bfq: Avoid use-after-free when moving processes between cgroups

index 2e0dd68a3cbe..6d122c28086e 100644
@@ -2895,9 +2895,12 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 				struct bfq_queue *new_bfqq =
 					bfq_setup_merge(bfqq, stable_merge_bfqq);
 
-				bic->stably_merged = true;
-				if (new_bfqq && new_bfqq->bic)
-					new_bfqq->bic->stably_merged = true;
+				if (new_bfqq) {
+					bic->stably_merged = true;
+					if (new_bfqq->bic)
+						new_bfqq->bic->stably_merged =
+									true;
+				}
 				return new_bfqq;
 			} else
 				return NULL;

index 6d122c28086e..7d00b21ebe5d 100644
@@ -2758,6 +2758,14 @@ bfq_setup_merge(struct bfq_queue *bfqq, struct bfq_queue *new_bfqq)
 	if (process_refs == 0 || new_process_refs == 0)
 		return NULL;
 
+	/*
+	 * Make sure merged queues belong to the same parent. Parents could
+	 * have changed since the time we decided the two queues are suitable
+	 * for merging.
+	 */
+	if (new_bfqq->entity.parent != bfqq->entity.parent)
+		return NULL;
+
 	bfq_log_bfqq(bfqq->bfqd, bfqq, "scheduling merge with queue %d",
 		new_bfqq->pid);
 

index 420eda2589c0..9352f3cc2377 100644
@@ -743,9 +743,39 @@ static struct bfq_group *__bfq_bic_change_cgroup(struct bfq_data *bfqd,
 	}
 
 	if (sync_bfqq) {
-		entity = &sync_bfqq->entity;
-		if (entity->sched_data != &bfqg->sched_data)
-			bfq_bfqq_move(bfqd, sync_bfqq, bfqg);
+		if (!sync_bfqq->new_bfqq && !bfq_bfqq_coop(sync_bfqq)) {
+			/* We are the only user of this bfqq, just move it */
+			if (sync_bfqq->entity.sched_data != &bfqg->sched_data)
+				bfq_bfqq_move(bfqd, sync_bfqq, bfqg);
+		} else {
+			struct bfq_queue *bfqq;
+
+			/*
+			 * The queue was merged to a different queue. Check
+			 * that the merge chain still belongs to the same
+			 * cgroup.
+			 */
+			for (bfqq = sync_bfqq; bfqq; bfqq = bfqq->new_bfqq)
+				if (bfqq->entity.sched_data !=
+				    &bfqg->sched_data)
+					break;
+			if (bfqq) {
+				/*
+				 * Some queue changed cgroup so the merge is
+				 * not valid anymore. We cannot easily just
+				 * cancel the merge (by clearing new_bfqq) as
+				 * there may be other processes using this
+				 * queue and holding refs to all queues below
+				 * sync_bfqq->new_bfqq. Similarly if the merge
+				 * already happened, we need to detach from
+				 * bfqq now so that we cannot merge bio to a
+				 * request from the old cgroup.
+				 */
+				bfq_put_cooperator(sync_bfqq);
+				bfq_release_process_ref(bfqd, sync_bfqq);
+				bic_set_bfqq(bic, NULL, 1);
+			}
+		}
 	}
 
 	return bfqg;

index 7d00b21ebe5d..89fe3f85eb3c 100644
@@ -5315,7 +5315,7 @@ static void bfq_put_stable_ref(struct bfq_queue *bfqq)
 	bfq_put_queue(bfqq);
 }
 
-static void bfq_put_cooperator(struct bfq_queue *bfqq)
+void bfq_put_cooperator(struct bfq_queue *bfqq)
 {
 	struct bfq_queue *__bfqq, *next;
 

index 3b83e3d1c2e5..a56763045d19 100644
@@ -979,6 +979,7 @@ void bfq_weights_tree_remove(struct bfq_data *bfqd,
 void bfq_bfqq_expire(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 		     bool compensate, enum bfqq_expiration reason);
 void bfq_put_queue(struct bfq_queue *bfqq);
+void bfq_put_cooperator(struct bfq_queue *bfqq);
 void bfq_end_wr_async_queues(struct bfq_data *bfqd, struct bfq_group *bfqg);
 void bfq_release_process_ref(struct bfq_data *bfqd, struct bfq_queue *bfqq);
 void bfq_schedule_dispatch(struct bfq_data *bfqd);

index 89fe3f85eb3c..1fc4d4628fba 100644
@@ -2457,10 +2457,17 @@ static bool bfq_bio_merge(struct request_queue *q, struct bio *bio,
 
 	spin_lock_irq(&bfqd->lock);
 
-	if (bic)
+	if (bic) {
+		/*
+		 * Make sure cgroup info is uptodate for current process before
+		 * considering the merge.
+		 */
+		bfq_bic_update_cgroup(bic, bio);
+
 		bfqd->bio_bfqq = bic_to_bfqq(bic, op_is_sync(bio->bi_opf));
-	else
+	} else {
 		bfqd->bio_bfqq = NULL;
+	}
 	bfqd->bio_bic = bic;
 
 	ret = blk_mq_sched_try_merge(q, bio, nr_segs, &free);

index 1fc4d4628fba..19082e14f3c1 100644
@@ -6150,11 +6150,8 @@ static void bfq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
 		return;
 	}
 
-	spin_unlock_irq(&bfqd->lock);
-
 	trace_block_rq_insert(rq);
 
-	spin_lock_irq(&bfqd->lock);
 	bfqq = bfq_init_rq(rq);
 	if (!bfqq || at_head) {
 		if (at_head)

index 19082e14f3c1..d7cf930b47bb 100644
@@ -2497,8 +2497,6 @@ static int bfq_request_merge(struct request_queue *q, struct request **req,
 	return ELEVATOR_NO_MERGE;
 }
 
-static struct bfq_queue *bfq_init_rq(struct request *rq);
-
 static void bfq_request_merged(struct request_queue *q, struct request *req,
 			       enum elv_merge type)
 {

@@ -2507,7 +2505,7 @@ static void bfq_request_merged(struct request_queue *q, struct request *req,
 	    blk_rq_pos(req) <
 	    blk_rq_pos(container_of(rb_prev(&req->rb_node),
 				    struct request, rb_node))) {
-		struct bfq_queue *bfqq = bfq_init_rq(req);
+		struct bfq_queue *bfqq = RQ_BFQQ(req);
 		struct bfq_data *bfqd;
 		struct request *prev, *next_rq;
 

@@ -2559,8 +2557,8 @@ static void bfq_request_merged(struct request_queue *q, struct request *req,
 static void bfq_requests_merged(struct request_queue *q, struct request *rq,
 				struct request *next)
 {
-	struct bfq_queue *bfqq = bfq_init_rq(rq),
-		*next_bfqq = bfq_init_rq(next);
+	struct bfq_queue *bfqq = RQ_BFQQ(rq),
+		*next_bfqq = RQ_BFQQ(next);
 
 	if (!bfqq)
 		goto remove;

@@ -6129,6 +6127,8 @@ static inline void bfq_update_insert_stats(struct request_queue *q,
 					   unsigned int cmd_flags) {}
 #endif /* CONFIG_BFQ_CGROUP_DEBUG */
 
+static struct bfq_queue *bfq_init_rq(struct request *rq);
+
 static void bfq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
 			       bool at_head)
 {

@@ -6144,6 +6144,7 @@ static void bfq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
 		bfqg_stats_update_legacy_io(q, rq);
 #endif
 	spin_lock_irq(&bfqd->lock);
+	bfqq = bfq_init_rq(rq);
 	if (blk_mq_sched_try_insert_merge(q, rq, &free)) {
 		spin_unlock_irq(&bfqd->lock);
 		blk_mq_free_requests(&free);

@@ -6152,7 +6153,6 @@ static void bfq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
 
 	trace_block_rq_insert(rq);
 
-	bfqq = bfq_init_rq(rq);
 	if (!bfqq || at_head) {
 		if (at_head)
 			list_add(&rq->queuelist, &bfqd->dispatch);

index 9352f3cc2377..879380c2bc7e 100644
@@ -557,6 +557,7 @@ static void bfq_pd_init(struct blkg_policy_data *pd)
 				   */
 	bfqg->bfqd = bfqd;
 	bfqg->active_entities = 0;
+	bfqg->online = true;
 	bfqg->rq_pos_tree = RB_ROOT;
 }
 

@@ -603,7 +604,6 @@ struct bfq_group *bfq_find_set_group(struct bfq_data *bfqd,
 	struct bfq_entity *entity;
 
 	bfqg = bfq_lookup_bfqg(bfqd, blkcg);
-
 	if (unlikely(!bfqg))
 		return NULL;
 

@@ -979,6 +979,7 @@ static void bfq_pd_offline(struct blkg_policy_data *pd)
 
 put_async_queues:
 	bfq_put_async_queues(bfqd, bfqg);
+	bfqg->online = false;
 
 	spin_unlock_irqrestore(&bfqd->lock, flags);
 	/*

index a56763045d19..4664e2f3e828 100644
@@ -928,6 +928,8 @@ struct bfq_group {
 
 	/* reference counter (see comments in bfq_bic_update_cgroup) */
 	int ref;
+	/* Is bfq_group still online? */
+	bool online;
 
 	struct bfq_entity entity;
 	struct bfq_sched_data sched_data;

index 879380c2bc7e..32d2c2a47480 100644
@@ -586,27 +586,11 @@ static void bfq_group_set_parent(struct bfq_group *bfqg,
 	entity->sched_data = &parent->sched_data;
 }
 
-static struct bfq_group *bfq_lookup_bfqg(struct bfq_data *bfqd,
-					 struct blkcg *blkcg)
+static void bfq_link_bfqg(struct bfq_data *bfqd, struct bfq_group *bfqg)
 {
-	struct blkcg_gq *blkg;
-
-	blkg = blkg_lookup(blkcg, bfqd->queue);
-	if (likely(blkg))
-		return blkg_to_bfqg(blkg);
-	return NULL;
-}
-
-struct bfq_group *bfq_find_set_group(struct bfq_data *bfqd,
-				     struct blkcg *blkcg)
-{
-	struct bfq_group *bfqg, *parent;
+	struct bfq_group *parent;
 	struct bfq_entity *entity;
 
-	bfqg = bfq_lookup_bfqg(bfqd, blkcg);
-	if (unlikely(!bfqg))
-		return NULL;
-
 	/*
 	 * Update chain of bfq_groups as we might be handling a leaf group
 	 * which, along with some of its relatives, has not been hooked yet

@@ -623,8 +607,15 @@ struct bfq_group *bfq_find_set_group(struct bfq_data *bfqd,
 			bfq_group_set_parent(curr_bfqg, parent);
 		}
 	}
+}
 
-	return bfqg;
+struct bfq_group *bfq_bio_bfqg(struct bfq_data *bfqd, struct bio *bio)
+{
+	struct blkcg_gq *blkg = bio->bi_blkg;
+
+	if (!blkg)
+		return bfqd->root_group;
+	return blkg_to_bfqg(blkg);
 }
 
 /**

@@ -714,25 +705,15 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_queue *bfqq,
  * Move bic to blkcg, assuming that bfqd->lock is held; which makes
  * sure that the reference to cgroup is valid across the call (see
  * comments in bfq_bic_update_cgroup on this issue)
- *
- * NOTE: an alternative approach might have been to store the current
- * cgroup in bfqq and getting a reference to it, reducing the lookup
- * time here, at the price of slightly more complex code.
  */
-static struct bfq_group *__bfq_bic_change_cgroup(struct bfq_data *bfqd,
-						struct bfq_io_cq *bic,
-						struct blkcg *blkcg)
+static void *__bfq_bic_change_cgroup(struct bfq_data *bfqd,
+				     struct bfq_io_cq *bic,
+				     struct bfq_group *bfqg)
 {
 	struct bfq_queue *async_bfqq = bic_to_bfqq(bic, 0);
 	struct bfq_queue *sync_bfqq = bic_to_bfqq(bic, 1);
-	struct bfq_group *bfqg;
 	struct bfq_entity *entity;
 
-	bfqg = bfq_find_set_group(bfqd, blkcg);
-
-	if (unlikely(!bfqg))
-		bfqg = bfqd->root_group;
-
 	if (async_bfqq) {
 		entity = &async_bfqq->entity;
 

@@ -784,20 +765,24 @@ static struct bfq_group *__bfq_bic_change_cgroup(struct bfq_data *bfqd,
 void bfq_bic_update_cgroup(struct bfq_io_cq *bic, struct bio *bio)
 {
 	struct bfq_data *bfqd = bic_to_bfqd(bic);
-	struct bfq_group *bfqg = NULL;
+	struct bfq_group *bfqg = bfq_bio_bfqg(bfqd, bio);
 	uint64_t serial_nr;
 
-	rcu_read_lock();
-	serial_nr = __bio_blkcg(bio)->css.serial_nr;
+	serial_nr = bfqg_to_blkg(bfqg)->blkcg->css.serial_nr;
 
 	/*
 	 * Check whether blkcg has changed.  The condition may trigger
 	 * spuriously on a newly created cic but there's no harm.
 	 */
 	if (unlikely(!bfqd) || likely(bic->blkcg_serial_nr == serial_nr))
-		goto out;
+		return;
 
-	bfqg = __bfq_bic_change_cgroup(bfqd, bic, __bio_blkcg(bio));
+	/*
+	 * New cgroup for this process. Make sure it is linked to bfq internal
+	 * cgroup hierarchy.
+	 */
+	bfq_link_bfqg(bfqd, bfqg);
+	__bfq_bic_change_cgroup(bfqd, bic, bfqg);
 	/*
 	 * Update blkg_path for bfq_log_* functions. We cache this
 	 * path, and update it here, for the following

@@ -850,8 +835,6 @@ void bfq_bic_update_cgroup(struct bfq_io_cq *bic, struct bio *bio)
 	 */
 	blkg_path(bfqg_to_blkg(bfqg), bfqg->blkg_path, sizeof(bfqg->blkg_path));
 	bic->blkcg_serial_nr = serial_nr;
-out:
-	rcu_read_unlock();
 }
 
 /**

@@ -1469,7 +1452,7 @@ void bfq_end_wr_async(struct bfq_data *bfqd)
 	bfq_end_wr_async_queues(bfqd, bfqd->root_group);
 }
 
-struct bfq_group *bfq_find_set_group(struct bfq_data *bfqd, struct blkcg *blkcg)
+struct bfq_group *bfq_bio_bfqg(struct bfq_data *bfqd, struct bio *bio)
 {
 	return bfqd->root_group;
 }

index d7cf930b47bb..e47c75f1fa0f 100644
@@ -5726,14 +5726,7 @@ static struct bfq_queue *bfq_get_queue(struct bfq_data *bfqd,
 	struct bfq_queue *bfqq;
 	struct bfq_group *bfqg;
 
-	rcu_read_lock();
-
-	bfqg = bfq_find_set_group(bfqd, __bio_blkcg(bio));
-	if (!bfqg) {
-		bfqq = &bfqd->oom_bfqq;
-		goto out;
-	}
-
+	bfqg = bfq_bio_bfqg(bfqd, bio);
 	if (!is_sync) {
 		async_bfqq = bfq_async_queue_prio(bfqd, bfqg, ioprio_class,
 						  ioprio);

@@ -5779,8 +5772,6 @@ static struct bfq_queue *bfq_get_queue(struct bfq_data *bfqd,
 
 	if (bfqq != &bfqd->oom_bfqq && is_sync && !respawn)
 		bfqq = bfq_do_or_sched_stable_merge(bfqd, bfqq, bic);
-
-	rcu_read_unlock();
 	return bfqq;
 }
 

index 4664e2f3e828..978ef5d6fe6a 100644
@@ -1009,8 +1009,7 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 void bfq_init_entity(struct bfq_entity *entity, struct bfq_group *bfqg);
 void bfq_bic_update_cgroup(struct bfq_io_cq *bic, struct bio *bio);
 void bfq_end_wr_async(struct bfq_data *bfqd);
-struct bfq_group *bfq_find_set_group(struct bfq_data *bfqd,
-				     struct blkcg *blkcg);
+struct bfq_group *bfq_bio_bfqg(struct bfq_data *bfqd, struct bio *bio);
 struct blkcg_gq *bfqg_to_blkg(struct bfq_group *bfqg);
 struct bfq_group *bfqq_group(struct bfq_queue *bfqq);
 struct bfq_group *bfq_create_group_hierarchy(struct bfq_data *bfqd, int node);

index 32d2c2a47480..09574af83566 100644
@@ -612,10 +612,19 @@ static void bfq_link_bfqg(struct bfq_data *bfqd, struct bfq_group *bfqg)
 struct bfq_group *bfq_bio_bfqg(struct bfq_data *bfqd, struct bio *bio)
 {
 	struct blkcg_gq *blkg = bio->bi_blkg;
+	struct bfq_group *bfqg;
 
-	if (!blkg)
-		return bfqd->root_group;
-	return blkg_to_bfqg(blkg);
+	while (blkg) {
+		bfqg = blkg_to_bfqg(blkg);
+		if (bfqg->online) {
+			bio_associate_blkg_from_css(bio, &blkg->blkcg->css);
+			return bfqg;
+		}
+		blkg = blkg->parent;
+	}
+	bio_associate_blkg_from_css(bio,
+				&bfqg_to_blkg(bfqd->root_group)->blkcg->css);
+	return bfqd->root_group;
 }
 
 /**