Re: [dpdk-dev] [PATCH v1 1/1] kernel/linux: introduce vfio_pf kernel module

[Jerin Jacob <jerinjacobk@gmail.com>]

On Thu, Nov 7, 2019 at 4:03 AM Alex Williamson
<alex.williamson@redhat.com> wrote:
>
> On Thu, 31 Oct 2019 18:03:53 +0100
> Thomas Monjalon <thomas@monjalon.net> wrote:
>
> > We don't get enough attention on this topic.
> > Let me rephrase the issue and the proposals with more people Cc'ed.
> >
> > We are talking about SR-IOV VFs in VMs
> > with a PF managed on the host by DPDK.
> > The PF driver is either a (1) bifurcated (Mellanox case),
> > or (2) bound to UIO with igb_uio, or (3) bound to VFIO.
> >
> > In case 1, the PF is still managed by a kernel driver, so no issue.
> >
> > In case 2, the PF is managed by UIO.
> > There is no SR-IOV support in upstream UIO,
> > but the out-of-tree module igb_uio works.
> > However we would like to drop this legacy module from DPDK.
> > Some (most) Linux distributions do not package igb_uio anyway.
> > The other issue is that igb_uio is using physical addressing,
> > which is not acceptable with OCTEON TX2 for performance reason.
> >
> > In case 3, the PF is managed by VFIO. This is the case we want to fix.
> > VFIO does not allow to create VFs.
> > The workaround is to create VFs before binding the PF to VFIO.
> > But since Linux 4.19, VFIO forbids any SR-IOV VF management.
> > There is a security concern about allowing userspace to manage SR-IOV
> > VF messages and taking the responsibility for VFs in the guest.
> >
> > It is desired to allow the system admin deciding the security levels,
> > by adding a flag in VFIO "let me manage VFs, I know what I am doing".
> > Reference of "recent" discussion: https://lkml.org/lkml/2018/3/6/855
> > For now, there is no upstream solution merged.
> >
> > This patch is proposing a solution using an out-of-tree module.
> > In this case, the admin will decide explicitly to bind the PF to vfio_pf.
> > Unfortunately this solution won't work in environments which
> > forbid any out-of-tree module.
> > Another concern is that it looks like DPDK-only solution.
> >
> > We have an issue but we do not want to propose a half-solution
> > which would harm other projects and users.
> > So the question is:
> > Do we accept this patch as a temporary solution?
> > Or can we get an agreement soon for an upstream kernel solution?
> >
> > Thanks for reading and giving your (clear) opinion.

Thanks, Alex for the feedback.

> I'm pretty appalled that anyone would consider shipping this module and
> actually claiming that it's supported in some way.  Seriously, it's

Actually DPDK already shipping with this hack using the igb_uio module for UIO.

https://git.dpdk.org/dpdk/tree/kernel/linux/igb_uio/igb_uio.c#n44.

> disturbing to see a driver that intentionally circumvents a security
> issue that we all seem to agree exists, but just hand wave that it
> doesn't apply to dpdk configurations.

Yes. There is a security issue wrt netdev VFs. That's the reason, I
was scared to
submit any patch in upstream on this front. Having said that, OVS-DPDK
kind of userland
programs would like to define the fate of the netdev VF packets of the guest
as it is vswitch application. So there is a perception change in who
is controlling the who.


> Ideas have been suggested
> upstream for for quarantining VFs generated from user owned PFs such
> that we require an opt-in to make use of them in this way.  Nobody
> seems to be pursuing such ideas upstream.  I don't even see upstream
> proposals to add a scary sounding module option to vfio-pci that would
> taint the kernel, but make this available.  If nothing else, please
> remove the vfio name from this abomination, it has nothing to do with
> vfio other than to try to subvert the security and isolation that vfio
> attempts to provide.

Thanks for the feedback. Let's hold on accepting this patch.

We would like to have an upstream solution so that DPDK needs to only focus
on userspace.

I will work on submitting a patch for the discussion in Linux upstream.
Let see how the discussion goes, Based on the that, We can revisit
fate of this module.


> Alex
>