From: Jason Wang <email@example.com>
To: "Jacob Pan (Jun)" <firstname.lastname@example.org>,
Jason Gunthorpe <email@example.com>
Cc: Jean-Philippe Brucker <firstname.lastname@example.org>,
email@example.com, "Raj, Ashok" <firstname.lastname@example.org>,
email@example.com, "Michael S. Tsirkin" <firstname.lastname@example.org>,
Alex Williamson <email@example.com>,
Subject: Re: [PATCH v7 00/16] vfio: expose virtual Shared Virtual Addressing to VMs
Date: Thu, 17 Sep 2020 11:53:49 +0800 [thread overview]
Message-ID: <firstname.lastname@example.org> (raw)
On 2020/9/17 上午7:09, Jacob Pan (Jun) wrote:
> Hi Jason,
> On Wed, 16 Sep 2020 15:38:41 -0300, Jason Gunthorpe <email@example.com>
>> On Wed, Sep 16, 2020 at 11:21:10AM -0700, Jacob Pan (Jun) wrote:
>>> Hi Jason,
>>> On Wed, 16 Sep 2020 14:01:13 -0300, Jason Gunthorpe <firstname.lastname@example.org>
>>>> On Wed, Sep 16, 2020 at 09:33:43AM -0700, Raj, Ashok wrote:
>>>>> On Wed, Sep 16, 2020 at 12:07:54PM -0300, Jason Gunthorpe
>>>>>> On Tue, Sep 15, 2020 at 05:22:26PM -0700, Jacob Pan (Jun)
>>>>>>>> If user space wants to bind page tables, create the PASID
>>>>>>>> with /dev/sva, use ioctls there to setup the page table
>>>>>>>> the way it wants, then pass the now configured PASID to a
>>>>>>>> driver that can use it.
>>>>>>> Are we talking about bare metal SVA?
>>>>>> What a weird term.
>>>>> Glad you noticed it at v7 :-)
>>>>> Any suggestions on something less weird than
>>>>> Shared Virtual Addressing? There is a reason why we moved from
>>>>> SVM to SVA.
>>>> SVA is fine, what is "bare metal" supposed to mean?
>>> What I meant here is sharing virtual address between DMA and host
>>> process. This requires devices perform DMA request with PASID and
>>> use IOMMU first level/stage 1 page tables.
>>> This can be further divided into 1) user SVA 2) supervisor SVA
>>> (sharing init_mm)
>>> My point is that /dev/sva is not useful here since the driver can
>>> perform PASID allocation while doing SVA bind.
>> No, you are thinking too small.
>> Look at VDPA, it has a SVA uAPI. Some HW might use PASID for the SVA.
> Could you point to me the SVA UAPI? I couldn't find it in the mainline.
> Seems VDPA uses VHOST interface?
It's the vhost_iotlb_msg defined in uapi/linux/vhost_types.h.
>> When VDPA is used by DPDK it makes sense that the PASID will be SVA
>> and 1:1 with the mm_struct.
> I still don't see why bare metal DPDK needs to get a handle of the
My understanding is that it may:
- have a unified uAPI with vSVA: alloc, bind, unbind, free
- leave the binding policy to userspace instead of the using a implied
one in the kenrel
> Perhaps the SVA patch would explain. Or are you talking about
> vDPA DPDK process that is used to support virtio-net-pmd in the guest?
>> When VDPA is used by qemu it makes sense that the PASID will be an
>> arbitary IOVA map constructed to be 1:1 with the guest vCPU physical
>> map. /dev/sva allows a single uAPI to do this kind of setup, and qemu
>> can support it while supporting a range of SVA kernel drivers. VDPA
>> and vfio-mdev are obvious initial targets.
>> *BOTH* are needed.
>> In general any uAPI for PASID should have the option to use either the
>> mm_struct SVA PASID *OR* a PASID from /dev/sva. It costs virtually
>> nothing to implement this in the driver as PASID is just a number, and
>> gives so much more flexability.
> Not really nothing in terms of PASID life cycles. For example, if user
> uses uacce interface to open an accelerator, it gets an FD_acc. Then it
> opens /dev/sva to allocate PASID then get another FD_pasid. Then we
> pass FD_pasid to the driver to bind page tables, perhaps multiple
> drivers. Now we have to worry about If FD_pasid gets closed before
> FD_acc(s) closed and all these race conditions.
I'm not sure I understand this. But this demonstrates the flexibility of
an unified uAPI. E.g it allows vDPA and VFIO device to use the same
PAISD which can be shared with a process in the guest.
For the race condition, it could be probably solved with refcnt.
> If we do not expose FD_pasid to the user, the teardown is much simpler
> and streamlined. Following each FD_acc close, PASID unbind is performed.
>>> Yi can correct me but this set is is about VFIO-PCI, VFIO-mdev will
>>> be introduced later.
>> Last patch is:
>> vfio/type1: Add vSVA support for IOMMU-backed mdevs
>> So pretty hard to see how this is not about vfio-mdev, at least a
iommu mailing list
next prev parent reply other threads:[~2020-09-17 3:54 UTC|newest]
Thread overview: 84+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-10 10:45 [PATCH v7 00/16] vfio: expose virtual Shared Virtual Addressing to VMs Liu Yi L
2020-09-10 10:45 ` [PATCH v7 01/16] iommu: Report domain nesting info Liu Yi L
2020-09-11 19:38 ` Alex Williamson
2020-09-10 10:45 ` [PATCH v7 02/16] iommu/smmu: Report empty " Liu Yi L
2021-01-12 6:50 ` Vivek Gautam
2021-01-12 9:21 ` Liu, Yi L
2021-01-12 11:05 ` Vivek Gautam
2021-01-13 5:56 ` Liu, Yi L
2021-01-19 10:03 ` Auger Eric
2021-01-23 8:59 ` Liu, Yi L
2021-02-12 7:14 ` Vivek Gautam
2021-02-12 9:57 ` Auger Eric
2021-02-12 10:18 ` Vivek Kumar Gautam
2021-02-12 11:01 ` Vivek Kumar Gautam
2021-03-03 9:44 ` Liu, Yi L
2020-09-10 10:45 ` [PATCH v7 03/16] vfio/type1: Report iommu nesting info to userspace Liu Yi L
2020-09-11 20:16 ` Alex Williamson
2020-09-12 8:24 ` Liu, Yi L
2020-09-10 10:45 ` [PATCH v7 04/16] vfio: Add PASID allocation/free support Liu Yi L
2020-09-11 20:54 ` Alex Williamson
2020-09-15 4:03 ` Liu, Yi L
2020-09-10 10:45 ` [PATCH v7 05/16] iommu/vt-d: Support setting ioasid set to domain Liu Yi L
2020-09-10 10:45 ` [PATCH v7 06/16] iommu/vt-d: Remove get_task_mm() in bind_gpasid() Liu Yi L
2020-09-10 10:45 ` [PATCH v7 07/16] vfio/type1: Add VFIO_IOMMU_PASID_REQUEST (alloc/free) Liu Yi L
2020-09-11 21:38 ` Alex Williamson
2020-09-12 6:17 ` Liu, Yi L
2020-09-10 10:45 ` [PATCH v7 08/16] iommu: Pass domain to sva_unbind_gpasid() Liu Yi L
2020-09-10 10:45 ` [PATCH v7 09/16] iommu/vt-d: Check ownership for PASIDs from user-space Liu Yi L
2020-09-10 10:45 ` [PATCH v7 10/16] vfio/type1: Support binding guest page tables to PASID Liu Yi L
2020-09-11 22:03 ` Alex Williamson
2020-09-12 6:02 ` Liu, Yi L
2020-09-10 10:45 ` [PATCH v7 11/16] vfio/type1: Allow invalidating first-level/stage IOMMU cache Liu Yi L
2020-09-10 10:45 ` [PATCH v7 12/16] vfio/type1: Add vSVA support for IOMMU-backed mdevs Liu Yi L
2020-09-10 10:45 ` [PATCH v7 13/16] vfio/pci: Expose PCIe PASID capability to guest Liu Yi L
2020-09-11 22:13 ` Alex Williamson
2020-09-12 7:17 ` Liu, Yi L
2020-09-10 10:45 ` [PATCH v7 14/16] vfio: Document dual stage control Liu Yi L
2020-09-10 10:45 ` [PATCH v7 15/16] iommu/vt-d: Only support nesting when nesting caps are consistent across iommu units Liu Yi L
2020-09-10 10:45 ` [PATCH v7 16/16] iommu/vt-d: Support reporting nesting capability info Liu Yi L
2020-09-14 4:20 ` [PATCH v7 00/16] vfio: expose virtual Shared Virtual Addressing to VMs Jason Wang
2020-09-14 8:01 ` Tian, Kevin
2020-09-14 8:57 ` Jason Wang
2020-09-14 10:38 ` Tian, Kevin
2020-09-14 11:38 ` Jason Gunthorpe
2020-09-14 13:31 ` Jean-Philippe Brucker
2020-09-14 13:47 ` Jason Gunthorpe
2020-09-14 16:22 ` Raj, Ashok
2020-09-14 16:33 ` Jason Gunthorpe
2020-09-14 16:58 ` Alex Williamson
2020-09-14 17:41 ` Jason Gunthorpe
2020-09-14 18:23 ` Alex Williamson
2020-09-14 19:00 ` Jason Gunthorpe
2020-09-14 22:33 ` Alex Williamson
2020-09-15 14:29 ` Jason Gunthorpe
2020-09-16 1:19 ` Tian, Kevin
2020-09-16 8:32 ` Jean-Philippe Brucker
2020-09-16 14:51 ` Jason Gunthorpe
2020-09-16 16:20 ` Jean-Philippe Brucker
2020-09-16 16:32 ` Jason Gunthorpe
2020-09-16 16:50 ` Auger Eric
2020-09-16 14:44 ` Jason Gunthorpe
2020-09-17 6:01 ` Tian, Kevin
2020-09-14 22:44 ` Raj, Ashok
2020-09-15 11:33 ` Jason Gunthorpe
2020-09-15 18:11 ` Raj, Ashok
2020-09-15 18:45 ` Jason Gunthorpe
2020-09-15 19:26 ` Raj, Ashok
2020-09-15 23:45 ` Jason Gunthorpe
2020-09-16 2:33 ` Jason Wang
2020-09-15 22:08 ` Jacob Pan
2020-09-15 23:51 ` Jason Gunthorpe
2020-09-16 0:22 ` Jacob Pan (Jun)
2020-09-16 1:46 ` Lu Baolu
2020-09-16 15:07 ` Jason Gunthorpe
2020-09-16 16:33 ` Raj, Ashok
2020-09-16 17:01 ` Jason Gunthorpe
2020-09-16 18:21 ` Jacob Pan (Jun)
2020-09-16 18:38 ` Jason Gunthorpe
2020-09-16 23:09 ` Jacob Pan (Jun)
2020-09-17 3:53 ` Jason Wang [this message]
2020-09-17 17:31 ` Jason Gunthorpe
2020-09-17 18:17 ` Jacob Pan (Jun)
2020-09-18 3:58 ` Jason Wang
2020-09-16 2:29 ` Jason Wang
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:
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
* 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).