[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [PATCH V2] virtio-net: introduce admin control virtqueue
On 2021/1/25 äå8:54, Cornelia Huck wrote:
On Mon, 25 Jan 2021 13:52:02 +0800 Jason Wang <jasowang@redhat.com> wrote:When implementing virtual devices like SR-IOV or sub-function. We're suffering from several issues: - There's no management interface for management device to configure features, attributes for a virtual device - Per virtual device control virtqueue could be very expensive as the number of virtual devices could be very large - Virtualize per virtual device's control virtqueue could be very challenge as we need the support of DMA isolation at queue level So this patch introduces the feature of VIRTIO_NET_CTRL_ADMIN_VQ. This allows the device to implement a single admin control virtqueue to manage the features and attributes for a specific virtual device. The idea is simple, a new virtual device id is introduced on top of the existing virtio_net_ctrl structure. This id is transport or device specific to uniquely identify a management or virtual device. With this, we get a way of using management device (PF) to configure per virtual device features and attributes. And since the admin control virtqueue belongs to management device (PF), the DMA is naturally isolated at device level instead of the queue level for per virtual device control vq. When the admin cvq is offered by management device and normal cvq is offered by virtual device. A new command class is introduced decide whether or not to accept commands from normal cvq for a virtual device. Signed-off-by: Jason Wang <jasowang@redhat.com> --- Changes from V1: - use 'virtual device' instead of 'function' - introuce trust command - clairfy that the admin cvq could be used to configure management device itself --- content.tex | 59 ++++++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 56 insertions(+), 3 deletions(-) diff --git a/content.tex b/content.tex index 620c0e2..989b4f6 100644 --- a/content.tex +++ b/content.tex @@ -2940,6 +2940,9 @@ \subsection{Feature bits}\label{sec:Device Types / Network Device / Feature bits \item[VIRTIO_NET_F_CTRL_MAC_ADDR(23)] Set MAC address through control channel.+\item[VIRTIO_NET_F_ADMIN_CTRL_VQ(56)] Admin control channel is+ available. + \item[VIRTIO_NET_F_HASH_REPORT(57)] Device can report per-packet hash value and a type of calculated hash.@@ -3840,11 +3843,12 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network\subsubsection{Control Virtqueue}\label{sec:Device Types / Network Device / Device Operation / Control Virtqueue}The driver uses the control virtqueue (if VIRTIO_NET_F_CTRL_VQ is-negotiated) to send commands to manipulate various features of -the device which would not easily map into the configuration -space. +negotiated but VIRTIO_NET_F_ADMIN_CTRL_VQ is not negotiated) to send +commands to manipulate various features of the device which would not +easily map into the configuration space.All commands are of the following form:+is not negotiated:Stray hunk.
Will fix.
\begin{lstlisting}struct virtio_net_ctrl { @@ -3864,6 +3868,29 @@ \subsubsection{Control Virtqueue}\label{sec:Device Types / Network Device / Devi do except issue a diagnostic if \field{ack} is not VIRTIO_NET_OK.+When both VIRTIO_NET_F_CTRL_VQ and VIRTIO_NET_F_ADMIN_CTRL_VQ are+negotiated, the driver can use the admin control virtqueue of the +management device to manipulate features of individual virtual devices +where the control virtqueue is not easily implemented. The definition +of management device and virtual device is transport or device +specific. E.g in the case of PCI SR-IOV, the management device is +implemented via the physical function (PF), then the virtual device is +the virtual function (VF) in this case."...if the management device is implemented via the physical function (PF), the virtual device is implemented via the virtual function (VF)." ?
Yes, looks better.
+ +All commands are of the following form: + +\begin{lstlisting} +struct virtio_net_admin_ctrl { + u32 virtual_device_id; + struct virtio_net_ctrl ctrl; +}; +\end{lstlisting} + +The \field{virtual_device_id} is an unique transport or device specific +identifier for a virtual device or management device. E.g for the case +of PCI SR-IOV, it's the PCI function id. Management device MUST"...it is the PCI function id of the virtual function." ? s/Management device/The management device/
Will fix.
+reserve 0 for \field{virtual_device_id} to identify itself."reserve", or rather "use"?
Fine.
That also implies that virtual_device_id must be != 0 for virtual devices, doesn't it?
Yes, that's for simplicity. I'm not sure how useful it we had multiple management device and in that case we need a transport specific way to know the id of management device itself.
+ \paragraph{Packet Receive Filtering}\label{sec:Device Types / Network Device / Device Operation / Control Virtqueue / Packet Receive Filtering} \label{sec:Device Types / Network Device / Device Operation / Control Virtqueue / Setting Promiscuous Mode}%old label for latexdiff@@ -4308,6 +4335,32 @@ \subsubsection{Control Virtqueue}\label{sec:Device Types / Network Device / Deviaccording to the native endian of the guest rather than (necessarily when not using the legacy interface) little-endian.+\paragraph{Setting Trust for Virtual Device}\label{sec:Device Types / Network Device / Device Operation / Control Virtqueue / Setting Trust for Virtual Device}+ +If the VIRTIO_NET_F_ADMIN_CTRL_VQ is negotiated, the management device +can accept operations via the control vq from the trusted virtual"can accept the following operations" ? (in addition to existing control commands?)
Fine.
+device. + +\begin{lstlisting} +#define VIRTIO_NET_ADMIN_CTRL_TRUST 0 + #define VIRTIO_NET_ADMIN_CTRL_TRUST_ENABLE 0 +\end{lstlisting} + +\devicenormative{\subparagraph}{Setting Trust for Virtual Device}{Device Types / Network Device / Device Operation / Control Virtqueue / Setting Trust for Virtual Device} + +If the VIRTIO_NET_F_ADMIN_CTRL_VQ has been negotiated, the device MUSTs/If the/If/+support VIRTIO_NET_ADMIN_CTRL_TRUST classs/support/support the/
Will fix.
+command. VIRTIO_NET_ADMIN_CTRL_TRUST_ENABLE turns trust for the +control virtqueue of a specific virtual device on and off. The command +specific data is one byte containing 0(off) or 1(on). When trust is +off the device MUST fail any operations from the control virtqueue of +the virtual device. The management device MUST NOT trust any virtual +devices after reset.A bit further up, it is stated that the management device can accept operations from a "trusted virtual device". From this statement here, I get that the trust state can be turned on and off by the virtual device.
There's some misunderstanding that needs to be clarified in the next version.
Firstly, for the virtual device, it's not expected to expose an admin cvq and the trust state could not be turned on and off by virtual device then.
The only want to turn on/off is the admin control vq belongs to management device.
Does the management device get to choose if it wants to trust a virtual device, i.e. can it fail enabling the trust?
Yes, the idea is to let management device to choose which virtual device is trust. For the possible failure, I think we should allow the device to fail the enabling, e.g when the device is lacking resources for virtual devices (e.g mac address tables etc).
Thanks
+ +\drivernormative{\subparagraph}{Setting Trust for Virtual Device}{Device Types / Network Device / Device Operation / Control Virtqueue / Setting Trust for Virtual Device} + +A driver SHOULD negotiate VIRTIO_NET_F_ADMIN_CTRL_VQ if the device +offers it.\subsubsection{Legacy Interface: Framing Requirements}\label{sec:DeviceTypes / Network Device / Legacy Interface: Framing Requirements}
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]