linux-kernel - Re: [PATCH 0/2] vfio/type1/pci: IOMMU PFNMAP invalidation

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 19 May 2020 21:23:52 -0300
From:   Jason Gunthorpe <jgg@...pe.ca>
To:     Alex Williamson <alex.williamson@...hat.com>
Cc:     kvm@...r.kernel.org, linux-kernel@...r.kernel.org,
        cohuck@...hat.com, peterx@...hat.com
Subject: Re: [PATCH 0/2] vfio/type1/pci: IOMMU PFNMAP invalidation

On Thu, May 14, 2020 at 10:51:46AM -0600, Alex Williamson wrote:
> This is a follow-on series to "vfio-pci: Block user access to disabled
> device MMIO"[1], which extends user access blocking of disabled MMIO
> ranges to include unmapping the ranges from the IOMMU.  The first patch
> adds an invalidation callback path, allowing vfio bus drivers to signal
> the IOMMU backend to unmap ranges with vma level granularity.  This
> signaling is done both when the MMIO range becomes inaccessible due to
> memory disabling, as well as when a vma is closed, making up for the
> lack of tracking or pinning for non-page backed vmas.  The second
> patch adds registration and testing interfaces such that the IOMMU
> backend driver can test whether a given PFNMAP vma is provided by a
> vfio bus driver supporting invalidation.  We can then implement more
> restricted semantics to only allow PFNMAP DMA mappings when we have
> such support, which becomes the new default.
> 
> Jason, if you'd like Suggested-by credit for the ideas here I'd be
> glad to add it.  Thanks,

Certainly a Reported-by would be OK

The only thing I don't like here is this makes some P2P DMA mapping
scheme for VMAs with invalidation that is completely private to vfio.

Many of us want this in other subsystems, and there are legimiate uses
for vfio to import BAR memory for P2P from other places than itself.

So I would really rather this be supported by the core kernel in some
way.

That said, this is a bug fix, and we still don't have much agreement
on what the core kernel version should look like, let alone how it
should work with an IOMMU. 

So maybe this goes ahead as is and we can figure out how to replace it
with something general later on?

Jason