[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [virtio-comment] About the plan of Admin Queue
On Wed, 2 Aug 2023 15:01:21 +0900, Yui Washizu <yui.washidu@gmail.com> wrote: > > On 2023/07/27 17:28, Jason Wang wrote: > > On Thu, Jul 27, 2023 at 4:20âPM Xuan Zhuo <xuanzhuo@linux.alibaba.com> wrote: > >> On Thu, 27 Jul 2023 16:03:56 +0800, Jason Wang <jasowang@redhat.com> wrote: > >>> On Thu, Jul 27, 2023 at 2:23âPM Xuan Zhuo <xuanzhuo@linux.alibaba.com> wrote: > >>>> On Thu, 27 Jul 2023 14:17:53 +0800, "Zhu, Lingshan" <lingshan.zhu@intel.com> wrote: > >>>>> > >>>>> On 7/27/2023 2:09 PM, Xuan Zhuo wrote: > >>>>>> On Thu, 27 Jul 2023 11:56:32 +0800, "Zhu, Lingshan" <lingshan.zhu@intel.com> wrote: > >>>>>>> On 7/27/2023 10:30 AM, Xuan Zhuo wrote: > >>>>>>>> On Mon, 3 Jul 2023 12:29:32 +0800, "Zhu, Lingshan" <lingshan.zhu@intel.com> wrote: > >>>>>>>>> On 6/30/2023 7:35 PM, Parav Pandit wrote: > >>>>>>>>>>> From: virtio-comment@lists.oasis-open.org <virtio-comment@lists.oasis- > >>>>>>>>>>> open.org> On Behalf Of Zhu, Lingshan > >>>>>>>>>>> Sent: Friday, June 30, 2023 6:33 AM > >>>>>>>>>>> > >>>>>>>>>>>>>> Can we let the DPU notify the driver to create a new devicer from the > >>>>>>>>>>> backend? > >>>>>>>>>> Yes, why not. > >>>>>>>>>> > >>>>>>>>>>>>>> The key point is who want to create a new device. > >>>>>>>>>>>>> DPU can come with a certain number of pre-created ADIs, just make > >>>>>>>>>>>>> sure the orchestration SW is aware of their device IDs. > >>>>>>>>>>>>> > >>>>>>>>>> Cloud often need these devices to be created dynamically, many a time after the host OS is booted. > >>>>>>>>>> To be more generic, those devices to be created and connected to the host regardless of the life cycle of the host. > >>>>>>>>>> Xuan partly explained it. > >>>>>>>>>> > >>>>>>>>>>>>> If you want the DPU randomly create ADIs and notify the driver, I > >>>>>>>>>>>>> think we need interrupt, e.g., re-use config interrupt. But why DPU > >>>>>>>>>>>>> wants to create and hot plug in a device to a guest? > >>>>>>>>>>>>> Shall the host handle that or DPU pre-create then expose to baremteal > >>>>>>>>>>>>> machines? > >>>>>>>>>>>> In your scenario, the supervisor is on the os, which controls the DPU > >>>>>>>>>>>> to create new devices. > >>>>>>>>>>>> > >>>>>>>>>>>> In the cloud scenario, the vendor manager is in the DPU, and the > >>>>>>>>>>>> entire host is for users. Of course, there are situations where the > >>>>>>>>>>>> vendor manager are in the HOST. But for bare metal machines, the host > >>>>>>>>>>>> belongs to the customer, the vendor manager is only in the DPU. > >>>>>>>>>>>> > >>>>>>>>>>>> So when the customers buy a new nic for the host, the vendor manager > >>>>>>>>>>>> will plug a device to the host from the DPU. > >>>>>>>>>>> I understand once a customer orders a new NIC, you wants to present the NIC > >>>>>>>>>>> to the host. > >>>>>>>>>>> However you only owns the DPU and the customer owns the host, that means > >>>>>>>>>>> this creation and hot plug must be transparent to the host and there may not be > >>>>>>>>>>> a host driver help handling an interrupt/probe. > >>>>>>>>>>> > >>>>>>>>>> That is ok. when driver is loaded, it would query about its child devices and probe it, if we strictly want to follow SIOV model. > >>>>>>>>>> > >>>>>>>>>>> However this is not PCI which has a tree/switch and can enumerate devices to > >>>>>>>>>>> the host by spanning the device across the PCI hierarchy. > >>>>>>>>>>> > >>>>>>>>>> Those enumeration is triggered by the parent PCI device and pci bridge and switch will also discover it. > >>>>>>>>>> > >>>>>>>>>>> To address an ADI, there is only a device_id. > >>>>>>>>>>> > >>>>>>>>>> SIOV device must have a unique identifier at PCI bus level for sure. > >>>>>>>>>> I cannot speak more about it in this forum due to other logistics issue. > >>>>>>>>>> But assume that there is PCI level unique identifier for SIOV device that switches on the path will learn about. > >>>>>>>>>> > >>>>>>>>>>> So, do you mind share how your DPU offload the device model? What kind of > >>>>>>>>>>> device your DPU provide to the host? Lets see whether DPU can mediate this by > >>>>>>>>>>> its own? > >>>>>>>>>>> > >>>>>>>>>> It is a virtio nic, blk and other virtio devices for us. > >>>>>>>>>> A DPU hotplugs a device, host side either gets interrupt or later gets to know about it when explicitly queries. > >>>>>>>>>> There is no mediation per say here, it is just a dpu based SIOV device like a regular PF. > >>>>>>>>>> > >>>>>>>>>> For non virtio DPU device, I implemented them in Linux for dpus 2 years ago. > >>>>>>>>>> You might find a Linux reference model useful at [1]. > >>>>>>>>>> A usage model already exists in one OS and in use for non virtio devices. > >>>>>>>>>> This certainly works without SIOV unique PCI device identifiers, because DPU (non-host) managed SIOV device spec still does not exist. > >>>>>>>>>> > >>>>>>>>>> For virtio, I think we should wait for this piece to be defined and leverage that, instead of virtio tc creating its own. > >>>>>>>>>> > >>>>>>>>>> [1] https://github.com/Mellanox/scalablefunctions/wiki > >>>>>>>>> well I see SF facing the similar challenge, I can add a command for the > >>>>>>>>> driver to query all existing SIOV ADIs of a device, > >>>>>>>>> and the device return ADIs id and status. Looks good? and work for you > >>>>>>>>> @Xuan? > >>>>>>>> Could I have your plan for this? > >>>>>>>> > >>>>>>>> If you do not mind, I'd like to add a command to query VF's info. Such > >>>>>>>> as mac, ip, etc. > >>>>>>> I think the query commands for SIOV is a little more complex, e.g., > >>>>>>> need to report device type and its scale(e.g., features, mq). > >>>>>>> There can be thousands of SIOV ADIs and we don't want output flood. > >>>>>>> > >>>>>>> We have discussed implementation a config interrupt to report new > >>>>>>> created / deleted > >>>>>>> ADIs on the DPU side, therefore there must be a cap contains related > >>>>>>> information, > >>>>>>> my rough approach of the process is: > >>>>>>> 1) a cap contains the total number of existing ADIs and the max dev id > >>>>>>> 2) driver queries detailed information of a certain ADI or a bunch of > >>>>>>> ADIs in a [dev_id....dev_id2] range. > >>>>>> Yes, Admin Queue can obtain the info of the specific one or more devices. > >>>>>> > >>>>>>> I am not sure whether a NIC stores its IP > >>>>>> IP is the other topic. I want the Admin Queue manage the switch. > >>>>>> So the switch know about the IP of every device, and the > >>>>>> Admin Queue will has the ability to config the IP of the device inside the > >>>>>> switch. > >>>>> DPU onboard switch? OVS? Does it beyond virtio spec? > >>>> YES. > >>> Adding Washizu. > >>> > >>> We can have a switch/dpa defined in the networking device for sure. > >> Yes, I think we should introduce that for the sr-iov. Or for other. > > This should be a general one as a switch should be transport independent. > > > >> I would like to know who is doing this? > > Washizu, could you confirm if you want to do this or not? > > > Does this mean adding a switch definition to the virtio spec? > > > If so, it will be necessary for the implementation of my plan, > > but it may take time (probably several months?) to get started, > > as I'm currently working on another task (virtio-net SR-IOV feature in > qemu). > > Anyone is welcome to work on adding the switch definition in the meantime, > > it's completely fine with me. > > I think I'll work on that if no one has finished the work. OK, I got. Because we have a need in this area, I will push the work in this area. Thanks. > > > > > > >> Another question, @Jason are you referring to a new device type or a > >> new virtio-net feature. > > Extending virtio-net should be fine, did you see any issues for this? > > > >>>> For SIOV, I think this is MUST. > >>> A learning bridge would be fine as a starter. It's better not to > >>> couple new scalable capability with any device specific features. > >>> > >>>> Maybe you have one simple implementation. > >>>> But you have to solve the IP steering. So admin queue should has the ability > >>>> to config the IP steering. > >>> I think not. Those L2/LN tables/filters are networking specific. > >> Let us assume that there is a switch/bridge firstly. > >> The VFs may be passed to different VMs. > >> > >> I also think this is the networking specific. But I want to config > >> the ip for every vf from the pf. > > What do you mean by ip here (e.g who is the user for this ip?) > > > >> Because the user of the vf may be unreliable. > >> We need a manager to config the ip for every vf. > > Did you mean you're using a tunnel or not? > > > >> > >>> Control virtqueue is better than admin virtqueue here. > >> by cq? > >> > >> What case? > > We've already used control virtqueue for steering. > > > > Thanks > > > >> Thanks. > >> > >>> Thanks > >>> > >>>> Thanks. > >>>> > >>>>>> > >>>>>> Thanks. > >>>>>> > >>>>>> > >>>>>>> Thanks > >>>>>>>> Thanks. > >>>>>>>> > >>>>>>>> > >>>>>>>> > >>>>>>>>> Thanks > >>>>>>>>> > >>>>>>>> This publicly archived list offers a means to provide input to the > >>>>>>>> OASIS Virtual I/O Device (VIRTIO) TC. > >>>>>>>> > >>>>>>>> In order to verify user consent to the Feedback License terms and > >>>>>>>> to minimize spam in the list archive, subscription is required > >>>>>>>> before posting. > >>>>>>>> > >>>>>>>> Subscribe: virtio-comment-subscribe@lists.oasis-open.org > >>>>>>>> Unsubscribe: virtio-comment-unsubscribe@lists.oasis-open.org > >>>>>>>> List help: virtio-comment-help@lists.oasis-open.org > >>>>>>>> List archive: https://lists.oasis-open.org/archives/virtio-comment/ > >>>>>>>> Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf > >>>>>>>> List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists > >>>>>>>> Committee: https://www.oasis-open.org/committees/virtio/ > >>>>>>>> Join OASIS: https://www.oasis-open.org/join/ > >>>>>>>> > >>>>> > >>>>> This publicly archived list offers a means to provide input to the > >>>>> OASIS Virtual I/O Device (VIRTIO) TC. > >>>>> > >>>>> In order to verify user consent to the Feedback License terms and > >>>>> to minimize spam in the list archive, subscription is required > >>>>> before posting. > >>>>> > >>>>> Subscribe: virtio-comment-subscribe@lists.oasis-open.org > >>>>> Unsubscribe: virtio-comment-unsubscribe@lists.oasis-open.org > >>>>> List help: virtio-comment-help@lists.oasis-open.org > >>>>> List archive: https://lists.oasis-open.org/archives/virtio-comment/ > >>>>> Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf > >>>>> List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists > >>>>> Committee: https://www.oasis-open.org/committees/virtio/ > >>>>> Join OASIS: https://www.oasis-open.org/join/ > >>>>> > >>>> This publicly archived list offers a means to provide input to the > >>>> OASIS Virtual I/O Device (VIRTIO) TC. > >>>> > >>>> In order to verify user consent to the Feedback License terms and > >>>> to minimize spam in the list archive, subscription is required > >>>> before posting. > >>>> > >>>> Subscribe: virtio-comment-subscribe@lists.oasis-open.org > >>>> Unsubscribe: virtio-comment-unsubscribe@lists.oasis-open.org > >>>> List help: virtio-comment-help@lists.oasis-open.org > >>>> List archive: https://lists.oasis-open.org/archives/virtio-comment/ > >>>> Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf > >>>> List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists > >>>> Committee: https://www.oasis-open.org/committees/virtio/ > >>>> Join OASIS: https://www.oasis-open.org/join/ > >>>>
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]