Re: [PATCH] usbip: vhci_sysfs: fix potential Spectre v1

[Greg Kroah-Hartman]

On Thu, May 17, 2018 at 12:57:49PM -0500, Gustavo A. R. Silva wrote:
> Hi Greg,
> 
> On 05/17/2018 01:51 AM, Greg Kroah-Hartman wrote:
> > On Wed, May 16, 2018 at 05:22:00PM -0500, Gustavo A. R. Silva wrote:
> > > pdev_nr and rhport can be controlled by user-space, hence leading to
> > > a potential exploitation of the Spectre variant 1 vulnerability.
> > > 
> > > This issue was detected with the help of Smatch:
> > > drivers/usb/usbip/vhci_sysfs.c:238 detach_store() warn: potential
> > > spectre issue 'vhcis'
> > > drivers/usb/usbip/vhci_sysfs.c:328 attach_store() warn: potential
> > > spectre issue 'vhcis'
> > > drivers/usb/usbip/vhci_sysfs.c:338 attach_store() warn: potential
> > > spectre issue 'vhci->vhci_hcd_ss->vdev'
> > > drivers/usb/usbip/vhci_sysfs.c:340 attach_store() warn: potential
> > > spectre issue 'vhci->vhci_hcd_hs->vdev'
> > 
> > Nit, no need to line-wrap long error messages from tools :)
> > 
> 
> Got it.
> 
> > > Fix this by sanitizing pdev_nr and rhport before using them to index
> > > vhcis and vhci->vhci_hcd_ss->vdev respectively.
> > > 
> > > Notice that given that speculation windows are large, the policy is
> > > to kill the speculation on the first load and not worry if it can be
> > > completed with a dependent load/store [1].
> > > 
> > > [1] https://marc.info/?l=linux-kernel&m=152449131114778&w=2
> > > 
> > > Cc: stable@xxxxxxxxxxxxxxx
> > > Signed-off-by: Gustavo A. R. Silva <gustavo@xxxxxxxxxxxxxx>
> > > ---
> > >   drivers/usb/usbip/vhci_sysfs.c | 6 ++++++
> > >   1 file changed, 6 insertions(+)
> > > 
> > > diff --git a/drivers/usb/usbip/vhci_sysfs.c b/drivers/usb/usbip/vhci_sysfs.c
> > > index 4880838..9045888 100644
> > > --- a/drivers/usb/usbip/vhci_sysfs.c
> > > +++ b/drivers/usb/usbip/vhci_sysfs.c
> > > @@ -10,6 +10,8 @@
> > >   #include <linux/platform_device.h>
> > >   #include <linux/slab.h>
> > > +#include <linux/nospec.h>
> > > +
> > >   #include "usbip_common.h"
> > >   #include "vhci.h"
> > > @@ -235,6 +237,8 @@ static ssize_t detach_store(struct device *dev, struct device_attribute *attr,
> > >   	if (!valid_port(pdev_nr, rhport))
> > >   		return -EINVAL;
> > > +	pdev_nr = array_index_nospec(pdev_nr, vhci_num_controllers);
> > > +	rhport = array_index_nospec(rhport, VHCI_HC_PORTS);
> > 
> > Shouldn't we just do this in one place, in the valid_port() function?
> > 
> > That way it keeps the range checking logic in one place (now it is in 3
> > places in the function), which should make maintenance much simpler.
> > 
> 
> Yep, I thought about that, the thing is: what happens if the hardware is
> "trained" to predict that valid_port always evaluates to false, and then
> malicious values are stored in pdev_nr and nhport?
> 
> It seems to me that under this scenario we need to serialize instructions in
> this place.
> 
> What do you think?

I don't understand, it should not matter where you put the barrier.  Be
it a function call back or right after it, it does the same thing, it
stops speculation from crossing that barrier.

So it _should_ work either way, if I understand the issue correctly.

If not, what am I missing?

thanks,

greg k-h