Merge remote-tracking branch 'jens/for-4.2/core' into dm-4.2

block/bio-integrity.c

@@ -361,7 +361,7 @@ static void bio_integrity_verify_fn(struct work_struct *work)
 
 	/* Restore original bio completion handler */
 	bio->bi_end_io = bip->bip_end_io;
-	bio_endio_nodec(bio, error);
+	bio_endio(bio, error);
 }
 
 /**
@@ -388,7 +388,7 @@ void bio_integrity_endio(struct bio *bio, int error)
 	 */
 	if (error) {
 		bio->bi_end_io = bip->bip_end_io;
-		bio_endio_nodec(bio, error);
+		bio_endio(bio, error);
 
 		return;
 	}

block/bio.c

@@ -270,8 +270,8 @@ void bio_init(struct bio *bio)
 {
 	memset(bio, 0, sizeof(*bio));
 	bio->bi_flags = 1 << BIO_UPTODATE;
-	atomic_set(&bio->bi_remaining, 1);
-	atomic_set(&bio->bi_cnt, 1);
+	atomic_set(&bio->__bi_remaining, 1);
+	atomic_set(&bio->__bi_cnt, 1);
 }
 EXPORT_SYMBOL(bio_init);
 
@@ -292,8 +292,8 @@ void bio_reset(struct bio *bio)
 	__bio_free(bio);
 
 	memset(bio, 0, BIO_RESET_BYTES);
-	bio->bi_flags = flags|(1 << BIO_UPTODATE);
-	atomic_set(&bio->bi_remaining, 1);
+	bio->bi_flags = flags | (1 << BIO_UPTODATE);
+	atomic_set(&bio->__bi_remaining, 1);
 }
 EXPORT_SYMBOL(bio_reset);
 
@@ -303,6 +303,17 @@ static void bio_chain_endio(struct bio *bio, int error)
 	bio_put(bio);
 }
 
+/*
+ * Increment chain count for the bio. Make sure the CHAIN flag update
+ * is visible before the raised count.
+ */
+static inline void bio_inc_remaining(struct bio *bio)
+{
+	bio->bi_flags |= (1 << BIO_CHAIN);
+	smp_mb__before_atomic();
+	atomic_inc(&bio->__bi_remaining);
+}
+
 /**
  * bio_chain - chain bio completions
  * @bio: the target bio
@@ -320,7 +331,7 @@ void bio_chain(struct bio *bio, struct bio *parent)
 
 	bio->bi_private = parent;
 	bio->bi_end_io	= bio_chain_endio;
-	atomic_inc(&parent->bi_remaining);
+	bio_inc_remaining(parent);
 }
 EXPORT_SYMBOL(bio_chain);
 
@@ -524,13 +535,17 @@ EXPORT_SYMBOL(zero_fill_bio);
  **/
 void bio_put(struct bio *bio)
 {
-	BIO_BUG_ON(!atomic_read(&bio->bi_cnt));
-
-	/*
-	 * last put frees it
-	 */
-	if (atomic_dec_and_test(&bio->bi_cnt))
+	if (!bio_flagged(bio, BIO_REFFED))
 		bio_free(bio);
+	else {
+		BIO_BUG_ON(!atomic_read(&bio->__bi_cnt));
+
+		/*
+		 * last put frees it
+		 */
+		if (atomic_dec_and_test(&bio->__bi_cnt))
+			bio_free(bio);
+	}
 }
 EXPORT_SYMBOL(bio_put);
 
@@ -1741,6 +1756,25 @@ void bio_flush_dcache_pages(struct bio *bi)
 EXPORT_SYMBOL(bio_flush_dcache_pages);
 #endif
 
+static inline bool bio_remaining_done(struct bio *bio)
+{
+	/*
+	 * If we're not chaining, then ->__bi_remaining is always 1 and
+	 * we always end io on the first invocation.
+	 */
+	if (!bio_flagged(bio, BIO_CHAIN))
+		return true;
+
+	BUG_ON(atomic_read(&bio->__bi_remaining) <= 0);
+
+	if (atomic_dec_and_test(&bio->__bi_remaining)) {
+		clear_bit(BIO_CHAIN, &bio->bi_flags);
+		return true;
+	}
+
+	return false;
+}
+
 /**
  * bio_endio - end I/O on a bio
  * @bio:	bio
@@ -1758,15 +1792,13 @@ EXPORT_SYMBOL(bio_flush_dcache_pages);
 void bio_endio(struct bio *bio, int error)
 {
 	while (bio) {
-		BUG_ON(atomic_read(&bio->bi_remaining) <= 0);
-
 		if (error)
 			clear_bit(BIO_UPTODATE, &bio->bi_flags);
 		else if (!test_bit(BIO_UPTODATE, &bio->bi_flags))
 			error = -EIO;
 
-		if (!atomic_dec_and_test(&bio->bi_remaining))
-			return;
+		if (unlikely(!bio_remaining_done(bio)))
+			break;
 
 		/*
 		 * Need to have a real endio function for chained bios,
@@ -1789,21 +1821,6 @@ void bio_endio(struct bio *bio, int error)
 }
 EXPORT_SYMBOL(bio_endio);
 
-/**
- * bio_endio_nodec - end I/O on a bio, without decrementing bi_remaining
- * @bio:	bio
- * @error:	error, if any
- *
- * For code that has saved and restored bi_end_io; thing hard before using this
- * function, probably you should've cloned the entire bio.
- **/
-void bio_endio_nodec(struct bio *bio, int error)
-{
-	atomic_inc(&bio->bi_remaining);
-	bio_endio(bio, error);
-}
-EXPORT_SYMBOL(bio_endio_nodec);
-
 /**
  * bio_split - split a bio
  * @bio:	bio to split

block/blk-core.c

@@ -117,7 +117,7 @@ EXPORT_SYMBOL(blk_rq_init);
 static void req_bio_endio(struct request *rq, struct bio *bio,
 			  unsigned int nbytes, int error)
 {
-	if (error)
+	if (error && !(rq->cmd_flags & REQ_CLONE))
 		clear_bit(BIO_UPTODATE, &bio->bi_flags);
 	else if (!test_bit(BIO_UPTODATE, &bio->bi_flags))
 		error = -EIO;
@@ -128,7 +128,8 @@ static void req_bio_endio(struct request *rq, struct bio *bio,
 	bio_advance(bio, nbytes);
 
 	/* don't actually finish bio if it's part of flush sequence */
-	if (bio->bi_iter.bi_size == 0 && !(rq->cmd_flags & REQ_FLUSH_SEQ))
+	if (bio->bi_iter.bi_size == 0 &&
+	    !(rq->cmd_flags & (REQ_FLUSH_SEQ|REQ_CLONE)))
 		bio_endio(bio, error);
 }
 
@@ -285,6 +286,7 @@ inline void __blk_run_queue_uncond(struct request_queue *q)
 	q->request_fn(q);
 	q->request_fn_active--;
 }
+EXPORT_SYMBOL_GPL(__blk_run_queue_uncond);
 
 /**
  * __blk_run_queue - run a single device queue
@@ -1525,7 +1527,8 @@ bool bio_attempt_front_merge(struct request_queue *q, struct request *req,
  * Caller must ensure !blk_queue_nomerges(q) beforehand.
  */
 bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio,
-			    unsigned int *request_count)
+			    unsigned int *request_count,
+			    struct request **same_queue_rq)
 {
 	struct blk_plug *plug;
 	struct request *rq;
@@ -1545,8 +1548,16 @@ bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio,
 	list_for_each_entry_reverse(rq, plug_list, queuelist) {
 		int el_ret;
 
-		if (rq->q == q)
+		if (rq->q == q) {
 			(*request_count)++;
+			/*
+			 * Only blk-mq multiple hardware queues case checks the
+			 * rq in the same queue, there should be only one such
+			 * rq in a queue
+			 **/
+			if (same_queue_rq)
+				*same_queue_rq = rq;
+		}
 
 		if (rq->q != q || !blk_rq_merge_ok(rq, bio))
 			continue;
@@ -1611,7 +1622,7 @@ static void blk_queue_bio(struct request_queue *q, struct bio *bio)
 	 * any locks.
 	 */
 	if (!blk_queue_nomerges(q) &&
-	    blk_attempt_plug_merge(q, bio, &request_count))
+	    blk_attempt_plug_merge(q, bio, &request_count, NULL))
 		return;
 
 	spin_lock_irq(q->queue_lock);
@@ -1718,8 +1729,6 @@ static void handle_bad_sector(struct bio *bio)
 			bio->bi_rw,
 			(unsigned long long)bio_end_sector(bio),
 			(long long)(i_size_read(bio->bi_bdev->bd_inode) >> 9));
-
-	set_bit(BIO_EOF, &bio->bi_flags);
 }
 
 #ifdef CONFIG_FAIL_MAKE_REQUEST
@@ -2904,95 +2913,22 @@ int blk_lld_busy(struct request_queue *q)
 }
 EXPORT_SYMBOL_GPL(blk_lld_busy);
 
-/**
- * blk_rq_unprep_clone - Helper function to free all bios in a cloned request
- * @rq: the clone request to be cleaned up
- *
- * Description:
- *     Free all bios in @rq for a cloned request.
- */
-void blk_rq_unprep_clone(struct request *rq)
-{
-	struct bio *bio;
-
-	while ((bio = rq->bio) != NULL) {
-		rq->bio = bio->bi_next;
-
-		bio_put(bio);
-	}
-}
-EXPORT_SYMBOL_GPL(blk_rq_unprep_clone);
-
-/*
- * Copy attributes of the original request to the clone request.
- * The actual data parts (e.g. ->cmd, ->sense) are not copied.
- */
-static void __blk_rq_prep_clone(struct request *dst, struct request *src)
+void blk_rq_prep_clone(struct request *dst, struct request *src)
 {
 	dst->cpu = src->cpu;
-	dst->cmd_flags |= (src->cmd_flags & REQ_CLONE_MASK) | REQ_NOMERGE;
+	dst->cmd_flags |= (src->cmd_flags & REQ_CLONE_MASK);
+	dst->cmd_flags |= REQ_NOMERGE | REQ_CLONE;
 	dst->cmd_type = src->cmd_type;
 	dst->__sector = blk_rq_pos(src);
 	dst->__data_len = blk_rq_bytes(src);
 	dst->nr_phys_segments = src->nr_phys_segments;
 	dst->ioprio = src->ioprio;
 	dst->extra_len = src->extra_len;
-}
-
-/**
- * blk_rq_prep_clone - Helper function to setup clone request
- * @rq: the request to be setup
- * @rq_src: original request to be cloned
- * @bs: bio_set that bios for clone are allocated from
- * @gfp_mask: memory allocation mask for bio
- * @bio_ctr: setup function to be called for each clone bio.
- *           Returns %0 for success, non %0 for failure.
- * @data: private data to be passed to @bio_ctr
- *
- * Description:
- *     Clones bios in @rq_src to @rq, and copies attributes of @rq_src to @rq.
- *     The actual data parts of @rq_src (e.g. ->cmd, ->sense)
- *     are not copied, and copying such parts is the caller's responsibility.
- *     Also, pages which the original bios are pointing to are not copied
- *     and the cloned bios just point same pages.
- *     So cloned bios must be completed before original bios, which means
- *     the caller must complete @rq before @rq_src.
- */
-int blk_rq_prep_clone(struct request *rq, struct request *rq_src,
-		      struct bio_set *bs, gfp_t gfp_mask,
-		      int (*bio_ctr)(struct bio *, struct bio *, void *),
-		      void *data)
-{
-	struct bio *bio, *bio_src;
-
-	if (!bs)
-		bs = fs_bio_set;
-
-	__rq_for_each_bio(bio_src, rq_src) {
-		bio = bio_clone_fast(bio_src, gfp_mask, bs);
-		if (!bio)
-			goto free_and_out;
-
-		if (bio_ctr && bio_ctr(bio, bio_src, data))
-			goto free_and_out;
-
-		if (rq->bio) {
-			rq->biotail->bi_next = bio;
-			rq->biotail = bio;
-		} else
-			rq->bio = rq->biotail = bio;
-	}
-
-	__blk_rq_prep_clone(rq, rq_src);
-
-	return 0;
-
-free_and_out:
-	if (bio)
-		bio_put(bio);
-	blk_rq_unprep_clone(rq);
-
-	return -ENOMEM;
+	dst->bio = src->bio;
+	dst->biotail = src->biotail;
+	dst->cmd = src->cmd;
+	dst->cmd_len = src->cmd_len;
+	dst->sense = src->sense;
 }
 EXPORT_SYMBOL_GPL(blk_rq_prep_clone);
 
@@ -3034,21 +2970,20 @@ void blk_start_plug(struct blk_plug *plug)
 {
 	struct task_struct *tsk = current;
 
+	/*
+	 * If this is a nested plug, don't actually assign it.
+	 */
+	if (tsk->plug)
+		return;
+
 	INIT_LIST_HEAD(&plug->list);
 	INIT_LIST_HEAD(&plug->mq_list);
 	INIT_LIST_HEAD(&plug->cb_list);
-
 	/*
-	 * If this is a nested plug, don't actually assign it. It will be
-	 * flushed on its own.
+	 * Store ordering should not be needed here, since a potential
+	 * preempt will imply a full memory barrier
 	 */
-	if (!tsk->plug) {
-		/*
-		 * Store ordering should not be needed here, since a potential
-		 * preempt will imply a full memory barrier
-		 */
-		tsk->plug = plug;
-	}
+	tsk->plug = plug;
 }
 EXPORT_SYMBOL(blk_start_plug);
 
@@ -3195,10 +3130,11 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule)
 
 void blk_finish_plug(struct blk_plug *plug)
 {
+	if (plug != current->plug)
+		return;
 	blk_flush_plug_list(plug, false);
 
-	if (plug == current->plug)
-		current->plug = NULL;
+	current->plug = NULL;
 }
 EXPORT_SYMBOL(blk_finish_plug);
 

block/blk-exec.c

@@ -53,7 +53,6 @@ void blk_execute_rq_nowait(struct request_queue *q, struct gendisk *bd_disk,
 			   rq_end_io_fn *done)
 {
 	int where = at_head ? ELEVATOR_INSERT_FRONT : ELEVATOR_INSERT_BACK;
-	bool is_pm_resume;
 
 	WARN_ON(irqs_disabled());
 	WARN_ON(rq->cmd_type == REQ_TYPE_FS);
@@ -70,12 +69,6 @@ void blk_execute_rq_nowait(struct request_queue *q, struct gendisk *bd_disk,
 		return;
 	}
 
-	/*
-	 * need to check this before __blk_run_queue(), because rq can
-	 * be freed before that returns.
-	 */
-	is_pm_resume = rq->cmd_type == REQ_TYPE_PM_RESUME;
-
 	spin_lock_irq(q->queue_lock);
 
 	if (unlikely(blk_queue_dying(q))) {
@@ -88,9 +81,6 @@ void blk_execute_rq_nowait(struct request_queue *q, struct gendisk *bd_disk,
 
 	__elv_add_request(q, rq, where);
 	__blk_run_queue(q);
-	/* the queue is stopped so it won't be run */
-	if (is_pm_resume)
-		__blk_run_queue_uncond(q);
 	spin_unlock_irq(q->queue_lock);
 }
 EXPORT_SYMBOL_GPL(blk_execute_rq_nowait);

block/blk-mq.c

@@ -89,7 +89,8 @@ static int blk_mq_queue_enter(struct request_queue *q, gfp_t gfp)
 			return -EBUSY;
 
 		ret = wait_event_interruptible(q->mq_freeze_wq,
-				!q->mq_freeze_depth || blk_queue_dying(q));
+				!atomic_read(&q->mq_freeze_depth) ||
+				blk_queue_dying(q));
 		if (blk_queue_dying(q))
 			return -ENODEV;
 		if (ret)
@@ -112,13 +113,10 @@ static void blk_mq_usage_counter_release(struct percpu_ref *ref)
 
 void blk_mq_freeze_queue_start(struct request_queue *q)
 {
-	bool freeze;
+	int freeze_depth;
 
-	spin_lock_irq(q->queue_lock);
-	freeze = !q->mq_freeze_depth++;
-	spin_unlock_irq(q->queue_lock);
-
-	if (freeze) {
+	freeze_depth = atomic_inc_return(&q->mq_freeze_depth);
+	if (freeze_depth == 1) {
 		percpu_ref_kill(&q->mq_usage_counter);
 		blk_mq_run_hw_queues(q, false);
 	}
@@ -143,13 +141,11 @@ EXPORT_SYMBOL_GPL(blk_mq_freeze_queue);
 
 void blk_mq_unfreeze_queue(struct request_queue *q)
 {
-	bool wake;
+	int freeze_depth;
 
-	spin_lock_irq(q->queue_lock);
-	wake = !--q->mq_freeze_depth;
-	WARN_ON_ONCE(q->mq_freeze_depth < 0);
-	spin_unlock_irq(q->queue_lock);
-	if (wake) {
+	freeze_depth = atomic_dec_return(&q->mq_freeze_depth);
+	WARN_ON_ONCE(freeze_depth < 0);
+	if (!freeze_depth) {
 		percpu_ref_reinit(&q->mq_usage_counter);
 		wake_up_all(&q->mq_freeze_wq);
 	}
@@ -1237,6 +1233,38 @@ static struct request *blk_mq_map_request(struct request_queue *q,
 	return rq;
 }
 
+static int blk_mq_direct_issue_request(struct request *rq)
+{
+	int ret;
+	struct request_queue *q = rq->q;
+	struct blk_mq_hw_ctx *hctx = q->mq_ops->map_queue(q,
+			rq->mq_ctx->cpu);
+	struct blk_mq_queue_data bd = {
+		.rq = rq,
+		.list = NULL,
+		.last = 1
+	};
+
+	/*
+	 * For OK queue, we are done. For error, kill it. Any other
+	 * error (busy), just add it to our list as we previously
+	 * would have done
+	 */
+	ret = q->mq_ops->queue_rq(hctx, &bd);
+	if (ret == BLK_MQ_RQ_QUEUE_OK)
+		return 0;
+	else {
+		__blk_mq_requeue_request(rq);
+
+		if (ret == BLK_MQ_RQ_QUEUE_ERROR) {
+			rq->errors = -EIO;
+			blk_mq_end_request(rq, rq->errors);
+			return 0;
+		}
+		return -1;
+	}
+}
+
 /*
  * Multiple hardware queue variant. This will not use per-process plugs,
  * but will attempt to bypass the hctx queueing if we can go straight to
@@ -1248,6 +1276,9 @@ static void blk_mq_make_request(struct request_queue *q, struct bio *bio)
 	const int is_flush_fua = bio->bi_rw & (REQ_FLUSH | REQ_FUA);
 	struct blk_map_ctx data;
 	struct request *rq;
+	unsigned int request_count = 0;
+	struct blk_plug *plug;
+	struct request *same_queue_rq = NULL;
 
 	blk_queue_bounce(q, &bio);
 
@@ -1256,6 +1287,10 @@ static void blk_mq_make_request(struct request_queue *q, struct bio *bio)
 		return;
 	}
 
+	if (!is_flush_fua && !blk_queue_nomerges(q) &&
+	    blk_attempt_plug_merge(q, bio, &request_count, &same_queue_rq))
+		return;
+
 	rq = blk_mq_map_request(q, bio, &data);
 	if (unlikely(!rq))
 		return;
@@ -1266,38 +1301,42 @@ static void blk_mq_make_request(struct request_queue *q, struct bio *bio)
 		goto run_queue;
 	}
 
+	plug = current->plug;
 	/*
 	 * If the driver supports defer issued based on 'last', then
 	 * queue it up like normal since we can potentially save some
 	 * CPU this way.
 	 */
-	if (is_sync && !(data.hctx->flags & BLK_MQ_F_DEFER_ISSUE)) {
-		struct blk_mq_queue_data bd = {
-			.rq = rq,
-			.list = NULL,
-			.last = 1
-		};
-		int ret;
+	if (((plug && !blk_queue_nomerges(q)) || is_sync) &&
+	    !(data.hctx->flags & BLK_MQ_F_DEFER_ISSUE)) {
+		struct request *old_rq = NULL;
 
 		blk_mq_bio_to_request(rq, bio);
 
 		/*
-		 * For OK queue, we are done. For error, kill it. Any other
-		 * error (busy), just add it to our list as we previously
-		 * would have done
+		 * we do limited pluging. If bio can be merged, do merge.
+		 * Otherwise the existing request in the plug list will be
+		 * issued. So the plug list will have one request at most
 		 */
-		ret = q->mq_ops->queue_rq(data.hctx, &bd);
-		if (ret == BLK_MQ_RQ_QUEUE_OK)
-			goto done;
-		else {
-			__blk_mq_requeue_request(rq);
-
-			if (ret == BLK_MQ_RQ_QUEUE_ERROR) {
-				rq->errors = -EIO;
-				blk_mq_end_request(rq, rq->errors);
-				goto done;
+		if (plug) {
+			/*
+			 * The plug list might get flushed before this. If that
+			 * happens, same_queue_rq is invalid and plug list is empty
+			 **/
+			if (same_queue_rq && !list_empty(&plug->mq_list)) {
+				old_rq = same_queue_rq;
+				list_del_init(&old_rq->queuelist);
 			}
-		}
+			list_add_tail(&rq->queuelist, &plug->mq_list);
+		} else /* is_sync */
+			old_rq = rq;
+		blk_mq_put_ctx(data.ctx);
+		if (!old_rq)
+			return;
+		if (!blk_mq_direct_issue_request(old_rq))
+			return;
+		blk_mq_insert_request(old_rq, false, true, true);
+		return;
 	}
 
 	if (!blk_mq_merge_queue_io(data.hctx, data.ctx, rq, bio)) {
@@ -1310,7 +1349,6 @@ static void blk_mq_make_request(struct request_queue *q, struct bio *bio)
 run_queue:
 		blk_mq_run_hw_queue(data.hctx, !is_sync || is_flush_fua);
 	}
-done:
 	blk_mq_put_ctx(data.ctx);
 }
 
@@ -1322,16 +1360,11 @@ static void blk_sq_make_request(struct request_queue *q, struct bio *bio)
 {
 	const int is_sync = rw_is_sync(bio->bi_rw);
 	const int is_flush_fua = bio->bi_rw & (REQ_FLUSH | REQ_FUA);
-	unsigned int use_plug, request_count = 0;
+	struct blk_plug *plug;
+	unsigned int request_count = 0;
 	struct blk_map_ctx data;
 	struct request *rq;
 
-	/*
-	 * If we have multiple hardware queues, just go directly to
-	 * one of those for sync IO.
-	 */
-	use_plug = !is_flush_fua && !is_sync;
-
 	blk_queue_bounce(q, &bio);
 
 	if (bio_integrity_enabled(bio) && bio_integrity_prep(bio)) {
@@ -1339,8 +1372,8 @@ static void blk_sq_make_request(struct request_queue *q, struct bio *bio)
 		return;
 	}
 
-	if (use_plug && !blk_queue_nomerges(q) &&
-	    blk_attempt_plug_merge(q, bio, &request_count))
+	if (!is_flush_fua && !blk_queue_nomerges(q) &&
+	    blk_attempt_plug_merge(q, bio, &request_count, NULL))
 		return;
 
 	rq = blk_mq_map_request(q, bio, &data);
@@ -1358,21 +1391,18 @@ static void blk_sq_make_request(struct request_queue *q, struct bio *bio)
 	 * utilize that to temporarily store requests until the task is
 	 * either done or scheduled away.
 	 */
-	if (use_plug) {
-		struct blk_plug *plug = current->plug;
-
-		if (plug) {
-			blk_mq_bio_to_request(rq, bio);
-			if (list_empty(&plug->mq_list))
-				trace_block_plug(q);
-			else if (request_count >= BLK_MAX_REQUEST_COUNT) {
-				blk_flush_plug_list(plug, false);
-				trace_block_plug(q);
-			}
-			list_add_tail(&rq->queuelist, &plug->mq_list);
-			blk_mq_put_ctx(data.ctx);
-			return;
+	plug = current->plug;
+	if (plug) {
+		blk_mq_bio_to_request(rq, bio);
+		if (list_empty(&plug->mq_list))
+			trace_block_plug(q);
+		else if (request_count >= BLK_MAX_REQUEST_COUNT) {
+			blk_flush_plug_list(plug, false);
+			trace_block_plug(q);
 		}
+		list_add_tail(&rq->queuelist, &plug->mq_list);
+		blk_mq_put_ctx(data.ctx);
+		return;
 	}
 
 	if (!blk_mq_merge_queue_io(data.hctx, data.ctx, rq, bio)) {
@@ -2052,7 +2082,7 @@ void blk_mq_free_queue(struct request_queue *q)
 /* Basically redo blk_mq_init_queue with queue frozen */
 static void blk_mq_queue_reinit(struct request_queue *q)
 {
-	WARN_ON_ONCE(!q->mq_freeze_depth);
+	WARN_ON_ONCE(!atomic_read(&q->mq_freeze_depth));
 
 	blk_mq_sysfs_unregister(q);
 

block/blk.h

@@ -78,7 +78,8 @@ bool bio_attempt_front_merge(struct request_queue *q, struct request *req,
 bool bio_attempt_back_merge(struct request_queue *q, struct request *req,
 			    struct bio *bio);
 bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio,
-			    unsigned int *request_count);
+			    unsigned int *request_count,
+			    struct request **same_queue_rq);
 
 void blk_account_io_start(struct request *req, bool new_io);
 void blk_account_io_completion(struct request *req, unsigned int bytes);
@@ -193,8 +194,6 @@ int blk_try_merge(struct request *rq, struct bio *bio);
 
 void blk_queue_congestion_threshold(struct request_queue *q);
 
-void __blk_run_queue_uncond(struct request_queue *q);
-
 int blk_dev_init(void);
 
 

block/bounce.c

@@ -128,9 +128,6 @@ static void bounce_end_io(struct bio *bio, mempool_t *pool, int err)
 	struct bio_vec *bvec, *org_vec;
 	int i;
 
-	if (test_bit(BIO_EOPNOTSUPP, &bio->bi_flags))
-		set_bit(BIO_EOPNOTSUPP, &bio_orig->bi_flags);
-
 	/*
 	 * free up bounce indirect pages used
 	 */

block/ioctl.c

@@ -150,21 +150,48 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user
 	}
 }
 
-static int blkdev_reread_part(struct block_device *bdev)
+/*
+ * This is an exported API for the block driver, and will not
+ * acquire bd_mutex. This API should be used in case that
+ * caller has held bd_mutex already.
+ */
+int __blkdev_reread_part(struct block_device *bdev)
 {
 	struct gendisk *disk = bdev->bd_disk;
-	int res;
 
 	if (!disk_part_scan_enabled(disk) || bdev != bdev->bd_contains)
 		return -EINVAL;
 	if (!capable(CAP_SYS_ADMIN))
 		return -EACCES;
-	if (!mutex_trylock(&bdev->bd_mutex))
-		return -EBUSY;
-	res = rescan_partitions(disk, bdev);
+
+	lockdep_assert_held(&bdev->bd_mutex);
+
+	return rescan_partitions(disk, bdev);
+}
+EXPORT_SYMBOL(__blkdev_reread_part);
+
+/*
+ * This is an exported API for the block driver, and will
+ * try to acquire bd_mutex. If bd_mutex has been held already
+ * in current context, please call __blkdev_reread_part().
+ *
+ * Make sure the held locks in current context aren't required
+ * in open()/close() handler and I/O path for avoiding ABBA deadlock:
+ * - bd_mutex is held before calling block driver's open/close
+ *   handler
+ * - reading partition table may submit I/O to the block device
+ */
+int blkdev_reread_part(struct block_device *bdev)
+{
+	int res;
+
+	mutex_lock(&bdev->bd_mutex);
+	res = __blkdev_reread_part(bdev);
 	mutex_unlock(&bdev->bd_mutex);
+
 	return res;
 }
+EXPORT_SYMBOL(blkdev_reread_part);
 
 static int blk_ioctl_discard(struct block_device *bdev, uint64_t start,
 			     uint64_t len, int secure)

drivers/block/nbd.c

@@ -230,29 +230,40 @@ static int nbd_send_req(struct nbd_device *nbd, struct request *req)
 	int result, flags;
 	struct nbd_request request;
 	unsigned long size = blk_rq_bytes(req);
+	u32 type;
+
+	if (req->cmd_type == REQ_TYPE_DRV_PRIV)
+		type = NBD_CMD_DISC;
+	else if (req->cmd_flags & REQ_DISCARD)
+		type = NBD_CMD_TRIM;
+	else if (req->cmd_flags & REQ_FLUSH)
+		type = NBD_CMD_FLUSH;
+	else if (rq_data_dir(req) == WRITE)
+		type = NBD_CMD_WRITE;
+	else
+		type = NBD_CMD_READ;
 
 	memset(&request, 0, sizeof(request));
 	request.magic = htonl(NBD_REQUEST_MAGIC);
-	request.type = htonl(nbd_cmd(req));
-
-	if (nbd_cmd(req) != NBD_CMD_FLUSH && nbd_cmd(req) != NBD_CMD_DISC) {
+	request.type = htonl(type);
+	if (type != NBD_CMD_FLUSH && type != NBD_CMD_DISC) {
 		request.from = cpu_to_be64((u64)blk_rq_pos(req) << 9);
 		request.len = htonl(size);
 	}
 	memcpy(request.handle, &req, sizeof(req));
 
 	dev_dbg(nbd_to_dev(nbd), "request %p: sending control (%s@%llu,%uB)\n",
-		req, nbdcmd_to_ascii(nbd_cmd(req)),
+		req, nbdcmd_to_ascii(type),
 		(unsigned long long)blk_rq_pos(req) << 9, blk_rq_bytes(req));
 	result = sock_xmit(nbd, 1, &request, sizeof(request),
-			(nbd_cmd(req) == NBD_CMD_WRITE) ? MSG_MORE : 0);
+			(type == NBD_CMD_WRITE) ? MSG_MORE : 0);
 	if (result <= 0) {
 		dev_err(disk_to_dev(nbd->disk),
 			"Send control failed (result %d)\n", result);
 		return -EIO;
 	}
 
-	if (nbd_cmd(req) == NBD_CMD_WRITE) {
+	if (type == NBD_CMD_WRITE) {
 		struct req_iterator iter;
 		struct bio_vec bvec;
 		/*
@@ -352,7 +363,7 @@ static struct request *nbd_read_stat(struct nbd_device *nbd)
 	}
 
 	dev_dbg(nbd_to_dev(nbd), "request %p: got reply\n", req);
-	if (nbd_cmd(req) == NBD_CMD_READ) {
+	if (rq_data_dir(req) != WRITE) {
 		struct req_iterator iter;
 		struct bio_vec bvec;
 
@@ -452,23 +463,11 @@ static void nbd_handle_req(struct nbd_device *nbd, struct request *req)
 	if (req->cmd_type != REQ_TYPE_FS)
 		goto error_out;
 
-	nbd_cmd(req) = NBD_CMD_READ;
-	if (rq_data_dir(req) == WRITE) {
-		if ((req->cmd_flags & REQ_DISCARD)) {
-			WARN_ON(!(nbd->flags & NBD_FLAG_SEND_TRIM));
-			nbd_cmd(req) = NBD_CMD_TRIM;
-		} else
-			nbd_cmd(req) = NBD_CMD_WRITE;
-		if (nbd->flags & NBD_FLAG_READ_ONLY) {
-			dev_err(disk_to_dev(nbd->disk),
-				"Write on read-only\n");
-			goto error_out;
-		}
-	}
-
-	if (req->cmd_flags & REQ_FLUSH) {
-		BUG_ON(unlikely(blk_rq_sectors(req)));
-		nbd_cmd(req) = NBD_CMD_FLUSH;
+	if (rq_data_dir(req) == WRITE &&
+	    (nbd->flags & NBD_FLAG_READ_ONLY)) {
+		dev_err(disk_to_dev(nbd->disk),
+			"Write on read-only\n");
+		goto error_out;
 	}
 
 	req->errors = 0;
@@ -592,8 +591,7 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
 		fsync_bdev(bdev);
 		mutex_lock(&nbd->tx_lock);
 		blk_rq_init(NULL, &sreq);
-		sreq.cmd_type = REQ_TYPE_SPECIAL;
-		nbd_cmd(&sreq) = NBD_CMD_DISC;
+		sreq.cmd_type = REQ_TYPE_DRV_PRIV;
 
 		/* Check again after getting mutex back.  */
 		if (!nbd->sock)

drivers/block/paride/pd.c

@@ -442,7 +442,7 @@ static char *pd_buf;		/* buffer for request in progress */
 
 static enum action do_pd_io_start(void)
 {
-	if (pd_req->cmd_type == REQ_TYPE_SPECIAL) {
+	if (pd_req->cmd_type == REQ_TYPE_DRV_PRIV) {
 		phase = pd_special;
 		return pd_special();
 	}
@@ -725,7 +725,7 @@ static int pd_special_command(struct pd_unit *disk,
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
-	rq->cmd_type = REQ_TYPE_SPECIAL;
+	rq->cmd_type = REQ_TYPE_DRV_PRIV;
 	rq->special = func;
 
 	err = blk_execute_rq(disk->gd->queue, disk->gd, rq, 0);

drivers/block/sx8.c

@@ -620,7 +620,7 @@ static int carm_array_info (struct carm_host *host, unsigned int array_idx)
 	spin_unlock_irq(&host->lock);
 
 	DPRINTK("blk_execute_rq_nowait, tag == %u\n", idx);
-	crq->rq->cmd_type = REQ_TYPE_SPECIAL;
+	crq->rq->cmd_type = REQ_TYPE_DRV_PRIV;
 	crq->rq->special = crq;
 	blk_execute_rq_nowait(host->oob_q, NULL, crq->rq, true, NULL);
 
@@ -661,7 +661,7 @@ static int carm_send_special (struct carm_host *host, carm_sspc_t func)
 	crq->msg_bucket = (u32) rc;
 
 	DPRINTK("blk_execute_rq_nowait, tag == %u\n", idx);
-	crq->rq->cmd_type = REQ_TYPE_SPECIAL;
+	crq->rq->cmd_type = REQ_TYPE_DRV_PRIV;
 	crq->rq->special = crq;
 	blk_execute_rq_nowait(host->oob_q, NULL, crq->rq, true, NULL);
 

drivers/block/virtio_blk.c

@@ -124,7 +124,7 @@ static inline void virtblk_request_done(struct request *req)
 		req->resid_len = virtio32_to_cpu(vblk->vdev, vbr->in_hdr.residual);
 		req->sense_len = virtio32_to_cpu(vblk->vdev, vbr->in_hdr.sense_len);
 		req->errors = virtio32_to_cpu(vblk->vdev, vbr->in_hdr.errors);
-	} else if (req->cmd_type == REQ_TYPE_SPECIAL) {
+	} else if (req->cmd_type == REQ_TYPE_DRV_PRIV) {
 		req->errors = (error != 0);
 	}
 
@@ -188,7 +188,7 @@ static int virtio_queue_rq(struct blk_mq_hw_ctx *hctx,
 			vbr->out_hdr.sector = 0;
 			vbr->out_hdr.ioprio = cpu_to_virtio32(vblk->vdev, req_get_ioprio(vbr->req));
 			break;
-		case REQ_TYPE_SPECIAL:
+		case REQ_TYPE_DRV_PRIV:
 			vbr->out_hdr.type = cpu_to_virtio32(vblk->vdev, VIRTIO_BLK_T_GET_ID);
 			vbr->out_hdr.sector = 0;
 			vbr->out_hdr.ioprio = cpu_to_virtio32(vblk->vdev, req_get_ioprio(vbr->req));
@@ -251,7 +251,7 @@ static int virtblk_get_id(struct gendisk *disk, char *id_str)
 		return PTR_ERR(req);
 	}
 
-	req->cmd_type = REQ_TYPE_SPECIAL;
+	req->cmd_type = REQ_TYPE_DRV_PRIV;
 	err = blk_execute_rq(vblk->disk->queue, vblk->disk, req, false);
 	blk_put_request(req);
 

drivers/ide/ide-atapi.c

@@ -93,7 +93,7 @@ int ide_queue_pc_tail(ide_drive_t *drive, struct gendisk *disk,
 	int error;
 
 	rq = blk_get_request(drive->queue, READ, __GFP_WAIT);
-	rq->cmd_type = REQ_TYPE_SPECIAL;
+	rq->cmd_type = REQ_TYPE_DRV_PRIV;
 	rq->special = (char *)pc;
 
 	if (buf && bufflen) {
@@ -191,7 +191,7 @@ void ide_prep_sense(ide_drive_t *drive, struct request *rq)
 
 	BUG_ON(sense_len > sizeof(*sense));
 
-	if (rq->cmd_type == REQ_TYPE_SENSE || drive->sense_rq_armed)
+	if (rq->cmd_type == REQ_TYPE_ATA_SENSE || drive->sense_rq_armed)
 		return;
 
 	memset(sense, 0, sizeof(*sense));
@@ -210,7 +210,7 @@ void ide_prep_sense(ide_drive_t *drive, struct request *rq)
 	sense_rq->rq_disk = rq->rq_disk;
 	sense_rq->cmd[0] = GPCMD_REQUEST_SENSE;
 	sense_rq->cmd[4] = cmd_len;
-	sense_rq->cmd_type = REQ_TYPE_SENSE;
+	sense_rq->cmd_type = REQ_TYPE_ATA_SENSE;
 	sense_rq->cmd_flags |= REQ_PREEMPT;
 
 	if (drive->media == ide_tape)
@@ -310,7 +310,7 @@ int ide_cd_get_xferlen(struct request *rq)
 	switch (rq->cmd_type) {
 	case REQ_TYPE_FS:
 		return 32768;
-	case REQ_TYPE_SENSE:
+	case REQ_TYPE_ATA_SENSE:
 	case REQ_TYPE_BLOCK_PC:
 	case REQ_TYPE_ATA_PC:
 		return blk_rq_bytes(rq);
@@ -477,7 +477,7 @@ static ide_startstop_t ide_pc_intr(ide_drive_t *drive)
 		if (uptodate == 0)
 			drive->failed_pc = NULL;
 
-		if (rq->cmd_type == REQ_TYPE_SPECIAL) {
+		if (rq->cmd_type == REQ_TYPE_DRV_PRIV) {
 			rq->errors = 0;
 			error = 0;
 		} else {

drivers/ide/ide-cd.c

@@ -210,7 +210,7 @@ static void cdrom_analyze_sense_data(ide_drive_t *drive,
 static void ide_cd_complete_failed_rq(ide_drive_t *drive, struct request *rq)
 {
 	/*
-	 * For REQ_TYPE_SENSE, "rq->special" points to the original
+	 * For REQ_TYPE_ATA_SENSE, "rq->special" points to the original
 	 * failed request.  Also, the sense data should be read
 	 * directly from rq which might be different from the original
 	 * sense buffer if it got copied during mapping.
@@ -285,7 +285,7 @@ static int cdrom_decode_status(ide_drive_t *drive, u8 stat)
 				  "stat 0x%x",
 				  rq->cmd[0], rq->cmd_type, err, stat);
 
-	if (rq->cmd_type == REQ_TYPE_SENSE) {
+	if (rq->cmd_type == REQ_TYPE_ATA_SENSE) {
 		/*
 		 * We got an error trying to get sense info from the drive
 		 * (probably while trying to recover from a former error).
@@ -526,7 +526,7 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive)
 	ide_expiry_t *expiry = NULL;
 	int dma_error = 0, dma, thislen, uptodate = 0;
 	int write = (rq_data_dir(rq) == WRITE) ? 1 : 0, rc = 0;
-	int sense = (rq->cmd_type == REQ_TYPE_SENSE);
+	int sense = (rq->cmd_type == REQ_TYPE_ATA_SENSE);
 	unsigned int timeout;
 	u16 len;
 	u8 ireason, stat;
@@ -791,7 +791,7 @@ static ide_startstop_t ide_cd_do_request(ide_drive_t *drive, struct request *rq,
 		if (cdrom_start_rw(drive, rq) == ide_stopped)
 			goto out_end;
 		break;
-	case REQ_TYPE_SENSE:
+	case REQ_TYPE_ATA_SENSE:
 	case REQ_TYPE_BLOCK_PC:
 	case REQ_TYPE_ATA_PC:
 		if (!rq->timeout)
@@ -799,7 +799,7 @@ static ide_startstop_t ide_cd_do_request(ide_drive_t *drive, struct request *rq,
 
 		cdrom_do_block_pc(drive, rq);
 		break;
-	case REQ_TYPE_SPECIAL:
+	case REQ_TYPE_DRV_PRIV:
 		/* right now this can only be a reset... */
 		uptodate = 1;
 		goto out_end;

drivers/ide/ide-cd_ioctl.c

@@ -304,7 +304,7 @@ int ide_cdrom_reset(struct cdrom_device_info *cdi)
 	int ret;
 
 	rq = blk_get_request(drive->queue, READ, __GFP_WAIT);
-	rq->cmd_type = REQ_TYPE_SPECIAL;
+	rq->cmd_type = REQ_TYPE_DRV_PRIV;
 	rq->cmd_flags = REQ_QUIET;
 	ret = blk_execute_rq(drive->queue, cd->disk, rq, 0);
 	blk_put_request(rq);

drivers/ide/ide-devsets.c

@@ -166,7 +166,7 @@ int ide_devset_execute(ide_drive_t *drive, const struct ide_devset *setting,
 		return setting->set(drive, arg);
 
 	rq = blk_get_request(q, READ, __GFP_WAIT);
-	rq->cmd_type = REQ_TYPE_SPECIAL;
+	rq->cmd_type = REQ_TYPE_DRV_PRIV;
 	rq->cmd_len = 5;
 	rq->cmd[0] = REQ_DEVSET_EXEC;
 	*(int *)&rq->cmd[1] = arg;

drivers/ide/ide-eh.c

@@ -129,7 +129,7 @@ ide_startstop_t ide_error(ide_drive_t *drive, const char *msg, u8 stat)
 
 			if (cmd)
 				ide_complete_cmd(drive, cmd, stat, err);
-		} else if (blk_pm_request(rq)) {
+		} else if (ata_pm_request(rq)) {
 			rq->errors = 1;
 			ide_complete_pm_rq(drive, rq);
 			return ide_stopped;
@@ -147,7 +147,7 @@ static inline void ide_complete_drive_reset(ide_drive_t *drive, int err)
 {
 	struct request *rq = drive->hwif->rq;
 
-	if (rq && rq->cmd_type == REQ_TYPE_SPECIAL &&
+	if (rq && rq->cmd_type == REQ_TYPE_DRV_PRIV &&
 	    rq->cmd[0] == REQ_DRIVE_RESET) {
 		if (err <= 0 && rq->errors == 0)
 			rq->errors = -EIO;

drivers/ide/ide-floppy.c

@@ -97,7 +97,7 @@ static int ide_floppy_callback(ide_drive_t *drive, int dsc)
 			       "Aborting request!\n");
 	}
 
-	if (rq->cmd_type == REQ_TYPE_SPECIAL)
+	if (rq->cmd_type == REQ_TYPE_DRV_PRIV)
 		rq->errors = uptodate ? 0 : IDE_DRV_ERROR_GENERAL;
 
 	return uptodate;
@@ -246,7 +246,7 @@ static ide_startstop_t ide_floppy_do_request(ide_drive_t *drive,
 		} else
 			printk(KERN_ERR PFX "%s: I/O error\n", drive->name);
 
-		if (rq->cmd_type == REQ_TYPE_SPECIAL) {
+		if (rq->cmd_type == REQ_TYPE_DRV_PRIV) {
 			rq->errors = 0;
 			ide_complete_rq(drive, 0, blk_rq_bytes(rq));
 			return ide_stopped;
@@ -265,8 +265,8 @@ static ide_startstop_t ide_floppy_do_request(ide_drive_t *drive,
 		pc = &floppy->queued_pc;
 		idefloppy_create_rw_cmd(drive, pc, rq, (unsigned long)block);
 		break;
-	case REQ_TYPE_SPECIAL:
-	case REQ_TYPE_SENSE:
+	case REQ_TYPE_DRV_PRIV:
+	case REQ_TYPE_ATA_SENSE:
 		pc = (struct ide_atapi_pc *)rq->special;
 		break;
 	case REQ_TYPE_BLOCK_PC:

drivers/ide/ide-io.c

@@ -135,7 +135,7 @@ EXPORT_SYMBOL(ide_complete_rq);
 
 void ide_kill_rq(ide_drive_t *drive, struct request *rq)
 {
-	u8 drv_req = (rq->cmd_type == REQ_TYPE_SPECIAL) && rq->rq_disk;
+	u8 drv_req = (rq->cmd_type == REQ_TYPE_DRV_PRIV) && rq->rq_disk;
 	u8 media = drive->media;
 
 	drive->failed_pc = NULL;
@@ -320,7 +320,7 @@ static ide_startstop_t start_request (ide_drive_t *drive, struct request *rq)
 		goto kill_rq;
 	}
 
-	if (blk_pm_request(rq))
+	if (ata_pm_request(rq))
 		ide_check_pm_state(drive, rq);
 
 	drive->hwif->tp_ops->dev_select(drive);
@@ -342,8 +342,8 @@ static ide_startstop_t start_request (ide_drive_t *drive, struct request *rq)
 
 		if (rq->cmd_type == REQ_TYPE_ATA_TASKFILE)
 			return execute_drive_cmd(drive, rq);
-		else if (blk_pm_request(rq)) {
-			struct request_pm_state *pm = rq->special;
+		else if (ata_pm_request(rq)) {
+			struct ide_pm_state *pm = rq->special;
 #ifdef DEBUG_PM
 			printk("%s: start_power_step(step: %d)\n",
 				drive->name, pm->pm_step);
@@ -353,7 +353,7 @@ static ide_startstop_t start_request (ide_drive_t *drive, struct request *rq)
 			    pm->pm_step == IDE_PM_COMPLETED)
 				ide_complete_pm_rq(drive, rq);
 			return startstop;
-		} else if (!rq->rq_disk && rq->cmd_type == REQ_TYPE_SPECIAL)
+		} else if (!rq->rq_disk && rq->cmd_type == REQ_TYPE_DRV_PRIV)
 			/*
 			 * TODO: Once all ULDs have been modified to
 			 * check for specific op codes rather than
@@ -538,7 +538,7 @@ repeat:
 		 * state machine.
 		 */
 		if ((drive->dev_flags & IDE_DFLAG_BLOCKED) &&
-		    blk_pm_request(rq) == 0 &&
+		    ata_pm_request(rq) == 0 &&
 		    (rq->cmd_flags & REQ_PREEMPT) == 0) {
 			/* there should be no pending command at this point */
 			ide_unlock_port(hwif);

drivers/ide/ide-ioctls.c

@@ -222,7 +222,7 @@ static int generic_drive_reset(ide_drive_t *drive)
 	int ret = 0;
 
 	rq = blk_get_request(drive->queue, READ, __GFP_WAIT);
-	rq->cmd_type = REQ_TYPE_SPECIAL;
+	rq->cmd_type = REQ_TYPE_DRV_PRIV;
 	rq->cmd_len = 1;
 	rq->cmd[0] = REQ_DRIVE_RESET;
 	if (blk_execute_rq(drive->queue, NULL, rq, 1))