f2fs: fix to avoid race in between aio and gc

We won't wait DIO synchronously when doing AIO, so there will be potential
IO reorder in between AIO and GC, which will cause data corruption.

This patch adds inode_dio_wait to serialize aio and data GC to avoid this
issue.

Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c
index 57bea2182f3035a37bd49ed8e6e3fcde60c60ae8..cd147e7c71e8fcd642da013e4fdde24413b55fca 100644 (file)
--- a/fs/f2fs/gc.c
+++ b/fs/f2fs/gc.c
@@ -875,6 +875,9 @@ next_step:
                                        continue;
                                }
                                locked = true;
+
+                               /* wait for all inflight aio data */
+                               inode_dio_wait(inode);
                        }
 
                        start_bidx = start_bidx_of_node(nofs, inode)