Skip to content

Commit

Permalink
loop: scale loop device by introducing per device lock
Browse files Browse the repository at this point in the history 
Currently, loop device has only one global lock: loop_ctl_mutex.

This becomes hot in scenarios where many loop devices are used.

Scale it by introducing per-device lock: lo_mutex that protects
modifications of all fields in struct loop_device.

Keep loop_ctl_mutex to protect global data: loop_index_idr, loop_lookup,
loop_add.

The new lock ordering requirement is that loop_ctl_mutex must be taken
before lo_mutex.

Signed-off-by: Pavel Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Tyler Hicks <tyhicks@linux.microsoft.com>
Reviewed-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
  • Loading branch information
Pavel Tatashin authored and Jens Axboe committed Jan 26, 2021
1 parent 767630c commit 6cc8e74
Show file tree
Hide file tree
Showing 2 changed files with 54 additions and 40 deletions.
93 changes: 53 additions & 40 deletions drivers/block/loop.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -704,7 +704,7 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev,
int error;
bool partscan;

error = mutex_lock_killable(&loop_ctl_mutex);
error = mutex_lock_killable(&lo->lo_mutex);
if (error)
return error;
error = -ENXIO;
Expand Down Expand Up @@ -743,9 +743,9 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev,
loop_update_dio(lo);
blk_mq_unfreeze_queue(lo->lo_queue);
partscan = lo->lo_flags & LO_FLAGS_PARTSCAN;
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
/*
* We must drop file reference outside of loop_ctl_mutex as dropping
* We must drop file reference outside of lo_mutex as dropping
* the file ref can take bd_mutex which creates circular locking
* dependency.
*/
Expand All @@ -755,7 +755,7 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev,
return 0;

out_err:
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
if (file)
fput(file);
return error;
Expand Down Expand Up @@ -1092,7 +1092,7 @@ static int loop_configure(struct loop_device *lo, fmode_t mode,
goto out_putf;
}

error = mutex_lock_killable(&loop_ctl_mutex);
error = mutex_lock_killable(&lo->lo_mutex);
if (error)
goto out_bdev;

Expand Down Expand Up @@ -1171,15 +1171,15 @@ static int loop_configure(struct loop_device *lo, fmode_t mode,
* put /dev/loopXX inode. Later in __loop_clr_fd() we bdput(bdev).
*/
bdgrab(bdev);
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
if (partscan)
loop_reread_partitions(lo, bdev);
if (!(mode & FMODE_EXCL))
bd_abort_claiming(bdev, loop_configure);
return 0;

out_unlock:
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
out_bdev:
if (!(mode & FMODE_EXCL))
bd_abort_claiming(bdev, loop_configure);
Expand All @@ -1200,7 +1200,7 @@ static int __loop_clr_fd(struct loop_device *lo, bool release)
bool partscan = false;
int lo_number;

mutex_lock(&loop_ctl_mutex);
mutex_lock(&lo->lo_mutex);
if (WARN_ON_ONCE(lo->lo_state != Lo_rundown)) {
err = -ENXIO;
goto out_unlock;
Expand Down Expand Up @@ -1253,7 +1253,7 @@ static int __loop_clr_fd(struct loop_device *lo, bool release)
lo_number = lo->lo_number;
loop_unprepare_queue(lo);
out_unlock:
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
if (partscan) {
/*
* bd_mutex has been held already in release path, so don't
Expand Down Expand Up @@ -1284,18 +1284,17 @@ static int __loop_clr_fd(struct loop_device *lo, bool release)
* protects us from all the other places trying to change the 'lo'
* device.
*/
mutex_lock(&loop_ctl_mutex);
mutex_lock(&lo->lo_mutex);
lo->lo_flags = 0;
if (!part_shift)
lo->lo_disk->flags |= GENHD_FL_NO_PART_SCAN;
lo->lo_state = Lo_unbound;
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);

/*
* Need not hold loop_ctl_mutex to fput backing file.
* Calling fput holding loop_ctl_mutex triggers a circular
* lock dependency possibility warning as fput can take
* bd_mutex which is usually taken before loop_ctl_mutex.
* Need not hold lo_mutex to fput backing file. Calling fput holding
* lo_mutex triggers a circular lock dependency possibility warning as
* fput can take bd_mutex which is usually taken before lo_mutex.
*/
if (filp)
fput(filp);
Expand All @@ -1306,11 +1305,11 @@ static int loop_clr_fd(struct loop_device *lo)
{
int err;

err = mutex_lock_killable(&loop_ctl_mutex);
err = mutex_lock_killable(&lo->lo_mutex);
if (err)
return err;
if (lo->lo_state != Lo_bound) {
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
return -ENXIO;
}
/*
Expand All @@ -1325,11 +1324,11 @@ static int loop_clr_fd(struct loop_device *lo)
*/
if (atomic_read(&lo->lo_refcnt) > 1) {
lo->lo_flags |= LO_FLAGS_AUTOCLEAR;
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
return 0;
}
lo->lo_state = Lo_rundown;
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);

return __loop_clr_fd(lo, false);
}
Expand All @@ -1344,7 +1343,7 @@ loop_set_status(struct loop_device *lo, const struct loop_info64 *info)
bool partscan = false;
bool size_changed = false;

err = mutex_lock_killable(&loop_ctl_mutex);
err = mutex_lock_killable(&lo->lo_mutex);
if (err)
return err;
if (lo->lo_encrypt_key_size &&
Expand Down Expand Up @@ -1411,7 +1410,7 @@ loop_set_status(struct loop_device *lo, const struct loop_info64 *info)
partscan = true;
}
out_unlock:
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
if (partscan)
loop_reread_partitions(lo, bdev);

Expand All @@ -1425,11 +1424,11 @@ loop_get_status(struct loop_device *lo, struct loop_info64 *info)
struct kstat stat;
int ret;

ret = mutex_lock_killable(&loop_ctl_mutex);
ret = mutex_lock_killable(&lo->lo_mutex);
if (ret)
return ret;
if (lo->lo_state != Lo_bound) {
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
return -ENXIO;
}

Expand All @@ -1448,10 +1447,10 @@ loop_get_status(struct loop_device *lo, struct loop_info64 *info)
lo->lo_encrypt_key_size);
}

/* Drop loop_ctl_mutex while we call into the filesystem. */
/* Drop lo_mutex while we call into the filesystem. */
path = lo->lo_backing_file->f_path;
path_get(&path);
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
ret = vfs_getattr(&path, &stat, STATX_INO, AT_STATX_SYNC_AS_STAT);
if (!ret) {
info->lo_device = huge_encode_dev(stat.dev);
Expand Down Expand Up @@ -1637,7 +1636,7 @@ static int lo_simple_ioctl(struct loop_device *lo, unsigned int cmd,
{
int err;

err = mutex_lock_killable(&loop_ctl_mutex);
err = mutex_lock_killable(&lo->lo_mutex);
if (err)
return err;
switch (cmd) {
Expand All @@ -1653,7 +1652,7 @@ static int lo_simple_ioctl(struct loop_device *lo, unsigned int cmd,
default:
err = lo->ioctl ? lo->ioctl(lo, cmd, arg) : -EINVAL;
}
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
return err;
}

Expand Down Expand Up @@ -1879,35 +1878,41 @@ static int lo_open(struct block_device *bdev, fmode_t mode)
struct loop_device *lo;
int err;

/*
* take loop_ctl_mutex to protect lo pointer from race with
* loop_control_ioctl(LOOP_CTL_REMOVE), however, to reduce contention
* release it prior to updating lo->lo_refcnt.
*/
err = mutex_lock_killable(&loop_ctl_mutex);
if (err)
return err;
lo = bdev->bd_disk->private_data;
if (!lo) {
err = -ENXIO;
goto out;
mutex_unlock(&loop_ctl_mutex);
return -ENXIO;
}

atomic_inc(&lo->lo_refcnt);
out:
err = mutex_lock_killable(&lo->lo_mutex);
mutex_unlock(&loop_ctl_mutex);
return err;
if (err)
return err;
atomic_inc(&lo->lo_refcnt);
mutex_unlock(&lo->lo_mutex);
return 0;
}

static void lo_release(struct gendisk *disk, fmode_t mode)
{
struct loop_device *lo;
struct loop_device *lo = disk->private_data;

mutex_lock(&loop_ctl_mutex);
lo = disk->private_data;
mutex_lock(&lo->lo_mutex);
if (atomic_dec_return(&lo->lo_refcnt))
goto out_unlock;

if (lo->lo_flags & LO_FLAGS_AUTOCLEAR) {
if (lo->lo_state != Lo_bound)
goto out_unlock;
lo->lo_state = Lo_rundown;
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
/*
* In autoclear mode, stop the loop thread
* and remove configuration after last close.
Expand All @@ -1924,7 +1929,7 @@ static void lo_release(struct gendisk *disk, fmode_t mode)
}

out_unlock:
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
}

static const struct block_device_operations lo_fops = {
Expand Down Expand Up @@ -1963,10 +1968,10 @@ static int unregister_transfer_cb(int id, void *ptr, void *data)
struct loop_device *lo = ptr;
struct loop_func_table *xfer = data;

mutex_lock(&loop_ctl_mutex);
mutex_lock(&lo->lo_mutex);
if (lo->lo_encryption == xfer)
loop_release_xfer(lo);
mutex_unlock(&loop_ctl_mutex);
mutex_unlock(&lo->lo_mutex);
return 0;
}

Expand Down Expand Up @@ -2152,6 +2157,7 @@ static int loop_add(struct loop_device **l, int i)
disk->flags |= GENHD_FL_NO_PART_SCAN;
disk->flags |= GENHD_FL_EXT_DEVT;
atomic_set(&lo->lo_refcnt, 0);
mutex_init(&lo->lo_mutex);
lo->lo_number = i;
spin_lock_init(&lo->lo_lock);
disk->major = LOOP_MAJOR;
Expand Down Expand Up @@ -2182,6 +2188,7 @@ static void loop_remove(struct loop_device *lo)
blk_cleanup_queue(lo->lo_queue);
blk_mq_free_tag_set(&lo->tag_set);
put_disk(lo->lo_disk);
mutex_destroy(&lo->lo_mutex);
kfree(lo);
}

Expand Down Expand Up @@ -2261,15 +2268,21 @@ static long loop_control_ioctl(struct file *file, unsigned int cmd,
ret = loop_lookup(&lo, parm);
if (ret < 0)
break;
ret = mutex_lock_killable(&lo->lo_mutex);
if (ret)
break;
if (lo->lo_state != Lo_unbound) {
ret = -EBUSY;
mutex_unlock(&lo->lo_mutex);
break;
}
if (atomic_read(&lo->lo_refcnt) > 0) {
ret = -EBUSY;
mutex_unlock(&lo->lo_mutex);
break;
}
lo->lo_disk->private_data = NULL;
mutex_unlock(&lo->lo_mutex);
idr_remove(&loop_index_idr, lo->lo_number);
loop_remove(lo);
break;
Expand Down
1 change: 1 addition & 0 deletions drivers/block/loop.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -62,6 +62,7 @@ struct loop_device {
struct request_queue *lo_queue;
struct blk_mq_tag_set tag_set;
struct gendisk *lo_disk;
struct mutex lo_mutex;
};

struct loop_cmd {
Expand Down

0 comments on commit 6cc8e74

Please sign in to comment.