sched: Handle priority boosted tasks proper in setscheduler()

Ronny reported that the following scenario is not handled correctly:

T1 (prio = 10)
lock(rtmutex);

T2 (prio = 20)
lock(rtmutex)
boost T1

T1 (prio = 20)
sys_set_scheduler(prio = 30)
T1 prio = 30
....
sys_set_scheduler(prio = 10)
T1 prio = 30

The last step is wrong as T1 should now be back at prio 20.

Commit c365c292d059 ("sched: Consider pi boosting in setscheduler()")
only handles the case where a boosted tasks tries to lower its
priority.

Fix it by taking the new effective priority into account for the
decision whether a change of the priority is required.

Reported-by: Ronny Meeus <[e-mail]>
Tested-by: Steven Rostedt <[e-mail]>
Signed-off-by: Thomas Gleixner <[e-mail]>
Signed-off-by: Peter Zijlstra (Intel) <[e-mail]>
Reviewed-by: Steven Rostedt <[e-mail]>
Cc: <[e-mail]>
Cc: Borislav Petkov <[e-mail]>
Cc: H. Peter Anvin <[e-mail]>
Cc: Mike Galbraith <[e-mail]>
Fixes: c365c292d059 ("sched: Consider pi boosting in setscheduler()")
Link: http://lkml.kernel.org/r/alpine.DEB.2.11.1505051806060.4225@nanos
Signed-off-by: Ingo Molnar <[e-mail]>

include/linux/sched/rt.h

···
 #ifdef CONFIG_RT_MUTEXES
 extern int rt_mutex_getprio(struct task_struct *p);
 extern void rt_mutex_setprio(struct task_struct *p, int prio);
-extern int rt_mutex_check_prio(struct task_struct *task, int newprio);
+extern int rt_mutex_get_effective_prio(struct task_struct *task, int newprio);
 extern struct task_struct *rt_mutex_get_top_task(struct task_struct *task);
 extern void rt_mutex_adjust_pi(struct task_struct *p);
 static inline bool tsk_is_pi_blocked(struct task_struct *tsk)
···
 	return p->normal_prio;
 }
 
-static inline int rt_mutex_check_prio(struct task_struct *task, int newprio)
+static inline int rt_mutex_get_effective_prio(struct task_struct *task,
+					      int newprio)
 {
-	return 0;
+	return newprio;
 }
 
 static inline struct task_struct *rt_mutex_get_top_task(struct task_struct *task)

kernel/locking/rtmutex.c

···
 }
 
 /*
- * Called by sched_setscheduler() to check whether the priority change
- * is overruled by a possible priority boosting.
+ * Called by sched_setscheduler() to get the priority which will be
+ * effective after the change.
  */
-int rt_mutex_check_prio(struct task_struct *task, int newprio)
+int rt_mutex_get_effective_prio(struct task_struct *task, int newprio)
 {
 	if (!task_has_pi_waiters(task))
-		return 0;
+		return newprio;
 
-	return task_top_pi_waiter(task)->task->prio <= newprio;
+	if (task_top_pi_waiter(task)->task->prio <= newprio)
+		return task_top_pi_waiter(task)->task->prio;
+	return newprio;
 }
 
 /*

kernel/sched/core.c

···
 
 /* Actually do priority change: must hold pi & rq lock. */
 static void __setscheduler(struct rq *rq, struct task_struct *p,
-			   const struct sched_attr *attr)
+			   const struct sched_attr *attr, bool keep_boost)
 {
 	__setscheduler_params(p, attr);
 
 	/*
-	 * If we get here, there was no pi waiters boosting the
-	 * task. It is safe to use the normal prio.
+	 * Keep a potential priority boosting if called from
+	 * sched_setscheduler().
 	 */
-	p->prio = normal_prio(p);
+	if (keep_boost)
+		p->prio = rt_mutex_get_effective_prio(p, normal_prio(p));
+	else
+		p->prio = normal_prio(p);
 
 	if (dl_prio(p->prio))
 		p->sched_class = &dl_sched_class;
···
 	int newprio = dl_policy(attr->sched_policy) ? MAX_DL_PRIO - 1 :
 		      MAX_RT_PRIO - 1 - attr->sched_priority;
 	int retval, oldprio, oldpolicy = -1, queued, running;
-	int policy = attr->sched_policy;
+	int new_effective_prio, policy = attr->sched_policy;
 	unsigned long flags;
 	const struct sched_class *prev_class;
 	struct rq *rq;
···
 	oldprio = p->prio;
 
 	/*
-	 * Special case for priority boosted tasks.
-	 *
-	 * If the new priority is lower or equal (user space view)
-	 * than the current (boosted) priority, we just store the new
+	 * Take priority boosted tasks into account. If the new
+	 * effective priority is unchanged, we just store the new
 	 * normal parameters and do not touch the scheduler class and
 	 * the runqueue. This will be done when the task deboost
 	 * itself.
 	 */
-	if (rt_mutex_check_prio(p, newprio)) {
+	new_effective_prio = rt_mutex_get_effective_prio(p, newprio);
+	if (new_effective_prio == oldprio) {
 		__setscheduler_params(p, attr);
 		task_rq_unlock(rq, p, &flags);
 		return 0;
···
 		put_prev_task(rq, p);
 
 	prev_class = p->sched_class;
-	__setscheduler(rq, p, attr);
+	__setscheduler(rq, p, attr, true);
 
 	if (running)
 		p->sched_class->set_curr_task(rq);
···
 	queued = task_on_rq_queued(p);
 	if (queued)
 		dequeue_task(rq, p, 0);
-	__setscheduler(rq, p, &attr);
+	__setscheduler(rq, p, &attr, false);
 	if (queued) {
 		enqueue_task(rq, p, 0);
 		resched_curr(rq);