qemu/docs/interop/vhost-user.rst

1565 lines
53 KiB
ReStructuredText
Raw Normal View History

===================
Vhost-user Protocol
===================
:Copyright: 2014 Virtual Open Systems Sarl.
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
:Copyright: 2019 Intel Corporation
:Licence: This work is licensed under the terms of the GNU GPL,
version 2 or later. See the COPYING file in the top-level
directory.
.. contents:: Table of Contents
Introduction
============
This protocol is aiming to complement the ``ioctl`` interface used to
control the vhost implementation in the Linux kernel. It implements
the control plane needed to establish virtqueue sharing with a user
space process on the same host. It uses communication over a Unix
domain socket to share file descriptors in the ancillary data of the
message.
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
the external process consuming the virtio queues, for example a
software Ethernet switch running in user space, such as Snabbswitch,
or a block device backend processing read & write to a virtual
disk. In order to facilitate interoperability between various backend
implementations, it is recommended to follow the :ref:`Backend program
conventions <backend_conventions>`.
*Master* and *slave* can be either a client (i.e. connecting) or
server (listening) in the socket communication.
Message Specification
=====================
.. Note:: All numbers are in the machine native byte order.
A vhost-user message consists of 3 header fields and a payload.
+---------+-------+------+---------+
| request | flags | size | payload |
+---------+-------+------+---------+
Header
------
: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 3 is the need_reply flag - see :ref:`REPLY_ACK <reply_ack>` for
details.
:size: 32-bit size of the payload
Payload
-------
Depending on the request type, **payload** can be:
A single 64-bit integer
^^^^^^^^^^^^^^^^^^^^^^^
+-----+
| u64 |
+-----+
:u64: a 64-bit unsigned integer
A vring state description
^^^^^^^^^^^^^^^^^^^^^^^^^
+-------+-----+
| index | num |
+-------+-----+
:index: a 32-bit index
:num: a 32-bit number
A vring address description
^^^^^^^^^^^^^^^^^^^^^^^^^^^
+-------+-------+------+------------+------+-----------+-----+
| index | flags | size | descriptor | used | available | log |
+-------+-------+------+------------+------+-----------+-----+
:index: a 32-bit vring index
:flags: a 32-bit vring flags
:descriptor: a 64-bit ring address of the vring descriptor table
:used: a 64-bit ring address of the vring used ring
:available: a 64-bit ring address of the vring available ring
:log: a 64-bit guest address for logging
Note that a ring address is an IOVA if ``VIRTIO_F_IOMMU_PLATFORM`` has
been negotiated. Otherwise it is a user address.
Memory regions description
^^^^^^^^^^^^^^^^^^^^^^^^^^
+-------------+---------+---------+-----+---------+
| num regions | padding | region0 | ... | region7 |
+-------------+---------+---------+-----+---------+
:num regions: a 32-bit number of regions
:padding: 32-bit
A region is:
+---------------+------+--------------+-------------+
| guest address | size | user address | mmap offset |
+---------------+------+--------------+-------------+
:guest address: a 64-bit guest address of the region
:size: a 64-bit size
:user address: a 64-bit user address
:mmap offset: 64-bit offset where region starts in the mapped memory
Log description
^^^^^^^^^^^^^^^
+----------+------------+
| log size | log offset |
+----------+------------+
:log size: size of area used for logging
:log offset: offset from start of supplied file descriptor where
logging starts (i.e. where guest address 0 would be
logged)
An IOTLB message
^^^^^^^^^^^^^^^^
+------+------+--------------+-------------------+------+
| iova | size | user address | permissions flags | type |
+------+------+--------------+-------------------+------+
:iova: a 64-bit I/O virtual address programmed by the guest
:size: a 64-bit size
:user address: a 64-bit user address
:permissions flags: an 8-bit value:
- 0: No access
- 1: Read access
- 2: Write access
- 3: Read/Write access
:type: an 8-bit IOTLB message type:
- 1: IOTLB miss
- 2: IOTLB update
- 3: IOTLB invalidate
- 4: IOTLB access fail
Virtio device config space
^^^^^^^^^^^^^^^^^^^^^^^^^^
+--------+------+-------+---------+
| offset | size | flags | payload |
+--------+------+-------+---------+
:offset: a 32-bit offset of virtio device's configuration space
:size: a 32-bit configuration space access size in bytes
:flags: a 32-bit value:
- 0: Vhost master messages used for writeable fields
- 1: Vhost master messages used for live migration
:payload: Size bytes array holding the contents of the virtio
device's configuration space
Vring area description
^^^^^^^^^^^^^^^^^^^^^^
+-----+------+--------+
| u64 | size | offset |
+-----+------+--------+
:u64: a 64-bit integer contains vring index and flags
:size: a 64-bit size of this area
:offset: a 64-bit offset of this area from the start of the
supplied file descriptor
Inflight description
^^^^^^^^^^^^^^^^^^^^
+-----------+-------------+------------+------------+
| mmap size | mmap offset | num queues | queue size |
+-----------+-------------+------------+------------+
:mmap size: a 64-bit size of area to track inflight I/O
:mmap offset: a 64-bit offset of this area from the start
of the supplied file descriptor
:num queues: a 16-bit number of virtqueues
:queue size: a 16-bit size of virtqueues
C structure
-----------
In QEMU the vhost-user message is implemented with the following struct:
.. code:: c
typedef struct VhostUserMsg {
VhostUserRequest request;
uint32_t flags;
uint32_t size;
union {
uint64_t u64;
struct vhost_vring_state state;
struct vhost_vring_addr addr;
VhostUserMemory memory;
VhostUserLog log;
struct vhost_iotlb_msg iotlb;
VhostUserConfig config;
VhostUserVringArea area;
VhostUserInflight inflight;
};
} QEMU_PACKED VhostUserMsg;
Communication
=============
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:
* ``VHOST_USER_GET_FEATURES``
* ``VHOST_USER_GET_PROTOCOL_FEATURES``
* ``VHOST_USER_GET_VRING_BASE``
* ``VHOST_USER_SET_LOG_BASE`` (if ``VHOST_USER_PROTOCOL_F_LOG_SHMFD``)
* ``VHOST_USER_GET_INFLIGHT_FD`` (if ``VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD``)
.. seealso::
:ref:`REPLY_ACK <reply_ack>`
The section on ``REPLY_ACK`` protocol extension.
There are several messages that the master sends with file descriptors passed
in the ancillary data:
* ``VHOST_USER_SET_MEM_TABLE``
* ``VHOST_USER_SET_LOG_BASE`` (if ``VHOST_USER_PROTOCOL_F_LOG_SHMFD``)
* ``VHOST_USER_SET_LOG_FD``
* ``VHOST_USER_SET_VRING_KICK``
* ``VHOST_USER_SET_VRING_CALL``
* ``VHOST_USER_SET_VRING_ERR``
* ``VHOST_USER_SET_SLAVE_REQ_FD``
* ``VHOST_USER_SET_INFLIGHT_FD`` (if ``VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD``)
If *master* is unable to send the full message or receives a wrong
reply it will close the connection. An optional reconnection mechanism
can be implemented.
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
If *slave* detects some error such as incompatible features, it may also
close the connection. This should only happen in exceptional circumstances.
Any protocol extensions are gated by protocol feature bits, which
allows full backwards compatibility on both master and slave. As
older slaves don't support negotiating protocol features, a feature
bit was dedicated for this purpose::
#define VHOST_USER_F_PROTOCOL_FEATURES 30
Starting and stopping rings
---------------------------
Client must only process each ring when it is started.
Client must only pass data between the ring and the backend, when the
ring is enabled.
If ring is started but disabled, client must process the ring without
talking to the backend.
For example, for a networking device, in the disabled state client
must not supply any new RX packets, but must process and discard any
TX packets.
If ``VHOST_USER_F_PROTOCOL_FEATURES`` has not been negotiated, the
ring is initialized in an enabled state.
If ``VHOST_USER_F_PROTOCOL_FEATURES`` has been negotiated, the ring is
initialized in a disabled state. Client must not pass data to/from the
backend until ring is enabled by ``VHOST_USER_SET_VRING_ENABLE`` with
parameter 1, or after it has been disabled by
``VHOST_USER_SET_VRING_ENABLE`` with parameter 0.
Each ring is initialized in a stopped state, client must not process
it until ring is started, or after it has been stopped.
Client must start ring upon receiving a kick (that is, detecting that
file descriptor is readable) on the descriptor specified by
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
``VHOST_USER_SET_VRING_KICK`` or receiving the in-band message
``VHOST_USER_VRING_KICK`` if negotiated, and stop ring upon receiving
``VHOST_USER_GET_VRING_BASE``.
While processing the rings (whether they are enabled or not), client
must support changing some configuration aspects on the fly.
Multiple queue support
----------------------
Many devices have a fixed number of virtqueues. In this case the master
already knows the number of available virtqueues without communicating with the
slave.
Some devices do not have a fixed number of virtqueues. Instead the maximum
number of virtqueues is chosen by the slave. The number can depend on host
resource availability or slave implementation details. Such devices are called
multiple queue devices.
Multiple queue support allows the slave to advertise the maximum number of
queues. This is treated as a protocol extension, hence the slave has to
implement protocol features first. The multiple queues feature is supported
only when the protocol feature ``VHOST_USER_PROTOCOL_F_MQ`` (bit 0) is set.
The max number of queues the slave supports can be queried with message
``VHOST_USER_GET_QUEUE_NUM``. Master should stop when the number of requested
queues is bigger than that.
As all queues share one connection, the master uses a unique index for each
queue in the sent message to identify a specified queue.
The master enables queues by sending message ``VHOST_USER_SET_VRING_ENABLE``.
vhost-user-net has historically automatically enabled the first queue pair.
Slaves should always implement the ``VHOST_USER_PROTOCOL_F_MQ`` protocol
feature, even for devices with a fixed number of virtqueues, since it is simple
to implement and offers a degree of introspection.
Masters must not rely on the ``VHOST_USER_PROTOCOL_F_MQ`` protocol feature for
devices with a fixed number of virtqueues. Only true multiqueue devices
require this protocol feature.
Migration
---------
During live migration, the master may need to track the modifications
the slave makes to the memory mapped regions. The client should mark
the dirty pages in a log. Once it complies to this logging, it may
declare the ``VHOST_F_LOG_ALL`` vhost feature.
To start/stop logging of data/used ring writes, server may send
messages ``VHOST_USER_SET_FEATURES`` with ``VHOST_F_LOG_ALL`` and
``VHOST_USER_SET_VRING_ADDR`` with ``VHOST_VRING_F_LOG`` in ring's
flags set to 1/0, respectively.
All the modifications to memory pointed by vring "descriptor" should
be marked. Modifications to "used" vring should be marked if
``VHOST_VRING_F_LOG`` is part of ring's flags.
Dirty pages are of size::
#define VHOST_LOG_PAGE 0x1000
The log memory fd is provided in the ancillary data of
``VHOST_USER_SET_LOG_BASE`` message when the slave has
``VHOST_USER_PROTOCOL_F_LOG_SHMFD`` protocol feature.
The size of the log is supplied as part of ``VhostUserMsg`` which
should be large enough to cover all known guest addresses. Log starts
at the supplied offset in the supplied file descriptor. The log
covers from address 0 to the maximum of guest regions. In pseudo-code,
to mark page at ``addr`` as dirty::
page = addr / VHOST_LOG_PAGE
log[page / 8] |= 1 << page % 8
Where ``addr`` is the guest physical address.
Use atomic operations, as the log may be concurrently manipulated.
Note that when logging modifications to the used ring (when
``VHOST_VRING_F_LOG`` is set for this ring), ``log_guest_addr`` should
be used to calculate the log offset: the write to first byte of the
used ring is logged at this offset from log start. Also note that this
value might be outside the legal guest physical address range
(i.e. does not have to be covered by the ``VhostUserMemory`` table), but
the bit offset of the last byte of the ring must fall within the size
supplied by ``VhostUserLog``.
``VHOST_USER_SET_LOG_FD`` is an optional message with an eventfd in
ancillary data, it may be used to inform the master that the log has
been modified.
Once the source has finished migration, rings will be stopped by the
source. No further update must be done before rings are restarted.
In postcopy migration the slave is started before all the memory has
been received from the source host, and care must be taken to avoid
accessing pages that have yet to be received. The slave opens a
'userfault'-fd and registers the memory with it; this fd is then
passed back over to the master. The master services requests on the
userfaultfd for pages that are accessed and when the page is available
it performs WAKE ioctl's on the userfaultfd to wake the stalled
slave. The client indicates support for this via the
``VHOST_USER_PROTOCOL_F_PAGEFAULT`` feature.
Memory access
-------------
The master sends a list of vhost memory regions to the slave using the
``VHOST_USER_SET_MEM_TABLE`` message. Each region has two base
addresses: a guest address and a user address.
Messages contain guest addresses and/or user addresses to reference locations
within the shared memory. The mapping of these addresses works as follows.
User addresses map to the vhost memory region containing that user address.
When the ``VIRTIO_F_IOMMU_PLATFORM`` feature has not been negotiated:
* Guest addresses map to the vhost memory region containing that guest
address.
When the ``VIRTIO_F_IOMMU_PLATFORM`` feature has been negotiated:
* Guest addresses are also called I/O virtual addresses (IOVAs). They are
translated to user addresses via the IOTLB.
* The vhost memory region guest address is not used.
IOMMU support
-------------
When the ``VIRTIO_F_IOMMU_PLATFORM`` feature has been negotiated, the
master sends IOTLB entries update & invalidation by sending
``VHOST_USER_IOTLB_MSG`` requests to the slave with a ``struct
vhost_iotlb_msg`` as payload. For update events, the ``iotlb`` payload
has to be filled with the update message type (2), the I/O virtual
address, the size, the user virtual address, and the permissions
flags. Addresses and size must be within vhost memory regions set via
the ``VHOST_USER_SET_MEM_TABLE`` request. For invalidation events, the
``iotlb`` payload has to be filled with the invalidation message type
(3), the I/O virtual address and the size. On success, the slave is
expected to reply with a zero payload, non-zero otherwise.
The slave relies on the slave communcation channel (see :ref:`Slave
communication <slave_communication>` section below) to send IOTLB miss
and access failure events, by sending ``VHOST_USER_SLAVE_IOTLB_MSG``
requests to the master with a ``struct vhost_iotlb_msg`` as
payload. For miss events, the iotlb payload has to be filled with the
miss message type (1), the I/O virtual address and the permissions
flags. For access failure event, the iotlb payload has to be filled
with the access failure message type (4), the I/O virtual address and
the permissions flags. For synchronization purpose, the slave may
rely on the reply-ack feature, so the master may send a reply when
operation is completed if the reply-ack feature is negotiated and
slaves requests a reply. For miss events, completed operation means
either master sent an update message containing the IOTLB entry
containing requested address and permission, or master sent nothing if
the IOTLB miss message is invalid (invalid IOVA or permission).
The master isn't expected to take the initiative to send IOTLB update
messages, as the slave sends IOTLB miss messages for the guest virtual
memory areas it needs to access.
.. _slave_communication:
Slave communication
-------------------
An optional communication channel is provided if the slave declares
``VHOST_USER_PROTOCOL_F_SLAVE_REQ`` protocol feature, to allow the
slave to make requests to the master.
The fd is provided via ``VHOST_USER_SET_SLAVE_REQ_FD`` ancillary data.
A slave may then send ``VHOST_USER_SLAVE_*`` messages to the master
using this fd communication channel.
If ``VHOST_USER_PROTOCOL_F_SLAVE_SEND_FD`` protocol feature is
negotiated, slave can send file descriptors (at most 8 descriptors in
each message) to master via ancillary data using this fd communication
channel.
Inflight I/O tracking
---------------------
To support reconnecting after restart or crash, slave may need to
resubmit inflight I/Os. If virtqueue is processed in order, we can
easily achieve that by getting the inflight descriptors from
descriptor table (split virtqueue) or descriptor ring (packed
virtqueue). However, it can't work when we process descriptors
out-of-order because some entries which store the information of
inflight descriptors in available ring (split virtqueue) or descriptor
ring (packed virtqueue) might be overrided by new entries. To solve
this problem, slave need to allocate an extra buffer to store this
information of inflight descriptors and share it with master for
persistent. ``VHOST_USER_GET_INFLIGHT_FD`` and
``VHOST_USER_SET_INFLIGHT_FD`` are used to transfer this buffer
between master and slave. And the format of this buffer is described
below:
+---------------+---------------+-----+---------------+
| queue0 region | queue1 region | ... | queueN region |
+---------------+---------------+-----+---------------+
N is the number of available virtqueues. Slave could get it from num
queues field of ``VhostUserInflight``.
For split virtqueue, queue region can be implemented as:
.. code:: c
typedef struct DescStateSplit {
/* Indicate whether this descriptor is inflight or not.
* Only available for head-descriptor. */
uint8_t inflight;
/* Padding */
uint8_t padding[5];
/* Maintain a list for the last batch of used descriptors.
* Only available when batching is used for submitting */
uint16_t next;
/* Used to preserve the order of fetching available descriptors.
* Only available for head-descriptor. */
uint64_t counter;
} DescStateSplit;
typedef struct QueueRegionSplit {
/* The feature flags of this region. Now it's initialized to 0. */
uint64_t features;
/* The version of this region. It's 1 currently.
* Zero value indicates an uninitialized buffer */
uint16_t version;
/* The size of DescStateSplit array. It's equal to the virtqueue
* size. Slave could get it from queue size field of VhostUserInflight. */
uint16_t desc_num;
/* The head of list that track the last batch of used descriptors. */
uint16_t last_batch_head;
/* Store the idx value of used ring */
uint16_t used_idx;
/* Used to track the state of each descriptor in descriptor table */
misc: Replace zero-length arrays with flexible array member (manual) Description copied from Linux kernel commit from Gustavo A. R. Silva (see [3]): --v-- description start --v-- The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member [1], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being unadvertenly introduced [2] to the Linux codebase from now on. --^-- description end --^-- Do the similar housekeeping in the QEMU codebase (which uses C99 since commit 7be41675f7cb). All these instances of code were found with the help of the following command (then manual analysis, without modifying structures only having a single flexible array member, such QEDTable in block/qed.h): git grep -F '[0];' [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=76497732932f [3] https://git.kernel.org/pub/scm/linux/kernel/git/gustavoars/linux.git/commit/?id=17642a2fbd2c1 Inspired-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Reviewed-by: David Hildenbrand <david@redhat.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2020-03-04 18:38:16 +03:00
DescStateSplit desc[];
} QueueRegionSplit;
To track inflight I/O, the queue region should be processed as follows:
When receiving available buffers from the driver:
#. Get the next available head-descriptor index from available ring, ``i``
#. Set ``desc[i].counter`` to the value of global counter
#. Increase global counter by 1
#. Set ``desc[i].inflight`` to 1
When supplying used buffers to the driver:
1. Get corresponding used head-descriptor index, i
2. Set ``desc[i].next`` to ``last_batch_head``
3. Set ``last_batch_head`` to ``i``
#. Steps 1,2,3 may be performed repeatedly if batching is possible
#. Increase the ``idx`` value of used ring by the size of the batch
#. Set the ``inflight`` field of each ``DescStateSplit`` entry in the batch to 0
#. Set ``used_idx`` to the ``idx`` value of used ring
When reconnecting:
#. If the value of ``used_idx`` does not match the ``idx`` value of
used ring (means the inflight field of ``DescStateSplit`` entries in
last batch may be incorrect),
a. Subtract the value of ``used_idx`` from the ``idx`` value of
used ring to get last batch size of ``DescStateSplit`` entries
#. Set the ``inflight`` field of each ``DescStateSplit`` entry to 0 in last batch
list which starts from ``last_batch_head``
#. Set ``used_idx`` to the ``idx`` value of used ring
#. Resubmit inflight ``DescStateSplit`` entries in order of their
counter value
For packed virtqueue, queue region can be implemented as:
.. code:: c
typedef struct DescStatePacked {
/* Indicate whether this descriptor is inflight or not.
* Only available for head-descriptor. */
uint8_t inflight;
/* Padding */
uint8_t padding;
/* Link to the next free entry */
uint16_t next;
/* Link to the last entry of descriptor list.
* Only available for head-descriptor. */
uint16_t last;
/* The length of descriptor list.
* Only available for head-descriptor. */
uint16_t num;
/* Used to preserve the order of fetching available descriptors.
* Only available for head-descriptor. */
uint64_t counter;
/* The buffer id */
uint16_t id;
/* The descriptor flags */
uint16_t flags;
/* The buffer length */
uint32_t len;
/* The buffer address */
uint64_t addr;
} DescStatePacked;
typedef struct QueueRegionPacked {
/* The feature flags of this region. Now it's initialized to 0. */
uint64_t features;
/* The version of this region. It's 1 currently.
* Zero value indicates an uninitialized buffer */
uint16_t version;
/* The size of DescStatePacked array. It's equal to the virtqueue
* size. Slave could get it from queue size field of VhostUserInflight. */
uint16_t desc_num;
/* The head of free DescStatePacked entry list */
uint16_t free_head;
/* The old head of free DescStatePacked entry list */
uint16_t old_free_head;
/* The used index of descriptor ring */
uint16_t used_idx;
/* The old used index of descriptor ring */
uint16_t old_used_idx;
/* Device ring wrap counter */
uint8_t used_wrap_counter;
/* The old device ring wrap counter */
uint8_t old_used_wrap_counter;
/* Padding */
uint8_t padding[7];
/* Used to track the state of each descriptor fetched from descriptor ring */
misc: Replace zero-length arrays with flexible array member (manual) Description copied from Linux kernel commit from Gustavo A. R. Silva (see [3]): --v-- description start --v-- The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member [1], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being unadvertenly introduced [2] to the Linux codebase from now on. --^-- description end --^-- Do the similar housekeeping in the QEMU codebase (which uses C99 since commit 7be41675f7cb). All these instances of code were found with the help of the following command (then manual analysis, without modifying structures only having a single flexible array member, such QEDTable in block/qed.h): git grep -F '[0];' [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=76497732932f [3] https://git.kernel.org/pub/scm/linux/kernel/git/gustavoars/linux.git/commit/?id=17642a2fbd2c1 Inspired-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Reviewed-by: David Hildenbrand <david@redhat.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2020-03-04 18:38:16 +03:00
DescStatePacked desc[];
} QueueRegionPacked;
To track inflight I/O, the queue region should be processed as follows:
When receiving available buffers from the driver:
#. Get the next available descriptor entry from descriptor ring, ``d``
#. If ``d`` is head descriptor,
a. Set ``desc[old_free_head].num`` to 0
#. Set ``desc[old_free_head].counter`` to the value of global counter
#. Increase global counter by 1
#. Set ``desc[old_free_head].inflight`` to 1
#. If ``d`` is last descriptor, set ``desc[old_free_head].last`` to
``free_head``
#. Increase ``desc[old_free_head].num`` by 1
#. Set ``desc[free_head].addr``, ``desc[free_head].len``,
``desc[free_head].flags``, ``desc[free_head].id`` to ``d.addr``,
``d.len``, ``d.flags``, ``d.id``
#. Set ``free_head`` to ``desc[free_head].next``
#. If ``d`` is last descriptor, set ``old_free_head`` to ``free_head``
When supplying used buffers to the driver:
1. Get corresponding used head-descriptor entry from descriptor ring,
``d``
2. Get corresponding ``DescStatePacked`` entry, ``e``
3. Set ``desc[e.last].next`` to ``free_head``
4. Set ``free_head`` to the index of ``e``
#. Steps 1,2,3,4 may be performed repeatedly if batching is possible
#. Increase ``used_idx`` by the size of the batch and update
``used_wrap_counter`` if needed
#. Update ``d.flags``
#. Set the ``inflight`` field of each head ``DescStatePacked`` entry
in the batch to 0
#. Set ``old_free_head``, ``old_used_idx``, ``old_used_wrap_counter``
to ``free_head``, ``used_idx``, ``used_wrap_counter``
When reconnecting:
#. If ``used_idx`` does not match ``old_used_idx`` (means the
``inflight`` field of ``DescStatePacked`` entries in last batch may
be incorrect),
a. Get the next descriptor ring entry through ``old_used_idx``, ``d``
#. Use ``old_used_wrap_counter`` to calculate the available flags
#. If ``d.flags`` is not equal to the calculated flags value (means
slave has submitted the buffer to guest driver before crash, so
it has to commit the in-progres update), set ``old_free_head``,
``old_used_idx``, ``old_used_wrap_counter`` to ``free_head``,
``used_idx``, ``used_wrap_counter``
#. Set ``free_head``, ``used_idx``, ``used_wrap_counter`` to
``old_free_head``, ``old_used_idx``, ``old_used_wrap_counter``
(roll back any in-progress update)
#. Set the ``inflight`` field of each ``DescStatePacked`` entry in
free list to 0
#. Resubmit inflight ``DescStatePacked`` entries in order of their
counter value
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
In-band notifications
---------------------
In some limited situations (e.g. for simulation) it is desirable to
have the kick, call and error (if used) signals done via in-band
messages instead of asynchronous eventfd notifications. This can be
done by negotiating the ``VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS``
protocol feature.
Note that due to the fact that too many messages on the sockets can
cause the sending application(s) to block, it is not advised to use
this feature unless absolutely necessary. It is also considered an
error to negotiate this feature without also negotiating
``VHOST_USER_PROTOCOL_F_SLAVE_REQ`` and ``VHOST_USER_PROTOCOL_F_REPLY_ACK``,
the former is necessary for getting a message channel from the slave
to the master, while the latter needs to be used with the in-band
notification messages to block until they are processed, both to avoid
blocking later and for proper processing (at least in the simulation
use case.) As it has no other way of signalling this error, the slave
should close the connection as a response to a
``VHOST_USER_SET_PROTOCOL_FEATURES`` message that sets the in-band
notifications feature flag without the other two.
Protocol features
-----------------
.. code:: c
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
#define VHOST_USER_PROTOCOL_F_MQ 0
#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_MTU 4
#define VHOST_USER_PROTOCOL_F_SLAVE_REQ 5
#define VHOST_USER_PROTOCOL_F_CROSS_ENDIAN 6
#define VHOST_USER_PROTOCOL_F_CRYPTO_SESSION 7
#define VHOST_USER_PROTOCOL_F_PAGEFAULT 8
#define VHOST_USER_PROTOCOL_F_CONFIG 9
#define VHOST_USER_PROTOCOL_F_SLAVE_SEND_FD 10
#define VHOST_USER_PROTOCOL_F_HOST_NOTIFIER 11
#define VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD 12
#define VHOST_USER_PROTOCOL_F_RESET_DEVICE 13
#define VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS 14
#define VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS 15
#define VHOST_USER_PROTOCOL_F_STATUS 16
Master message types
--------------------
``VHOST_USER_GET_FEATURES``
:id: 1
:equivalent ioctl: ``VHOST_GET_FEATURES``
:master payload: N/A
:slave payload: ``u64``
Get from the underlying vhost implementation the features bitmask.
Feature bit ``VHOST_USER_F_PROTOCOL_FEATURES`` signals slave support
for ``VHOST_USER_GET_PROTOCOL_FEATURES`` and
``VHOST_USER_SET_PROTOCOL_FEATURES``.
``VHOST_USER_SET_FEATURES``
:id: 2
:equivalent ioctl: ``VHOST_SET_FEATURES``
:master payload: ``u64``
Enable features in the underlying vhost implementation using a
bitmask. Feature bit ``VHOST_USER_F_PROTOCOL_FEATURES`` signals
slave support for ``VHOST_USER_GET_PROTOCOL_FEATURES`` and
``VHOST_USER_SET_PROTOCOL_FEATURES``.
``VHOST_USER_GET_PROTOCOL_FEATURES``
:id: 15
:equivalent ioctl: ``VHOST_GET_FEATURES``
:master payload: N/A
:slave payload: ``u64``
Get the protocol feature bitmask from the underlying vhost
implementation. Only legal if feature bit
``VHOST_USER_F_PROTOCOL_FEATURES`` is present in
``VHOST_USER_GET_FEATURES``.
.. Note::
Slave that reported ``VHOST_USER_F_PROTOCOL_FEATURES`` must
support this message even before ``VHOST_USER_SET_FEATURES`` was
called.
``VHOST_USER_SET_PROTOCOL_FEATURES``
:id: 16
:equivalent ioctl: ``VHOST_SET_FEATURES``
:master payload: ``u64``
Enable protocol features in the underlying vhost implementation.
Only legal if feature bit ``VHOST_USER_F_PROTOCOL_FEATURES`` is present in
``VHOST_USER_GET_FEATURES``.
.. Note::
Slave that reported ``VHOST_USER_F_PROTOCOL_FEATURES`` must support
this message even before ``VHOST_USER_SET_FEATURES`` was called.
``VHOST_USER_SET_OWNER``
:id: 3
:equivalent ioctl: ``VHOST_SET_OWNER``
:master payload: N/A
Issued when a new connection is established. It sets the current
*master* as an owner of the session. This can be used on the *slave*
as a "session start" flag.
``VHOST_USER_RESET_OWNER``
:id: 4
:master payload: N/A
.. admonition:: Deprecated
This is no longer used. Used to be sent to request disabling all
rings, but some clients interpreted it to also discard connection
state (this interpretation would lead to bugs). It is recommended
that clients either ignore this message, or use it to disable all
rings.
``VHOST_USER_SET_MEM_TABLE``
:id: 5
:equivalent ioctl: ``VHOST_SET_MEM_TABLE``
:master payload: memory regions description
:slave payload: (postcopy only) memory regions description
Sets the memory map regions on the slave so it can translate the
vring addresses. In the ancillary data there is an array of file
descriptors for each memory mapped region. The size and ordering of
the fds matches the number and ordering of memory regions.
When ``VHOST_USER_POSTCOPY_LISTEN`` has been received,
``SET_MEM_TABLE`` replies with the bases of the memory mapped
regions to the master. The slave must have mmap'd the regions but
not yet accessed them and should not yet generate a userfault
event.
.. Note::
``NEED_REPLY_MASK`` is not set in this case. QEMU will then
reply back to the list of mappings with an empty
``VHOST_USER_SET_MEM_TABLE`` as an acknowledgement; only upon
reception of this message may the guest start accessing the memory
and generating faults.
``VHOST_USER_SET_LOG_BASE``
:id: 6
:equivalent ioctl: ``VHOST_SET_LOG_BASE``
:master payload: u64
:slave payload: N/A
Sets logging shared memory space.
When slave has ``VHOST_USER_PROTOCOL_F_LOG_SHMFD`` protocol feature,
the log memory fd is provided in the ancillary data of
``VHOST_USER_SET_LOG_BASE`` message, the size and offset of shared
memory area provided in the message.
``VHOST_USER_SET_LOG_FD``
:id: 7
:equivalent ioctl: ``VHOST_SET_LOG_FD``
:master payload: N/A
Sets the logging file descriptor, which is passed as ancillary data.
``VHOST_USER_SET_VRING_NUM``
:id: 8
:equivalent ioctl: ``VHOST_SET_VRING_NUM``
:master payload: vring state description
Set the size of the queue.
``VHOST_USER_SET_VRING_ADDR``
:id: 9
:equivalent ioctl: ``VHOST_SET_VRING_ADDR``
:master payload: vring address description
:slave payload: N/A
Sets the addresses of the different aspects of the vring.
``VHOST_USER_SET_VRING_BASE``
:id: 10
:equivalent ioctl: ``VHOST_SET_VRING_BASE``
:master payload: vring state description
Sets the base offset in the available vring.
``VHOST_USER_GET_VRING_BASE``
:id: 11
:equivalent ioctl: ``VHOST_USER_GET_VRING_BASE``
:master payload: vring state description
:slave payload: vring state description
Get the available vring base offset.
``VHOST_USER_SET_VRING_KICK``
:id: 12
:equivalent ioctl: ``VHOST_SET_VRING_KICK``
:master payload: ``u64``
Set the event file descriptor for adding buffers to the vring. It is
passed in the ancillary data.
Bits (0-7) of the payload contain the vring index. Bit 8 is the
invalid FD flag. This flag is set when there is no file descriptor
in the ancillary data. This signals that polling should be used
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
instead of waiting for the kick. Note that if the protocol feature
``VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS`` has been negotiated
this message isn't necessary as the ring is also started on the
``VHOST_USER_VRING_KICK`` message, it may however still be used to
set an event file descriptor (which will be preferred over the
message) or to enable polling.
``VHOST_USER_SET_VRING_CALL``
:id: 13
:equivalent ioctl: ``VHOST_SET_VRING_CALL``
:master payload: ``u64``
Set the event file descriptor to signal when buffers are used. It is
passed in the ancillary data.
Bits (0-7) of the payload contain the vring index. Bit 8 is the
invalid FD flag. This flag is set when there is no file descriptor
in the ancillary data. This signals that polling will be used
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
instead of waiting for the call. Note that if the protocol features
``VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS`` and
``VHOST_USER_PROTOCOL_F_SLAVE_REQ`` have been negotiated this message
isn't necessary as the ``VHOST_USER_SLAVE_VRING_CALL`` message can be
used, it may however still be used to set an event file descriptor
or to enable polling.
``VHOST_USER_SET_VRING_ERR``
:id: 14
:equivalent ioctl: ``VHOST_SET_VRING_ERR``
:master payload: ``u64``
Set the event file descriptor to signal when error occurs. It is
passed in the ancillary data.
Bits (0-7) of the payload contain the vring index. Bit 8 is the
invalid FD flag. This flag is set when there is no file descriptor
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
in the ancillary data. Note that if the protocol features
``VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS`` and
``VHOST_USER_PROTOCOL_F_SLAVE_REQ`` have been negotiated this message
isn't necessary as the ``VHOST_USER_SLAVE_VRING_ERR`` message can be
used, it may however still be used to set an event file descriptor
(which will be preferred over the message).
``VHOST_USER_GET_QUEUE_NUM``
:id: 17
:equivalent ioctl: N/A
:master payload: N/A
:slave payload: u64
Query how many queues the backend supports.
This request should be sent only when ``VHOST_USER_PROTOCOL_F_MQ``
is set in queried protocol features by
``VHOST_USER_GET_PROTOCOL_FEATURES``.
``VHOST_USER_SET_VRING_ENABLE``
:id: 18
:equivalent ioctl: N/A
:master payload: vring state description
Signal slave to enable or disable corresponding vring.
This request should be sent only when
``VHOST_USER_F_PROTOCOL_FEATURES`` has been negotiated.
``VHOST_USER_SEND_RARP``
:id: 19
:equivalent ioctl: N/A
:master payload: ``u64``
Ask vhost user backend to broadcast a fake RARP to notify the migration
is terminated for guest that does not support GUEST_ANNOUNCE.
Only legal if feature bit ``VHOST_USER_F_PROTOCOL_FEATURES`` is
present in ``VHOST_USER_GET_FEATURES`` and protocol feature bit
``VHOST_USER_PROTOCOL_F_RARP`` is present in
``VHOST_USER_GET_PROTOCOL_FEATURES``. 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_NET_SET_MTU``
:id: 20
:equivalent ioctl: N/A
:master payload: ``u64``
Set host MTU value exposed to the guest.
This request should be sent only when ``VIRTIO_NET_F_MTU`` feature
has been successfully negotiated, ``VHOST_USER_F_PROTOCOL_FEATURES``
is present in ``VHOST_USER_GET_FEATURES`` and protocol feature bit
``VHOST_USER_PROTOCOL_F_NET_MTU`` is present in
``VHOST_USER_GET_PROTOCOL_FEATURES``.
If ``VHOST_USER_PROTOCOL_F_REPLY_ACK`` is negotiated, slave must
respond with zero in case the specified MTU is valid, or non-zero
otherwise.
``VHOST_USER_SET_SLAVE_REQ_FD``
:id: 21
:equivalent ioctl: N/A
:master payload: N/A
Set the socket file descriptor for slave initiated requests. It is passed
in the ancillary data.
This request should be sent only when
``VHOST_USER_F_PROTOCOL_FEATURES`` has been negotiated, and protocol
feature bit ``VHOST_USER_PROTOCOL_F_SLAVE_REQ`` bit is present in
``VHOST_USER_GET_PROTOCOL_FEATURES``. If
``VHOST_USER_PROTOCOL_F_REPLY_ACK`` is negotiated, slave must
respond with zero for success, non-zero otherwise.
``VHOST_USER_IOTLB_MSG``
:id: 22
:equivalent ioctl: N/A (equivalent to ``VHOST_IOTLB_MSG`` message type)
:master payload: ``struct vhost_iotlb_msg``
:slave payload: ``u64``
Send IOTLB messages with ``struct vhost_iotlb_msg`` as payload.
Master sends such requests to update and invalidate entries in the
device IOTLB. The slave has to acknowledge the request with sending
zero as ``u64`` payload for success, non-zero otherwise.
This request should be send only when ``VIRTIO_F_IOMMU_PLATFORM``
feature has been successfully negotiated.
``VHOST_USER_SET_VRING_ENDIAN``
:id: 23
:equivalent ioctl: ``VHOST_SET_VRING_ENDIAN``
:master payload: vring state description
Set the endianness of a VQ for legacy devices. Little-endian is
indicated with state.num set to 0 and big-endian is indicated with
state.num set to 1. Other values are invalid.
This request should be sent only when
``VHOST_USER_PROTOCOL_F_CROSS_ENDIAN`` has been negotiated.
Backends that negotiated this feature should handle both
endiannesses and expect this message once (per VQ) during device
configuration (ie. before the master starts the VQ).
``VHOST_USER_GET_CONFIG``
:id: 24
:equivalent ioctl: N/A
:master payload: virtio device config space
:slave payload: virtio device config space
When ``VHOST_USER_PROTOCOL_F_CONFIG`` is negotiated, this message is
submitted by the vhost-user master to fetch the contents of the
virtio device configuration space, vhost-user slave's payload size
MUST match master's request, vhost-user slave uses zero length of
payload to indicate an error to vhost-user master. The vhost-user
master may cache the contents to avoid repeated
``VHOST_USER_GET_CONFIG`` calls.
``VHOST_USER_SET_CONFIG``
:id: 25
:equivalent ioctl: N/A
:master payload: virtio device config space
:slave payload: N/A
When ``VHOST_USER_PROTOCOL_F_CONFIG`` is negotiated, this message is
submitted by the vhost-user master when the Guest changes the virtio
device configuration space and also can be used for live migration
on the destination host. The vhost-user slave must check the flags
field, and slaves MUST NOT accept SET_CONFIG for read-only
configuration space fields unless the live migration bit is set.
``VHOST_USER_CREATE_CRYPTO_SESSION``
:id: 26
:equivalent ioctl: N/A
:master payload: crypto session description
:slave payload: crypto session description
Create a session for crypto operation. The server side must return
the session id, 0 or positive for success, negative for failure.
This request should be sent only when
``VHOST_USER_PROTOCOL_F_CRYPTO_SESSION`` feature has been
successfully negotiated. It's a required feature for crypto
devices.
``VHOST_USER_CLOSE_CRYPTO_SESSION``
:id: 27
:equivalent ioctl: N/A
:master payload: ``u64``
Close a session for crypto operation which was previously
created by ``VHOST_USER_CREATE_CRYPTO_SESSION``.
This request should be sent only when
``VHOST_USER_PROTOCOL_F_CRYPTO_SESSION`` feature has been
successfully negotiated. It's a required feature for crypto
devices.
``VHOST_USER_POSTCOPY_ADVISE``
:id: 28
:master payload: N/A
:slave payload: userfault fd
When ``VHOST_USER_PROTOCOL_F_PAGEFAULT`` is supported, the master
advises slave that a migration with postcopy enabled is underway,
the slave must open a userfaultfd for later use. Note that at this
stage the migration is still in precopy mode.
``VHOST_USER_POSTCOPY_LISTEN``
:id: 29
:master payload: N/A
Master advises slave that a transition to postcopy mode has
happened. The slave must ensure that shared memory is registered
with userfaultfd to cause faulting of non-present pages.
This is always sent sometime after a ``VHOST_USER_POSTCOPY_ADVISE``,
and thus only when ``VHOST_USER_PROTOCOL_F_PAGEFAULT`` is supported.
``VHOST_USER_POSTCOPY_END``
:id: 30
:slave payload: ``u64``
Master advises that postcopy migration has now completed. The slave
must disable the userfaultfd. The response is an acknowledgement
only.
When ``VHOST_USER_PROTOCOL_F_PAGEFAULT`` is supported, this message
is sent at the end of the migration, after
``VHOST_USER_POSTCOPY_LISTEN`` was previously sent.
The value returned is an error indication; 0 is success.
``VHOST_USER_GET_INFLIGHT_FD``
:id: 31
:equivalent ioctl: N/A
:master payload: inflight description
When ``VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD`` protocol feature has
been successfully negotiated, this message is submitted by master to
get a shared buffer from slave. The shared buffer will be used to
track inflight I/O by slave. QEMU should retrieve a new one when vm
reset.
``VHOST_USER_SET_INFLIGHT_FD``
:id: 32
:equivalent ioctl: N/A
:master payload: inflight description
When ``VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD`` protocol feature has
been successfully negotiated, this message is submitted by master to
send the shared inflight buffer back to slave so that slave could
get inflight I/O after a crash or restart.
``VHOST_USER_GPU_SET_SOCKET``
:id: 33
:equivalent ioctl: N/A
:master payload: N/A
Sets the GPU protocol socket file descriptor, which is passed as
ancillary data. The GPU protocol is used to inform the master of
rendering state and updates. See vhost-user-gpu.rst for details.
``VHOST_USER_RESET_DEVICE``
:id: 34
:equivalent ioctl: N/A
:master payload: N/A
:slave payload: N/A
Ask the vhost user backend to disable all rings and reset all
internal device state to the initial state, ready to be
reinitialized. The backend retains ownership of the device
throughout the reset operation.
Only valid if the ``VHOST_USER_PROTOCOL_F_RESET_DEVICE`` protocol
feature is set by the backend.
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
``VHOST_USER_VRING_KICK``
:id: 35
:equivalent ioctl: N/A
:slave payload: vring state description
:master payload: N/A
When the ``VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS`` protocol
feature has been successfully negotiated, this message may be
submitted by the master to indicate that a buffer was added to
the vring instead of signalling it using the vring's kick file
descriptor or having the slave rely on polling.
The state.num field is currently reserved and must be set to 0.
``VHOST_USER_GET_MAX_MEM_SLOTS``
:id: 36
:equivalent ioctl: N/A
:slave payload: u64
When the ``VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS`` protocol
feature has been successfully negotiated, this message is submitted
by master to the slave. The slave should return the message with a
u64 payload containing the maximum number of memory slots for
QEMU to expose to the guest. The value returned by the backend
will be capped at the maximum number of ram slots which can be
supported by the target platform.
Transmit vhost-user memory regions individually With this change, when the VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS protocol feature has been negotiated, Qemu no longer sends the backend all the memory regions in a single message. Rather, when the memory tables are set or updated, a series of VHOST_USER_ADD_MEM_REG and VHOST_USER_REM_MEM_REG messages are sent to transmit the regions to map and/or unmap instead of sending send all the regions in one fixed size VHOST_USER_SET_MEM_TABLE message. The vhost_user struct maintains a shadow state of the VM’s memory regions. When the memory tables are modified, the vhost_user_set_mem_table() function compares the new device memory state to the shadow state and only sends regions which need to be unmapped or mapped in. The regions which must be unmapped are sent first, followed by the new regions to be mapped in. After all the messages have been sent, the shadow state is set to the current virtual device state. Existing backends which do not support VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS are unaffected. Signed-off-by: Raphael Norwitz <raphael.norwitz@nutanix.com> Signed-off-by: Swapnil Ingle <swapnil.ingle@nutanix.com> Signed-off-by: Peter Turschmid <peter.turschm@nutanix.com> Suggested-by: Mike Cui <cui@nutanix.com> Message-Id: <1588533678-23450-5-git-send-email-raphael.norwitz@nutanix.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2020-05-21 08:00:35 +03:00
``VHOST_USER_ADD_MEM_REG``
:id: 37
:equivalent ioctl: N/A
:slave payload: memory region
When the ``VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS`` protocol
feature has been successfully negotiated, this message is submitted
by the master to the slave. The message payload contains a memory
region descriptor struct, describing a region of guest memory which
the slave device must map in. When the
``VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS`` protocol feature has
been successfully negotiated, along with the
``VHOST_USER_REM_MEM_REG`` message, this message is used to set and
update the memory tables of the slave device.
``VHOST_USER_REM_MEM_REG``
:id: 38
:equivalent ioctl: N/A
:slave payload: memory region
When the ``VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS`` protocol
feature has been successfully negotiated, this message is submitted
by the master to the slave. The message payload contains a memory
region descriptor struct, describing a region of guest memory which
the slave device must unmap. When the
``VHOST_USER_PROTOCOL_F_CONFIGURE_MEM_SLOTS`` protocol feature has
been successfully negotiated, along with the
``VHOST_USER_ADD_MEM_REG`` message, this message is used to set and
update the memory tables of the slave device.
``VHOST_USER_SET_STATUS``
:id: 39
:equivalent ioctl: VHOST_VDPA_SET_STATUS
:slave payload: N/A
:master payload: ``u64``
When the ``VHOST_USER_PROTOCOL_F_STATUS`` protocol feature has been
successfully negotiated, this message is submitted by the master to
notify the backend with updated device status as defined in the Virtio
specification.
``VHOST_USER_GET_STATUS``
:id: 40
:equivalent ioctl: VHOST_VDPA_GET_STATUS
:slave payload: ``u64``
:master payload: N/A
When the ``VHOST_USER_PROTOCOL_F_STATUS`` protocol feature has been
successfully negotiated, this message is submitted by the master to
query the backend for its device status as defined in the Virtio
specification.
Slave message types
-------------------
``VHOST_USER_SLAVE_IOTLB_MSG``
:id: 1
:equivalent ioctl: N/A (equivalent to ``VHOST_IOTLB_MSG`` message type)
:slave payload: ``struct vhost_iotlb_msg``
:master payload: N/A
Send IOTLB messages with ``struct vhost_iotlb_msg`` as payload.
Slave sends such requests to notify of an IOTLB miss, or an IOTLB
access failure. If ``VHOST_USER_PROTOCOL_F_REPLY_ACK`` is
negotiated, and slave set the ``VHOST_USER_NEED_REPLY`` flag, master
must respond with zero when operation is successfully completed, or
non-zero otherwise. This request should be send only when
``VIRTIO_F_IOMMU_PLATFORM`` feature has been successfully
negotiated.
``VHOST_USER_SLAVE_CONFIG_CHANGE_MSG``
:id: 2
:equivalent ioctl: N/A
:slave payload: N/A
:master payload: N/A
When ``VHOST_USER_PROTOCOL_F_CONFIG`` is negotiated, vhost-user
slave sends such messages to notify that the virtio device's
configuration space has changed, for those host devices which can
support such feature, host driver can send ``VHOST_USER_GET_CONFIG``
message to slave to get the latest content. If
``VHOST_USER_PROTOCOL_F_REPLY_ACK`` is negotiated, and slave set the
``VHOST_USER_NEED_REPLY`` flag, master must respond with zero when
operation is successfully completed, or non-zero otherwise.
``VHOST_USER_SLAVE_VRING_HOST_NOTIFIER_MSG``
:id: 3
:equivalent ioctl: N/A
:slave payload: vring area description
:master payload: N/A
Sets host notifier for a specified queue. The queue index is
contained in the ``u64`` field of the vring area description. The
host notifier is described by the file descriptor (typically it's a
VFIO device fd) which is passed as ancillary data and the size
(which is mmap size and should be the same as host page size) and
offset (which is mmap offset) carried in the vring area
description. QEMU can mmap the file descriptor based on the size and
offset to get a memory range. Registering a host notifier means
mapping this memory range to the VM as the specified queue's notify
MMIO region. Slave sends this request to tell QEMU to de-register
the existing notifier if any and register the new notifier if the
request is sent with a file descriptor.
This request should be sent only when
``VHOST_USER_PROTOCOL_F_HOST_NOTIFIER`` protocol feature has been
successfully negotiated.
docs: vhost-user: add in-band kick/call messages For good reason, vhost-user is currently built asynchronously, that way better performance can be obtained. However, for certain use cases such as simulation, this is problematic. Consider an event-based simulation in which both the device and CPU have scheduled according to a simulation "calendar". Now, consider the CPU sending I/O to the device, over a vring in the vhost-user protocol. In this case, the CPU must wait for the vring interrupt to have been processed by the device, so that the device is able to put an entry onto the simulation calendar to obtain time to handle the interrupt. Note that this doesn't mean the I/O is actually done at this time, it just means that the handling of it is scheduled before the CPU can continue running. This cannot be done with the asynchronous eventfd based vring kick and call design. Extend the protocol slightly, so that a message can be used for kick and call instead, if VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS is negotiated. This in itself doesn't guarantee synchronisation, but both sides can also negotiate VHOST_USER_PROTOCOL_F_REPLY_ACK and thus get a reply to this message by setting the need_reply flag, and ensure synchronisation this way. To really use it in both directions, VHOST_USER_PROTOCOL_F_SLAVE_REQ is also needed. Since it is used for simulation purposes and too many messages on the socket can lock up the virtual machine, document that this should only be used together with the mentioned features. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Message-Id: <20200123081708.7817-6-johannes@sipsolutions.net> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2020-01-23 11:17:07 +03:00
``VHOST_USER_SLAVE_VRING_CALL``
:id: 4
:equivalent ioctl: N/A
:slave payload: vring state description
:master payload: N/A
When the ``VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS`` protocol
feature has been successfully negotiated, this message may be
submitted by the slave to indicate that a buffer was used from
the vring instead of signalling this using the vring's call file
descriptor or having the master relying on polling.
The state.num field is currently reserved and must be set to 0.
``VHOST_USER_SLAVE_VRING_ERR``
:id: 5
:equivalent ioctl: N/A
:slave payload: vring state description
:master payload: N/A
When the ``VHOST_USER_PROTOCOL_F_INBAND_NOTIFICATIONS`` protocol
feature has been successfully negotiated, this message may be
submitted by the slave to indicate that an error occurred on the
specific vring, instead of signalling the error file descriptor
set by the master via ``VHOST_USER_SET_VRING_ERR``.
The state.num field is currently reserved and must be set to 0.
.. _reply_ack:
VHOST_USER_PROTOCOL_F_REPLY_ACK
-------------------------------
The original vhost-user specification only demands replies for certain
commands. This differs from the vhost protocol implementation where
commands are sent over an ``ioctl()`` call and block until the client
has completed.
With this protocol extension negotiated, the sender (QEMU) can set the
``need_reply`` [Bit 3] flag to any command. This indicates that the
client MUST respond with a Payload ``VhostUserMsg`` indicating success
or failure. The payload should be set to zero on success or non-zero
on failure, unless the message already has an explicit reply body.
The response payload gives QEMU a deterministic indication of the result
of the command. Today, QEMU is expected to terminate the main vhost-user
loop upon receiving such errors. In future, qemu could be taught to be more
resilient for selective requests.
For the message types that already solicit a reply from the client,
the presence of ``VHOST_USER_PROTOCOL_F_REPLY_ACK`` or need_reply bit
being set brings no behavioural change. (See the Communication_
section for details.)
.. _backend_conventions:
Backend program conventions
===========================
vhost-user backends can provide various devices & services and may
need to be configured manually depending on the use case. However, it
is a good idea to follow the conventions listed here when
possible. Users, QEMU or libvirt, can then rely on some common
behaviour to avoid heterogenous configuration and management of the
backend programs and facilitate interoperability.
Each backend installed on a host system should come with at least one
JSON file that conforms to the vhost-user.json schema. Each file
informs the management applications about the backend type, and binary
location. In addition, it defines rules for management apps for
picking the highest priority backend when multiple match the search
criteria (see ``@VhostUserBackend`` documentation in the schema file).
If the backend is not capable of enabling a requested feature on the
host (such as 3D acceleration with virgl), or the initialization
failed, the backend should fail to start early and exit with a status
!= 0. It may also print a message to stderr for further details.
The backend program must not daemonize itself, but it may be
daemonized by the management layer. It may also have a restricted
access to the system.
File descriptors 0, 1 and 2 will exist, and have regular
stdin/stdout/stderr usage (they may have been redirected to /dev/null
by the management layer, or to a log handler).
The backend program must end (as quickly and cleanly as possible) when
the SIGTERM signal is received. Eventually, it may receive SIGKILL by
the management layer after a few seconds.
The following command line options have an expected behaviour. They
are mandatory, unless explicitly said differently:
--socket-path=PATH
This option specify the location of the vhost-user Unix domain socket.
It is incompatible with --fd.
--fd=FDNUM
When this argument is given, the backend program is started with the
vhost-user socket as file descriptor FDNUM. It is incompatible with
--socket-path.
--print-capabilities
Output to stdout the backend capabilities in JSON format, and then
exit successfully. Other options and arguments should be ignored, and
the backend program should not perform its normal function. The
capabilities can be reported dynamically depending on the host
capabilities.
The JSON output is described in the ``vhost-user.json`` schema, by
```@VHostUserBackendCapabilities``. Example:
.. code:: json
{
"type": "foo",
"features": [
"feature-a",
"feature-b"
]
}
vhost-user-input
----------------
Command line options:
--evdev-path=PATH
Specify the linux input device.
(optional)
--no-grab
Do no request exclusive access to the input device.
(optional)
vhost-user-gpu
--------------
Command line options:
--render-node=PATH
Specify the GPU DRM render node.
(optional)
--virgl
Enable virgl rendering support.
(optional)
vhost-user-blk
--------------
Command line options:
--blk-file=PATH
Specify block device or file path.
(optional)
--read-only
Enable read-only.
(optional)