xfs: fix inode size update overflow in xfs_map_direct()

Both direct IO and DAX pass an offset and count into get_blocks that
will overflow a s64 variable when an IO goes into the last supported
block in a file (i.e. at offset 2^63 - 1FSB bytes). This can be seen
from the tracing:

xfs_get_blocks_alloc: [...] offset 0x7ffffffffffff000 count 4096
xfs_gbmap_direct:     [...] offset 0x7ffffffffffff000 count 4096
xfs_gbmap_direct_none:[...] offset 0x7ffffffffffff000 count 4096

0x7ffffffffffff000 + 4096 = 0x8000000000000000, and hence that
overflows the s64 offset and we fail to detect the need for a
filesize update and an ioend is not allocated.

This is *mostly* avoided for direct IO because such extending IOs
occur with full block allocation, and so the "IS_UNWRITTEN()" check
still evaluates as true and we get an ioend that way. However, doing
single sector extending IOs to this last block will expose the fact
that file size updates will not occur after the first allocating
direct IO as the overflow will then be exposed.

There is one further complexity: the DAX page fault path also
exposes the same issue in block allocation. However, page faults
cannot extend the file size, so in this case we want to allocate the
block but do not want to allocate an ioend to enable file size
update at IO completion. Hence we now need to distinguish between
the direct IO patch allocation and dax fault path allocation to
avoid leaking ioend structures.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>

1 files changed, 44 insertions, 6 deletions

diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c
index 50ab287..e747d6a 100644
--- a/fs/xfs/xfs_aops.c
+++ b/fs/xfs/xfs_aops.c
@@ -1250,13 +1250,28 @@ xfs_vm_releasepage(
  * the DIO. There is only going to be one reference to the ioend and its life
  * cycle is constrained by the DIO completion code. hence we don't need
  * reference counting here.
+ *
+ * Note that for DIO, an IO to the highest supported file block offset (i.e.
+ * 2^63 - 1FSB bytes) will result in the offset + count overflowing a signed 64
+ * bit variable. Hence if we see this overflow, we have to assume that the IO is
+ * extending the file size. We won't know for sure until IO completion is run
+ * and the actual max write offset is communicated to the IO completion
+ * routine.
+ *
+ * For DAX page faults, we are preparing to never see unwritten extents here,
+ * nor should we ever extend the inode size. Hence we will soon have nothing to
+ * do here for this case, ensuring we don't have to provide an IO completion
+ * callback to free an ioend that we don't actually need for a fault into the
+ * page at offset (2^63 - 1FSB) bytes.
  */
+
 static void
 xfs_map_direct(
 	struct inode		*inode,
 	struct buffer_head	*bh_result,
 	struct xfs_bmbt_irec	*imap,
-	xfs_off_t		offset)
+	xfs_off_t		offset,
+	bool			dax_fault)
 {
 	struct xfs_ioend	*ioend;
 	xfs_off_t		size = bh_result->b_size;
@@ -1269,6 +1284,16 @@ xfs_map_direct(
 
 	trace_xfs_gbmap_direct(XFS_I(inode), offset, size, type, imap);
 
+	/* XXX: preparation for removing unwritten extents in DAX */
+#if 0
+	if (dax_fault) {
+		ASSERT(type == XFS_IO_OVERWRITE);
+		trace_xfs_gbmap_direct_none(XFS_I(inode), offset, size, type,
+					    imap);
+		return;
+	}
+#endif
+
 	if (bh_result->b_private) {
 		ioend = bh_result->b_private;
 		ASSERT(ioend->io_size > 0);
@@ -1283,7 +1308,8 @@ xfs_map_direct(
 					      ioend->io_size, ioend->io_type,
 					      imap);
 	} else if (type == XFS_IO_UNWRITTEN ||
-		   offset + size > i_size_read(inode)) {
+		   offset + size > i_size_read(inode) ||
+		   offset + size < 0) {
 		ioend = xfs_alloc_ioend(inode, type);
 		ioend->io_offset = offset;
 		ioend->io_size = size;
@@ -1345,7 +1371,8 @@ __xfs_get_blocks(
 	sector_t		iblock,
 	struct buffer_head	*bh_result,
 	int			create,
-	bool			direct)
+	bool			direct,
+	bool			dax_fault)
 {
 	struct xfs_inode	*ip = XFS_I(inode);
 	struct xfs_mount	*mp = ip->i_mount;
@@ -1458,7 +1485,8 @@ __xfs_get_blocks(
 			set_buffer_unwritten(bh_result);
 		/* direct IO needs special help */
 		if (create && direct)
-			xfs_map_direct(inode, bh_result, &imap, offset);
+			xfs_map_direct(inode, bh_result, &imap, offset,
+				       dax_fault);
 	}
 
 	/*
@@ -1505,7 +1533,7 @@ xfs_get_blocks(
 	struct buffer_head	*bh_result,
 	int			create)
 {
-	return __xfs_get_blocks(inode, iblock, bh_result, create, false);
+	return __xfs_get_blocks(inode, iblock, bh_result, create, false, false);
 }
 
 int
@@ -1515,7 +1543,17 @@ xfs_get_blocks_direct(
 	struct buffer_head	*bh_result,
 	int			create)
 {
-	return __xfs_get_blocks(inode, iblock, bh_result, create, true);
+	return __xfs_get_blocks(inode, iblock, bh_result, create, true, false);
+}
+
+int
+xfs_get_blocks_dax_fault(
+	struct inode		*inode,
+	sector_t		iblock,
+	struct buffer_head	*bh_result,
+	int			create)
+{
+	return __xfs_get_blocks(inode, iblock, bh_result, create, true, true);
 }
 
 static void