On Fri, Mar 15, 2013 at 06:16:07PM +0100, Johan Hovold wrote:
> On Wed, Feb 27, 2013 at 01:52:27PM +0100, Johan Hovold wrote:
> > Make sure to check the serial disconnected flag before accessing port
> > private data after waking up.
> >
> > This fixes a use after free in the ftdi_sio introduced by commit
> > 876ae50d94b ("USB: ftdi_sio: fix race condition in TIOCMIWAIT, and abort
> > of TIOCMIWAIT when the device is removed").
> >
> > When switching to tty ports, some lifetime assumptions where changed.
> > Specifically, close can now be called before the final tty reference is
> > dropped as part of hangup at device disconnect. Even with the ftdi
> > private-data refcounting this means that the port private data can be
> > freed while a process is sleeping on modem-status changes and thus
> > cannot be relied on to detect disconnects when woken up.
>
> Greg, those changed life-times introduced a second use after free here
> as well: the wait queue itself. This affects all usb-serial drivers with
> private wait queues.
>
> My third series with the TIOCMIWAIT-rework fixes this problem, but
> I'll submit something that can more easily be backported to stable
> first.
>
> Can you hold back this patch and the two follow up series (or if you
> prefer all three of my USB-series) and I'll respin and resubmit them
> shortly?
Ok, I'm totally confused :)
Care to resend me what you want to have applied to 3.9-final as one
series, and then, anything you want to have for 3.10, as a separate
series? The second one can be "on top" of the first one, if you need it
to be.
For now, consider all patches you have sent to me previously, that I
have not applied, as dropped from my todo queues.
thanks,
greg k-h
--
To unsubscribe from this list: send the line "unsubscribe stable" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html