On Sun, 2019-03-31 at 23:28 +0800, weiping zhang wrote:
> On Tue, Jan 08, 2019 at 04:57:40PM -0800, Bart Van Assche wrote:
> > > @@ -942,6 +961,14 @@ int blk_register_queue(struct gendisk *disk)
> > > goto unlock;
> > > }
> > >
> > > + ret = sysfs_create_group(&q->kobj, &queue_attr_group);
> > > + if (ret) {
> > > + kobject_del(&q->kobj);
> > > + blk_trace_remove_sysfs(dev);
> > > + kobject_put(&dev->kobj);
> > > + goto unlock;
> > > + }
> >
> > Are you sure the "goto unlock" is OK here? Shouldn't kobject_del() be called
> > to undo the kobject_add() call if sysfs_create_group() fails?
>
> Sorry, can you tell me why it's may be not safe, if goto unlock here,
> if failed to call sysfs_create_group, I think we should call
> kobject_del.
Can you address the other comments and repost your patch? I may have misread
your patch when I wrote the above comment.
Thanks,
Bart.
