On Sat, May 30, 2020 at 09:14:50AM -0700, Kees Cook wrote:
> On Sat, May 30, 2020 at 04:13:29PM +0200, Christian Brauner wrote:
> > On Fri, May 29, 2020 at 10:47:12PM -0700, Kees Cook wrote:
> > > Oh hey! Look at scm_detach_fds_compat(). It needs this too. (And it's
> > > missing the cgroup tracking.) That would fix:
> > >
> > > 48a87cc26c13 ("net: netprio: fd passed in SCM_RIGHTS datagram not set correctly")
> > > d84295067fc7 ("net: net_cls: fd passed in SCM_RIGHTS datagram not set correctly")
> > >
> > > So, yes, let's get this fixed up. I'd say first fix the missing sock
> > > update in the compat path (so it can be CCed stable). Then fix the missing
> >
> > send this patch to net.
> >
> > > sock update in pidfd_getfd() (so it can be CCed stable), then write the
> >
> > send this patch to me.
> >
> > > helper with a refactoring of scm_detach_fds(), scm_detach_fds_compat(),
> >
> > this would be net-next most likely.
> >
> > > and pidfd_getfd(). And then add the addfd seccomp user_notif ioctl cmd.
> >
> > If you do this first, I'd suggest you resend the series here after all
> > this has been merged. We're not in a rush since this won't make it for
> > the 5.8 merge window anyway. By the time the changes land Kees might've
> > applied my changes to his tree so you can rebase yours on top of it
> > relieving Kees from fixing up merge conflicts.
> >
> > About your potential net and net-next changes. Just in case you don't
> > know - otherwise ignore this - please read and treat
> > https://www.kernel.org/doc/Documentation/networking/netdev-FAQ.txt
> > as the gospel. Also note, that after this Sunday - assuming Linus
> > releases - net-next will be closed until the merge window is closed,
> > i.e. for _at least_ 2 weeks. After the merge window closes you can check
> > http://vger.kernel.org/~davem/net-next.html
> > which either has a picture saying "Come In We're Open" or a sign saying
> > "Sorry, We're Closed". Only send when the first sign is up or the wrath
> > of Dave might hit you. :)
>
> Yeah, timing is awkward here. I was originally thinking it could all
> just land via seccomp (with appropriate Acks). Hmmm.
I don't particularly care so sure. :)
Christian
_______________________________________________
Containers mailing list
Containers@xxxxxxxxxxxxxxxxxxxxxxxxxx
https://lists.linuxfoundation.org/mailman/listinfo/containers
