[PATCH] ceph: truncate page cache when doing DIO in encrypted inodes

["Luís Henriques" <lhenriques@suse.de>]

When doing DIO on an encrypted node, we need to truncate the page cache in
the range being written to, otherwise the cache will include invalid data.

Signed-off-by: Luís Henriques <lhenriques@suse.de>
---
 fs/ceph/file.c | 5 +++++
 1 file changed, 5 insertions(+)

This patch should fix generic/647 fstest when run with test_dummy_encryption.

diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 5072570c2203..0f31c4d352a4 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -1895,6 +1895,11 @@ ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos,
 		req->r_inode = inode;
 		req->r_mtime = mtime;
 
+		if (IS_ENCRYPTED(inode) && (iocb->ki_flags & IOCB_DIRECT))
+			truncate_inode_pages_range(
+				inode->i_mapping, write_pos,
+				PAGE_ALIGN(write_pos + write_len) - 1);
+
 		/* Set up the assertion */
 		if (rmw) {
 			/*