Re: [Qemu-devel] [PATCH 2/6] crypto: clear out buffer after timing pbkdf algorithm

[Daniel P. Berrange]

On Thu, Sep 08, 2016 at 12:47:43PM -0500, Eric Blake wrote:
> On 09/08/2016 11:27 AM, Daniel P. Berrange wrote:
> > The 'out' buffer will hold a key derived from master
> > password, so it is best practice to clear this buffer
> > when no longer required.
> > 
> > Signed-off-by: Daniel P. Berrange <address@hidden>
> > ---
> >  crypto/pbkdf.c | 15 ++++++++++-----
> >  1 file changed, 10 insertions(+), 5 deletions(-)
> 
> Reviewed-by: Eric Blake <address@hidden>
> 
> It still doesn't prevent the memory from being copied elsewhere (such as
> the stack being paged out), unless we go to extraordinary lengths to
> explicitly request volatile memory that can't be paged out.  I don't
> know if we need to worry about that, though.  Do any of our crypto
> libraries provide APIs for allocating local-use-only memory for
> sensitive data?

AFAICT, while gcrypt uses such APIs internally, it doesn't expose them
to users. Nettle avoids malloc entirely in its API. So if we wanted
that we'd basically need to roll our own.  I don't think this is a big
deal, but I just wanted to add the memset() as a sanity check really.

Regards,
Daniel
-- 
|: http://berrange.com      -o-    http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org              -o-             http://virt-manager.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org       -o-       http://live.gnome.org/gtk-vnc :|