On Thu, Apr 11, 2019 at 07:02:15PM +0200, Cornelia Huck wrote:> On Thu, 11 Apr 2019 14:01:54 +0300 > Yuval Shaia <yuval.shaia at oracle.com> wrote: > > > Data center backends use more and more RDMA or RoCE devices and more and > > more software runs in virtualized environment. > > There is a need for a standard to enable RDMA/RoCE on Virtual Machines. > > > > Virtio is the optimal solution since is the de-facto para-virtualizaton > > technology and also because the Virtio specification > > allows Hardware Vendors to support Virtio protocol natively in order to > > achieve bare metal performance. > > > > This RFC is an effort to addresses challenges in defining the RDMA/RoCE > > Virtio Specification and a look forward on possible implementation > > techniques. > > > > Open issues/Todo list: > > List is huge, this is only start point of the project. > > Anyway, here is one example of item in the list: > > - Multi VirtQ: Every QP has two rings and every CQ has one. This means that > > in order to support for example 32K QPs we will need 64K VirtQ. Not sure > > that this is reasonable so one option is to have one for all and > > multiplex the traffic on it. This is not good approach as by design it > > introducing an optional starvation. Another approach would be multi > > queues and round-robin (for example) between them. > > > > Expectations from this posting: > > In general, any comment is welcome, starting from hey, drop this as it is a > > very bad idea, to yeah, go ahead, we really want it. > > Idea here is that since it is not a minor effort i first want to know if > > there is some sort interest in the community for such device. > > My first reaction is: Sounds sensible, but it would be good to have a > spec for this :)I'm unclear why you'd want to have a virtio queue for anything other that some kind of command channel. I'm not sure a QP or CQ benefits from this?? Jason
On Thu, Apr 11, 2019 at 05:24:08PM +0000, Jason Gunthorpe wrote:> On Thu, Apr 11, 2019 at 07:02:15PM +0200, Cornelia Huck wrote: > > On Thu, 11 Apr 2019 14:01:54 +0300 > > Yuval Shaia <yuval.shaia at oracle.com> wrote: > > > > > Data center backends use more and more RDMA or RoCE devices and more and > > > more software runs in virtualized environment. > > > There is a need for a standard to enable RDMA/RoCE on Virtual Machines. > > > > > > Virtio is the optimal solution since is the de-facto para-virtualizaton > > > technology and also because the Virtio specification > > > allows Hardware Vendors to support Virtio protocol natively in order to > > > achieve bare metal performance. > > > > > > This RFC is an effort to addresses challenges in defining the RDMA/RoCE > > > Virtio Specification and a look forward on possible implementation > > > techniques. > > > > > > Open issues/Todo list: > > > List is huge, this is only start point of the project. > > > Anyway, here is one example of item in the list: > > > - Multi VirtQ: Every QP has two rings and every CQ has one. This means that > > > in order to support for example 32K QPs we will need 64K VirtQ. Not sure > > > that this is reasonable so one option is to have one for all and > > > multiplex the traffic on it. This is not good approach as by design it > > > introducing an optional starvation. Another approach would be multi > > > queues and round-robin (for example) between them. > > > > > > Expectations from this posting: > > > In general, any comment is welcome, starting from hey, drop this as it is a > > > very bad idea, to yeah, go ahead, we really want it. > > > Idea here is that since it is not a minor effort i first want to know if > > > there is some sort interest in the community for such device. > > > > My first reaction is: Sounds sensible, but it would be good to have a > > spec for this :) > > I'm unclear why you'd want to have a virtio queue for anything other > that some kind of command channel. > > I'm not sure a QP or CQ benefits from this??Virtqueue is a standard mechanism to pass data from guest to host. By saying that - it really sounds like QP send and recv rings. So my thought is to use a standard way for rings. As i've learned this is how it is used by other virtio devices ex virtio-net.> > Jason
On Thu, Apr 11, 2019 at 08:34:20PM +0300, Yuval Shaia wrote:> On Thu, Apr 11, 2019 at 05:24:08PM +0000, Jason Gunthorpe wrote: > > On Thu, Apr 11, 2019 at 07:02:15PM +0200, Cornelia Huck wrote: > > > On Thu, 11 Apr 2019 14:01:54 +0300 > > > Yuval Shaia <yuval.shaia at oracle.com> wrote: > > > > > > > Data center backends use more and more RDMA or RoCE devices and more and > > > > more software runs in virtualized environment. > > > > There is a need for a standard to enable RDMA/RoCE on Virtual Machines. > > > > > > > > Virtio is the optimal solution since is the de-facto para-virtualizaton > > > > technology and also because the Virtio specification > > > > allows Hardware Vendors to support Virtio protocol natively in order to > > > > achieve bare metal performance. > > > > > > > > This RFC is an effort to addresses challenges in defining the RDMA/RoCE > > > > Virtio Specification and a look forward on possible implementation > > > > techniques. > > > > > > > > Open issues/Todo list: > > > > List is huge, this is only start point of the project. > > > > Anyway, here is one example of item in the list: > > > > - Multi VirtQ: Every QP has two rings and every CQ has one. This means that > > > > in order to support for example 32K QPs we will need 64K VirtQ. Not sure > > > > that this is reasonable so one option is to have one for all and > > > > multiplex the traffic on it. This is not good approach as by design it > > > > introducing an optional starvation. Another approach would be multi > > > > queues and round-robin (for example) between them. > > > > > > > > Expectations from this posting: > > > > In general, any comment is welcome, starting from hey, drop this as it is a > > > > very bad idea, to yeah, go ahead, we really want it. > > > > Idea here is that since it is not a minor effort i first want to know if > > > > there is some sort interest in the community for such device. > > > > > > My first reaction is: Sounds sensible, but it would be good to have a > > > spec for this :) > > > > I'm unclear why you'd want to have a virtio queue for anything other > > that some kind of command channel. > > > > I'm not sure a QP or CQ benefits from this?? > > Virtqueue is a standard mechanism to pass data from guest to host. By > saying that - it really sounds like QP send and recv rings. So my thought > is to use a standard way for rings. As i've learned this is how it is used > by other virtio devices ex virtio-net.I doubt you can use virtio queues from userspace securely? Usually needs a dedicated page for each user space process Jason
On Thu, Apr 11, 2019 at 08:34:20PM +0300, Yuval Shaia wrote:> On Thu, Apr 11, 2019 at 05:24:08PM +0000, Jason Gunthorpe wrote: > > On Thu, Apr 11, 2019 at 07:02:15PM +0200, Cornelia Huck wrote: > > > On Thu, 11 Apr 2019 14:01:54 +0300 > > > Yuval Shaia <yuval.shaia at oracle.com> wrote: > > > > > > > Data center backends use more and more RDMA or RoCE devices and more and > > > > more software runs in virtualized environment. > > > > There is a need for a standard to enable RDMA/RoCE on Virtual Machines. > > > > > > > > Virtio is the optimal solution since is the de-facto para-virtualizaton > > > > technology and also because the Virtio specification > > > > allows Hardware Vendors to support Virtio protocol natively in order to > > > > achieve bare metal performance. > > > > > > > > This RFC is an effort to addresses challenges in defining the RDMA/RoCE > > > > Virtio Specification and a look forward on possible implementation > > > > techniques. > > > > > > > > Open issues/Todo list: > > > > List is huge, this is only start point of the project. > > > > Anyway, here is one example of item in the list: > > > > - Multi VirtQ: Every QP has two rings and every CQ has one. This means that > > > > in order to support for example 32K QPs we will need 64K VirtQ. Not sure > > > > that this is reasonable so one option is to have one for all and > > > > multiplex the traffic on it. This is not good approach as by design it > > > > introducing an optional starvation. Another approach would be multi > > > > queues and round-robin (for example) between them. > > > > > > > > Expectations from this posting: > > > > In general, any comment is welcome, starting from hey, drop this as it is a > > > > very bad idea, to yeah, go ahead, we really want it. > > > > Idea here is that since it is not a minor effort i first want to know if > > > > there is some sort interest in the community for such device. > > > > > > My first reaction is: Sounds sensible, but it would be good to have a > > > spec for this :) > > > > I'm unclear why you'd want to have a virtio queue for anything other > > that some kind of command channel. > > > > I'm not sure a QP or CQ benefits from this?? > > Virtqueue is a standard mechanism to pass data from guest to host. ByAnd vice versa (CQ?)> saying that - it really sounds like QP send and recv rings. So my thought > is to use a standard way for rings. As i've learned this is how it is used > by other virtio devices ex virtio-net. > > > > > Jason >
On Fri, Apr 12, 2019 at 03:21:56PM +0530, Devesh Sharma wrote:> On Thu, Apr 11, 2019 at 11:11 PM Yuval Shaia <yuval.shaia at oracle.com> wrote: > > > > On Thu, Apr 11, 2019 at 08:34:20PM +0300, Yuval Shaia wrote: > > > On Thu, Apr 11, 2019 at 05:24:08PM +0000, Jason Gunthorpe wrote: > > > > On Thu, Apr 11, 2019 at 07:02:15PM +0200, Cornelia Huck wrote: > > > > > On Thu, 11 Apr 2019 14:01:54 +0300 > > > > > Yuval Shaia <yuval.shaia at oracle.com> wrote: > > > > > > > > > > > Data center backends use more and more RDMA or RoCE devices and more and > > > > > > more software runs in virtualized environment. > > > > > > There is a need for a standard to enable RDMA/RoCE on Virtual Machines. > > > > > > > > > > > > Virtio is the optimal solution since is the de-facto para-virtualizaton > > > > > > technology and also because the Virtio specification > > > > > > allows Hardware Vendors to support Virtio protocol natively in order to > > > > > > achieve bare metal performance. > > > > > > > > > > > > This RFC is an effort to addresses challenges in defining the RDMA/RoCE > > > > > > Virtio Specification and a look forward on possible implementation > > > > > > techniques. > > > > > > > > > > > > Open issues/Todo list: > > > > > > List is huge, this is only start point of the project. > > > > > > Anyway, here is one example of item in the list: > > > > > > - Multi VirtQ: Every QP has two rings and every CQ has one. This means that > > > > > > in order to support for example 32K QPs we will need 64K VirtQ. Not sure > > > > > > that this is reasonable so one option is to have one for all and > > > > > > multiplex the traffic on it. This is not good approach as by design it > > > > > > introducing an optional starvation. Another approach would be multi > > > > > > queues and round-robin (for example) between them. > > > > > > > > > > > > Expectations from this posting: > > > > > > In general, any comment is welcome, starting from hey, drop this as it is a > > > > > > very bad idea, to yeah, go ahead, we really want it. > > > > > > Idea here is that since it is not a minor effort i first want to know if > > > > > > there is some sort interest in the community for such device. > > > > > > > > > > My first reaction is: Sounds sensible, but it would be good to have a > > > > > spec for this :) > > > > > > > > I'm unclear why you'd want to have a virtio queue for anything other > > > > that some kind of command channel. > > > > > > > > I'm not sure a QP or CQ benefits from this?? > > > > > > Virtqueue is a standard mechanism to pass data from guest to host. By > > > > And vice versa (CQ?) > > > > > saying that - it really sounds like QP send and recv rings. So my thought > > > is to use a standard way for rings. As i've learned this is how it is used > > > by other virtio devices ex virtio-net. > > > > > > > > > > > Jason > > > > I would like to ask more basic question, how virtio queue will glue > with actual h/w qps? I may be to naive though.Have to admit - I have no idea. This work is based on emulated device so i'm my case - the emulated device is creating the virtqueue. I guess that HW device will create a QP and expose a virtqueue interface to it. The same driver should serve both the SW and HW devices. One of the objectives of this RFC is to collaborate an effort and implementation notes/ideas from HW vendors.> > -Regards > Devesh