A set of scheduler updates:

- Prevent a NULL pointer dereference in the migration_stop_cpu()
     mechanims
 
   - Prevent self concurrency of affine_move_task()
 
   - Small fixes and cleanups related to task migration/affinity setting
 
   - Ensure that sync_runqueues_membarrier_state() is invoked on the current
     CPU when it is in the cpu mask
 -----BEGIN PGP SIGNATURE-----
 
 iQJHBAABCgAxFiEEQp8+kY+LLUocC4bMphj1TA10mKEFAmBOLHQTHHRnbHhAbGlu
 dXRyb25peC5kZQAKCRCmGPVMDXSYoWqTEACQLldMda63sEBPzEh4s0y+s4BqUsUM
 Pn5wVK1J91PZg1ofv4vLjIzKfjNuIbNNTswhux9kfb29LO0/KBd9BTYi442q4A+P
 chMi0Amfp4AGYlwo5+RNwEFNDFr33TD2Ax83cJ6FIDlJzLj8DRfzyxtwBvXfBG5R
 EZLTtKL30g20Y8N3nmQjvCInGvh0J1igr4lWXKtmvist7Ie3hW5jpvc8hF+VI0f+
 C1JfHg6GRw2eSCVFaF9EEeqX8+Wce+MrWIjwwB363vIX82lc/XC2XVbvrsgpA2P8
 sJaZz4KsOcXJLg9DWcN/OrpiMsgjnpKdMMsa3H2Uza8V3URtshpacb0wBWUpa1IA
 R55oCv4aRst6hNcCW1ayOLSEOcR2A2qAW2/ktiWYDqerIqkSCezMktunmrOc/vrW
 tmnEjlkYf0TNV54XREQ0Hr6OEnSIxqc9WrjbHUFbpv50YURqOCaHr19L0aOsemMJ
 g1pJCNkQhv4gZSenM6Fgo5ucbWB2Nvzu/Y6g7B2VFcpa3K7fmRJZW2uU5FvhwbeQ
 3ngvEwxMf3Rb6D7SpJyU41TYV9SqdOmoO4/UAFJ8YOlKp8biHCPmGh4+/QYza4Ky
 BIfPKtpr7MnSuYayo0wYYcKG0nE+rRJrj0Y0MAtz+6SfRCEc5Vd0NfIPQeAxDTvp
 oAITUrOuePiBrA==
 =WSUq
 -----END PGP SIGNATURE-----

Merge tag 'sched-urgent-2021-03-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull scheduler fixes from Thomas Gleixner:
 "A set of scheduler updates:

   - Prevent a NULL pointer dereference in the migration_stop_cpu()
     mechanims

   - Prevent self concurrency of affine_move_task()

   - Small fixes and cleanups related to task migration/affinity setting

   - Ensure that sync_runqueues_membarrier_state() is invoked on the
     current CPU when it is in the cpu mask"

* tag 'sched-urgent-2021-03-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched/membarrier: fix missing local execution of ipi_sync_rq_state()
  sched: Simplify set_affinity_pending refcounts
  sched: Fix affine_move_task() self-concurrency
  sched: Optimize migration_cpu_stop()
  sched: Collate affine_move_task() stoppers
  sched: Simplify migration_cpu_stop()
  sched: Fix migration_cpu_stop() requeueing

kernel/sched/core.c

@@ -1862,8 +1862,13 @@ struct migration_arg {
 	struct set_affinity_pending	*pending;
 };
 
+/*
+ * @refs: number of wait_for_completion()
+ * @stop_pending: is @stop_work in use
+ */
 struct set_affinity_pending {
 	refcount_t		refs;
+	unsigned int		stop_pending;
 	struct completion	done;
 	struct cpu_stop_work	stop_work;
 	struct migration_arg	arg;
@@ -1898,8 +1903,8 @@ static struct rq *__migrate_task(struct rq *rq, struct rq_flags *rf,
  */
 static int migration_cpu_stop(void *data)
 {
-	struct set_affinity_pending *pending;
 	struct migration_arg *arg = data;
+	struct set_affinity_pending *pending = arg->pending;
 	struct task_struct *p = arg->task;
 	int dest_cpu = arg->dest_cpu;
 	struct rq *rq = this_rq();
@@ -1921,7 +1926,6 @@ static int migration_cpu_stop(void *data)
 	raw_spin_lock(&p->pi_lock);
 	rq_lock(rq, &rf);
 
-	pending = p->migration_pending;
 	/*
 	 * If task_rq(p) != rq, it cannot be migrated here, because we're
 	 * holding rq->lock, if p->on_rq == 0 it cannot get enqueued because
@@ -1932,21 +1936,14 @@ static int migration_cpu_stop(void *data)
 			goto out;
 
 		if (pending) {
-			p->migration_pending = NULL;
+			if (p->migration_pending == pending)
+				p->migration_pending = NULL;
 			complete = true;
 		}
 
-		/* migrate_enable() --  we must not race against SCA */
 		if (dest_cpu < 0) {
-			/*
+			if (cpumask_test_cpu(task_cpu(p), &p->cpus_mask))
-			 * When this was migrate_enable() but we no longer
-			 * have a @pending, a concurrent SCA 'fixed' things
-			 * and we should be valid again. Nothing to do.
-			 */
-			if (!pending) {
-				WARN_ON_ONCE(!cpumask_test_cpu(task_cpu(p), &p->cpus_mask));
 				goto out;
-			}
 
 			dest_cpu = cpumask_any_distribute(&p->cpus_mask);
 		}
@@ -1956,7 +1953,14 @@ static int migration_cpu_stop(void *data)
 		else
 			p->wake_cpu = dest_cpu;
 
-	} else if (dest_cpu < 0 || pending) {
+		/*
+		 * XXX __migrate_task() can fail, at which point we might end
+		 * up running on a dodgy CPU, AFAICT this can only happen
+		 * during CPU hotplug, at which point we'll get pushed out
+		 * anyway, so it's probably not a big deal.
+		 */
+
+	} else if (pending) {
 		/*
 		 * This happens when we get migrated between migrate_enable()'s
 		 * preempt_enable() and scheduling the stopper task. At that
@@ -1971,43 +1975,32 @@ static int migration_cpu_stop(void *data)
 		 * ->pi_lock, so the allowed mask is stable - if it got
 		 * somewhere allowed, we're done.
 		 */
-		if (pending && cpumask_test_cpu(task_cpu(p), p->cpus_ptr)) {
+		if (cpumask_test_cpu(task_cpu(p), p->cpus_ptr)) {
-			p->migration_pending = NULL;
+			if (p->migration_pending == pending)
+				p->migration_pending = NULL;
 			complete = true;
 			goto out;
 		}
 
-		/*
-		 * When this was migrate_enable() but we no longer have an
-		 * @pending, a concurrent SCA 'fixed' things and we should be
-		 * valid again. Nothing to do.
-		 */
-		if (!pending) {
-			WARN_ON_ONCE(!cpumask_test_cpu(task_cpu(p), &p->cpus_mask));
-			goto out;
-		}
-
 		/*
 		 * When migrate_enable() hits a rq mis-match we can't reliably
 		 * determine is_migration_disabled() and so have to chase after
 		 * it.
 		 */
+		WARN_ON_ONCE(!pending->stop_pending);
 		task_rq_unlock(rq, p, &rf);
 		stop_one_cpu_nowait(task_cpu(p), migration_cpu_stop,
 				    &pending->arg, &pending->stop_work);
 		return 0;
 	}
 out:
+	if (pending)
+		pending->stop_pending = false;
 	task_rq_unlock(rq, p, &rf);
 
 	if (complete)
 		complete_all(&pending->done);
 
-	/* For pending->{arg,stop_work} */
-	pending = arg->pending;
-	if (pending && refcount_dec_and_test(&pending->refs))
-		wake_up_var(&pending->refs);
-
 	return 0;
 }
 
@@ -2194,11 +2187,7 @@ static int affine_move_task(struct rq *rq, struct task_struct *p, struct rq_flag
 			    int dest_cpu, unsigned int flags)
 {
 	struct set_affinity_pending my_pending = { }, *pending = NULL;
-	struct migration_arg arg = {
+	bool stop_pending, complete = false;
-		.task = p,
-		.dest_cpu = dest_cpu,
-	};
-	bool complete = false;
 
 	/* Can the task run on the task's current CPU? If so, we're done */
 	if (cpumask_test_cpu(task_cpu(p), &p->cpus_mask)) {
@@ -2210,12 +2199,16 @@ static int affine_move_task(struct rq *rq, struct task_struct *p, struct rq_flag
 			push_task = get_task_struct(p);
 		}
 
+		/*
+		 * If there are pending waiters, but no pending stop_work,
+		 * then complete now.
+		 */
 		pending = p->migration_pending;
-		if (pending) {
+		if (pending && !pending->stop_pending) {
-			refcount_inc(&pending->refs);
 			p->migration_pending = NULL;
 			complete = true;
 		}
+
 		task_rq_unlock(rq, p, rf);
 
 		if (push_task) {
@@ -2224,7 +2217,7 @@ static int affine_move_task(struct rq *rq, struct task_struct *p, struct rq_flag
 		}
 
 		if (complete)
-			goto do_complete;
+			complete_all(&pending->done);
 
 		return 0;
 	}
@@ -2235,6 +2228,12 @@ static int affine_move_task(struct rq *rq, struct task_struct *p, struct rq_flag
 			/* Install the request */
 			refcount_set(&my_pending.refs, 1);
 			init_completion(&my_pending.done);
+			my_pending.arg = (struct migration_arg) {
+				.task = p,
+				.dest_cpu = -1,		/* any */
+				.pending = &my_pending,
+			};
+
 			p->migration_pending = &my_pending;
 		} else {
 			pending = p->migration_pending;
@@ -2259,45 +2258,41 @@ static int affine_move_task(struct rq *rq, struct task_struct *p, struct rq_flag
 		return -EINVAL;
 	}
 
-	if (flags & SCA_MIGRATE_ENABLE) {
-
-		refcount_inc(&pending->refs); /* pending->{arg,stop_work} */
-		p->migration_flags &= ~MDF_PUSH;
-		task_rq_unlock(rq, p, rf);
-
-		pending->arg = (struct migration_arg) {
-			.task = p,
-			.dest_cpu = -1,
-			.pending = pending,
-		};
-
-		stop_one_cpu_nowait(cpu_of(rq), migration_cpu_stop,
-				    &pending->arg, &pending->stop_work);
-
-		return 0;
-	}
-
 	if (task_running(rq, p) || p->state == TASK_WAKING) {
 		/*
-		 * Lessen races (and headaches) by delegating
+		 * MIGRATE_ENABLE gets here because 'p == current', but for
-		 * is_migration_disabled(p) checks to the stopper, which will
+		 * anything else we cannot do is_migration_disabled(), punt
-		 * run on the same CPU as said p.
+		 * and have the stopper function handle it all race-free.
 		 */
-		task_rq_unlock(rq, p, rf);
+		stop_pending = pending->stop_pending;
-		stop_one_cpu(cpu_of(rq), migration_cpu_stop, &arg);
+		if (!stop_pending)
+			pending->stop_pending = true;
 
+		if (flags & SCA_MIGRATE_ENABLE)
+			p->migration_flags &= ~MDF_PUSH;
+
+		task_rq_unlock(rq, p, rf);
+
+		if (!stop_pending) {
+			stop_one_cpu_nowait(cpu_of(rq), migration_cpu_stop,
+					    &pending->arg, &pending->stop_work);
+		}
+
+		if (flags & SCA_MIGRATE_ENABLE)
+			return 0;
 	} else {
 
 		if (!is_migration_disabled(p)) {
 			if (task_on_rq_queued(p))
 				rq = move_queued_task(rq, rf, p, dest_cpu);
 
-			p->migration_pending = NULL;
+			if (!pending->stop_pending) {
-			complete = true;
+				p->migration_pending = NULL;
+				complete = true;
+			}
 		}
 		task_rq_unlock(rq, p, rf);
 
-do_complete:
 		if (complete)
 			complete_all(&pending->done);
 	}
@@ -2305,7 +2300,7 @@ do_complete:
 	wait_for_completion(&pending->done);
 
 	if (refcount_dec_and_test(&pending->refs))
-		wake_up_var(&pending->refs);
+		wake_up_var(&pending->refs); /* No UaF, just an address */
 
 	/*
 	 * Block the original owner of &pending until all subsequent callers
@@ -2313,6 +2308,9 @@ do_complete:
 	 */
 	wait_var_event(&my_pending.refs, !refcount_read(&my_pending.refs));
 
+	/* ARGH */
+	WARN_ON_ONCE(my_pending.stop_pending);
+
 	return 0;
 }
 

kernel/sched/membarrier.c

@@ -471,9 +471,7 @@ static int sync_runqueues_membarrier_state(struct mm_struct *mm)
 	}
 	rcu_read_unlock();
 
-	preempt_disable();
+	on_each_cpu_mask(tmpmask, ipi_sync_rq_state, mm, true);
-	smp_call_function_many(tmpmask, ipi_sync_rq_state, mm, 1);
-	preempt_enable();
 
 	free_cpumask_var(tmpmask);
 	cpus_read_unlock();