ocfs2: fix zero out valid data
authorJunxiao Bi <junxiao.bi@oracle.com>
Thu, 29 Jul 2021 21:53:38 +0000 (14:53 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 4 Aug 2021 09:58:02 +0000 (11:58 +0200)
commit f267aeb6dea5e468793e5b8eb6a9c72c0020d418 upstream.

If append-dio feature is enabled, direct-io write and fallocate could
run in parallel to extend file size, fallocate used "orig_isize" to
record i_size before taking "ip_alloc_sem", when
ocfs2_zeroout_partial_cluster() zeroout EOF blocks, i_size maybe already
extended by ocfs2_dio_end_io_write(), that will cause valid data zeroed
out.

Link: https://lkml.kernel.org/r/20210722054923.24389-1-junxiao.bi@oracle.com
Fixes: 6bba4471f0cc ("ocfs2: fix data corruption by fallocate")
Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Jun Piao <piaojun@huawei.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/ocfs2/file.c

index 3aa4441f5ab50796706ce6923d86bbad283ae1a3..3cb0d9ef1cb2da6d6438a227f63348c7597e3cab 100644 (file)
@@ -1918,7 +1918,6 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode,
                goto out_inode_unlock;
        }
 
-       orig_isize = i_size_read(inode);
        switch (sr->l_whence) {
        case 0: /*SEEK_SET*/
                break;
@@ -1926,7 +1925,7 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode,
                sr->l_start += f_pos;
                break;
        case 2: /*SEEK_END*/
-               sr->l_start += orig_isize;
+               sr->l_start += i_size_read(inode);
                break;
        default:
                ret = -EINVAL;
@@ -1981,6 +1980,7 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode,
                ret = -EINVAL;
        }
 
+       orig_isize = i_size_read(inode);
        /* zeroout eof blocks in the cluster. */
        if (!ret && change_size && orig_isize < size) {
                ret = ocfs2_zeroout_partial_cluster(inode, orig_isize,