[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <871qxsrfal.fsf@miraculix.mork.no>
Date: Tue, 19 Apr 2022 22:25:54 +0200
From: Bjørn Mork <bjorn@...k.no>
To: Oliver Neukum <oneukum@...e.com>
Cc: Andy Shevchenko <andy.shevchenko@...il.com>,
Johan Hovold <johan@...nel.org>,
Oleksij Rempel <linux@...pel-privat.de>,
Dongliang Mu <dzm91@...t.edu.cn>,
Oliver Neukum <oliver@...kum.org>,
"David S. Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>,
Paolo Abeni <pabeni@...hat.com>,
Dongliang Mu <mudongliangabcd@...il.com>,
syzbot+eabbf2aaa999cc507108@...kaller.appspotmail.com,
USB <linux-usb@...r.kernel.org>, netdev <netdev@...r.kernel.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] driver: usb: nullify dangling pointer in cdc_ncm_free
Oliver Neukum <oneukum@...e.com> writes:
> It seems to me that fundamentally the order of actions to handle
> a hotunplug must mirror the order in a hotplug. We can add more hooks
> if that turns out to be necessary for some drivers, but the basic
> reverse mirrored order must be supported and I very much favor
> restoring it as default.
FWIW, I agree 100% with this. Please go ahead with the revert of commit
2c9d6c2b871d ("usbnet: run unbind() before unregister_netdev()").
AFAICS, your proposed new hook should solve the original problem just
fine.
Bjørn
Powered by blists - more mailing lists