From: "Michael S. Tsirkin" <mst@redhat.com>
To: Wei Wang <wei.w.wang@intel.com>
Cc: virtio-comment@lists.oasis-open.org, qemu-devel@nongnu.org,
marcandre.lureau@gmail.com, stefanha@redhat.com,
pbonzini@redhat.com
Subject: Re: [Qemu-devel] [PATCH v1] docs/vhost-user: extend the vhost-user protocol to support the vhost-pci based inter-vm communication
Date: Sun, 30 Oct 2016 20:02:43 +0200 [thread overview]
Message-ID: <20161030200125-mutt-send-email-mst@kernel.org> (raw)
In-Reply-To: <1477293002-144213-1-git-send-email-wei.w.wang@intel.com>
On Mon, Oct 24, 2016 at 03:10:02PM +0800, Wei Wang wrote:
> Signed-off-by: Wei Wang <wei.w.wang@intel.com>
As the implementation is not going to be merged
for this QEMU release, I think we shouldn't
drop the doc either.
I'll review and respond a bit later so we can finalize
the protocol meanwhile.
> ---
> docs/specs/vhost-user.txt | 81 +++++++++++++++++++++++++++++++++++++++++------
> 1 file changed, 72 insertions(+), 9 deletions(-)
>
> diff --git a/docs/specs/vhost-user.txt b/docs/specs/vhost-user.txt
> index 7890d71..173f693 100644
> --- a/docs/specs/vhost-user.txt
> +++ b/docs/specs/vhost-user.txt
> @@ -17,28 +17,37 @@ The protocol defines 2 sides of the communication, master and slave. Master is
> the application that shares its virtqueues, in our case QEMU. Slave is the
> consumer of the virtqueues.
>
> -In the current implementation QEMU is the Master, and the Slave is intended to
> +In the traditional implementation QEMU is the Master, and the Slave is intended to
> be a software Ethernet switch running in user space, such as Snabbswitch.
>
> Master and slave can be either a client (i.e. connecting) or server (listening)
> in the socket communication.
>
> +The current vhost-user protocol is extended to support the vhost-pci based inter-VM
> +communication. In this case, Slave is a QEMU which runs a vhost-pci server, and
> +Master is another QEMU which runs a vhost-pci client.
> +
> Message Specification
> ---------------------
>
> Note that all numbers are in the machine native byte order. A vhost-user message
> -consists of 3 header fields and a payload:
> +consists of 4 header fields and a payload:
>
> -------------------------------------
> -| request | flags | size | payload |
> -------------------------------------
> +----------------------------------------------
> +| request | flags | conn_id | size | payload |
> +----------------------------------------------
>
> * Request: 32-bit type of the request
> * Flags: 32-bit bit field:
> - Lower 2 bits are the version (currently 0x01)
> - - Bit 2 is the reply flag - needs to be sent on each reply from the slave
> + - Bit 2 is the reply flag - needs to be sent on each reply
> - Bit 3 is the need_reply flag - see VHOST_USER_PROTOCOL_F_REPLY_ACK for
> details.
> + * Conn_id: 64-bit connection id to indentify a client socket connection. It is
> + introduced in version 0x02 to support the "1-server-N-client" model
> + and an asynchronous client read implementation. The connection id,
> + 0xFFFFFFFFFFFFFFFF, is used by an anonymous client (e.g. a client who
> + has not got its connection id from the server in the initial talk)
> * Size - 32-bit size of the payload
>
>
> @@ -97,6 +106,13 @@ Depending on the request type, payload can be:
> log offset: offset from start of supplied file descriptor
> where logging starts (i.e. where guest address 0 would be logged)
>
> +* Device info
> + --------------------
> + | virito id | uuid |
> + --------------------
> + Virtio id: 16-bit virtio id of the device
> + UUID: 128-bit UUID to identify the QEMU instance that creates the device
> +
> In QEMU the vhost-user message is implemented with the following struct:
>
> typedef struct VhostUserMsg {
> @@ -109,6 +125,7 @@ typedef struct VhostUserMsg {
> struct vhost_vring_addr addr;
> VhostUserMemory memory;
> VhostUserLog log;
> + DeviceInfo dev_info;
> };
> } QEMU_PACKED VhostUserMsg;
>
> @@ -119,17 +136,25 @@ The protocol for vhost-user is based on the existing implementation of vhost
> for the Linux Kernel. Most messages that can be sent via the Unix domain socket
> implementing vhost-user have an equivalent ioctl to the kernel implementation.
>
> -The communication consists of master sending message requests and slave sending
> -message replies. Most of the requests don't require replies. Here is a list of
> -the ones that do:
> +Traditionally, the communication consists of master sending message requests
> +and slave sending message replies. Most of the requests don't require replies.
> +Here is a list of the ones that do:
>
> * VHOST_GET_FEATURES
> * VHOST_GET_PROTOCOL_FEATURES
> * VHOST_GET_VRING_BASE
> * VHOST_SET_LOG_BASE (if VHOST_USER_PROTOCOL_F_LOG_SHMFD)
> + * VHOST_USER_GET_CONN_ID
> + * VHOST_USER_SET_PEER_CONNECTION
>
> [ Also see the section on REPLY_ACK protocol extension. ]
>
> +Currently, the communication also supports the Slave (server) sending messages
> +to the Master (client). Here is a list of them:
> + * VHOST_USER_SET_FEATURES
> + * VHOST_USER_SET_PEER_CONNECTION (the serve may actively request to disconnect
> + with the client)
> +
> There are several messages that the master sends with file descriptors passed
> in the ancillary data:
>
> @@ -259,6 +284,7 @@ Protocol features
> #define VHOST_USER_PROTOCOL_F_LOG_SHMFD 1
> #define VHOST_USER_PROTOCOL_F_RARP 2
> #define VHOST_USER_PROTOCOL_F_REPLY_ACK 3
> +#define VHOST_USER_PROTOCOL_F_VHOST_PCI 4
>
> Message types
> -------------
> @@ -470,6 +496,43 @@ Message types
> The first 6 bytes of the payload contain the mac address of the guest to
> allow the vhost user backend to construct and broadcast the fake RARP.
>
> + * VHOST_USER_GET_CONN_ID
> +
> + Id: 20
> + Equivalent ioctl: N/A
> + Master payload: u64
> +
> + The client sends this message to the server to ask for its connection id.
> + The connection id is then put into the message header (the conn_id field),
> + so that the server can always know who it is talking with.
> +
> +* VHOST_USER_SET_DEV_INFO
> +
> + Id: 21
> + Equivalent ioctl: N/A
> + Master payload: dev info
> +
> + The client sends the producer device info to the server.
> + This request should be sent only when VHOST_USER_PROTOCOL_F_VHOST_PCI has
> + been negotiated.
> +
> +* VHOST_USER_SET_PEER_CONNECTION
> +
> + Id: 22
> + Equivalent ioctl: N/A
> + Master payload: u64
> +
> + The producer device requests to connect or disconnect to the consumer device.
> + The consumer device may request to disconnect to the producer device. This
> + request should be sent only when VHOST_USER_PROTOCOL_F_VHOST_PCI has been
> + negotiated.
> + Connection request: If the reply message indicates "success", the vhost-pci based
> + inter-VM communication channel has been established.
> + Disconnection request: If the reply message indicates "success", the vhost-pci based
> + inter-VM communication channel has been destroyed.
> + #define VHOST_USER_SET_PEER_CONNECTION_F_OFF 0
> + #define VHOST_USER_SET_PEER_CONNECTION_F_ON 1
> +
> VHOST_USER_PROTOCOL_F_REPLY_ACK:
> -------------------------------
> The original vhost-user specification only demands replies for certain
> --
> 1.9.1
next prev parent reply other threads:[~2016-10-30 18:02 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-10-24 7:10 [Qemu-devel] [PATCH v1] docs/vhost-user: extend the vhost-user protocol to support the vhost-pci based inter-vm communication Wei Wang
2016-10-30 18:02 ` Michael S. Tsirkin [this message]
2016-11-07 5:43 ` Wang, Wei W
2016-11-08 7:47 ` Marc-André Lureau
2016-11-08 11:35 ` [Qemu-devel] [virtio-comment] " Wei Wang
2016-11-08 11:59 ` Wang, Wei W
2016-11-08 12:17 ` Marc-André Lureau
2016-11-09 8:32 ` Wei Wang
2016-11-10 12:26 ` Marc-André Lureau
2016-11-11 8:24 ` Wei Wang
2016-11-11 17:12 ` Michael S. Tsirkin
Reply instructions:
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:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20161030200125-mutt-send-email-mst@kernel.org \
--to=mst@redhat.com \
--cc=marcandre.lureau@gmail.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@redhat.com \
--cc=virtio-comment@lists.oasis-open.org \
--cc=wei.w.wang@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* 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).