block: unify disk/part handling and improve ext devt

index 304ec73..b9a6ed1 100644
@@ -310,7 +310,7 @@ int blk_register_queue(struct gendisk *disk)
 	if (!q->request_fn)
 		return 0;
 
-	ret = kobject_add(&q->kobj, kobject_get(&disk->dev.kobj),
+	ret = kobject_add(&q->kobj, kobject_get(&disk_to_dev(disk)->kobj),
 			  "%s", "queue");
 	if (ret < 0)
 		return ret;

@@ -339,6 +339,6 @@ void blk_unregister_queue(struct gendisk *disk)
 
 		kobject_uevent(&q->kobj, KOBJ_REMOVE);
 		kobject_del(&q->kobj);
-		kobject_put(&disk->dev.kobj);
+		kobject_put(&disk_to_dev(disk)->kobj);
 	}
 }

index e2deec2..c54f694 100644
@@ -59,7 +59,7 @@ struct hd_struct *disk_get_part(struct gendisk *disk, int partno)
 	rcu_read_lock();
 	part = rcu_dereference(disk->__part[partno - 1]);
 	if (part)
-		get_device(&part->dev);
+		get_device(part_to_dev(part));
 	rcu_read_unlock();
 
 	return part;

@@ -130,7 +130,7 @@ struct hd_struct *disk_part_iter_next(struct disk_part_iter *piter)
 		if (!(piter->flags & DISK_PITER_INCL_EMPTY) && !part->nr_sects)
 			continue;
 
-		get_device(&part->dev);
+		get_device(part_to_dev(part));
 		piter->part = part;
 		piter->idx++;
 		break;

@@ -434,7 +434,7 @@ static struct kobject *exact_match(dev_t devt, int *partno, void *data)
 {
 	struct gendisk *p = data;
 
-	return &p->dev.kobj;
+	return &disk_to_dev(p)->kobj;
 }
 
 static int exact_lock(dev_t devt, void *data)

@@ -458,7 +458,7 @@ void add_disk(struct gendisk *disk)
 	struct backing_dev_info *bdi;
 
 	disk->flags |= GENHD_FL_UP;
-	disk->dev.devt = MKDEV(disk->major, disk->first_minor);
+	disk_to_dev(disk)->devt = MKDEV(disk->major, disk->first_minor);
 	blk_register_region(disk_devt(disk), disk->minors, NULL,
 			    exact_match, exact_lock, disk);
 	register_disk(disk);

@@ -466,7 +466,7 @@ void add_disk(struct gendisk *disk)
 
 	bdi = &disk->queue->backing_dev_info;
 	bdi_register_dev(bdi, disk_devt(disk));
-	sysfs_create_link(&disk->dev.kobj, &bdi->dev->kobj, "bdi");
+	sysfs_create_link(&disk_to_dev(disk)->kobj, &bdi->dev->kobj, "bdi");
 }
 
 EXPORT_SYMBOL(add_disk);

@@ -474,7 +474,7 @@ EXPORT_SYMBOL(del_gendisk);	/* in partitions/check.c */
 
 void unlink_gendisk(struct gendisk *disk)
 {
-	sysfs_remove_link(&disk->dev.kobj, "bdi");
+	sysfs_remove_link(&disk_to_dev(disk)->kobj, "bdi");
 	bdi_unregister(&disk->queue->backing_dev_info);
 	blk_unregister_queue(disk);
 	blk_unregister_region(disk_devt(disk), disk->minors);

@@ -623,7 +623,7 @@ static void *part_next(struct seq_file *seqf, void *v, loff_t *pos)
 	struct gendisk *gp = v;
 	struct device *dev;
 	++*pos;
-	list_for_each_entry(dev, &gp->dev.node, node) {
+	list_for_each_entry(dev, &disk_to_dev(gp)->node, node) {
 		if (&dev->node == &block_class.devices)
 			return NULL;
 		if (dev->type == &disk_type)

@@ -644,7 +644,7 @@ static int show_partition(struct seq_file *seqf, void *v)
 	struct hd_struct *part;
 	char buf[BDEVNAME_SIZE];
 
-	if (&sgp->dev.node == block_class.devices.next)
+	if (&disk_to_dev(sgp)->node == block_class.devices.next)
 		seq_puts(seqf, "major   minor   #blocks  name\n\n");
 
 	/* Don't show non-partitionable removeable devices or empty devices */

@@ -881,7 +881,7 @@ static void *diskstats_next(struct seq_file *seqf, void *v, loff_t *pos)
 	struct device *dev;
 
 	++*pos;
-	list_for_each_entry(dev, &gp->dev.node, node) {
+	list_for_each_entry(dev, &disk_to_dev(gp)->node, node) {
 		if (&dev->node == &block_class.devices)
 			return NULL;
 		if (dev->type == &disk_type)

@@ -903,7 +903,7 @@ static int diskstats_show(struct seq_file *seqf, void *v)
 	char buf[BDEVNAME_SIZE];
 
 	/*
-	if (&gp->dev.kobj.entry == block_class.devices.next)
+	if (&disk_to_dev(gp)->kobj.entry == block_class.devices.next)
 		seq_puts(seqf,	"major minor name"
 				"     rio rmerge rsect ruse wio wmerge "
 				"wsect wuse running use aveq"

@@ -971,7 +971,7 @@ static void media_change_notify_thread(struct work_struct *work)
 	 * set enviroment vars to indicate which event this is for
 	 * so that user space will know to go check the media status.
 	 */
-	kobject_uevent_env(&gd->dev.kobj, KOBJ_CHANGE, envp);
+	kobject_uevent_env(&disk_to_dev(gd)->kobj, KOBJ_CHANGE, envp);
 	put_device(gd->driverfs_dev);
 }
 

@@ -1063,9 +1063,9 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 		disk->minors = minors;
 		disk->ext_minors = ext_minors;
 		rand_initialize_disk(disk);
-		disk->dev.class = &block_class;
-		disk->dev.type = &disk_type;
-		device_initialize(&disk->dev);
+		disk_to_dev(disk)->class = &block_class;
+		disk_to_dev(disk)->type = &disk_type;
+		device_initialize(disk_to_dev(disk));
 		INIT_WORK(&disk->async_notify,
 			media_change_notify_thread);
 	}

@@ -1087,7 +1087,7 @@ struct kobject *get_disk(struct gendisk *disk)
 	owner = disk->fops->owner;
 	if (owner && !try_module_get(owner))
 		return NULL;
-	kobj = kobject_get(&disk->dev.kobj);
+	kobj = kobject_get(&disk_to_dev(disk)->kobj);
 	if (kobj == NULL) {
 		module_put(owner);
 		return NULL;

@@ -1101,7 +1101,7 @@ EXPORT_SYMBOL(get_disk);
 void put_disk(struct gendisk *disk)
 {
 	if (disk)
-		kobject_put(&disk->dev.kobj);
+		kobject_put(&disk_to_dev(disk)->kobj);
 }
 
 EXPORT_SYMBOL(put_disk);

index 0c39782..3edb6cb 100644
@@ -109,12 +109,12 @@ static const struct attribute_group attr_group = {
 static int
 aoedisk_add_sysfs(struct aoedev *d)
 {
-	return sysfs_create_group(&d->gd->dev.kobj, &attr_group);
+	return sysfs_create_group(&disk_to_dev(d->gd)->kobj, &attr_group);
 }
 void
 aoedisk_rm_sysfs(struct aoedev *d)
 {
-	sysfs_remove_group(&d->gd->dev.kobj, &attr_group);
+	sysfs_remove_group(&disk_to_dev(d->gd)->kobj, &attr_group);
 }
 
 static int

index ad98dda..00337c8 100644
@@ -403,7 +403,7 @@ static int nbd_do_it(struct nbd_device *lo)
 	BUG_ON(lo->magic != LO_MAGIC);
 
 	lo->pid = current->pid;
-	ret = sysfs_create_file(&lo->disk->dev.kobj, &pid_attr.attr);
+	ret = sysfs_create_file(&disk_to_dev(lo->disk)->kobj, &pid_attr.attr);
 	if (ret) {
 		printk(KERN_ERR "nbd: sysfs_create_file failed!");
 		return ret;

@@ -412,7 +412,7 @@ static int nbd_do_it(struct nbd_device *lo)
 	while ((req = nbd_read_stat(lo)) != NULL)
 		nbd_end_request(req);
 
-	sysfs_remove_file(&lo->disk->dev.kobj, &pid_attr.attr);
+	sysfs_remove_file(&disk_to_dev(lo->disk)->kobj, &pid_attr.attr);
 	return 0;
 }
 

index 26e68b6..2e020de 100644
@@ -1133,7 +1133,7 @@ static struct kobject *exact_match(dev_t dev, int *part, void *data)
 {
 	struct gendisk *p = data;
 	*part &= (1 << PARTN_BITS) - 1;
-	return &p->dev.kobj;
+	return &disk_to_dev(p)->kobj;
 }
 
 static int exact_lock(dev_t dev, void *data)

index f76fb06..47d92be 100644
@@ -1136,7 +1136,7 @@ static void event_callback(void *context)
 	list_splice_init(&md->uevent_list, &uevents);
 	spin_unlock_irqrestore(&md->uevent_lock, flags);
 
-	dm_send_uevents(&uevents, &md->disk->dev.kobj);
+	dm_send_uevents(&uevents, &disk_to_dev(md->disk)->kobj);
 
 	atomic_inc(&md->event_nr);
 	wake_up(&md->eventq);

@@ -1593,7 +1593,7 @@ out:
  *---------------------------------------------------------------*/
 void dm_kobject_uevent(struct mapped_device *md)
 {
-	kobject_uevent(&md->disk->dev.kobj, KOBJ_CHANGE);
+	kobject_uevent(&disk_to_dev(md->disk)->kobj, KOBJ_CHANGE);
 }
 
 uint32_t dm_next_uevent_seq(struct mapped_device *md)

index 2580ac1..4d6ef34 100644
@@ -1416,9 +1416,9 @@ static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 		goto fail;
 
 	if (rdev->bdev->bd_part)
-		ko = &rdev->bdev->bd_part->dev.kobj;
+		ko = &part_to_dev(rdev->bdev->bd_part)->kobj;
 	else
-		ko = &rdev->bdev->bd_disk->dev.kobj;
+		ko = &disk_to_dev(rdev->bdev->bd_disk)->kobj;
 	if ((err = sysfs_create_link(&rdev->kobj, ko, "block"))) {
 		kobject_del(&rdev->kobj);
 		goto fail;

@@ -3327,8 +3327,8 @@ static struct kobject *md_probe(dev_t dev, int *part, void *data)
 	add_disk(disk);
 	mddev->gendisk = disk;
 	mutex_unlock(&disks_mutex);
-	error = kobject_init_and_add(&mddev->kobj, &md_ktype, &disk->dev.kobj,
-				     "%s", "md");
+	error = kobject_init_and_add(&mddev->kobj, &md_ktype,
+				     &disk_to_dev(disk)->kobj, "%s", "md");
 	if (error)
 		printk(KERN_WARNING "md: cannot register %s/md - name in use\n",
 		       disk->disk_name);

@@ -3608,7 +3608,7 @@ static int do_md_run(mddev_t * mddev)
 
 	mddev->changed = 1;
 	md_new_event(mddev);
-	kobject_uevent(&mddev->gendisk->dev.kobj, KOBJ_CHANGE);
+	kobject_uevent(&disk_to_dev(mddev->gendisk)->kobj, KOBJ_CHANGE);
 	return 0;
 }
 

index 5d38203..5f177bd 100644
@@ -543,9 +543,9 @@ EXPORT_SYMBOL(bd_release);
 static struct kobject *bdev_get_kobj(struct block_device *bdev)
 {
 	if (bdev->bd_contains != bdev)
-		return kobject_get(&bdev->bd_part->dev.kobj);
+		return kobject_get(&part_to_dev(bdev->bd_part)->kobj);
 	else
-		return kobject_get(&bdev->bd_disk->dev.kobj);
+		return kobject_get(&disk_to_dev(bdev->bd_disk)->kobj);
 }
 
 static struct kobject *bdev_get_holder(struct block_device *bdev)

179 lines not on this page · raw message

index cb6d9a7..96d7ac5 100644
@@ -15,9 +15,11 @@
 
 #ifdef CONFIG_BLOCK
 
-#define kobj_to_dev(k) container_of(k, struct device, kobj)
-#define dev_to_disk(device) container_of(device, struct gendisk, dev)
-#define dev_to_part(device) container_of(device, struct hd_struct, dev)
+#define kobj_to_dev(k)		container_of((k), struct device, kobj)
+#define dev_to_disk(device)	container_of((device), struct gendisk, __dev)
+#define dev_to_part(device)	container_of((device), struct hd_struct, __dev)
+#define disk_to_dev(disk)	(&((disk)->__dev))
+#define part_to_dev(part)	(&((part)->__dev))
 
 extern struct device_type part_type;
 extern struct kobject *block_depr;

@@ -88,7 +90,7 @@ struct disk_stats {
 struct hd_struct {
 	sector_t start_sect;
 	sector_t nr_sects;
-	struct device dev;
+	struct device __dev;
 	struct kobject *holder_dir;
 	int policy, partno;
 #ifdef CONFIG_FAIL_MAKE_REQUEST

@@ -140,7 +142,7 @@ struct gendisk {
 
 	int flags;
 	struct device *driverfs_dev;  // FIXME: remove
-	struct device dev;
+	struct device __dev;
 	struct kobject *holder_dir;
 	struct kobject *slave_dir;
 

@@ -161,7 +163,7 @@ struct gendisk {
 static inline struct gendisk *part_to_disk(struct hd_struct *part)
 {
 	if (likely(part))
-		return dev_to_disk((part)->dev.parent);
+		return dev_to_disk(part_to_dev(part)->parent);
 	return NULL;
 }
 

@@ -172,12 +174,12 @@ static inline int disk_max_parts(struct gendisk *disk)
 
 static inline dev_t disk_devt(struct gendisk *disk)
 {
-	return disk->dev.devt;
+	return disk_to_dev(disk)->devt;
 }
 
 static inline dev_t part_devt(struct hd_struct *part)
 {
-	return part->dev.devt;
+	return part_to_dev(part)->devt;
 }
 
 extern struct hd_struct *disk_get_part(struct gendisk *disk, int partno);

@@ -185,7 +187,7 @@ extern struct hd_struct *disk_get_part(struct gendisk *disk, int partno);
 static inline void disk_put_part(struct hd_struct *part)
 {
 	if (likely(part))
-		put_device(&part->dev);
+		put_device(part_to_dev(part));
 }
 
 /*

index c54f694..07f1597 100644
@@ -54,10 +54,10 @@ struct hd_struct *disk_get_part(struct gendisk *disk, int partno)
 {
 	struct hd_struct *part;
 
-	if (unlikely(partno < 1 || partno > disk_max_parts(disk)))
+	if (unlikely(partno < 0 || partno >= disk_max_parts(disk)))
 		return NULL;
 	rcu_read_lock();
-	part = rcu_dereference(disk->__part[partno - 1]);
+	part = rcu_dereference(disk->__part[partno]);
 	if (part)
 		get_device(part_to_dev(part));
 	rcu_read_unlock();

@@ -85,8 +85,10 @@ void disk_part_iter_start(struct disk_part_iter *piter, struct gendisk *disk,
 
 	if (flags & DISK_PITER_REVERSE)
 		piter->idx = disk_max_parts(piter->disk) - 1;
-	else
+	else if (flags & DISK_PITER_INCL_PART0)
 		piter->idx = 0;
+	else
+		piter->idx = 1;
 
 	piter->flags = flags;
 }

@@ -114,7 +116,10 @@ struct hd_struct *disk_part_iter_next(struct disk_part_iter *piter)
 	/* determine iteration parameters */
 	if (piter->flags & DISK_PITER_REVERSE) {
 		inc = -1;
-		end = -1;
+		if (piter->flags & DISK_PITER_INCL_PART0)
+			end = -1;
+		else
+			end = 0;
 	} else {
 		inc = 1;
 		end = disk_max_parts(piter->disk);

@@ -177,7 +182,7 @@ struct hd_struct *disk_map_sector_rcu(struct gendisk *disk, sector_t sector)
 {
 	int i;
 
-	for (i = 0; i < disk_max_parts(disk); i++) {
+	for (i = 1; i < disk_max_parts(disk); i++) {
 		struct hd_struct *part = rcu_dereference(disk->__part[i]);
 
 		if (part && part->start_sect <= sector &&

@@ -648,7 +653,7 @@ static int show_partition(struct seq_file *seqf, void *v)
 		seq_puts(seqf, "major   minor   #blocks  name\n\n");
 
 	/* Don't show non-partitionable removeable devices or empty devices */
-	if (!get_capacity(sgp) || (!disk_max_parts(sgp) &&
+	if (!get_capacity(sgp) || (!disk_partitionable(sgp) &&
 				   (sgp->flags & GENHD_FL_REMOVABLE)))
 		return 0;
 	if (sgp->flags & GENHD_FL_SUPPRESS_PARTITION_INFO)

@@ -718,7 +723,7 @@ static ssize_t disk_ext_range_show(struct device *dev,
 {
 	struct gendisk *disk = dev_to_disk(dev);
 
-	return sprintf(buf, "%d\n", disk_max_parts(disk) + 1);
+	return sprintf(buf, "%d\n", disk_max_parts(disk));
 }
 
 static ssize_t disk_removable_show(struct device *dev,

@@ -997,7 +1002,7 @@ dev_t blk_lookup_devt(const char *name, int partno)
 			continue;
 		if (strcmp(dev->bus_id, name))
 			continue;
-		if (partno < 0 || partno > disk_max_parts(disk))
+		if (partno < 0 || partno >= disk_max_parts(disk))
 			continue;
 
 		if (partno == 0)

@@ -1045,21 +1050,22 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 				GFP_KERNEL | __GFP_ZERO, node_id);
 	if (disk) {
 		int tot_minors = minors + ext_minors;
+		int size = tot_minors * sizeof(struct hd_struct *);
 
 		if (!init_disk_stats(disk)) {
 			kfree(disk);
 			return NULL;
 		}
-		if (tot_minors > 1) {
-			int size = (tot_minors - 1) * sizeof(struct hd_struct *);
-			disk->__part = kmalloc_node(size,
-				GFP_KERNEL | __GFP_ZERO, node_id);
-			if (!disk->__part) {
-				free_disk_stats(disk);
-				kfree(disk);
-				return NULL;
-			}
+
+		disk->__part = kmalloc_node(size, GFP_KERNEL | __GFP_ZERO,
+					    node_id);
+		if (!disk->__part) {
+			free_disk_stats(disk);
+			kfree(disk);
+			return NULL;
 		}
+		disk->__part[0] = &disk->part0;
+
 		disk->minors = minors;
 		disk->ext_minors = ext_minors;
 		rand_initialize_disk(disk);

index 75aeead..67d2813 100644
@@ -30,7 +30,7 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user
 	if (bdev != bdev->bd_contains)
 		return -EINVAL;
 	partno = p.pno;
-	if (partno <= 0 || partno > disk_max_parts(disk))
+	if (partno <= 0 || partno >= disk_max_parts(disk))
 		return -EINVAL;
 	switch (a.op) {
 		case BLKPG_ADD_PARTITION:

@@ -104,7 +104,7 @@ static int blkdev_reread_part(struct block_device *bdev)
 	struct gendisk *disk = bdev->bd_disk;
 	int res;
 
-	if (!disk_max_parts(disk) || bdev != bdev->bd_contains)
+	if (!disk_partitionable(disk) || bdev != bdev->bd_contains)
 		return -EINVAL;
 	if (!capable(CAP_SYS_ADMIN))
 		return -EACCES;

index 5f177bd..4fd822d 100644
@@ -892,7 +892,7 @@ int check_disk_change(struct block_device *bdev)
 
 	if (bdops->revalidate_disk)
 		bdops->revalidate_disk(bdev->bd_disk);
-	if (disk_max_parts(bdev->bd_disk))
+	if (disk_partitionable(bdev->bd_disk))
 		bdev->bd_invalidated = 1;
 	return 1;
 }

index d12ac7d..d4e677d 100644
@@ -173,7 +173,7 @@ check_partition(struct gendisk *hd, struct block_device *bdev)
 	if (isdigit(state->name[strlen(state->name)-1]))
 		sprintf(state->name, "p");
 
-	state->limit = disk_max_parts(hd) + 1;
+	state->limit = disk_max_parts(hd);
 	i = res = err = 0;
 	while (!res && check_part[i]) {
 		memset(&state->parts, 0, sizeof(state->parts));

@@ -338,12 +338,12 @@ void delete_partition(struct gendisk *disk, int partno)
 {
 	struct hd_struct *part;
 
-	part = disk->__part[partno-1];
+	part = disk->__part[partno];
 	if (!part)
 		return;
 
 	blk_free_devt(part_devt(part));
-	rcu_assign_pointer(disk->__part[partno-1], NULL);
+	rcu_assign_pointer(disk->__part[partno], NULL);
 	kobject_put(part->holder_dir);
 	device_del(part_to_dev(part));
 

@@ -368,7 +368,7 @@ int add_partition(struct gendisk *disk, int partno,
 	const char *dname;
 	int err;
 
-	if (disk->__part[partno - 1]) {
+	if (disk->__part[partno]) {
 		err = -EBUSY;
 		goto fail;
 	}

@@ -403,7 +403,7 @@ int add_partition(struct gendisk *disk, int partno,
 	pdev->devt = devt;
 
 	INIT_RCU_HEAD(&p->rcu_head);
-	rcu_assign_pointer(disk->__part[partno-1], p);
+	rcu_assign_pointer(disk->__part[partno], p);
 
 	/* delay uevent until 'holders' subdir is created */
 	pdev->uevent_suppress = 1;

@@ -466,7 +466,7 @@ void register_disk(struct gendisk *disk)
 	disk_sysfs_add_subdirs(disk);
 
 	/* No minors to use for partitions */
-	if (!disk_max_parts(disk))
+	if (!disk_partitionable(disk))
 		goto exit;
 
 	/* No such device (e.g., media were just removed) */

index 96d7ac5..9018b11 100644
@@ -128,12 +128,13 @@ struct gendisk {
 
 	char disk_name[32];		/* name of major driver */
 
-	/* Array of pointers to partitions indexed by partno - 1.
+	/* Array of pointers to partitions indexed by partno.
 	 * Protected with matching bdev lock but stat and other
 	 * non-critical accesses use RCU.  Always access through
 	 * helpers.
 	 */
 	struct hd_struct **__part;
+	struct hd_struct part0;
 
 	struct block_device_operations *fops;
 	struct request_queue *queue;

@@ -169,7 +170,12 @@ static inline struct gendisk *part_to_disk(struct hd_struct *part)
 
 static inline int disk_max_parts(struct gendisk *disk)
 {
-	return disk->minors + disk->ext_minors - 1;
+	return disk->minors + disk->ext_minors;
+}
+
+static inline bool disk_partitionable(struct gendisk *disk)
+{
+	return disk_max_parts(disk) > 1;
 }
 
 static inline dev_t disk_devt(struct gendisk *disk)

@@ -195,6 +201,7 @@ static inline void disk_put_part(struct hd_struct *part)
  */
 #define DISK_PITER_REVERSE	(1 << 0) /* iterate in the reverse direction */
 #define DISK_PITER_INCL_EMPTY	(1 << 1) /* include 0-sized parts */
+#define DISK_PITER_INCL_PART0	(1 << 2) /* include partition 0 */
 
 struct disk_part_iter {
 	struct gendisk		*disk;

index 3edb6cb..aa69556 100644
@@ -276,7 +276,7 @@ aoeblk_gdalloc(void *vp)
 	gd->first_minor = d->sysminor * AOE_PARTITIONS;
 	gd->fops = &aoe_bdops;
 	gd->private_data = d;
-	gd->capacity = d->ssize;
+	set_capacity(gd, d->ssize);
 	snprintf(gd->disk_name, sizeof gd->disk_name, "etherd/e%ld.%d",
 		d->aoemajor, d->aoeminor);
 

index 6710df3..b13e0b7 100644
@@ -645,7 +645,7 @@ aoecmd_sleepwork(struct work_struct *work)
 		unsigned long flags;
 		u64 ssize;
 
-		ssize = d->gd->capacity;
+		ssize = get_capacity(d->gd);
 		bd = bdget_disk(d->gd, 0);
 
 		if (bd) {

@@ -707,7 +707,7 @@ ataid_complete(struct aoedev *d, struct aoetgt *t, unsigned char *id)
 	if (d->flags & (DEVFL_GDALLOC|DEVFL_NEWSIZE))
 		return;
 	if (d->gd != NULL) {
-		d->gd->capacity = ssize;
+		set_capacity(d->gd, ssize);
 		d->flags |= DEVFL_NEWSIZE;
 	} else
 		d->flags |= DEVFL_GDALLOC;

index a1d813a..6a8038d 100644
@@ -91,7 +91,7 @@ aoedev_downdev(struct aoedev *d)
 	}
 
 	if (d->gd)
-		d->gd->capacity = 0;
+		set_capacity(d->gd, 0);
 
 	d->flags &= ~DEVFL_UP;
 }

index d4e677d..d504c1a 100644
@@ -582,7 +582,7 @@ void del_gendisk(struct gendisk *disk)
 		delete_partition(disk, part->partno);
 	}
 	invalidate_partition(disk, 0);
-	disk->capacity = 0;
+	set_capacity(disk, 0);
 	disk->flags &= ~GENHD_FL_UP;
 	unlink_gendisk(disk);
 	disk_stat_set_all(disk, 0);

index 9018b11..0c15733 100644
@@ -139,7 +139,6 @@ struct gendisk {
 	struct block_device_operations *fops;
 	struct request_queue *queue;
 	void *private_data;
-	sector_t capacity;
 
 	int flags;
 	struct device *driverfs_dev;  // FIXME: remove

@@ -397,11 +396,11 @@ static inline sector_t get_start_sect(struct block_device *bdev)
 }
 static inline sector_t get_capacity(struct gendisk *disk)
 {
-	return disk->capacity;
+	return disk->part0.nr_sects;
 }
 static inline void set_capacity(struct gendisk *disk, sector_t size)
 {
-	disk->capacity = size;
+	disk->part0.nr_sects = size;
 }
 
 #ifdef CONFIG_SOLARIS_X86_PARTITION

index 07f1597..25f76d5 100644
@@ -532,22 +532,15 @@ struct gendisk *get_gendisk(dev_t devt, int *partno)
  */
 extern struct block_device *bdget_disk(struct gendisk *disk, int partno)
 {
-	dev_t devt = MKDEV(0, 0);
+	struct hd_struct *part;
+	struct block_device *bdev = NULL;
 
-	if (partno == 0)
-		devt = disk_devt(disk);
-	else {
-		struct hd_struct *part;
+	part = disk_get_part(disk, partno);
+	if (part && part->nr_sects)
+		bdev = bdget(part_devt(part));
+	disk_put_part(part);
 
-		part = disk_get_part(disk, partno);
-		if (part && part->nr_sects)
-			devt = part_devt(part);
-		disk_put_part(part);
-	}
-
-	if (likely(devt != MKDEV(0, 0)))
-		return bdget(devt);
-	return NULL;
+	return bdev;
 }
 EXPORT_SYMBOL(bdget_disk);
 

@@ -997,29 +990,20 @@ dev_t blk_lookup_devt(const char *name, int partno)
 	down(&block_class.sem);
 	list_for_each_entry(dev, &block_class.devices, node) {
 		struct gendisk *disk = dev_to_disk(dev);
+		struct hd_struct *part;
 
 		if (dev->type != &disk_type)
 			continue;
 		if (strcmp(dev->bus_id, name))
 			continue;
-		if (partno < 0 || partno >= disk_max_parts(disk))
-			continue;
-
-		if (partno == 0)
-			devt = disk_devt(disk);
-		else {
-			struct hd_struct *part;
-
-			part = disk_get_part(disk, partno);
-			if (!part || !part->nr_sects) {
-				disk_put_part(part);
-				continue;
-			}
 
+		part = disk_get_part(disk, partno);
+		if (part && part->nr_sects) {
 			devt = part_devt(part);
 			disk_put_part(part);
+			break;
 		}
-		break;
+		disk_put_part(part);
 	}
 	up(&block_class.sem);
 

index 0c15733..c677bae 100644
@@ -16,9 +16,9 @@
 #ifdef CONFIG_BLOCK
 
 #define kobj_to_dev(k)		container_of((k), struct device, kobj)
-#define dev_to_disk(device)	container_of((device), struct gendisk, __dev)
+#define dev_to_disk(device)	container_of((device), struct gendisk, part0.__dev)
 #define dev_to_part(device)	container_of((device), struct hd_struct, __dev)
-#define disk_to_dev(disk)	(&((disk)->__dev))
+#define disk_to_dev(disk)	(&(disk)->part0.__dev)
 #define part_to_dev(part)	(&((part)->__dev))
 
 extern struct device_type part_type;

@@ -142,7 +142,6 @@ struct gendisk {
 
 	int flags;
 	struct device *driverfs_dev;  // FIXME: remove
-	struct device __dev;
 	struct kobject *holder_dir;
 	struct kobject *slave_dir;
 

@@ -162,8 +161,12 @@ struct gendisk {
 
 static inline struct gendisk *part_to_disk(struct hd_struct *part)
 {
-	if (likely(part))
-		return dev_to_disk(part_to_dev(part)->parent);
+	if (likely(part)) {
+		if (part->partno)
+			return dev_to_disk(part_to_dev(part)->parent);
+		else
+			return dev_to_disk(part_to_dev(part));
+	}
 	return NULL;
 }
 

index 07f1597..25f76d5 100644
@@ -532,22 +532,15 @@ struct gendisk *get_gendisk(dev_t devt, int *partno)
  */
 extern struct block_device *bdget_disk(struct gendisk *disk, int partno)
 {
-	dev_t devt = MKDEV(0, 0);
+	struct hd_struct *part;
+	struct block_device *bdev = NULL;
 
-	if (partno == 0)
-		devt = disk_devt(disk);
-	else {
-		struct hd_struct *part;
+	part = disk_get_part(disk, partno);
+	if (part && part->nr_sects)
+		bdev = bdget(part_devt(part));
+	disk_put_part(part);
 
-		part = disk_get_part(disk, partno);
-		if (part && part->nr_sects)
-			devt = part_devt(part);
-		disk_put_part(part);
-	}
-
-	if (likely(devt != MKDEV(0, 0)))
-		return bdget(devt);
-	return NULL;
+	return bdev;
 }
 EXPORT_SYMBOL(bdget_disk);
 

@@ -997,29 +990,20 @@ dev_t blk_lookup_devt(const char *name, int partno)
 	down(&block_class.sem);
 	list_for_each_entry(dev, &block_class.devices, node) {
 		struct gendisk *disk = dev_to_disk(dev);
+		struct hd_struct *part;
 
 		if (dev->type != &disk_type)
 			continue;
 		if (strcmp(dev->bus_id, name))
 			continue;
-		if (partno < 0 || partno >= disk_max_parts(disk))
-			continue;
-
-		if (partno == 0)
-			devt = disk_devt(disk);
-		else {
-			struct hd_struct *part;
-
-			part = disk_get_part(disk, partno);
-			if (!part || !part->nr_sects) {
-				disk_put_part(part);
-				continue;
-			}
 
+		part = disk_get_part(disk, partno);
+		if (part && part->nr_sects) {
 			devt = part_devt(part);
 			disk_put_part(part);
+			break;
 		}
-		break;
+		disk_put_part(part);
 	}
 	up(&block_class.sem);
 

index 0c15733..c677bae 100644
@@ -16,9 +16,9 @@
 #ifdef CONFIG_BLOCK
 
 #define kobj_to_dev(k)		container_of((k), struct device, kobj)
-#define dev_to_disk(device)	container_of((device), struct gendisk, __dev)
+#define dev_to_disk(device)	container_of((device), struct gendisk, part0.__dev)
 #define dev_to_part(device)	container_of((device), struct hd_struct, __dev)
-#define disk_to_dev(disk)	(&((disk)->__dev))
+#define disk_to_dev(disk)	(&(disk)->part0.__dev)
 #define part_to_dev(part)	(&((part)->__dev))
 
 extern struct device_type part_type;

@@ -142,7 +142,6 @@ struct gendisk {
 
 	int flags;
 	struct device *driverfs_dev;  // FIXME: remove
-	struct device __dev;
 	struct kobject *holder_dir;
 	struct kobject *slave_dir;
 

@@ -162,8 +161,12 @@ struct gendisk {
 
 static inline struct gendisk *part_to_disk(struct hd_struct *part)
 {
-	if (likely(part))
-		return dev_to_disk(part_to_dev(part)->parent);
+	if (likely(part)) {
+		if (part->partno)
+			return dev_to_disk(part_to_dev(part)->parent);
+		else
+			return dev_to_disk(part_to_dev(part));
+	}
 	return NULL;
 }
 

index 25f76d5..c5b45e9 100644
@@ -728,14 +728,6 @@ static ssize_t disk_removable_show(struct device *dev,
 		       (disk->flags & GENHD_FL_REMOVABLE ? 1 : 0));
 }
 
-static ssize_t disk_size_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
-{
-	struct gendisk *disk = dev_to_disk(dev);
-
-	return sprintf(buf, "%llu\n", (unsigned long long)get_capacity(disk));
-}
-
 static ssize_t disk_capability_show(struct device *dev,
 				    struct device_attribute *attr, char *buf)
 {

@@ -801,7 +793,7 @@ static ssize_t disk_fail_store(struct device *dev,
 static DEVICE_ATTR(range, S_IRUGO, disk_range_show, NULL);
 static DEVICE_ATTR(ext_range, S_IRUGO, disk_ext_range_show, NULL);
 static DEVICE_ATTR(removable, S_IRUGO, disk_removable_show, NULL);
-static DEVICE_ATTR(size, S_IRUGO, disk_size_show, NULL);
+static DEVICE_ATTR(size, S_IRUGO, part_size_show, NULL);
 static DEVICE_ATTR(capability, S_IRUGO, disk_capability_show, NULL);
 static DEVICE_ATTR(stat, S_IRUGO, disk_stat_show, NULL);
 #ifdef CONFIG_FAIL_MAKE_REQUEST

index d504c1a..bc4453a 100644
@@ -208,8 +208,8 @@ static ssize_t part_start_show(struct device *dev,
 	return sprintf(buf, "%llu\n",(unsigned long long)p->start_sect);
 }
 
-static ssize_t part_size_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
+ssize_t part_size_show(struct device *dev,
+		       struct device_attribute *attr, char *buf)
 {
 	struct hd_struct *p = dev_to_part(dev);
 	return sprintf(buf, "%llu\n",(unsigned long long)p->nr_sects);

index c677bae..8933ba8 100644
@@ -577,6 +577,9 @@ extern void blk_register_region(dev_t devt, unsigned long range,
 			void *data);
 extern void blk_unregister_region(dev_t devt, unsigned long range);
 
+extern ssize_t part_size_show(struct device *dev,
+			      struct device_attribute *attr, char *buf);
+
 #else /* CONFIG_BLOCK */
 
 static inline void printk_all_partitions(void) { }

index 25f76d5..c5b45e9 100644
@@ -728,14 +728,6 @@ static ssize_t disk_removable_show(struct device *dev,
 		       (disk->flags & GENHD_FL_REMOVABLE ? 1 : 0));
 }
 
-static ssize_t disk_size_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
-{
-	struct gendisk *disk = dev_to_disk(dev);
-
-	return sprintf(buf, "%llu\n", (unsigned long long)get_capacity(disk));
-}
-
 static ssize_t disk_capability_show(struct device *dev,
 				    struct device_attribute *attr, char *buf)
 {

@@ -801,7 +793,7 @@ static ssize_t disk_fail_store(struct device *dev,
 static DEVICE_ATTR(range, S_IRUGO, disk_range_show, NULL);
 static DEVICE_ATTR(ext_range, S_IRUGO, disk_ext_range_show, NULL);
 static DEVICE_ATTR(removable, S_IRUGO, disk_removable_show, NULL);
-static DEVICE_ATTR(size, S_IRUGO, disk_size_show, NULL);
+static DEVICE_ATTR(size, S_IRUGO, part_size_show, NULL);
 static DEVICE_ATTR(capability, S_IRUGO, disk_capability_show, NULL);
 static DEVICE_ATTR(stat, S_IRUGO, disk_stat_show, NULL);
 #ifdef CONFIG_FAIL_MAKE_REQUEST

index d504c1a..bc4453a 100644
@@ -208,8 +208,8 @@ static ssize_t part_start_show(struct device *dev,
 	return sprintf(buf, "%llu\n",(unsigned long long)p->start_sect);
 }
 
-static ssize_t part_size_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
+ssize_t part_size_show(struct device *dev,
+		       struct device_attribute *attr, char *buf)
 {
 	struct hd_struct *p = dev_to_part(dev);
 	return sprintf(buf, "%llu\n",(unsigned long long)p->nr_sects);

index c677bae..8933ba8 100644
@@ -577,6 +577,9 @@ extern void blk_register_region(dev_t devt, unsigned long range,
 			void *data);
 extern void blk_unregister_region(dev_t devt, unsigned long range);
 
+extern ssize_t part_size_show(struct device *dev,
+			      struct device_attribute *attr, char *buf);
+
 #else /* CONFIG_BLOCK */
 
 static inline void printk_all_partitions(void) { }

index c5b45e9..ff33b8c 100644
@@ -1090,10 +1090,7 @@ EXPORT_SYMBOL(put_disk);
 
 void set_device_ro(struct block_device *bdev, int flag)
 {
-	if (bdev->bd_contains != bdev)
-		bdev->bd_part->policy = flag;
-	else
-		bdev->bd_disk->policy = flag;
+	bdev->bd_part->policy = flag;
 }
 
 EXPORT_SYMBOL(set_device_ro);

@@ -1103,8 +1100,8 @@ void set_disk_ro(struct gendisk *disk, int flag)
 	struct disk_part_iter piter;
 	struct hd_struct *part;
 
-	disk->policy = flag;
-	disk_part_iter_start(&piter, disk, DISK_PITER_INCL_EMPTY);
+	disk_part_iter_start(&piter, disk,
+			     DISK_PITER_INCL_EMPTY | DISK_PITER_INCL_PART0);
 	while ((part = disk_part_iter_next(&piter)))
 		part->policy = flag;
 	disk_part_iter_stop(&piter);

@@ -1116,10 +1113,7 @@ int bdev_read_only(struct block_device *bdev)
 {
 	if (!bdev)
 		return 0;
-	else if (bdev->bd_contains != bdev)
-		return bdev->bd_part->policy;
-	else
-		return bdev->bd_disk->policy;
+	return bdev->bd_part->policy;
 }
 
 EXPORT_SYMBOL(bdev_read_only);

index 68e7f19..86752cb 100644
@@ -1126,7 +1126,7 @@ static ide_startstop_t cdrom_start_rw(ide_drive_t *drive, struct request *rq)
 
 	if (write) {
 		/* disk has become write protected */
-		if (cd->disk->policy) {
+		if (get_disk_ro(cd->disk)) {
 			cdrom_end_request(drive, 0);
 			return ide_stopped;
 		}

index c3de311..5b91915 100644
@@ -548,7 +548,7 @@ static int __dev_status(struct mapped_device *md, struct dm_ioctl *param)
 	 */
 	param->open_count = dm_open_count(md);
 
-	if (disk->policy)
+	if (get_disk_ro(disk))
 		param->flags |= DM_READONLY_FLAG;
 
 	param->event_nr = dm_get_event_nr(md);

index bc4453a..bb969a0 100644
@@ -384,7 +384,7 @@ int add_partition(struct gendisk *disk, int partno,
 	p->start_sect = start;
 	p->nr_sects = len;
 	p->partno = partno;
-	p->policy = disk->policy;
+	p->policy = get_disk_ro(disk);
 
 	dname = dev_name(ddev);
 	if (isdigit(dname[strlen(dname) - 1]))

index 8933ba8..0edef09 100644
@@ -146,7 +146,6 @@ struct gendisk {
 	struct kobject *slave_dir;
 
 	struct timer_rand_state *random;
-	int policy;
 
 	atomic_t sync_io;		/* RAID */
 	unsigned long stamp;

@@ -389,6 +388,11 @@ extern struct block_device *bdget_disk(struct gendisk *disk, int partno);
 extern void set_device_ro(struct block_device *bdev, int flag);
 extern void set_disk_ro(struct gendisk *disk, int flag);
 
+static inline int get_disk_ro(struct gendisk *disk)
+{
+	return disk->part0.policy;
+}
+
 /* drivers/char/random.c */
 extern void add_disk_randomness(struct gendisk *disk);
 extern void rand_initialize_disk(struct gendisk *disk);

index 4fd822d..a563e71 100644
@@ -548,14 +548,6 @@ static struct kobject *bdev_get_kobj(struct block_device *bdev)
 		return kobject_get(&disk_to_dev(bdev->bd_disk)->kobj);
 }
 
-static struct kobject *bdev_get_holder(struct block_device *bdev)
-{
-	if (bdev->bd_contains != bdev)
-		return kobject_get(bdev->bd_part->holder_dir);
-	else
-		return kobject_get(bdev->bd_disk->holder_dir);
-}
-
 static int add_symlink(struct kobject *from, struct kobject *to)
 {
 	if (!from || !to)

@@ -608,7 +600,7 @@ static int bd_holder_grab_dirs(struct block_device *bdev,
 	if (!bo->sdev)
 		goto fail_put_hdev;
 
-	bo->hdir = bdev_get_holder(bdev);
+	bo->hdir = kobject_get(bdev->bd_part->holder_dir);
 	if (!bo->hdir)
 		goto fail_put_sdev;
 

index bb969a0..e247b81 100644
@@ -304,23 +304,17 @@ struct device_type part_type = {
 	.release	= part_release,
 };
 
-static inline void partition_sysfs_add_subdir(struct hd_struct *p)
+static inline void partition_sysfs_add_subdir(struct hd_struct *part)
 {
-	struct kobject *k;
-
-	k = kobject_get(&part_to_dev(p)->kobj);
-	p->holder_dir = kobject_create_and_add("holders", k);
-	kobject_put(k);
+	part->holder_dir = kobject_create_and_add("holders",
+						  &part_to_dev(part)->kobj);
 }
 
 static inline void disk_sysfs_add_subdirs(struct gendisk *disk)
 {
-	struct kobject *k;
-
-	k = kobject_get(&disk_to_dev(disk)->kobj);
-	disk->holder_dir = kobject_create_and_add("holders", k);
-	disk->slave_dir = kobject_create_and_add("slaves", k);
-	kobject_put(k);
+	partition_sysfs_add_subdir(&disk->part0);
+	disk->slave_dir = kobject_create_and_add("slaves",
+						 &disk_to_dev(disk)->kobj);
 }
 
 static void delete_partition_rcu_cb(struct rcu_head *head)

@@ -588,7 +582,7 @@ void del_gendisk(struct gendisk *disk)
 	disk_stat_set_all(disk, 0);
 	disk->stamp = 0;
 
-	kobject_put(disk->holder_dir);
+	kobject_put(disk->part0.holder_dir);
 	kobject_put(disk->slave_dir);
 	disk->driverfs_dev = NULL;
 #ifndef CONFIG_SYSFS_DEPRECATED

index 0edef09..bf3d7e3 100644
@@ -142,7 +142,6 @@ struct gendisk {
 
 	int flags;
 	struct device *driverfs_dev;  // FIXME: remove
-	struct kobject *holder_dir;
 	struct kobject *slave_dir;
 
 	struct timer_rand_state *random;

index 4fd822d..a563e71 100644
@@ -548,14 +548,6 @@ static struct kobject *bdev_get_kobj(struct block_device *bdev)
 		return kobject_get(&disk_to_dev(bdev->bd_disk)->kobj);
 }
 
-static struct kobject *bdev_get_holder(struct block_device *bdev)
-{
-	if (bdev->bd_contains != bdev)
-		return kobject_get(bdev->bd_part->holder_dir);
-	else
-		return kobject_get(bdev->bd_disk->holder_dir);
-}
-
 static int add_symlink(struct kobject *from, struct kobject *to)
 {
 	if (!from || !to)

@@ -608,7 +600,7 @@ static int bd_holder_grab_dirs(struct block_device *bdev,
 	if (!bo->sdev)
 		goto fail_put_hdev;
 
-	bo->hdir = bdev_get_holder(bdev);
+	bo->hdir = kobject_get(bdev->bd_part->holder_dir);
 	if (!bo->hdir)
 		goto fail_put_sdev;
 

index bb969a0..e247b81 100644
@@ -304,23 +304,17 @@ struct device_type part_type = {
 	.release	= part_release,
 };
 
-static inline void partition_sysfs_add_subdir(struct hd_struct *p)
+static inline void partition_sysfs_add_subdir(struct hd_struct *part)
 {
-	struct kobject *k;
-
-	k = kobject_get(&part_to_dev(p)->kobj);
-	p->holder_dir = kobject_create_and_add("holders", k);
-	kobject_put(k);
+	part->holder_dir = kobject_create_and_add("holders",
+						  &part_to_dev(part)->kobj);
 }
 
 static inline void disk_sysfs_add_subdirs(struct gendisk *disk)
 {
-	struct kobject *k;
-
-	k = kobject_get(&disk_to_dev(disk)->kobj);
-	disk->holder_dir = kobject_create_and_add("holders", k);
-	disk->slave_dir = kobject_create_and_add("slaves", k);
-	kobject_put(k);
+	partition_sysfs_add_subdir(&disk->part0);
+	disk->slave_dir = kobject_create_and_add("slaves",
+						 &disk_to_dev(disk)->kobj);
 }
 
 static void delete_partition_rcu_cb(struct rcu_head *head)

@@ -588,7 +582,7 @@ void del_gendisk(struct gendisk *disk)
 	disk_stat_set_all(disk, 0);
 	disk->stamp = 0;
 
-	kobject_put(disk->holder_dir);
+	kobject_put(disk->part0.holder_dir);
 	kobject_put(disk->slave_dir);
 	disk->driverfs_dev = NULL;
 #ifndef CONFIG_SYSFS_DEPRECATED

index 0edef09..bf3d7e3 100644
@@ -142,7 +142,6 @@ struct gendisk {
 
 	int flags;
 	struct device *driverfs_dev;  // FIXME: remove
-	struct kobject *holder_dir;
 	struct kobject *slave_dir;
 
 	struct timer_rand_state *random;

index 1da936f..511b8c7 100644
@@ -1263,7 +1263,7 @@ __setup("fail_make_request=", setup_fail_make_request);
 static int should_fail_request(struct bio *bio)
 {
 	if ((bio->bi_bdev->bd_disk->flags & GENHD_FL_FAIL) ||
-	    (bio->bi_bdev->bd_part && bio->bi_bdev->bd_part->make_it_fail))
+	    bio->bi_bdev->bd_part->make_it_fail)
 		return should_fail(&fail_make_request, bio->bi_size);
 
 	return 0;

index 4d6ef34..cff446f 100644
@@ -1415,10 +1415,7 @@ static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 	if ((err = kobject_add(&rdev->kobj, &mddev->kobj, "dev-%s", b)))
 		goto fail;
 
-	if (rdev->bdev->bd_part)
-		ko = &part_to_dev(rdev->bdev->bd_part)->kobj;
-	else
-		ko = &disk_to_dev(rdev->bdev->bd_disk)->kobj;
+	ko = &part_to_dev(rdev->bdev->bd_part)->kobj;
 	if ((err = sysfs_create_link(&rdev->kobj, ko, "block"))) {
 		kobject_del(&rdev->kobj);
 		goto fail;

146 lines not on this page · raw message

index e247b81..931d366 100644
@@ -134,12 +134,7 @@ char *disk_name(struct gendisk *hd, int partno, char *buf)
 
 const char *bdevname(struct block_device *bdev, char *buf)
 {
-	int partno = 0;
-
-	if (bdev->bd_part)
-		partno = bdev->bd_part->partno;
-
-	return disk_name(bdev->bd_disk, partno, buf);
+	return disk_name(bdev->bd_disk, bdev->bd_part->partno, buf);
 }
 
 EXPORT_SYMBOL(bdevname);

index bf3d7e3..8c94f17 100644
@@ -398,7 +398,7 @@ extern void rand_initialize_disk(struct gendisk *disk);
 
 static inline sector_t get_start_sect(struct block_device *bdev)
 {
-	return bdev->bd_contains == bdev ? 0 : bdev->bd_part->start_sect;
+	return bdev->bd_part->start_sect;
 }
 static inline sector_t get_capacity(struct gendisk *disk)
 {

index 511b8c7..332d695 100644
@@ -1262,8 +1262,9 @@ __setup("fail_make_request=", setup_fail_make_request);
 
 static int should_fail_request(struct bio *bio)
 {
-	if ((bio->bi_bdev->bd_disk->flags & GENHD_FL_FAIL) ||
-	    bio->bi_bdev->bd_part->make_it_fail)
+	struct hd_struct *part = bio->bi_bdev->bd_part;
+
+	if (part_to_disk(part)->part0.make_it_fail || part->make_it_fail)
 		return should_fail(&fail_make_request, bio->bi_size);
 
 	return 0;

index ff33b8c..21b8535 100644
@@ -762,34 +762,6 @@ static ssize_t disk_stat_show(struct device *dev,
 		jiffies_to_msecs(disk_stat_read(disk, time_in_queue)));
 }
 
-#ifdef CONFIG_FAIL_MAKE_REQUEST
-static ssize_t disk_fail_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
-{
-	struct gendisk *disk = dev_to_disk(dev);
-
-	return sprintf(buf, "%d\n", disk->flags & GENHD_FL_FAIL ? 1 : 0);
-}
-
-static ssize_t disk_fail_store(struct device *dev,
-			       struct device_attribute *attr,
-			       const char *buf, size_t count)
-{
-	struct gendisk *disk = dev_to_disk(dev);
-	int i;
-
-	if (count > 0 && sscanf(buf, "%d", &i) > 0) {
-		if (i == 0)
-			disk->flags &= ~GENHD_FL_FAIL;
-		else
-			disk->flags |= GENHD_FL_FAIL;
-	}
-
-	return count;
-}
-
-#endif
-
 static DEVICE_ATTR(range, S_IRUGO, disk_range_show, NULL);
 static DEVICE_ATTR(ext_range, S_IRUGO, disk_ext_range_show, NULL);
 static DEVICE_ATTR(removable, S_IRUGO, disk_removable_show, NULL);

@@ -798,7 +770,7 @@ static DEVICE_ATTR(capability, S_IRUGO, disk_capability_show, NULL);
 static DEVICE_ATTR(stat, S_IRUGO, disk_stat_show, NULL);
 #ifdef CONFIG_FAIL_MAKE_REQUEST
 static struct device_attribute dev_attr_fail =
-	__ATTR(make-it-fail, S_IRUGO|S_IWUSR, disk_fail_show, disk_fail_store);
+	__ATTR(make-it-fail, S_IRUGO|S_IWUSR, part_fail_show, part_fail_store);
 #endif
 
 static struct attribute *disk_attrs[] = {

index 931d366..ddd0258 100644
@@ -237,17 +237,17 @@ static ssize_t part_stat_show(struct device *dev,
 }
 
 #ifdef CONFIG_FAIL_MAKE_REQUEST
-static ssize_t part_fail_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
+ssize_t part_fail_show(struct device *dev,
+		       struct device_attribute *attr, char *buf)
 {
 	struct hd_struct *p = dev_to_part(dev);
 
 	return sprintf(buf, "%d\n", p->make_it_fail);
 }
 
-static ssize_t part_fail_store(struct device *dev,
-			       struct device_attribute *attr,
-			       const char *buf, size_t count)
+ssize_t part_fail_store(struct device *dev,
+			struct device_attribute *attr,
+			const char *buf, size_t count)
 {
 	struct hd_struct *p = dev_to_part(dev);
 	int i;

index 8c94f17..61448a9 100644
@@ -112,8 +112,6 @@ struct hd_struct {
 #define GENHD_FL_CD				8
 #define GENHD_FL_UP				16
 #define GENHD_FL_SUPPRESS_PARTITION_INFO	32
-#define GENHD_FL_FAIL				64
-
 
 struct gendisk {
 	/* major, first_minor, minors and ext_minors are input

@@ -582,6 +580,13 @@ extern void blk_unregister_region(dev_t devt, unsigned long range);
 
 extern ssize_t part_size_show(struct device *dev,
 			      struct device_attribute *attr, char *buf);
+#ifdef CONFIG_FAIL_MAKE_REQUEST
+extern ssize_t part_fail_show(struct device *dev,
+			      struct device_attribute *attr, char *buf);
+extern ssize_t part_fail_store(struct device *dev,
+			       struct device_attribute *attr,
+			       const char *buf, size_t count);
+#endif /* CONFIG_FAIL_MAKE_REQUEST */
 
 #else /* CONFIG_BLOCK */
 

index 332d695..d4186f3 100644
@@ -64,14 +64,10 @@ static void drive_stat_acct(struct request *rq, int new_io)
 
 	part = disk_map_sector_rcu(rq->rq_disk, rq->sector);
 	if (!new_io)
-		all_stat_inc(rq->rq_disk, part, merges[rw], rq->sector);
+		part_stat_inc(part, merges[rw]);
 	else {
-		disk_round_stats(rq->rq_disk);
-		rq->rq_disk->in_flight++;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight++;
-		}
+		part_round_stats(part);
+		part_inc_in_flight(part);
 	}
 
 	rcu_read_unlock_preempt();

@@ -966,8 +962,21 @@ static inline void add_request(struct request_queue *q, struct request *req)
 	__elv_add_request(q, req, ELEVATOR_INSERT_SORT, 0);
 }
 
+static void part_round_stats_single(struct hd_struct *part, unsigned long now)
+{
+	if (now == part->stamp)
+		return;
+
+	if (part->in_flight) {
+		__part_stat_add(part, time_in_queue,
+				part->in_flight * (now - part->stamp));
+		__part_stat_add(part, io_ticks, (now - part->stamp));
+	}
+	part->stamp = now;
+}
+
 /**
- * disk_round_stats()	- Round off the performance stats on a struct
+ * part_round_stats()	- Round off the performance stats on a struct
  * disk_stats.
  *
  * The average IO queue length and utilisation statistics are maintained

@@ -984,45 +993,15 @@ static inline void add_request(struct request_queue *q, struct request *req)
  * CONTEXT:
  * Preemption disabled.
  */
-void disk_round_stats(struct gendisk *disk)
-{
-	unsigned long now = jiffies;
-
-	if (now == disk->stamp)
-		return;
-
-	if (disk->in_flight) {
-		disk_stat_add(disk, time_in_queue,
-			      disk->in_flight * (now - disk->stamp));
-		disk_stat_add(disk, io_ticks, (now - disk->stamp));
-	}
-	disk->stamp = now;
-}
-EXPORT_SYMBOL_GPL(disk_round_stats);
-
-/**
- * part_round_stats()	- Round off the performance stats on a struct
- * disk_stats.
- *
- * Please see disk_round_stats() for more information.
- *
- * CONTEXT:
- * Preemption disabled.
- */
 void part_round_stats(struct hd_struct *part)
 {
 	unsigned long now = jiffies;
 
-	if (now == part->stamp)
-		return;
-
-	if (part->in_flight) {
-		part_stat_add(part, time_in_queue,
-			      part->in_flight * (now - part->stamp));
-		part_stat_add(part, io_ticks, (now - part->stamp));
-	}
-	part->stamp = now;
+	if (part->partno)
+		part_round_stats_single(&part_to_disk(part)->part0, now);
+	part_round_stats_single(part, now);
 }
+EXPORT_SYMBOL_GPL(part_round_stats);
 
 /*
  * queue lock must be held

@@ -1557,8 +1536,7 @@ static int __end_that_request_first(struct request *req, int error,
 
 		rcu_read_lock_preempt();
 		part = disk_map_sector_rcu(req->rq_disk, req->sector);
-		all_stat_add(req->rq_disk, part, sectors[rw],
-			     nr_bytes >> 9, req->sector);
+		part_stat_add(part, sectors[rw], nr_bytes >> 9);
 		rcu_read_unlock_preempt();
 	}
 

@@ -1748,15 +1726,10 @@ static void end_that_request_last(struct request *req, int error)
 		rcu_read_lock_preempt();
 
 		part = disk_map_sector_rcu(disk, req->sector);
-
-		all_stat_inc(disk, part, ios[rw], req->sector);
-		all_stat_add(disk, part, ticks[rw], duration, req->sector);
-		disk_round_stats(disk);
-		disk->in_flight--;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight--;
-		}
+		part_stat_inc(part, ios[rw]);
+		part_stat_add(part, ticks[rw], duration);
+		part_round_stats(part);
+		part_dec_in_flight(part);
 
 		rcu_read_unlock_preempt();
 	}

index 55456c8..a25a9f7 100644
@@ -472,12 +472,8 @@ static int attempt_merge(struct request_queue *q, struct request *req,
 		rcu_read_lock_preempt();
 
 		part = disk_map_sector_rcu(req->rq_disk, req->sector);
-		disk_round_stats(req->rq_disk);
-		req->rq_disk->in_flight--;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight--;
-		}
+		part_round_stats(part);
+		part_dec_in_flight(part);
 
 		rcu_read_unlock_preempt();
 	}

154 lines not on this page · raw message

index b13e0b7..5dac67c 100644
@@ -760,10 +760,10 @@ diskstats(struct gendisk *disk, struct bio *bio, ulong duration, sector_t sector
 	rcu_read_lock_preempt();
 
 	part = disk_map_sector_rcu(disk, sector);
-	all_stat_inc(disk, part, ios[rw], sector);
-	all_stat_add(disk, part, ticks[rw], duration, sector);
-	all_stat_add(disk, part, sectors[rw], n_sect, sector);
-	all_stat_add(disk, part, io_ticks, duration, sector);
+	part_stat_inc(part, ios[rw]);
+	part_stat_add(part, ticks[rw], duration);
+	part_stat_add(part, sectors[rw], n_sect);
+	part_stat_add(part, io_ticks, duration);
 
 	rcu_read_unlock_preempt();
 }

index 47d92be..75751ac 100644
@@ -381,9 +381,9 @@ static void start_io_acct(struct dm_io *io)
 	io->start_time = jiffies;
 
 	preempt_disable();
-	disk_round_stats(dm_disk(md));
+	part_round_stats(&dm_disk(md)->part0);
 	preempt_enable();
-	dm_disk(md)->in_flight = atomic_inc_return(&md->pending);
+	dm_disk(md)->part0.in_flight = atomic_inc_return(&md->pending);
 }
 
 static int end_io_acct(struct dm_io *io)

@@ -395,11 +395,12 @@ static int end_io_acct(struct dm_io *io)
 	int rw = bio_data_dir(bio);
 
 	preempt_disable();
-	disk_round_stats(dm_disk(md));
-	disk_stat_add(dm_disk(md), ticks[rw], duration);
+	part_round_stats(&dm_disk(md)->part0);
+	part_stat_add(&dm_disk(md)->part0, ticks[rw], duration);
 	preempt_enable();
 
-	dm_disk(md)->in_flight = pending = atomic_dec_return(&md->pending);
+	dm_disk(md)->part0.in_flight = pending =
+		atomic_dec_return(&md->pending);
 
 	return !pending;
 }

@@ -851,8 +852,8 @@ static int dm_request(struct request_queue *q, struct bio *bio)
 	down_read(&md->io_lock);
 
 	preempt_disable();
-	disk_stat_inc(dm_disk(md), ios[rw]);
-	disk_stat_add(dm_disk(md), sectors[rw], bio_sectors(bio));
+	part_stat_inc(&dm_disk(md)->part0, ios[rw]);
+	part_stat_add(&dm_disk(md)->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	/*

index 7341196..0c01f5a 100644
@@ -323,8 +323,8 @@ static int linear_make_request (struct request_queue *q, struct bio *bio)
 	}
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	tmp_dev = which_dev(mddev, bio->bi_sector);

index cff446f..ac85e47 100644
@@ -5395,8 +5395,8 @@ static int is_mddev_idle(mddev_t *mddev)
 	idle = 1;
 	rdev_for_each(rdev, tmp, mddev) {
 		struct gendisk *disk = rdev->bdev->bd_contains->bd_disk;
-		curr_events = disk_stat_read(disk, sectors[0]) + 
-				disk_stat_read(disk, sectors[1]) - 
+		curr_events = part_stat_read(&disk->part0, sectors[0]) +
+				part_stat_read(&disk->part0, sectors[1]) -
 				atomic_read(&disk->sync_io);
 		/* sync IO will cause sync_io to increase before the disk_stats
 		 * as sync_io is counted when a request starts, and

index 5db0a45..30d1c20 100644
@@ -159,8 +159,8 @@ static int multipath_make_request (struct request_queue *q, struct bio * bio)
 	mp_bh->mddev = mddev;
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	mp_bh->path = multipath_map(conf);

index 2602b05..f7e08bc 100644
@@ -404,8 +404,8 @@ static int raid0_make_request (struct request_queue *q, struct bio *bio)
 	}
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	chunk_size = mddev->chunk_size >> 10;

index 1fd2c3d..8ff3aaa 100644
@@ -805,8 +805,8 @@ static int make_request(struct request_queue *q, struct bio * bio)
 	bitmap = mddev->bitmap;
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	/*

index 2521b73..735c167 100644
@@ -839,8 +839,8 @@ static int make_request(struct request_queue *q, struct bio * bio)
 	wait_barrier(conf);
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	r10_bio = mempool_alloc(conf->r10bio_pool, GFP_NOIO);

index bfd8b6c..8131245 100644
@@ -3581,8 +3581,8 @@ static int make_request(struct request_queue *q, struct bio * bi)
 	md_write_start(mddev, bi);
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bi));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bi));
 	preempt_enable();
 
 	if (rw == READ &&

index ddd0258..e4d4d5c 100644
@@ -210,8 +210,8 @@ ssize_t part_size_show(struct device *dev,
 	return sprintf(buf, "%llu\n",(unsigned long long)p->nr_sects);
 }
 
-static ssize_t part_stat_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
+ssize_t part_stat_show(struct device *dev,
+		       struct device_attribute *attr, char *buf)
 {
 	struct hd_struct *p = dev_to_part(dev);
 

@@ -574,8 +574,8 @@ void del_gendisk(struct gendisk *disk)
 	set_capacity(disk, 0);
 	disk->flags &= ~GENHD_FL_UP;
 	unlink_gendisk(disk);
-	disk_stat_set_all(disk, 0);
-	disk->stamp = 0;
+	part_stat_set_all(&disk->part0, 0);
+	disk->part0.stamp = 0;
 
 	kobject_put(disk->part0.holder_dir);
 	kobject_put(disk->slave_dir);

213 lines not on this page · raw message

index 332d695..d4186f3 100644
@@ -64,14 +64,10 @@ static void drive_stat_acct(struct request *rq, int new_io)
 
 	part = disk_map_sector_rcu(rq->rq_disk, rq->sector);
 	if (!new_io)
-		all_stat_inc(rq->rq_disk, part, merges[rw], rq->sector);
+		part_stat_inc(part, merges[rw]);
 	else {
-		disk_round_stats(rq->rq_disk);
-		rq->rq_disk->in_flight++;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight++;
-		}
+		part_round_stats(part);
+		part_inc_in_flight(part);
 	}
 
 	rcu_read_unlock_preempt();

@@ -966,8 +962,21 @@ static inline void add_request(struct request_queue *q, struct request *req)
 	__elv_add_request(q, req, ELEVATOR_INSERT_SORT, 0);
 }
 
+static void part_round_stats_single(struct hd_struct *part, unsigned long now)
+{
+	if (now == part->stamp)
+		return;
+
+	if (part->in_flight) {
+		__part_stat_add(part, time_in_queue,
+				part->in_flight * (now - part->stamp));
+		__part_stat_add(part, io_ticks, (now - part->stamp));
+	}
+	part->stamp = now;
+}
+
 /**
- * disk_round_stats()	- Round off the performance stats on a struct
+ * part_round_stats()	- Round off the performance stats on a struct
  * disk_stats.
  *
  * The average IO queue length and utilisation statistics are maintained

@@ -984,45 +993,15 @@ static inline void add_request(struct request_queue *q, struct request *req)
  * CONTEXT:
  * Preemption disabled.
  */
-void disk_round_stats(struct gendisk *disk)
-{
-	unsigned long now = jiffies;
-
-	if (now == disk->stamp)
-		return;
-
-	if (disk->in_flight) {
-		disk_stat_add(disk, time_in_queue,
-			      disk->in_flight * (now - disk->stamp));
-		disk_stat_add(disk, io_ticks, (now - disk->stamp));
-	}
-	disk->stamp = now;
-}
-EXPORT_SYMBOL_GPL(disk_round_stats);
-
-/**
- * part_round_stats()	- Round off the performance stats on a struct
- * disk_stats.
- *
- * Please see disk_round_stats() for more information.
- *
- * CONTEXT:
- * Preemption disabled.
- */
 void part_round_stats(struct hd_struct *part)
 {
 	unsigned long now = jiffies;
 
-	if (now == part->stamp)
-		return;
-
-	if (part->in_flight) {
-		part_stat_add(part, time_in_queue,
-			      part->in_flight * (now - part->stamp));
-		part_stat_add(part, io_ticks, (now - part->stamp));
-	}
-	part->stamp = now;
+	if (part->partno)
+		part_round_stats_single(&part_to_disk(part)->part0, now);
+	part_round_stats_single(part, now);
 }
+EXPORT_SYMBOL_GPL(part_round_stats);
 
 /*
  * queue lock must be held

@@ -1557,8 +1536,7 @@ static int __end_that_request_first(struct request *req, int error,
 
 		rcu_read_lock_preempt();
 		part = disk_map_sector_rcu(req->rq_disk, req->sector);
-		all_stat_add(req->rq_disk, part, sectors[rw],
-			     nr_bytes >> 9, req->sector);
+		part_stat_add(part, sectors[rw], nr_bytes >> 9);
 		rcu_read_unlock_preempt();
 	}
 

@@ -1748,15 +1726,10 @@ static void end_that_request_last(struct request *req, int error)
 		rcu_read_lock_preempt();
 
 		part = disk_map_sector_rcu(disk, req->sector);
-
-		all_stat_inc(disk, part, ios[rw], req->sector);
-		all_stat_add(disk, part, ticks[rw], duration, req->sector);
-		disk_round_stats(disk);
-		disk->in_flight--;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight--;
-		}
+		part_stat_inc(part, ios[rw]);
+		part_stat_add(part, ticks[rw], duration);
+		part_round_stats(part);
+		part_dec_in_flight(part);
 
 		rcu_read_unlock_preempt();
 	}

index 55456c8..a25a9f7 100644
@@ -472,12 +472,8 @@ static int attempt_merge(struct request_queue *q, struct request *req,
 		rcu_read_lock_preempt();
 
 		part = disk_map_sector_rcu(req->rq_disk, req->sector);
-		disk_round_stats(req->rq_disk);
-		req->rq_disk->in_flight--;
-		if (part) {
-			part_round_stats(part);
-			part->in_flight--;
-		}
+		part_round_stats(part);
+		part_dec_in_flight(part);
 
 		rcu_read_unlock_preempt();
 	}

154 lines not on this page · raw message

index b13e0b7..5dac67c 100644
@@ -760,10 +760,10 @@ diskstats(struct gendisk *disk, struct bio *bio, ulong duration, sector_t sector
 	rcu_read_lock_preempt();
 
 	part = disk_map_sector_rcu(disk, sector);
-	all_stat_inc(disk, part, ios[rw], sector);
-	all_stat_add(disk, part, ticks[rw], duration, sector);
-	all_stat_add(disk, part, sectors[rw], n_sect, sector);
-	all_stat_add(disk, part, io_ticks, duration, sector);
+	part_stat_inc(part, ios[rw]);
+	part_stat_add(part, ticks[rw], duration);
+	part_stat_add(part, sectors[rw], n_sect);
+	part_stat_add(part, io_ticks, duration);
 
 	rcu_read_unlock_preempt();
 }

index 47d92be..75751ac 100644
@@ -381,9 +381,9 @@ static void start_io_acct(struct dm_io *io)
 	io->start_time = jiffies;
 
 	preempt_disable();
-	disk_round_stats(dm_disk(md));
+	part_round_stats(&dm_disk(md)->part0);
 	preempt_enable();
-	dm_disk(md)->in_flight = atomic_inc_return(&md->pending);
+	dm_disk(md)->part0.in_flight = atomic_inc_return(&md->pending);
 }
 
 static int end_io_acct(struct dm_io *io)

@@ -395,11 +395,12 @@ static int end_io_acct(struct dm_io *io)
 	int rw = bio_data_dir(bio);
 
 	preempt_disable();
-	disk_round_stats(dm_disk(md));
-	disk_stat_add(dm_disk(md), ticks[rw], duration);
+	part_round_stats(&dm_disk(md)->part0);
+	part_stat_add(&dm_disk(md)->part0, ticks[rw], duration);
 	preempt_enable();
 
-	dm_disk(md)->in_flight = pending = atomic_dec_return(&md->pending);
+	dm_disk(md)->part0.in_flight = pending =
+		atomic_dec_return(&md->pending);
 
 	return !pending;
 }

@@ -851,8 +852,8 @@ static int dm_request(struct request_queue *q, struct bio *bio)
 	down_read(&md->io_lock);
 
 	preempt_disable();
-	disk_stat_inc(dm_disk(md), ios[rw]);
-	disk_stat_add(dm_disk(md), sectors[rw], bio_sectors(bio));
+	part_stat_inc(&dm_disk(md)->part0, ios[rw]);
+	part_stat_add(&dm_disk(md)->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	/*

index 7341196..0c01f5a 100644
@@ -323,8 +323,8 @@ static int linear_make_request (struct request_queue *q, struct bio *bio)
 	}
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	tmp_dev = which_dev(mddev, bio->bi_sector);

index cff446f..ac85e47 100644
@@ -5395,8 +5395,8 @@ static int is_mddev_idle(mddev_t *mddev)
 	idle = 1;
 	rdev_for_each(rdev, tmp, mddev) {
 		struct gendisk *disk = rdev->bdev->bd_contains->bd_disk;
-		curr_events = disk_stat_read(disk, sectors[0]) + 
-				disk_stat_read(disk, sectors[1]) - 
+		curr_events = part_stat_read(&disk->part0, sectors[0]) +
+				part_stat_read(&disk->part0, sectors[1]) -
 				atomic_read(&disk->sync_io);
 		/* sync IO will cause sync_io to increase before the disk_stats
 		 * as sync_io is counted when a request starts, and

index 5db0a45..30d1c20 100644
@@ -159,8 +159,8 @@ static int multipath_make_request (struct request_queue *q, struct bio * bio)
 	mp_bh->mddev = mddev;
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	mp_bh->path = multipath_map(conf);

index 2602b05..f7e08bc 100644
@@ -404,8 +404,8 @@ static int raid0_make_request (struct request_queue *q, struct bio *bio)
 	}
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	chunk_size = mddev->chunk_size >> 10;

index 1fd2c3d..8ff3aaa 100644
@@ -805,8 +805,8 @@ static int make_request(struct request_queue *q, struct bio * bio)
 	bitmap = mddev->bitmap;
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	/*

index 2521b73..735c167 100644
@@ -839,8 +839,8 @@ static int make_request(struct request_queue *q, struct bio * bio)
 	wait_barrier(conf);
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bio));
 	preempt_enable();
 
 	r10_bio = mempool_alloc(conf->r10bio_pool, GFP_NOIO);

index bfd8b6c..8131245 100644
@@ -3581,8 +3581,8 @@ static int make_request(struct request_queue *q, struct bio * bi)
 	md_write_start(mddev, bi);
 
 	preempt_disable();
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bi));
+	part_stat_inc(&mddev->gendisk->part0, ios[rw]);
+	part_stat_add(&mddev->gendisk->part0, sectors[rw], bio_sectors(bi));
 	preempt_enable();
 
 	if (rw == READ &&

index ddd0258..e4d4d5c 100644
@@ -210,8 +210,8 @@ ssize_t part_size_show(struct device *dev,
 	return sprintf(buf, "%llu\n",(unsigned long long)p->nr_sects);
 }
 
-static ssize_t part_stat_show(struct device *dev,
-			      struct device_attribute *attr, char *buf)
+ssize_t part_stat_show(struct device *dev,
+		       struct device_attribute *attr, char *buf)
 {
 	struct hd_struct *p = dev_to_part(dev);
 

@@ -574,8 +574,8 @@ void del_gendisk(struct gendisk *disk)
 	set_capacity(disk, 0);
 	disk->flags &= ~GENHD_FL_UP;
 	unlink_gendisk(disk);
-	disk_stat_set_all(disk, 0);
-	disk->stamp = 0;
+	part_stat_set_all(&disk->part0, 0);
+	disk->part0.stamp = 0;
 
 	kobject_put(disk->part0.holder_dir);
 	kobject_put(disk->slave_dir);

213 lines not on this page · raw message

index 5d0f5d1..72aef36 100644
@@ -52,14 +52,21 @@ static struct device_type disk_type;
  */
 struct hd_struct *disk_get_part(struct gendisk *disk, int partno)
 {
-	struct hd_struct *part;
+	struct hd_struct *part = NULL;
+	struct disk_part_tbl *ptbl;
 
-	if (unlikely(partno < 0 || partno >= disk_max_parts(disk)))
+	if (unlikely(partno < 0))
 		return NULL;
+
 	rcu_read_lock();
-	part = rcu_dereference(disk->__part[partno]);
-	if (part)
-		get_device(part_to_dev(part));
+
+	ptbl = rcu_dereference(disk->part_tbl);
+	if (likely(partno < ptbl->len)) {
+		part = rcu_dereference(ptbl->part[partno]);
+		if (part)
+			get_device(part_to_dev(part));
+	}
+
 	rcu_read_unlock();
 
 	return part;

@@ -80,17 +87,24 @@ EXPORT_SYMBOL_GPL(disk_get_part);
 void disk_part_iter_start(struct disk_part_iter *piter, struct gendisk *disk,
 			  unsigned int flags)
 {
+	struct disk_part_tbl *ptbl;
+
+	rcu_read_lock();
+	ptbl = rcu_dereference(disk->part_tbl);
+
 	piter->disk = disk;
 	piter->part = NULL;
 
 	if (flags & DISK_PITER_REVERSE)
-		piter->idx = disk_max_parts(piter->disk) - 1;
+		piter->idx = ptbl->len - 1;
 	else if (flags & DISK_PITER_INCL_PART0)
 		piter->idx = 0;
 	else
 		piter->idx = 1;
 
 	piter->flags = flags;
+
+	rcu_read_unlock();
 }
 EXPORT_SYMBOL_GPL(disk_part_iter_start);
 

@@ -105,13 +119,16 @@ EXPORT_SYMBOL_GPL(disk_part_iter_start);
  */
 struct hd_struct *disk_part_iter_next(struct disk_part_iter *piter)
 {
+	struct disk_part_tbl *ptbl;
 	int inc, end;
 
 	/* put the last partition */
 	disk_put_part(piter->part);
 	piter->part = NULL;
 
+	/* get part_tbl */
 	rcu_read_lock();
+	ptbl = rcu_dereference(piter->disk->part_tbl);
 
 	/* determine iteration parameters */
 	if (piter->flags & DISK_PITER_REVERSE) {

@@ -122,14 +139,14 @@ struct hd_struct *disk_part_iter_next(struct disk_part_iter *piter)
 			end = 0;
 	} else {
 		inc = 1;
-		end = disk_max_parts(piter->disk);
+		end = ptbl->len;
 	}
 
 	/* iterate to the next partition */
 	for (; piter->idx != end; piter->idx += inc) {
 		struct hd_struct *part;
 
-		part = rcu_dereference(piter->disk->__part[piter->idx]);
+		part = rcu_dereference(ptbl->part[piter->idx]);
 		if (!part)
 			continue;
 		if (!(piter->flags & DISK_PITER_INCL_EMPTY) && !part->nr_sects)

@@ -180,10 +197,13 @@ EXPORT_SYMBOL_GPL(disk_part_iter_stop);
  */
 struct hd_struct *disk_map_sector_rcu(struct gendisk *disk, sector_t sector)
 {
+	struct disk_part_tbl *ptbl;
 	int i;
 
-	for (i = 1; i < disk_max_parts(disk); i++) {
-		struct hd_struct *part = rcu_dereference(disk->__part[i]);
+	ptbl = rcu_dereference(disk->part_tbl);
+
+	for (i = 1; i < ptbl->len; i++) {
+		struct hd_struct *part = rcu_dereference(ptbl->part[i]);
 
 		if (part && part->start_sect <= sector &&
 		    sector < part->start_sect + part->nr_sects)

@@ -763,12 +783,86 @@ static struct attribute_group *disk_attr_groups[] = {
 	NULL
 };
 
+static void disk_free_ptbl_rcu_cb(struct rcu_head *head)
+{
+	struct disk_part_tbl *ptbl =
+		container_of(head, struct disk_part_tbl, rcu_head);
+
+	kfree(ptbl);
+}
+
+/**
+ * disk_replace_part_tbl - replace disk->part_tbl in RCU-safe way
+ * @disk: disk to replace part_tbl for
+ * @new_ptbl: new part_tbl to install
+ *
+ * Replace disk->part_tbl with @new_ptbl in RCU-safe way.  The
+ * original ptbl is freed using RCU callback.
+ *
+ * LOCKING:
+ * Matching bd_mutx locked.
+ */
+static void disk_replace_part_tbl(struct gendisk *disk,
+				  struct disk_part_tbl *new_ptbl)
+{
+	struct disk_part_tbl *old_ptbl = disk->part_tbl;
+
+	rcu_assign_pointer(disk->part_tbl, new_ptbl);
+	if (old_ptbl)
+		call_rcu(&old_ptbl->rcu_head, disk_free_ptbl_rcu_cb);
+}
+
+/**
+ * disk_expand_part_tbl - expand disk->part_tbl
+ * @disk: disk to expand part_tbl for
+ * @partno: expand such that this partno can fit in
+ *
+ * Expand disk->part_tbl such that @partno can fit in.  disk->part_tbl
+ * uses RCU to allow unlocked dereferencing for stats and other stuff.
+ *
+ * LOCKING:
+ * Matching bd_mutex locked, might sleep.
+ *
+ * RETURNS:
+ * 0 on success, -errno on failure.
+ */
+int disk_expand_part_tbl(struct gendisk *disk, int partno)
+{
+	struct disk_part_tbl *old_ptbl = disk->part_tbl;
+	struct disk_part_tbl *new_ptbl;
+	int len = old_ptbl ? old_ptbl->len : 0;
+	int target = partno + 1;
+	size_t size;
+	int i;
+
+	/* disk_max_parts() is zero during initialization, ignore if so */
+	if (disk_max_parts(disk) && target > disk_max_parts(disk))
+		return -EINVAL;
+
+	if (target <= len)
+		return 0;
+
+	size = sizeof(*new_ptbl) + target * sizeof(new_ptbl->part[0]);
+	new_ptbl = kzalloc_node(size, GFP_KERNEL, disk->node_id);
+	if (!new_ptbl)
+		return -ENOMEM;
+
+	INIT_RCU_HEAD(&new_ptbl->rcu_head);
+	new_ptbl->len = target;
+
+	for (i = 0; i < len; i++)
+		rcu_assign_pointer(new_ptbl->part[i], old_ptbl->part[i]);
+
+	disk_replace_part_tbl(disk, new_ptbl);
+	return 0;
+}
+
 static void disk_release(struct device *dev)
 {
 	struct gendisk *disk = dev_to_disk(dev);
 
 	kfree(disk->random);
-	kfree(disk->__part);
+	disk_replace_part_tbl(disk, NULL);
 	free_part_stats(&disk->part0);
 	kfree(disk);
 }

@@ -949,22 +1043,16 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 	disk = kmalloc_node(sizeof(struct gendisk),
 				GFP_KERNEL | __GFP_ZERO, node_id);
 	if (disk) {
-		int tot_minors = minors + ext_minors;
-		int size = tot_minors * sizeof(struct hd_struct *);
-
 		if (!init_part_stats(&disk->part0)) {
 			kfree(disk);
 			return NULL;
 		}
-
-		disk->__part = kmalloc_node(size, GFP_KERNEL | __GFP_ZERO,
-					    node_id);
-		if (!disk->__part) {
-				free_part_stats(&disk->part0);
+		if (disk_expand_part_tbl(disk, 0)) {
+			free_part_stats(&disk->part0);
 			kfree(disk);
 			return NULL;
 		}
-		disk->__part[0] = &disk->part0;
+		disk->part_tbl->part[0] = &disk->part0;
 
 		disk->minors = minors;
 		disk->ext_minors = ext_minors;

@@ -974,6 +1062,7 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 		device_initialize(disk_to_dev(disk));
 		INIT_WORK(&disk->async_notify,
 			media_change_notify_thread);
+		disk->node_id = node_id;
 	}
 	return disk;
 }

index 67d2813..a933656 100644
@@ -30,7 +30,7 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user
 	if (bdev != bdev->bd_contains)
 		return -EINVAL;
 	partno = p.pno;
-	if (partno <= 0 || partno >= disk_max_parts(disk))
+	if (partno <= 0)
 		return -EINVAL;
 	switch (a.op) {
 		case BLKPG_ADD_PARTITION:

index e4d4d5c..825da63 100644
@@ -325,14 +325,18 @@ static void delete_partition_rcu_cb(struct rcu_head *head)
 
 void delete_partition(struct gendisk *disk, int partno)
 {
+	struct disk_part_tbl *ptbl = disk->part_tbl;
 	struct hd_struct *part;
 
-	part = disk->__part[partno];
+	if (partno >= ptbl->len)
+		return;
+
+	part = ptbl->part[partno];
 	if (!part)
 		return;
 
 	blk_free_devt(part_devt(part));
-	rcu_assign_pointer(disk->__part[partno], NULL);
+	rcu_assign_pointer(ptbl->part[partno], NULL);
 	kobject_put(part->holder_dir);
 	device_del(part_to_dev(part));
 

@@ -354,10 +358,16 @@ int add_partition(struct gendisk *disk, int partno,
 	dev_t devt = MKDEV(0, 0);
 	struct device *ddev = disk_to_dev(disk);
 	struct device *pdev;
+	struct disk_part_tbl *ptbl;
 	const char *dname;
 	int err;
 
-	if (disk->__part[partno]) {
+	err = disk_expand_part_tbl(disk, partno);
+	if (err)
+		return err;
+	ptbl = disk->part_tbl;
+
+	if (ptbl->part[partno]) {
 		err = -EBUSY;
 		goto fail;
 	}

@@ -392,7 +402,7 @@ int add_partition(struct gendisk *disk, int partno,
 	pdev->devt = devt;
 
 	INIT_RCU_HEAD(&p->rcu_head);
-	rcu_assign_pointer(disk->__part[partno], p);
+	rcu_assign_pointer(ptbl->part[partno], p);
 
 	/* delay uevent until 'holders' subdir is created */
 	pdev->uevent_suppress = 1;

@@ -489,7 +499,7 @@ int rescan_partitions(struct gendisk *disk, struct block_device *bdev)
 	struct disk_part_iter piter;
 	struct hd_struct *part;
 	struct parsed_partitions *state;
-	int p, res;
+	int p, highest, res;
 
 	if (bdev->bd_part_count)
 		return -EBUSY;

@@ -513,6 +523,17 @@ int rescan_partitions(struct gendisk *disk, struct block_device *bdev)
 	/* tell userspace that the media / partition table may have changed */
 	kobject_uevent(&disk_to_dev(disk)->kobj, KOBJ_CHANGE);
 
+	/* Detect the highest partition number and preallocate
+	 * disk->part_tbl.  This is an optimization and not strictly
+	 * necessary.
+	 */
+	for (p = 1, highest = 0; p < state->limit; p++)
+		if (state->parts[p].size)
+			highest = p;
+
+	disk_expand_part_tbl(disk, highest);
+
+	/* add partitions */
 	for (p = 1; p < state->limit; p++) {
 		sector_t size = state->parts[p].size;
 		sector_t from = state->parts[p].from;

index d5d2ef9..c7d9251 100644
@@ -113,6 +113,12 @@ struct hd_struct {
 #define GENHD_FL_UP				16
 #define GENHD_FL_SUPPRESS_PARTITION_INFO	32
 
+struct disk_part_tbl {
+	struct rcu_head rcu_head;
+	int len;
+	struct hd_struct *part[];
+};
+
 struct gendisk {
 	/* major, first_minor, minors and ext_minors are input
 	 * parameters only, don't use directly.  Use disk_devt() and

@@ -131,7 +137,7 @@ struct gendisk {
 	 * non-critical accesses use RCU.  Always access through
 	 * helpers.
 	 */
-	struct hd_struct **__part;
+	struct disk_part_tbl *part_tbl;
 	struct hd_struct part0;
 
 	struct block_device_operations *fops;

@@ -146,6 +152,8 @@ struct gendisk {
 
 	atomic_t sync_io;		/* RAID */
 	struct work_struct async_notify;
+
+	int node_id;
 };
 
 static inline struct gendisk *part_to_disk(struct hd_struct *part)

@@ -491,6 +499,7 @@ extern void blk_free_devt(dev_t devt);
 extern dev_t blk_lookup_devt(const char *name, int partno);
 extern char *disk_name (struct gendisk *hd, int partno, char *buf);
 
+extern int disk_expand_part_tbl(struct gendisk *disk, int target);
 extern int rescan_partitions(struct gendisk *disk, struct block_device *bdev);
 extern int add_partition(struct gendisk *, int, sector_t, sector_t, int);
 extern void delete_partition(struct gendisk *, int);

index 5d0f5d1..72aef36 100644
@@ -52,14 +52,21 @@ static struct device_type disk_type;
  */
 struct hd_struct *disk_get_part(struct gendisk *disk, int partno)
 {
-	struct hd_struct *part;
+	struct hd_struct *part = NULL;
+	struct disk_part_tbl *ptbl;
 
-	if (unlikely(partno < 0 || partno >= disk_max_parts(disk)))
+	if (unlikely(partno < 0))
 		return NULL;
+
 	rcu_read_lock();
-	part = rcu_dereference(disk->__part[partno]);
-	if (part)
-		get_device(part_to_dev(part));
+
+	ptbl = rcu_dereference(disk->part_tbl);
+	if (likely(partno < ptbl->len)) {
+		part = rcu_dereference(ptbl->part[partno]);
+		if (part)
+			get_device(part_to_dev(part));
+	}
+
 	rcu_read_unlock();
 
 	return part;

@@ -80,17 +87,24 @@ EXPORT_SYMBOL_GPL(disk_get_part);
 void disk_part_iter_start(struct disk_part_iter *piter, struct gendisk *disk,
 			  unsigned int flags)
 {
+	struct disk_part_tbl *ptbl;
+
+	rcu_read_lock();
+	ptbl = rcu_dereference(disk->part_tbl);
+
 	piter->disk = disk;
 	piter->part = NULL;
 
 	if (flags & DISK_PITER_REVERSE)
-		piter->idx = disk_max_parts(piter->disk) - 1;
+		piter->idx = ptbl->len - 1;
 	else if (flags & DISK_PITER_INCL_PART0)
 		piter->idx = 0;
 	else
 		piter->idx = 1;
 
 	piter->flags = flags;
+
+	rcu_read_unlock();
 }
 EXPORT_SYMBOL_GPL(disk_part_iter_start);
 

@@ -105,13 +119,16 @@ EXPORT_SYMBOL_GPL(disk_part_iter_start);
  */
 struct hd_struct *disk_part_iter_next(struct disk_part_iter *piter)
 {
+	struct disk_part_tbl *ptbl;
 	int inc, end;
 
 	/* put the last partition */
 	disk_put_part(piter->part);
 	piter->part = NULL;
 
+	/* get part_tbl */
 	rcu_read_lock();
+	ptbl = rcu_dereference(piter->disk->part_tbl);
 
 	/* determine iteration parameters */
 	if (piter->flags & DISK_PITER_REVERSE) {

@@ -122,14 +139,14 @@ struct hd_struct *disk_part_iter_next(struct disk_part_iter *piter)
 			end = 0;
 	} else {
 		inc = 1;
-		end = disk_max_parts(piter->disk);
+		end = ptbl->len;
 	}
 
 	/* iterate to the next partition */
 	for (; piter->idx != end; piter->idx += inc) {
 		struct hd_struct *part;
 
-		part = rcu_dereference(piter->disk->__part[piter->idx]);
+		part = rcu_dereference(ptbl->part[piter->idx]);
 		if (!part)
 			continue;
 		if (!(piter->flags & DISK_PITER_INCL_EMPTY) && !part->nr_sects)

@@ -180,10 +197,13 @@ EXPORT_SYMBOL_GPL(disk_part_iter_stop);
  */
 struct hd_struct *disk_map_sector_rcu(struct gendisk *disk, sector_t sector)
 {
+	struct disk_part_tbl *ptbl;
 	int i;
 
-	for (i = 1; i < disk_max_parts(disk); i++) {
-		struct hd_struct *part = rcu_dereference(disk->__part[i]);
+	ptbl = rcu_dereference(disk->part_tbl);
+
+	for (i = 1; i < ptbl->len; i++) {
+		struct hd_struct *part = rcu_dereference(ptbl->part[i]);
 
 		if (part && part->start_sect <= sector &&
 		    sector < part->start_sect + part->nr_sects)

@@ -763,12 +783,86 @@ static struct attribute_group *disk_attr_groups[] = {
 	NULL
 };
 
+static void disk_free_ptbl_rcu_cb(struct rcu_head *head)
+{
+	struct disk_part_tbl *ptbl =
+		container_of(head, struct disk_part_tbl, rcu_head);
+
+	kfree(ptbl);
+}
+
+/**
+ * disk_replace_part_tbl - replace disk->part_tbl in RCU-safe way
+ * @disk: disk to replace part_tbl for
+ * @new_ptbl: new part_tbl to install
+ *
+ * Replace disk->part_tbl with @new_ptbl in RCU-safe way.  The
+ * original ptbl is freed using RCU callback.
+ *
+ * LOCKING:
+ * Matching bd_mutx locked.
+ */
+static void disk_replace_part_tbl(struct gendisk *disk,
+				  struct disk_part_tbl *new_ptbl)
+{
+	struct disk_part_tbl *old_ptbl = disk->part_tbl;
+
+	rcu_assign_pointer(disk->part_tbl, new_ptbl);
+	if (old_ptbl)
+		call_rcu(&old_ptbl->rcu_head, disk_free_ptbl_rcu_cb);
+}
+
+/**
+ * disk_expand_part_tbl - expand disk->part_tbl
+ * @disk: disk to expand part_tbl for
+ * @partno: expand such that this partno can fit in
+ *
+ * Expand disk->part_tbl such that @partno can fit in.  disk->part_tbl
+ * uses RCU to allow unlocked dereferencing for stats and other stuff.
+ *
+ * LOCKING:
+ * Matching bd_mutex locked, might sleep.
+ *
+ * RETURNS:
+ * 0 on success, -errno on failure.
+ */
+int disk_expand_part_tbl(struct gendisk *disk, int partno)
+{
+	struct disk_part_tbl *old_ptbl = disk->part_tbl;
+	struct disk_part_tbl *new_ptbl;
+	int len = old_ptbl ? old_ptbl->len : 0;
+	int target = partno + 1;
+	size_t size;
+	int i;
+
+	/* disk_max_parts() is zero during initialization, ignore if so */
+	if (disk_max_parts(disk) && target > disk_max_parts(disk))
+		return -EINVAL;
+
+	if (target <= len)
+		return 0;
+
+	size = sizeof(*new_ptbl) + target * sizeof(new_ptbl->part[0]);
+	new_ptbl = kzalloc_node(size, GFP_KERNEL, disk->node_id);
+	if (!new_ptbl)
+		return -ENOMEM;
+
+	INIT_RCU_HEAD(&new_ptbl->rcu_head);
+	new_ptbl->len = target;
+
+	for (i = 0; i < len; i++)
+		rcu_assign_pointer(new_ptbl->part[i], old_ptbl->part[i]);
+
+	disk_replace_part_tbl(disk, new_ptbl);
+	return 0;
+}
+
 static void disk_release(struct device *dev)
 {
 	struct gendisk *disk = dev_to_disk(dev);
 
 	kfree(disk->random);
-	kfree(disk->__part);
+	disk_replace_part_tbl(disk, NULL);
 	free_part_stats(&disk->part0);
 	kfree(disk);
 }

@@ -949,22 +1043,16 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 	disk = kmalloc_node(sizeof(struct gendisk),
 				GFP_KERNEL | __GFP_ZERO, node_id);
 	if (disk) {
-		int tot_minors = minors + ext_minors;
-		int size = tot_minors * sizeof(struct hd_struct *);
-
 		if (!init_part_stats(&disk->part0)) {
 			kfree(disk);
 			return NULL;
 		}
-
-		disk->__part = kmalloc_node(size, GFP_KERNEL | __GFP_ZERO,
-					    node_id);
-		if (!disk->__part) {
-				free_part_stats(&disk->part0);
+		if (disk_expand_part_tbl(disk, 0)) {
+			free_part_stats(&disk->part0);
 			kfree(disk);
 			return NULL;
 		}
-		disk->__part[0] = &disk->part0;
+		disk->part_tbl->part[0] = &disk->part0;
 
 		disk->minors = minors;
 		disk->ext_minors = ext_minors;

@@ -974,6 +1062,7 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 		device_initialize(disk_to_dev(disk));
 		INIT_WORK(&disk->async_notify,
 			media_change_notify_thread);
+		disk->node_id = node_id;
 	}
 	return disk;
 }

index 67d2813..a933656 100644
@@ -30,7 +30,7 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user
 	if (bdev != bdev->bd_contains)
 		return -EINVAL;
 	partno = p.pno;
-	if (partno <= 0 || partno >= disk_max_parts(disk))
+	if (partno <= 0)
 		return -EINVAL;
 	switch (a.op) {
 		case BLKPG_ADD_PARTITION:

index e4d4d5c..825da63 100644
@@ -325,14 +325,18 @@ static void delete_partition_rcu_cb(struct rcu_head *head)
 
 void delete_partition(struct gendisk *disk, int partno)
 {
+	struct disk_part_tbl *ptbl = disk->part_tbl;
 	struct hd_struct *part;
 
-	part = disk->__part[partno];
+	if (partno >= ptbl->len)
+		return;
+
+	part = ptbl->part[partno];
 	if (!part)
 		return;
 
 	blk_free_devt(part_devt(part));
-	rcu_assign_pointer(disk->__part[partno], NULL);
+	rcu_assign_pointer(ptbl->part[partno], NULL);
 	kobject_put(part->holder_dir);
 	device_del(part_to_dev(part));
 

@@ -354,10 +358,16 @@ int add_partition(struct gendisk *disk, int partno,
 	dev_t devt = MKDEV(0, 0);
 	struct device *ddev = disk_to_dev(disk);
 	struct device *pdev;
+	struct disk_part_tbl *ptbl;
 	const char *dname;
 	int err;
 
-	if (disk->__part[partno]) {
+	err = disk_expand_part_tbl(disk, partno);
+	if (err)
+		return err;
+	ptbl = disk->part_tbl;
+
+	if (ptbl->part[partno]) {
 		err = -EBUSY;
 		goto fail;
 	}

@@ -392,7 +402,7 @@ int add_partition(struct gendisk *disk, int partno,
 	pdev->devt = devt;
 
 	INIT_RCU_HEAD(&p->rcu_head);
-	rcu_assign_pointer(disk->__part[partno], p);
+	rcu_assign_pointer(ptbl->part[partno], p);
 
 	/* delay uevent until 'holders' subdir is created */
 	pdev->uevent_suppress = 1;

@@ -489,7 +499,7 @@ int rescan_partitions(struct gendisk *disk, struct block_device *bdev)
 	struct disk_part_iter piter;
 	struct hd_struct *part;
 	struct parsed_partitions *state;
-	int p, res;
+	int p, highest, res;
 
 	if (bdev->bd_part_count)
 		return -EBUSY;

@@ -513,6 +523,17 @@ int rescan_partitions(struct gendisk *disk, struct block_device *bdev)
 	/* tell userspace that the media / partition table may have changed */
 	kobject_uevent(&disk_to_dev(disk)->kobj, KOBJ_CHANGE);
 
+	/* Detect the highest partition number and preallocate
+	 * disk->part_tbl.  This is an optimization and not strictly
+	 * necessary.
+	 */
+	for (p = 1, highest = 0; p < state->limit; p++)
+		if (state->parts[p].size)
+			highest = p;
+
+	disk_expand_part_tbl(disk, highest);
+
+	/* add partitions */
 	for (p = 1; p < state->limit; p++) {
 		sector_t size = state->parts[p].size;
 		sector_t from = state->parts[p].from;

index d5d2ef9..c7d9251 100644
@@ -113,6 +113,12 @@ struct hd_struct {
 #define GENHD_FL_UP				16
 #define GENHD_FL_SUPPRESS_PARTITION_INFO	32
 
+struct disk_part_tbl {
+	struct rcu_head rcu_head;
+	int len;
+	struct hd_struct *part[];
+};
+
 struct gendisk {
 	/* major, first_minor, minors and ext_minors are input
 	 * parameters only, don't use directly.  Use disk_devt() and

@@ -131,7 +137,7 @@ struct gendisk {
 	 * non-critical accesses use RCU.  Always access through
 	 * helpers.
 	 */
-	struct hd_struct **__part;
+	struct disk_part_tbl *part_tbl;
 	struct hd_struct part0;
 
 	struct block_device_operations *fops;

@@ -146,6 +152,8 @@ struct gendisk {
 
 	atomic_t sync_io;		/* RAID */
 	struct work_struct async_notify;
+
+	int node_id;
 };
 
 static inline struct gendisk *part_to_disk(struct hd_struct *part)

@@ -491,6 +499,7 @@ extern void blk_free_devt(dev_t devt);
 extern dev_t blk_lookup_devt(const char *name, int partno);
 extern char *disk_name (struct gendisk *hd, int partno, char *buf);
 
+extern int disk_expand_part_tbl(struct gendisk *disk, int target);
 extern int rescan_partitions(struct gendisk *disk, struct block_device *bdev);
 extern int add_partition(struct gendisk *, int, sector_t, sector_t, int);
 extern void delete_partition(struct gendisk *, int);

index 72aef36..b38c235 100644
@@ -1025,19 +1025,10 @@ struct gendisk *alloc_disk(int minors)
 {
 	return alloc_disk_node(minors, -1);
 }
+EXPORT_SYMBOL(alloc_disk);
 
 struct gendisk *alloc_disk_node(int minors, int node_id)
 {
-	return alloc_disk_ext_node(minors, 0, node_id);
-}
-
-struct gendisk *alloc_disk_ext(int minors, int ext_minors)
-{
-	return alloc_disk_ext_node(minors, ext_minors, -1);
-}
-
-struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
-{
 	struct gendisk *disk;
 
 	disk = kmalloc_node(sizeof(struct gendisk),

@@ -1055,7 +1046,6 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 		disk->part_tbl->part[0] = &disk->part0;
 
 		disk->minors = minors;
-		disk->ext_minors = ext_minors;
 		rand_initialize_disk(disk);
 		disk_to_dev(disk)->class = &block_class;
 		disk_to_dev(disk)->type = &disk_type;

@@ -1066,11 +1056,7 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 	}
 	return disk;
 }
-
-EXPORT_SYMBOL(alloc_disk);
 EXPORT_SYMBOL(alloc_disk_node);
-EXPORT_SYMBOL(alloc_disk_ext);
-EXPORT_SYMBOL(alloc_disk_ext_node);
 
 struct kobject *get_disk(struct gendisk *disk)
 {

index 3c6f0fa..6d7cf15 100644
@@ -41,16 +41,12 @@
 #include <asm/io.h>
 #include <asm/div64.h>
 
-#define IDE_DISK_PARTS		(1 << PARTN_BITS)
-
 #if !defined(CONFIG_DEBUG_BLOCK_EXT_DEVT)
-#define IDE_DISK_MINORS		IDE_DISK_PARTS
+#define IDE_DISK_MINORS		(1 << PARTN_BITS)
 #else
 #define IDE_DISK_MINORS		1
 #endif
 
-#define IDE_DISK_EXT_MINORS	(IDE_DISK_PARTS - IDE_DISK_MINORS)
-
 struct ide_disk_obj {
 	ide_drive_t	*drive;
 	ide_driver_t	*driver;

@@ -1168,8 +1164,7 @@ static int ide_disk_probe(ide_drive_t *drive)
 	if (!idkp)
 		goto failed;
 
-	g = alloc_disk_ext_node(IDE_DISK_MINORS, IDE_DISK_EXT_MINORS,
-				hwif_to_node(drive->hwif));
+	g = alloc_disk_node(IDE_DISK_MINORS, hwif_to_node(drive->hwif));
 	if (!g)
 		goto out_free_idkp;
 

@@ -1196,9 +1191,10 @@ static int ide_disk_probe(ide_drive_t *drive)
 		drive->attach = 1;
 
 	g->minors = IDE_DISK_MINORS;
-	g->ext_minors = IDE_DISK_EXT_MINORS;
 	g->driverfs_dev = &drive->gendev;
-	g->flags = drive->removable ? GENHD_FL_REMOVABLE : 0;
+	g->flags |= GENHD_FL_EXT_DEVT;
+	if (drive->removable)
+		g->flags |= GENHD_FL_REMOVABLE;
 	set_capacity(g, idedisk_capacity(drive));
 	g->fops = &idedisk_ops;
 	add_disk(g);

index f355a39..1f41e3d 100644
@@ -86,16 +86,12 @@ MODULE_ALIAS_SCSI_DEVICE(TYPE_DISK);
 MODULE_ALIAS_SCSI_DEVICE(TYPE_MOD);
 MODULE_ALIAS_SCSI_DEVICE(TYPE_RBC);
 
-#define SD_PARTS	64
-
 #if !defined(CONFIG_DEBUG_BLOCK_EXT_DEVT)
 #define SD_MINORS	16
 #else
 #define SD_MINORS	1
 #endif
 
-#define SD_EXT_MINORS	(SD_PARTS - SD_MINORS)
-
 static int  sd_revalidate_disk(struct gendisk *);
 static int  sd_probe(struct device *);
 static int  sd_remove(struct device *);

@@ -1651,7 +1647,7 @@ static int sd_probe(struct device *dev)
 	if (!sdkp)
 		goto out;
 
-	gd = alloc_disk_ext(SD_MINORS, SD_EXT_MINORS);
+	gd = alloc_disk(SD_MINORS);
 	if (!gd)
 		goto out_free;
 

@@ -1696,7 +1692,6 @@ static int sd_probe(struct device *dev)
 	gd->major = sd_major((index & 0xf0) >> 4);
 	gd->first_minor = ((index & 0xf) << 4) | (index & 0xfff00);
 	gd->minors = SD_MINORS;
-	gd->ext_minors = SD_EXT_MINORS;
 	gd->fops = &sd_fops;
 
 	if (index < 26) {

@@ -1720,7 +1715,7 @@ static int sd_probe(struct device *dev)
 	blk_queue_prep_rq(sdp->request_queue, sd_prep_fn);
 
 	gd->driverfs_dev = &sdp->sdev_gendev;
-	gd->flags = GENHD_FL_DRIVERFS;
+	gd->flags = GENHD_FL_EXT_DEVT | GENHD_FL_DRIVERFS;
 	if (sdp->removable)
 		gd->flags |= GENHD_FL_REMOVABLE;
 

index 17ae8ec..98dbe1a 100644
@@ -5,15 +5,13 @@
  * add_gd_partition adds a partitions details to the devices partition
  * description.
  */
-enum { MAX_PART = 256 };
-
 struct parsed_partitions {
 	char name[BDEVNAME_SIZE];
 	struct {
 		sector_t from;
 		sector_t size;
 		int flags;
-	} parts[MAX_PART];
+	} parts[DISK_MAX_PARTS];
 	int next;
 	int limit;
 };

index c7d9251..a48f350 100644
@@ -58,6 +58,8 @@ enum {
 	UNIXWARE_PARTITION = 0x63,	/* Same as GNU_HURD and SCO Unix */
 };
 
+#define DISK_MAX_PARTS			256
+
 #include <linux/major.h>
 #include <linux/device.h>
 #include <linux/smp.h>

@@ -112,6 +114,7 @@ struct hd_struct {
 #define GENHD_FL_CD				8
 #define GENHD_FL_UP				16
 #define GENHD_FL_SUPPRESS_PARTITION_INFO	32
+#define GENHD_FL_EXT_DEVT			64 /* allow extended devt */
 
 struct disk_part_tbl {
 	struct rcu_head rcu_head;

@@ -120,15 +123,13 @@ struct disk_part_tbl {
 };
 
 struct gendisk {
-	/* major, first_minor, minors and ext_minors are input
-	 * parameters only, don't use directly.  Use disk_devt() and
-	 * disk_max_parts().
+	/* major, first_minor and minors are input parameters only,
+	 * don't use directly.  Use disk_devt() and disk_max_parts().
 	 */
 	int major;			/* major number of driver */
 	int first_minor;
 	int minors;                     /* maximum number of minors, =1 for
                                          * disks that can't be partitioned. */
-	int ext_minors;			/* number of extended dynamic minors */
 
 	char disk_name[32];		/* name of major driver */
 

@@ -169,7 +170,9 @@ static inline struct gendisk *part_to_disk(struct hd_struct *part)
 
 static inline int disk_max_parts(struct gendisk *disk)
 {
-	return disk->minors + disk->ext_minors;
+	if (disk->flags & GENHD_FL_EXT_DEVT)
+		return DISK_MAX_PARTS;
+	return disk->minors;
 }
 
 static inline bool disk_partitionable(struct gendisk *disk)

@@ -507,9 +510,6 @@ extern void printk_all_partitions(void);
 
 extern struct gendisk *alloc_disk_node(int minors, int node_id);
 extern struct gendisk *alloc_disk(int minors);
-extern struct gendisk *alloc_disk_ext_node(int minors, int ext_minrs,
-					   int node_id);
-extern struct gendisk *alloc_disk_ext(int minors, int ext_minors);
 extern struct kobject *get_disk(struct gendisk *disk);
 extern void put_disk(struct gendisk *disk);
 extern void blk_register_region(dev_t devt, unsigned long range,

index 72aef36..b38c235 100644
@@ -1025,19 +1025,10 @@ struct gendisk *alloc_disk(int minors)
 {
 	return alloc_disk_node(minors, -1);
 }
+EXPORT_SYMBOL(alloc_disk);
 
 struct gendisk *alloc_disk_node(int minors, int node_id)
 {
-	return alloc_disk_ext_node(minors, 0, node_id);
-}
-
-struct gendisk *alloc_disk_ext(int minors, int ext_minors)
-{
-	return alloc_disk_ext_node(minors, ext_minors, -1);
-}
-
-struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
-{
 	struct gendisk *disk;
 
 	disk = kmalloc_node(sizeof(struct gendisk),

@@ -1055,7 +1046,6 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 		disk->part_tbl->part[0] = &disk->part0;
 
 		disk->minors = minors;
-		disk->ext_minors = ext_minors;
 		rand_initialize_disk(disk);
 		disk_to_dev(disk)->class = &block_class;
 		disk_to_dev(disk)->type = &disk_type;

@@ -1066,11 +1056,7 @@ struct gendisk *alloc_disk_ext_node(int minors, int ext_minors, int node_id)
 	}
 	return disk;
 }
-
-EXPORT_SYMBOL(alloc_disk);
 EXPORT_SYMBOL(alloc_disk_node);
-EXPORT_SYMBOL(alloc_disk_ext);
-EXPORT_SYMBOL(alloc_disk_ext_node);
 
 struct kobject *get_disk(struct gendisk *disk)
 {

index 3c6f0fa..6d7cf15 100644
@@ -41,16 +41,12 @@
 #include <asm/io.h>
 #include <asm/div64.h>
 
-#define IDE_DISK_PARTS		(1 << PARTN_BITS)
-
 #if !defined(CONFIG_DEBUG_BLOCK_EXT_DEVT)
-#define IDE_DISK_MINORS		IDE_DISK_PARTS
+#define IDE_DISK_MINORS		(1 << PARTN_BITS)
 #else
 #define IDE_DISK_MINORS		1
 #endif
 
-#define IDE_DISK_EXT_MINORS	(IDE_DISK_PARTS - IDE_DISK_MINORS)
-
 struct ide_disk_obj {
 	ide_drive_t	*drive;
 	ide_driver_t	*driver;

@@ -1168,8 +1164,7 @@ static int ide_disk_probe(ide_drive_t *drive)
 	if (!idkp)
 		goto failed;
 
-	g = alloc_disk_ext_node(IDE_DISK_MINORS, IDE_DISK_EXT_MINORS,
-				hwif_to_node(drive->hwif));
+	g = alloc_disk_node(IDE_DISK_MINORS, hwif_to_node(drive->hwif));
 	if (!g)
 		goto out_free_idkp;
 

@@ -1196,9 +1191,10 @@ static int ide_disk_probe(ide_drive_t *drive)
 		drive->attach = 1;
 
 	g->minors = IDE_DISK_MINORS;
-	g->ext_minors = IDE_DISK_EXT_MINORS;
 	g->driverfs_dev = &drive->gendev;
-	g->flags = drive->removable ? GENHD_FL_REMOVABLE : 0;
+	g->flags |= GENHD_FL_EXT_DEVT;
+	if (drive->removable)
+		g->flags |= GENHD_FL_REMOVABLE;
 	set_capacity(g, idedisk_capacity(drive));
 	g->fops = &idedisk_ops;
 	add_disk(g);

index f355a39..1f41e3d 100644
@@ -86,16 +86,12 @@ MODULE_ALIAS_SCSI_DEVICE(TYPE_DISK);
 MODULE_ALIAS_SCSI_DEVICE(TYPE_MOD);
 MODULE_ALIAS_SCSI_DEVICE(TYPE_RBC);
 
-#define SD_PARTS	64
-
 #if !defined(CONFIG_DEBUG_BLOCK_EXT_DEVT)
 #define SD_MINORS	16
 #else
 #define SD_MINORS	1
 #endif
 
-#define SD_EXT_MINORS	(SD_PARTS - SD_MINORS)
-
 static int  sd_revalidate_disk(struct gendisk *);
 static int  sd_probe(struct device *);
 static int  sd_remove(struct device *);

@@ -1651,7 +1647,7 @@ static int sd_probe(struct device *dev)
 	if (!sdkp)
 		goto out;
 
-	gd = alloc_disk_ext(SD_MINORS, SD_EXT_MINORS);
+	gd = alloc_disk(SD_MINORS);
 	if (!gd)
 		goto out_free;
 

@@ -1696,7 +1692,6 @@ static int sd_probe(struct device *dev)
 	gd->major = sd_major((index & 0xf0) >> 4);
 	gd->first_minor = ((index & 0xf) << 4) | (index & 0xfff00);
 	gd->minors = SD_MINORS;
-	gd->ext_minors = SD_EXT_MINORS;
 	gd->fops = &sd_fops;
 
 	if (index < 26) {

@@ -1720,7 +1715,7 @@ static int sd_probe(struct device *dev)
 	blk_queue_prep_rq(sdp->request_queue, sd_prep_fn);
 
 	gd->driverfs_dev = &sdp->sdev_gendev;
-	gd->flags = GENHD_FL_DRIVERFS;
+	gd->flags = GENHD_FL_EXT_DEVT | GENHD_FL_DRIVERFS;
 	if (sdp->removable)
 		gd->flags |= GENHD_FL_REMOVABLE;
 

index 17ae8ec..98dbe1a 100644
@@ -5,15 +5,13 @@
  * add_gd_partition adds a partitions details to the devices partition
  * description.
  */
-enum { MAX_PART = 256 };
-
 struct parsed_partitions {
 	char name[BDEVNAME_SIZE];
 	struct {
 		sector_t from;
 		sector_t size;
 		int flags;
-	} parts[MAX_PART];
+	} parts[DISK_MAX_PARTS];
 	int next;
 	int limit;
 };

index c7d9251..a48f350 100644
@@ -58,6 +58,8 @@ enum {
 	UNIXWARE_PARTITION = 0x63,	/* Same as GNU_HURD and SCO Unix */
 };
 
+#define DISK_MAX_PARTS			256
+
 #include <linux/major.h>
 #include <linux/device.h>
 #include <linux/smp.h>

@@ -112,6 +114,7 @@ struct hd_struct {
 #define GENHD_FL_CD				8
 #define GENHD_FL_UP				16
 #define GENHD_FL_SUPPRESS_PARTITION_INFO	32
+#define GENHD_FL_EXT_DEVT			64 /* allow extended devt */
 
 struct disk_part_tbl {
 	struct rcu_head rcu_head;

@@ -120,15 +123,13 @@ struct disk_part_tbl {
 };
 
 struct gendisk {
-	/* major, first_minor, minors and ext_minors are input
-	 * parameters only, don't use directly.  Use disk_devt() and
-	 * disk_max_parts().
+	/* major, first_minor and minors are input parameters only,
+	 * don't use directly.  Use disk_devt() and disk_max_parts().
 	 */
 	int major;			/* major number of driver */
 	int first_minor;
 	int minors;                     /* maximum number of minors, =1 for
                                          * disks that can't be partitioned. */
-	int ext_minors;			/* number of extended dynamic minors */
 
 	char disk_name[32];		/* name of major driver */
 

@@ -169,7 +170,9 @@ static inline struct gendisk *part_to_disk(struct hd_struct *part)
 
 static inline int disk_max_parts(struct gendisk *disk)
 {
-	return disk->minors + disk->ext_minors;
+	if (disk->flags & GENHD_FL_EXT_DEVT)
+		return DISK_MAX_PARTS;
+	return disk->minors;
 }
 
 static inline bool disk_partitionable(struct gendisk *disk)

@@ -507,9 +510,6 @@ extern void printk_all_partitions(void);
 
 extern struct gendisk *alloc_disk_node(int minors, int node_id);
 extern struct gendisk *alloc_disk(int minors);
-extern struct gendisk *alloc_disk_ext_node(int minors, int ext_minrs,
-					   int node_id);
-extern struct gendisk *alloc_disk_ext(int minors, int ext_minors);
 extern struct kobject *get_disk(struct gendisk *disk);
 extern void put_disk(struct gendisk *disk);
 extern void blk_register_region(dev_t devt, unsigned long range,

index b38c235..a4f3117 100644
@@ -477,13 +477,40 @@ static int exact_lock(dev_t devt, void *data)
  *
  * This function registers the partitioning information in @disk
  * with the kernel.
+ *
+ * FIXME: error handling
  */
 void add_disk(struct gendisk *disk)
 {
 	struct backing_dev_info *bdi;
+	dev_t devt;
+	int rc;
+
+	/* minors == 0 indicates to use ext devt from part0 and should
+	 * be accompanied with EXT_DEVT flag.  Make sure all
+	 * parameters make sense.
+	 */
+	WARN_ON(disk->minors && !(disk->major || disk->first_minor));
+	WARN_ON(!disk->minors && !(disk->flags & GENHD_FL_EXT_DEVT));
 
 	disk->flags |= GENHD_FL_UP;
-	disk_to_dev(disk)->devt = MKDEV(disk->major, disk->first_minor);
+
+	if (disk->minors)
+		devt = MKDEV(disk->major, disk->first_minor);
+	else {
+		rc = blk_alloc_devt(&disk->part0, &devt);
+		if (rc)
+			return;
+
+		/* ->major and ->first_minor aren't supposed to be
+		 * referenced from here on, but set them just in case.
+		 */
+		disk->major = MAJOR(devt);
+		disk->first_minor = MINOR(devt);
+	}
+
+	disk_to_dev(disk)->devt = devt;
+
 	blk_register_region(disk_devt(disk), disk->minors, NULL,
 			    exact_match, exact_lock, disk);
 	register_disk(disk);