Re: [Qemu-devel] [virtio-dev] Re: Vhost-pci RFC2.0

[Wei Wang <wei.w.wang@intel.com>]

On 04/19/2017 05:31 PM, Jan Kiszka wrote:
> On 2017-04-19 11:09, Wei Wang wrote:
>> On 04/19/2017 04:49 PM, Jan Kiszka wrote:
>>> On 2017-04-19 10:42, Wei Wang wrote:
>>>> On 04/19/2017 03:35 PM, Jan Kiszka wrote:
>>>>> On 2017-04-19 08:38, Wang, Wei W wrote:
>>>>>> Hi,
>>>>>>     We made some design changes to the original vhost-pci design,
>>>>>> and want
>>>>>> to open
>>>>>> a discussion about the latest design (labelled 2.0) and its extension
>>>>>> (2.1).
>>>>>> 2.0 design: One VM shares the entire memory of another VM
>>>>>> 2.1 design: One VM uses an intermediate memory shared with another VM
>>>>>> for
>>>>>>                         packet transmission.
>>>>>>     For the convenience of discussion, I have some pictures
>>>>>> presented at
>>>>>> this link:
>>>>>> _https://github.com/wei-w-wang/vhost-pci-discussion/blob/master/vhost-pci-rfc2.0.pdf_
>>>>>>
>>>>>>
>>>>>>     Fig. 1 shows the common driver frame that we want use to build
>>>>>> the 2.0
>>>>>> and 2.1
>>>>>> design. A TX/RX engine consists of a local ring and an exotic ring.
>>>>>> Local ring:
>>>>>> 1) allocated by the driver itself;
>>>>>> 2) registered with the device (i.e. virtio_add_queue())
>>>>>> Exotic ring:
>>>>>> 1) ring memory comes from the outside (of the driver), and exposed to
>>>>>> the driver
>>>>>>         via a BAR MMIO;
>>>>> Small additional requirement: In order to make this usable with
>>>>> Jailhouse as well, we need [also] a side-channel configuration for the
>>>>> regions, i.e. likely via a PCI capability. There are too few BARs, and
>>>>> they suggest relocatablity, which is not available under Jailhouse for
>>>>> simplicity reasons (IOW, the shared regions are statically mapped by
>>>>> the
>>>>> hypervisor into the affected guest address spaces).
>>>> What kind of configuration would you need for the regions?
>>>> I think adding a PCI capability should be easy.
>>> Basically address and size, see
>>> https://github.com/siemens/jailhouse/blob/wip/ivshmem2/Documentation/ivshmem-v2-specification.md#vendor-specific-capability-id-09h
>>>
>> Got it, thanks. That should be easy to add to 2.1.
>>
>>>>>> 2) does not have a registration in the device, so no ioeventfd/irqfd,
>>>>>> configuration
>>>>>> registers allocated in the device
>>>>>>     Fig. 2 shows how the driver frame is used to build the 2.0 design.
>>>>>> 1) Asymmetric: vhost-pci-net <-> virtio-net
>>>>>> 2) VM1 shares the entire memory of VM2, and the exotic rings are the
>>>>>> rings
>>>>>>        from VM2.
>>>>>> 3) Performance (in terms of copies between VMs):
>>>>>>        TX: 0-copy (packets are put to VM2’s RX ring directly)
>>>>>>        RX: 1-copy (the green arrow line in the VM1’s RX engine)
>>>>>>     Fig. 3 shows how the driver frame is used to build the 2.1 design.
>>>>>> 1) Symmetric: vhost-pci-net <-> vhost-pci-net
>>>>> This is interesting!
>>>>>
>>>>>> 2) Share an intermediate memory, allocated by VM1’s vhost-pci device,
>>>>>> for data exchange, and the exotic rings are built on the shared memory
>>>>>> 3) Performance:
>>>>>>        TX: 1-copy
>>>>>> RX: 1-copy
>>>>> I'm not yet sure I to this right: there are two different MMIO regions
>>>>> involved, right? One is used for VM1's RX / VM2's TX, and the other for
>>>>> the reverse path? Would allow our requirement to have those regions
>>>>> mapped with asymmetric permissions (RX read-only, TX read/write).
>>>> The design presented here intends to use only one BAR to expose
>>>> both TX and RX. The two VMs share an intermediate memory
>>>> here, why couldn't we give the same permission to TX and RX?
>>>>
>>> For security and/or safety reasons: the TX side can then safely prepare
>>> and sign a message in-place because the RX side cannot mess around with
>>> it while not yet being signed (or check-summed). Saves one copy from a
>>> secure place into the shared memory.
>> If we allow guest1 to write to RX, what safety issue would it cause to
>> guest2?
> This way, guest1 could trick guest2, in a race condition, to sign a
> modified message instead of the original one.
>
Just align the context that we are talking about: RX is the intermediate
shared ring that guest1 uses to receive packets and guest2 uses to send
packet.

Seems the issue is that guest1 will receive a hacked message from RX
(modified by itself). How would it affect guest2?

>>
>>>>>>     Fig. 4 shows the inter-VM notification path for 2.0 (2.1 is
>>>>>> similar).
>>>>>> The four eventfds are allocated by virtio-net, and shared with
>>>>>> vhost-pci-net:
>>>>>> Uses virtio-net’s TX/RX kickfd as the vhost-pci-net’s RX/TX callfd
>>>>>> Uses virtio-net’s TX/RX callfd as the vhost-pci-net’s RX/TX kickfd
>>>>>> Example of how it works:
>>>>>> After packets are put into vhost-pci-net’s TX, the driver kicks TX,
>>>>>> which
>>>>>> causes the an interrupt associated with fd3 to be injected to
>>>>>> virtio-net
>>>>>>     The draft code of the 2.0 design is ready, and can be found here:
>>>>>> Qemu: _https://github.com/wei-w-wang/vhost-pci-device_
>>>>>> Guest driver: _https://github.com/wei-w-wang/vhost-pci-driver_
>>>>>>     We tested the 2.0 implementation using the Spirent packet
>>>>>> generator to transmit 64B packets, the results show that the
>>>>>> throughput of vhost-pci reaches around 1.8Mpps, which is around
>>>>>> two times larger than the legacy OVS+DPDK. Also, vhost-pci shows
>>>>>> better scalability than OVS+DPDK.
>>>>>>     
>>>>> Do you have numbers for the symmetric 2.1 case as well? Or is the
>>>>> driver
>>>>> not yet ready for that yet? Otherwise, I could try to make it work over
>>>>> a simplistic vhost-pci 2.1 version in Jailhouse as well. That would
>>>>> give
>>>>> a better picture of how much additional complexity this would mean
>>>>> compared to our ivshmem 2.0.
>>>>>
>>>> Implementation of 2.1 is not ready yet. We can extend it to 2.1 after
>>>> the common driver frame is reviewed.
>>> Can you you assess the needed effort?
>>>
>>> For us, this is a critical feature, because we need to decide if
>>> vhost-pci can be an option at all. In fact, the "exotic ring" will be
>>> the only way to provide secure inter-partition communication on
>>> Jailhouse.
>>>
>> If what is here for 2.0 is suitable to be upstream-ed, I think it will
>> be easy
>> to extend it to 2.1 (probably within 1 month).
> Unfortunate ordering here, though. Specifically if we need to modify
> existing things instead of just adding something. We will need 2.1 prior
> to committing to 2.0 being the right thing.
>

If you want, we can get the common part of design ready first,
then we can start to build on the common part at the same time.
The draft code of 2.0 is ready. I can clean it up, making it easier for
us to continue and change.

Best,
Wei