Merge tag 'block-5.14-2021-07-08' of git://git.kernel.dk/linux-block

Pull more block updates from Jens Axboe:
 "A combination of changes that ended up depending on both the driver
  and core branch (and/or the IDE removal), and a few late arriving
  fixes. In detail:

   - Fix io ticks wrap-around issue (Chunguang)

   - nvme-tcp sock locking fix (Maurizio)

   - s390-dasd fixes (Kees, Christoph)

   - blk_execute_rq polling support (Keith)

   - blk-cgroup RCU iteration fix (Yu)

   - nbd backend ID addition (Prasanna)

   - Partition deletion fix (Yufen)

   - Use blk_mq_alloc_disk for mmc, mtip32xx, ubd (Christoph)

   - Removal of now dead block request types due to IDE removal
     (Christoph)

   - Loop probing and control device cleanups (Christoph)

   - Device uevent fix (Christoph)

   - Misc cleanups/fixes (Tetsuo, Christoph)"

* tag 'block-5.14-2021-07-08' of git://git.kernel.dk/linux-block: (34 commits)
  blk-cgroup: prevent rcu_sched detected stalls warnings while iterating blkgs
  block: fix the problem of io_ticks becoming smaller
  nvme-tcp: can't set sk_user_data without write_lock
  loop: remove unused variable in loop_set_status()
  block: remove the bdgrab in blk_drop_partitions
  block: grab a device refcount in disk_uevent
  s390/dasd: Avoid field over-reading memcpy()
  dasd: unexport dasd_set_target_state
  block: check disk exist before trying to add partition
  ubd: remove dead code in ubd_setup_common
  nvme: use return value from blk_execute_rq()
  block: return errors from blk_execute_rq()
  nvme: use blk_execute_rq() for passthrough commands
  block: support polling through blk_execute_rq
  block: remove REQ_OP_SCSI_{IN,OUT}
  block: mark blk_mq_init_queue_data static
  loop: rewrite loop_exit using idr_for_each_entry
  loop: split loop_lookup
  loop: don't allow deleting an unspecified loop device
  loop: move loop_ctl_mutex locking into loop_add
  ...

Author: torvalds

arch/um/drivers/ubd_kern.c

@@ -125,9 +125,7 @@ static const struct block_device_operations ubd_blops = {
 };
 
 /* Protected by ubd_lock */
-static int fake_major = UBD_MAJOR;
 static struct gendisk *ubd_gendisk[MAX_DEV];
-static struct gendisk *fake_gendisk[MAX_DEV];
 
 #ifdef CONFIG_BLK_DEV_UBD_SYNC
 #define OPEN_FLAGS ((struct openflags) { .r = 1, .w = 1, .s = 1, .c = 0, \
@@ -197,54 +195,19 @@ struct ubd {
 /* Protected by ubd_lock */
 static struct ubd ubd_devs[MAX_DEV] = { [0 ... MAX_DEV - 1] = DEFAULT_UBD };
 
-/* Only changed by fake_ide_setup which is a setup */
-static int fake_ide = 0;
-static struct proc_dir_entry *proc_ide_root = NULL;
-static struct proc_dir_entry *proc_ide = NULL;
-
 static blk_status_t ubd_queue_rq(struct blk_mq_hw_ctx *hctx,
 				 const struct blk_mq_queue_data *bd);
 
-static void make_proc_ide(void)
-{
-	proc_ide_root = proc_mkdir("ide", NULL);
-	proc_ide = proc_mkdir("ide0", proc_ide_root);
-}
-
-static int fake_ide_media_proc_show(struct seq_file *m, void *v)
-{
-	seq_puts(m, "disk\n");
-	return 0;
-}
-
-static void make_ide_entries(const char *dev_name)
-{
-	struct proc_dir_entry *dir, *ent;
-	char name[64];
-
-	if(proc_ide_root == NULL) make_proc_ide();
-
-	dir = proc_mkdir(dev_name, proc_ide);
-	if(!dir) return;
-
-	ent = proc_create_single("media", S_IRUGO, dir,
-			fake_ide_media_proc_show);
-	if(!ent) return;
-	snprintf(name, sizeof(name), "ide0/%s", dev_name);
-	proc_symlink(dev_name, proc_ide_root, name);
-}
-
 static int fake_ide_setup(char *str)
 {
-	fake_ide = 1;
+	pr_warn("The fake_ide option has been removed\n");
 	return 1;
 }
-
 __setup("fake_ide", fake_ide_setup);
 
 __uml_help(fake_ide_setup,
 "fake_ide\n"
-"    Create ide0 entries that map onto ubd devices.\n\n"
+"    Obsolete stub.\n\n"
 );
 
 static int parse_unit(char **ptr)
@@ -280,36 +243,14 @@ static int ubd_setup_common(char *str, int *index_out, char **error_out)
 	if(index_out) *index_out = -1;
 	n = *str;
 	if(n == '='){
-		char *end;
-		int major;
-
 		str++;
 		if(!strcmp(str, "sync")){
 			global_openflags = of_sync(global_openflags);
 			return err;
 		}
 
-		err = -EINVAL;
-		major = simple_strtoul(str, &end, 0);
-		if((*end != '\0') || (end == str)){
-			*error_out = "Didn't parse major number";
-			return err;
-		}
-
-		mutex_lock(&ubd_lock);
-		if (fake_major != UBD_MAJOR) {
-			*error_out = "Can't assign a fake major twice";
-			goto out1;
-		}
-
-		fake_major = major;
-
-		printk(KERN_INFO "Setting extra ubd major number to %d\n",
-		       major);
-		err = 0;
-	out1:
-		mutex_unlock(&ubd_lock);
-		return err;
+		pr_warn("fake major not supported any more\n");
+		return 0;
 	}
 
 	n = parse_unit(&str);
@@ -874,7 +815,6 @@ static void ubd_device_release(struct device *dev)
 {
 	struct ubd *ubd_dev = dev_get_drvdata(dev);
 
-	blk_cleanup_queue(ubd_dev->queue);
 	blk_mq_free_tag_set(&ubd_dev->tag_set);
 	*ubd_dev = ((struct ubd) DEFAULT_UBD);
 }
@@ -914,41 +854,25 @@ static const struct attribute_group *ubd_attr_groups[] = {
 	NULL,
 };
 
-static int ubd_disk_register(int major, u64 size, int unit,
-			     struct gendisk **disk_out)
+static void ubd_disk_register(int major, u64 size, int unit,
+			      struct gendisk *disk)
 {
-	struct device *parent = NULL;
-	struct gendisk *disk;
-
-	disk = alloc_disk(1 << UBD_SHIFT);
-	if(disk == NULL)
-		return -ENOMEM;
-
 	disk->major = major;
 	disk->first_minor = unit << UBD_SHIFT;
+	disk->minors = 1 << UBD_SHIFT;
 	disk->fops = &ubd_blops;
 	set_capacity(disk, size / 512);
-	if (major == UBD_MAJOR)
-		sprintf(disk->disk_name, "ubd%c", 'a' + unit);
-	else
-		sprintf(disk->disk_name, "ubd_fake%d", unit);
-
-	/* sysfs register (not for ide fake devices) */
-	if (major == UBD_MAJOR) {
-		ubd_devs[unit].pdev.id   = unit;
-		ubd_devs[unit].pdev.name = DRIVER_NAME;
-		ubd_devs[unit].pdev.dev.release = ubd_device_release;
-		dev_set_drvdata(&ubd_devs[unit].pdev.dev, &ubd_devs[unit]);
-		platform_device_register(&ubd_devs[unit].pdev);
-		parent = &ubd_devs[unit].pdev.dev;
-	}
+	sprintf(disk->disk_name, "ubd%c", 'a' + unit);
+
+	ubd_devs[unit].pdev.id   = unit;
+	ubd_devs[unit].pdev.name = DRIVER_NAME;
+	ubd_devs[unit].pdev.dev.release = ubd_device_release;
+	dev_set_drvdata(&ubd_devs[unit].pdev.dev, &ubd_devs[unit]);
+	platform_device_register(&ubd_devs[unit].pdev);
 
 	disk->private_data = &ubd_devs[unit];
 	disk->queue = ubd_devs[unit].queue;
-	device_add_disk(parent, disk, ubd_attr_groups);
-
-	*disk_out = disk;
-	return 0;
+	device_add_disk(&ubd_devs[unit].pdev.dev, disk, ubd_attr_groups);
 }
 
 #define ROUND_BLOCK(n) ((n + (SECTOR_SIZE - 1)) & (-SECTOR_SIZE))
@@ -960,6 +884,7 @@ static const struct blk_mq_ops ubd_mq_ops = {
 static int ubd_add(int n, char **error_out)
 {
 	struct ubd *ubd_dev = &ubd_devs[n];
+	struct gendisk *disk;
 	int err = 0;
 
 	if(ubd_dev->file == NULL)
@@ -984,43 +909,24 @@ static int ubd_add(int n, char **error_out)
 	if (err)
 		goto out;
 
-	ubd_dev->queue = blk_mq_init_queue(&ubd_dev->tag_set);
-	if (IS_ERR(ubd_dev->queue)) {
-		err = PTR_ERR(ubd_dev->queue);
+	disk = blk_mq_alloc_disk(&ubd_dev->tag_set, ubd_dev);
+	if (IS_ERR(disk)) {
+		err = PTR_ERR(disk);
 		goto out_cleanup_tags;
 	}
+	ubd_dev->queue = disk->queue;
 
-	ubd_dev->queue->queuedata = ubd_dev;
 	blk_queue_write_cache(ubd_dev->queue, true, false);
-
 	blk_queue_max_segments(ubd_dev->queue, MAX_SG);
 	blk_queue_segment_boundary(ubd_dev->queue, PAGE_SIZE - 1);
-	err = ubd_disk_register(UBD_MAJOR, ubd_dev->size, n, &ubd_gendisk[n]);
-	if(err){
-		*error_out = "Failed to register device";
-		goto out_cleanup_tags;
-	}
-
-	if (fake_major != UBD_MAJOR)
-		ubd_disk_register(fake_major, ubd_dev->size, n,
-				  &fake_gendisk[n]);
-
-	/*
-	 * Perhaps this should also be under the "if (fake_major)" above
-	 * using the fake_disk->disk_name
-	 */
-	if (fake_ide)
-		make_ide_entries(ubd_gendisk[n]->disk_name);
-
-	err = 0;
-out:
-	return err;
+	ubd_disk_register(UBD_MAJOR, ubd_dev->size, n, disk);
+	ubd_gendisk[n] = disk;
+	return 0;
 
 out_cleanup_tags:
 	blk_mq_free_tag_set(&ubd_dev->tag_set);
-	if (!(IS_ERR(ubd_dev->queue)))
-		blk_cleanup_queue(ubd_dev->queue);
-	goto out;
+out:
+	return err;
 }
 
 static int ubd_config(char *str, char **error_out)
@@ -1123,13 +1029,7 @@ static int ubd_remove(int n, char **error_out)
 	ubd_gendisk[n] = NULL;
 	if(disk != NULL){
 		del_gendisk(disk);
-		put_disk(disk);
-	}
-
-	if(fake_gendisk[n] != NULL){
-		del_gendisk(fake_gendisk[n]);
-		put_disk(fake_gendisk[n]);
-		fake_gendisk[n] = NULL;
+		blk_cleanup_disk(disk);
 	}
 
 	err = 0;
@@ -1188,14 +1088,6 @@ static int __init ubd_init(void)
 	if (register_blkdev(UBD_MAJOR, "ubd"))
 		return -1;
 
-	if (fake_major != UBD_MAJOR) {
-		char name[sizeof("ubd_nnn\0")];
-
-		snprintf(name, sizeof(name), "ubd_%d", fake_major);
-		if (register_blkdev(fake_major, "ubd"))
-			return -1;
-	}
-
 	irq_req_buffer = kmalloc_array(UBD_REQ_BUFFER_SIZE,
 				       sizeof(struct io_thread_req *),
 				       GFP_KERNEL

block/blk-cgroup.c

@@ -56,6 +56,8 @@ static LIST_HEAD(all_blkcgs);		/* protected by blkcg_pol_mutex */
 bool blkcg_debug_stats = false;
 static struct workqueue_struct *blkcg_punt_bio_wq;
 
+#define BLKG_DESTROY_BATCH_SIZE  64
+
 static bool blkcg_policy_enabled(struct request_queue *q,
 				 const struct blkcg_policy *pol)
 {
@@ -422,14 +424,27 @@ static void blkg_destroy(struct blkcg_gq *blkg)
 static void blkg_destroy_all(struct request_queue *q)
 {
 	struct blkcg_gq *blkg, *n;
+	int count = BLKG_DESTROY_BATCH_SIZE;
 
+restart:
 	spin_lock_irq(&q->queue_lock);
 	list_for_each_entry_safe(blkg, n, &q->blkg_list, q_node) {
 		struct blkcg *blkcg = blkg->blkcg;
 
 		spin_lock(&blkcg->lock);
 		blkg_destroy(blkg);
 		spin_unlock(&blkcg->lock);
+
+		/*
+		 * in order to avoid holding the spin lock for too long, release
+		 * it when a batch of blkgs are destroyed.
+		 */
+		if (!(--count)) {
+			count = BLKG_DESTROY_BATCH_SIZE;
+			spin_unlock_irq(&q->queue_lock);
+			cond_resched();
+			goto restart;
+		}
 	}
 
 	q->root_blkg = NULL;

block/blk-core.c

@@ -142,8 +142,6 @@ static const char *const blk_op_name[] = {
 	REQ_OP_NAME(ZONE_APPEND),
 	REQ_OP_NAME(WRITE_SAME),
 	REQ_OP_NAME(WRITE_ZEROES),
-	REQ_OP_NAME(SCSI_IN),
-	REQ_OP_NAME(SCSI_OUT),
 	REQ_OP_NAME(DRV_IN),
 	REQ_OP_NAME(DRV_OUT),
 };
@@ -1243,7 +1241,7 @@ static void update_io_ticks(struct block_device *part, unsigned long now,
 	unsigned long stamp;
 again:
 	stamp = READ_ONCE(part->bd_stamp);
-	if (unlikely(stamp != now)) {
+	if (unlikely(time_after(now, stamp))) {
 		if (likely(cmpxchg(&part->bd_stamp, stamp, now) == stamp))
 			__part_stat_add(part, io_ticks, end ? now - stamp : 1);
 	}

block/blk-exec.c

@@ -21,7 +21,7 @@ static void blk_end_sync_rq(struct request *rq, blk_status_t error)
 {
 	struct completion *waiting = rq->end_io_data;
 
-	rq->end_io_data = NULL;
+	rq->end_io_data = (void *)(uintptr_t)error;
 
 	/*
 	 * complete last, if this is a stack request the process (and thus
@@ -63,6 +63,19 @@ void blk_execute_rq_nowait(struct gendisk *bd_disk, struct request *rq,
 }
 EXPORT_SYMBOL_GPL(blk_execute_rq_nowait);
 
+static bool blk_rq_is_poll(struct request *rq)
+{
+	return rq->mq_hctx && rq->mq_hctx->type == HCTX_TYPE_POLL;
+}
+
+static void blk_rq_poll_completion(struct request *rq, struct completion *wait)
+{
+	do {
+		blk_poll(rq->q, request_to_qc_t(rq->mq_hctx, rq), true);
+		cond_resched();
+	} while (!completion_done(wait));
+}
+
 /**
  * blk_execute_rq - insert a request into queue for execution
  * @bd_disk:	matching gendisk
@@ -72,8 +85,9 @@ EXPORT_SYMBOL_GPL(blk_execute_rq_nowait);
  * Description:
  *    Insert a fully prepared request at the back of the I/O scheduler queue
  *    for execution and wait for completion.
+ * Return: The blk_status_t result provided to blk_mq_end_request().
  */
-void blk_execute_rq(struct gendisk *bd_disk, struct request *rq, int at_head)
+blk_status_t blk_execute_rq(struct gendisk *bd_disk, struct request *rq, int at_head)
 {
 	DECLARE_COMPLETION_ONSTACK(wait);
 	unsigned long hang_check;
@@ -83,9 +97,14 @@ void blk_execute_rq(struct gendisk *bd_disk, struct request *rq, int at_head)
 
 	/* Prevent hang_check timer from firing at us during very long I/O */
 	hang_check = sysctl_hung_task_timeout_secs;
-	if (hang_check)
+
+	if (blk_rq_is_poll(rq))
+		blk_rq_poll_completion(rq, &wait);
+	else if (hang_check)
 		while (!wait_for_completion_io_timeout(&wait, hang_check * (HZ/2)));
 	else
 		wait_for_completion_io(&wait);
+
+	return (blk_status_t)(uintptr_t)rq->end_io_data;
 }
 EXPORT_SYMBOL(blk_execute_rq);

block/blk-mq.c

@@ -3109,7 +3109,7 @@ void blk_mq_release(struct request_queue *q)
 	blk_mq_sysfs_deinit(q);
 }
 
-struct request_queue *blk_mq_init_queue_data(struct blk_mq_tag_set *set,
+static struct request_queue *blk_mq_init_queue_data(struct blk_mq_tag_set *set,
 		void *queuedata)
 {
 	struct request_queue *q;
@@ -3126,7 +3126,6 @@ struct request_queue *blk_mq_init_queue_data(struct blk_mq_tag_set *set,
 	}
 	return q;
 }
-EXPORT_SYMBOL_GPL(blk_mq_init_queue_data);
 
 struct request_queue *blk_mq_init_queue(struct blk_mq_tag_set *set)
 {

block/bsg-lib.c

@@ -45,7 +45,7 @@ static int bsg_transport_fill_hdr(struct request *rq, struct sg_io_v4 *hdr,
 		return PTR_ERR(job->request);
 
 	if (hdr->dout_xfer_len && hdr->din_xfer_len) {
-		job->bidi_rq = blk_get_request(rq->q, REQ_OP_SCSI_IN, 0);
+		job->bidi_rq = blk_get_request(rq->q, REQ_OP_DRV_IN, 0);
 		if (IS_ERR(job->bidi_rq)) {
 			ret = PTR_ERR(job->bidi_rq);
 			goto out;