diff --git a/[refs] b/[refs]
index 819ede1627d0..df652938b6b3 100644
--- a/[refs]
+++ b/[refs]
@@ -1,2 +1,2 @@
 ---
-refs/heads/master: e780c452cfea10cbfae1c96dc16e506758138b6f
+refs/heads/master: 08928e7aea930e6822ce8f1b20068bf857ecf20d
diff --git a/trunk/arch/arm/include/asm/mutex.h b/trunk/arch/arm/include/asm/mutex.h
index b1479fd04a95..87c044910fe0 100644
--- a/trunk/arch/arm/include/asm/mutex.h
+++ b/trunk/arch/arm/include/asm/mutex.h
@@ -9,8 +9,13 @@
 #define _ASM_MUTEX_H
 /*
  * On pre-ARMv6 hardware this results in a swp-based implementation,
- * which is the most efficient. For ARMv6+, we emit a pair of exclusive
- * accesses instead.
+ * which is the most efficient. For ARMv6+, we have exclusive memory
+ * accessors and use atomic_dec to avoid the extra xchg operations
+ * on the locking slowpaths.
  */
+#if __LINUX_ARM_ARCH__ < 6
 #include <asm-generic/mutex-xchg.h>
+#else
+#include <asm-generic/mutex-dec.h>
 #endif
+#endif	/* _ASM_MUTEX_H */