block: switch gendisk lookup to a simple xarray

Now that bdev_map is only used for finding gendisks, we can use
a simple xarray instead of the regions tracking structure for it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This commit is contained in:
Christoph Hellwig 2020-10-29 15:58:41 +01:00 committed by Jens Axboe
parent 76487f0241
commit e418de3abc
2 changed files with 36 additions and 177 deletions

View File

@ -27,15 +27,7 @@
static struct kobject *block_depr;
struct bdev_map {
struct bdev_map *next;
dev_t dev;
unsigned long range;
struct module *owner;
struct kobject *(*probe)(dev_t, int *, void *);
int (*lock)(dev_t, void *);
void *data;
} *bdev_map[255];
static DEFINE_XARRAY(bdev_map);
static DEFINE_MUTEX(bdev_map_lock);
/* for extended dynamic devt allocation, currently only one major is used */
@ -649,85 +641,26 @@ static char *bdevt_str(dev_t devt, char *buf)
return buf;
}
/*
* Register device numbers dev..(dev+range-1)
* range must be nonzero
* The hash chain is sorted on range, so that subranges can override.
*/
void blk_register_region(dev_t devt, unsigned long range, struct module *module,
struct kobject *(*probe)(dev_t, int *, void *),
int (*lock)(dev_t, void *), void *data)
static void blk_register_region(struct gendisk *disk)
{
unsigned n = MAJOR(devt + range - 1) - MAJOR(devt) + 1;
unsigned index = MAJOR(devt);
unsigned i;
struct bdev_map *p;
n = min(n, 255u);
p = kmalloc_array(n, sizeof(struct bdev_map), GFP_KERNEL);
if (p == NULL)
return;
for (i = 0; i < n; i++, p++) {
p->owner = module;
p->probe = probe;
p->lock = lock;
p->dev = devt;
p->range = range;
p->data = data;
}
int i;
mutex_lock(&bdev_map_lock);
for (i = 0, p -= n; i < n; i++, p++, index++) {
struct bdev_map **s = &bdev_map[index % 255];
while (*s && (*s)->range < range)
s = &(*s)->next;
p->next = *s;
*s = p;
for (i = 0; i < disk->minors; i++) {
if (xa_insert(&bdev_map, disk_devt(disk) + i, disk, GFP_KERNEL))
WARN_ON_ONCE(1);
}
mutex_unlock(&bdev_map_lock);
}
EXPORT_SYMBOL(blk_register_region);
void blk_unregister_region(dev_t devt, unsigned long range)
static void blk_unregister_region(struct gendisk *disk)
{
unsigned n = MAJOR(devt + range - 1) - MAJOR(devt) + 1;
unsigned index = MAJOR(devt);
unsigned i;
struct bdev_map *found = NULL;
int i;
mutex_lock(&bdev_map_lock);
for (i = 0; i < min(n, 255u); i++, index++) {
struct bdev_map **s;
for (s = &bdev_map[index % 255]; *s; s = &(*s)->next) {
struct bdev_map *p = *s;
if (p->dev == devt && p->range == range) {
*s = p->next;
if (!found)
found = p;
break;
}
}
}
for (i = 0; i < disk->minors; i++)
xa_erase(&bdev_map, disk_devt(disk) + i);
mutex_unlock(&bdev_map_lock);
kfree(found);
}
EXPORT_SYMBOL(blk_unregister_region);
static struct kobject *exact_match(dev_t devt, int *partno, void *data)
{
struct gendisk *p = data;
return &disk_to_dev(p)->kobj;
}
static int exact_lock(dev_t devt, void *data)
{
struct gendisk *p = data;
if (!get_disk_and_module(p))
return -1;
return 0;
}
static void disk_scan_partitions(struct gendisk *disk)
@ -873,8 +806,7 @@ static void __device_add_disk(struct device *parent, struct gendisk *disk,
ret = bdi_register(bdi, "%u:%u", MAJOR(devt), MINOR(devt));
WARN_ON(ret);
bdi_set_owner(bdi, dev);
blk_register_region(disk_devt(disk), disk->minors, NULL,
exact_match, exact_lock, disk);
blk_register_region(disk);
}
register_disk(parent, disk, groups);
if (register_queue)
@ -987,7 +919,7 @@ void del_gendisk(struct gendisk *disk)
blk_unregister_queue(disk);
if (!(disk->flags & GENHD_FL_HIDDEN))
blk_unregister_region(disk_devt(disk), disk->minors);
blk_unregister_region(disk);
/*
* Remove gendisk pointer from idr so that it cannot be looked up
* while RCU period before freeing gendisk is running to prevent
@ -1053,54 +985,22 @@ static void request_gendisk_module(dev_t devt)
request_module("block-major-%d", MAJOR(devt));
}
static struct gendisk *lookup_gendisk(dev_t dev, int *partno)
static bool get_disk_and_module(struct gendisk *disk)
{
struct kobject *kobj;
struct bdev_map *p;
unsigned long best = ~0UL;
struct module *owner;
retry:
mutex_lock(&bdev_map_lock);
for (p = bdev_map[MAJOR(dev) % 255]; p; p = p->next) {
struct kobject *(*probe)(dev_t, int *, void *);
struct module *owner;
void *data;
if (p->dev > dev || p->dev + p->range - 1 < dev)
continue;
if (p->range - 1 >= best)
break;
if (!try_module_get(p->owner))
continue;
owner = p->owner;
data = p->data;
probe = p->probe;
best = p->range - 1;
*partno = dev - p->dev;
if (!probe) {
mutex_unlock(&bdev_map_lock);
module_put(owner);
request_gendisk_module(dev);
goto retry;
}
if (p->lock && p->lock(dev, data) < 0) {
module_put(owner);
continue;
}
mutex_unlock(&bdev_map_lock);
kobj = probe(dev, partno, data);
/* Currently ->owner protects _only_ ->probe() itself. */
if (!disk->fops)
return false;
owner = disk->fops->owner;
if (owner && !try_module_get(owner))
return false;
if (!kobject_get_unless_zero(&disk_to_dev(disk)->kobj)) {
module_put(owner);
if (kobj)
return dev_to_disk(kobj_to_dev(kobj));
goto retry;
return false;
}
mutex_unlock(&bdev_map_lock);
return NULL;
}
return true;
}
/**
* get_gendisk - get partitioning information for a given device
@ -1119,7 +1019,19 @@ struct gendisk *get_gendisk(dev_t devt, int *partno)
might_sleep();
if (MAJOR(devt) != BLOCK_EXT_MAJOR) {
disk = lookup_gendisk(devt, partno);
mutex_lock(&bdev_map_lock);
disk = xa_load(&bdev_map, devt);
if (!disk) {
mutex_unlock(&bdev_map_lock);
request_gendisk_module(devt);
mutex_lock(&bdev_map_lock);
disk = xa_load(&bdev_map, devt);
}
if (disk && !get_disk_and_module(disk))
disk = NULL;
if (disk)
*partno = devt - disk_devt(disk);
mutex_unlock(&bdev_map_lock);
} else {
struct hd_struct *part;
@ -1323,21 +1235,6 @@ static const struct seq_operations partitions_op = {
};
#endif
static void bdev_map_init(void)
{
struct bdev_map *base;
int i;
base = kzalloc(sizeof(*base), GFP_KERNEL);
if (!base)
panic("cannot allocate bdev_map");
base->dev = 1;
base->range = ~0 ;
for (i = 0; i < 255; i++)
bdev_map[i] = base;
}
static int __init genhd_device_init(void)
{
int error;
@ -1346,7 +1243,6 @@ static int __init genhd_device_init(void)
error = class_register(&block_class);
if (unlikely(error))
return error;
bdev_map_init();
blk_dev_init();
register_blkdev(BLOCK_EXT_MAJOR, "blkext");
@ -1895,35 +1791,6 @@ out_free_disk:
}
EXPORT_SYMBOL(__alloc_disk_node);
/**
* get_disk_and_module - increments the gendisk and gendisk fops module refcount
* @disk: the struct gendisk to increment the refcount for
*
* This increments the refcount for the struct gendisk, and the gendisk's
* fops module owner.
*
* Context: Any context.
*/
struct kobject *get_disk_and_module(struct gendisk *disk)
{
struct module *owner;
struct kobject *kobj;
if (!disk->fops)
return NULL;
owner = disk->fops->owner;
if (owner && !try_module_get(owner))
return NULL;
kobj = kobject_get_unless_zero(&disk_to_dev(disk)->kobj);
if (kobj == NULL) {
module_put(owner);
return NULL;
}
return kobj;
}
EXPORT_SYMBOL(get_disk_and_module);
/**
* put_disk - decrements the gendisk refcount
* @disk: the struct gendisk to decrement the refcount for
@ -1960,7 +1827,6 @@ void put_disk_and_module(struct gendisk *disk)
module_put(owner);
}
}
EXPORT_SYMBOL(put_disk_and_module);
static void set_disk_ro_uevent(struct gendisk *gd, int ro)
{

View File

@ -339,15 +339,8 @@ int blk_add_partitions(struct gendisk *disk, struct block_device *bdev);
int blk_drop_partitions(struct block_device *bdev);
extern struct gendisk *__alloc_disk_node(int minors, int node_id);
extern struct kobject *get_disk_and_module(struct gendisk *disk);
extern void put_disk(struct gendisk *disk);
extern void put_disk_and_module(struct gendisk *disk);
extern void blk_register_region(dev_t devt, unsigned long range,
struct module *module,
struct kobject *(*probe)(dev_t, int *, void *),
int (*lock)(dev_t, void *),
void *data);
extern void blk_unregister_region(dev_t devt, unsigned long range);
#define alloc_disk_node(minors, node_id) \
({ \