On Mon, 18 May 2009, David Woodhouse wrote: > On Mon, 2009-05-18 at 01:31 +0200, Daniel Mack wrote: > > Ok, good - thanks for testing. Would you care to send that with a > > commit log and your sign-off? > > I suppose so, if I drop the addition of cancel_delayed_work_sync() to > ftdi_sio_port_remove(). That doesn't need to be there, AFAICT -- there > can't be any work outstanding unless the device is open anyway. And the > private data structure isn't going away until the device is closed... at > which point the pending work gets flushed appropriately. If you say so... I haven't followed the driver's logic all the way through. The mere fact that the delayed_work gets cancelled _eventually_ might not be good enough. In particular, the driver isn't supposed to submit any URBs after usb_serial_disconnect() returns, regardless of whether or not the device file is open. So if there's a chance the delayed_work routine could submit an URB, it really should be cancelled during ftdi_sio_port_remove(). 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
