[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Pine.LNX.4.44L0.1909031303500.1859-100000@iolanthe.rowland.org>
Date: Tue, 3 Sep 2019 13:06:07 -0400 (EDT)
From: Alan Stern <stern@...land.harvard.edu>
To: David Howells <dhowells@...hat.com>
cc: Guenter Roeck <linux@...ck-us.net>, <viro@...iv.linux.org.uk>,
Casey Schaufler <casey@...aufler-ca.com>,
Stephen Smalley <sds@...ho.nsa.gov>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
<nicolas.dichtel@...nd.com>, <raven@...maw.net>,
Christian Brauner <christian@...uner.io>,
<keyrings@...r.kernel.org>, <linux-usb@...r.kernel.org>,
<linux-security-module@...r.kernel.org>,
<linux-fsdevel@...r.kernel.org>, <linux-api@...r.kernel.org>,
<linux-block@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 08/11] usb: Add USB subsystem notifications [ver #7]
On Tue, 3 Sep 2019, David Howells wrote:
> Guenter Roeck <linux@...ck-us.net> wrote:
>
> > > > This added call to usbdev_remove() results in a crash when running
> > > > the qemu "tosa" emulation. Removing the call fixes the problem.
> > >
> > > Yeah - I'm going to drop the bus notification messages for now.
> > >
> > It is not the bus notification itself causing problems. It is the
> > call to usbdev_remove().
>
> Unfortunately, I don't know how to fix it and don't have much time to
> investigate it right now - and it's something that can be added back later.
The cause of your problem is quite simple:
static int usbdev_notify(struct notifier_block *self,
unsigned long action, void *dev)
{
switch (action) {
case USB_DEVICE_ADD:
+ post_usb_device_notification(dev, NOTIFY_USB_DEVICE_ADD, 0);
break;
case USB_DEVICE_REMOVE:
+ post_usb_device_notification(dev, NOTIFY_USB_DEVICE_REMOVE, 0);
+ usbdev_remove(dev);
+ break;
+ case USB_BUS_ADD:
+ post_usb_bus_notification(dev, NOTIFY_USB_BUS_ADD, 0);
+ break;
+ case USB_BUS_REMOVE:
+ post_usb_bus_notification(dev, NOTIFY_USB_BUS_REMOVE, 0);
usbdev_remove(dev);
break;
}
The original code had usbdev_remove(dev) under the USB_DEVICE_REMOVE
case. The patch mistakenly moves it, putting it under the
USB_BUS_REMOVE case.
If the usbdev_remove() call were left where it was originally, the
problem would be solved.
Alan Stern
Powered by blists - more mailing lists