Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Fri, 27 Sep 2019 10:21:00 -0400 (EDT)
From:   Alan Stern <stern@rowland.harvard.edu>
To:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>
cc:     Andrey Konovalov <andreyknvl@google.com>,
        "Rafael J. Wysocki" <rafael@kernel.org>,
        syzbot <syzbot+e74a998ca8f1df9cc332@syzkaller.appspotmail.com>,
        LKML <linux-kernel@vger.kernel.org>,
        USB list <linux-usb@vger.kernel.org>,
        syzkaller-bugs <syzkaller-bugs@googlegroups.com>
Subject: Re: WARNING in pvr2_i2c_core_done
In-Reply-To: <20190927051042.GA1767219@kroah.com>
Message-ID: <Pine.LNX.4.44L0.1909271015510.1698-100000@iolanthe.rowland.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, 27 Sep 2019, Greg Kroah-Hartman wrote:

> > It turns out the reason for this error is simple: The driver 
> > unregisters its subdevices in the release handler instead of in the 
> > disconnect handler.  There probably is documentation about this 
> > somewhere, but I don't know exactly where -- maybe Greg remembers.
> 
> Nope, I don't remember.  It should happen in the disconnect handler, odd
> of it to be in release, but maybe that's the "easiest" way for v4l to
> handle this?

This isn't a question of "easiest".  Unregistering child devices in a
release handler is just _wrong_, plain and simple.  That's what gives
rise to the

"sysfs group 'power' not found for kobject 'i2c-0'"

warning in the kernel log.  The group can't be found because it has 
already been removed; it gets destroyed when the parent USB interface 
device is unregistered, because unregistering a device also removes 
from sysfs everything below that device.

Alan Stern