kernfs: fix regression in kernfs_fop_write caused by wrong type
authorIvan Vecera <ivecera@redhat.com>
Fri, 19 Jan 2018 08:18:54 +0000 (09:18 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 16 Feb 2018 19:22:59 +0000 (20:22 +0100)
commit650d3d8512c6c542d7c11ece049b46d5f2337ce8
tree3df33c5ce7fd40b83b168fcf446605e64e4e4f6a
parentf1e31607e73d5f4546ff8d1e40e1153230392188
kernfs: fix regression in kernfs_fop_write caused by wrong type

commit ba87977a49913129962af8ac35b0e13e0fa4382d upstream.

Commit b7ce40cff0b9 ("kernfs: cache atomic_write_len in
kernfs_open_file") changes type of local variable 'len' from ssize_t
to size_t. This change caused that the *ppos value is updated also
when the previous write callback failed.

Mentioned snippet:
...
len = ops->write(...); <- return value can be negative
...
if (len > 0)           <- true here in this case
        *ppos += len;
...

Fixes: b7ce40cff0b9 ("kernfs: cache atomic_write_len in kernfs_open_file")
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Ivan Vecera <ivecera@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/kernfs/file.c