Re: CVE request: kernel: memory disclosure in SO_BSDCOMPAT gsopt

[Florian Weimer <fw () deneb enyo de>]

* Eugene Teo:

> Eugene Teo wrote:
> > On Tue, Mar 3, 2009 at 6:49 AM, Steven M. Christey
> > <coley () linus mitre org> wrote:
> > > On Wed, 25 Feb 2009, Eugene Teo wrote:
> > >
> > > > Eugene Teo wrote:
> > > > > [...]
> > > > > The fix for CVE-2009-0676 (upstream commit df0bca04) is incomplete. Note
> > > > > that the same problem of leaking kernel memory will reappear if someone
> > > > > on some architecture uses struct timeval with some internal padding (for
> > > > > example tv_sec 64-bit and tv_usec 32-bit) --- then, you are going to
> > > > > leak the padded bytes to userspace.
> > > Is this going to require a separate CVE identifier?  If a new minor
> > > version of the kernel wasn't released yet, then I'd consider the fix to be
> > > little more than a couple patch-discussion messages in a single Bugzilla
> > > entry.
> >
> > No, it shouldn't. Please use the same CVE name. Thanks.
>
> But you might want to add the link to the new CVE-2009-0676 patch in:
> http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2009-0676

BTW, the reproducer I saw in your bug tracker doesn't initialize the
len field.  It only worked for me after I fixed that.