lock debug: sit tight when we are already in a panic
authorAndrew Morton <akpm@linux-foundation.org>
Fri, 12 Sep 2008 00:02:58 +0000 (17:02 -0700)
committerIngo Molnar <mingo@elte.hu>
Fri, 12 Sep 2008 09:11:21 +0000 (11:11 +0200)
in:

  > http://bugzilla.kernel.org/show_bug.cgi?id=11543

The panic code called the kexec code which called mutex_trylock() which
called spin_lock_mutex() which then stupidly went and blurted a load of
debug stuff because of in_interrupt().

Keep the lock debug code from escallating an already crappy situation.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
include/linux/debug_locks.h

index 4aaa4afb1cb99f849bb345a8189153e65e94cbcd..096476f1fb356a2c17ab5d267be9a2295f55295a 100644 (file)
@@ -17,7 +17,7 @@ extern int debug_locks_off(void);
 ({                                                                     \
        int __ret = 0;                                                  \
                                                                        \
-       if (unlikely(c)) {                                              \
+       if (!oops_in_progress && unlikely(c)) {                         \
                if (debug_locks_off() && !debug_locks_silent)           \
                        WARN_ON(1);                                     \
                __ret = 1;                                              \