x86, kaslr: fix module lock ordering problem
authorKees Cook <keescook@chromium.org>
Mon, 10 Mar 2014 20:42:48 +0000 (13:42 -0700)
committerH. Peter Anvin <hpa@linux.intel.com>
Mon, 24 Mar 2014 17:18:26 +0000 (10:18 -0700)
commit9dd721c6dbfc310f94306902611f86dda87a45fa
tree56dbfcb8ae3c05f26be75dba824c5fb18b1ac25f
parente2b32e6785138d92d2a40e0d0473575c8c7310a2
x86, kaslr: fix module lock ordering problem

There was a potential lock ordering problem with the module kASLR patch
("x86, kaslr: randomize module base load address"). This patch removes
the usage of the module_mutex and creates a new mutex to protect the
module base address offset value.

Chain exists of:
  text_mutex --> kprobe_insn_slots.mutex --> module_mutex

[    0.515561]  Possible unsafe locking scenario:
[    0.515561]
[    0.515561]        CPU0                    CPU1
[    0.515561]        ----                    ----
[    0.515561]   lock(module_mutex);
[    0.515561]                                lock(kprobe_insn_slots.mutex);
[    0.515561]                                lock(module_mutex);
[    0.515561]   lock(text_mutex);
[    0.515561]
[    0.515561]  *** DEADLOCK ***

Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Andy Honig <ahonig@google.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
arch/x86/kernel/module.c