[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <95b22758709c7e3bec2e9b3e31a94333b363a353.camel@kernel.crashing.org>
Date: Tue, 10 Jul 2018 09:50:29 +1000
From: Benjamin Herrenschmidt <benh@...nel.crashing.org>
To: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>,
"Eric W. Biederman" <ebiederm@...ssion.com>,
Joel Stanley <joel@....id.au>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 1/2] drivers: core: Don't try to use a dead glue_dir
On Sat, 2018-07-07 at 18:51 +0200, Greg Kroah-Hartman wrote:
>
> > diff --git a/drivers/base/core.c b/drivers/base/core.c
> > index b610816eb887..e9eff2099896 100644
> > --- a/drivers/base/core.c
> > +++ b/drivers/base/core.c
> > @@ -1517,11 +1517,13 @@ static struct kobject *get_device_parent(struct device *dev,
> >
> > /* find our class-directory at the parent and reference it */
> > spin_lock(&dev->class->p->glue_dirs.list_lock);
> > - list_for_each_entry(k, &dev->class->p->glue_dirs.list, entry)
> > + list_for_each_entry(k, &dev->class->p->glue_dirs.list, entry) {
> > if (k->parent == parent_kobj) {
> > - kobj = kobject_get(k);
> > - break;
> > + kobj = kobject_get_unless_zero(k);
> > + if (kobj)
> > + break;
>
> A parent directory _should_ not ever be able to be removed before the
> object being removed was, as we should have had a reference to it,
> right? So I don't see how this can get hit "in real life".
>
> Yes, enabling kobject debugging does keep objects around for a long time
> in order to try to help figure out where people are messing up their
> usage of them. What subsystem is doing this in a way that causes
> problems here? Shouldn't we fix that up instead?
The broken subsystem is the driver core itself :-) See the descriptions
here and in patch 2/2.
Note: This is a more generic problem with ksets vs relying on the magic
sysfs cleanup happening in kobject_release().
Any kobject that is a member of a kset and doesn't get explicitely
removed from sysfs with kobject_del() prior to dropping the last
reference with kobject_put() (and thus relies instead on the automatic
cleanup done by kobject_release()) will be exposed to the race:
The last kobject_put() will drop the refcount to 0 while the object is
still in the kset. Only some amount of time later (which can be very
short or very long if you enable kobject debugging), will
kobject_release() take it out of sysfs and out of the kset.
Thus the object will be visible, with a 0 refcount, to anything that
"walks" the kset during that period.
This is exactly what happens with the gluedirs in the device core, but
it could happen elsewhere for all I know.
Cheers,
Ben.
Powered by blists - more mailing lists