On Thu, 3 Feb 2011, Greg KH wrote: > On Tue, Dec 28, 2010 at 10:25:47PM +0200, Ionut Nicu wrote: > > Hi, > > > > On Tue, 2010-12-28 at 22:21 +0200, Ionut Nicu wrote: > > > If usb_deregister() is called after usb_serial_deregister() when > > > the device is plugged in, the following Oops occurs: > > > > This is true for all usb serial drivers. I tested it also with pl2303 > > and I get the same crash. > > > > I wasn't really sure if sending this patch is the right way to fix it. > > This driver seems to be the only one calling these two functions in the > > order that causes the Oops. > > > > Maybe the functions are designed to be called this way, or maybe there's > > a bug that needs to be fixed in the usb serial core rather than applying > > this patch. I'm waiting for your thoughts on this. > > Hm, something must have changed as it used to be that we had to > unregister with the usb core before the serial one. Any chance in > trying to track this down to the real problem that is happening here? Isn't it pretty clear? The situation is just as both of you have said: Drivers must unregister with the USB core before the usb-serial core. The ti_usb driver does this in the wrong order and consequently crashes. The fix is to switch the order of unregistrations in ti_usb_3410_5052.c, which IIRC is what the patch did. Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
