vfs: Test for and handle paths that are unreachable from their mnt_root
authorEric W. Biederman <ebiederm@xmission.com>
Sun, 16 Aug 2015 01:27:13 +0000 (20:27 -0500)
committerBen Hutchings <ben@decadent.org.uk>
Tue, 13 Oct 2015 02:46:04 +0000 (03:46 +0100)
commit6d84ade2c8242ab83fbc5bacb66eb81a8d1ca6db
tree406e247bf5400b4fa331ab6fe9f7e668c880698b
parent722632af3c2b4828e79f143e356489c6761035ec
vfs: Test for and handle paths that are unreachable from their mnt_root

commit 397d425dc26da728396e66d392d5dcb8dac30c37 upstream.

In rare cases a directory can be renamed out from under a bind mount.
In those cases without special handling it becomes possible to walk up
the directory tree to the root dentry of the filesystem and down
from the root dentry to every other file or directory on the filesystem.

Like division by zero .. from an unconnected path can not be given
a useful semantic as there is no predicting at which path component
the code will realize it is unconnected.  We certainly can not match
the current behavior as the current behavior is a security hole.

Therefore when encounting .. when following an unconnected path
return -ENOENT.

- Add a function path_connected to verify path->dentry is reachable
  from path->mnt.mnt_root.  AKA to validate that rename did not do
  something nasty to the bind mount.

  To avoid races path_connected must be called after following a path
  component to it's next path component.

Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
fs/namei.c