libceph: validate con->state at the top of try_write()
authorIlya Dryomov <idryomov@gmail.com>
Tue, 24 Apr 2018 17:10:55 +0000 (19:10 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Tue, 1 May 2018 22:13:09 +0000 (15:13 -0700)
commit30c85553d822c2d882ed5fe1652be5acdf405509
tree3ced39ac04023644e55882fccfc5e1cd5cec3d2b
parent6044c69297adb224e4a2ec1bc8680b2e51428cc9
libceph: validate con->state at the top of try_write()

commit 9c55ad1c214d9f8c4594ac2c3fa392c1c32431a7 upstream.

ceph_con_workfn() validates con->state before calling try_read() and
then try_write().  However, try_read() temporarily releases con->mutex,
notably in process_message() and ceph_con_in_msg_alloc(), opening the
window for ceph_con_close() to sneak in, close the connection and
release con->sock.  When try_write() is called on the assumption that
con->state is still valid (i.e. not STANDBY or CLOSED), a NULL sock
gets passed to the networking stack:

  BUG: unable to handle kernel NULL pointer dereference at 0000000000000020
  IP: selinux_socket_sendmsg+0x5/0x20

Make sure con->state is valid at the top of try_write() and add an
explicit BUG_ON for this, similar to try_read().

Cc: stable@vger.kernel.org
Link: https://tracker.ceph.com/issues/23706
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Jason Dillaman <dillaman@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net/ceph/messenger.c