Merge tag 'for-5.3/block-20190708' of git://git.kernel.dk/linux-block

Pull block updates from Jens Axboe:
 "This is the main block updates for 5.3. Nothing earth shattering or
  major in here, just fixes, additions, and improvements all over the
  map. This contains:

   - Series of documentation fixes (Bart)

   - Optimization of the blk-mq ctx get/put (Bart)

   - null_blk removal race condition fix (Bob)

   - req/bio_op() cleanups (Chaitanya)

   - Series cleaning up the segment accounting, and request/bio mapping
     (Christoph)

   - Series cleaning up the page getting/putting for bios (Christoph)

   - block cgroup cleanups and moving it to where it is used (Christoph)

   - block cgroup fixes (Tejun)

   - Series of fixes and improvements to bcache, most notably a write
     deadlock fix (Coly)

   - blk-iolatency STS_AGAIN and accounting fixes (Dennis)

   - Series of improvements and fixes to BFQ (Douglas, Paolo)

   - debugfs_create() return value check removal for drbd (Greg)

   - Use struct_size(), where appropriate (Gustavo)

   - Two lighnvm fixes (Heiner, Geert)

   - MD fixes, including a read balance and corruption fix (Guoqing,
     Marcos, Xiao, Yufen)

   - block opal shadow mbr additions (Jonas, Revanth)

   - sbitmap compare-and-exhange improvemnts (Pavel)

   - Fix for potential bio->bi_size overflow (Ming)

   - NVMe pull requests:
       - improved PCIe suspent support (Keith Busch)
       - error injection support for the admin queue (Akinobu Mita)
       - Fibre Channel discovery improvements (James Smart)
       - tracing improvements including nvmetc tracing support (Minwoo Im)
       - misc fixes and cleanups (Anton Eidelman, Minwoo Im, Chaitanya
         Kulkarni)"

   - Various little fixes and improvements to drivers and core"

* tag 'for-5.3/block-20190708' of git://git.kernel.dk/linux-block: (153 commits)
  blk-iolatency: fix STS_AGAIN handling
  block: nr_phys_segments needs to be zero for REQ_OP_WRITE_ZEROES
  blk-mq: simplify blk_mq_make_request()
  blk-mq: remove blk_mq_put_ctx()
  sbitmap: Replace cmpxchg with xchg
  block: fix .bi_size overflow
  block: sed-opal: check size of shadow mbr
  block: sed-opal: ioctl for writing to shadow mbr
  block: sed-opal: add ioctl for done-mark of shadow mbr
  block: never take page references for ITER_BVEC
  direct-io: use bio_release_pages in dio_bio_complete
  block_dev: use bio_release_pages in bio_unmap_user
  block_dev: use bio_release_pages in blkdev_bio_end_io
  iomap: use bio_release_pages in iomap_dio_bio_end_io
  block: use bio_release_pages in bio_map_user_iov
  block: use bio_release_pages in bio_unmap_user
  block: optionally mark pages dirty in bio_release_pages
  block: move the BIO_NO_PAGE_REF check into bio_release_pages
  block: skd_main.c: Remove call to memset after dma_alloc_coherent
  block: mtip32xx: Remove call to memset after dma_alloc_coherent
  ...

22 files changed, 1342 insertions, 818 deletions

diff --git a/block/Kconfig.iosched b/block/Kconfig.iosched
index 4626b88b2d5a..7a6b2f29a582 100644
--- a/block/Kconfig.iosched
+++ b/block/Kconfig.iosched
@@ -36,6 +36,13 @@ config BFQ_GROUP_IOSCHED
        Enable hierarchical scheduling in BFQ, using the blkio
        (cgroups-v1) or io (cgroups-v2) controller.
 
+config BFQ_CGROUP_DEBUG
+	bool "BFQ IO controller debugging"
+	depends on BFQ_GROUP_IOSCHED
+	---help---
+	Enable some debugging help. Currently it exports additional stat
+	files in a cgroup which can be useful for debugging.
+
 endmenu
 
 endif
diff --git a/block/bfq-cgroup.c b/block/bfq-cgroup.c
index b3796a40a61a..0f6cd688924f 100644
--- a/block/bfq-cgroup.c
+++ b/block/bfq-cgroup.c
@@ -15,7 +15,83 @@
 
 #include "bfq-iosched.h"
 
-#if defined(CONFIG_BFQ_GROUP_IOSCHED) &&  defined(CONFIG_DEBUG_BLK_CGROUP)
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
+static int bfq_stat_init(struct bfq_stat *stat, gfp_t gfp)
+{
+	int ret;
+
+	ret = percpu_counter_init(&stat->cpu_cnt, 0, gfp);
+	if (ret)
+		return ret;
+
+	atomic64_set(&stat->aux_cnt, 0);
+	return 0;
+}
+
+static void bfq_stat_exit(struct bfq_stat *stat)
+{
+	percpu_counter_destroy(&stat->cpu_cnt);
+}
+
+/**
+ * bfq_stat_add - add a value to a bfq_stat
+ * @stat: target bfq_stat
+ * @val: value to add
+ *
+ * Add @val to @stat.  The caller must ensure that IRQ on the same CPU
+ * don't re-enter this function for the same counter.
+ */
+static inline void bfq_stat_add(struct bfq_stat *stat, uint64_t val)
+{
+	percpu_counter_add_batch(&stat->cpu_cnt, val, BLKG_STAT_CPU_BATCH);
+}
+
+/**
+ * bfq_stat_read - read the current value of a bfq_stat
+ * @stat: bfq_stat to read
+ */
+static inline uint64_t bfq_stat_read(struct bfq_stat *stat)
+{
+	return percpu_counter_sum_positive(&stat->cpu_cnt);
+}
+
+/**
+ * bfq_stat_reset - reset a bfq_stat
+ * @stat: bfq_stat to reset
+ */
+static inline void bfq_stat_reset(struct bfq_stat *stat)
+{
+	percpu_counter_set(&stat->cpu_cnt, 0);
+	atomic64_set(&stat->aux_cnt, 0);
+}
+
+/**
+ * bfq_stat_add_aux - add a bfq_stat into another's aux count
+ * @to: the destination bfq_stat
+ * @from: the source
+ *
+ * Add @from's count including the aux one to @to's aux count.
+ */
+static inline void bfq_stat_add_aux(struct bfq_stat *to,
+				     struct bfq_stat *from)
+{
+	atomic64_add(bfq_stat_read(from) + atomic64_read(&from->aux_cnt),
+		     &to->aux_cnt);
+}
+
+/**
+ * blkg_prfill_stat - prfill callback for bfq_stat
+ * @sf: seq_file to print to
+ * @pd: policy private data of interest
+ * @off: offset to the bfq_stat in @pd
+ *
+ * prfill callback for printing a bfq_stat.
+ */
+static u64 blkg_prfill_stat(struct seq_file *sf, struct blkg_policy_data *pd,
+		int off)
+{
+	return __blkg_prfill_u64(sf, pd, bfq_stat_read((void *)pd + off));
+}
 
 /* bfqg stats flags */
 enum bfqg_stats_flags {
@@ -53,7 +129,7 @@ static void bfqg_stats_update_group_wait_time(struct bfqg_stats *stats)
 
 	now = ktime_get_ns();
 	if (now > stats->start_group_wait_time)
-		blkg_stat_add(&stats->group_wait_time,
+		bfq_stat_add(&stats->group_wait_time,
 			      now - stats->start_group_wait_time);
 	bfqg_stats_clear_waiting(stats);
 }
@@ -82,14 +158,14 @@ static void bfqg_stats_end_empty_time(struct bfqg_stats *stats)
 
 	now = ktime_get_ns();
 	if (now > stats->start_empty_time)
-		blkg_stat_add(&stats->empty_time,
+		bfq_stat_add(&stats->empty_time,
 			      now - stats->start_empty_time);
 	bfqg_stats_clear_empty(stats);
 }
 
 void bfqg_stats_update_dequeue(struct bfq_group *bfqg)
 {
-	blkg_stat_add(&bfqg->stats.dequeue, 1);
+	bfq_stat_add(&bfqg->stats.dequeue, 1);
 }
 
 void bfqg_stats_set_start_empty_time(struct bfq_group *bfqg)
@@ -119,7 +195,7 @@ void bfqg_stats_update_idle_time(struct bfq_group *bfqg)
 		u64 now = ktime_get_ns();
 
 		if (now > stats->start_idle_time)
-			blkg_stat_add(&stats->idle_time,
+			bfq_stat_add(&stats->idle_time,
 				      now - stats->start_idle_time);
 		bfqg_stats_clear_idling(stats);
 	}
@@ -137,9 +213,9 @@ void bfqg_stats_update_avg_queue_size(struct bfq_group *bfqg)
 {
 	struct bfqg_stats *stats = &bfqg->stats;
 
-	blkg_stat_add(&stats->avg_queue_size_sum,
+	bfq_stat_add(&stats->avg_queue_size_sum,
 		      blkg_rwstat_total(&stats->queued));
-	blkg_stat_add(&stats->avg_queue_size_samples, 1);
+	bfq_stat_add(&stats->avg_queue_size_samples, 1);
 	bfqg_stats_update_group_wait_time(stats);
 }
 
@@ -176,7 +252,7 @@ void bfqg_stats_update_completion(struct bfq_group *bfqg, u64 start_time_ns,
 				io_start_time_ns - start_time_ns);
 }
 
-#else /* CONFIG_BFQ_GROUP_IOSCHED && CONFIG_DEBUG_BLK_CGROUP */
+#else /* CONFIG_BFQ_CGROUP_DEBUG */
 
 void bfqg_stats_update_io_add(struct bfq_group *bfqg, struct bfq_queue *bfqq,
 			      unsigned int op) { }
@@ -190,7 +266,7 @@ void bfqg_stats_update_idle_time(struct bfq_group *bfqg) { }
 void bfqg_stats_set_start_idle_time(struct bfq_group *bfqg) { }
 void bfqg_stats_update_avg_queue_size(struct bfq_group *bfqg) { }
 
-#endif /* CONFIG_BFQ_GROUP_IOSCHED && CONFIG_DEBUG_BLK_CGROUP */
+#endif /* CONFIG_BFQ_CGROUP_DEBUG */
 
 #ifdef CONFIG_BFQ_GROUP_IOSCHED
 
@@ -274,18 +350,18 @@ void bfqg_and_blkg_put(struct bfq_group *bfqg)
 /* @stats = 0 */
 static void bfqg_stats_reset(struct bfqg_stats *stats)
 {
-#ifdef CONFIG_DEBUG_BLK_CGROUP
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
 	/* queued stats shouldn't be cleared */
 	blkg_rwstat_reset(&stats->merged);
 	blkg_rwstat_reset(&stats->service_time);
 	blkg_rwstat_reset(&stats->wait_time);
-	blkg_stat_reset(&stats->time);
-	blkg_stat_reset(&stats->avg_queue_size_sum);
-	blkg_stat_reset(&stats->avg_queue_size_samples);
-	blkg_stat_reset(&stats->dequeue);
-	blkg_stat_reset(&stats->group_wait_time);
-	blkg_stat_reset(&stats->idle_time);
-	blkg_stat_reset(&stats->empty_time);
+	bfq_stat_reset(&stats->time);
+	bfq_stat_reset(&stats->avg_queue_size_sum);
+	bfq_stat_reset(&stats->avg_queue_size_samples);
+	bfq_stat_reset(&stats->dequeue);
+	bfq_stat_reset(&stats->group_wait_time);
+	bfq_stat_reset(&stats->idle_time);
+	bfq_stat_reset(&stats->empty_time);
 #endif
 }
 
@@ -295,19 +371,19 @@ static void bfqg_stats_add_aux(struct bfqg_stats *to, struct bfqg_stats *from)
 	if (!to || !from)
 		return;
 
-#ifdef CONFIG_DEBUG_BLK_CGROUP
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
 	/* queued stats shouldn't be cleared */
 	blkg_rwstat_add_aux(&to->merged, &from->merged);
 	blkg_rwstat_add_aux(&to->service_time, &from->service_time);
 	blkg_rwstat_add_aux(&to->wait_time, &from->wait_time);
-	blkg_stat_add_aux(&from->time, &from->time);
-	blkg_stat_add_aux(&to->avg_queue_size_sum, &from->avg_queue_size_sum);
-	blkg_stat_add_aux(&to->avg_queue_size_samples,
+	bfq_stat_add_aux(&from->time, &from->time);
+	bfq_stat_add_aux(&to->avg_queue_size_sum, &from->avg_queue_size_sum);
+	bfq_stat_add_aux(&to->avg_queue_size_samples,
 			  &from->avg_queue_size_samples);
-	blkg_stat_add_aux(&to->dequeue, &from->dequeue);
-	blkg_stat_add_aux(&to->group_wait_time, &from->group_wait_time);
-	blkg_stat_add_aux(&to->idle_time, &from->idle_time);
-	blkg_stat_add_aux(&to->empty_time, &from->empty_time);
+	bfq_stat_add_aux(&to->dequeue, &from->dequeue);
+	bfq_stat_add_aux(&to->group_wait_time, &from->group_wait_time);
+	bfq_stat_add_aux(&to->idle_time, &from->idle_time);
+	bfq_stat_add_aux(&to->empty_time, &from->empty_time);
 #endif
 }
 
@@ -355,35 +431,35 @@ void bfq_init_entity(struct bfq_entity *entity, struct bfq_group *bfqg)
 
 static void bfqg_stats_exit(struct bfqg_stats *stats)
 {
-#ifdef CONFIG_DEBUG_BLK_CGROUP
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
 	blkg_rwstat_exit(&stats->merged);
 	blkg_rwstat_exit(&stats->service_time);
 	blkg_rwstat_exit(&stats->wait_time);
 	blkg_rwstat_exit(&stats->queued);
-	blkg_stat_exit(&stats->time);
-	blkg_stat_exit(&stats->avg_queue_size_sum);
-	blkg_stat_exit(&stats->avg_queue_size_samples);
-	blkg_stat_exit(&stats->dequeue);
-	blkg_stat_exit(&stats->group_wait_time);
-	blkg_stat_exit(&stats->idle_time);
-	blkg_stat_exit(&stats->empty_time);
+	bfq_stat_exit(&stats->time);
+	bfq_stat_exit(&stats->avg_queue_size_sum);
+	bfq_stat_exit(&stats->avg_queue_size_samples);
+	bfq_stat_exit(&stats->dequeue);
+	bfq_stat_exit(&stats->group_wait_time);
+	bfq_stat_exit(&stats->idle_time);
+	bfq_stat_exit(&stats->empty_time);
 #endif
 }
 
 static int bfqg_stats_init(struct bfqg_stats *stats, gfp_t gfp)
 {
-#ifdef CONFIG_DEBUG_BLK_CGROUP
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
 	if (blkg_rwstat_init(&stats->merged, gfp) ||
 	    blkg_rwstat_init(&stats->service_time, gfp) ||
 	    blkg_rwstat_init(&stats->wait_time, gfp) ||
 	    blkg_rwstat_init(&stats->queued, gfp) ||
-	    blkg_stat_init(&stats->time, gfp) ||
-	    blkg_stat_init(&stats->avg_queue_size_sum, gfp) ||
-	    blkg_stat_init(&stats->avg_queue_size_samples, gfp) ||
-	    blkg_stat_init(&stats->dequeue, gfp) ||
-	    blkg_stat_init(&stats->group_wait_time, gfp) ||
-	    blkg_stat_init(&stats->idle_time, gfp) ||
-	    blkg_stat_init(&stats->empty_time, gfp)) {
+	    bfq_stat_init(&stats->time, gfp) ||
+	    bfq_stat_init(&stats->avg_queue_size_sum, gfp) ||
+	    bfq_stat_init(&stats->avg_queue_size_samples, gfp) ||
+	    bfq_stat_init(&stats->dequeue, gfp) ||
+	    bfq_stat_init(&stats->group_wait_time, gfp) ||
+	    bfq_stat_init(&stats->idle_time, gfp) ||
+	    bfq_stat_init(&stats->empty_time, gfp)) {
 		bfqg_stats_exit(stats);
 		return -ENOMEM;
 	}
@@ -909,7 +985,7 @@ static ssize_t bfq_io_set_weight(struct kernfs_open_file *of,
 	return ret ?: nbytes;
 }
 
-#ifdef CONFIG_DEBUG_BLK_CGROUP
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
 static int bfqg_print_stat(struct seq_file *sf, void *v)
 {
 	blkcg_print_blkgs(sf, css_to_blkcg(seq_css(sf)), blkg_prfill_stat,
@@ -927,17 +1003,34 @@ static int bfqg_print_rwstat(struct seq_file *sf, void *v)
 static u64 bfqg_prfill_stat_recursive(struct seq_file *sf,
 				      struct blkg_policy_data *pd, int off)
 {
-	u64 sum = blkg_stat_recursive_sum(pd_to_blkg(pd),
-					  &blkcg_policy_bfq, off);
+	struct blkcg_gq *blkg = pd_to_blkg(pd);
+	struct blkcg_gq *pos_blkg;
+	struct cgroup_subsys_state *pos_css;
+	u64 sum = 0;
+
+	lockdep_assert_held(&blkg->q->queue_lock);
+
+	rcu_read_lock();
+	blkg_for_each_descendant_pre(pos_blkg, pos_css, blkg) {
+		struct bfq_stat *stat;
+
+		if (!pos_blkg->online)
+			continue;
+
+		stat = (void *)blkg_to_pd(pos_blkg, &blkcg_policy_bfq) + off;
+		sum += bfq_stat_read(stat) + atomic64_read(&stat->aux_cnt);
+	}
+	rcu_read_unlock();
+
 	return __blkg_prfill_u64(sf, pd, sum);
 }
 
 static u64 bfqg_prfill_rwstat_recursive(struct seq_file *sf,
 					struct blkg_policy_data *pd, int off)
 {
-	struct blkg_rwstat sum = blkg_rwstat_recursive_sum(pd_to_blkg(pd),
-							   &blkcg_policy_bfq,
-							   off);
+	struct blkg_rwstat_sample sum;
+
+	blkg_rwstat_recursive_sum(pd_to_blkg(pd), &blkcg_policy_bfq, off, &sum);
 	return __blkg_prfill_rwstat(sf, pd, &sum);
 }
 
@@ -975,12 +1068,13 @@ static int bfqg_print_stat_sectors(struct seq_file *sf, void *v)
 static u64 bfqg_prfill_sectors_recursive(struct seq_file *sf,
 					 struct blkg_policy_data *pd, int off)
 {
-	struct blkg_rwstat tmp = blkg_rwstat_recursive_sum(pd->blkg, NULL,
-					offsetof(struct blkcg_gq, stat_bytes));
-	u64 sum = atomic64_read(&tmp.aux_cnt[BLKG_RWSTAT_READ]) +
-		atomic64_read(&tmp.aux_cnt[BLKG_RWSTAT_WRITE]);
+	struct blkg_rwstat_sample tmp;
 
-	return __blkg_prfill_u64(sf, pd, sum >> 9);
+	blkg_rwstat_recursive_sum(pd->blkg, NULL,
+			offsetof(struct blkcg_gq, stat_bytes), &tmp);
+
+	return __blkg_prfill_u64(sf, pd,
+		(tmp.cnt[BLKG_RWSTAT_READ] + tmp.cnt[BLKG_RWSTAT_WRITE]) >> 9);
 }
 
 static int bfqg_print_stat_sectors_recursive(struct seq_file *sf, void *v)
@@ -995,11 +1089,11 @@ static u64 bfqg_prfill_avg_queue_size(struct seq_file *sf,
 				      struct blkg_policy_data *pd, int off)
 {
 	struct bfq_group *bfqg = pd_to_bfqg(pd);
-	u64 samples = blkg_stat_read(&bfqg->stats.avg_queue_size_samples);
+	u64 samples = bfq_stat_read(&bfqg->stats.avg_queue_size_samples);
 	u64 v = 0;
 
 	if (samples) {
-		v = blkg_stat_read(&bfqg->stats.avg_queue_size_sum);
+		v = bfq_stat_read(&bfqg->stats.avg_queue_size_sum);
 		v = div64_u64(v, samples);
 	}
 	__blkg_prfill_u64(sf, pd, v);
@@ -1014,7 +1108,7 @@ static int bfqg_print_avg_queue_size(struct seq_file *sf, void *v)
 			  0, false);
 	return 0;
 }
-#endif /* CONFIG_DEBUG_BLK_CGROUP */
+#endif /* CONFIG_BFQ_CGROUP_DEBUG */
 
 struct bfq_group *bfq_create_group_hierarchy(struct bfq_data *bfqd, int node)
 {
@@ -1062,7 +1156,7 @@ struct cftype bfq_blkcg_legacy_files[] = {
 		.private = (unsigned long)&blkcg_policy_bfq,
 		.seq_show = blkg_print_stat_ios,
 	},
-#ifdef CONFIG_DEBUG_BLK_CGROUP
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
 	{
 		.name = "bfq.time",
 		.private = offsetof(struct bfq_group, stats.time),
@@ -1092,7 +1186,7 @@ struct cftype bfq_blkcg_legacy_files[] = {
 		.private = offsetof(struct bfq_group, stats.queued),
 		.seq_show = bfqg_print_rwstat,
 	},
-#endif /* CONFIG_DEBUG_BLK_CGROUP */
+#endif /* CONFIG_BFQ_CGROUP_DEBUG */
 
 	/* the same statistics which cover the bfqg and its descendants */
 	{
@@ -1105,7 +1199,7 @@ struct cftype bfq_blkcg_legacy_files[] = {
 		.private = (unsigned long)&blkcg_policy_bfq,
 		.seq_show = blkg_print_stat_ios_recursive,
 	},
-#ifdef CONFIG_DEBUG_BLK_CGROUP
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
 	{
 		.name = "bfq.time_recursive",
 		.private = offsetof(struct bfq_group, stats.time),
@@ -1159,7 +1253,7 @@ struct cftype bfq_blkcg_legacy_files[] = {
 		.private = offsetof(struct bfq_group, stats.dequeue),
 		.seq_show = bfqg_print_stat,
 	},
-#endif	/* CONFIG_DEBUG_BLK_CGROUP */
+#endif	/* CONFIG_BFQ_CGROUP_DEBUG */
 	{ }	/* terminate */
 };
 
diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
index f9269ae6da9c..50c9d2598500 100644
--- a/block/bfq-iosched.c
+++ b/block/bfq-iosched.c
@@ -157,6 +157,7 @@ BFQ_BFQQ_FNS(in_large_burst);
 BFQ_BFQQ_FNS(coop);
 BFQ_BFQQ_FNS(split_coop);
 BFQ_BFQQ_FNS(softrt_update);
+BFQ_BFQQ_FNS(has_waker);
 #undef BFQ_BFQQ_FNS						\
 
 /* Expiration time of sync (0) and async (1) requests, in ns. */
@@ -1427,17 +1428,19 @@ static int bfq_min_budget(struct bfq_data *bfqd)
  * mechanism may be re-designed in such a way to make it possible to
  * know whether preemption is needed without needing to update service
  * trees). In addition, queue preemptions almost always cause random
- * I/O, and thus loss of throughput. Because of these facts, the next
- * function adopts the following simple scheme to avoid both costly
- * operations and too frequent preemptions: it requests the expiration
- * of the in-service queue (unconditionally) only for queues that need
- * to recover a hole, or that either are weight-raised or deserve to
- * be weight-raised.
+ * I/O, which may in turn cause loss of throughput. Finally, there may
+ * even be no in-service queue when the next function is invoked (so,
+ * no queue to compare timestamps with). Because of these facts, the
+ * next function adopts the following simple scheme to avoid costly
+ * operations, too frequent preemptions and too many dependencies on
+ * the state of the scheduler: it requests the expiration of the
+ * in-service queue (unconditionally) only for queues that need to
+ * recover a hole. Then it delegates to other parts of the code the
+ * responsibility of handling the above case 2.
  */
 static bool bfq_bfqq_update_budg_for_activation(struct bfq_data *bfqd,
 						struct bfq_queue *bfqq,
-						bool arrived_in_time,
-						bool wr_or_deserves_wr)
+						bool arrived_in_time)
 {
 	struct bfq_entity *entity = &bfqq->entity;
 
@@ -1492,7 +1495,7 @@ static bool bfq_bfqq_update_budg_for_activation(struct bfq_data *bfqd,
 	entity->budget = max_t(unsigned long, bfqq->max_budget,
 			       bfq_serv_to_charge(bfqq->next_rq, bfqq));
 	bfq_clear_bfqq_non_blocking_wait_rq(bfqq);
-	return wr_or_deserves_wr;
+	return false;
 }
 
 /*
@@ -1610,6 +1613,36 @@ static bool bfq_bfqq_idle_for_long_time(struct bfq_data *bfqd,
 			bfqd->bfq_wr_min_idle_time);
 }
 
+
+/*
+ * Return true if bfqq is in a higher priority class, or has a higher
+ * weight than the in-service queue.
+ */
+static bool bfq_bfqq_higher_class_or_weight(struct bfq_queue *bfqq,
+					    struct bfq_queue *in_serv_bfqq)
+{
+	int bfqq_weight, in_serv_weight;
+
+	if (bfqq->ioprio_class < in_serv_bfqq->ioprio_class)
+		return true;
+
+	if (in_serv_bfqq->entity.parent == bfqq->entity.parent) {
+		bfqq_weight = bfqq->entity.weight;
+		in_serv_weight = in_serv_bfqq->entity.weight;
+	} else {
+		if (bfqq->entity.parent)
+			bfqq_weight = bfqq->entity.parent->weight;
+		else
+			bfqq_weight = bfqq->entity.weight;
+		if (in_serv_bfqq->entity.parent)
+			in_serv_weight = in_serv_bfqq->entity.parent->weight;
+		else
+			in_serv_weight = in_serv_bfqq->entity.weight;
+	}
+
+	return bfqq_weight > in_serv_weight;
+}
+
 static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
 					     struct bfq_queue *bfqq,
 					     int old_wr_coeff,
@@ -1654,8 +1687,7 @@ static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
 	 */
 	bfqq_wants_to_preempt =
 		bfq_bfqq_update_budg_for_activation(bfqd, bfqq,
-						    arrived_in_time,
-						    wr_or_deserves_wr);
+						    arrived_in_time);
 
 	/*
 	 * If bfqq happened to be activated in a burst, but has been
@@ -1720,21 +1752,111 @@ static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
 
 	/*
 	 * Expire in-service queue only if preemption may be needed
-	 * for guarantees. In this respect, the function
-	 * next_queue_may_preempt just checks a simple, necessary
-	 * condition, and not a sufficient condition based on
-	 * timestamps. In fact, for the latter condition to be
-	 * evaluated, timestamps would need first to be updated, and
-	 * this operation is quite costly (see the comments on the
-	 * function bfq_bfqq_update_budg_for_activation).
+	 * for guarantees. In particular, we care only about two
+	 * cases. The first is that bfqq has to recover a service
+	 * hole, as explained in the comments on
+	 * bfq_bfqq_update_budg_for_activation(), i.e., that
+	 * bfqq_wants_to_preempt is true. However, if bfqq does not
+	 * carry time-critical I/O, then bfqq's bandwidth is less
+	 * important than that of queues that carry time-critical I/O.
+	 * So, as a further constraint, we consider this case only if
+	 * bfqq is at least as weight-raised, i.e., at least as time
+	 * critical, as the in-service queue.
+	 *
+	 * The second case is that bfqq is in a higher priority class,
+	 * or has a higher weight than the in-service queue. If this
+	 * condition does not hold, we don't care because, even if
+	 * bfqq does not start to be served immediately, the resulting
+	 * delay for bfqq's I/O is however lower or much lower than
+	 * the ideal completion time to be guaranteed to bfqq's I/O.
+	 *
+	 * In both cases, preemption is needed only if, according to
+	 * the timestamps of both bfqq and of the in-service queue,
+	 * bfqq actually is the next queue to serve. So, to reduce
+	 * useless preemptions, the return value of
+	 * next_queue_may_preempt() is considered in the next compound
+	 * condition too. Yet next_queue_may_preempt() just checks a
+	 * simple, necessary condition for bfqq to be the next queue
+	 * to serve. In fact, to evaluate a sufficient condition, the
+	 * timestamps of the in-service queue would need to be
+	 * updated, and this operation is quite costly (see the
+	 * comments on bfq_bfqq_update_budg_for_activation()).
 	 */
-	if (bfqd->in_service_queue && bfqq_wants_to_preempt &&
-	    bfqd->in_service_queue->wr_coeff < bfqq->wr_coeff &&
+	if (bfqd->in_service_queue &&
+	    ((bfqq_wants_to_preempt &&
+	      bfqq->wr_coeff >= bfqd->in_service_queue->wr_coeff) ||
+	     bfq_bfqq_higher_class_or_weight(bfqq, bfqd->in_service_queue)) &&
 	    next_queue_may_preempt(bfqd))
 		bfq_bfqq_expire(bfqd, bfqd->in_service_queue,
 				false, BFQQE_PREEMPTED);
 }
 
+static void bfq_reset_inject_limit(struct bfq_data *bfqd,
+				   struct bfq_queue *bfqq)
+{
+	/* invalidate baseline total service time */
+	bfqq->last_serv_time_ns = 0;
+
+	/*
+	 * Reset pointer in case we are waiting for
+	 * some request completion.
+	 */
+	bfqd->waited_rq = NULL;
+
+	/*
+	 * If bfqq has a short think time, then start by setting the
+	 * inject limit to 0 prudentially, because the service time of
+	 * an injected I/O request may be higher than the think time
+	 * of bfqq, and therefore, if one request was injected when
+	 * bfqq remains empty, this injected request might delay the
+	 * service of the next I/O request for bfqq significantly. In
+	 * case bfqq can actually tolerate some injection, then the
+	 * adaptive update will however raise the limit soon. This
+	 * lucky circumstance holds exactly because bfqq has a short
+	 * think time, and thus, after remaining empty, is likely to
+	 * get new I/O enqueued---and then completed---before being
+	 * expired. This is the very pattern that gives the
+	 * limit-update algorithm the chance to measure the effect of
+	 * injection on request service times, and then to update the
+	 * limit accordingly.
+	 *
+	 * However, in the following special case, the inject limit is
+	 * left to 1 even if the think time is short: bfqq's I/O is
+	 * synchronized with that of some other queue, i.e., bfqq may
+	 * receive new I/O only after the I/O of the other queue is
+	 * completed. Keeping the inject limit to 1 allows the
+	 * blocking I/O to be served while bfqq is in service. And
+	 * this is very convenient both for bfqq and for overall
+	 * throughput, as explained in detail in the comments in
+	 * bfq_update_has_short_ttime().
+	 *
+	 * On the opposite end, if bfqq has a long think time, then
+	 * start directly by 1, because:
+	 * a) on the bright side, keeping at most one request in
+	 * service in the drive is unlikely to cause any harm to the
+	 * latency of bfqq's requests, as the service time of a single
+	 * request is likely to be lower than the think time of bfqq;
+	 * b) on the downside, after becoming empty, bfqq is likely to
+	 * expire before getting its next request. With this request
+	 * arrival pattern, it is very hard to sample total service
+	 * times and update the inject limit accordingly (see comments
+	 * on bfq_update_inject_limit()). So the limit is likely to be
+	 * never, or at least seldom, updated.  As a consequence, by
+	 * setting the limit to 1, we avoid that no injection ever
+	 * occurs with bfqq. On the downside, this proactive step
+	 * further reduces chances to actually compute the baseline
+	 * total service time. Thus it reduces chances to execute the
+	 * limit-update algorithm and possibly raise the limit to more
+	 * than 1.
+	 */
+	if (bfq_bfqq_has_short_ttime(bfqq))
+		bfqq->inject_limit = 0;
+	else
+		bfqq->inject_limit = 1;
+
+	bfqq->decrease_time_jif = jiffies;
+}
+
 static void bfq_add_request(struct request *rq)
 {
 	struct bfq_queue *bfqq = RQ_BFQQ(rq);
@@ -1749,77 +1871,119 @@ static void bfq_add_request(struct request *rq)
 
 	if (RB_EMPTY_ROOT(&bfqq->sort_list) && bfq_bfqq_sync(bfqq)) {
 		/*
+		 * Detect whether bfqq's I/O seems synchronized with
+		 * that of some other queue, i.e., whether bfqq, after
+		 * remaining empty, happens to receive new I/O only
+		 * right after some I/O request of the other queue has
+		 * been completed. We call waker queue the other
+		 * queue, and we assume, for simplicity, that bfqq may
+		 * have at most one waker queue.
+		 *
+		 * A remarkable throughput boost can be reached by
+		 * unconditionally injecting the I/O of the waker
+		 * queue, every time a new bfq_dispatch_request
+		 * happens to be invoked while I/O is being plugged
+		 * for bfqq.  In addition to boosting throughput, this
+		 * unblocks bfqq's I/O, thereby improving bandwidth
+		 * and latency for bfqq. Note that these same results
+		 * may be achieved with the general injection
+		 * mechanism, but less effectively. For details on
+		 * this aspect, see the comments on the choice of the
+		 * queue for injection in bfq_select_queue().
+		 *
+		 * Turning back to the detection of a waker queue, a
+		 * queue Q is deemed as a waker queue for bfqq if, for
+		 * two consecutive times, bfqq happens to become non
+		 * empty right after a request of Q has been
+		 * completed. In particular, on the first time, Q is
+		 * tentatively set as a candidate waker queue, while
+		 * on the second time, the flag
+		 * bfq_bfqq_has_waker(bfqq) is set to confirm that Q
+		 * is a waker queue for bfqq. These detection steps
+		 * are performed only if bfqq has a long think time,
+		 * so as to make it more likely that bfqq's I/O is
+		 * actually being blocked by a synchronization. This
+		 * last filter, plus the above two-times requirement,
+		 * make false positives less likely.
+		 *
+		 * NOTE
+		 *
+		 * The sooner a waker queue is detected, the sooner
+		 * throughput can be boosted by injecting I/O from the
+		 * waker queue. Fortunately, detection is likely to be
+		 * actually fast, for the following reasons. While
+		 * blocked by synchronization, bfqq has a long think
+		 * time. This implies that bfqq's inject limit is at
+		 * least equal to 1 (see the comments in
+		 * bfq_update_inject_limit()). So, thanks to
+		 * injection, the waker queue is likely to be served
+		 * during the very first I/O-plugging time interval
+		 * for bfqq. This triggers the first step of the
+		 * detection mechanism. Thanks again to injection, the
+		 * candidate waker queue is then likely to be
+		 * confirmed no later than during the next
+		 * I/O-plugging interval for bfqq.
+		 */
+		if (!bfq_bfqq_has_short_ttime(bfqq) &&
+		    ktime_get_ns() - bfqd->last_completion <
+		    200 * NSEC_PER_USEC) {
+			if (bfqd->last_completed_rq_bfqq != bfqq &&
+				   bfqd->last_completed_rq_bfqq !=
+				   bfqq->waker_bfqq) {
+				/*
+				 * First synchronization detected with
+				 * a candidate waker queue, or with a
+				 * different candidate waker queue
+				 * from the current one.
+				 */
+				bfqq->waker_bfqq = bfqd->last_completed_rq_bfqq;
+
+				/*
+				 * If the waker queue disappears, then
+				 * bfqq->waker_bfqq must be reset. To
+				 * this goal, we maintain in each
+				 * waker queue a list, woken_list, of
+				 * all the queues that reference the
+				 * waker queue through their
+				 * waker_bfqq pointer. When the waker
+				 * queue exits, the waker_bfqq pointer
+				 * of all the queues in the woken_list
+				 * is reset.
+				 *
+				 * In addition, if bfqq is already in
+				 * the woken_list of a waker queue,
+				 * then, before being inserted into
+				 * the woken_list of a new waker
+				 * queue, bfqq must be removed from
+				 * the woken_list of the old waker
+				 * queue.
+				 */
+				if (!hlist_unhashed(&bfqq->woken_list_node))
+					hlist_del_init(&bfqq->woken_list_node);
+				hlist_add_head(&bfqq->woken_list_node,
+				    &bfqd->last_completed_rq_bfqq->woken_list);
+
+				bfq_clear_bfqq_has_waker(bfqq);
+			} else if (bfqd->last_completed_rq_bfqq ==
+				   bfqq->waker_bfqq &&
+				   !bfq_bfqq_has_waker(bfqq)) {
+				/*
+				 * synchronization with waker_bfqq
+				 * seen for the second time
+				 */
+				bfq_mark_bfqq_has_waker(bfqq);
+			}
+		}
+
+		/*
 		 * Periodically reset inject limit, to make sure that
 		 * the latter eventually drops in case workload
 		 * changes, see step (3) in the comments on
 		 * bfq_update_inject_limit().
 		 */
 		if (time_is_before_eq_jiffies(bfqq->decrease_time_jif +
-					     msecs_to_jiffies(1000))) {
-			/* invalidate baseline total service time */
-			bfqq->last_serv_time_ns = 0;