On Mon, 7 Dec 2009, Oliver Neukum wrote:
> Am Montag, 7. Dezember 2009 18:58:00 schrieb Alan Cox:
> > > + if (acm->port.count) {
> > > + tty = tty_port_tty_get(&acm->port);
> > > + if (tty) {
> > > + tty_hangup(tty);
> > > + tty_kref_put(tty);
> > > + }
> >
> > What locking do you have against a parallel open, hangup or close event on
> > another CPU ?
>
> Alan S.,
>
> this touches the discussion we recently had on reset_resume() for devices
> that require remote wakeup. The need for locking here depends on it.
> Are you planning to submit any patches in that regard in this merge window?
Nothing that hasn't already appeared on the mailing list.
In this case, it seems likely that no locking is needed. (But note
that I am almost completely unfamiliar with the cdc-acm driver.) Any
necessary locking should already be provided by the TTY layer or
cdc-acm itself. Aren't there other places where it does a hangup? The
same sort of locking should be needed in both places.
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
