On Wed, Apr 8, 2015 at 2:23 PM, Jarod Wilson <[email protected]> wrote: > This patch exports blkdev_reread_part() for block drivers, also > introduce __blkdev_reread_part(), a lockless version.
Generally the term of lockless is used when lock is avoided, and that isn't the case of __blkdev_reread_part(), because bd_mutex is definitely required here. > > For some drivers, such as loop, a reread of partitions can be run > from the release path, and bd_mutex may already be held prior to > calling ioctl_by_bdev(bdev, BLKRRPART, 0), so introduce a lockless > path for use in such cases. > > CC: Christoph Hellwig <[email protected]> > CC: Jens Axboe <[email protected]> > CC: Tejun Heo <[email protected]> > CC: Alexander Viro <[email protected]> > CC: Markus Pargmann <[email protected]> > CC: Stefan Weinhuber <[email protected]> > CC: Stefan Haberland <[email protected]> > CC: Sebastian Ott <[email protected]> > CC: Fabian Frederick <[email protected]> > CC: Ming Lei <[email protected]> > CC: David Herrmann <[email protected]> > CC: Mike Galbraith <[email protected]> > CC: Andrew Morton <[email protected]> > CC: Peter Zijlstra <[email protected]> > CC: [email protected] > CC: [email protected] > Signed-off-by: Jarod Wilson <[email protected]> > --- > block/ioctl.c | 26 +++++++++++++++++++++++--- > include/linux/fs.h | 3 +++ > 2 files changed, 26 insertions(+), 3 deletions(-) > > diff --git a/block/ioctl.c b/block/ioctl.c > index 7d8befd..64a4fcb 100644 > --- a/block/ioctl.c > +++ b/block/ioctl.c > @@ -150,21 +150,41 @@ static int blkpg_ioctl(struct block_device *bdev, > struct blkpg_ioctl_arg __user > } > } > > -static int blkdev_reread_part(struct block_device *bdev) > +/* > + * This is an exported API for the block driver, and will not > + * acquire bd_mutex, leaving it up to the caller to handle > + * any necessary locking. Actually, the function is introduced and should be used in case that bd_mutex has been held already, such as clearing fd in loop release(). > + */ > +int __blkdev_reread_part(struct block_device *bdev) > { > struct gendisk *disk = bdev->bd_disk; > - int res; > > if (!disk_part_scan_enabled(disk) || bdev != bdev->bd_contains) > return -EINVAL; > if (!capable(CAP_SYS_ADMIN)) > return -EACCES; > + > + return rescan_partitions(disk, bdev); > +} > +EXPORT_SYMBOL(__blkdev_reread_part); > + > +/* > + * This is an exported API for the block driver, and will > + * acquire bd_mutex. Make sure you aren't calling it with > + * bd_mutex already held, or we'll return -EBUSY. Strictly speaking, it should be "Make sure you aren't calling it with bd_mutex already held in current context". > + */ > +int blkdev_reread_part(struct block_device *bdev) > +{ > + int res; > + > if (!mutex_trylock(&bdev->bd_mutex)) > return -EBUSY; > - res = rescan_partitions(disk, bdev); > + res = __blkdev_reread_part(bdev); > mutex_unlock(&bdev->bd_mutex); > + > return res; > } > +EXPORT_SYMBOL(blkdev_reread_part); > > static int blk_ioctl_discard(struct block_device *bdev, uint64_t start, > uint64_t len, int secure) > diff --git a/include/linux/fs.h b/include/linux/fs.h > index f4131e8..11398e3 100644 > --- a/include/linux/fs.h > +++ b/include/linux/fs.h > @@ -2245,6 +2245,9 @@ extern struct block_device *blkdev_get_by_path(const > char *path, fmode_t mode, > extern struct block_device *blkdev_get_by_dev(dev_t dev, fmode_t mode, > void *holder); > extern void blkdev_put(struct block_device *bdev, fmode_t mode); > +extern int __blkdev_reread_part(struct block_device *bdev); > +extern int blkdev_reread_part(struct block_device *bdev); > + > #ifdef CONFIG_SYSFS > extern int bd_link_disk_holder(struct block_device *bdev, struct gendisk > *disk); > extern void bd_unlink_disk_holder(struct block_device *bdev, > -- > 1.8.3.1 > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
