This patch by Kaigai Kohei fixes a bug in the SELinux sidtab code, where we
do a spin_unlock_irq() while nested under another irq lock, which enables
interrupts and allows a deadlock to happen:

  sidtab_set() is called between POLICY_WRLOCK and POLICY_WRUNLOCK in
  services.c:1092.  sidtab_set() uses SIDTAB_LOCK()/SIDTAB_UNLOCK(), but
  SIDTAB_UNLOCK() enables any interruptions because it's defined as
  spin_unlock_irq().  If an interruption occurs between SIDTAB_UNLOCK() and
  POLICY_WRUNLOCK, and interruption context try to hold the POLICY_RDLOCK,
  then a deadlock happen in the result.

The solution is to save & restore flags on the inner lock, per the patch
below.
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Stephen Smalley <sds@epoch.ncsc.mil>
Signed-off-by: Kaigai Kohei <kaigai@ak.jp.nec.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>