sched: align rq to cacheline boundary

Align the per cpu runqueue to the cacheline boundary. This will minimize
the number of cachelines touched during remote wakeup.

Signed-off-by: Suresh Siddha <[e-mail]>
Acked-by: Ingo Molnar <[e-mail]>
Cc: Ravikiran G Thirumalai <[e-mail]>
Cc: Nick Piggin <[e-mail]>
Signed-off-by: Andrew Morton <[e-mail]>
Signed-off-by: Linus Torvalds <[e-mail]>

kernel/sched.c

···
 	struct lock_class_key rq_lock_key;
 };
 
-static DEFINE_PER_CPU(struct rq, runqueues);
+static DEFINE_PER_CPU(struct rq, runqueues) ____cacheline_aligned_in_smp;
 
 static inline int cpu_of(struct rq *rq)
 {