cifs: don't instantiate new dentries in readdir for inodes that need to be revalidate...
authorJeff Layton <jlayton@redhat.com>
Wed, 7 Aug 2013 14:29:08 +0000 (10:29 -0400)
committerBen Hutchings <ben@decadent.org.uk>
Tue, 10 Sep 2013 00:57:20 +0000 (01:57 +0100)
commiteb4a22ba43d91d58a35bc4fce11d8e6065bb4e1d
tree9f65607dd47da322192c048cf19a35e213115088
parent4f440a7abb7c4b81192be3e192af8feec8e02ff3
cifs: don't instantiate new dentries in readdir for inodes that need to be revalidated immediately

commit 757c4f6260febff982276818bb946df89c1105aa upstream.

David reported that commit c2b93e06 (cifs: only set ops for inodes in
I_NEW state) caused a regression with mfsymlinks. Prior to that patch,
if a mfsymlink dentry was instantiated at readdir time, the inode would
get a new set of ops when it was revalidated. After that patch, this
did not occur.

This patch addresses this by simply skipping instantiating dentries in
the readdir codepath when we know that they will need to be immediately
revalidated. The next attempt to use that dentry will cause a new lookup
to occur (which is basically what we want to happen anyway).

Cc: "Stefan (metze) Metzmacher" <metze@samba.org>
Cc: Sachin Prabhu <sprabhu@redhat.com>
Reported-and-Tested-by: David McBride <dwm37@cam.ac.uk>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
[bwh: Backported to 3.2: need to return NULL]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
fs/cifs/readdir.c