Merge branch 'for-linus' of git://git.kernel.dk/linux-block

Pull block layer fixes from Jens Axboe:
 "A smaller collection of fixes that have come up since the initial
  merge window pull request.  This contains:

   - error handling cleanup and support for larger than 16 byte cdbs in
     sg_io() from Christoph.  The latter just matches what bsg and
     friends support, sg_io() got left out in the merge.

   - an option for brd to expose partitions in /proc/partitions.  They
     are hidden by default for compat reasons.  From Dmitry Monakhov.

   - a few blk-mq fixes from me - killing a dead/unused flag, fix for
     merging happening even if turned off, and correction of a few
     comments.

   - removal of unnecessary ->owner setting in systemace.  From Michal
     Simek.

   - two related fixes for a problem with nesting freezing of queues in
     blk-mq.  One from Ming Lei removing an unecessary freeze operation,
     and another from Tejun fixing the nesting regression introduced in
     the merge window.

   - fix for a BUG_ON() at bio_endio time when protection info is
     attached and the IO has an error.  From Sagi Grimberg.

   - two scsi_ioctl bug fixes for regressions with scsi-mq from Tony
     Battersby.

   - a cfq weight update fix and subsequent comment update from Toshiaki
     Makita"

* 'for-linus' of git://git.kernel.dk/linux-block:
  cfq-iosched: Add comments on update timing of weight
  cfq-iosched: Fix wrong children_weight calculation
  block: fix error handling in sg_io
  fix regression in SCSI_IOCTL_SEND_COMMAND
  scsi-mq: fix requests that use a separate CDB buffer
  block: support > 16 byte CDBs for SG_IO
  block: cleanup error handling in sg_io
  brd: add ram disk visibility option
  block: systemace: Remove .owner field for driver
  blk-mq: blk_mq_freeze_queue() should allow nesting
  blk-mq: correct a few wrong/bad comments
  block: Fix BUG_ON when pi errors occur
  blk-mq: don't allow merges if turned off for the queue
  blk-mq: get rid of unused BLK_MQ_F_SHOULD_SORT flag
  blk-mq: fix WARNING "percpu_ref_kill() called more than once!"

9 files changed, 74 insertions, 39 deletions

diff --git a/block/bio-integrity.c b/block/bio-integrity.c
index bc423f7b02da..f14b4abbebd8 100644
--- a/block/bio-integrity.c
+++ b/block/bio-integrity.c
@@ -520,7 +520,7 @@ void bio_integrity_endio(struct bio *bio, int error)
 	 */
 	if (error) {
 		bio->bi_end_io = bip->bip_end_io;
-		bio_endio(bio, error);
+		bio_endio_nodec(bio, error);
 
 		return;
 	}
diff --git a/block/blk-core.c b/block/blk-core.c
index c359d72e9d76..bf930f481d43 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1252,7 +1252,6 @@ void blk_rq_set_block_pc(struct request *rq)
 	rq->__sector = (sector_t) -1;
 	rq->bio = rq->biotail = NULL;
 	memset(rq->__cmd, 0, sizeof(rq->__cmd));
-	rq->cmd = rq->__cmd;
 }
 EXPORT_SYMBOL(blk_rq_set_block_pc);
 
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 5189cb1e478a..4aac82615a46 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -112,18 +112,22 @@ static void blk_mq_usage_counter_release(struct percpu_ref *ref)
  */
 void blk_mq_freeze_queue(struct request_queue *q)
 {
+	bool freeze;
+
 	spin_lock_irq(q->queue_lock);
-	q->mq_freeze_depth++;
+	freeze = !q->mq_freeze_depth++;
 	spin_unlock_irq(q->queue_lock);
 
-	percpu_ref_kill(&q->mq_usage_counter);
-	blk_mq_run_queues(q, false);
+	if (freeze) {
+		percpu_ref_kill(&q->mq_usage_counter);
+		blk_mq_run_queues(q, false);
+	}
 	wait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->mq_usage_counter));
 }
 
 static void blk_mq_unfreeze_queue(struct request_queue *q)
 {
-	bool wake = false;
+	bool wake;
 
 	spin_lock_irq(q->queue_lock);
 	wake = !--q->mq_freeze_depth;
@@ -172,6 +176,8 @@ static void blk_mq_rq_ctx_init(struct request_queue *q, struct blk_mq_ctx *ctx,
 	/* tag was already set */
 	rq->errors = 0;
 
+	rq->cmd = rq->__cmd;
+
 	rq->extra_len = 0;
 	rq->sense_len = 0;
 	rq->resid_len = 0;
@@ -1068,13 +1074,17 @@ static void blk_mq_bio_to_request(struct request *rq, struct bio *bio)
 		blk_account_io_start(rq, 1);
 }
 
+static inline bool hctx_allow_merges(struct blk_mq_hw_ctx *hctx)
+{
+	return (hctx->flags & BLK_MQ_F_SHOULD_MERGE) &&
+		!blk_queue_nomerges(hctx->queue);
+}
+
 static inline bool blk_mq_merge_queue_io(struct blk_mq_hw_ctx *hctx,
 					 struct blk_mq_ctx *ctx,
 					 struct request *rq, struct bio *bio)
 {
-	struct request_queue *q = hctx->queue;
-
-	if (!(hctx->flags & BLK_MQ_F_SHOULD_MERGE)) {
+	if (!hctx_allow_merges(hctx)) {
 		blk_mq_bio_to_request(rq, bio);
 		spin_lock(&ctx->lock);
 insert_rq:
@@ -1082,6 +1092,8 @@ insert_rq:
 		spin_unlock(&ctx->lock);
 		return false;
 	} else {
+		struct request_queue *q = hctx->queue;
+
 		spin_lock(&ctx->lock);
 		if (!blk_mq_attempt_merge(q, ctx, bio)) {
 			blk_mq_bio_to_request(rq, bio);
@@ -1574,7 +1586,7 @@ static int blk_mq_init_hw_queues(struct request_queue *q,
 		hctx->tags = set->tags[i];
 
 		/*
-		 * Allocate space for all possible cpus to avoid allocation in
+		 * Allocate space for all possible cpus to avoid allocation at
 		 * runtime
 		 */
 		hctx->ctxs = kmalloc_node(nr_cpu_ids * sizeof(void *),
@@ -1662,8 +1674,8 @@ static void blk_mq_map_swqueue(struct request_queue *q)
 
 	queue_for_each_hw_ctx(q, hctx, i) {
 		/*
-		 * If not software queues are mapped to this hardware queue,
-		 * disable it and free the request entries
+		 * If no software queues are mapped to this hardware queue,
+		 * disable it and free the request entries.
 		 */
 		if (!hctx->nr_ctx) {
 			struct blk_mq_tag_set *set = q->tag_set;
@@ -1713,14 +1725,10 @@ static void blk_mq_del_queue_tag_set(struct request_queue *q)
 {
 	struct blk_mq_tag_set *set = q->tag_set;
 
-	blk_mq_freeze_queue(q);
-
 	mutex_lock(&set->tag_list_lock);
 	list_del_init(&q->tag_set_list);
 	blk_mq_update_tag_set_depth(set);
 	mutex_unlock(&set->tag_list_lock);
-
-	blk_mq_unfreeze_queue(q);
 }
 
 static void blk_mq_add_queue_tag_set(struct blk_mq_tag_set *set,
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index cadc37841744..3f31cf9508e6 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -1272,15 +1272,22 @@ __cfq_group_service_tree_add(struct cfq_rb_root *st, struct cfq_group *cfqg)
 	rb_insert_color(&cfqg->rb_node, &st->rb);
 }
 
+/*
+ * This has to be called only on activation of cfqg
+ */
 static void
 cfq_update_group_weight(struct cfq_group *cfqg)
 {
-	BUG_ON(!RB_EMPTY_NODE(&cfqg->rb_node));
-
 	if (cfqg->new_weight) {
 		cfqg->weight = cfqg->new_weight;
 		cfqg->new_weight = 0;
 	}
+}
+
+static void
+cfq_update_group_leaf_weight(struct cfq_group *cfqg)
+{
+	BUG_ON(!RB_EMPTY_NODE(&cfqg->rb_node));
 
 	if (cfqg->new_leaf_weight) {
 		cfqg->leaf_weight = cfqg->new_leaf_weight;
@@ -1299,7 +1306,12 @@ cfq_group_service_tree_add(struct cfq_rb_root *st, struct cfq_group *cfqg)
 	/* add to the service tree */
 	BUG_ON(!RB_EMPTY_NODE(&cfqg->rb_node));
 
-	cfq_update_group_weight(cfqg);
+	/*
+	 * Update leaf_weight.  We cannot update weight at this point
+	 * because cfqg might already have been activated and is
+	 * contributing its current weight to the parent's child_weight.
+	 */
+	cfq_update_group_leaf_weight(cfqg);
 	__cfq_group_service_tree_add(st, cfqg);
 
 	/*
@@ -1323,6 +1335,7 @@ cfq_group_service_tree_add(struct cfq_rb_root *st, struct cfq_group *cfqg)
 	 */
 	while ((parent = cfqg_parent(pos))) {
 		if (propagate) {
+			cfq_update_group_weight(pos);
 			propagate = !parent->nr_active++;
 			parent->children_weight += pos->weight;
 		}
diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c
index 51bf5155ee75..9b8eaeca6a79 100644
--- a/block/scsi_ioctl.c
+++ b/block/scsi_ioctl.c
@@ -279,7 +279,6 @@ static int blk_complete_sghdr_rq(struct request *rq, struct sg_io_hdr *hdr,
 	r = blk_rq_unmap_user(bio);
 	if (!ret)
 		ret = r;
-	blk_put_request(rq);
 
 	return ret;
 }
@@ -297,8 +296,6 @@ static int sg_io(struct request_queue *q, struct gendisk *bd_disk,
 
 	if (hdr->interface_id != 'S')
 		return -EINVAL;
-	if (hdr->cmd_len > BLK_MAX_CDB)
-		return -EINVAL;
 
 	if (hdr->dxfer_len > (queue_max_hw_sectors(q) << 9))
 		return -EIO;
@@ -317,16 +314,23 @@ static int sg_io(struct request_queue *q, struct gendisk *bd_disk,
 	if (hdr->flags & SG_FLAG_Q_AT_HEAD)
 		at_head = 1;
 
+	ret = -ENOMEM;
 	rq = blk_get_request(q, writing ? WRITE : READ, GFP_KERNEL);
 	if (!rq)
-		return -ENOMEM;
+		goto out;
 	blk_rq_set_block_pc(rq);
 
-	if (blk_fill_sghdr_rq(q, rq, hdr, mode)) {
-		blk_put_request(rq);
-		return -EFAULT;
+	if (hdr->cmd_len > BLK_MAX_CDB) {
+		rq->cmd = kzalloc(hdr->cmd_len, GFP_KERNEL);
+		if (!rq->cmd)
+			goto out_put_request;
 	}
 
+	ret = -EFAULT;
+	if (blk_fill_sghdr_rq(q, rq, hdr, mode))
+		goto out_free_cdb;
+
+	ret = 0;
 	if (hdr->iovec_count) {
 		size_t iov_data_len;
 		struct iovec *iov = NULL;
@@ -335,7 +339,7 @@ static int sg_io(struct request_queue *q, struct gendisk *bd_disk,
 					    0, NULL, &iov);
 		if (ret < 0) {
 			kfree(iov);
-			goto out;
+			goto out_free_cdb;
 		}
 
 		iov_data_len = ret;
@@ -358,7 +362,7 @@ static int sg_io(struct request_queue *q, struct gendisk *bd_disk,
 				      GFP_KERNEL);
 
 	if (ret)
-		goto out;
+		goto out_free_cdb;
 
 	bio = rq->bio;
 	memset(sense, 0, sizeof(sense));
@@ -376,9 +380,14 @@ static int sg_io(struct request_queue *q, struct gendisk *bd_disk,
 
 	hdr->duration = jiffies_to_msecs(jiffies - start_time);
 
-	return blk_complete_sghdr_rq(rq, hdr, bio);
-out:
+	ret = blk_complete_sghdr_rq(rq, hdr, bio);
+
+out_free_cdb:
+	if (rq->cmd != rq->__cmd)
+		kfree(rq->cmd);
+out_put_request:
 	blk_put_request(rq);
+out:
 	return ret;
 }
 
@@ -448,6 +457,11 @@ int sg_scsi_ioctl(struct request_queue *q, struct gendisk *disk, fmode_t mode,
 	}
 
 	rq = blk_get_request(q, in_len ? WRITE : READ, __GFP_WAIT);
+	if (!rq) {
+		err = -ENOMEM;
+		goto error;
+	}
+	blk_rq_set_block_pc(rq);
 
 	cmdlen = COMMAND_SIZE(opcode);
 
@@ -501,7 +515,6 @@ int sg_scsi_ioctl(struct request_queue *q, struct gendisk *disk, fmode_t mode,
 	memset(sense, 0, sizeof(sense));
 	rq->sense = sense;
 	rq->sense_len = 0;
-	blk_rq_set_block_pc(rq);
 
 	blk_execute_rq(q, disk, rq, 0);
 
@@ -521,7 +534,8 @@ out:
 	
 error:
 	kfree(buffer);
-	blk_put_request(rq);
+	if (rq)
+		blk_put_request(rq);
 	return err;
 }
 EXPORT_SYMBOL_GPL(sg_scsi_ioctl);
diff --git a/drivers/block/brd.c b/drivers/block/brd.c
index c7d138eca731..3598110d2cef 100644
--- a/drivers/block/brd.c
+++ b/drivers/block/brd.c
@@ -442,12 +442,15 @@ static int rd_nr;
 int rd_size = CONFIG_BLK_DEV_RAM_SIZE;
 static int max_part;
 static int part_shift;
+static int part_show = 0;
 module_param(rd_nr, int, S_IRUGO);
 MODULE_PARM_DESC(rd_nr, "Maximum number of brd devices");
 module_param(rd_size, int, S_IRUGO);
 MODULE_PARM_DESC(rd_size, "Size of each RAM disk in kbytes.");
 module_param(max_part, int, S_IRUGO);
 MODULE_PARM_DESC(max_part, "Maximum number of partitions per RAM disk");
+module_param(part_show, int, S_IRUGO);
+MODULE_PARM_DESC(part_show, "Control RAM disk visibility in /proc/partitions");
 MODULE_LICENSE("GPL");
 MODULE_ALIAS_BLOCKDEV_MAJOR(RAMDISK_MAJOR);
 MODULE_ALIAS("rd");
@@ -501,7 +504,8 @@ static struct brd_device *brd_alloc(int i)
 	disk->fops		= &brd_fops;
 	disk->private_data	= brd;
 	disk->queue		= brd->brd_queue;
-	disk->flags |= GENHD_FL_SUPPRESS_PARTITION_INFO;
+	if (!part_show)
+		disk->flags |= GENHD_FL_SUPPRESS_PARTITION_INFO;
 	sprintf(disk->disk_name, "ram%d", i);
 	set_capacity(disk, rd_size * 2);
 
diff --git a/drivers/block/xsysace.c b/drivers/block/xsysace.c
index ab3ea62e5dfc..c4328d9d9981 100644
--- a/drivers/block/xsysace.c
+++ b/drivers/block/xsysace.c
@@ -1203,7 +1203,6 @@ static struct platform_driver ace_platform_driver = {
 	.probe = ace_probe,
 	.remove = ace_remove,
 	.driver = {
-		.owner = THIS_MODULE,
 		.name = "xsysace",
 		.of_match_table = ace_of_match,
 	},
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index ce62e8798cc8..d837dc180522 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -1808,7 +1808,6 @@ static int scsi_mq_prep_fn(struct request *req)
 
 	cmd->tag = req->tag;
 
-	req->cmd = req->__cmd;
 	cmd->cmnd = req->cmd;
 	cmd->prot_op = SCSI_PROT_NORMAL;
 
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index eb726b9c5762..a1e31f274fcd 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -127,10 +127,9 @@ enum {
 	BLK_MQ_RQ_QUEUE_ERROR	= 2,	/* end IO with error */
 
 	BLK_MQ_F_SHOULD_MERGE	= 1 << 0,
-	BLK_MQ_F_SHOULD_SORT	= 1 << 1,
-	BLK_MQ_F_TAG_SHARED	= 1 << 2,
-	BLK_MQ_F_SG_MERGE	= 1 << 3,
-	BLK_MQ_F_SYSFS_UP	= 1 << 4,
+	BLK_MQ_F_TAG_SHARED	= 1 << 1,
+	BLK_MQ_F_SG_MERGE	= 1 << 2,
+	BLK_MQ_F_SYSFS_UP	= 1 << 3,
 
 	BLK_MQ_S_STOPPED	= 0,
 	BLK_MQ_S_TAG_ACTIVE	= 1,