1 files changed, 60 insertions, 89 deletions

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index d054b73b56fb..f5effa68e037 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -210,62 +210,21 @@ xfs_file_dio_aio_read(
 	struct kiocb		*iocb,
 	struct iov_iter		*to)
 {
-	struct address_space	*mapping = iocb->ki_filp->f_mapping;
-	struct inode		*inode = mapping->host;
-	struct xfs_inode	*ip = XFS_I(inode);
-	loff_t			isize = i_size_read(inode);
+	struct xfs_inode	*ip = XFS_I(file_inode(iocb->ki_filp));
 	size_t			count = iov_iter_count(to);
-	loff_t			end = iocb->ki_pos + count - 1;
-	struct iov_iter		data;
-	struct xfs_buftarg	*target;
-	ssize_t			ret = 0;
+	ssize_t			ret;
 
 	trace_xfs_file_direct_read(ip, count, iocb->ki_pos);
 
 	if (!count)
 		return 0; /* skip atime */
 
-	if (XFS_IS_REALTIME_INODE(ip))
-		target = ip->i_mount->m_rtdev_targp;
-	else
-		target = ip->i_mount->m_ddev_targp;
-
-	/* DIO must be aligned to device logical sector size */
-	if ((iocb->ki_pos | count) & target->bt_logical_sectormask) {
-		if (iocb->ki_pos == isize)
-			return 0;
-		return -EINVAL;
-	}
-
 	file_accessed(iocb->ki_filp);
 
 	xfs_ilock(ip, XFS_IOLOCK_SHARED);
-	if (mapping->nrpages) {
-		ret = filemap_write_and_wait_range(mapping, iocb->ki_pos, end);
-		if (ret)
-			goto out_unlock;
-
-		/*
-		 * Invalidate whole pages. This can return an error if we fail
-		 * to invalidate a page, but this should never happen on XFS.
-		 * Warn if it does fail.
-		 */
-		ret = invalidate_inode_pages2_range(mapping,
-				iocb->ki_pos >> PAGE_SHIFT, end >> PAGE_SHIFT);
-		WARN_ON_ONCE(ret);
-		ret = 0;
-	}
-
-	data = *to;
-	ret = __blockdev_direct_IO(iocb, inode, target->bt_bdev, &data,
-			xfs_get_blocks_direct, NULL, NULL, 0);
-	if (ret >= 0) {
-		iocb->ki_pos += ret;
-		iov_iter_advance(to, ret);
-	}
-
-out_unlock:
+	ret = iomap_dio_rw(iocb, to, &xfs_iomap_ops, NULL);
 	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+
 	return ret;
 }
 
@@ -465,6 +424,58 @@ restart:
 	return 0;
 }
 
+static int
+xfs_dio_write_end_io(
+	struct kiocb		*iocb,
+	ssize_t			size,
+	unsigned		flags)
+{
+	struct inode		*inode = file_inode(iocb->ki_filp);
+	struct xfs_inode	*ip = XFS_I(inode);
+	loff_t			offset = iocb->ki_pos;
+	bool			update_size = false;
+	int			error = 0;
+
+	trace_xfs_end_io_direct_write(ip, offset, size);
+
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
+		return -EIO;
+
+	if (size <= 0)
+		return size;
+
+	/*
+	 * We need to update the in-core inode size here so that we don't end up
+	 * with the on-disk inode size being outside the in-core inode size. We
+	 * have no other method of updating EOF for AIO, so always do it here
+	 * if necessary.
+	 *
+	 * We need to lock the test/set EOF update as we can be racing with
+	 * other IO completions here to update the EOF. Failing to serialise
+	 * here can result in EOF moving backwards and Bad Things Happen when
+	 * that occurs.
+	 */
+	spin_lock(&ip->i_flags_lock);
+	if (offset + size > i_size_read(inode)) {
+		i_size_write(inode, offset + size);
+		update_size = true;
+	}
+	spin_unlock(&ip->i_flags_lock);
+
+	if (flags & IOMAP_DIO_COW) {
+		error = xfs_reflink_end_cow(ip, offset, size);
+		if (error)
+			return error;
+	}
+
+	if (flags & IOMAP_DIO_UNWRITTEN)
+		error = xfs_iomap_write_unwritten(ip, offset, size);
+	else if (update_size)
+		error = xfs_setfilesize(ip, offset, size);
+
+	return error;
+}
+
 /*
  * xfs_file_dio_aio_write - handle direct IO writes
  *
@@ -504,9 +515,7 @@ xfs_file_dio_aio_write(
 	int			unaligned_io = 0;
 	int			iolock;
 	size_t			count = iov_iter_count(from);
-	loff_t			end;
-	struct iov_iter		data;
-	struct xfs_buftarg	*target = XFS_IS_REALTIME_INODE(ip) ?
+	struct xfs_buftarg      *target = XFS_IS_REALTIME_INODE(ip) ?
 					mp->m_rtdev_targp : mp->m_ddev_targp;
 
 	/* DIO must be aligned to device logical sector size */
@@ -534,23 +543,6 @@ xfs_file_dio_aio_write(
 	if (ret)
 		goto out;
 	count = iov_iter_count(from);
-	end = iocb->ki_pos + count - 1;
-
-	if (mapping->nrpages) {
-		ret = filemap_write_and_wait_range(mapping, iocb->ki_pos, end);
-		if (ret)
-			goto out;
-
-		/*
-		 * Invalidate whole pages. This can return an error if we fail
-		 * to invalidate a page, but this should never happen on XFS.
-		 * Warn if it does fail.
-		 */
-		ret = invalidate_inode_pages2_range(mapping,
-				iocb->ki_pos >> PAGE_SHIFT, end >> PAGE_SHIFT);
-		WARN_ON_ONCE(ret);
-		ret = 0;
-	}
 
 	/*
 	 * If we are doing unaligned IO, wait for all other IO to drain,
@@ -573,22 +565,7 @@ xfs_file_dio_aio_write(
 			goto out;
 	}
 
-	data = *from;
-	ret = __blockdev_direct_IO(iocb, inode, target->bt_bdev, &data,
-			xfs_get_blocks_direct, xfs_end_io_direct_write,
-			NULL, DIO_ASYNC_EXTEND);
-
-	/* see generic_file_direct_write() for why this is necessary */
-	if (mapping->nrpages) {
-		invalidate_inode_pages2_range(mapping,
-					      iocb->ki_pos >> PAGE_SHIFT,
-					      end >> PAGE_SHIFT);
-	}
-
-	if (ret > 0) {
-		iocb->ki_pos += ret;
-		iov_iter_advance(from, ret);
-	}
+	ret = iomap_dio_rw(iocb, from, &xfs_iomap_ops, xfs_dio_write_end_io);
 out:
 	xfs_iunlock(ip, iolock);
 
@@ -1468,15 +1445,9 @@ xfs_filemap_fault(
 		return xfs_filemap_page_mkwrite(vma, vmf);
 
 	xfs_ilock(XFS_I(inode), XFS_MMAPLOCK_SHARED);
-	if (IS_DAX(inode)) {
-		/*
-		 * we do not want to trigger unwritten extent conversion on read
-		 * faults - that is unnecessary overhead and would also require
-		 * changes to xfs_get_blocks_direct() to map unwritten extent
-		 * ioend for conversion on read-only mappings.
-		 */
+	if (IS_DAX(inode))
 		ret = dax_iomap_fault(vma, vmf, &xfs_iomap_ops);
-	} else
+	else
 		ret = filemap_fault(vma, vmf);
 	xfs_iunlock(XFS_I(inode), XFS_MMAPLOCK_SHARED);