From: Jan Kiszka <jan.kiszka@siemens.com>
To: "Michael S. Tsirkin" <mst@redhat.com>,
David Kiarie <davidkiarie4@gmail.com>
Cc: marcel@redhat.com, valentine.sinitsyn@gmail.com, qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [V6 0/4] AMD IOMMU
Date: Tue, 1 Mar 2016 14:48:19 +0100 [thread overview]
Message-ID: <56D59DA3.3040002@siemens.com> (raw)
In-Reply-To: <20160301134419-mutt-send-email-mst@redhat.com>
On 2016-03-01 14:07, Michael S. Tsirkin wrote:
> On Sun, Feb 21, 2016 at 09:10:56PM +0300, David Kiarie wrote:
>> Hello there,
>>
>> Repost, AMD IOMMU patches version 6.
>>
>> Changes since version 5
>> -Fixed macro formating issues
>> -changed occurences of IO MMU to IOMMU for consistency
>> -Fixed capability registers duplication
>> -Rebased to current master
>>
>> David Kiarie (4):
>> hw/i386: Introduce AMD IOMMU
>> hw/core: Add AMD IOMMU to machine properties
>> hw/i386: ACPI table for AMD IOMMU
>> hw/pci-host: Emulate AMD IOMMU
>
> I went over AMD IOMMU spec.
> I'm concerned that it appears that there's no chance for it to
> work correctly if host caches invalid PTE entries.
>
> The spec vaguely discusses write-protecting such PTEs but
> that would be very complex if it can be made to work at all.
>
> This means that this can't work with e.g. VFIO.
> It can only work with emulated devices.
You mean it can't work if we program a real IOMMU (for VFIO) with
translated data from the emulated one but cannot track any updates of
the related page tables because the guest is not required to issue
traceable flush requests? Hmm, too bad.
>
> OTOH VTD can easily support PTE shadowing by setting a flag.
Do you mean RWBF=1 in the CAP register? Given that "Newer hardware
implementations are expected to NOT require explicit software flushing
of write buffers and report RWBF=0 in the Capability register", we may
eventually run into guests that no longer check that flag if we expose
something that looks like a "newer" implementation.
However, this flag is not set right now in our VT-d model.
>
> I'd like us to find some way to avoid possibility
> of user error creating a configuration mixing e.g.
> vfio with the amd iommu.
>
> I'm not sure how to do this.
>
> Any idea?
There is likely no way around write-protecting the IOMMU page tables (in
KVM mode) once we evaluated and cached them somewhere. For now, I would
simply deny vfio while an IOMMU is active on x86.
Jan
--
Siemens AG, Corporate Technology, CT RDA ITP SES-DE
Corporate Competence Center Embedded Linux
next prev parent reply other threads:[~2016-03-01 13:48 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-02-21 18:10 [Qemu-devel] [V6 0/4] AMD IOMMU David Kiarie
2016-02-21 18:10 ` [Qemu-devel] [V6 1/4] hw/i386: Introduce " David Kiarie
2016-02-25 15:43 ` Marcel Apfelbaum
2016-02-26 6:23 ` David Kiarie
2016-03-02 4:00 ` David Kiarie
2016-03-02 4:08 ` David Kiarie
2016-03-03 9:40 ` Marcel Apfelbaum
2016-03-03 9:34 ` Marcel Apfelbaum
2016-03-02 19:11 ` David Kiarie
2016-03-03 12:16 ` Marcel Apfelbaum
2016-02-21 18:10 ` [Qemu-devel] [V6 2/4] hw/core: Add AMD IOMMU to machine properties David Kiarie
2016-02-21 20:09 ` Jan Kiszka
2016-03-02 20:51 ` David Kiarie
2016-03-03 9:28 ` Marcel Apfelbaum
2016-03-11 13:20 ` Michael S. Tsirkin
2016-02-21 18:10 ` [Qemu-devel] [V6 3/4] hw/i386: ACPI table for AMD IOMMU David Kiarie
2016-02-21 18:20 ` Jan Kiszka
2016-02-21 19:00 ` David Kiarie
2016-02-21 18:11 ` [Qemu-devel] [V6 4/4] hw/pci-host: Emulate " David Kiarie
2016-02-22 11:22 ` Marcel Apfelbaum
[not found] ` <56D75688.1020500@gmail.com>
2016-03-02 21:17 ` Michael S. Tsirkin
2016-03-02 22:04 ` David Kiarie
2016-03-03 9:49 ` Michael S. Tsirkin
2016-03-03 11:47 ` David Kiarie
2016-03-03 12:02 ` Marcel Apfelbaum
2016-03-03 12:06 ` Marcel Apfelbaum
2016-03-03 12:18 ` David Kiarie
2016-03-03 12:58 ` Michael S. Tsirkin
2016-03-08 17:15 ` David Kiarie
2016-03-11 13:22 ` Michael S. Tsirkin
2016-03-13 0:14 ` David Kiarie
2016-03-13 13:59 ` Michael S. Tsirkin
2016-02-21 20:20 ` [Qemu-devel] [V6 0/4] " Jan Kiszka
2016-02-22 5:57 ` David Kiarie
2016-02-22 7:29 ` Jan Kiszka
2016-02-22 11:05 ` David Kiarie
2016-02-22 11:12 ` Jan Kiszka
2016-03-01 13:07 ` Michael S. Tsirkin
2016-03-01 13:48 ` Jan Kiszka [this message]
2016-03-01 13:55 ` Michael S. Tsirkin
2016-03-01 14:12 ` Jan Kiszka
2016-03-01 14:18 ` Jan Kiszka
2016-03-01 14:30 ` Michael S. Tsirkin
2016-03-01 14:35 ` Jan Kiszka
2016-03-01 14:19 ` Michael S. Tsirkin
2016-03-01 14:00 ` Jan Kiszka
2016-03-01 20:11 ` Michael S. Tsirkin
2016-03-01 20:17 ` Jan Kiszka
2016-03-01 20:39 ` Michael S. Tsirkin
2016-03-01 21:23 ` Jan Kiszka
2016-03-01 22:35 ` Michael S. Tsirkin
2016-03-02 21:17 ` David Kiarie
2016-03-02 21:32 ` Michael S. Tsirkin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=56D59DA3.3040002@siemens.com \
--to=jan.kiszka@siemens.com \
--cc=davidkiarie4@gmail.com \
--cc=marcel@redhat.com \
--cc=mst@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=valentine.sinitsyn@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).