1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2010-2016 Intel Corporation.
7 The vhost library implements a user space virtio net server allowing the user
8 to manipulate the virtio ring directly. In another words, it allows the user
9 to fetch/put packets from/to the VM virtio net device. To achieve this, a
10 vhost library should be able to:
12 * Access the guest memory:
14 For QEMU, this is done by using the ``-object memory-backend-file,share=on,...``
15 option. Which means QEMU will create a file to serve as the guest RAM.
16 The ``share=on`` option allows another process to map that file, which
17 means it can access the guest RAM.
19 * Know all the necessary information about the vring:
21 Information such as where the available ring is stored. Vhost defines some
22 messages (passed through a Unix domain socket file) to tell the backend all
23 the information it needs to know how to manipulate the vring.
29 The following is an overview of some key Vhost API functions:
31 * ``rte_vhost_driver_register(path, flags)``
33 This function registers a vhost driver into the system. ``path`` specifies
34 the Unix domain socket file path.
36 Currently supported flags are:
38 - ``RTE_VHOST_USER_CLIENT``
40 DPDK vhost-user will act as the client when this flag is given. See below
43 - ``RTE_VHOST_USER_NO_RECONNECT``
45 When DPDK vhost-user acts as the client it will keep trying to reconnect
46 to the server (QEMU) until it succeeds. This is useful in two cases:
48 * When QEMU is not started yet.
49 * When QEMU restarts (for example due to a guest OS reboot).
51 This reconnect option is enabled by default. However, it can be turned off
54 - ``RTE_VHOST_USER_IOMMU_SUPPORT``
56 IOMMU support will be enabled when this flag is set. It is disabled by
59 Enabling this flag makes possible to use guest vIOMMU to protect vhost
60 from accessing memory the virtio device isn't allowed to, when the feature
61 is negotiated and an IOMMU device is declared.
63 - ``RTE_VHOST_USER_POSTCOPY_SUPPORT``
65 Postcopy live-migration support will be enabled when this flag is set.
66 It is disabled by default.
68 Enabling this flag should only be done when the calling application does
69 not pre-fault the guest shared memory, otherwise migration would fail.
71 - ``RTE_VHOST_USER_LINEARBUF_SUPPORT``
73 Enabling this flag forces vhost dequeue function to only provide linear
74 pktmbuf (no multi-segmented pktmbuf).
76 The vhost library by default provides a single pktmbuf for given a
77 packet, but if for some reason the data doesn't fit into a single
78 pktmbuf (e.g., TSO is enabled), the library will allocate additional
79 pktmbufs from the same mempool and chain them together to create a
80 multi-segmented pktmbuf.
82 However, the vhost application needs to support multi-segmented format.
83 If the vhost application does not support that format and requires large
84 buffers to be dequeue, this flag should be enabled to force only linear
85 buffers (see RTE_VHOST_USER_EXTBUF_SUPPORT) or drop the packet.
87 It is disabled by default.
89 - ``RTE_VHOST_USER_EXTBUF_SUPPORT``
91 Enabling this flag allows vhost dequeue function to allocate and attach
92 an external buffer to a pktmbuf if the pkmbuf doesn't provide enough
93 space to store all data.
95 This is useful when the vhost application wants to support large packets
96 but doesn't want to increase the default mempool object size nor to
97 support multi-segmented mbufs (non-linear). In this case, a fresh buffer
98 is allocated using rte_malloc() which gets attached to a pktmbuf using
99 rte_pktmbuf_attach_extbuf().
101 See RTE_VHOST_USER_LINEARBUF_SUPPORT as well to disable multi-segmented
102 mbufs for application that doesn't support chained mbufs.
104 It is disabled by default.
106 - ``RTE_VHOST_USER_ASYNC_COPY``
108 Asynchronous data path will be enabled when this flag is set. Async data
109 path allows applications to register async copy devices (typically
110 hardware DMA channels) to the vhost queues. Vhost leverages the copy
111 device registered to free CPU from memory copy operations. A set of
112 async data path APIs are defined for DPDK applications to make use of
113 the async capability. Only packets enqueued/dequeued by async APIs are
114 processed through the async data path.
116 Currently this feature is only implemented on split ring enqueue data
119 It is disabled by default.
121 * ``rte_vhost_driver_set_features(path, features)``
123 This function sets the feature bits the vhost-user driver supports. The
124 vhost-user driver could be vhost-user net, yet it could be something else,
125 say, vhost-user SCSI.
127 * ``rte_vhost_driver_callback_register(path, vhost_device_ops)``
129 This function registers a set of callbacks, to let DPDK applications take
130 the appropriate action when some events happen. The following events are
133 * ``new_device(int vid)``
135 This callback is invoked when a virtio device becomes ready. ``vid``
136 is the vhost device ID.
138 * ``destroy_device(int vid)``
140 This callback is invoked when a virtio device is paused or shut down.
142 * ``vring_state_changed(int vid, uint16_t queue_id, int enable)``
144 This callback is invoked when a specific queue's state is changed, for
145 example to enabled or disabled.
147 * ``features_changed(int vid, uint64_t features)``
149 This callback is invoked when the features is changed. For example,
150 ``VHOST_F_LOG_ALL`` will be set/cleared at the start/end of live
151 migration, respectively.
153 * ``new_connection(int vid)``
155 This callback is invoked on new vhost-user socket connection. If DPDK
156 acts as the server the device should not be deleted before
157 ``destroy_connection`` callback is received.
159 * ``destroy_connection(int vid)``
161 This callback is invoked when vhost-user socket connection is closed.
162 It indicates that device with id ``vid`` is no longer in use and can be
165 * ``rte_vhost_driver_disable/enable_features(path, features))``
167 This function disables/enables some features. For example, it can be used to
168 disable mergeable buffers and TSO features, which both are enabled by
171 * ``rte_vhost_driver_start(path)``
173 This function triggers the vhost-user negotiation. It should be invoked at
174 the end of initializing a vhost-user driver.
176 * ``rte_vhost_enqueue_burst(vid, queue_id, pkts, count)``
178 Transmits (enqueues) ``count`` packets from host to guest.
180 * ``rte_vhost_dequeue_burst(vid, queue_id, mbuf_pool, pkts, count)``
182 Receives (dequeues) ``count`` packets from guest, and stored them at ``pkts``.
184 * ``rte_vhost_crypto_create(vid, cryptodev_id, sess_mempool, socket_id)``
186 As an extension of new_device(), this function adds virtio-crypto workload
187 acceleration capability to the device. All crypto workload is processed by
188 DPDK cryptodev with the device ID of ``cryptodev_id``.
190 * ``rte_vhost_crypto_free(vid)``
192 Frees the memory and vhost-user message handlers created in
193 rte_vhost_crypto_create().
195 * ``rte_vhost_crypto_fetch_requests(vid, queue_id, ops, nb_ops)``
197 Receives (dequeues) ``nb_ops`` virtio-crypto requests from guest, parses
198 them to DPDK Crypto Operations, and fills the ``ops`` with parsing results.
200 * ``rte_vhost_crypto_finalize_requests(queue_id, ops, nb_ops)``
202 After the ``ops`` are dequeued from Cryptodev, finalizes the jobs and
203 notifies the guest(s).
205 * ``rte_vhost_crypto_set_zero_copy(vid, option)``
207 Enable or disable zero copy feature of the vhost crypto backend.
209 * ``rte_vhost_async_channel_register(vid, queue_id, features, ops)``
211 Register a vhost queue with async copy device channel.
212 Following device ``features`` must be specified together with the
217 Async copy device can guarantee the ordering of copy completion
218 sequence. Copies are completed in the same order with that at
221 Currently, only ``async_inorder`` capable device is supported by vhost.
223 * ``async_threshold``
225 The copy length (in bytes) below which CPU copy will be used even if
226 applications call async vhost APIs to enqueue/dequeue data.
228 Typical value is 512~1024 depending on the async device capability.
230 Applications must provide following ``ops`` callbacks for vhost lib to
231 work with the async copy devices:
233 * ``transfer_data(vid, queue_id, descs, opaque_data, count)``
235 vhost invokes this function to submit copy data to the async devices.
236 For non-async_inorder capable devices, ``opaque_data`` could be used
237 for identifying the completed packets.
239 * ``check_completed_copies(vid, queue_id, opaque_data, max_packets)``
241 vhost invokes this function to get the copy data completed by async
244 * ``rte_vhost_async_channel_unregister(vid, queue_id)``
246 Unregister the async copy device channel from a vhost queue.
248 * ``rte_vhost_submit_enqueue_burst(vid, queue_id, pkts, count, comp_pkts, comp_count)``
250 Submit an enqueue request to transmit ``count`` packets from host to guest
251 by async data path. Successfully enqueued packets can be transfer completed
252 or being occupied by DMA engines; transfer completed packets are returned in
253 ``comp_pkts``, but others are not guaranteed to finish, when this API
256 Applications must not free the packets submitted for enqueue until the
257 packets are completed.
259 * ``rte_vhost_poll_enqueue_completed(vid, queue_id, pkts, count)``
261 Poll enqueue completion status from async data path. Completed packets
262 are returned to applications through ``pkts``.
264 Vhost-user Implementations
265 --------------------------
267 Vhost-user uses Unix domain sockets for passing messages. This means the DPDK
268 vhost-user implementation has two options:
270 * DPDK vhost-user acts as the server.
272 DPDK will create a Unix domain socket server file and listen for
273 connections from the frontend.
275 Note, this is the default mode, and the only mode before DPDK v16.07.
278 * DPDK vhost-user acts as the client.
280 Unlike the server mode, this mode doesn't create the socket file;
281 it just tries to connect to the server (which responses to create the
284 When the DPDK vhost-user application restarts, DPDK vhost-user will try to
285 connect to the server again. This is how the "reconnect" feature works.
288 * The "reconnect" feature requires **QEMU v2.7** (or above).
290 * The vhost supported features must be exactly the same before and
291 after the restart. For example, if TSO is disabled and then enabled,
292 nothing will work and issues undefined might happen.
294 No matter which mode is used, once a connection is established, DPDK
295 vhost-user will start receiving and processing vhost messages from QEMU.
297 For messages with a file descriptor, the file descriptor can be used directly
298 in the vhost process as it is already installed by the Unix domain socket.
300 The supported vhost messages are:
302 * ``VHOST_SET_MEM_TABLE``
303 * ``VHOST_SET_VRING_KICK``
304 * ``VHOST_SET_VRING_CALL``
305 * ``VHOST_SET_LOG_FD``
306 * ``VHOST_SET_VRING_ERR``
308 For ``VHOST_SET_MEM_TABLE`` message, QEMU will send information for each
309 memory region and its file descriptor in the ancillary data of the message.
310 The file descriptor is used to map that region.
312 ``VHOST_SET_VRING_KICK`` is used as the signal to put the vhost device into
313 the data plane, and ``VHOST_GET_VRING_BASE`` is used as the signal to remove
314 the vhost device from the data plane.
316 When the socket connection is closed, vhost will destroy the device.
318 Guest memory requirement
319 ------------------------
321 * Memory pre-allocation
323 For non-async data path, guest memory pre-allocation is not a
324 must. This can help save of memory. If users really want the guest memory
325 to be pre-allocated (e.g., for performance reason), we can add option
326 ``-mem-prealloc`` when starting QEMU. Or, we can lock all memory at vhost
327 side which will force memory to be allocated when mmap at vhost side;
328 option --mlockall in ovs-dpdk is an example in hand.
330 For async data path, we force the VM memory to be pre-allocated at vhost
331 lib when mapping the guest memory; and also we need to lock the memory to
332 prevent pages being swapped out to disk.
336 Make sure ``share=on`` QEMU option is given. vhost-user will not work with
337 a QEMU version without shared memory mapping.
339 Vhost supported vSwitch reference
340 ---------------------------------
342 For more vhost details and how to support vhost in vSwitch, please refer to
343 the vhost example in the DPDK Sample Applications Guide.
345 Vhost data path acceleration (vDPA)
346 -----------------------------------
348 vDPA supports selective datapath in vhost-user lib by enabling virtio ring
349 compatible devices to serve virtio driver directly for datapath acceleration.
351 ``rte_vhost_driver_attach_vdpa_device`` is used to configure the vhost device
352 with accelerated backend.
354 Also vhost device capabilities are made configurable to adopt various devices.
355 Such capabilities include supported features, protocol features, queue number.
357 Finally, a set of device ops is defined for device specific operations:
361 Called to get supported queue number of the device.
365 Called to get supported features of the device.
367 * ``get_protocol_features``
369 Called to get supported protocol features of the device.
373 Called to configure the actual device when the virtio device becomes ready.
377 Called to close the actual device when the virtio device is stopped.
379 * ``set_vring_state``
381 Called to change the state of the vring in the actual device when vring state
386 Called to set the negotiated features to device.
390 Called to allow the device to response to RARP sending.
392 * ``get_vfio_group_fd``
394 Called to get the VFIO group fd of the device.
396 * ``get_vfio_device_fd``
398 Called to get the VFIO device fd of the device.
400 * ``get_notify_area``
402 Called to get the notify area info of the queue.