioend(ioend);
 	} else {
 		ASSERT(!xfs_ioend_is_append(ioend));
 	}
@@ -241,17 +237,6 @@ xfs_end_io(
 	xfs_destroy_ioend(ioend);
 }
 
-/*
- * Call IO completion handling in caller context on the final put of an ioend.
- */
-STATIC void
-xfs_finish_ioend_sync(
-	struct xfs_ioend	*ioend)
-{
-	if (atomic_dec_and_test(&ioend->io_remaining))
-		xfs_end_io(&ioend->io_work);
-}
-
 /*
  * Allocate and initialise an IO completion structure.
  * We need to track unwritten extent write completion here initially.
@@ -273,7 +258,6 @@ xfs_alloc_ioend(
 	 * all the I/O from calling the completion routine too early.
 	 */
 	atomic_set(&ioend->io_remaining, 1);
-	ioend->io_isdirect = 0;
 	ioend->io_error = 0;
 	ioend->io_list = NULL;
 	ioend->io_type = type;
@@ -1459,11 +1443,7 @@ xfs_get_blocks_direct(
  *
  * If the private argument is non-NULL __xfs_get_blocks signals us that we
  * need to issue a transaction to convert the range from unwritten to written
- * extents.  In case this is regular synchronous I/O we just call xfs_end_io
- * to do this and we are done.  But in case this was a successful AIO
- * request this handler is called from interrupt context, from which we
- * can't start transactions.  In that case offload the I/O completion to
- * the workqueues we also use for buffered I/O completion.
+ * extents.
  */
 STATIC void
 xfs_end_io_direct_write(
@@ -1472,7 +1452,12 @@ xfs_end_io_direct_write(
 	ssize_t			size,
 	void			*private)
 {
-	struct xfs_ioend	*ioend = iocb->private;
+	struct inode		*inode = file_inode(iocb->ki_filp);
+	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_mount	*mp = ip->i_mount;
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return;
 
 	/*
 	 * While the generic direct I/O code updates the inode size, it does
@@ -1480,22 +1465,33 @@ xfs_end_io_direct_write(
 	 * end_io handler thinks the on-disk size is outside the in-core
 	 * size.  To prevent this just update it a little bit earlier here.
 	 */
-	if (offset + size > i_size_read(ioend->io_inode))
-		i_size_write(ioend->io_inode, offset + size);
+	if (offset + size > i_size_read(inode))
+		i_size_write(inode, offset + size);
 
 	/*
-	 * blockdev_direct_IO can return an error even after the I/O
-	 * completion handler was called.  Thus we need to protect
-	 * against double-freeing.
+	 * For direct I/O we do not know if we need to allocate blocks or not,
+	 * so we can't preallocate an append transaction, as that results in
+	 * nested reservations and log space deadlocks. Hence allocate the
+	 * transaction here. While this is sub-optimal and can block IO
+	 * completion for some time, we're stuck with doing it this way until
+	 * we can pass the ioend to the direct IO allocation callbacks and
+	 * avoid nesting that way.
 	 */
-	iocb->private = NULL;
-
-	ioend->io_offset = offset;
-	ioend->io_size = size;
-	if (private && size > 0)
-		ioend->io_type = XFS_IO_UNWRITTEN;
+	if (private && size > 0) {
+		xfs_iomap_write_unwritten(ip, offset, size);
+	} else if (offset + size > ip->i_d.di_size) {
+		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, 0);
+			return;
+		}
 
-	xfs_finish_ioend_sync(ioend);
+		xfs_setfilesize(ip, tp, offset, size);
+	}
 }
 
 STATIC ssize_t
@@ -1507,39 +1503,16 @@ xfs_vm_direct_IO(
 {
 	struct inode		*inode = iocb->ki_filp->f_mapping->host;
 	struct block_device	*bdev = xfs_find_bdev_for_inode(inode);
-	struct xfs_ioend	*ioend = NULL;
-	ssize_t			ret;
 
 	if (rw & WRITE) {
-		size_t size = iov_iter_count(iter);
-
-		/*
-		 * We cannot preallocate a size update transaction here as we
-		 * don't know whether allocation is necessary or not. Hence we
-		 * can only tell IO completion that one is necessary if we are
-		 * not doing unwritten extent conversion.
-		 */
-		iocb->private = ioend = xfs_alloc_ioend(inode, XFS_IO_DIRECT);
-		if (offset + size > XFS_I(inode)->i_d.di_size)
-			ioend->io_isdirect = 1;
-
-		ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iter,
+		return __blockdev_direct_IO(rw, iocb, inode, bdev, iter,
 					    offset, xfs_get_blocks_direct,
 					    xfs_end_io_direct_write, NULL,
 					    DIO_ASYNC_EXTEND);
-		if (ret != -EIOCBQUEUED && iocb->private)
-			goto out_destroy_ioend;
-	} else {
-		ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iter,
-					    offset, xfs_get_blocks_direct,
-					    NULL, NULL, 0);
 	}
-
-	return ret;
-
-out_destroy_ioend:
-	xfs_destroy_ioend(ioend);
-	return ret;
+	return __blockdev_direct_IO(rw, iocb, inode, bdev, iter,
+				    offset, xfs_get_blocks_direct,
+				    NULL, NULL, 0);
 }
 
 /*

fs/xfs/xfs_aops.h

@@ -24,14 +24,12 @@ extern mempool_t *xfs_ioend_pool;
  * Types of I/O for bmap clustering and I/O completion tracking.
  */
 enum {
-	XFS_IO_DIRECT = 0,	/* special case for direct I/O ioends */
 	XFS_IO_DELALLOC,	/* covers delalloc region */
 	XFS_IO_UNWRITTEN,	/* covers allocated but uninitialized data */
 	XFS_IO_OVERWRITE,	/* covers already allocated extent */
 };
 
 #define XFS_IO_TYPES \
-	{ 0,			"" }, \
 	{ XFS_IO_DELALLOC,		"delalloc" }, \
 	{ XFS_IO_UNWRITTEN,		"unwritten" }, \
 	{ XFS_IO_OVERWRITE,		"overwrite" }
@@ -45,7 +43,6 @@ typedef struct xfs_ioend {
 	unsigned int		io_type;	/* delalloc / unwritten */
 	int			io_error;	/* I/O error code */
 	atomic_t		io_remaining;	/* hold count */
-	unsigned int		io_isdirect : 1;/* direct I/O */
 	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 */

fs/xfs/xfs_bmap_util.h

@@ -26,43 +26,8 @@ struct xfs_ifork;
 struct xfs_inode;
 struct xfs_mount;
 struct xfs_trans;
+struct xfs_bmalloca;
 
-/*
- * Argument structure for xfs_bmap_alloc.
- */
-struct xfs_bmalloca {
-	xfs_fsblock_t		*firstblock; /* i/o first block allocated */
-	struct xfs_bmap_free	*flist;	/* bmap freelist */
-	struct xfs_trans	*tp;	/* transaction pointer */
-	struct xfs_inode	*ip;	/* incore inode pointer */
-	struct xfs_bmbt_irec	prev;	/* extent before the new one */
-	struct xfs_bmbt_irec	got;	/* extent after, or delayed */
-
-	xfs_fileoff_t		offset;	/* offset in file filling in */
-	xfs_extlen_t		length;	/* i/o length asked/allocated */
-	xfs_fsblock_t		blkno;	/* starting block of new extent */
-
-	struct xfs_btree_cur	*cur;	/* btree cursor */
-	xfs_extnum_t		idx;	/* current extent index */
-	int			nallocs;/* number of extents alloc'd */
-	int			logflags;/* flags for transaction logging */
-
-	xfs_extlen_t		total;	/* total blocks needed for xaction */
-	xfs_extlen_t		minlen;	/* minimum allocation size (blocks) */
-	xfs_extlen_t		minleft; /* amount must be left after alloc */
-	bool			eof;	/* set if allocating past last extent */
-	bool			wasdel;	/* replacing a delayed allocation */
-	bool			userdata;/* set if is user data */
-	bool			aeof;	/* allocated space at eof */
-	bool			conv;	/* overwriting unwritten extents */
-	int			flags;
-	struct completion	*done;
-	struct work_struct	work;
-	int			result;
-};
-
-int	xfs_bmap_finish(struct xfs_trans **tp, struct xfs_bmap_free *flist,
-			int *committed);
 int	xfs_bmap_rtalloc(struct xfs_bmalloca *ap);
 int	xfs_bmap_eof(struct xfs_inode *ip, xfs_fileoff_t endoff,
 		     int whichfork, int *eof);

fs/xfs/xfs_buf_item.c

@@ -319,6 +319,10 @@ xfs_buf_item_format(
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 	ASSERT((bip->bli_flags & XFS_BLI_LOGGED) ||
 	       (bip->bli_flags & XFS_BLI_STALE));
+	ASSERT((bip->bli_flags & XFS_BLI_STALE) ||
+	       (xfs_blft_from_flags(&bip->__bli_format) > XFS_BLFT_UNKNOWN_BUF
+	        && xfs_blft_from_flags(&bip->__bli_format) < XFS_BLFT_MAX_BUF));
+
 
 	/*
 	 * If it is an inode buffer, transfer the in-memory state to the
@@ -535,7 +539,7 @@ xfs_buf_item_push(
 	if ((bp->b_flags & XBF_WRITE_FAIL) &&
 	    ___ratelimit(&xfs_buf_write_fail_rl_state, "XFS:")) {
 		xfs_warn(bp->b_target->bt_mount,
-"Detected failing async write on buffer block 0x%llx. Retrying async write.\n",
+"Detected failing async write on buffer block 0x%llx. Retrying async write.",
 			 (long long)bp->b_bn);
 	}
 

fs/xfs/xfs_dquot.h

@@ -86,7 +86,7 @@ static inline void xfs_dqflock(xfs_dquot_t *dqp)
 	wait_for_completion(&dqp->q_flush);
 }
 
-static inline int xfs_dqflock_nowait(xfs_dquot_t *dqp)
+static inline bool xfs_dqflock_nowait(xfs_dquot_t *dqp)
 {
 	return try_wait_for_completion(&dqp->q_flush);
 }

fs/xfs/xfs_file.c

@@ -127,6 +127,42 @@ xfs_iozero(
 	return (-status);
 }
 
+int
+xfs_update_prealloc_flags(
+	struct xfs_inode	*ip,
+	enum xfs_prealloc_flags	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, 0);
+		return error;
+	}
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+
+	if (!(flags & XFS_PREALLOC_INVISIBLE)) {
+		ip->i_d.di_mode &= ~S_ISUID;
+		if (ip->i_d.di_mode & S_IXGRP)
+			ip->i_d.di_mode &= ~S_ISGID;
+		xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
+	}
+
+	if (flags & XFS_PREALLOC_SET)
+		ip->i_d.di_flags |= XFS_DIFLAG_PREALLOC;
+	if (flags & XFS_PREALLOC_CLEAR)
+		ip->i_d.di_flags &= ~XFS_DIFLAG_PREALLOC;
+
+	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+	if (flags & XFS_PREALLOC_SYNC)
+		xfs_trans_set_sync(tp);
+	return xfs_trans_commit(tp, 0);
+}
+
 /*
  * Fsync operations on directories are much simpler than on regular files,
  * as there is no file data to flush, and thus also no need for explicit
@@ -784,8 +820,8 @@ xfs_file_fallocate(
 {
 	struct inode		*inode = file_inode(file);
 	struct xfs_inode	*ip = XFS_I(inode);
-	struct xfs_trans	*tp;
 	long			error;
+	enum xfs_prealloc_flags	flags = 0;
 	loff_t			new_size = 0;
 
 	if (!S_ISREG(inode->i_mode))
@@ -822,6 +858,8 @@ xfs_file_fallocate(
 		if (error)
 			goto out_unlock;
 	} else {
+		flags |= XFS_PREALLOC_SET;
+
 		if (!(mode & FALLOC_FL_KEEP_SIZE) &&
 		    offset + len > i_size_read(inode)) {
 			new_size = offset + len;
@@ -839,28 +877,10 @@ xfs_file_fallocate(
 			goto out_unlock;
 	}
 
-	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, 0);
-		goto out_unlock;
-	}
-
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
-	ip->i_d.di_mode &= ~S_ISUID;
-	if (ip->i_d.di_mode & S_IXGRP)
-		ip->i_d.di_mode &= ~S_ISGID;
-
-	if (!(mode & (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_COLLAPSE_RANGE)))
-		ip->i_d.di_flags |= XFS_DIFLAG_PREALLOC;
-
-	xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
-	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-
 	if (file->f_flags & O_DSYNC)
-		xfs_trans_set_sync(tp);
-	error = xfs_trans_commit(tp, 0);
+		flags |= XFS_PREALLOC_SYNC;
+
+	error = xfs_update_prealloc_flags(ip, flags);
 	if (error)
 		goto out_unlock;
 

fs/xfs/xfs_fsops.c

@@ -488,6 +488,7 @@ xfs_growfs_data_private(
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_FDBLOCKS, nfree);
 	if (dpct)
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_IMAXPCT, dpct);
+	xfs_trans_set_sync(tp);
 	error = xfs_trans_commit(tp, 0);
 	if (error)
 		return error;
@@ -541,7 +542,7 @@ xfs_growfs_data_private(
 			saved_error = error;
 			continue;
 		}
-		xfs_sb_to_disk(XFS_BUF_TO_SBP(bp), &mp->m_sb, XFS_SB_ALL_BITS);
+		xfs_sb_to_disk(XFS_BUF_TO_SBP(bp), &mp->m_sb);
 
 		error = xfs_bwrite(bp);
 		xfs_buf_relse(bp);
@@ -756,37 +757,6 @@ xfs_reserve_blocks(
 	return 0;
 }
 
-/*
- * Dump a transaction into the log that contains no real change. This is needed
- * to be able to make the log dirty or stamp the current tail LSN into the log
- * during the covering operation.
- *
- * We cannot use an inode here for this - that will push dirty state back up
- * into the VFS and then periodic inode flushing will prevent log covering from
- * making progress. Hence we log a field in the superblock instead and use a
- * synchronous transaction to ensure the superblock is immediately unpinned
- * and can be written back.
- */
-int
-xfs_fs_log_dummy(
-	xfs_mount_t	*mp)
-{
-	xfs_trans_t	*tp;
-	int		error;
-
-	tp = _xfs_trans_alloc(mp, XFS_TRANS_DUMMY1, KM_SLEEP);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_sb, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp, 0);
-		return error;
-	}
-
-	/* log the UUID because it is an unchanging field */
-	xfs_mod_sb(tp, XFS_SB_UUID);
-	xfs_trans_set_sync(tp);
-	return xfs_trans_commit(tp, 0);
-}
-
 int
 xfs_fs_goingdown(
 	xfs_mount_t	*mp,

fs/xfs/xfs_inode.c

@@ -1995,6 +1995,7 @@ xfs_iunlink(
 	agi->agi_unlinked[bucket_index] = cpu_to_be32(agino);
 	offset = offsetof(xfs_agi_t, agi_unlinked) +
 		(sizeof(xfs_agino_t) * bucket_index);
+	xfs_trans_buf_set_type(tp, agibp, XFS_BLFT_AGI_BUF);
 	xfs_trans_log_buf(tp, agibp, offset,
 			  (offset + sizeof(xfs_agino_t) - 1));
 	return 0;
@@ -2086,6 +2087,7 @@ xfs_iunlink_remove(
 		agi->agi_unlinked[bucket_index] = cpu_to_be32(next_agino);
 		offset = offsetof(xfs_agi_t, agi_unlinked) +
 			(sizeof(xfs_agino_t) * bucket_index);
+		xfs_trans_buf_set_type(tp, agibp, XFS_BLFT_AGI_BUF);
 		xfs_trans_log_buf(tp, agibp, offset,
 				  (offset + sizeof(xfs_agino_t) - 1));
 	} else {
@@ -2655,6 +2657,124 @@ xfs_sort_for_rename(
 	}
 }
 
+/*
+ * xfs_cross_rename()
+ *
+ * responsible for handling RENAME_EXCHANGE flag in renameat2() sytemcall
+ */
+STATIC int
+xfs_cross_rename(
+	struct xfs_trans	*tp,
+	struct xfs_inode	*dp1,
+	struct xfs_name		*name1,
+	struct xfs_inode	*ip1,
+	struct xfs_inode	*dp2,
+	struct xfs_name		*name2,
+	struct xfs_inode	*ip2,
+	struct xfs_bmap_free	*free_list,
+	xfs_fsblock_t		*first_block,
+	int			spaceres)
+{
+	int		error = 0;
+	int		ip1_flags = 0;
+	int		ip2_flags = 0;
+	int		dp2_flags = 0;
+
+	/* Swap inode number for dirent in first parent */
+	error = xfs_dir_replace(tp, dp1, name1,
+				ip2->i_ino,
+				first_block, free_list, spaceres);
+	if (error)
+		goto out;
+
+	/* Swap inode number for dirent in second parent */
+	error = xfs_dir_replace(tp, dp2, name2,
+				ip1->i_ino,
+				first_block, free_list, spaceres);
+	if (error)
+		goto out;
+
+	/*
+	 * If we're renaming one or more directories across different parents,
+	 * update the respective ".." entries (and link counts) to match the new
+	 * parents.
+	 */
+	if (dp1 != dp2) {
+		dp2_flags = XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG;
+
+		if (S_ISDIR(ip2->i_d.di_mode)) {
+			error = xfs_dir_replace(tp, ip2, &xfs_name_dotdot,
+						dp1->i_ino, first_block,
+						free_list, spaceres);
+			if (error)
+				goto out;
+
+			/* transfer ip2 ".." reference to dp1 */
+			if (!S_ISDIR(ip1->i_d.di_mode)) {
+				error = xfs_droplink(tp, dp2);
+				if (error)
+					goto out;
+				error = xfs_bumplink(tp, dp1);
+				if (error)
+					goto out;
+			}
+
+			/*
+			 * Although ip1 isn't changed here, userspace needs
+			 * to be warned about the change, so that applications
+			 * relying on it (like backup ones), will properly
+			 * notify the change
+			 */
+			ip1_flags |= XFS_ICHGTIME_CHG;
+			ip2_flags |= XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG;
+		}
+
+		if (S_ISDIR(ip1->i_d.di_mode)) {
+			error = xfs_dir_replace(tp, ip1, &xfs_name_dotdot,
+						dp2->i_ino, first_block,
+						free_list, spaceres);
+			if (error)
+				goto out;
+
+			/* transfer ip1 ".." reference to dp2 */
+			if (!S_ISDIR(ip2->i_d.di_mode)) {
+				error = xfs_droplink(tp, dp1);
+				if (error)
+					goto out;
+				error = xfs_bumplink(tp, dp2);
+				if (error)
+					goto out;
+			}
+
+			/*
+			 * Although ip2 isn't changed here, userspace needs
+			 * to be warned about the change, so that applications
+			 * relying on it (like backup ones), will properly
+			 * notify the change
+			 */
+			ip1_flags |= XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG;
+			ip2_flags |= XFS_ICHGTIME_CHG;
+		}
+	}
+
+	if (ip1_flags) {
+		xfs_trans_ichgtime(tp, ip1, ip1_flags);
+		xfs_trans_log_inode(tp, ip1, XFS_ILOG_CORE);
+	}
+	if (ip2_flags) {
+		xfs_trans_ichgtime(tp, ip2, ip2_flags);
+		xfs_trans_log_inode(tp, ip2, XFS_ILOG_CORE);
+	}
+	if (dp2_flags) {
+		xfs_trans_ichgtime(tp, dp2, dp2_flags);
+		xfs_trans_log_inode(tp, dp2, XFS_ILOG_CORE);
+	}
+	xfs_trans_ichgtime(tp, dp1, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
+	xfs_trans_log_inode(tp, dp1, XFS_ILOG_CORE);
+out:
+	return error;
+}
+
 /*
  * xfs_rename
  */
@@ -2665,7 +2785,8 @@ xfs_rename(
 	xfs_inode_t	*src_ip,
 	xfs_inode_t	*target_dp,
 	struct xfs_name	*target_name,
-	xfs_inode_t	*target_ip)
+	xfs_inode_t	*target_ip,
+	unsigned int	flags)
 {
 	xfs_trans_t	*tp = NULL;
 	xfs_mount_t	*mp = src_dp->i_mount;
@@ -2742,6 +2863,18 @@ xfs_rename(
 		goto error_return;
 	}
 
+	/*
+	 * Handle RENAME_EXCHANGE flags
+	 */
+	if (flags & RENAME_EXCHANGE) {
+		error = xfs_cross_rename(tp, src_dp, src_name, src_ip,
+					 target_dp, target_name, target_ip,
+					 &free_list, &first_block, spaceres);
+		if (error)
+			goto abort_return;
+		goto finish_rename;
+	}
+
 	/*
 	 * Set up the target.
 	 */
@@ -2881,6 +3014,7 @@ xfs_rename(
 	if (new_parent)
 		xfs_trans_log_inode(tp, target_dp, XFS_ILOG_CORE);
 
+finish_rename:
 	/*
 	 * If this is a synchronous mount, make sure that the
 	 * rename transaction goes to disk before returning to

fs/xfs/xfs_inode.h

@@ -338,7 +338,7 @@ int		xfs_link(struct xfs_inode *tdp, struct xfs_inode *sip,
 int		xfs_rename(struct xfs_inode *src_dp, struct xfs_name *src_name,
 			   struct xfs_inode *src_ip, struct xfs_inode *target_dp,
 			   struct xfs_name *target_name,
-			   struct xfs_inode *target_ip);
+			   struct xfs_inode *target_ip, unsigned int flags);
 
 void		xfs_ilock(xfs_inode_t *, uint);
 int		xfs_ilock_nowait(xfs_inode_t *, uint);
@@ -377,6 +377,15 @@ int		xfs_droplink(struct xfs_trans *, struct xfs_inode *);
 int		xfs_bumplink(struct xfs_trans *, struct xfs_inode *);
 
 /* from xfs_file.c */
+enum xfs_prealloc_flags {
+	XFS_PREALLOC_SET	= (1 << 1),
+	XFS_PREALLOC_CLEAR	= (1 << 2),
+	XFS_PREALLOC_SYNC	= (1 << 3),
+	XFS_PREALLOC_INVISIBLE	= (1 << 4),
+};
+
+int		xfs_update_prealloc_flags(struct xfs_inode *,
+			enum xfs_prealloc_flags);
 int		xfs_zero_eof(struct xfs_inode *, xfs_off_t, xfs_fsize_t);
 int		xfs_iozero(struct xfs_inode *, loff_t, size_t);
 

fs/xfs/xfs_ioctl.c

@@ -606,11 +606,8 @@ xfs_ioc_space(
 	unsigned int		cmd,
 	xfs_flock64_t		*bf)
 {
-	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_trans	*tp;
 	struct iattr		iattr;
-	bool			setprealloc = false;
-	bool			clrprealloc = false;
+	enum xfs_prealloc_flags	flags = 0;
 	int			error;
 
 	/*
@@ -630,6 +627,11 @@ xfs_ioc_space(
 	if (!S_ISREG(inode->i_mode))
 		return -EINVAL;
 
+	if (filp->f_flags & O_DSYNC)
+		flags |= XFS_PREALLOC_SYNC;
+	if (ioflags & XFS_IO_INVIS)	
+		flags |= XFS_PREALLOC_INVISIBLE;
+
 	error = mnt_want_write_file(filp);
 	if (error)
 		return error;
@@ -673,25 +675,23 @@ xfs_ioc_space(
 	}
 
 	if (bf->l_start < 0 ||
-	    bf->l_start > mp->m_super->s_maxbytes ||
+	    bf->l_start > inode->i_sb->s_maxbytes ||
 	    bf->l_start + bf->l_len < 0 ||
-	    bf->l_start + bf->l_len >= mp->m_super->s_maxbytes) {
+	    bf->l_start + bf->l_len >= inode->i_sb->s_maxbytes) {
 		error = -EINVAL;
 		goto out_unlock;
 	}
 
 	switch (cmd) {
 	case XFS_IOC_ZERO_RANGE:
+		flags |= XFS_PREALLOC_SET;
 		error = xfs_zero_file_space(ip, bf->l_start, bf->l_len);
-		if (!error)
-			setprealloc = true;
 		break;
 	case XFS_IOC_RESVSP:
 	case XFS_IOC_RESVSP64:
+		flags |= XFS_PREALLOC_SET;
 		error = xfs_alloc_file_space(ip, bf->l_start, bf->l_len,
 						XFS_BMAPI_PREALLOC);
-		if (!error)
-			setprealloc = true;
 		break;
 	case XFS_IOC_UNRESVSP:
 	case XFS_IOC_UNRESVSP64:
@@ -701,6 +701,7 @@ xfs_ioc_space(
 	case XFS_IOC_ALLOCSP64:
 	case XFS_IOC_FREESP:
 	case XFS_IOC_FREESP64:
+		flags |= XFS_PREALLOC_CLEAR;
 		if (bf->l_start > XFS_ISIZE(ip)) {
 			error = xfs_alloc_file_space(ip, XFS_ISIZE(ip),
 					bf->l_start - XFS_ISIZE(ip), 0);
@@ -712,8 +713,6 @@ xfs_ioc_space(
 		iattr.ia_size = bf->l_start;
 
 		error = xfs_setattr_size(ip, &iattr);
-		if (!error)
-			clrprealloc = true;
 		break;
 	default:
 		ASSERT(0);
@@ -723,32 +722,7 @@ xfs_ioc_space(
 	if (error)
 		goto out_unlock;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_WRITEID);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_writeid, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp, 0);
-		goto out_unlock;
-	}
-
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
-
-	if (!(ioflags & XFS_IO_INVIS)) {
-		ip->i_d.di_mode &= ~S_ISUID;
-		if (ip->i_d.di_mode & S_IXGRP)
-			ip->i_d.di_mode &= ~S_ISGID;
-		xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
-	}
-
-	if (setprealloc)
-		ip->i_d.di_flags |= XFS_DIFLAG_PREALLOC;
-	else if (clrprealloc)
-		ip->i_d.di_flags &= ~XFS_DIFLAG_PREALLOC;
-
-	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-	if (filp->f_flags & O_DSYNC)
-		xfs_trans_set_sync(tp);
-	error = xfs_trans_commit(tp, 0);
+	error = xfs_update_prealloc_flags(ip, flags);
 
 out_unlock:
 	xfs_iunlock(ip, XFS_IOLOCK_EXCL);
@@ -1013,20 +987,182 @@ xfs_diflags_to_linux(
 		inode->i_flags &= ~S_NOATIME;
 }
 
-#define FSX_PROJID	1
-#define FSX_EXTSIZE	2
-#define FSX_XFLAGS	4
-#define FSX_NONBLOCK	8
+static int
+xfs_ioctl_setattr_xflags(
+	struct xfs_trans	*tp,
+	struct xfs_inode	*ip,
+	struct fsxattr		*fa)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+
+	/* Can't change realtime flag if any extents are allocated. */
+	if ((ip->i_d.di_nextents || ip->i_delayed_blks) &&
+	    XFS_IS_REALTIME_INODE(ip) != (fa->fsx_xflags & XFS_XFLAG_REALTIME))
+		return -EINVAL;
+
+	/* If realtime flag is set then must have realtime device */
+	if (fa->fsx_xflags & XFS_XFLAG_REALTIME) {
+		if (mp->m_sb.sb_rblocks == 0 || mp->m_sb.sb_rextsize == 0 ||
+		    (ip->i_d.di_extsize % mp->m_sb.sb_rextsize))
+			return -EINVAL;
+	}
+
+	/*
+	 * Can't modify an immutable/append-only file unless
+	 * we have appropriate permission.
+	 */
+	if (((ip->i_d.di_flags & (XFS_DIFLAG_IMMUTABLE | XFS_DIFLAG_APPEND)) ||
+	     (fa->fsx_xflags & (XFS_XFLAG_IMMUTABLE | XFS_XFLAG_APPEND))) &&
+	    !capable(CAP_LINUX_IMMUTABLE))
+		return -EPERM;
+
+	xfs_set_diflags(ip, fa->fsx_xflags);
+	xfs_diflags_to_linux(ip);
+	xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG);
+	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+	XFS_STATS_INC(xs_ig_attrchg);
+	return 0;
+}
+
+/*
+ * Set up the transaction structure for the setattr operation, checking that we
+ * have permission to do so. On success, return a clean transaction and the
+ * inode locked exclusively ready for further operation specific checks. On
+ * failure, return an error without modifying or locking the inode.
+ */
+static struct xfs_trans *
+xfs_ioctl_setattr_get_trans(
+	struct xfs_inode	*ip)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_trans	*tp;
+	int			error;
+
+	if (mp->m_flags & XFS_MOUNT_RDONLY)
+		return ERR_PTR(-EROFS);
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return ERR_PTR(-EIO);
+
+	tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
+	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0);
+	if (error)
+		goto out_cancel;
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+
+	/*
+	 * CAP_FOWNER overrides the following restrictions:
+	 *
+	 * The user ID of the calling process must be equal to the file owner
+	 * ID, except in cases where the CAP_FSETID capability is applicable.
+	 */
+	if (!inode_owner_or_capable(VFS_I(ip))) {
+		error = -EPERM;
+		goto out_cancel;
+	}
+
+	if (mp->m_flags & XFS_MOUNT_WSYNC)
+		xfs_trans_set_sync(tp);
+
+	return tp;
+
+out_cancel:
+	xfs_trans_cancel(tp, 0);
+	return ERR_PTR(error);
+}
+
+/*
+ * extent size hint validation is somewhat cumbersome. Rules are:
+ *
+ * 1. extent size hint is only valid for directories and regular files
+ * 2. XFS_XFLAG_EXTSIZE is only valid for regular files
+ * 3. XFS_XFLAG_EXTSZINHERIT is only valid for directories.
+ * 4. can only be changed on regular files if no extents are allocated
+ * 5. can be changed on directories at any time
+ * 6. extsize hint of 0 turns off hints, clears inode flags.
+ * 7. Extent size must be a multiple of the appropriate block size.
+ * 8. for non-realtime files, the extent size hint must be limited
+ *    to half the AG size to avoid alignment extending the extent beyond the
+ *    limits of the AG.
+ */
+static int
+xfs_ioctl_setattr_check_extsize(
+	struct xfs_inode	*ip,
+	struct fsxattr		*fa)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+
+	if ((fa->fsx_xflags & XFS_XFLAG_EXTSIZE) && !S_ISREG(ip->i_d.di_mode))
+		return -EINVAL;
+
+	if ((fa->fsx_xflags & XFS_XFLAG_EXTSZINHERIT) &&
+	    !S_ISDIR(ip->i_d.di_mode))
+		return -EINVAL;
+
+	if (S_ISREG(ip->i_d.di_mode) && ip->i_d.di_nextents &&
+	    ((ip->i_d.di_extsize << mp->m_sb.sb_blocklog) != fa->fsx_extsize))
+		return -EINVAL;
+
+	if (fa->fsx_extsize != 0) {
+		xfs_extlen_t    size;
+		xfs_fsblock_t   extsize_fsb;
+
+		extsize_fsb = XFS_B_TO_FSB(mp, fa->fsx_extsize);
+		if (extsize_fsb > MAXEXTLEN)
+			return -EINVAL;
+
+		if (XFS_IS_REALTIME_INODE(ip) ||
+		    (fa->fsx_xflags & XFS_XFLAG_REALTIME)) {
+			size = mp->m_sb.sb_rextsize << mp->m_sb.sb_blocklog;
+		} else {
+			size = mp->m_sb.sb_blocksize;
+			if (extsize_fsb > mp->m_sb.sb_agblocks / 2)
+				return -EINVAL;
+		}
+
+		if (fa->fsx_extsize % size)
+			return -EINVAL;
+	} else
+		fa->fsx_xflags &= ~(XFS_XFLAG_EXTSIZE | XFS_XFLAG_EXTSZINHERIT);
+
+	return 0;
+}
+
+static int
+xfs_ioctl_setattr_check_projid(
+	struct xfs_inode	*ip,
+	struct fsxattr		*fa)
+{
+	/* Disallow 32bit project ids if projid32bit feature is not enabled. */
+	if (fa->fsx_projid > (__uint16_t)-1 &&
+	    !xfs_sb_version_hasprojid32bit(&ip->i_mount->m_sb))
+		return -EINVAL;
+
+	/*
+	 * Project Quota ID state is only allowed to change from within the init
+	 * namespace. Enforce that restriction only if we are trying to change
+	 * the quota ID state. Everything else is allowed in user namespaces.
+	 */
+	if (current_user_ns() == &init_user_ns)
+		return 0;
+
+	if (xfs_get_projid(ip) != fa->fsx_projid)
+		return -EINVAL;
+	if ((fa->fsx_xflags & XFS_XFLAG_PROJINHERIT) !=
+	    (ip->i_d.di_flags & XFS_DIFLAG_PROJINHERIT))
+		return -EINVAL;
+
+	return 0;
+}
 
 STATIC int
 xfs_ioctl_setattr(
 	xfs_inode_t		*ip,
-	struct fsxattr		*fa,
-	int			mask)
+	struct fsxattr		*fa)
 {
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_trans	*tp;
-	unsigned int		lock_flags = 0;
 	struct xfs_dquot	*udqp = NULL;
 	struct xfs_dquot	*pdqp = NULL;
 	struct xfs_dquot	*olddquot = NULL;
@@ -1034,17 +1170,9 @@ xfs_ioctl_setattr(
 
 	trace_xfs_ioctl_setattr(ip);
 
-	if (mp->m_flags & XFS_MOUNT_RDONLY)
-		return -EROFS;
-	if (XFS_FORCED_SHUTDOWN(mp))
-		return -EIO;
-
-	/*
-	 * Disallow 32bit project ids when projid32bit feature is not enabled.
-	 */
-	if ((mask & FSX_PROJID) && (fa->fsx_projid > (__uint16_t)-1) &&
-			!xfs_sb_version_hasprojid32bit(&ip->i_mount->m_sb))
-		return -EINVAL;
+	code = xfs_ioctl_setattr_check_projid(ip, fa);
+	if (code)
+		return code;
 
 	/*
 	 * If disk quotas is on, we make sure that the dquots do exist on disk,
@@ -1054,7 +1182,7 @@ xfs_ioctl_setattr(
 	 * If the IDs do change before we take the ilock, we're covered
 	 * because the i_*dquot fields will get updated anyway.
 	 */
-	if (XFS_IS_QUOTA_ON(mp) && (mask & FSX_PROJID)) {
+	if (XFS_IS_QUOTA_ON(mp)) {
 		code = xfs_qm_vop_dqalloc(ip, ip->i_d.di_uid,
 					 ip->i_d.di_gid, fa->fsx_projid,
 					 XFS_QMOPT_PQUOTA, &udqp, NULL, &pdqp);
@@ -1062,175 +1190,49 @@ xfs_ioctl_setattr(
 			return code;
 	}
 
-	/*
-	 * For the other attributes, we acquire the inode lock and
-	 * first do an error checking pass.
-	 */
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
-	code = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0);
-	if (code)
-		goto error_return;
-
-	lock_flags = XFS_ILOCK_EXCL;
-	xfs_ilock(ip, lock_flags);
-
-	/*
-	 * CAP_FOWNER overrides the following restrictions:
-	 *
-	 * The user ID of the calling process must be equal
-	 * to the file owner ID, except in cases where the
-	 * CAP_FSETID capability is applicable.
-	 */
-	if (!inode_owner_or_capable(VFS_I(ip))) {
-		code = -EPERM;
-		goto error_return;
-	}
-
-	/*
-	 * Do a quota reservation only if projid is actually going to change.
-	 * Only allow changing of projid from init_user_ns since it is a
-	 * non user namespace aware identifier.
-	 */
-	if (mask & FSX_PROJID) {
-		if (current_user_ns() != &init_user_ns) {
-			code = -EINVAL;
-			goto error_return;
-		}
-
-		if (XFS_IS_QUOTA_RUNNING(mp) &&
-		    XFS_IS_PQUOTA_ON(mp) &&
-		    xfs_get_projid(ip) != fa->fsx_projid) {
-			ASSERT(tp);
-			code = xfs_qm_vop_chown_reserve(tp, ip, udqp, NULL,
-						pdqp, capable(CAP_FOWNER) ?
-						XFS_QMOPT_FORCE_RES : 0);
-			if (code)	/* out of quota */
-				goto error_return;
-		}
+	tp = xfs_ioctl_setattr_get_trans(ip);
+	if (IS_ERR(tp)) {
+		code = PTR_ERR(tp);
+		goto error_free_dquots;
 	}
 
-	if (mask & FSX_EXTSIZE) {
-		/*
-		 * Can't change extent size if any extents are allocated.
-		 */
-		if (ip->i_d.di_nextents &&
-		    ((ip->i_d.di_extsize << mp->m_sb.sb_blocklog) !=
-		     fa->fsx_extsize)) {
-			code = -EINVAL;	/* EFBIG? */
-			goto error_return;
-		}
 
-		/*
-		 * Extent size must be a multiple of the appropriate block
-		 * size, if set at all. It must also be smaller than the
-		 * maximum extent size supported by the filesystem.
-		 *
-		 * Also, for non-realtime files, limit the extent size hint to
-		 * half the size of the AGs in the filesystem so alignment
-		 * doesn't result in extents larger than an AG.
-		 */
-		if (fa->fsx_extsize != 0) {
-			xfs_extlen_t    size;
-			xfs_fsblock_t   extsize_fsb;
-
-			extsize_fsb = XFS_B_TO_FSB(mp, fa->fsx_extsize);
-			if (extsize_fsb > MAXEXTLEN) {
-				code = -EINVAL;
-				goto error_return;
-			}
-
-			if (XFS_IS_REALTIME_INODE(ip) ||
-			    ((mask & FSX_XFLAGS) &&
-			    (fa->fsx_xflags & XFS_XFLAG_REALTIME))) {
-				size = mp->m_sb.sb_rextsize <<
-				       mp->m_sb.sb_blocklog;
-			} else {
-				size = mp->m_sb.sb_blocksize;
-				if (extsize_fsb > mp->m_sb.sb_agblocks / 2) {
-					code = -EINVAL;
-					goto error_return;
-				}
-			}
-
-			if (fa->fsx_extsize % size) {
-				code = -EINVAL;
-				goto error_return;
-			}
-		}
+	if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_PQUOTA_ON(mp) &&
+	    xfs_get_projid(ip) != fa->fsx_projid) {
+		code = xfs_qm_vop_chown_reserve(tp, ip, udqp, NULL, pdqp,
+				capable(CAP_FOWNER) ?  XFS_QMOPT_FORCE_RES : 0);
+		if (code)	/* out of quota */
+			goto error_trans_cancel;
 	}
 
+	code = xfs_ioctl_setattr_check_extsize(ip, fa);
+	if (code)
+		goto error_trans_cancel;
 
-	if (mask & FSX_XFLAGS) {
-		/*
-		 * Can't change realtime flag if any extents are allocated.
-		 */
-		if ((ip->i_d.di_nextents || ip->i_delayed_blks) &&
-		    (XFS_IS_REALTIME_INODE(ip)) !=
-		    (fa->fsx_xflags & XFS_XFLAG_REALTIME)) {
-			code = -EINVAL;	/* EFBIG? */
-			goto error_return;
-		}
-
-		/*
-		 * If realtime flag is set then must have realtime data.
-		 */
-		if ((fa->fsx_xflags & XFS_XFLAG_REALTIME)) {
-			if ((mp->m_sb.sb_rblocks == 0) ||
-			    (mp->m_sb.sb_rextsize == 0) ||
-			    (ip->i_d.di_extsize % mp->m_sb.sb_rextsize)) {
-				code = -EINVAL;
-				goto error_return;
-			}
-		}
-
-		/*
-		 * Can't modify an immutable/append-only file unless
-		 * we have appropriate permission.
-		 */
-		if ((ip->i_d.di_flags &
-				(XFS_DIFLAG_IMMUTABLE|XFS_DIFLAG_APPEND) ||
-		     (fa->fsx_xflags &
-				(XFS_XFLAG_IMMUTABLE | XFS_XFLAG_APPEND))) &&
-		    !capable(CAP_LINUX_IMMUTABLE)) {
-			code = -EPERM;
-			goto error_return;
-		}
-	}
-
-	xfs_trans_ijoin(tp, ip, 0);
+	code = xfs_ioctl_setattr_xflags(tp, ip, fa);
+	if (code)
+		goto error_trans_cancel;
 
 	/*
-	 * Change file ownership.  Must be the owner or privileged.
+	 * Change file ownership.  Must be the owner or privileged.  CAP_FSETID
+	 * overrides the following restrictions:
+	 *
+	 * The set-user-ID and set-group-ID bits of a file will be cleared upon
+	 * successful return from chown()
 	 */
-	if (mask & FSX_PROJID) {
-		/*
-		 * CAP_FSETID overrides the following restrictions:
-		 *
-		 * The set-user-ID and set-group-ID bits of a file will be
-		 * cleared upon successful return from chown()
-		 */
-		if ((ip->i_d.di_mode & (S_ISUID|S_ISGID)) &&
-		    !capable_wrt_inode_uidgid(VFS_I(ip), CAP_FSETID))
-			ip->i_d.di_mode &= ~(S_ISUID|S_ISGID);
-
-		/*
-		 * Change the ownerships and register quota modifications
-		 * in the transaction.
-		 */
-		if (xfs_get_projid(ip) != fa->fsx_projid) {
-			if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_PQUOTA_ON(mp)) {
-				olddquot = xfs_qm_vop_chown(tp, ip,
-							&ip->i_pdquot, pdqp);
-			}
-			ASSERT(ip->i_d.di_version > 1);
-			xfs_set_projid(ip, fa->fsx_projid);
-		}
 
-	}
+	if ((ip->i_d.di_mode & (S_ISUID|S_ISGID)) &&
+	    !capable_wrt_inode_uidgid(VFS_I(ip), CAP_FSETID))
+		ip->i_d.di_mode &= ~(S_ISUID|S_ISGID);
 
-	if (mask & FSX_XFLAGS) {
-		xfs_set_diflags(ip, fa->fsx_xflags);
-		xfs_diflags_to_linux(ip);
+	/* Change the ownerships and register project quota modifications */
+	if (xfs_get_projid(ip) != fa->fsx_projid) {
+		if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_PQUOTA_ON(mp)) {
+			olddquot = xfs_qm_vop_chown(tp, ip,
+						&ip->i_pdquot, pdqp);
+		}
+		ASSERT(ip->i_d.di_version > 1);
+		xfs_set_projid(ip, fa->fsx_projid);
 	}
 
 	/*
@@ -1238,34 +1240,12 @@ xfs_ioctl_setattr(
 	 * extent size hint should be set on the inode. If no extent size flags
 	 * are set on the inode then unconditionally clear the extent size hint.
 	 */
-	if (mask & FSX_EXTSIZE) {
-		int	extsize = 0;
-
-		if (ip->i_d.di_flags &
-				(XFS_DIFLAG_EXTSIZE | XFS_DIFLAG_EXTSZINHERIT))
-			extsize = fa->fsx_extsize >> mp->m_sb.sb_blocklog;
-		ip->i_d.di_extsize = extsize;
-	}
-
-	xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG);
-	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-
-	XFS_STATS_INC(xs_ig_attrchg);
+	if (ip->i_d.di_flags & (XFS_DIFLAG_EXTSIZE | XFS_DIFLAG_EXTSZINHERIT))
+		ip->i_d.di_extsize = fa->fsx_extsize >> mp->m_sb.sb_blocklog;
+	else
+		ip->i_d.di_extsize = 0;
 
-	/*
-	 * If this is a synchronous mount, make sure that the
-	 * transaction goes to disk before returning to the user.
-	 * This is slightly sub-optimal in that truncates require
-	 * two sync transactions instead of one for wsync filesystems.
-	 * One for the truncate and one for the timestamps since we
-	 * don't want to change the timestamps unless we're sure the
-	 * truncate worked.  Truncates are less than 1% of the laddis
-	 * mix so this probably isn't worth the trouble to optimize.
-	 */
-	if (mp->m_flags & XFS_MOUNT_WSYNC)
-		xfs_trans_set_sync(tp);
 	code = xfs_trans_commit(tp, 0);
-	xfs_iunlock(ip, lock_flags);
 
 	/*
 	 * Release any dquot(s) the inode had kept before chown.
@@ -1276,12 +1256,11 @@ xfs_ioctl_setattr(
 
 	return code;
 
- error_return:
+error_trans_cancel:
+	xfs_trans_cancel(tp, 0);
+error_free_dquots:
 	xfs_qm_dqrele(udqp);
 	xfs_qm_dqrele(pdqp);
-	xfs_trans_cancel(tp, 0);
-	if (lock_flags)
-		xfs_iunlock(ip, lock_flags);
 	return code;
 }
 
@@ -1292,20 +1271,15 @@ xfs_ioc_fssetxattr(
 	void			__user *arg)
 {
 	struct fsxattr		fa;
-	unsigned int		mask;
 	int error;
 
 	if (copy_from_user(&fa, arg, sizeof(fa)))
 		return -EFAULT;
 
-	mask = FSX_XFLAGS | FSX_EXTSIZE | FSX_PROJID;
-	if (filp->f_flags & (O_NDELAY|O_NONBLOCK))
-		mask |= FSX_NONBLOCK;
-
 	error = mnt_want_write_file(filp);
 	if (error)
 		return error;
-	error = xfs_ioctl_setattr(ip, &fa, mask);
+	error = xfs_ioctl_setattr(ip, &fa);
 	mnt_drop_write_file(filp);
 	return error;
 }
@@ -1325,14 +1299,14 @@ xfs_ioc_getxflags(
 
 STATIC int
 xfs_ioc_setxflags(
-	xfs_inode_t		*ip,
+	struct xfs_inode	*ip,
 	struct file		*filp,
 	void			__user *arg)
 {
+	struct xfs_trans	*tp;
 	struct fsxattr		fa;
 	unsigned int		flags;
-	unsigned int		mask;
-	int error;
+	int			error;
 
 	if (copy_from_user(&flags, arg, sizeof(flags)))
 		return -EFAULT;
@@ -1342,15 +1316,26 @@ xfs_ioc_setxflags(
 		      FS_SYNC_FL))
 		return -EOPNOTSUPP;
 
-	mask = FSX_XFLAGS;
-	if (filp->f_flags & (O_NDELAY|O_NONBLOCK))
-		mask |= FSX_NONBLOCK;
 	fa.fsx_xflags = xfs_merge_ioc_xflags(flags, xfs_ip2xflags(ip));
 
 	error = mnt_want_write_file(filp);
 	if (error)
 		return error;
-	error = xfs_ioctl_setattr(ip, &fa, mask);
+
+	tp = xfs_ioctl_setattr_get_trans(ip);
+	if (IS_ERR(tp)) {
+		error = PTR_ERR(tp);
+		goto out_drop_write;
+	}
+
+	error = xfs_ioctl_setattr_xflags(tp, ip, &fa);
+	if (error) {
+		xfs_trans_cancel(tp, 0);
+		goto out_drop_write;
+	}
+
+	error = xfs_trans_commit(tp, 0);
+out_drop_write:
 	mnt_drop_write_file(filp);
 	return error;
 }

fs/xfs/xfs_ioctl32.c

@@ -423,7 +423,7 @@ xfs_compat_attrmulti_by_handle(
 
 	ops = memdup_user(compat_ptr(am_hreq.ops), size);
 	if (IS_ERR(ops)) {
-		error = -PTR_ERR(ops);
+		error = PTR_ERR(ops);
 		goto out_dput;
 	}
 

fs/xfs/xfs_iomap.c

@@ -802,7 +802,7 @@ int
 xfs_iomap_write_unwritten(
 	xfs_inode_t	*ip,
 	xfs_off_t	offset,
-	size_t		count)
+	xfs_off_t	count)
 {
 	xfs_mount_t	*mp = ip->i_mount;
 	xfs_fileoff_t	offset_fsb;

fs/xfs/xfs_iomap.h

@@ -27,6 +27,6 @@ int xfs_iomap_write_delay(struct xfs_inode *, xfs_off_t, size_t,
 			struct xfs_bmbt_irec *);
 int xfs_iomap_write_allocate(struct xfs_inode *, xfs_off_t,
 			struct xfs_bmbt_irec *);
-int xfs_iomap_write_unwritten(struct xfs_inode *, xfs_off_t, size_t);
+int xfs_iomap_write_unwritten(struct xfs_inode *, xfs_off_t, xfs_off_t);
 
 #endif /* __XFS_IOMAP_H__*/

fs/xfs/xfs_iops.c

@@ -380,18 +380,27 @@ xfs_vn_rename(
 	struct inode	*odir,
 	struct dentry	*odentry,
 	struct inode	*ndir,
-	struct dentry	*ndentry)
+	struct dentry	*ndentry,
+	unsigned int	flags)
 {
 	struct inode	*new_inode = ndentry->d_inode;
+	int		omode = 0;
 	struct xfs_name	oname;
 	struct xfs_name	nname;
 
-	xfs_dentry_to_name(&oname, odentry, 0);
+	if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE))
+		return -EINVAL;
+
+	/* if we are exchanging files, we need to set i_mode of both files */
+	if (flags & RENAME_EXCHANGE)
+		omode = ndentry->d_inode->i_mode;
+
+	xfs_dentry_to_name(&oname, odentry, omode);
 	xfs_dentry_to_name(&nname, ndentry, odentry->d_inode->i_mode);
 
 	return xfs_rename(XFS_I(odir), &oname, XFS_I(odentry->d_inode),
-			  XFS_I(ndir), &nname, new_inode ?
-						XFS_I(new_inode) : NULL);
+			  XFS_I(ndir), &nname,
+			  new_inode ? XFS_I(new_inode) : NULL, flags);
 }
 
 /*
@@ -1144,7 +1153,7 @@ static const struct inode_operations xfs_dir_inode_operations = {
 	 */
 	.rmdir			= xfs_vn_unlink,
 	.mknod			= xfs_vn_mknod,
-	.rename			= xfs_vn_rename,
+	.rename2		= xfs_vn_rename,
 	.get_acl		= xfs_get_acl,
 	.set_acl		= xfs_set_acl,
 	.getattr		= xfs_vn_getattr,
@@ -1172,7 +1181,7 @@ static const struct inode_operations xfs_dir_ci_inode_operations = {
 	 */
 	.rmdir			= xfs_vn_unlink,
 	.mknod			= xfs_vn_mknod,
-	.rename			= xfs_vn_rename,
+	.rename2		= xfs_vn_rename,
 	.get_acl		= xfs_get_acl,
 	.set_acl		= xfs_set_acl,
 	.getattr		= xfs_vn_getattr,

fs/xfs/xfs_log.c

@@ -33,6 +33,7 @@
 #include "xfs_fsops.h"
 #include "xfs_cksum.h"
 #include "xfs_sysfs.h"
+#include "xfs_sb.h"
 
 kmem_zone_t	*xfs_log_ticket_zone;
 
@@ -1290,9 +1291,20 @@ xfs_log_worker(
 	struct xfs_mount	*mp = log->l_mp;
 
 	/* dgc: errors ignored - not fatal and nowhere to report them */
-	if (xfs_log_need_covered(mp))
-		xfs_fs_log_dummy(mp);
-	else
+	if (xfs_log_need_covered(mp)) {
+		/*
+		 * Dump a transaction into the log that contains no real change.
+		 * This is needed to stamp the current tail LSN into the log
+		 * during the covering operation.
+		 *
+		 * We cannot use an inode here for this - that will push dirty
+		 * state back up into the VFS and then periodic inode flushing
+		 * will prevent log covering from making progress. Hence we
+		 * synchronously log the superblock instead to ensure the
+		 * superblock is immediately unpinned and can be written back.
+		 */
+		xfs_sync_sb(mp, true);
+	} else
 		xfs_log_force(mp, 0);
 
 	/* start pushing all the metadata that is currently dirty */
@@ -1395,6 +1407,8 @@ xlog_alloc_log(
 	ASSERT(xfs_buf_islocked(bp));
 	xfs_buf_unlock(bp);
 
+	/* use high priority wq for log I/O completion */
+	bp->b_ioend_wq = mp->m_log_workqueue;
 	bp->b_iodone = xlog_iodone;
 	log->l_xbuf = bp;
 
@@ -1427,6 +1441,8 @@ xlog_alloc_log(
 		ASSERT(xfs_buf_islocked(bp));
 		xfs_buf_unlock(bp);
 
+		/* use high priority wq for log I/O completion */
+		bp->b_ioend_wq = mp->m_log_workqueue;
 		bp->b_iodone = xlog_iodone;
 		iclog->ic_bp = bp;
 		iclog->ic_data = bp->b_addr;
@@ -1806,8 +1822,6 @@ xlog_sync(
 	XFS_BUF_ZEROFLAGS(bp);
 	XFS_BUF_ASYNC(bp);
 	bp->b_flags |= XBF_SYNCIO;
-	/* use high priority completion wq */
-	bp->b_ioend_wq = log->l_mp->m_log_workqueue;
 
 	if (log->l_mp->m_flags & XFS_MOUNT_BARRIER) {
 		bp->b_flags |= XBF_FUA;
@@ -1856,8 +1870,6 @@ xlog_sync(
 		bp->b_flags |= XBF_SYNCIO;
 		if (log->l_mp->m_flags & XFS_MOUNT_BARRIER)
 			bp->b_flags |= XBF_FUA;
-		/* use high priority completion wq */
-		bp->b_ioend_wq = log->l_mp->m_log_workqueue;
 
 		ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1);
 		ASSERT(XFS_BUF_ADDR(bp) + BTOBB(count) <= log->l_logBBsize);
@@ -2027,7 +2039,7 @@ xlog_print_tic_res(
 		"  total reg   = %u bytes (o/flow = %u bytes)\n"
 		"  ophdrs      = %u (ophdr space = %u bytes)\n"
 		"  ophdr + reg = %u bytes\n"
-		"  num regions = %u\n",
+		"  num regions = %u",
 		((ticket->t_trans_type <= 0 ||
 		  ticket->t_trans_type > XFS_TRANS_TYPE_MAX) ?
 		  "bad-trans-type" : trans_type_str[ticket->t_trans_type-1]),

fs/xfs/xfs_mount.c

@@ -408,11 +408,11 @@ xfs_update_alignment(xfs_mount_t *mp)
 		if (xfs_sb_version_hasdalign(sbp)) {
 			if (sbp->sb_unit != mp->m_dalign) {
 				sbp->sb_unit = mp->m_dalign;
-				mp->m_update_flags |= XFS_SB_UNIT;
+				mp->m_update_sb = true;
 			}
 			if (sbp->sb_width != mp->m_swidth) {
 				sbp->sb_width = mp->m_swidth;
-				mp->m_update_flags |= XFS_SB_WIDTH;
+				mp->m_update_sb = true;
 			}
 		} else {
 			xfs_warn(mp,
@@ -583,38 +583,19 @@ int
 xfs_mount_reset_sbqflags(
 	struct xfs_mount	*mp)
 {
-	int			error;
-	struct xfs_trans	*tp;
-
 	mp->m_qflags = 0;
 
-	/*
-	 * It is OK to look at sb_qflags here in mount path,
-	 * without m_sb_lock.
-	 */
+	/* It is OK to look at sb_qflags in the mount path without m_sb_lock. */
 	if (mp->m_sb.sb_qflags == 0)
 		return 0;
 	spin_lock(&mp->m_sb_lock);
 	mp->m_sb.sb_qflags = 0;
 	spin_unlock(&mp->m_sb_lock);
 
-	/*
-	 * If the fs is readonly, let the incore superblock run
-	 * with quotas off but don't flush the update out to disk
-	 */
-	if (mp->m_flags & XFS_MOUNT_RDONLY)
+	if (!xfs_fs_writable(mp, SB_FREEZE_WRITE))
 		return 0;
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SBCHANGE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_sbchange, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp, 0);
-		xfs_alert(mp, "%s: Superblock update failed!", __func__);
-		return error;
-	}
-
-	xfs_mod_sb(tp, XFS_SB_QFLAGS);
-	return xfs_trans_commit(tp, 0);
+	return xfs_sync_sb(mp, false);
 }
 
 __uint64_t
@@ -659,26 +640,25 @@ xfs_mountfs(
 	xfs_sb_mount_common(mp, sbp);
 
 	/*
-	 * Check for a mismatched features2 values.  Older kernels
-	 * read & wrote into the wrong sb offset for sb_features2
-	 * on some platforms due to xfs_sb_t not being 64bit size aligned
-	 * when sb_features2 was added, which made older superblock
-	 * reading/writing routines swap it as a 64-bit value.
+	 * Check for a mismatched features2 values.  Older kernels read & wrote
+	 * into the wrong sb offset for sb_features2 on some platforms due to
+	 * xfs_sb_t not being 64bit size aligned when sb_features2 was added,
+	 * which made older superblock reading/writing routines swap it as a
+	 * 64-bit value.
 	 *
 	 * For backwards compatibility, we make both slots equal.
 	 *
-	 * If we detect a mismatched field, we OR the set bits into the
-	 * existing features2 field in case it has already been modified; we
-	 * don't want to lose any features.  We then update the bad location
-	 * with the ORed value so that older kernels will see any features2
-	 * flags, and mark the two fields as needing updates once the
-	 * transaction subsystem is online.
+	 * If we detect a mismatched field, we OR the set bits into the existing
+	 * features2 field in case it has already been modified; we don't want
+	 * to lose any features.  We then update the bad location with the ORed
+	 * value so that older kernels will see any features2 flags. The
+	 * superblock writeback code ensures the new sb_features2 is copied to
+	 * sb_bad_features2 before it is logged or written to disk.
 	 */
 	if (xfs_sb_has_mismatched_features2(sbp)) {
 		xfs_warn(mp, "correcting sb_features alignment problem");
 		sbp->sb_features2 |= sbp->sb_bad_features2;
-		sbp->sb_bad_features2 = sbp->sb_features2;
-		mp->m_update_flags |= XFS_SB_FEATURES2 | XFS_SB_BAD_FEATURES2;
+		mp->m_update_sb = true;
 
 		/*
 		 * Re-check for ATTR2 in case it was found in bad_features2
@@ -692,17 +672,17 @@ xfs_mountfs(
 	if (xfs_sb_version_hasattr2(&mp->m_sb) &&
 	   (mp->m_flags & XFS_MOUNT_NOATTR2)) {
 		xfs_sb_version_removeattr2(&mp->m_sb);
-		mp->m_update_flags |= XFS_SB_FEATURES2;
+		mp->m_update_sb = true;
 
 		/* update sb_versionnum for the clearing of the morebits */
 		if (!sbp->sb_features2)
-			mp->m_update_flags |= XFS_SB_VERSIONNUM;
+			mp->m_update_sb = true;
 	}
 
 	/* always use v2 inodes by default now */
 	if (!(mp->m_sb.sb_versionnum & XFS_SB_VERSION_NLINKBIT)) {
 		mp->m_sb.sb_versionnum |= XFS_SB_VERSION_NLINKBIT;
-		mp->m_update_flags |= XFS_SB_VERSIONNUM;
+		mp->m_update_sb = true;
 	}
 
 	/*
@@ -895,8 +875,8 @@ xfs_mountfs(
 	 * the next remount into writeable mode.  Otherwise we would never
 	 * perform the update e.g. for the root filesystem.
 	 */
-	if (mp->m_update_flags && !(mp->m_flags & XFS_MOUNT_RDONLY)) {
-		error = xfs_mount_log_sb(mp, mp->m_update_flags);
+	if (mp->m_update_sb && !(mp->m_flags & XFS_MOUNT_RDONLY)) {
+		error = xfs_sync_sb(mp, false);
 		if (error) {
 			xfs_warn(mp, "failed to write sb changes");
 			goto out_rtunmount;
@@ -1103,9 +1083,6 @@ xfs_fs_writable(
 int
 xfs_log_sbcount(xfs_mount_t *mp)
 {
-	xfs_trans_t	*tp;
-	int		error;
-
 	/* allow this to proceed during the freeze sequence... */
 	if (!xfs_fs_writable(mp, SB_FREEZE_COMPLETE))
 		return 0;
@@ -1119,17 +1096,7 @@ xfs_log_sbcount(xfs_mount_t *mp)
 	if (!xfs_sb_version_haslazysbcount(&mp->m_sb))
 		return 0;
 
-	tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_COUNT, KM_SLEEP);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_sb, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp, 0);
-		return error;
-	}
-
-	xfs_mod_sb(tp, XFS_SB_IFREE | XFS_SB_ICOUNT | XFS_SB_FDBLOCKS);
-	xfs_trans_set_sync(tp);
-	error = xfs_trans_commit(tp, 0);
-	return error;
+	return xfs_sync_sb(mp, true);
 }
 
 /*
@@ -1422,34 +1389,6 @@ xfs_freesb(
 	xfs_buf_relse(bp);
 }
 
-/*
- * Used to log changes to the superblock unit and width fields which could
- * be altered by the mount options, as well as any potential sb_features2
- * fixup. Only the first superblock is updated.
- */
-int
-xfs_mount_log_sb(
-	xfs_mount_t	*mp,
-	__int64_t	fields)
-{
-	xfs_trans_t	*tp;
-	int		error;
-
-	ASSERT(fields & (XFS_SB_UNIT | XFS_SB_WIDTH | XFS_SB_UUID |
-			 XFS_SB_FEATURES2 | XFS_SB_BAD_FEATURES2 |
-			 XFS_SB_VERSIONNUM));
-
-	tp = xfs_trans_alloc(mp, XFS_TRANS_SB_UNIT);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_sb, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp, 0);
-		return error;
-	}
-	xfs_mod_sb(tp, fields);
-	error = xfs_trans_commit(tp, 0);
-	return error;
-}
-
 /*
  * If the underlying (data/log/rt) device is readonly, there are some
  * operations that cannot proceed.

fs/xfs/xfs_mount.h

@@ -162,8 +162,7 @@ typedef struct xfs_mount {
 	struct delayed_work	m_reclaim_work;	/* background inode reclaim */
 	struct delayed_work	m_eofblocks_work; /* background eof blocks
 						     trimming */
-	__int64_t		m_update_flags;	/* sb flags we need to update
-						   on the next remount,rw */
+	bool			m_update_sb;	/* sb needs update in mount */
 	int64_t			m_low_space[XFS_LOWSP_MAX];
 						/* low free space thresholds */
 	struct xfs_kobj		m_kobj;
@@ -378,7 +377,7 @@ extern void	xfs_unmountfs(xfs_mount_t *);
 extern int	xfs_mod_incore_sb(xfs_mount_t *, xfs_sb_field_t, int64_t, int);
 extern int	xfs_mod_incore_sb_batch(xfs_mount_t *, xfs_mod_sb_t *,
 			uint, int);
-extern int	xfs_mount_log_sb(xfs_mount_t *, __int64_t);
+extern int	xfs_mount_log_sb(xfs_mount_t *);
 extern struct xfs_buf *xfs_getsb(xfs_mount_t *, int);
 extern int	xfs_readsb(xfs_mount_t *, int);
 extern void	xfs_freesb(xfs_mount_t *);

fs/xfs/xfs_qm.c

@@ -714,7 +714,6 @@ STATIC int
 xfs_qm_qino_alloc(
 	xfs_mount_t	*mp,
 	xfs_inode_t	**ip,
-	__int64_t	sbfields,
 	uint		flags)
 {
 	xfs_trans_t	*tp;
@@ -777,11 +776,6 @@ xfs_qm_qino_alloc(
 	spin_lock(&mp->m_sb_lock);
 	if (flags & XFS_QMOPT_SBVERSION) {
 		ASSERT(!xfs_sb_version_hasquota(&mp->m_sb));
-		ASSERT((sbfields & (XFS_SB_VERSIONNUM | XFS_SB_UQUOTINO |
-			XFS_SB_GQUOTINO | XFS_SB_PQUOTINO | XFS_SB_QFLAGS)) ==
-				(XFS_SB_VERSIONNUM | XFS_SB_UQUOTINO |
-				 XFS_SB_GQUOTINO | XFS_SB_PQUOTINO |
-				 XFS_SB_QFLAGS));
 
 		xfs_sb_version_addquota(&mp->m_sb);
 		mp->m_sb.sb_uquotino = NULLFSINO;
@@ -798,7 +792,7 @@ xfs_qm_qino_alloc(
 	else
 		mp->m_sb.sb_pquotino = (*ip)->i_ino;
 	spin_unlock(&mp->m_sb_lock);
-	xfs_mod_sb(tp, sbfields);
+	xfs_log_sb(tp);
 
 	if ((error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES))) {
 		xfs_alert(mp, "%s failed (error %d)!", __func__, error);
@@ -1451,7 +1445,7 @@ xfs_qm_mount_quotas(
 	spin_unlock(&mp->m_sb_lock);
 
 	if (sbf != (mp->m_qflags & XFS_MOUNT_QUOTA_ALL)) {
-		if (xfs_qm_write_sb_changes(mp, XFS_SB_QFLAGS)) {
+		if (xfs_sync_sb(mp, false)) {
 			/*
 			 * We could only have been turning quotas off.
 			 * We aren't in very good shape actually because
@@ -1482,7 +1476,6 @@ xfs_qm_init_quotainos(
 	struct xfs_inode	*gip = NULL;
 	struct xfs_inode	*pip = NULL;
 	int			error;
-	__int64_t		sbflags = 0;
 	uint			flags = 0;
 
 	ASSERT(mp->m_quotainfo);
@@ -1517,9 +1510,6 @@ xfs_qm_init_quotainos(
 		}
 	} else {
 		flags |= XFS_QMOPT_SBVERSION;
-		sbflags |= (XFS_SB_VERSIONNUM | XFS_SB_UQUOTINO |
-			    XFS_SB_GQUOTINO | XFS_SB_PQUOTINO |
-			    XFS_SB_QFLAGS);
 	}
 
 	/*
@@ -1530,7 +1520,6 @@ xfs_qm_init_quotainos(
 	 */
 	if (XFS_IS_UQUOTA_ON(mp) && uip == NULL) {
 		error = xfs_qm_qino_alloc(mp, &uip,
-					      sbflags | XFS_SB_UQUOTINO,
 					      flags | XFS_QMOPT_UQUOTA);
 		if (error)
 			goto error_rele;
@@ -1539,7 +1528,6 @@ xfs_qm_init_quotainos(
 	}
 	if (XFS_IS_GQUOTA_ON(mp) && gip == NULL) {
 		error = xfs_qm_qino_alloc(mp, &gip,
-					  sbflags | XFS_SB_GQUOTINO,
 					  flags | XFS_QMOPT_GQUOTA);
 		if (error)
 			goto error_rele;
@@ -1548,7 +1536,6 @@ xfs_qm_init_quotainos(
 	}
 	if (XFS_IS_PQUOTA_ON(mp) && pip == NULL) {
 		error = xfs_qm_qino_alloc(mp, &pip,
-					  sbflags | XFS_SB_PQUOTINO,
 					  flags | XFS_QMOPT_PQUOTA);
 		if (error)
 			goto error_rele;
@@ -1587,32 +1574,6 @@ xfs_qm_dqfree_one(
 	xfs_qm_dqdestroy(dqp);
 }
 
-/*
- * Start a transaction and write the incore superblock changes to
- * disk. flags parameter indicates which fields have changed.
- */
-int
-xfs_qm_write_sb_changes(
-	xfs_mount_t	*mp,
-	__int64_t	flags)
-{
-	xfs_trans_t	*tp;
-	int		error;
-
-	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SBCHANGE);
-	error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_sbchange, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp, 0);
-		return error;
-	}
-
-	xfs_mod_sb(tp, flags);
-	error = xfs_trans_commit(tp, 0);
-
-	return error;
-}
-
-
 /* --------------- utility functions for vnodeops ---------------- */
 
 

fs/xfs/xfs_qm.h

@@ -157,7 +157,6 @@ struct xfs_dquot_acct {
 #define XFS_QM_RTBWARNLIMIT	5
 
 extern void		xfs_qm_destroy_quotainfo(struct xfs_mount *);
-extern int		xfs_qm_write_sb_changes(struct xfs_mount *, __int64_t);
 
 /* dquot stuff */
 extern void		xfs_qm_dqpurge_all(struct xfs_mount *, uint);

fs/xfs/xfs_qm_syscalls.c

@@ -91,8 +91,7 @@ xfs_qm_scall_quotaoff(
 		mutex_unlock(&q->qi_quotaofflock);
 
 		/* XXX what to do if error ? Revert back to old vals incore ? */
-		error = xfs_qm_write_sb_changes(mp, XFS_SB_QFLAGS);
-		return error;
+		return xfs_sync_sb(mp, false);
 	}
 
 	dqtype = 0;
@@ -313,7 +312,6 @@ xfs_qm_scall_quotaon(
 {
 	int		error;
 	uint		qf;
-	__int64_t	sbflags;
 
 	flags &= (XFS_ALL_QUOTA_ACCT | XFS_ALL_QUOTA_ENFD);
 	/*
@@ -321,8 +319,6 @@ xfs_qm_scall_quotaon(
 	 */
 	flags &= ~(XFS_ALL_QUOTA_ACCT);
 
-	sbflags = 0;
-
 	if (flags == 0) {
 		xfs_debug(mp, "%s: zero flags, m_qflags=%x",
 			__func__, mp->m_qflags);
@@ -363,11 +359,11 @@ xfs_qm_scall_quotaon(
 	/*
 	 * There's nothing to change if it's the same.
 	 */
-	if ((qf & flags) == flags && sbflags == 0)
+	if ((qf & flags) == flags)
 		return -EEXIST;
-	sbflags |= XFS_SB_QFLAGS;
 
-	if ((error = xfs_qm_write_sb_changes(mp, sbflags)))
+	error = xfs_sync_sb(mp, false);
+	if (error)
 		return error;
 	/*
 	 * If we aren't trying to switch on quota enforcement, we are done.
@@ -774,7 +770,7 @@ xfs_qm_log_quotaoff(
 	mp->m_sb.sb_qflags = (mp->m_qflags & ~(flags)) & XFS_MOUNT_QUOTA_ALL;
 	spin_unlock(&mp->m_sb_lock);
 
-	xfs_mod_sb(tp, XFS_SB_QFLAGS);
+	xfs_log_sb(tp);
 
 	/*
 	 * We have to make sure that the transaction is secure on disk before we

fs/xfs/xfs_super.c

@@ -685,7 +685,7 @@ xfs_blkdev_get(
 				    mp);
 	if (IS_ERR(*bdevp)) {
 		error = PTR_ERR(*bdevp);
-		xfs_warn(mp, "Invalid device [%s], error=%d\n", name, error);
+		xfs_warn(mp, "Invalid device [%s], error=%d", name, error);
 	}
 
 	return error;
@@ -1111,6 +1111,11 @@ xfs_fs_statfs(
 					statp->f_files,
 					mp->m_maxicount);
 
+	/* If sb_icount overshot maxicount, report actual allocation */
+	statp->f_files = max_t(typeof(statp->f_files),
+					statp->f_files,
+					sbp->sb_icount);
+
 	/* make sure statp->f_ffree does not underflow */
 	ffree = statp->f_files - (sbp->sb_icount - sbp->sb_ifree);
 	statp->f_ffree = max_t(__int64_t, ffree, 0);
@@ -1257,13 +1262,13 @@ xfs_fs_remount(
 		 * If this is the first remount to writeable state we
 		 * might have some superblock changes to update.
 		 */
-		if (mp->m_update_flags) {
-			error = xfs_mount_log_sb(mp, mp->m_update_flags);
+		if (mp->m_update_sb) {
+			error = xfs_sync_sb(mp, false);
 			if (error) {
 				xfs_warn(mp, "failed to write sb changes");
 				return error;
 			}
-			mp->m_update_flags = 0;
+			mp->m_update_sb = false;
 		}
 
 		/*
@@ -1293,8 +1298,9 @@ xfs_fs_remount(
 
 /*
  * Second stage of a freeze. The data is already frozen so we only
- * need to take care of the metadata. Once that's done write a dummy
- * record to dirty the log in case of a crash while frozen.
+ * need to take care of the metadata. Once that's done sync the superblock
+ * to the log to dirty it in case of a crash while frozen. This ensures that we
+ * will recover the unlinked inode lists on the next mount.
  */
 STATIC int
 xfs_fs_freeze(
@@ -1304,7 +1310,7 @@ xfs_fs_freeze(
 
 	xfs_save_resvblks(mp);
 	xfs_quiesce_attr(mp);
-	return xfs_fs_log_dummy(mp);
+	return xfs_sync_sb(mp, true);
 }
 
 STATIC int

fs/xfs/xfs_sysctl.c

@@ -148,24 +148,6 @@ static struct ctl_table xfs_table[] = {
 		.extra1		= &xfs_params.inherit_noatim.min,
 		.extra2		= &xfs_params.inherit_noatim.max
 	},
-	{
-		.procname	= "xfsbufd_centisecs",
-		.data		= &xfs_params.xfs_buf_timer.val,
-		.maxlen		= sizeof(int),
-		.mode		= 0644,
-		.proc_handler	= proc_dointvec_minmax,
-		.extra1		= &xfs_params.xfs_buf_timer.min,
-		.extra2		= &xfs_params.xfs_buf_timer.max
-	},
-	{
-		.procname	= "age_buffer_centisecs",
-		.data		= &xfs_params.xfs_buf_age.val,
-		.maxlen		= sizeof(int),
-		.mode		= 0644,
-		.proc_handler	= proc_dointvec_minmax,
-		.extra1		= &xfs_params.xfs_buf_age.min,
-		.extra2		= &xfs_params.xfs_buf_age.max
-	},
 	{
 		.procname	= "inherit_nosymlinks",
 		.data		= &xfs_params.inherit_nosym.val,

fs/xfs/xfs_trans.c

@@ -472,6 +472,7 @@ xfs_trans_apply_sb_deltas(
 		whole = 1;
 	}
 
+	xfs_trans_buf_set_type(tp, bp, XFS_BLFT_SB_BUF);
 	if (whole)
 		/*
 		 * Log the whole thing, the fields are noncontiguous.

fs/xfs/xfs_trans_buf.c

@@ -327,9 +327,10 @@ xfs_trans_read_buf_map(
 		return -EIO;
 	}
 
-	if (tp)
+	if (tp) {
 		_xfs_trans_bjoin(tp, bp, 1);
-	trace_xfs_trans_read_buf(bp->b_fspriv);
+		trace_xfs_trans_read_buf(bp->b_fspriv);
+	}
 	*bpp = bp;
 	return 0;