Merge tag 'xfs-for-linus-3.15-rc1' of git://oss.sgi.com/xfs/xfs

Pull xfs update from Dave Chinner:
 "There are a couple of new fallocate features in this request - it was
  decided that it was easiest to push them through the XFS tree using
  topic branches and have the ext4 support be based on those branches.
  Hence you may see some overlap with the ext4 tree merge depending on
  how they including those topic branches into their tree.  Other than
  that, there is O_TMPFILE support, some cleanups and bug fixes.

  The main changes in the XFS tree for 3.15-rc1 are:

   - O_TMPFILE support
   - allowing AIO+DIO writes beyond EOF
   - FALLOC_FL_COLLAPSE_RANGE support for fallocate syscall and XFS
     implementation
   - FALLOC_FL_ZERO_RANGE support for fallocate syscall and XFS
     implementation
   - IO verifier cleanup and rework
   - stack usage reduction changes
   - vm_map_ram NOIO context fixes to remove lockdep warings
   - various bug fixes and cleanups"

* tag 'xfs-for-linus-3.15-rc1' of git://oss.sgi.com/xfs/xfs: (34 commits)
  xfs: fix directory hash ordering bug
  xfs: extra semi-colon breaks a condition
  xfs: Add support for FALLOC_FL_ZERO_RANGE
  fs: Introduce FALLOC_FL_ZERO_RANGE flag for fallocate
  xfs: inode log reservations are still too small
  xfs: xfs_check_page_type buffer checks need help
  xfs: avoid AGI/AGF deadlock scenario for inode chunk allocation
  xfs: use NOIO contexts for vm_map_ram
  xfs: don't leak EFSBADCRC to userspace
  xfs: fix directory inode iolock lockdep false positive
  xfs: allocate xfs_da_args to reduce stack footprint
  xfs: always do log forces via the workqueue
  xfs: modify verifiers to differentiate CRC from other errors
  xfs: print useful caller information in xfs_error_report
  xfs: add xfs_verifier_error()
  xfs: add helper for updating checksums on xfs_bufs
  xfs: add helper for verifying checksums on xfs_bufs
  xfs: Use defines for CRC offsets in all cases
  xfs: skip pointless CRC updates after verifier failures
  xfs: Add support FALLOC_FL_COLLAPSE_RANGE for fallocate
  ...

fs/direct-io.c

@@ -1193,13 +1193,19 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 	}
 
 	/*
-	 * For file extending writes updating i_size before data
-	 * writeouts complete can expose uninitialized blocks. So
-	 * even for AIO, we need to wait for i/o to complete before
-	 * returning in this case.
+	 * For file extending writes updating i_size before data writeouts
+	 * complete can expose uninitialized blocks in dumb filesystems.
+	 * In that case we need to wait for I/O completion even if asked
+	 * for an asynchronous write.
 	 */
-	dio->is_async = !is_sync_kiocb(iocb) && !((rw & WRITE) &&
-		(end > i_size_read(inode)));
+	if (is_sync_kiocb(iocb))
+		dio->is_async = false;
+	else if (!(dio->flags & DIO_ASYNC_EXTEND) &&
+            (rw & WRITE) && end > i_size_read(inode))
+		dio->is_async = false;
+	else
+		dio->is_async = true;
+
 	dio->inode = inode;
 	dio->rw = rw;
 

fs/open.c

@@ -231,7 +231,13 @@ int do_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
 		return -EINVAL;
 
 	/* Return error if mode is not supported */
-	if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
+	if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE |
+		     FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE))
+		return -EOPNOTSUPP;
+
+	/* Punch hole and zero range are mutually exclusive */
+	if ((mode & (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_ZERO_RANGE)) ==
+	    (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_ZERO_RANGE))
 		return -EOPNOTSUPP;
 
 	/* Punch hole must have keep size set */
@@ -239,11 +245,20 @@ int do_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
 	    !(mode & FALLOC_FL_KEEP_SIZE))
 		return -EOPNOTSUPP;
 
+	/* Collapse range should only be used exclusively. */
+	if ((mode & FALLOC_FL_COLLAPSE_RANGE) &&
+	    (mode & ~FALLOC_FL_COLLAPSE_RANGE))
+		return -EINVAL;
+
 	if (!(file->f_mode & FMODE_WRITE))
 		return -EBADF;
 
-	/* It's not possible punch hole on append only file */
-	if (mode & FALLOC_FL_PUNCH_HOLE && IS_APPEND(inode))
+	/*
+	 * It's not possible to punch hole or perform collapse range
+	 * on append only file
+	 */
+	if (mode & (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_COLLAPSE_RANGE)
+	    && IS_APPEND(inode))
 		return -EPERM;
 
 	if (IS_IMMUTABLE(inode))
@@ -271,6 +286,14 @@ int do_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
 	if (((offset + len) > inode->i_sb->s_maxbytes) || ((offset + len) < 0))
 		return -EFBIG;
 
+	/*
+	 * There is no need to overlap collapse range with EOF, in which case
+	 * it is effectively a truncate operation
+	 */
+	if ((mode & FALLOC_FL_COLLAPSE_RANGE) &&
+	    (offset + len >= i_size_read(inode)))
+		return -EINVAL;
+
 	if (!file->f_op->fallocate)
 		return -EOPNOTSUPP;
 

fs/xfs/kmem.c

@@ -65,12 +65,31 @@ kmem_alloc(size_t size, xfs_km_flags_t flags)
 void *
 kmem_zalloc_large(size_t size, xfs_km_flags_t flags)
 {
+	unsigned noio_flag = 0;
 	void	*ptr;
+	gfp_t	lflags;
 
 	ptr = kmem_zalloc(size, flags | KM_MAYFAIL);
 	if (ptr)
 		return ptr;
-	return vzalloc(size);
+
+	/*
+	 * __vmalloc() will allocate data pages and auxillary structures (e.g.
+	 * pagetables) with GFP_KERNEL, yet we may be under GFP_NOFS context
+	 * here. Hence we need to tell memory reclaim that we are in such a
+	 * context via PF_MEMALLOC_NOIO to prevent memory reclaim re-entering
+	 * the filesystem here and potentially deadlocking.
+	 */
+	if ((current->flags & PF_FSTRANS) || (flags & KM_NOFS))
+		noio_flag = memalloc_noio_save();
+
+	lflags = kmem_flags_convert(flags);
+	ptr = __vmalloc(size, lflags | __GFP_HIGHMEM | __GFP_ZERO, PAGE_KERNEL);
+
+	if ((current->flags & PF_FSTRANS) || (flags & KM_NOFS))
+		memalloc_noio_restore(noio_flag);
+
+	return ptr;
 }
 
 void

fs/xfs/xfs_acl.c

@@ -281,7 +281,7 @@ xfs_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 	if (!acl)
 		goto set_acl;
 
-	error = -EINVAL;
+	error = -E2BIG;
 	if (acl->a_count > XFS_ACL_MAX_ENTRIES(XFS_M(inode->i_sb)))
 		return error;
 

fs/xfs/xfs_ag.h

@@ -89,6 +89,8 @@ typedef struct xfs_agf {
 	/* structure must be padded to 64 bit alignment */
 } xfs_agf_t;
 
+#define XFS_AGF_CRC_OFF		offsetof(struct xfs_agf, agf_crc)
+
 #define	XFS_AGF_MAGICNUM	0x00000001
 #define	XFS_AGF_VERSIONNUM	0x00000002
 #define	XFS_AGF_SEQNO		0x00000004
@@ -167,6 +169,8 @@ typedef struct xfs_agi {
 	/* structure must be padded to 64 bit alignment */
 } xfs_agi_t;
 
+#define XFS_AGI_CRC_OFF		offsetof(struct xfs_agi, agi_crc)
+
 #define	XFS_AGI_MAGICNUM	0x00000001
 #define	XFS_AGI_VERSIONNUM	0x00000002
 #define	XFS_AGI_SEQNO		0x00000004
@@ -222,6 +226,8 @@ typedef struct xfs_agfl {
 	__be32		agfl_bno[];	/* actually XFS_AGFL_SIZE(mp) */
 } xfs_agfl_t;
 
+#define XFS_AGFL_CRC_OFF	offsetof(struct xfs_agfl, agfl_crc)
+
 /*
  * tags for inode radix tree
  */

fs/xfs/xfs_alloc.c

@@ -474,7 +474,6 @@ xfs_agfl_read_verify(
 	struct xfs_buf	*bp)
 {
 	struct xfs_mount *mp = bp->b_target->bt_mount;
-	int		agfl_ok = 1;
 
 	/*
 	 * There is no verification of non-crc AGFLs because mkfs does not
@@ -485,15 +484,13 @@ xfs_agfl_read_verify(
 	if (!xfs_sb_version_hascrc(&mp->m_sb))
 		return;
 
-	agfl_ok = xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-				   offsetof(struct xfs_agfl, agfl_crc));
-
-	agfl_ok = agfl_ok && xfs_agfl_verify(bp);
-
-	if (!agfl_ok) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (!xfs_buf_verify_cksum(bp, XFS_AGFL_CRC_OFF))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_agfl_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 static void
@@ -508,16 +505,15 @@ xfs_agfl_write_verify(
 		return;
 
 	if (!xfs_agfl_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
 	if (bip)
 		XFS_BUF_TO_AGFL(bp)->agfl_lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
-			 offsetof(struct xfs_agfl, agfl_crc));
+	xfs_buf_update_cksum(bp, XFS_AGFL_CRC_OFF);
 }
 
 const struct xfs_buf_ops xfs_agfl_buf_ops = {
@@ -2238,19 +2234,17 @@ xfs_agf_read_verify(
 	struct xfs_buf	*bp)
 {
 	struct xfs_mount *mp = bp->b_target->bt_mount;
-	int		agf_ok = 1;
-
-	if (xfs_sb_version_hascrc(&mp->m_sb))
-		agf_ok = xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					  offsetof(struct xfs_agf, agf_crc));
 
-	agf_ok = agf_ok && xfs_agf_verify(mp, bp);
-
-	if (unlikely(XFS_TEST_ERROR(!agf_ok, mp, XFS_ERRTAG_ALLOC_READ_AGF,
-			XFS_RANDOM_ALLOC_READ_AGF))) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (xfs_sb_version_hascrc(&mp->m_sb) &&
+	    !xfs_buf_verify_cksum(bp, XFS_AGF_CRC_OFF))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (XFS_TEST_ERROR(!xfs_agf_verify(mp, bp), mp,
+				XFS_ERRTAG_ALLOC_READ_AGF,
+				XFS_RANDOM_ALLOC_READ_AGF))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 static void
@@ -2261,8 +2255,8 @@ xfs_agf_write_verify(
 	struct xfs_buf_log_item	*bip = bp->b_fspriv;
 
 	if (!xfs_agf_verify(mp, bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -2272,8 +2266,7 @@ xfs_agf_write_verify(
 	if (bip)
 		XFS_BUF_TO_AGF(bp)->agf_lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
-			 offsetof(struct xfs_agf, agf_crc));
+	xfs_buf_update_cksum(bp, XFS_AGF_CRC_OFF);
 }
 
 const struct xfs_buf_ops xfs_agf_buf_ops = {

fs/xfs/xfs_alloc_btree.c

@@ -355,12 +355,14 @@ static void
 xfs_allocbt_read_verify(
 	struct xfs_buf	*bp)
 {
-	if (!(xfs_btree_sblock_verify_crc(bp) &&
-	      xfs_allocbt_verify(bp))) {
-		trace_xfs_btree_corrupt(bp, _RET_IP_);
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
-				     bp->b_target->bt_mount, bp->b_addr);
+	if (!xfs_btree_sblock_verify_crc(bp))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_allocbt_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+
+	if (bp->b_error) {
+		trace_xfs_btree_corrupt(bp, _RET_IP_);
+		xfs_verifier_error(bp);
 	}
 }
 
@@ -370,9 +372,9 @@ xfs_allocbt_write_verify(
 {
 	if (!xfs_allocbt_verify(bp)) {
 		trace_xfs_btree_corrupt(bp, _RET_IP_);
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
-				     bp->b_target->bt_mount, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
+		return;
 	}
 	xfs_btree_sblock_calc_crc(bp);
 

fs/xfs/xfs_aops.c

@@ -632,38 +632,46 @@ xfs_map_at_offset(
 }
 
 /*
- * Test if a given page is suitable for writing as part of an unwritten
- * or delayed allocate extent.
+ * Test if a given page contains at least one buffer of a given @type.
+ * If @check_all_buffers is true, then we walk all the buffers in the page to
+ * try to find one of the type passed in. If it is not set, then the caller only
+ * needs to check the first buffer on the page for a match.
  */
-STATIC int
+STATIC bool
 xfs_check_page_type(
 	struct page		*page,
-	unsigned int		type)
+	unsigned int		type,
+	bool			check_all_buffers)
 {
-	if (PageWriteback(page))
-		return 0;
+	struct buffer_head	*bh;
+	struct buffer_head	*head;
 
-	if (page->mapping && page_has_buffers(page)) {
-		struct buffer_head	*bh, *head;
-		int			acceptable = 0;
+	if (PageWriteback(page))
+		return false;
+	if (!page->mapping)
+		return false;
+	if (!page_has_buffers(page))
+		return false;
 
-		bh = head = page_buffers(page);
-		do {
-			if (buffer_unwritten(bh))
-				acceptable += (type == XFS_IO_UNWRITTEN);
-			else if (buffer_delay(bh))
-				acceptable += (type == XFS_IO_DELALLOC);
-			else if (buffer_dirty(bh) && buffer_mapped(bh))
-				acceptable += (type == XFS_IO_OVERWRITE);
-			else
-				break;
-		} while ((bh = bh->b_this_page) != head);
+	bh = head = page_buffers(page);
+	do {
+		if (buffer_unwritten(bh)) {
+			if (type == XFS_IO_UNWRITTEN)
+				return true;
+		} else if (buffer_delay(bh)) {
+			if (type == XFS_IO_DELALLOC)
+				return true;
+		} else if (buffer_dirty(bh) && buffer_mapped(bh)) {
+			if (type == XFS_IO_OVERWRITE)
+				return true;
+		}
 
-		if (acceptable)
-			return 1;
-	}
+		/* If we are only checking the first buffer, we are done now. */
+		if (!check_all_buffers)
+			break;
+	} while ((bh = bh->b_this_page) != head);
 
-	return 0;
+	return false;
 }
 
 /*
@@ -697,7 +705,7 @@ xfs_convert_page(
 		goto fail_unlock_page;
 	if (page->mapping != inode->i_mapping)
 		goto fail_unlock_page;
-	if (!xfs_check_page_type(page, (*ioendp)->io_type))
+	if (!xfs_check_page_type(page, (*ioendp)->io_type, false))
 		goto fail_unlock_page;
 
 	/*
@@ -742,6 +750,15 @@ xfs_convert_page(
 	p_offset = p_offset ? roundup(p_offset, len) : PAGE_CACHE_SIZE;
 	page_dirty = p_offset / len;
 
+	/*
+	 * The moment we find a buffer that doesn't match our current type
+	 * specification or can't be written, abort the loop and start
+	 * writeback. As per the above xfs_imap_valid() check, only
+	 * xfs_vm_writepage() can handle partial page writeback fully - we are
+	 * limited here to the buffers that are contiguous with the current
+	 * ioend, and hence a buffer we can't write breaks that contiguity and
+	 * we have to defer the rest of the IO to xfs_vm_writepage().
+	 */
 	bh = head = page_buffers(page);
 	do {
 		if (offset >= end_offset)
@@ -750,7 +767,7 @@ xfs_convert_page(
 			uptodate = 0;
 		if (!(PageUptodate(page) || buffer_uptodate(bh))) {
 			done = 1;
-			continue;
+			break;
 		}
 
 		if (buffer_unwritten(bh) || buffer_delay(bh) ||
@@ -762,10 +779,11 @@ xfs_convert_page(
 			else
 				type = XFS_IO_OVERWRITE;
 
-			if (!xfs_imap_valid(inode, imap, offset)) {
-				done = 1;
-				continue;
-			}
+			/*
+			 * imap should always be valid because of the above
+			 * partial page end_offset check on the imap.
+			 */
+			ASSERT(xfs_imap_valid(inode, imap, offset));
 
 			lock_buffer(bh);
 			if (type != XFS_IO_OVERWRITE)
@@ -777,6 +795,7 @@ xfs_convert_page(
 			count++;
 		} else {
 			done = 1;
+			break;
 		}
 	} while (offset += len, (bh = bh->b_this_page) != head);
 
@@ -868,7 +887,7 @@ xfs_aops_discard_page(
 	struct buffer_head	*bh, *head;
 	loff_t			offset = page_offset(page);
 
-	if (!xfs_check_page_type(page, XFS_IO_DELALLOC))
+	if (!xfs_check_page_type(page, XFS_IO_DELALLOC, true))
 		goto out_invalidate;
 
 	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
@@ -1441,7 +1460,8 @@ xfs_vm_direct_IO(
 		ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov,
 					    offset, nr_segs,
 					    xfs_get_blocks_direct,
-					    xfs_end_io_direct_write, NULL, 0);
+					    xfs_end_io_direct_write, NULL,
+					    DIO_ASYNC_EXTEND);
 		if (ret != -EIOCBQUEUED && iocb->private)
 			goto out_destroy_ioend;
 	} else {

fs/xfs/xfs_attr_leaf.c

@@ -213,8 +213,8 @@ xfs_attr3_leaf_write_verify(
 	struct xfs_attr3_leaf_hdr *hdr3 = bp->b_addr;
 
 	if (!xfs_attr3_leaf_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -224,7 +224,7 @@ xfs_attr3_leaf_write_verify(
 	if (bip)
 		hdr3->info.lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_ATTR3_LEAF_CRC_OFF);
+	xfs_buf_update_cksum(bp, XFS_ATTR3_LEAF_CRC_OFF);
 }
 
 /*
@@ -239,13 +239,14 @@ xfs_attr3_leaf_read_verify(
 {
 	struct xfs_mount	*mp = bp->b_target->bt_mount;
 
-	if ((xfs_sb_version_hascrc(&mp->m_sb) &&
-	     !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					  XFS_ATTR3_LEAF_CRC_OFF)) ||
-	    !xfs_attr3_leaf_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (xfs_sb_version_hascrc(&mp->m_sb) &&
+	     !xfs_buf_verify_cksum(bp, XFS_ATTR3_LEAF_CRC_OFF))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_attr3_leaf_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 const struct xfs_buf_ops xfs_attr3_leaf_buf_ops = {

fs/xfs/xfs_attr_remote.c

@@ -125,7 +125,6 @@ xfs_attr3_rmt_read_verify(
 	struct xfs_mount *mp = bp->b_target->bt_mount;
 	char		*ptr;
 	int		len;
-	bool		corrupt = false;
 	xfs_daddr_t	bno;
 
 	/* no verification of non-crc buffers */
@@ -140,11 +139,11 @@ xfs_attr3_rmt_read_verify(
 	while (len > 0) {
 		if (!xfs_verify_cksum(ptr, XFS_LBSIZE(mp),
 				      XFS_ATTR3_RMT_CRC_OFF)) {
-			corrupt = true;
+			xfs_buf_ioerror(bp, EFSBADCRC);
 			break;
 		}
 		if (!xfs_attr3_rmt_verify(mp, ptr, XFS_LBSIZE(mp), bno)) {
-			corrupt = true;
+			xfs_buf_ioerror(bp, EFSCORRUPTED);
 			break;
 		}
 		len -= XFS_LBSIZE(mp);
@@ -152,10 +151,9 @@ xfs_attr3_rmt_read_verify(
 		bno += mp->m_bsize;
 	}
 
-	if (corrupt) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
-		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	} else
+	if (bp->b_error)
+		xfs_verifier_error(bp);
+	else
 		ASSERT(len == 0);
 }
 
@@ -180,9 +178,8 @@ xfs_attr3_rmt_write_verify(
 
 	while (len > 0) {
 		if (!xfs_attr3_rmt_verify(mp, ptr, XFS_LBSIZE(mp), bno)) {
-			XFS_CORRUPTION_ERROR(__func__,
-					    XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 			xfs_buf_ioerror(bp, EFSCORRUPTED);
+			xfs_verifier_error(bp);
 			return;
 		}
 		if (bip) {

fs/xfs/xfs_bmap.c

@@ -5378,3 +5378,196 @@ error0:
 	}
 	return error;
 }
+
+/*
+ * Shift extent records to the left to cover a hole.
+ *
+ * The maximum number of extents to be shifted in a single operation
+ * is @num_exts, and @current_ext keeps track of the current extent
+ * index we have shifted. @offset_shift_fsb is the length by which each
+ * extent is shifted. If there is no hole to shift the extents
+ * into, this will be considered invalid operation and we abort immediately.
+ */
+int
+xfs_bmap_shift_extents(
+	struct xfs_trans	*tp,
+	struct xfs_inode	*ip,
+	int			*done,
+	xfs_fileoff_t		start_fsb,
+	xfs_fileoff_t		offset_shift_fsb,
+	xfs_extnum_t		*current_ext,
+	xfs_fsblock_t		*firstblock,
+	struct xfs_bmap_free	*flist,
+	int			num_exts)
+{
+	struct xfs_btree_cur		*cur;
+	struct xfs_bmbt_rec_host	*gotp;
+	struct xfs_bmbt_irec            got;
+	struct xfs_bmbt_irec		left;
+	struct xfs_mount		*mp = ip->i_mount;
+	struct xfs_ifork		*ifp;
+	xfs_extnum_t			nexts = 0;
+	xfs_fileoff_t			startoff;
+	int				error = 0;
+	int				i;
+	int				whichfork = XFS_DATA_FORK;
+	int				logflags;
+	xfs_filblks_t			blockcount = 0;
+
+	if (unlikely(XFS_TEST_ERROR(
+	    (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
+	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
+	     mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
+		XFS_ERROR_REPORT("xfs_bmap_shift_extents",
+				 XFS_ERRLEVEL_LOW, mp);
+		return XFS_ERROR(EFSCORRUPTED);
+	}
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return XFS_ERROR(EIO);
+
+	ASSERT(current_ext != NULL);
+
+	ifp = XFS_IFORK_PTR(ip, whichfork);
+
+	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+		/* Read in all the extents */
+		error = xfs_iread_extents(tp, ip, whichfork);
+		if (error)
+			return error;
+	}
+
+	/*
+	 * If *current_ext is 0, we would need to lookup the extent
+	 * from where we would start shifting and store it in gotp.
+	 */
+	if (!*current_ext) {
+		gotp = xfs_iext_bno_to_ext(ifp, start_fsb, current_ext);
+		/*
+		 * gotp can be null in 2 cases: 1) if there are no extents
+		 * or 2) start_fsb lies in a hole beyond which there are
+		 * no extents. Either way, we are done.
+		 */
+		if (!gotp) {
+			*done = 1;
+			return 0;
+		}
+	}
+
+	/* We are going to change core inode */
+	logflags = XFS_ILOG_CORE;
+
+	if (ifp->if_flags & XFS_IFBROOT) {
+		cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
+		cur->bc_private.b.firstblock = *firstblock;
+		cur->bc_private.b.flist = flist;
+		cur->bc_private.b.flags = 0;
+	} else {
+		cur = NULL;
+		logflags |= XFS_ILOG_DEXT;
+	}
+
+	while (nexts++ < num_exts &&
+	       *current_ext <  XFS_IFORK_NEXTENTS(ip, whichfork)) {
+
+		gotp = xfs_iext_get_ext(ifp, *current_ext);
+		xfs_bmbt_get_all(gotp, &got);
+		startoff = got.br_startoff - offset_shift_fsb;
+
+		/*
+		 * Before shifting extent into hole, make sure that the hole
+		 * is large enough to accomodate the shift.
+		 */
+		if (*current_ext) {
+			xfs_bmbt_get_all(xfs_iext_get_ext(ifp,
+						*current_ext - 1), &left);
+
+			if (startoff < left.br_startoff + left.br_blockcount)
+				error = XFS_ERROR(EINVAL);
+		} else if (offset_shift_fsb > got.br_startoff) {
+			/*
+			 * When first extent is shifted, offset_shift_fsb
+			 * should be less than the stating offset of
+			 * the first extent.
+			 */
+			error = XFS_ERROR(EINVAL);
+		}
+
+		if (error)
+			goto del_cursor;
+
+		if (cur) {
+			error = xfs_bmbt_lookup_eq(cur, got.br_startoff,
+						   got.br_startblock,
+						   got.br_blockcount,
+						   &i);
+			if (error)
+				goto del_cursor;
+			XFS_WANT_CORRUPTED_GOTO(i == 1, del_cursor);
+		}
+
+		/* Check if we can merge 2 adjacent extents */
+		if (*current_ext &&
+		    left.br_startoff + left.br_blockcount == startoff &&
+		    left.br_startblock + left.br_blockcount ==
+				got.br_startblock &&
+		    left.br_state == got.br_state &&
+		    left.br_blockcount + got.br_blockcount <= MAXEXTLEN) {
+			blockcount = left.br_blockcount +
+				got.br_blockcount;
+			xfs_iext_remove(ip, *current_ext, 1, 0);
+			if (cur) {
+				error = xfs_btree_delete(cur, &i);
+				if (error)
+					goto del_cursor;
+				XFS_WANT_CORRUPTED_GOTO(i == 1, del_cursor);
+			}
+			XFS_IFORK_NEXT_SET(ip, whichfork,
+				XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
+			gotp = xfs_iext_get_ext(ifp, --*current_ext);
+			xfs_bmbt_get_all(gotp, &got);
+
+			/* Make cursor point to the extent we will update */
+			if (cur) {
+				error = xfs_bmbt_lookup_eq(cur, got.br_startoff,
+							   got.br_startblock,
+							   got.br_blockcount,
+							   &i);
+				if (error)
+					goto del_cursor;
+				XFS_WANT_CORRUPTED_GOTO(i == 1, del_cursor);
+			}
+
+			xfs_bmbt_set_blockcount(gotp, blockcount);
+			got.br_blockcount = blockcount;
+		} else {
+			/* We have to update the startoff */
+			xfs_bmbt_set_startoff(gotp, startoff);
+			got.br_startoff = startoff;
+		}
+
+		if (cur) {
+			error = xfs_bmbt_update(cur, got.br_startoff,
+						got.br_startblock,
+						got.br_blockcount,
+						got.br_state);
+			if (error)
+				goto del_cursor;
+		}
+
+		(*current_ext)++;
+	}
+
+	/* Check if we are done */
+	if (*current_ext ==  XFS_IFORK_NEXTENTS(ip, whichfork))
+		*done = 1;
+
+del_cursor:
+	if (cur)
+		xfs_btree_del_cursor(cur,
+			error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
+
+	xfs_trans_log_inode(tp, ip, logflags);
+
+	return error;
+}

fs/xfs/xfs_bmap.h

@@ -127,6 +127,16 @@ static inline void xfs_bmap_init(xfs_bmap_free_t *flp, xfs_fsblock_t *fbp)
 	{ BMAP_RIGHT_FILLING,	"RF" }, \
 	{ BMAP_ATTRFORK,	"ATTR" }
 
+
+/*
+ * This macro is used to determine how many extents will be shifted
+ * in one write transaction. We could require two splits,
+ * an extent move on the first and an extent merge on the second,
+ * So it is proper that one extent is shifted inside write transaction
+ * at a time.
+ */
+#define XFS_BMAP_MAX_SHIFT_EXTENTS	1
+
 #ifdef DEBUG
 void	xfs_bmap_trace_exlist(struct xfs_inode *ip, xfs_extnum_t cnt,
 		int whichfork, unsigned long caller_ip);
@@ -169,5 +179,10 @@ int	xfs_bunmapi(struct xfs_trans *tp, struct xfs_inode *ip,
 int	xfs_check_nostate_extents(struct xfs_ifork *ifp, xfs_extnum_t idx,
 		xfs_extnum_t num);
 uint	xfs_default_attroffset(struct xfs_inode *ip);
+int	xfs_bmap_shift_extents(struct xfs_trans *tp, struct xfs_inode *ip,
+		int *done, xfs_fileoff_t start_fsb,
+		xfs_fileoff_t offset_shift_fsb, xfs_extnum_t *current_ext,
+		xfs_fsblock_t *firstblock, struct xfs_bmap_free	*flist,
+		int num_exts);
 
 #endif	/* __XFS_BMAP_H__ */

fs/xfs/xfs_bmap_btree.c

@@ -780,12 +780,14 @@ static void
 xfs_bmbt_read_verify(
 	struct xfs_buf	*bp)
 {
-	if (!(xfs_btree_lblock_verify_crc(bp) &&
-	      xfs_bmbt_verify(bp))) {
-		trace_xfs_btree_corrupt(bp, _RET_IP_);
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
-				     bp->b_target->bt_mount, bp->b_addr);
+	if (!xfs_btree_lblock_verify_crc(bp))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_bmbt_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+
+	if (bp->b_error) {
+		trace_xfs_btree_corrupt(bp, _RET_IP_);
+		xfs_verifier_error(bp);
 	}
 }
 
@@ -794,11 +796,9 @@ xfs_bmbt_write_verify(
 	struct xfs_buf	*bp)
 {
 	if (!xfs_bmbt_verify(bp)) {
-		xfs_warn(bp->b_target->bt_mount, "bmbt daddr 0x%llx failed", bp->b_bn);
 		trace_xfs_btree_corrupt(bp, _RET_IP_);
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
-				     bp->b_target->bt_mount, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 	xfs_btree_lblock_calc_crc(bp);

fs/xfs/xfs_bmap_util.c

@@ -1349,7 +1349,6 @@ xfs_free_file_space(
 		 * the freeing of the space succeeds at ENOSPC.
 		 */
 		tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
-		tp->t_flags |= XFS_TRANS_RESERVE;
 		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write, resblks, 0);
 
 		/*
@@ -1467,6 +1466,102 @@ out:
 
 }
 
+/*
+ * xfs_collapse_file_space()
+ *	This routine frees disk space and shift extent for the given file.
+ *	The first thing we do is to free data blocks in the specified range
+ *	by calling xfs_free_file_space(). It would also sync dirty data
+ *	and invalidate page cache over the region on which collapse range
+ *	is working. And Shift extent records to the left to cover a hole.
+ * RETURNS:
+ *	0 on success
+ *	errno on error
+ *
+ */
+int
+xfs_collapse_file_space(
+	struct xfs_inode	*ip,
+	xfs_off_t		offset,
+	xfs_off_t		len)
+{
+	int			done = 0;
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_trans	*tp;
+	int			error;
+	xfs_extnum_t		current_ext = 0;
+	struct xfs_bmap_free	free_list;
+	xfs_fsblock_t		first_block;
+	int			committed;
+	xfs_fileoff_t		start_fsb;
+	xfs_fileoff_t		shift_fsb;
+
+	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
+
+	trace_xfs_collapse_file_space(ip);
+
+	start_fsb = XFS_B_TO_FSB(mp, offset + len);
+	shift_fsb = XFS_B_TO_FSB(mp, len);
+
+	error = xfs_free_file_space(ip, offset, len);
+	if (error)
+		return error;
+
+	while (!error && !done) {
+		tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
+		tp->t_flags |= XFS_TRANS_RESERVE;
+		/*
+		 * 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) {
+			ASSERT(error == ENOSPC || XFS_FORCED_SHUTDOWN(mp));
+			xfs_trans_cancel(tp, 0);
+			break;
+		}
+
+		xfs_ilock(ip, XFS_ILOCK_EXCL);
+		error = xfs_trans_reserve_quota(tp, mp, ip->i_udquot,
+				ip->i_gdquot, ip->i_pdquot,
+				XFS_DIOSTRAT_SPACE_RES(mp, 0), 0,
+				XFS_QMOPT_RES_REGBLKS);
+		if (error)
+			goto out;
+
+		xfs_trans_ijoin(tp, ip, 0);
+
+		xfs_bmap_init(&free_list, &first_block);
+
+		/*
+		 * We are using the write transaction in which max 2 bmbt
+		 * updates are allowed
+		 */
+		error = xfs_bmap_shift_extents(tp, ip, &done, start_fsb,
+					       shift_fsb, &current_ext,
+					       &first_block, &free_list,
+					       XFS_BMAP_MAX_SHIFT_EXTENTS);
+		if (error)
+			goto out;
+
+		error = xfs_bmap_finish(&tp, &free_list, &committed);
+		if (error)
+			goto out;
+
+		error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	}
+
+	return error;
+
+out:
+	xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT);
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	return error;
+}
+
 /*
  * We need to check that the format of the data fork in the temporary inode is
  * valid for the target inode before doing the swap. This is not a problem with

fs/xfs/xfs_bmap_util.h

@@ -99,6 +99,8 @@ int	xfs_free_file_space(struct xfs_inode *ip, xfs_off_t offset,
 			    xfs_off_t len);
 int	xfs_zero_file_space(struct xfs_inode *ip, xfs_off_t offset,
 			    xfs_off_t len);
+int	xfs_collapse_file_space(struct xfs_inode *, xfs_off_t offset,
+				xfs_off_t len);
 
 /* EOF block manipulation functions */
 bool	xfs_can_free_eofblocks(struct xfs_inode *ip, bool force);

fs/xfs/xfs_btree.c

@@ -234,8 +234,7 @@ xfs_btree_lblock_calc_crc(
 		return;
 	if (bip)
 		block->bb_u.l.bb_lsn = cpu_to_be64(bip->bli_item.li_lsn);
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
-			 XFS_BTREE_LBLOCK_CRC_OFF);
+	xfs_buf_update_cksum(bp, XFS_BTREE_LBLOCK_CRC_OFF);
 }
 
 bool
@@ -243,8 +242,8 @@ xfs_btree_lblock_verify_crc(
 	struct xfs_buf		*bp)
 {
 	if (xfs_sb_version_hascrc(&bp->b_target->bt_mount->m_sb))
-		return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					XFS_BTREE_LBLOCK_CRC_OFF);
+		return xfs_buf_verify_cksum(bp, XFS_BTREE_LBLOCK_CRC_OFF);
+
 	return true;
 }
 
@@ -267,8 +266,7 @@ xfs_btree_sblock_calc_crc(
 		return;
 	if (bip)
 		block->bb_u.s.bb_lsn = cpu_to_be64(bip->bli_item.li_lsn);
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
-			 XFS_BTREE_SBLOCK_CRC_OFF);
+	xfs_buf_update_cksum(bp, XFS_BTREE_SBLOCK_CRC_OFF);
 }
 
 bool
@@ -276,8 +274,8 @@ xfs_btree_sblock_verify_crc(
 	struct xfs_buf		*bp)
 {
 	if (xfs_sb_version_hascrc(&bp->b_target->bt_mount->m_sb))
-		return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					XFS_BTREE_SBLOCK_CRC_OFF);
+		return xfs_buf_verify_cksum(bp, XFS_BTREE_SBLOCK_CRC_OFF);
+
 	return true;
 }
 

fs/xfs/xfs_buf.c

@@ -396,7 +396,17 @@ _xfs_buf_map_pages(
 		bp->b_addr = NULL;
 	} else {
 		int retried = 0;
+		unsigned noio_flag;
 
+		/*
+		 * vm_map_ram() will allocate auxillary structures (e.g.
+		 * pagetables) with GFP_KERNEL, yet we are likely to be under
+		 * GFP_NOFS context here. Hence we need to tell memory reclaim
+		 * that we are in such a context via PF_MEMALLOC_NOIO to prevent
+		 * memory reclaim re-entering the filesystem here and
+		 * potentially deadlocking.
+		 */
+		noio_flag = memalloc_noio_save();
 		do {
 			bp->b_addr = vm_map_ram(bp->b_pages, bp->b_page_count,
 						-1, PAGE_KERNEL);
@@ -404,6 +414,7 @@ _xfs_buf_map_pages(
 				break;
 			vm_unmap_aliases();
 		} while (retried++ <= 1);
+		memalloc_noio_restore(noio_flag);
 
 		if (!bp->b_addr)
 			return -ENOMEM;

fs/xfs/xfs_buf.h

@@ -369,6 +369,20 @@ static inline void xfs_buf_relse(xfs_buf_t *bp)
 	xfs_buf_rele(bp);
 }
 
+static inline int
+xfs_buf_verify_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
+{
+	return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+				cksum_offset);
+}
+
+static inline void
+xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
+{
+	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
+			 cksum_offset);
+}
+
 /*
  *	Handling of buftargs.
  */

fs/xfs/xfs_buf_item.c

@@ -796,20 +796,6 @@ xfs_buf_item_init(
 		bip->bli_formats[i].blf_map_size = map_size;
 	}
 
-#ifdef XFS_TRANS_DEBUG
-	/*
-	 * Allocate the arrays for tracking what needs to be logged
-	 * and what our callers request to be logged.  bli_orig
-	 * holds a copy of the original, clean buffer for comparison
-	 * against, and bli_logged keeps a 1 bit flag per byte in
-	 * the buffer to indicate which bytes the callers have asked
-	 * to have logged.
-	 */
-	bip->bli_orig = kmem_alloc(BBTOB(bp->b_length), KM_SLEEP);
-	memcpy(bip->bli_orig, bp->b_addr, BBTOB(bp->b_length));
-	bip->bli_logged = kmem_zalloc(BBTOB(bp->b_length) / NBBY, KM_SLEEP);
-#endif
-
 	/*
 	 * Put the buf item into the list of items attached to the
 	 * buffer at the front.
@@ -957,11 +943,6 @@ STATIC void
 xfs_buf_item_free(
 	xfs_buf_log_item_t	*bip)
 {
-#ifdef XFS_TRANS_DEBUG
-	kmem_free(bip->bli_orig);
-	kmem_free(bip->bli_logged);
-#endif /* XFS_TRANS_DEBUG */
-
 	xfs_buf_item_free_format(bip);
 	kmem_zone_free(xfs_buf_item_zone, bip);
 }

fs/xfs/xfs_da_btree.c

@@ -185,8 +185,8 @@ xfs_da3_node_write_verify(
 	struct xfs_da3_node_hdr *hdr3 = bp->b_addr;
 
 	if (!xfs_da3_node_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -196,7 +196,7 @@ xfs_da3_node_write_verify(
 	if (bip)
 		hdr3->info.lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DA3_NODE_CRC_OFF);
+	xfs_buf_update_cksum(bp, XFS_DA3_NODE_CRC_OFF);
 }
 
 /*
@@ -209,18 +209,20 @@ static void
 xfs_da3_node_read_verify(
 	struct xfs_buf		*bp)
 {
-	struct xfs_mount	*mp = bp->b_target->bt_mount;
 	struct xfs_da_blkinfo	*info = bp->b_addr;
 
 	switch (be16_to_cpu(info->magic)) {
 		case XFS_DA3_NODE_MAGIC:
-			if (!xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					      XFS_DA3_NODE_CRC_OFF))
+			if (!xfs_buf_verify_cksum(bp, XFS_DA3_NODE_CRC_OFF)) {
+				xfs_buf_ioerror(bp, EFSBADCRC);
 				break;
+			}
 			/* fall through */
 		case XFS_DA_NODE_MAGIC:
-			if (!xfs_da3_node_verify(bp))
+			if (!xfs_da3_node_verify(bp)) {
+				xfs_buf_ioerror(bp, EFSCORRUPTED);
 				break;
+			}
 			return;
 		case XFS_ATTR_LEAF_MAGIC:
 		case XFS_ATTR3_LEAF_MAGIC:
@@ -237,8 +239,7 @@ xfs_da3_node_read_verify(
 	}
 
 	/* corrupt block */
-	XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
-	xfs_buf_ioerror(bp, EFSCORRUPTED);
+	xfs_verifier_error(bp);
 }
 
 const struct xfs_buf_ops xfs_da3_node_buf_ops = {
@@ -1295,7 +1296,7 @@ xfs_da3_fixhashpath(
 		node = blk->bp->b_addr;
 		dp->d_ops->node_hdr_from_disk(&nodehdr, node);
 		btree = dp->d_ops->node_tree_p(node);
-		if (be32_to_cpu(btree->hashval) == lasthash)
+		if (be32_to_cpu(btree[blk->index].hashval) == lasthash)
 			break;
 		blk->hashval = lasthash;
 		btree[blk->index].hashval = cpu_to_be32(lasthash);

fs/xfs/xfs_dinode.h

@@ -89,6 +89,8 @@ typedef struct xfs_dinode {
 	/* structure must be padded to 64 bit alignment */
 } xfs_dinode_t;
 
+#define XFS_DINODE_CRC_OFF	offsetof(struct xfs_dinode, di_crc)
+
 #define DI_MAX_FLUSH 0xffff
 
 /*

fs/xfs/xfs_dir2.c

@@ -180,16 +180,23 @@ xfs_dir_init(
 	xfs_inode_t	*dp,
 	xfs_inode_t	*pdp)
 {
-	xfs_da_args_t	args;
+	struct xfs_da_args *args;
 	int		error;
 
-	memset((char *)&args, 0, sizeof(args));
-	args.dp = dp;
-	args.trans = tp;
 	ASSERT(S_ISDIR(dp->i_d.di_mode));
-	if ((error = xfs_dir_ino_validate(tp->t_mountp, pdp->i_ino)))
+	error = xfs_dir_ino_validate(tp->t_mountp, pdp->i_ino);
+	if (error)
 		return error;
-	return xfs_dir2_sf_create(&args, pdp->i_ino);
+
+	args = kmem_zalloc(sizeof(*args), KM_SLEEP | KM_NOFS);
+	if (!args)
+		return ENOMEM;
+
+	args->dp = dp;
+	args->trans = tp;
+	error = xfs_dir2_sf_create(args, pdp->i_ino);
+	kmem_free(args);
+	return error;
 }
 
 /*
@@ -205,41 +212,56 @@ xfs_dir_createname(
 	xfs_bmap_free_t		*flist,		/* bmap's freeblock list */
 	xfs_extlen_t		total)		/* bmap's total block count */
 {
-	xfs_da_args_t		args;
+	struct xfs_da_args	*args;
 	int			rval;
 	int			v;		/* type-checking value */
 
 	ASSERT(S_ISDIR(dp->i_d.di_mode));
-	if ((rval = xfs_dir_ino_validate(tp->t_mountp, inum)))
+	rval = xfs_dir_ino_validate(tp->t_mountp, inum);
+	if (rval)
 		return rval;
 	XFS_STATS_INC(xs_dir_create);
 
-	memset(&args, 0, sizeof(xfs_da_args_t));
-	args.name = name->name;
-	args.namelen = name->len;
-	args.filetype = name->type;
-	args.hashval = dp->i_mount->m_dirnameops->hashname(name);
-	args.inumber = inum;
-	args.dp = dp;
-	args.firstblock = first;
-	args.flist = flist;
-	args.total = total;
-	args.whichfork = XFS_DATA_FORK;
-	args.trans = tp;
-	args.op_flags = XFS_DA_OP_ADDNAME | XFS_DA_OP_OKNOENT;
-
-	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL)
-		rval = xfs_dir2_sf_addname(&args);
-	else if ((rval = xfs_dir2_isblock(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_block_addname(&args);
-	else if ((rval = xfs_dir2_isleaf(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_leaf_addname(&args);
+	args = kmem_zalloc(sizeof(*args), KM_SLEEP | KM_NOFS);
+	if (!args)
+		return ENOMEM;
+
+	args->name = name->name;
+	args->namelen = name->len;
+	args->filetype = name->type;
+	args->hashval = dp->i_mount->m_dirnameops->hashname(name);
+	args->inumber = inum;
+	args->dp = dp;
+	args->firstblock = first;
+	args->flist = flist;
+	args->total = total;
+	args->whichfork = XFS_DATA_FORK;
+	args->trans = tp;
+	args->op_flags = XFS_DA_OP_ADDNAME | XFS_DA_OP_OKNOENT;
+
+	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL) {
+		rval = xfs_dir2_sf_addname(args);
+		goto out_free;
+	}
+
+	rval = xfs_dir2_isblock(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v) {
+		rval = xfs_dir2_block_addname(args);
+		goto out_free;
+	}
+
+	rval = xfs_dir2_isleaf(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v)
+		rval = xfs_dir2_leaf_addname(args);
 	else
-		rval = xfs_dir2_node_addname(&args);
+		rval = xfs_dir2_node_addname(args);
+
+out_free:
+	kmem_free(args);
 	return rval;
 }
 
@@ -282,46 +304,66 @@ xfs_dir_lookup(
 	xfs_ino_t	*inum,		/* out: inode number */
 	struct xfs_name *ci_name)	/* out: actual name if CI match */
 {
-	xfs_da_args_t	args;
+	struct xfs_da_args *args;
 	int		rval;
 	int		v;		/* type-checking value */
 
 	ASSERT(S_ISDIR(dp->i_d.di_mode));
 	XFS_STATS_INC(xs_dir_lookup);
 
-	memset(&args, 0, sizeof(xfs_da_args_t));
-	args.name = name->name;
-	args.namelen = name->len;
-	args.filetype = name->type;
-	args.hashval = dp->i_mount->m_dirnameops->hashname(name);
-	args.dp = dp;
-	args.whichfork = XFS_DATA_FORK;
-	args.trans = tp;
-	args.op_flags = XFS_DA_OP_OKNOENT;
+	/*
+	 * We need to use KM_NOFS here so that lockdep will not throw false
+	 * positive deadlock warnings on a non-transactional lookup path. It is
+	 * safe to recurse into inode recalim in that case, but lockdep can't
+	 * easily be taught about it. Hence KM_NOFS avoids having to add more
+	 * lockdep Doing this avoids having to add a bunch of lockdep class
+	 * annotations into the reclaim path for the ilock.
+	 */
+	args = kmem_zalloc(sizeof(*args), KM_SLEEP | KM_NOFS);
+	args->name = name->name;
+	args->namelen = name->len;
+	args->filetype = name->type;
+	args->hashval = dp->i_mount->m_dirnameops->hashname(name);
+	args->dp = dp;
+	args->whichfork = XFS_DATA_FORK;
+	args->trans = tp;
+	args->op_flags = XFS_DA_OP_OKNOENT;
 	if (ci_name)
-		args.op_flags |= XFS_DA_OP_CILOOKUP;
+		args->op_flags |= XFS_DA_OP_CILOOKUP;
 
-	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL)
-		rval = xfs_dir2_sf_lookup(&args);
-	else if ((rval = xfs_dir2_isblock(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_block_lookup(&args);
-	else if ((rval = xfs_dir2_isleaf(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_leaf_lookup(&args);
+	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL) {
+		rval = xfs_dir2_sf_lookup(args);
+		goto out_check_rval;
+	}
+
+	rval = xfs_dir2_isblock(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v) {
+		rval = xfs_dir2_block_lookup(args);
+		goto out_check_rval;
+	}
+
+	rval = xfs_dir2_isleaf(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v)
+		rval = xfs_dir2_leaf_lookup(args);
 	else
-		rval = xfs_dir2_node_lookup(&args);
+		rval = xfs_dir2_node_lookup(args);
+
+out_check_rval:
 	if (rval == EEXIST)
 		rval = 0;
 	if (!rval) {
-		*inum = args.inumber;
+		*inum = args->inumber;
 		if (ci_name) {
-			ci_name->name = args.value;
-			ci_name->len = args.valuelen;
+			ci_name->name = args->value;
+			ci_name->len = args->valuelen;
 		}
 	}
+out_free:
+	kmem_free(args);
 	return rval;
 }
 
@@ -338,38 +380,51 @@ xfs_dir_removename(
 	xfs_bmap_free_t	*flist,		/* bmap's freeblock list */
 	xfs_extlen_t	total)		/* bmap's total block count */
 {
-	xfs_da_args_t	args;
+	struct xfs_da_args *args;
 	int		rval;
 	int		v;		/* type-checking value */
 
 	ASSERT(S_ISDIR(dp->i_d.di_mode));
 	XFS_STATS_INC(xs_dir_remove);
 
-	memset(&args, 0, sizeof(xfs_da_args_t));
-	args.name = name->name;
-	args.namelen = name->len;
-	args.filetype = name->type;
-	args.hashval = dp->i_mount->m_dirnameops->hashname(name);
-	args.inumber = ino;
-	args.dp = dp;
-	args.firstblock = first;
-	args.flist = flist;
-	args.total = total;
-	args.whichfork = XFS_DATA_FORK;
-	args.trans = tp;
-
-	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL)
-		rval = xfs_dir2_sf_removename(&args);
-	else if ((rval = xfs_dir2_isblock(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_block_removename(&args);
-	else if ((rval = xfs_dir2_isleaf(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_leaf_removename(&args);
+	args = kmem_zalloc(sizeof(*args), KM_SLEEP | KM_NOFS);
+	if (!args)
+		return ENOMEM;
+
+	args->name = name->name;
+	args->namelen = name->len;
+	args->filetype = name->type;
+	args->hashval = dp->i_mount->m_dirnameops->hashname(name);
+	args->inumber = ino;
+	args->dp = dp;
+	args->firstblock = first;
+	args->flist = flist;
+	args->total = total;
+	args->whichfork = XFS_DATA_FORK;
+	args->trans = tp;
+
+	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL) {
+		rval = xfs_dir2_sf_removename(args);
+		goto out_free;
+	}
+
+	rval = xfs_dir2_isblock(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v) {
+		rval = xfs_dir2_block_removename(args);
+		goto out_free;
+	}
+
+	rval = xfs_dir2_isleaf(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v)
+		rval = xfs_dir2_leaf_removename(args);
 	else
-		rval = xfs_dir2_node_removename(&args);
+		rval = xfs_dir2_node_removename(args);
+out_free:
+	kmem_free(args);
 	return rval;
 }
 
@@ -386,40 +441,54 @@ xfs_dir_replace(
 	xfs_bmap_free_t	*flist,		/* bmap's freeblock list */
 	xfs_extlen_t	total)		/* bmap's total block count */
 {
-	xfs_da_args_t	args;
+	struct xfs_da_args *args;
 	int		rval;
 	int		v;		/* type-checking value */
 
 	ASSERT(S_ISDIR(dp->i_d.di_mode));
 
-	if ((rval = xfs_dir_ino_validate(tp->t_mountp, inum)))
+	rval = xfs_dir_ino_validate(tp->t_mountp, inum);
+	if (rval)
 		return rval;
 
-	memset(&args, 0, sizeof(xfs_da_args_t));
-	args.name = name->name;
-	args.namelen = name->len;
-	args.filetype = name->type;
-	args.hashval = dp->i_mount->m_dirnameops->hashname(name);
-	args.inumber = inum;
-	args.dp = dp;
-	args.firstblock = first;
-	args.flist = flist;
-	args.total = total;
-	args.whichfork = XFS_DATA_FORK;
-	args.trans = tp;
-
-	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL)
-		rval = xfs_dir2_sf_replace(&args);
-	else if ((rval = xfs_dir2_isblock(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_block_replace(&args);
-	else if ((rval = xfs_dir2_isleaf(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_leaf_replace(&args);
+	args = kmem_zalloc(sizeof(*args), KM_SLEEP | KM_NOFS);
+	if (!args)
+		return ENOMEM;
+
+	args->name = name->name;
+	args->namelen = name->len;
+	args->filetype = name->type;
+	args->hashval = dp->i_mount->m_dirnameops->hashname(name);
+	args->inumber = inum;
+	args->dp = dp;
+	args->firstblock = first;
+	args->flist = flist;
+	args->total = total;
+	args->whichfork = XFS_DATA_FORK;
+	args->trans = tp;
+
+	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL) {
+		rval = xfs_dir2_sf_replace(args);
+		goto out_free;
+	}
+
+	rval = xfs_dir2_isblock(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v) {
+		rval = xfs_dir2_block_replace(args);
+		goto out_free;
+	}
+
+	rval = xfs_dir2_isleaf(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v)
+		rval = xfs_dir2_leaf_replace(args);
 	else
-		rval = xfs_dir2_node_replace(&args);
+		rval = xfs_dir2_node_replace(args);
+out_free:
+	kmem_free(args);
 	return rval;
 }
 
@@ -434,7 +503,7 @@ xfs_dir_canenter(
 	struct xfs_name	*name,		/* name of entry to add */
 	uint		resblks)
 {
-	xfs_da_args_t	args;
+	struct xfs_da_args *args;
 	int		rval;
 	int		v;		/* type-checking value */
 
@@ -443,29 +512,42 @@ xfs_dir_canenter(
 
 	ASSERT(S_ISDIR(dp->i_d.di_mode));
 
-	memset(&args, 0, sizeof(xfs_da_args_t));
-	args.name = name->name;
-	args.namelen = name->len;
-	args.filetype = name->type;
-	args.hashval = dp->i_mount->m_dirnameops->hashname(name);
-	args.dp = dp;
-	args.whichfork = XFS_DATA_FORK;
-	args.trans = tp;
-	args.op_flags = XFS_DA_OP_JUSTCHECK | XFS_DA_OP_ADDNAME |
+	args = kmem_zalloc(sizeof(*args), KM_SLEEP | KM_NOFS);
+	if (!args)
+		return ENOMEM;
+
+	args->name = name->name;
+	args->namelen = name->len;
+	args->filetype = name->type;
+	args->hashval = dp->i_mount->m_dirnameops->hashname(name);
+	args->dp = dp;
+	args->whichfork = XFS_DATA_FORK;
+	args->trans = tp;
+	args->op_flags = XFS_DA_OP_JUSTCHECK | XFS_DA_OP_ADDNAME |
 							XFS_DA_OP_OKNOENT;
 
-	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL)
-		rval = xfs_dir2_sf_addname(&args);
-	else if ((rval = xfs_dir2_isblock(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_block_addname(&args);
-	else if ((rval = xfs_dir2_isleaf(tp, dp, &v)))
-		return rval;
-	else if (v)
-		rval = xfs_dir2_leaf_addname(&args);
+	if (dp->i_d.di_format == XFS_DINODE_FMT_LOCAL) {
+		rval = xfs_dir2_sf_addname(args);
+		goto out_free;
+	}
+
+	rval = xfs_dir2_isblock(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v) {
+		rval = xfs_dir2_block_addname(args);
+		goto out_free;
+	}
+
+	rval = xfs_dir2_isleaf(tp, dp, &v);
+	if (rval)
+		goto out_free;
+	if (v)
+		rval = xfs_dir2_leaf_addname(args);
 	else
-		rval = xfs_dir2_node_addname(&args);
+		rval = xfs_dir2_node_addname(args);
+out_free:
+	kmem_free(args);
 	return rval;
 }
 

fs/xfs/xfs_dir2_block.c

@@ -89,13 +89,14 @@ xfs_dir3_block_read_verify(
 {
 	struct xfs_mount	*mp = bp->b_target->bt_mount;
 
-	if ((xfs_sb_version_hascrc(&mp->m_sb) &&
-	     !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					  XFS_DIR3_DATA_CRC_OFF)) ||
-	    !xfs_dir3_block_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (xfs_sb_version_hascrc(&mp->m_sb) &&
+	     !xfs_buf_verify_cksum(bp, XFS_DIR3_DATA_CRC_OFF))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_dir3_block_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 static void
@@ -107,8 +108,8 @@ xfs_dir3_block_write_verify(
 	struct xfs_dir3_blk_hdr	*hdr3 = bp->b_addr;
 
 	if (!xfs_dir3_block_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -118,7 +119,7 @@ xfs_dir3_block_write_verify(
 	if (bip)
 		hdr3->lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_DATA_CRC_OFF);
+	xfs_buf_update_cksum(bp, XFS_DIR3_DATA_CRC_OFF);
 }
 
 const struct xfs_buf_ops xfs_dir3_block_buf_ops = {

fs/xfs/xfs_dir2_data.c

@@ -241,7 +241,6 @@ static void
 xfs_dir3_data_reada_verify(
 	struct xfs_buf		*bp)
 {
-	struct xfs_mount	*mp = bp->b_target->bt_mount;
 	struct xfs_dir2_data_hdr *hdr = bp->b_addr;
 
 	switch (hdr->magic) {
@@ -255,8 +254,8 @@ xfs_dir3_data_reada_verify(
 		xfs_dir3_data_verify(bp);
 		return;
 	default:
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		break;
 	}
 }
@@ -267,13 +266,14 @@ xfs_dir3_data_read_verify(
 {
 	struct xfs_mount	*mp = bp->b_target->bt_mount;
 
-	if ((xfs_sb_version_hascrc(&mp->m_sb) &&
-	     !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					  XFS_DIR3_DATA_CRC_OFF)) ||
-	    !xfs_dir3_data_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (xfs_sb_version_hascrc(&mp->m_sb) &&
+	     !xfs_buf_verify_cksum(bp, XFS_DIR3_DATA_CRC_OFF))
+		 xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_dir3_data_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 static void
@@ -285,8 +285,8 @@ xfs_dir3_data_write_verify(
 	struct xfs_dir3_blk_hdr	*hdr3 = bp->b_addr;
 
 	if (!xfs_dir3_data_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -296,7 +296,7 @@ xfs_dir3_data_write_verify(
 	if (bip)
 		hdr3->lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_DATA_CRC_OFF);
+	xfs_buf_update_cksum(bp, XFS_DIR3_DATA_CRC_OFF);
 }
 
 const struct xfs_buf_ops xfs_dir3_data_buf_ops = {

fs/xfs/xfs_dir2_leaf.c

@@ -179,13 +179,14 @@ __read_verify(
 {
 	struct xfs_mount	*mp = bp->b_target->bt_mount;
 
-	if ((xfs_sb_version_hascrc(&mp->m_sb) &&
-	     !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					  XFS_DIR3_LEAF_CRC_OFF)) ||
-	    !xfs_dir3_leaf_verify(bp, magic)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (xfs_sb_version_hascrc(&mp->m_sb) &&
+	     !xfs_buf_verify_cksum(bp, XFS_DIR3_LEAF_CRC_OFF))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_dir3_leaf_verify(bp, magic))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 static void
@@ -198,8 +199,8 @@ __write_verify(
 	struct xfs_dir3_leaf_hdr *hdr3 = bp->b_addr;
 
 	if (!xfs_dir3_leaf_verify(bp, magic)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -209,7 +210,7 @@ __write_verify(
 	if (bip)
 		hdr3->info.lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_LEAF_CRC_OFF);
+	xfs_buf_update_cksum(bp, XFS_DIR3_LEAF_CRC_OFF);
 }
 
 static void

fs/xfs/xfs_dir2_node.c

@@ -115,13 +115,14 @@ xfs_dir3_free_read_verify(
 {
 	struct xfs_mount	*mp = bp->b_target->bt_mount;
 
-	if ((xfs_sb_version_hascrc(&mp->m_sb) &&
-	     !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					  XFS_DIR3_FREE_CRC_OFF)) ||
-	    !xfs_dir3_free_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (xfs_sb_version_hascrc(&mp->m_sb) &&
+	    !xfs_buf_verify_cksum(bp, XFS_DIR3_FREE_CRC_OFF))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_dir3_free_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 static void
@@ -133,8 +134,8 @@ xfs_dir3_free_write_verify(
 	struct xfs_dir3_blk_hdr	*hdr3 = bp->b_addr;
 
 	if (!xfs_dir3_free_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -144,7 +145,7 @@ xfs_dir3_free_write_verify(
 	if (bip)
 		hdr3->lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_FREE_CRC_OFF);
+	xfs_buf_update_cksum(bp, XFS_DIR3_FREE_CRC_OFF);
 }
 
 const struct xfs_buf_ops xfs_dir3_free_buf_ops = {

fs/xfs/xfs_dquot.c

@@ -615,7 +615,7 @@ xfs_qm_dqread(
 
 	if (flags & XFS_QMOPT_DQALLOC) {
 		tp = xfs_trans_alloc(mp, XFS_TRANS_QM_DQALLOC);
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_attrsetm,
+		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_dqalloc,
 					  XFS_QM_DQALLOC_SPACE_RES(mp), 0);
 		if (error)
 			goto error1;

fs/xfs/xfs_dquot_buf.c

@@ -257,10 +257,13 @@ xfs_dquot_buf_read_verify(
 {
 	struct xfs_mount	*mp = bp->b_target->bt_mount;
 
-	if (!xfs_dquot_buf_verify_crc(mp, bp) || !xfs_dquot_buf_verify(mp, bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (!xfs_dquot_buf_verify_crc(mp, bp))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_dquot_buf_verify(mp, bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 /*
@@ -275,8 +278,8 @@ xfs_dquot_buf_write_verify(
 	struct xfs_mount	*mp = bp->b_target->bt_mount;
 
 	if (!xfs_dquot_buf_verify(mp, bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 }

fs/xfs/xfs_error.c

@@ -156,7 +156,7 @@ xfs_error_report(
 {
 	if (level <= xfs_error_level) {
 		xfs_alert_tag(mp, XFS_PTAG_ERROR_REPORT,
-		"Internal error %s at line %d of file %s.  Caller 0x%p",
+		"Internal error %s at line %d of file %s.  Caller %pF",
 			    tag, linenum, filename, ra);
 
 		xfs_stack_trace();
@@ -178,3 +178,28 @@ xfs_corruption_error(
 	xfs_error_report(tag, level, mp, filename, linenum, ra);
 	xfs_alert(mp, "Corruption detected. Unmount and run xfs_repair");
 }
+
+/*
+ * Warnings specifically for verifier errors.  Differentiate CRC vs. invalid
+ * values, and omit the stack trace unless the error level is tuned high.
+ */
+void
+xfs_verifier_error(
+	struct xfs_buf		*bp)
+{
+	struct xfs_mount *mp = bp->b_target->bt_mount;
+
+	xfs_alert(mp, "Metadata %s detected at %pF, block 0x%llx",
+		  bp->b_error == EFSBADCRC ? "CRC error" : "corruption",
+		  __return_address, bp->b_bn);
+
+	xfs_alert(mp, "Unmount and run xfs_repair");
+
+	if (xfs_error_level >= XFS_ERRLEVEL_LOW) {
+		xfs_alert(mp, "First 64 bytes of corrupted metadata buffer:");
+		xfs_hex_dump(xfs_buf_offset(bp, 0), 64);
+	}
+
+	if (xfs_error_level >= XFS_ERRLEVEL_HIGH)
+		xfs_stack_trace();
+}

fs/xfs/xfs_error.h

@@ -34,6 +34,7 @@ extern void xfs_error_report(const char *tag, int level, struct xfs_mount *mp,
 extern void xfs_corruption_error(const char *tag, int level,
 			struct xfs_mount *mp, void *p, const char *filename,
 			int linenum, inst_t *ra);
+extern void xfs_verifier_error(struct xfs_buf *bp);
 
 #define	XFS_ERROR_REPORT(e, lvl, mp)	\
 	xfs_error_report(e, lvl, mp, __FILE__, __LINE__, __return_address)

fs/xfs/xfs_file.c

@@ -823,7 +823,8 @@ xfs_file_fallocate(
 
 	if (!S_ISREG(inode->i_mode))
 		return -EINVAL;
-	if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
+	if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE |
+		     FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE))
 		return -EOPNOTSUPP;
 
 	xfs_ilock(ip, XFS_IOLOCK_EXCL);
@@ -831,6 +832,20 @@ xfs_file_fallocate(
 		error = xfs_free_file_space(ip, offset, len);
 		if (error)
 			goto out_unlock;
+	} else if (mode & FALLOC_FL_COLLAPSE_RANGE) {
+		unsigned blksize_mask = (1 << inode->i_blkbits) - 1;
+
+		if (offset & blksize_mask || len & blksize_mask) {
+			error = -EINVAL;
+			goto out_unlock;
+		}
+
+		ASSERT(offset + len < i_size_read(inode));
+		new_size = i_size_read(inode) - len;
+
+		error = xfs_collapse_file_space(ip, offset, len);
+		if (error)
+			goto out_unlock;
 	} else {
 		if (!(mode & FALLOC_FL_KEEP_SIZE) &&
 		    offset + len > i_size_read(inode)) {
@@ -840,8 +855,11 @@ xfs_file_fallocate(
 				goto out_unlock;
 		}
 
-		error = xfs_alloc_file_space(ip, offset, len,
-					     XFS_BMAPI_PREALLOC);
+		if (mode & FALLOC_FL_ZERO_RANGE)
+			error = xfs_zero_file_space(ip, offset, len);
+		else
+			error = xfs_alloc_file_space(ip, offset, len,
+						     XFS_BMAPI_PREALLOC);
 		if (error)
 			goto out_unlock;
 	}
@@ -859,7 +877,7 @@ xfs_file_fallocate(
 	if (ip->i_d.di_mode & S_IXGRP)
 		ip->i_d.di_mode &= ~S_ISGID;
 
-	if (!(mode & FALLOC_FL_PUNCH_HOLE))
+	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);

fs/xfs/xfs_format.h

@@ -145,6 +145,8 @@ struct xfs_dsymlink_hdr {
 	__be64	sl_lsn;
 };
 
+#define XFS_SYMLINK_CRC_OFF	offsetof(struct xfs_dsymlink_hdr, sl_crc)
+
 /*
  * The maximum pathlen is 1024 bytes. Since the minimum file system
  * blocksize is 512 bytes, we can get a max of 3 extents back from

fs/xfs/xfs_ialloc.c

@@ -363,6 +363,18 @@ xfs_ialloc_ag_alloc(
 		args.minleft = args.mp->m_in_maxlevels - 1;
 		if ((error = xfs_alloc_vextent(&args)))
 			return error;
+
+		/*
+		 * This request might have dirtied the transaction if the AG can
+		 * satisfy the request, but the exact block was not available.
+		 * If the allocation did fail, subsequent requests will relax
+		 * the exact agbno requirement and increase the alignment
+		 * instead. It is critical that the total size of the request
+		 * (len + alignment + slop) does not increase from this point
+		 * on, so reset minalignslop to ensure it is not included in
+		 * subsequent requests.
+		 */
+		args.minalignslop = 0;
 	} else
 		args.fsbno = NULLFSBLOCK;
 
@@ -1568,18 +1580,17 @@ xfs_agi_read_verify(
 	struct xfs_buf	*bp)
 {
 	struct xfs_mount *mp = bp->b_target->bt_mount;
-	int		agi_ok = 1;
-
-	if (xfs_sb_version_hascrc(&mp->m_sb))
-		agi_ok = xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-					  offsetof(struct xfs_agi, agi_crc));
-	agi_ok = agi_ok && xfs_agi_verify(bp);
 
-	if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IALLOC_READ_AGI,
-			XFS_RANDOM_IALLOC_READ_AGI))) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (xfs_sb_version_hascrc(&mp->m_sb) &&
+	    !xfs_buf_verify_cksum(bp, XFS_AGI_CRC_OFF))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (XFS_TEST_ERROR(!xfs_agi_verify(bp), mp,
+				XFS_ERRTAG_IALLOC_READ_AGI,
+				XFS_RANDOM_IALLOC_READ_AGI))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 static void
@@ -1590,8 +1601,8 @@ xfs_agi_write_verify(
 	struct xfs_buf_log_item	*bip = bp->b_fspriv;
 
 	if (!xfs_agi_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -1600,8 +1611,7 @@ xfs_agi_write_verify(
 
 	if (bip)
 		XFS_BUF_TO_AGI(bp)->agi_lsn = cpu_to_be64(bip->bli_item.li_lsn);
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
-			 offsetof(struct xfs_agi, agi_crc));
+	xfs_buf_update_cksum(bp, XFS_AGI_CRC_OFF);
 }
 
 const struct xfs_buf_ops xfs_agi_buf_ops = {

fs/xfs/xfs_ialloc_btree.c

@@ -243,12 +243,14 @@ static void
 xfs_inobt_read_verify(
 	struct xfs_buf	*bp)
 {
-	if (!(xfs_btree_sblock_verify_crc(bp) &&
-	      xfs_inobt_verify(bp))) {
-		trace_xfs_btree_corrupt(bp, _RET_IP_);
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
-				     bp->b_target->bt_mount, bp->b_addr);
+	if (!xfs_btree_sblock_verify_crc(bp))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_inobt_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+
+	if (bp->b_error) {
+		trace_xfs_btree_corrupt(bp, _RET_IP_);
+		xfs_verifier_error(bp);
 	}
 }
 
@@ -258,9 +260,9 @@ xfs_inobt_write_verify(
 {
 	if (!xfs_inobt_verify(bp)) {
 		trace_xfs_btree_corrupt(bp, _RET_IP_);
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
-				     bp->b_target->bt_mount, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
+		return;
 	}
 	xfs_btree_sblock_calc_crc(bp);
 

fs/xfs/xfs_inode.c

@@ -42,7 +42,6 @@
 #include "xfs_bmap_util.h"
 #include "xfs_error.h"
 #include "xfs_quota.h"
-#include "xfs_dinode.h"
 #include "xfs_filestream.h"
 #include "xfs_cksum.h"
 #include "xfs_trace.h"
@@ -62,6 +61,8 @@ kmem_zone_t *xfs_inode_zone;
 
 STATIC int xfs_iflush_int(xfs_inode_t *, xfs_buf_t *);
 
+STATIC int xfs_iunlink_remove(xfs_trans_t *, xfs_inode_t *);
+
 /*
  * helper function to extract extent size hint from inode
  */
@@ -1115,7 +1116,7 @@ xfs_bumplink(
 {
 	xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG);
 
-	ASSERT(ip->i_d.di_nlink > 0);
+	ASSERT(ip->i_d.di_nlink > 0 || (VFS_I(ip)->i_state & I_LINKABLE));
 	ip->i_d.di_nlink++;
 	inc_nlink(VFS_I(ip));
 	if ((ip->i_d.di_version == 1) &&
@@ -1165,10 +1166,7 @@ xfs_create(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return XFS_ERROR(EIO);
 
-	if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
-		prid = xfs_get_projid(dp);
-	else
-		prid = XFS_PROJID_DEFAULT;
+	prid = xfs_get_initial_prid(dp);
 
 	/*
 	 * Make sure that we have allocated dquot(s) on disk.
@@ -1332,6 +1330,113 @@ xfs_create(
 	return error;
 }
 
+int
+xfs_create_tmpfile(
+	struct xfs_inode	*dp,
+	struct dentry		*dentry,
+	umode_t			mode)
+{
+	struct xfs_mount	*mp = dp->i_mount;
+	struct xfs_inode	*ip = NULL;
+	struct xfs_trans	*tp = NULL;
+	int			error;
+	uint			cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
+	prid_t                  prid;
+	struct xfs_dquot	*udqp = NULL;
+	struct xfs_dquot	*gdqp = NULL;
+	struct xfs_dquot	*pdqp = NULL;
+	struct xfs_trans_res	*tres;
+	uint			resblks;
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return XFS_ERROR(EIO);
+
+	prid = xfs_get_initial_prid(dp);
+
+	/*
+	 * Make sure that we have allocated dquot(s) on disk.
+	 */
+	error = xfs_qm_vop_dqalloc(dp, xfs_kuid_to_uid(current_fsuid()),
+				xfs_kgid_to_gid(current_fsgid()), prid,
+				XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT,
+				&udqp, &gdqp, &pdqp);
+	if (error)
+		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);
+	if (error == ENOSPC) {
+		/* No space at all so try a "no-allocation" reservation */
+		resblks = 0;
+		error = xfs_trans_reserve(tp, tres, 0, 0);
+	}
+	if (error) {
+		cancel_flags = 0;
+		goto out_trans_cancel;
+	}
+
+	error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp,
+						pdqp, resblks, 1, 0);
+	if (error)
+		goto out_trans_cancel;
+
+	error = xfs_dir_ialloc(&tp, dp, mode, 1, 0,
+				prid, resblks > 0, &ip, NULL);
+	if (error) {
+		if (error == ENOSPC)
+			goto out_trans_cancel;
+		goto out_trans_abort;
+	}
+
+	if (mp->m_flags & XFS_MOUNT_WSYNC)
+		xfs_trans_set_sync(tp);
+
+	/*
+	 * Attach the dquot(s) to the inodes and modify them incore.
+	 * These ids of the inode couldn't have changed since the new
+	 * inode has been locked ever since it was created.
+	 */
+	xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp, pdqp);
+
+	ip->i_d.di_nlink--;
+	d_tmpfile(dentry, VFS_I(ip));
+	error = xfs_iunlink(tp, ip);
+	if (error)
+		goto out_trans_abort;
+
+	error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
+	if (error)
+		goto out_release_inode;
+
+	xfs_qm_dqrele(udqp);
+	xfs_qm_dqrele(gdqp);
+	xfs_qm_dqrele(pdqp);
+
+	return 0;
+
+ out_trans_abort:
+	cancel_flags |= XFS_TRANS_ABORT;
+ out_trans_cancel:
+	xfs_trans_cancel(tp, cancel_flags);
+ out_release_inode:
+	/*
+	 * Wait until after the current transaction is aborted to
+	 * release the inode.  This prevents recursive transactions
+	 * and deadlocks from xfs_inactive.
+	 */
+	if (ip)
+		IRELE(ip);
+
+	xfs_qm_dqrele(udqp);
+	xfs_qm_dqrele(gdqp);
+	xfs_qm_dqrele(pdqp);
+
+	return error;
+}
+
 int
 xfs_link(
 	xfs_inode_t		*tdp,
@@ -1397,6 +1502,12 @@ xfs_link(
 
 	xfs_bmap_init(&free_list, &first_block);
 
+	if (sip->i_d.di_nlink == 0) {
+		error = xfs_iunlink_remove(tp, sip);
+		if (error)
+			goto abort_return;
+	}
+
 	error = xfs_dir_createname(tp, tdp, target_name, sip->i_ino,
 					&first_block, &free_list, resblks);
 	if (error)

fs/xfs/xfs_inode.h

@@ -20,6 +20,7 @@
 
 #include "xfs_inode_buf.h"
 #include "xfs_inode_fork.h"
+#include "xfs_dinode.h"
 
 /*
  * Kernel only inode definitions
@@ -192,6 +193,15 @@ xfs_set_projid(struct xfs_inode *ip,
 	ip->i_d.di_projid_lo = (__uint16_t) (projid & 0xffff);
 }
 
+static inline prid_t
+xfs_get_initial_prid(struct xfs_inode *dp)
+{
+	if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
+		return xfs_get_projid(dp);
+
+	return XFS_PROJID_DEFAULT;
+}
+
 /*
  * In-core inode flags.
  */
@@ -323,6 +333,8 @@ int		xfs_lookup(struct xfs_inode *dp, struct xfs_name *name,
 			   struct xfs_inode **ipp, struct xfs_name *ci_name);
 int		xfs_create(struct xfs_inode *dp, struct xfs_name *name,
 			   umode_t mode, xfs_dev_t rdev, struct xfs_inode **ipp);
+int		xfs_create_tmpfile(struct xfs_inode *dp, struct dentry *dentry,
+			   umode_t mode);
 int		xfs_remove(struct xfs_inode *dp, struct xfs_name *name,
 			   struct xfs_inode *ip);
 int		xfs_link(struct xfs_inode *tdp, struct xfs_inode *sip,

fs/xfs/xfs_inode_buf.c

@@ -102,8 +102,7 @@ xfs_inode_buf_verify(
 			}
 
 			xfs_buf_ioerror(bp, EFSCORRUPTED);
-			XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_HIGH,
-					     mp, dip);
+			xfs_verifier_error(bp);
 #ifdef DEBUG
 			xfs_alert(mp,
 				"bad inode magic/vsn daddr %lld #%d (magic=%x)",
@@ -306,7 +305,7 @@ xfs_dinode_verify(
 	if (!xfs_sb_version_hascrc(&mp->m_sb))
 		return false;
 	if (!xfs_verify_cksum((char *)dip, mp->m_sb.sb_inodesize,
-			      offsetof(struct xfs_dinode, di_crc)))
+			      XFS_DINODE_CRC_OFF))
 		return false;
 	if (be64_to_cpu(dip->di_ino) != ip->i_ino)
 		return false;
@@ -327,7 +326,7 @@ xfs_dinode_calc_crc(
 
 	ASSERT(xfs_sb_version_hascrc(&mp->m_sb));
 	crc = xfs_start_cksum((char *)dip, mp->m_sb.sb_inodesize,
-			      offsetof(struct xfs_dinode, di_crc));
+			      XFS_DINODE_CRC_OFF);
 	dip->di_crc = xfs_end_cksum(crc);
 }
 

fs/xfs/xfs_iomap.c

@@ -128,7 +128,6 @@ xfs_iomap_write_direct(
 	xfs_fsblock_t	firstfsb;
 	xfs_extlen_t	extsz, temp;
 	int		nimaps;
-	int		bmapi_flag;
 	int		quota_flag;
 	int		rt;
 	xfs_trans_t	*tp;
@@ -200,18 +199,15 @@ xfs_iomap_write_direct(
 
 	xfs_trans_ijoin(tp, ip, 0);
 
-	bmapi_flag = 0;
-	if (offset < XFS_ISIZE(ip) || extsz)
-		bmapi_flag |= XFS_BMAPI_PREALLOC;
-
 	/*
 	 * From this point onwards we overwrite the imap pointer that the
 	 * caller gave to us.
 	 */
 	xfs_bmap_init(&free_list, &firstfsb);
 	nimaps = 1;
-	error = xfs_bmapi_write(tp, ip, offset_fsb, count_fsb, bmapi_flag,
-				&firstfsb, 0, imap, &nimaps, &free_list);
+	error = xfs_bmapi_write(tp, ip, offset_fsb, count_fsb,
+				XFS_BMAPI_PREALLOC, &firstfsb, 0,
+				imap, &nimaps, &free_list);
 	if (error)
 		goto out_bmap_cancel;
 

fs/xfs/xfs_iops.c

@@ -39,6 +39,7 @@
 #include "xfs_da_btree.h"
 #include "xfs_dir2_priv.h"
 #include "xfs_dinode.h"
+#include "xfs_trans_space.h"
 
 #include <linux/capability.h>
 #include <linux/xattr.h>
@@ -48,6 +49,18 @@
 #include <linux/fiemap.h>
 #include <linux/slab.h>
 
+/*
+ * Directories have different lock order w.r.t. mmap_sem compared to regular
+ * files. This is due to readdir potentially triggering page faults on a user
+ * buffer inside filldir(), and this happens with the ilock on the directory
+ * held. For regular files, the lock order is the other way around - the
+ * mmap_sem is taken during the page fault, and then we lock the ilock to do
+ * block mapping. Hence we need a different class for the directory ilock so
+ * that lockdep can tell them apart.
+ */
+static struct lock_class_key xfs_nondir_ilock_class;
+static struct lock_class_key xfs_dir_ilock_class;
+
 static int
 xfs_initxattrs(
 	struct inode		*inode,
@@ -1034,6 +1047,19 @@ xfs_vn_fiemap(
 	return 0;
 }
 
+STATIC int
+xfs_vn_tmpfile(
+	struct inode	*dir,
+	struct dentry	*dentry,
+	umode_t		mode)
+{
+	int		error;
+
+	error = xfs_create_tmpfile(XFS_I(dir), dentry, mode);
+
+	return -error;
+}
+
 static const struct inode_operations xfs_inode_operations = {
 	.get_acl		= xfs_get_acl,
 	.set_acl		= xfs_set_acl,
@@ -1072,6 +1098,7 @@ static const struct inode_operations xfs_dir_inode_operations = {
 	.removexattr		= generic_removexattr,
 	.listxattr		= xfs_vn_listxattr,
 	.update_time		= xfs_vn_update_time,
+	.tmpfile		= xfs_vn_tmpfile,
 };
 
 static const struct inode_operations xfs_dir_ci_inode_operations = {
@@ -1099,6 +1126,7 @@ static const struct inode_operations xfs_dir_ci_inode_operations = {
 	.removexattr		= generic_removexattr,
 	.listxattr		= xfs_vn_listxattr,
 	.update_time		= xfs_vn_update_time,
+	.tmpfile		= xfs_vn_tmpfile,
 };
 
 static const struct inode_operations xfs_symlink_inode_operations = {
@@ -1191,6 +1219,7 @@ xfs_setup_inode(
 	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;
@@ -1198,6 +1227,7 @@ xfs_setup_inode(
 		inode->i_mapping->a_ops = &xfs_address_space_operations;
 		break;
 	case S_IFDIR:
+		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

fs/xfs/xfs_linux.h

@@ -119,6 +119,7 @@ typedef __uint64_t __psunsigned_t;
 #include "xfs_iops.h"
 #include "xfs_aops.h"
 #include "xfs_super.h"
+#include "xfs_cksum.h"
 #include "xfs_buf.h"
 #include "xfs_message.h"
 
@@ -178,6 +179,7 @@ typedef __uint64_t __psunsigned_t;
 #define ENOATTR		ENODATA		/* Attribute not found */
 #define EWRONGFS	EINVAL		/* Mount with wrong filesystem type */
 #define EFSCORRUPTED	EUCLEAN		/* Filesystem is corrupted */
+#define EFSBADCRC	EBADMSG		/* Bad CRC detected */
 
 #define SYNCHRONIZE()	barrier()
 #define __return_address __builtin_return_address(0)

fs/xfs/xfs_log.h

@@ -175,7 +175,7 @@ void	  xlog_iodone(struct xfs_buf *);
 struct xlog_ticket *xfs_log_ticket_get(struct xlog_ticket *ticket);
 void	  xfs_log_ticket_put(struct xlog_ticket *ticket);
 
-int	xfs_log_commit_cil(struct xfs_mount *mp, struct xfs_trans *tp,
+void	xfs_log_commit_cil(struct xfs_mount *mp, struct xfs_trans *tp,
 				xfs_lsn_t *commit_lsn, int flags);
 bool	xfs_log_item_in_current_chkpt(struct xfs_log_item *lip);
 

fs/xfs/xfs_log_cil.c

@@ -498,13 +498,6 @@ xlog_cil_push(
 	new_ctx->cil = cil;
 	cil->xc_ctx = new_ctx;
 
-	/*
-	 * mirror the new sequence into the cil structure so that we can do
-	 * unlocked checks against the current sequence in log forces without
-	 * risking deferencing a freed context pointer.
-	 */
-	cil->xc_current_sequence = new_ctx->sequence;
-
 	/*
 	 * The switch is now done, so we can drop the context lock and move out
 	 * of a shared context. We can't just go straight to the commit record,
@@ -523,8 +516,15 @@ xlog_cil_push(
 	 * Hence we need to add this context to the committing context list so
 	 * that higher sequences will wait for us to write out a commit record
 	 * before they do.
+	 *
+	 * xfs_log_force_lsn requires us to mirror the new sequence into the cil
+	 * structure atomically with the addition of this sequence to the
+	 * committing list. This also ensures that we can do unlocked checks
+	 * against the current sequence in log forces without risking
+	 * deferencing a freed context pointer.
 	 */
 	spin_lock(&cil->xc_push_lock);
+	cil->xc_current_sequence = new_ctx->sequence;
 	list_add(&ctx->committing, &cil->xc_committing);
 	spin_unlock(&cil->xc_push_lock);
 	up_write(&cil->xc_ctx_lock);
@@ -662,8 +662,14 @@ xlog_cil_push_background(
 
 }
 
+/*
+ * xlog_cil_push_now() is used to trigger an immediate CIL push to the sequence
+ * number that is passed. When it returns, the work will be queued for
+ * @push_seq, but it won't be completed. The caller is expected to do any
+ * waiting for push_seq to complete if it is required.
+ */
 static void
-xlog_cil_push_foreground(
+xlog_cil_push_now(
 	struct xlog	*log,
 	xfs_lsn_t	push_seq)
 {
@@ -688,10 +694,8 @@ xlog_cil_push_foreground(
 	}
 
 	cil->xc_push_seq = push_seq;
+	queue_work(log->l_mp->m_cil_workqueue, &cil->xc_push_work);
 	spin_unlock(&cil->xc_push_lock);
-
-	/* do the push now */
-	xlog_cil_push(log);
 }
 
 bool
@@ -721,7 +725,7 @@ xlog_cil_empty(
  * background commit, returns without it held once background commits are
  * allowed again.
  */
-int
+void
 xfs_log_commit_cil(
 	struct xfs_mount	*mp,
 	struct xfs_trans	*tp,
@@ -767,7 +771,6 @@ xfs_log_commit_cil(
 	xlog_cil_push_background(log);
 
 	up_read(&cil->xc_ctx_lock);
-	return 0;
 }
 
 /*
@@ -796,7 +799,8 @@ xlog_cil_force_lsn(
 	 * xlog_cil_push() handles racing pushes for the same sequence,
 	 * so no need to deal with it here.
 	 */
-	xlog_cil_push_foreground(log, sequence);
+restart:
+	xlog_cil_push_now(log, sequence);
 
 	/*
 	 * See if we can find a previous sequence still committing.
@@ -804,7 +808,6 @@ xlog_cil_force_lsn(
 	 * before allowing the force of push_seq to go ahead. Hence block
 	 * on commits for those as well.
 	 */
-restart:
 	spin_lock(&cil->xc_push_lock);
 	list_for_each_entry(ctx, &cil->xc_committing, committing) {
 		if (ctx->sequence > sequence)
@@ -822,6 +825,28 @@ restart:
 		/* found it! */
 		commit_lsn = ctx->commit_lsn;
 	}
+
+	/*
+	 * The call to xlog_cil_push_now() executes the push in the background.
+	 * Hence by the time we have got here it our sequence may not have been
+	 * pushed yet. This is true if the current sequence still matches the
+	 * push sequence after the above wait loop and the CIL still contains
+	 * dirty objects.
+	 *
+	 * When the push occurs, it will empty the CIL and
+	 * atomically increment the currect sequence past the push sequence and
+	 * move it into the committing list. Of course, if the CIL is clean at
+	 * the time of the push, it won't have pushed the CIL at all, so in that
+	 * case we should try the push for this sequence again from the start
+	 * just in case.
+	 */
+
+	if (sequence == cil->xc_current_sequence &&
+	    !list_empty(&cil->xc_cil)) {
+		spin_unlock(&cil->xc_push_lock);
+		goto restart;
+	}
+
 	spin_unlock(&cil->xc_push_lock);
 	return commit_lsn;
 }

fs/xfs/xfs_mount.c

@@ -314,6 +314,9 @@ reread:
 		error = bp->b_error;
 		if (loud)
 			xfs_warn(mp, "SB validate failed with error %d.", error);
+		/* bad CRC means corrupted metadata */
+		if (error == EFSBADCRC)
+			error = EFSCORRUPTED;
 		goto release_buf;
 	}
 

fs/xfs/xfs_rtalloc.c

@@ -842,7 +842,7 @@ xfs_growfs_rt_alloc(
 		/*
 		 * Reserve space & log for one extent added to the file.
 		 */
-		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_growdata,
+		error = xfs_trans_reserve(tp, &M_RES(mp)->tr_growrtalloc,
 					  resblks, 0);
 		if (error)
 			goto error_cancel;

fs/xfs/xfs_sb.c

@@ -288,6 +288,7 @@ xfs_mount_validate_sb(
 	    sbp->sb_inodelog < XFS_DINODE_MIN_LOG			||
 	    sbp->sb_inodelog > XFS_DINODE_MAX_LOG			||
 	    sbp->sb_inodesize != (1 << sbp->sb_inodelog)		||
+	    sbp->sb_inopblock != howmany(sbp->sb_blocksize,sbp->sb_inodesize) ||
 	    (sbp->sb_blocklog - sbp->sb_inodelog != sbp->sb_inopblog)	||
 	    (sbp->sb_rextsize * sbp->sb_blocksize > XFS_MAX_RTEXTSIZE)	||
 	    (sbp->sb_rextsize * sbp->sb_blocksize < XFS_MIN_RTEXTSIZE)	||
@@ -610,12 +611,11 @@ xfs_sb_read_verify(
 						XFS_SB_VERSION_5) ||
 	     dsb->sb_crc != 0)) {
 
-		if (!xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-				      offsetof(struct xfs_sb, sb_crc))) {
+		if (!xfs_buf_verify_cksum(bp, XFS_SB_CRC_OFF)) {
 			/* Only fail bad secondaries on a known V5 filesystem */
 			if (bp->b_bn == XFS_SB_DADDR ||
 			    xfs_sb_version_hascrc(&mp->m_sb)) {
-				error = EFSCORRUPTED;
+				error = EFSBADCRC;
 				goto out_error;
 			}
 		}
@@ -624,10 +624,9 @@ xfs_sb_read_verify(
 
 out_error:
 	if (error) {
-		if (error == EFSCORRUPTED)
-			XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
-					     mp, bp->b_addr);
 		xfs_buf_ioerror(bp, error);
+		if (error == EFSCORRUPTED || error == EFSBADCRC)
+			xfs_verifier_error(bp);
 	}
 }
 
@@ -662,9 +661,8 @@ xfs_sb_write_verify(
 
 	error = xfs_sb_verify(bp, false);
 	if (error) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
-				     mp, bp->b_addr);
 		xfs_buf_ioerror(bp, error);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -674,8 +672,7 @@ xfs_sb_write_verify(
 	if (bip)
 		XFS_BUF_TO_SBP(bp)->sb_lsn = cpu_to_be64(bip->bli_item.li_lsn);
 
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
-			 offsetof(struct xfs_sb, sb_crc));
+	xfs_buf_update_cksum(bp, XFS_SB_CRC_OFF);
 }
 
 const struct xfs_buf_ops xfs_sb_buf_ops = {

fs/xfs/xfs_sb.h

@@ -182,6 +182,8 @@ typedef struct xfs_sb {
 	/* must be padded to 64 bit alignment */
 } xfs_sb_t;
 
+#define XFS_SB_CRC_OFF		offsetof(struct xfs_sb, sb_crc)
+
 /*
  * Superblock - on disk version.  Must match the in core version above.
  * Must be padded to 64 bit alignment.

fs/xfs/xfs_shared.h

@@ -104,7 +104,8 @@ extern const struct xfs_buf_ops xfs_symlink_buf_ops;
 #define	XFS_TRANS_SB_COUNT		41
 #define	XFS_TRANS_CHECKPOINT		42
 #define	XFS_TRANS_ICREATE		43
-#define	XFS_TRANS_TYPE_MAX		43
+#define	XFS_TRANS_CREATE_TMPFILE	44
+#define	XFS_TRANS_TYPE_MAX		44
 /* new transaction types need to be reflected in xfs_logprint(8) */
 
 #define XFS_TRANS_TYPES \
@@ -112,6 +113,7 @@ 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" }, \

fs/xfs/xfs_symlink.c

@@ -80,6 +80,10 @@ xfs_readlink_bmap(
 		if (error) {
 			xfs_buf_ioerror_alert(bp, __func__);
 			xfs_buf_relse(bp);
+
+			/* bad CRC means corrupted metadata */
+			if (error == EFSBADCRC)
+				error = EFSCORRUPTED;
 			goto out;
 		}
 		byte_cnt = XFS_SYMLINK_BUF_SPACE(mp, byte_cnt);
@@ -208,10 +212,7 @@ xfs_symlink(
 		return XFS_ERROR(ENAMETOOLONG);
 
 	udqp = gdqp = NULL;
-	if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
-		prid = xfs_get_projid(dp);
-	else
-		prid = XFS_PROJID_DEFAULT;
+	prid = xfs_get_initial_prid(dp);
 
 	/*
 	 * Make sure that we have allocated dquot(s) on disk.

fs/xfs/xfs_symlink_remote.c

@@ -133,12 +133,13 @@ xfs_symlink_read_verify(
 	if (!xfs_sb_version_hascrc(&mp->m_sb))
 		return;
 
-	if (!xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
-				  offsetof(struct xfs_dsymlink_hdr, sl_crc)) ||
-	    !xfs_symlink_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+	if (!xfs_buf_verify_cksum(bp, XFS_SYMLINK_CRC_OFF))
+		xfs_buf_ioerror(bp, EFSBADCRC);
+	else if (!xfs_symlink_verify(bp))
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+
+	if (bp->b_error)
+		xfs_verifier_error(bp);
 }
 
 static void
@@ -153,8 +154,8 @@ xfs_symlink_write_verify(
 		return;
 
 	if (!xfs_symlink_verify(bp)) {
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
 		xfs_buf_ioerror(bp, EFSCORRUPTED);
+		xfs_verifier_error(bp);
 		return;
 	}
 
@@ -162,8 +163,7 @@ xfs_symlink_write_verify(
 		struct xfs_dsymlink_hdr *dsl = bp->b_addr;
 		dsl->sl_lsn = cpu_to_be64(bip->bli_item.li_lsn);
 	}
-	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
-			 offsetof(struct xfs_dsymlink_hdr, sl_crc));
+	xfs_buf_update_cksum(bp, XFS_SYMLINK_CRC_OFF);
 }
 
 const struct xfs_buf_ops xfs_symlink_buf_ops = {

fs/xfs/xfs_trace.h

@@ -603,6 +603,7 @@ DEFINE_INODE_EVENT(xfs_readlink);
 DEFINE_INODE_EVENT(xfs_inactive_symlink);
 DEFINE_INODE_EVENT(xfs_alloc_file_space);
 DEFINE_INODE_EVENT(xfs_free_file_space);
+DEFINE_INODE_EVENT(xfs_collapse_file_space);
 DEFINE_INODE_EVENT(xfs_readdir);
 #ifdef CONFIG_XFS_POSIX_ACL
 DEFINE_INODE_EVENT(xfs_get_acl);

fs/xfs/xfs_trans.c

@@ -887,12 +887,7 @@ xfs_trans_commit(
 		xfs_trans_apply_sb_deltas(tp);
 	xfs_trans_apply_dquot_deltas(tp);
 
-	error = xfs_log_commit_cil(mp, tp, &commit_lsn, flags);
-	if (error == ENOMEM) {
-		xfs_force_shutdown(mp, SHUTDOWN_LOG_IO_ERROR);
-		error = XFS_ERROR(EIO);
-		goto out_unreserve;
-	}
+	xfs_log_commit_cil(mp, tp, &commit_lsn, flags);
 
 	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
 	xfs_trans_free(tp);
@@ -902,10 +897,7 @@ xfs_trans_commit(
 	 * log out now and wait for it.
 	 */
 	if (sync) {
-		if (!error) {
-			error = _xfs_log_force_lsn(mp, commit_lsn,
-				      XFS_LOG_SYNC, NULL);
-		}
+		error = _xfs_log_force_lsn(mp, commit_lsn, XFS_LOG_SYNC, NULL);
 		XFS_STATS_INC(xs_trans_sync);
 	} else {
 		XFS_STATS_INC(xs_trans_async);

fs/xfs/xfs_trans_buf.c

@@ -275,6 +275,10 @@ xfs_trans_read_buf_map(
 			XFS_BUF_UNDONE(bp);
 			xfs_buf_stale(bp);
 			xfs_buf_relse(bp);
+
+			/* bad CRC means corrupted metadata */
+			if (error == EFSBADCRC)
+				error = EFSCORRUPTED;
 			return error;
 		}
 #ifdef DEBUG
@@ -338,6 +342,9 @@ xfs_trans_read_buf_map(
 				if (tp->t_flags & XFS_TRANS_DIRTY)
 					xfs_force_shutdown(tp->t_mountp,
 							SHUTDOWN_META_IO_ERROR);
+				/* bad CRC means corrupted metadata */
+				if (error == EFSBADCRC)
+					error = EFSCORRUPTED;
 				return error;
 			}
 		}
@@ -375,6 +382,10 @@ xfs_trans_read_buf_map(
 		if (tp->t_flags & XFS_TRANS_DIRTY)
 			xfs_force_shutdown(tp->t_mountp, SHUTDOWN_META_IO_ERROR);
 		xfs_buf_relse(bp);
+
+		/* bad CRC means corrupted metadata */
+		if (error == EFSBADCRC)
+			error = EFSCORRUPTED;
 		return error;
 	}
 #ifdef DEBUG

fs/xfs/xfs_trans_resv.c

@@ -81,20 +81,28 @@ xfs_calc_buf_res(
  * on disk. Hence we need an inode reservation function that calculates all this
  * correctly. So, we log:
  *
- * - log op headers for object
+ * - 4 log op headers for object
+ *	- for the ilf, the inode core and 2 forks
  * - inode log format object
- * - the entire inode contents (core + 2 forks)
- * - two bmap btree block headers
+ * - the inode core
+ * - two inode forks containing bmap btree root blocks.
+ *	- the btree data contained by both forks will fit into the inode size,
+ *	  hence when combined with the inode core above, we have a total of the
+ *	  actual inode size.
+ *	- the BMBT headers need to be accounted separately, as they are
+ *	  additional to the records and pointers that fit inside the inode
+ *	  forks.
  */
 STATIC uint
 xfs_calc_inode_res(
 	struct xfs_mount	*mp,
 	uint			ninodes)
 {
-	return ninodes * (sizeof(struct xlog_op_header) +
-			  sizeof(struct xfs_inode_log_format) +
-			  mp->m_sb.sb_inodesize +
-			  2 * XFS_BMBT_BLOCK_LEN(mp));
+	return ninodes *
+		(4 * sizeof(struct xlog_op_header) +
+		 sizeof(struct xfs_inode_log_format) +
+		 mp->m_sb.sb_inodesize +
+		 2 * XFS_BMBT_BLOCK_LEN(mp));
 }
 
 /*
@@ -203,6 +211,19 @@ xfs_calc_rename_reservation(
 				      XFS_FSB_TO_B(mp, 1))));
 }
 
+/*
+ * For removing an inode from unlinked list at first, we can modify:
+ *    the agi hash list and counters: sector size
+ *    the on disk inode before ours in the agi hash list: inode cluster size
+ */
+STATIC uint
+xfs_calc_iunlink_remove_reservation(
+	struct xfs_mount        *mp)
+{
+	return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
+	       max_t(uint, XFS_FSB_TO_B(mp, 1), mp->m_inode_cluster_size);
+}
+
 /*
  * For creating a link to an inode:
  *    the parent directory inode: inode size
@@ -220,6 +241,7 @@ xfs_calc_link_reservation(
 	struct xfs_mount	*mp)
 {
 	return XFS_DQUOT_LOGRES(mp) +
+		xfs_calc_iunlink_remove_reservation(mp) +
 		MAX((xfs_calc_inode_res(mp, 2) +
 		     xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp),
 				      XFS_FSB_TO_B(mp, 1))),
@@ -228,6 +250,18 @@ xfs_calc_link_reservation(
 				      XFS_FSB_TO_B(mp, 1))));
 }
 
+/*
+ * For adding an inode to unlinked list we can modify:
+ *    the agi hash list: sector size
+ *    the unlinked inode: inode size
+ */
+STATIC uint
+xfs_calc_iunlink_add_reservation(xfs_mount_t *mp)
+{
+	return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
+		xfs_calc_inode_res(mp, 1);
+}
+
 /*
  * For removing a directory entry we can modify:
  *    the parent directory inode: inode size
@@ -245,10 +279,11 @@ xfs_calc_remove_reservation(
 	struct xfs_mount	*mp)
 {
 	return XFS_DQUOT_LOGRES(mp) +
-		MAX((xfs_calc_inode_res(mp, 2) +
+		xfs_calc_iunlink_add_reservation(mp) +
+		MAX((xfs_calc_inode_res(mp, 1) +
 		     xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp),
 				      XFS_FSB_TO_B(mp, 1))),
-		    (xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) +
+		    (xfs_calc_buf_res(4, mp->m_sb.sb_sectsize) +
 		     xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 2),
 				      XFS_FSB_TO_B(mp, 1))));
 }
@@ -343,6 +378,20 @@ xfs_calc_create_reservation(
 
 }
 
+STATIC uint
+xfs_calc_create_tmpfile_reservation(
+	struct xfs_mount        *mp)
+{
+	uint	res = XFS_DQUOT_LOGRES(mp);
+
+	if (xfs_sb_version_hascrc(&mp->m_sb))
+		res += xfs_calc_icreate_resv_alloc(mp);
+	else
+		res += xfs_calc_create_resv_alloc(mp);
+
+	return res + xfs_calc_iunlink_add_reservation(mp);
+}
+
 /*
  * Making a new directory is the same as creating a new file.
  */
@@ -383,9 +432,9 @@ xfs_calc_ifree_reservation(
 {
 	return XFS_DQUOT_LOGRES(mp) +
 		xfs_calc_inode_res(mp, 1) +
-		xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) +
+		xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) +
 		xfs_calc_buf_res(1, XFS_FSB_TO_B(mp, 1)) +
-		max_t(uint, XFS_FSB_TO_B(mp, 1), mp->m_inode_cluster_size) +
+		xfs_calc_iunlink_remove_reservation(mp) +
 		xfs_calc_buf_res(1, 0) +
 		xfs_calc_buf_res(2 + mp->m_ialloc_blks +
 				 mp->m_in_maxlevels, 0) +
@@ -644,15 +693,14 @@ xfs_calc_qm_setqlim_reservation(
 
 /*
  * Allocating quota on disk if needed.
- *	the write transaction log space: M_RES(mp)->tr_write.tr_logres
+ *	the write transaction log space for quota file extent allocation
  *	the unit of quota allocation: one system block size
  */
 STATIC uint
 xfs_calc_qm_dqalloc_reservation(
 	struct xfs_mount	*mp)
 {
-	ASSERT(M_RES(mp)->tr_write.tr_logres);
-	return M_RES(mp)->tr_write.tr_logres +
+	return xfs_calc_write_reservation(mp) +
 		xfs_calc_buf_res(1,
 			XFS_FSB_TO_B(mp, XFS_DQUOT_CLUSTER_SIZE_FSB) - 1);
 }
@@ -729,6 +777,11 @@ xfs_trans_resv_calc(
 	resp->tr_create.tr_logcount = XFS_CREATE_LOG_COUNT;
 	resp->tr_create.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
 
+	resp->tr_create_tmpfile.tr_logres =
+			xfs_calc_create_tmpfile_reservation(mp);
+	resp->tr_create_tmpfile.tr_logcount = XFS_CREATE_TMPFILE_LOG_COUNT;
+	resp->tr_create_tmpfile.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
 	resp->tr_mkdir.tr_logres = xfs_calc_mkdir_reservation(mp);
 	resp->tr_mkdir.tr_logcount = XFS_MKDIR_LOG_COUNT;
 	resp->tr_mkdir.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
@@ -784,7 +837,6 @@ xfs_trans_resv_calc(
 	/* The following transaction are logged in logical format */
 	resp->tr_ichange.tr_logres = xfs_calc_ichange_reservation(mp);
 	resp->tr_growdata.tr_logres = xfs_calc_growdata_reservation(mp);
-	resp->tr_swrite.tr_logres = xfs_calc_swrite_reservation(mp);
 	resp->tr_fsyncts.tr_logres = xfs_calc_swrite_reservation(mp);
 	resp->tr_writeid.tr_logres = xfs_calc_writeid_reservation(mp);
 	resp->tr_attrsetrt.tr_logres = xfs_calc_attrsetrt_reservation(mp);

fs/xfs/xfs_trans_resv.h

@@ -38,11 +38,11 @@ struct xfs_trans_resv {
 	struct xfs_trans_res	tr_remove;	/* unlink trans */
 	struct xfs_trans_res	tr_symlink;	/* symlink trans */
 	struct xfs_trans_res	tr_create;	/* create trans */
+	struct xfs_trans_res	tr_create_tmpfile; /* create O_TMPFILE trans */
 	struct xfs_trans_res	tr_mkdir;	/* mkdir trans */
 	struct xfs_trans_res	tr_ifree;	/* inode free trans */
 	struct xfs_trans_res	tr_ichange;	/* inode update trans */
 	struct xfs_trans_res	tr_growdata;	/* fs data section grow trans */
-	struct xfs_trans_res	tr_swrite;	/* sync write inode trans */
 	struct xfs_trans_res	tr_addafork;	/* add inode attr fork trans */
 	struct xfs_trans_res	tr_writeid;	/* write setuid/setgid file */
 	struct xfs_trans_res	tr_attrinval;	/* attr fork buffer
@@ -100,6 +100,7 @@ struct xfs_trans_resv {
 #define	XFS_ITRUNCATE_LOG_COUNT		2
 #define XFS_INACTIVE_LOG_COUNT		2
 #define	XFS_CREATE_LOG_COUNT		2
+#define	XFS_CREATE_TMPFILE_LOG_COUNT	2
 #define	XFS_MKDIR_LOG_COUNT		3
 #define	XFS_SYMLINK_LOG_COUNT		3
 #define	XFS_REMOVE_LOG_COUNT		2

include/linux/fs.h

@@ -2550,6 +2550,9 @@ enum {
 
 	/* filesystem does not support filling holes */
 	DIO_SKIP_HOLES	= 0x02,
+
+	/* filesystem can handle aio writes beyond i_size */
+	DIO_ASYNC_EXTEND = 0x04,
 };
 
 void dio_end_io(struct bio *bio, int error);

include/uapi/linux/falloc.h

@@ -5,5 +5,40 @@
 #define FALLOC_FL_PUNCH_HOLE	0x02 /* de-allocates range */
 #define FALLOC_FL_NO_HIDE_STALE	0x04 /* reserved codepoint */
 
+/*
+ * FALLOC_FL_COLLAPSE_RANGE is used to remove a range of a file
+ * without leaving a hole in the file. The contents of the file beyond
+ * the range being removed is appended to the start offset of the range
+ * being removed (i.e. the hole that was punched is "collapsed"),
+ * resulting in a file layout that looks like the range that was
+ * removed never existed. As such collapsing a range of a file changes
+ * the size of the file, reducing it by the same length of the range
+ * that has been removed by the operation.
+ *
+ * Different filesystems may implement different limitations on the
+ * granularity of the operation. Most will limit operations to
+ * filesystem block size boundaries, but this boundary may be larger or
+ * smaller depending on the filesystem and/or the configuration of the
+ * filesystem or file.
+ *
+ * Attempting to collapse a range that crosses the end of the file is
+ * considered an illegal operation - just use ftruncate(2) if you need
+ * to collapse a range that crosses EOF.
+ */
+#define FALLOC_FL_COLLAPSE_RANGE	0x08
+
+/*
+ * FALLOC_FL_ZERO_RANGE is used to convert a range of file to zeros preferably
+ * without issuing data IO. Blocks should be preallocated for the regions that
+ * span holes in the file, and the entire range is preferable converted to
+ * unwritten extents - even though file system may choose to zero out the
+ * extent or do whatever which will result in reading zeros from the range
+ * while the range remains allocated for the file.
+ *
+ * This can be also used to preallocate blocks past EOF in the same way as
+ * with fallocate. Flag FALLOC_FL_KEEP_SIZE should cause the inode
+ * size to remain the same.
+ */
+#define FALLOC_FL_ZERO_RANGE		0x10
 
 #endif /* _UAPI_FALLOC_H_ */