Merge tag 'v5.4.47' into 5.4-2.1.x-imx

This is the 5.4.47 stable release

All conflicts resolved in favour of HEAD, i.e. 5.4-2.1.x-imx

Conflicts:
	arch/arm/boot/dts/imx6qdl.dtsi
	arch/arm/mach-imx/Kconfig
	arch/arm/mach-imx/common.h
	arch/arm/mach-imx/suspend-imx6.S
	arch/arm64/boot/dts/freescale/imx8qxp-mek.dts
	arch/powerpc/include/asm/cacheflush.h
	drivers/cpufreq/imx6q-cpufreq.c
	drivers/dma/imx-sdma.c
	drivers/edac/synopsys_edac.c
	drivers/firmware/imx/imx-scu.c
	drivers/net/ethernet/freescale/fec.h
	drivers/net/ethernet/freescale/fec_main.c
	drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c
	drivers/net/phy/phy_device.c
	drivers/perf/fsl_imx8_ddr_perf.c
	drivers/usb/cdns3/gadget.c
	drivers/usb/dwc3/gadget.c
	include/uapi/linux/dma-buf.h

Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com>

12 files changed, 243 insertions, 119 deletions

diff --git a/block/bfq-cgroup.c b/block/bfq-cgroup.c
index 86a607cf19a1..12b707a4e52f 100644
--- a/block/bfq-cgroup.c
+++ b/block/bfq-cgroup.c
@@ -332,7 +332,7 @@ static void bfqg_put(struct bfq_group *bfqg)
 		kfree(bfqg);
 }
 
-static void bfqg_and_blkg_get(struct bfq_group *bfqg)
+void bfqg_and_blkg_get(struct bfq_group *bfqg)
 {
 	/* see comments in bfq_bic_update_cgroup for why refcounting bfqg */
 	bfqg_get(bfqg);
@@ -593,12 +593,13 @@ struct bfq_group *bfq_find_set_group(struct bfq_data *bfqd,
 	 */
 	entity = &bfqg->entity;
 	for_each_entity(entity) {
-		bfqg = container_of(entity, struct bfq_group, entity);
-		if (bfqg != bfqd->root_group) {
-			parent = bfqg_parent(bfqg);
+		struct bfq_group *curr_bfqg = container_of(entity,
+						struct bfq_group, entity);
+		if (curr_bfqg != bfqd->root_group) {
+			parent = bfqg_parent(curr_bfqg);
 			if (!parent)
 				parent = bfqd->root_group;
-			bfq_group_set_parent(bfqg, parent);
+			bfq_group_set_parent(curr_bfqg, parent);
 		}
 	}
 
@@ -624,6 +625,12 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 {
 	struct bfq_entity *entity = &bfqq->entity;
 
+	/*
+	 * Get extra reference to prevent bfqq from being freed in
+	 * next possible expire or deactivate.
+	 */
+	bfqq->ref++;
+
 	/* If bfqq is empty, then bfq_bfqq_expire also invokes
 	 * bfq_del_bfqq_busy, thereby removing bfqq and its entity
 	 * from data structures related to current group. Otherwise we
@@ -653,6 +660,8 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 
 	if (!bfqd->in_service_queue && !bfqd->rq_in_driver)
 		bfq_schedule_dispatch(bfqd);
+	/* release extra ref taken above, bfqq may happen to be freed now */
+	bfq_put_queue(bfqq);
 }
 
 /**
@@ -688,10 +697,7 @@ static struct bfq_group *__bfq_bic_change_cgroup(struct bfq_data *bfqd,
 
 		if (entity->sched_data != &bfqg->sched_data) {
 			bic_set_bfqq(bic, NULL, 0);
-			bfq_log_bfqq(bfqd, async_bfqq,
-				     "bic_change_group: %p %d",
-				     async_bfqq, async_bfqq->ref);
-			bfq_put_queue(async_bfqq);
+			bfq_release_process_ref(bfqd, async_bfqq);
 		}
 	}
 
@@ -792,39 +798,53 @@ static void bfq_flush_idle_tree(struct bfq_service_tree *st)
 /**
  * bfq_reparent_leaf_entity - move leaf entity to the root_group.
  * @bfqd: the device data structure with the root group.
- * @entity: the entity to move.
+ * @entity: the entity to move, if entity is a leaf; or the parent entity
+ *	    of an active leaf entity to move, if entity is not a leaf.
  */
 static void bfq_reparent_leaf_entity(struct bfq_data *bfqd,
-				     struct bfq_entity *entity)
+				     struct bfq_entity *entity,
+				     int ioprio_class)
 {
-	struct bfq_queue *bfqq = bfq_entity_to_bfqq(entity);
+	struct bfq_queue *bfqq;
+	struct bfq_entity *child_entity = entity;
+
+	while (child_entity->my_sched_data) { /* leaf not reached yet */
+		struct bfq_sched_data *child_sd = child_entity->my_sched_data;
+		struct bfq_service_tree *child_st = child_sd->service_tree +
+			ioprio_class;
+		struct rb_root *child_active = &child_st->active;
 
+		child_entity = bfq_entity_of(rb_first(child_active));
+
+		if (!child_entity)
+			child_entity = child_sd->in_service_entity;
+	}
+
+	bfqq = bfq_entity_to_bfqq(child_entity);
 	bfq_bfqq_move(bfqd, bfqq, bfqd->root_group);
 }
 
 /**
- * bfq_reparent_active_entities - move to the root group all active
- *                                entities.
+ * bfq_reparent_active_queues - move to the root group all active queues.
  * @bfqd: the device data structure with the root group.
  * @bfqg: the group to move from.
- * @st: the service tree with the entities.
+ * @st: the service tree to start the search from.
  */
-static void bfq_reparent_active_entities(struct bfq_data *bfqd,
-					 struct bfq_group *bfqg,
-					 struct bfq_service_tree *st)
+static void bfq_reparent_active_queues(struct bfq_data *bfqd,
+				       struct bfq_group *bfqg,
+				       struct bfq_service_tree *st,
+				       int ioprio_class)
 {
 	struct rb_root *active = &st->active;
-	struct bfq_entity *entity = NULL;
-
-	if (!RB_EMPTY_ROOT(&st->active))
-		entity = bfq_entity_of(rb_first(active));
+	struct bfq_entity *entity;
 
-	for (; entity ; entity = bfq_entity_of(rb_first(active)))
-		bfq_reparent_leaf_entity(bfqd, entity);
+	while ((entity = bfq_entity_of(rb_first(active))))
+		bfq_reparent_leaf_entity(bfqd, entity, ioprio_class);
 
 	if (bfqg->sched_data.in_service_entity)
 		bfq_reparent_leaf_entity(bfqd,
-			bfqg->sched_data.in_service_entity);
+					 bfqg->sched_data.in_service_entity,
+					 ioprio_class);
 }
 
 /**
@@ -857,13 +877,6 @@ static void bfq_pd_offline(struct blkg_policy_data *pd)
 		st = bfqg->sched_data.service_tree + i;
 
 		/*
-		 * The idle tree may still contain bfq_queues belonging
-		 * to exited task because they never migrated to a different
-		 * cgroup from the one being destroyed now.
-		 */
-		bfq_flush_idle_tree(st);
-
-		/*
 		 * It may happen that some queues are still active
 		 * (busy) upon group destruction (if the corresponding
 		 * processes have been forced to terminate). We move
@@ -875,7 +888,20 @@ static void bfq_pd_offline(struct blkg_policy_data *pd)
 		 * There is no need to put the sync queues, as the
 		 * scheduler has taken no reference.
 		 */
-		bfq_reparent_active_entities(bfqd, bfqg, st);
+		bfq_reparent_active_queues(bfqd, bfqg, st, i);
+
+		/*
+		 * The idle tree may still contain bfq_queues
+		 * belonging to exited task because they never
+		 * migrated to a different cgroup from the one being
+		 * destroyed now. In addition, even
+		 * bfq_reparent_active_queues() may happen to add some
+		 * entities to the idle tree. It happens if, in some
+		 * of the calls to bfq_bfqq_move() performed by
+		 * bfq_reparent_active_queues(), the queue to move is
+		 * empty and gets expired.
+		 */
+		bfq_flush_idle_tree(st);
 	}
 
 	__bfq_deactivate_entity(entity, false);
@@ -1380,6 +1406,10 @@ struct bfq_group *bfqq_group(struct bfq_queue *bfqq)
 	return bfqq->bfqd->root_group;
 }
 
+void bfqg_and_blkg_get(struct bfq_group *bfqg) {}
+
+void bfqg_and_blkg_put(struct bfq_group *bfqg) {}
+
 struct bfq_group *bfq_create_group_hierarchy(struct bfq_data *bfqd, int node)
 {
 	struct bfq_group *bfqg;
diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
index 5498d05b873d..88497bff1135 100644
--- a/block/bfq-iosched.c
+++ b/block/bfq-iosched.c
@@ -614,6 +614,10 @@ bfq_pos_tree_add_move(struct bfq_data *bfqd, struct bfq_queue *bfqq)
 		bfqq->pos_root = NULL;
 	}
 
+	/* oom_bfqq does not participate in queue merging */
+	if (bfqq == &bfqd->oom_bfqq)
+		return;
+
 	/*
 	 * bfqq cannot be merged any longer (see comments in
 	 * bfq_setup_cooperator): no point in adding bfqq into the
@@ -2713,8 +2717,6 @@ static void bfq_bfqq_save_state(struct bfq_queue *bfqq)
 	}
 }
 
-
-static
 void bfq_release_process_ref(struct bfq_data *bfqd, struct bfq_queue *bfqq)
 {
 	/*
@@ -4822,9 +4824,7 @@ void bfq_put_queue(struct bfq_queue *bfqq)
 {
 	struct bfq_queue *item;
 	struct hlist_node *n;
-#ifdef CONFIG_BFQ_GROUP_IOSCHED
 	struct bfq_group *bfqg = bfqq_group(bfqq);
-#endif
 
 	if (bfqq->bfqd)
 		bfq_log_bfqq(bfqq->bfqd, bfqq, "put_queue: %p %d",
@@ -4897,9 +4897,7 @@ void bfq_put_queue(struct bfq_queue *bfqq)
 		bfqq->bfqd->last_completed_rq_bfqq = NULL;
 
 	kmem_cache_free(bfq_pool, bfqq);
-#ifdef CONFIG_BFQ_GROUP_IOSCHED
 	bfqg_and_blkg_put(bfqg);
-#endif
 }
 
 static void bfq_put_cooperator(struct bfq_queue *bfqq)
@@ -6210,20 +6208,28 @@ static struct bfq_queue *bfq_init_rq(struct request *rq)
 	return bfqq;
 }
 
-static void bfq_idle_slice_timer_body(struct bfq_queue *bfqq)
+static void
+bfq_idle_slice_timer_body(struct bfq_data *bfqd, struct bfq_queue *bfqq)
 {
-	struct bfq_data *bfqd = bfqq->bfqd;
 	enum bfqq_expiration reason;
 	unsigned long flags;
 
 	spin_lock_irqsave(&bfqd->lock, flags);
-	bfq_clear_bfqq_wait_request(bfqq);
 
+	/*
+	 * Considering that bfqq may be in race, we should firstly check
+	 * whether bfqq is in service before doing something on it. If
+	 * the bfqq in race is not in service, it has already been expired
+	 * through __bfq_bfqq_expire func and its wait_request flags has
+	 * been cleared in __bfq_bfqd_reset_in_service func.
+	 */
 	if (bfqq != bfqd->in_service_queue) {
 		spin_unlock_irqrestore(&bfqd->lock, flags);
 		return;
 	}
 
+	bfq_clear_bfqq_wait_request(bfqq);
+
 	if (bfq_bfqq_budget_timeout(bfqq))
 		/*
 		 * Also here the queue can be safely expired
@@ -6268,7 +6274,7 @@ static enum hrtimer_restart bfq_idle_slice_timer(struct hrtimer *timer)
 	 * early.
 	 */
 	if (bfqq)
-		bfq_idle_slice_timer_body(bfqq);
+		bfq_idle_slice_timer_body(bfqd, bfqq);
 
 	return HRTIMER_NORESTART;
 }
@@ -6383,10 +6389,10 @@ static void bfq_exit_queue(struct elevator_queue *e)
 
 	hrtimer_cancel(&bfqd->idle_slice_timer);
 
-#ifdef CONFIG_BFQ_GROUP_IOSCHED
 	/* release oom-queue reference to root group */
 	bfqg_and_blkg_put(bfqd->root_group);
 
+#ifdef CONFIG_BFQ_GROUP_IOSCHED
 	blkcg_deactivate_policy(bfqd->queue, &blkcg_policy_bfq);
 #else
 	spin_lock_irq(&bfqd->lock);
diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h
index 5d1a519640f6..c0232975075d 100644
--- a/block/bfq-iosched.h
+++ b/block/bfq-iosched.h
@@ -916,6 +916,7 @@ struct bfq_group {
 
 #else
 struct bfq_group {
+	struct bfq_entity entity;
 	struct bfq_sched_data sched_data;
 
 	struct bfq_queue *async_bfqq[2][IOPRIO_BE_NR];
@@ -949,6 +950,7 @@ 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_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);
 void bfq_put_async_queues(struct bfq_data *bfqd, struct bfq_group *bfqg);
 
@@ -978,6 +980,7 @@ struct bfq_group *bfq_find_set_group(struct bfq_data *bfqd,
 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);
+void bfqg_and_blkg_get(struct bfq_group *bfqg);
 void bfqg_and_blkg_put(struct bfq_group *bfqg);
 
 #ifdef CONFIG_BFQ_GROUP_IOSCHED
diff --git a/block/bfq-wf2q.c b/block/bfq-wf2q.c
index 05f0bf4a1144..44079147e396 100644
--- a/block/bfq-wf2q.c
+++ b/block/bfq-wf2q.c
@@ -536,7 +536,9 @@ static void bfq_get_entity(struct bfq_entity *entity)
 		bfqq->ref++;
 		bfq_log_bfqq(bfqq->bfqd, bfqq, "get_entity: %p %d",
 			     bfqq, bfqq->ref);
-	}
+	} else
+		bfqg_and_blkg_get(container_of(entity, struct bfq_group,
+					       entity));
 }
 
 /**
@@ -650,8 +652,14 @@ static void bfq_forget_entity(struct bfq_service_tree *st,
 
 	entity->on_st = false;
 	st->wsum -= entity->weight;
-	if (bfqq && !is_in_service)
+	if (is_in_service)
+		return;
+
+	if (bfqq)
 		bfq_put_queue(bfqq);
+	else
+		bfqg_and_blkg_put(container_of(entity, struct bfq_group,
+					       entity));
 }
 
 /**
diff --git a/block/blk-core.c b/block/blk-core.c
index 1075aaff606d..d5e668ec751b 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -886,14 +886,11 @@ generic_make_request_checks(struct bio *bio)
 	}
 
 	/*
-	 * Non-mq queues do not honor REQ_NOWAIT, so complete a bio
-	 * with BLK_STS_AGAIN status in order to catch -EAGAIN and
-	 * to give a chance to the caller to repeat request gracefully.
+	 * For a REQ_NOWAIT based request, return -EOPNOTSUPP
+	 * if queue is not a request based queue.
 	 */
-	if ((bio->bi_opf & REQ_NOWAIT) && !queue_is_mq(q)) {
-		status = BLK_STS_AGAIN;
-		goto end_io;
-	}
+	if ((bio->bi_opf & REQ_NOWAIT) && !queue_is_mq(q))
+		goto not_supported;
 
 	if (should_fail_bio(bio))
 		goto end_io;
diff --git a/block/blk-flush.c b/block/blk-flush.c
index b1f0a1ac505c..5aa6fada2259 100644
--- a/block/blk-flush.c
+++ b/block/blk-flush.c
@@ -399,7 +399,7 @@ void blk_insert_flush(struct request *rq)
 	 */
 	if ((policy & REQ_FSEQ_DATA) &&
 	    !(policy & (REQ_FSEQ_PREFLUSH | REQ_FSEQ_POSTFLUSH))) {
-		blk_mq_request_bypass_insert(rq, false);
+		blk_mq_request_bypass_insert(rq, false, false);
 		return;
 	}
 
diff --git a/block/blk-ioc.c b/block/blk-ioc.c
index 5ed59ac6ae58..9df50fb507ca 100644
--- a/block/blk-ioc.c
+++ b/block/blk-ioc.c
@@ -84,6 +84,7 @@ static void ioc_destroy_icq(struct io_cq *icq)
 	 * making it impossible to determine icq_cache.  Record it in @icq.
 	 */
 	icq->__rcu_icq_cache = et->icq_cache;
+	icq->flags |= ICQ_DESTROYED;
 	call_rcu(&icq->__rcu_head, icq_free_icq_rcu);
 }
 
@@ -212,15 +213,21 @@ static void __ioc_clear_queue(struct list_head *icq_list)
 {
 	unsigned long flags;
 
+	rcu_read_lock();
 	while (!list_empty(icq_list)) {
 		struct io_cq *icq = list_entry(icq_list->next,
 						struct io_cq, q_node);
 		struct io_context *ioc = icq->ioc;
 
 		spin_lock_irqsave(&ioc->lock, flags);
+		if (icq->flags & ICQ_DESTROYED) {
+			spin_unlock_irqrestore(&ioc->lock, flags);
+			continue;
+		}
 		ioc_destroy_icq(icq);
 		spin_unlock_irqrestore(&ioc->lock, flags);
 	}
+	rcu_read_unlock();
 }
 
 /**
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index 27ca68621137..d083f7704082 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -469,7 +469,7 @@ struct ioc_gq {
 	 */
 	atomic64_t			vtime;
 	atomic64_t			done_vtime;
-	atomic64_t			abs_vdebt;
+	u64				abs_vdebt;
 	u64				last_vtime;
 
 	/*
@@ -1145,7 +1145,7 @@ static void iocg_kick_waitq(struct ioc_gq *iocg, struct ioc_now *now)
 	struct iocg_wake_ctx ctx = { .iocg = iocg };
 	u64 margin_ns = (u64)(ioc->period_us *
 			      WAITQ_TIMER_MARGIN_PCT / 100) * NSEC_PER_USEC;
-	u64 abs_vdebt, vdebt, vshortage, expires, oexpires;
+	u64 vdebt, vshortage, expires, oexpires;
 	s64 vbudget;
 	u32 hw_inuse;
 
@@ -1155,18 +1155,15 @@ static void iocg_kick_waitq(struct ioc_gq *iocg, struct ioc_now *now)
 	vbudget = now->vnow - atomic64_read(&iocg->vtime);
 
 	/* pay off debt */
-	abs_vdebt = atomic64_read(&iocg->abs_vdebt);
-	vdebt = abs_cost_to_cost(abs_vdebt, hw_inuse);
+	vdebt = abs_cost_to_cost(iocg->abs_vdebt, hw_inuse);
 	if (vdebt && vbudget > 0) {
 		u64 delta = min_t(u64, vbudget, vdebt);
 		u64 abs_delta = min(cost_to_abs_cost(delta, hw_inuse),
-				    abs_vdebt);
+				    iocg->abs_vdebt);
 
 		atomic64_add(delta, &iocg->vtime);
 		atomic64_add(delta, &iocg->done_vtime);
-		atomic64_sub(abs_delta, &iocg->abs_vdebt);
-		if (WARN_ON_ONCE(atomic64_read(&iocg->abs_vdebt) < 0))
-			atomic64_set(&iocg->abs_vdebt, 0);
+		iocg->abs_vdebt -= abs_delta;
 	}
 
 	/*
@@ -1222,12 +1219,18 @@ static bool iocg_kick_delay(struct ioc_gq *iocg, struct ioc_now *now, u64 cost)
 	u64 expires, oexpires;
 	u32 hw_inuse;
 
+	lockdep_assert_held(&iocg->waitq.lock);
+
 	/* debt-adjust vtime */
 	current_hweight(iocg, NULL, &hw_inuse);
-	vtime += abs_cost_to_cost(atomic64_read(&iocg->abs_vdebt), hw_inuse);
+	vtime += abs_cost_to_cost(iocg->abs_vdebt, hw_inuse);
 
-	/* clear or maintain depending on the overage */
-	if (time_before_eq64(vtime, now->vnow)) {
+	/*
+	 * Clear or maintain depending on the overage. Non-zero vdebt is what
+	 * guarantees that @iocg is online and future iocg_kick_delay() will
+	 * clear use_delay. Don't leave it on when there's no vdebt.
+	 */
+	if (!iocg->abs_vdebt || time_before_eq64(vtime, now->vnow)) {
 		blkcg_clear_delay(blkg);
 		return false;
 	}
@@ -1261,9 +1264,12 @@ static enum hrtimer_restart iocg_delay_timer_fn(struct hrtimer *timer)
 {
 	struct ioc_gq *iocg = container_of(timer, struct ioc_gq, delay_timer);
 	struct ioc_now now;
+	unsigned long flags;
 
+	spin_lock_irqsave(&iocg->waitq.lock, flags);
 	ioc_now(iocg->ioc, &now);
 	iocg_kick_delay(iocg, &now, 0);
+	spin_unlock_irqrestore(&iocg->waitq.lock, flags);
 
 	return HRTIMER_NORESTART;
 }
@@ -1318,7 +1324,7 @@ static bool iocg_is_idle(struct ioc_gq *iocg)
 		return false;
 
 	/* is something in flight? */
-	if (atomic64_read(&iocg->done_vtime) < atomic64_read(&iocg->vtime))
+	if (atomic64_read(&iocg->done_vtime) != atomic64_read(&iocg->vtime))
 		return false;
 
 	return true;
@@ -1371,14 +1377,13 @@ static void ioc_timer_fn(struct timer_list *timer)
 	 * should have woken up in the last period and expire idle iocgs.
 	 */
 	list_for_each_entry_safe(iocg, tiocg, &ioc->active_iocgs, active_list) {
-		if (!waitqueue_active(&iocg->waitq) &&
-		    !atomic64_read(&iocg->abs_vdebt) && !iocg_is_idle(iocg))
+		if (!waitqueue_active(&iocg->waitq) && iocg->abs_vdebt &&
+		    !iocg_is_idle(iocg))
 			continue;
 
 		spin_lock(&iocg->waitq.lock);
 
-		if (waitqueue_active(&iocg->waitq) ||
-		    atomic64_read(&iocg->abs_vdebt)) {
+		if (waitqueue_active(&iocg->waitq) || iocg->abs_vdebt) {
 			/* might be oversleeping vtime / hweight changes, kick */
 			iocg_kick_waitq(iocg, &now);
 			iocg_kick_delay(iocg, &now, 0);
@@ -1594,7 +1599,7 @@ skip_surplus_transfers:
 				      vrate_min, vrate_max);
 		}
 
-		trace_iocost_ioc_vrate_adj(ioc, vrate, &missed_ppm, rq_wait_pct,
+		trace_iocost_ioc_vrate_adj(ioc, vrate, missed_ppm, rq_wait_pct,
 					   nr_lagging, nr_shortages,
 					   nr_surpluses);
 
@@ -1603,7 +1608,7 @@ skip_surplus_transfers:
 			ioc->period_us * vrate * INUSE_MARGIN_PCT, 100);
 	} else if (ioc->busy_level != prev_busy_level || nr_lagging) {
 		trace_iocost_ioc_vrate_adj(ioc, atomic64_read(&ioc->vtime_rate),
-					   &missed_ppm, rq_wait_pct, nr_lagging,
+					   missed_ppm, rq_wait_pct, nr_lagging,
 					   nr_shortages, nr_surpluses);
 	}
 
@@ -1721,28 +1726,49 @@ static void ioc_rqos_throttle(struct rq_qos *rqos, struct bio *bio)
 	 * tests are racy but the races aren't systemic - we only miss once
 	 * in a while which is fine.
 	 */
-	if (!waitqueue_active(&iocg->waitq) &&
-	    !atomic64_read(&iocg->abs_vdebt) &&
+	if (!waitqueue_active(&iocg->waitq) && !iocg->abs_vdebt &&
 	    time_before_eq64(vtime + cost, now.vnow)) {
 		iocg_commit_bio(iocg, bio, cost);
 		return;
 	}
 
 	/*
-	 * We're over budget.  If @bio has to be issued regardless,
-	 * remember the abs_cost instead of advancing vtime.
-	 * iocg_kick_waitq() will pay off the debt before waking more IOs.
+	 * We activated above but w/o any synchronization. Deactivation is
+	 * synchronized with waitq.lock and we won't get deactivated as long
+	 * as we're waiting or has debt, so we're good if we're activated
+	 * here. In the unlikely case that we aren't, just issue the IO.
+	 */
+	spin_lock_irq(&iocg->waitq.lock);
+
+	if (unlikely(list_empty(&iocg->active_list))) {
+		spin_unlock_irq(&iocg->waitq.lock);
+		iocg_commit_bio(iocg, bio, cost);
+		return;
+	}
+
+	/*
+	 * We're over budget. If @bio has to be issued regardless, remember
+	 * the abs_cost instead of advancing vtime. iocg_kick_waitq() will pay
+	 * off the debt before waking more IOs.
+	 *
 	 * This way, the debt is continuously paid off each period with the
-	 * actual budget available to the cgroup.  If we just wound vtime,
-	 * we would incorrectly use the current hw_inuse for the entire
-	 * amount which, for example, can lead to the cgroup staying
-	 * blocked for a long time even with substantially raised hw_inuse.
+	 * actual budget available to the cgroup. If we just wound vtime, we
+	 * would incorrectly use the current hw_inuse for the entire amount
+	 * which, for example, can lead to the cgroup staying blocked for a
+	 * long time even with substantially raised hw_inuse.
+	 *
+	 * An iocg with vdebt should stay online so that the timer can keep
+	 * deducting its vdebt and [de]activate use_delay mechanism
+	 * accordingly. We don't want to race against the timer trying to
+	 * clear them and leave @iocg inactive w/ dangling use_delay heavily
+	 * penalizing the cgroup and its descendants.
 	 */
 	if (bio_issue_as_root_blkg(bio) || fatal_signal_pending(current)) {
-		atomic64_add(abs_cost, &iocg->abs_vdebt);
+		iocg->abs_vdebt += abs_cost;
 		if (iocg_kick_delay(iocg, &now, cost))
 			blkcg_schedule_throttle(rqos->q,
 					(bio->bi_opf & REQ_SWAP) == REQ_SWAP);
+		spin_unlock_irq(&iocg->waitq.lock);
 		return;
 	}
 
@@ -1759,20 +1785,6 @@ static void ioc_rqos_throttle(struct rq_qos *rqos, struct bio *bio)
 	 * All waiters are on iocg->waitq and the wait states are
 	 * synchronized using waitq.lock.
 	 */
-	spin_lock_irq(&iocg->waitq.lock);
-
-	/*
-	 * We activated above but w/o any synchronization.  Deactivation is
-	 * synchronized with waitq.lock and we won't get deactivated as
-	 * long as we're waiting, so we're good if we're activated here.
-	 * In the unlikely case that we are deactivated, just issue the IO.
-	 */
-	if (unlikely(list_empty(&iocg->active_list))) {
-		spin_unlock_irq(&iocg->waitq.lock);
-		iocg_commit_bio(iocg, bio, cost);
-		return;
-	}
-
 	init_waitqueue_func_entry(&wait.wait, iocg_wake_fn);
 	wait.wait.private = current;
 	wait.bio = bio;
@@ -1804,6 +1816,7 @@ static void ioc_rqos_merge(struct rq_qos *rqos, struct request *rq,
 	struct ioc_now now;
 	u32 hw_inuse;
 	u64 abs_cost, cost;
+	unsigned long flags;
 
 	/* bypass if disabled or for root cgroup */
 	if (!ioc->enabled || !iocg->level)
@@ -1823,15 +1836,28 @@ static void ioc_rqos_merge(struct rq_qos *rqos, struct request *rq,
 		iocg->cursor = bio_end;
 
 	/*
-	 * Charge if there's enough vtime budget and the existing request
-	 * has cost assigned.  Otherwise, account it as debt.  See debt
-	 * handling in ioc_rqos_throttle() for details.
+	 * Charge if there's enough vtime budget and the existing request has
+	 * cost assigned.
 	 */
 	if (rq->bio && rq->bio->bi_iocost_cost &&
-	    time_before_eq64(atomic64_read(&iocg->vtime) + cost, now.vnow))
+	    time_before_eq64(atomic64_read(&iocg->vtime) + cost, now.vnow)) {
 		iocg_commit_bio(iocg, bio, cost);
-	else
-		atomic64_add(abs_cost, &iocg->abs_vdebt);
+		return;
+	}
+
+	/*
+	 * Otherwise, account it as debt if @iocg is online, which it should
+	 * be for the vast majority of cases. See debt handling in
+	 * ioc_rqos_throttle() for details.
+	 */
+	spin_lock_irqsave(&iocg->waitq.lock, flags);
+	if (likely(!list_empty(&iocg->active_list))) {
+		iocg->abs_vdebt += abs_cost;
+		iocg_kick_delay(iocg, &now, cost);
+	} else {
+		iocg_commit_bio(iocg, bio, cost);
+	}
+	spin_unlock_irqrestore(&iocg->waitq.lock, flags);
 }
 
 static void ioc_rqos_done_bio(struct rq_qos *rqos, struct bio *bio)
@@ -2001,7 +2027,6 @@ static void ioc_pd_init(struct blkg_policy_data *pd)
 	iocg->ioc = ioc;
 	atomic64_set(&iocg->vtime, now.vnow);
 	atomic64_set(&iocg->done_vtime, now.vnow);
-	atomic64_set(&iocg->abs_vdebt, 0);
 	atomic64_set(&iocg->active_period, atomic64_read(&ioc->cur_period));
 	INIT_LIST_HEAD(&iocg->active_list);
 	iocg->hweight_active = HWEIGHT_WHOLE;
diff --git a/block/blk-mq-sched.c b/block/blk-mq-sched.c
index ca22afd47b3d..74cedea56034 100644
--- a/block/blk-mq-sched.c
+++ b/block/blk-mq-sched.c
@@ -361,13 +361,19 @@ static bool blk_mq_sched_bypass_insert(struct blk_mq_hw_ctx *hctx,
 				       bool has_sched,
 				       struct request *rq)
 {
-	/* dispatch flush rq directly */
-	if (rq->rq_flags & RQF_FLUSH_SEQ) {
-		spin_lock(&hctx->lock);
-		list_add(&rq->queuelist, &hctx->dispatch);
-		spin_unlock(&hctx->lock);
+	/*
+	 * dispatch flush and passthrough rq directly
+	 *
+	 * passthrough request has to be added to hctx->dispatch directly.
+	 * For some reason, device may be in one situation which can't
+	 * handle FS request, so STS_RESOURCE is always returned and the
+	 * FS request will be added to hctx->dispatch. However passthrough
+	 * request may be required at that time for fixing the problem. If
+	 * passthrough request is added to scheduler queue, there isn't any
+	 * chance to dispatch it given we prioritize requests in hctx->dispatch.
+	 */
+	if ((rq->rq_flags & RQF_FLUSH_SEQ) || blk_rq_is_passthrough(rq))
 		return true;
-	}
 
 	if (has_sched)
 		rq->rq_flags |= RQF_SORTED;
@@ -391,8 +397,32 @@ void blk_mq_sched_insert_request(struct request *rq, bool at_head,
 
 	WARN_ON(e && (rq->tag != -1));
 
-	if (blk_mq_sched_bypass_insert(hctx, !!e, rq))
+	if (blk_mq_sched_bypass_insert(hctx, !!e, rq)) {
+		/*
+		 * Firstly normal IO request is inserted to scheduler queue or
+		 * sw queue, meantime we add flush request to dispatch queue(
+		 * hctx->dispatch) directly and there is at most one in-flight
+		 * flush request for each hw queue, so it doesn't matter to add
+		 * flush request to tail or front of the dispatch queue.
+		 *
+		 * Secondly in case of NCQ, flush request belongs to non-NCQ
+		 * command, and queueing it will fail when there is any
+		 * in-flight normal IO request(NCQ command). When adding flush
+		 * rq to the front of hctx->dispatch, it is easier to introduce
+		 * extra time to flush rq's latency because of S_SCHED_RESTART
+		 * compared with adding to the tail of dispatch queue, then
+		 * chance of flush merge is increased, and less flush requests
+		 * will be issued to controller. It is observed that ~10% time
+		 * is saved in blktests block/004 on disk attached to AHCI/NCQ
+		 * drive when adding flush rq to the front of hctx->dispatch.
+		 *
+		 * Simply queue flush rq to the front of hctx->dispatch so that
+		 * intensive flush workloads can benefit in case of NCQ HW.
+		 */
+		at_head = (rq->rq_flags & RQF_FLUSH_SEQ) ? true : at_head;
+		blk_mq_request_bypass_insert(rq, at_head, false);
 		goto run;
+	}
 
 	if (e && e->type->ops.insert_requests) {
 		LIST_HEAD(list);
diff --git a/block/blk-mq.c b/block/blk-mq.c
index ec791156e9cc..757c0fd9f0cc 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -761,7 +761,7 @@ static void blk_mq_requeue_work(struct work_struct *work)
 		 * merge.
 		 */
 		if (rq->rq_flags & RQF_DONTPREP)
-			blk_mq_request_bypass_insert(rq, false);
+			blk_mq_request_bypass_insert(rq, false, false);
 		else
 			blk_mq_sched_insert_request(rq, true, false, false);
 	}
@@ -1232,8 +1232,10 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list,
 		rq = list_first_entry(list, struct request, queuelist);
 
 		hctx = rq->mq_hctx;
-		if (!got_budget && !blk_mq_get_dispatch_budget(hctx))
+		if (!got_budget && !blk_mq_get_dispatch_budget(hctx)) {
+			blk_mq_put_driver_tag(rq);
 			break;
+		}
 
 		if (!blk_mq_get_driver_tag(rq)) {
 			/*
@@ -1313,7 +1315,7 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list,
 			q->mq_ops->commit_rqs(hctx);
 
 		spin_lock(&hctx->lock);
-		list_splice_init(list, &hctx->dispatch);
+		list_splice_tail_init(list, &hctx->dispatch);
 		spin_unlock(&hctx->lock);
 
 		/*
@@ -1668,12 +1670,16 @@ void __blk_mq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
  * Should only be used carefully, when the caller knows we want to
  * bypass a potential IO scheduler on the target device.
  */
-void blk_mq_request_bypass_insert(struct request *rq, bool run_queue)
+void blk_mq_request_bypass_insert(struct request *rq, bool at_head,
+				  bool run_queue)
 {
 	struct blk_mq_hw_ctx *hctx = rq->mq_hctx;
 
 	spin_lock(&hctx->lock);
-	list_add_tail(&rq->queuelist, &hctx->dispatch);
+	if (at_head)
+		list_add(&rq->queuelist, &hctx->dispatch);
+	else
+		list_add_tail(&rq->queuelist, &hctx->dispatch);
 	spin_unlock(&hctx->lock);
 
 	if (run_queue)
@@ -1863,7 +1869,7 @@ insert:
 	if (bypass_insert)
 		return BLK_STS_RESOURCE;
 
-	blk_mq_request_bypass_insert(rq, run_queue);
+	blk_mq_request_bypass_insert(rq, false, run_queue);
 	return BLK_STS_OK;
 }
 
@@ -1879,7 +1885,7 @@ static void blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx,
 
 	ret = __blk_mq_try_issue_directly(hctx, rq, cookie, false, true);
 	if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE)
-		blk_mq_request_bypass_insert(rq, true);
+		blk_mq_request_bypass_insert(rq, false, true);
 	else if (ret != BLK_STS_OK)
 		blk_mq_end_request(rq, ret);
 
@@ -1913,7 +1919,7 @@ void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx,
 		if (ret != BLK_STS_OK) {
 			if (ret == BLK_STS_RESOURCE ||
 					ret == BLK_STS_DEV_RESOURCE) {
-				blk_mq_request_bypass_insert(rq,
+				blk_mq_request_bypass_insert(rq, false,
 							list_empty(list));
 				break;
 			}
@@ -3003,6 +3009,14 @@ static int blk_mq_alloc_rq_maps(struct blk_mq_tag_set *set)
 
 static int blk_mq_update_queue_map(struct blk_mq_tag_set *set)
 {
+	/*
+	 * blk_mq_map_queues() and multiple .map_queues() implementations
+	 * expect that set->map[HCTX_TYPE_DEFAULT].nr_queues is set to the
+	 * number of hardware queues.
+	 */
+	if (set->nr_maps == 1)
+		set->map[HCTX_TYPE_DEFAULT].nr_queues = set->nr_hw_queues;
+
 	if (set->ops->map_queues && !is_kdump_kernel()) {
 		int i;
 
diff --git a/block/blk-mq.h b/block/blk-mq.h
index 32c62c64e6c2..f2075978db50 100644
--- a/block/blk-mq.h
+++ b/block/blk-mq.h
@@ -66,7 +66,8 @@ int blk_mq_alloc_rqs(struct blk_mq_tag_set *set, struct blk_mq_tags *tags,
  */
 void __blk_mq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
 				bool at_head);
-void blk_mq_request_bypass_insert(struct request *rq, bool run_queue);
+void blk_mq_request_bypass_insert(struct request *rq, bool at_head,
+				  bool run_queue);
 void blk_mq_insert_requests(struct blk_mq_hw_ctx *hctx, struct blk_mq_ctx *ctx,
 				struct list_head *list);
 
diff --git a/block/blk-settings.c b/block/blk-settings.c
index c8eda2e7b91e..be1dca0103a4 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -664,6 +664,9 @@ void disk_stack_limits(struct gendisk *disk, struct block_device *bdev,
 		printk(KERN_NOTICE "%s: Warning: Device %s is misaligned\n",
 		       top, bottom);
 	}
+
+	t->backing_dev_info->io_pages =
+		t->limits.max_sectors >> (PAGE_SHIFT - 9);
 }
 EXPORT_SYMBOL(disk_stack_limits);