nilfs2: fix oops due to inconsistent state in page with discrete b-tree nodes

Andrea Gelmini gave me a report that a kernel oops hit on a nilfs
filesystem with a 1KB block size when doing rsync.

This turned out to be caused by an inconsistency of dirty state
between a page and its buffers storing b-tree node blocks.

If the page had multiple buffers split over multiple logs, and if the
logs were written at a time, a dirty flag remained in the page even
every dirty flag in the buffers was cleared.

This will fix the failure by dropping the dirty flag properly for
pages with the discrete multiple b-tree nodes.

Reported-by: Andrea Gelmini <andrea.gelmini@gmail.com>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Tested-by: Andrea Gelmini <andrea.gelmini@gmail.com>
Cc: stable@kernel.org

1 files changed, 15 insertions, 1 deletions

diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
index 8b5e4778cf28..51ff3d0a4ee2 100644
--- a/fs/nilfs2/segment.c
+++ b/fs/nilfs2/segment.c
@@ -1859,12 +1859,26 @@ static void nilfs_end_page_io(struct page *page, int err)
 	if (!page)
 		return;
 
-	if (buffer_nilfs_node(page_buffers(page)) && !PageWriteback(page))
+	if (buffer_nilfs_node(page_buffers(page)) && !PageWriteback(page)) {
 		/*
 		 * For b-tree node pages, this function may be called twice
 		 * or more because they might be split in a segment.
 		 */
+		if (PageDirty(page)) {
+			/*
+			 * For pages holding split b-tree node buffers, dirty
+			 * flag on the buffers may be cleared discretely.
+			 * In that case, the page is once redirtied for
+			 * remaining buffers, and it must be cancelled if
+			 * all the buffers get cleaned later.
+			 */
+			lock_page(page);
+			if (nilfs_page_buffers_clean(page))
+				__nilfs_clear_page_dirty(page);
+			unlock_page(page);
+		}
 		return;
+	}
 
 	__nilfs_end_page_io(page, err);
 }