4 * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
11 * * Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * * Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in
15 * the documentation and/or other materials provided with the
17 * * Neither the name of Intel Corporation nor the names of its
18 * contributors may be used to endorse or promote products derived
19 * from this software without specific prior written permission.
21 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
22 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
23 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
24 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
25 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
26 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
27 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
28 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
29 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
30 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
31 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
34 #ifndef _VIRTIO_NET_H_
35 #define _VIRTIO_NET_H_
39 * Interface to vhost net
43 #include <linux/vhost.h>
44 #include <linux/virtio_ring.h>
45 #include <linux/virtio_net.h>
46 #include <sys/eventfd.h>
47 #include <sys/socket.h>
50 #include <rte_memory.h>
51 #include <rte_mempool.h>
52 #include <rte_ether.h>
56 #define VHOST_MEMORY_MAX_NREGIONS 8
58 /* Used to indicate that the device is running on a data core */
59 #define VIRTIO_DEV_RUNNING 1
61 /* Backend value set by guest. */
62 #define VIRTIO_DEV_STOPPED -1
65 /* Enum for virtqueue management. */
66 enum {VIRTIO_RXQ, VIRTIO_TXQ, VIRTIO_QNUM};
68 #define BUF_VECTOR_MAX 256
71 * Structure contains buffer address, length and descriptor index
72 * from vring to do scatter RX.
81 * Structure contains variables relevant to RX/TX virtqueues.
83 struct vhost_virtqueue {
84 struct vring_desc *desc; /**< Virtqueue descriptor ring. */
85 struct vring_avail *avail; /**< Virtqueue available ring. */
86 struct vring_used *used; /**< Virtqueue used ring. */
87 uint32_t size; /**< Size of descriptor ring. */
88 int backend; /**< Backend value to determine if device should started/stopped. */
89 uint16_t vhost_hlen; /**< Vhost header length (varies depending on RX merge buffers. */
90 volatile uint16_t last_used_idx; /**< Last index used on the available ring */
91 volatile uint16_t last_used_idx_res; /**< Used for multiple devices reserving buffers. */
92 #define VIRTIO_INVALID_EVENTFD (-1)
93 #define VIRTIO_UNINITIALIZED_EVENTFD (-2)
94 int callfd; /**< Used to notify the guest (trigger interrupt). */
95 int kickfd; /**< Currently unused as polling mode is enabled. */
97 uint64_t log_guest_addr; /**< Physical address of used ring, for logging */
98 uint64_t reserved[15]; /**< Reserve some spaces for future extension. */
99 struct buf_vector buf_vec[BUF_VECTOR_MAX]; /**< for scatter RX. */
100 } __rte_cache_aligned;
102 /* Old kernels have no such macro defined */
103 #ifndef VIRTIO_NET_F_GUEST_ANNOUNCE
104 #define VIRTIO_NET_F_GUEST_ANNOUNCE 21
109 * Make an extra wrapper for VIRTIO_NET_F_MQ and
110 * VIRTIO_NET_CTRL_MQ_VQ_PAIRS_MAX as they are
111 * introduced since kernel v3.8. This makes our
112 * code buildable for older kernel.
114 #ifdef VIRTIO_NET_F_MQ
115 #define VHOST_MAX_QUEUE_PAIRS VIRTIO_NET_CTRL_MQ_VQ_PAIRS_MAX
116 #define VHOST_SUPPORTS_MQ (1ULL << VIRTIO_NET_F_MQ)
118 #define VHOST_MAX_QUEUE_PAIRS 1
119 #define VHOST_SUPPORTS_MQ 0
123 * Define virtio 1.0 for older kernels
125 #ifndef VIRTIO_F_VERSION_1
126 #define VIRTIO_F_VERSION_1 32
130 * Device structure contains all configuration information relating to the device.
133 struct virtio_memory *mem; /**< QEMU memory and memory region information. */
134 uint64_t features; /**< Negotiated feature set. */
135 uint64_t protocol_features; /**< Negotiated protocol feature set. */
136 int vid; /**< device identifier. */
137 uint32_t flags; /**< Device flags. Only used to check if device is running on data core. */
138 #define IF_NAME_SZ (PATH_MAX > IFNAMSIZ ? PATH_MAX : IFNAMSIZ)
139 char ifname[IF_NAME_SZ]; /**< Name of the tap device or socket path. */
140 uint32_t virt_qp_nb; /**< number of queue pair we have allocated */
141 void *priv; /**< private context */
142 uint64_t log_size; /**< Size of log area */
143 uint64_t log_base; /**< Where dirty pages are logged */
144 struct ether_addr mac; /**< MAC address */
145 rte_atomic16_t broadcast_rarp; /**< A flag to tell if we need broadcast rarp packet */
146 uint64_t reserved[61]; /**< Reserve some spaces for future extension. */
147 struct vhost_virtqueue *virtqueue[VHOST_MAX_QUEUE_PAIRS * 2]; /**< Contains all virtqueue information. */
148 } __rte_cache_aligned;
151 * Information relating to memory regions including offsets to addresses in QEMUs memory file.
153 struct virtio_memory_regions {
154 uint64_t guest_phys_address; /**< Base guest physical address of region. */
155 uint64_t guest_phys_address_end; /**< End guest physical address of region. */
156 uint64_t memory_size; /**< Size of region. */
157 uint64_t userspace_address; /**< Base userspace address of region. */
158 uint64_t address_offset; /**< Offset of region for address translation. */
163 * Memory structure includes region and mapping information.
165 struct virtio_memory {
166 uint64_t base_address; /**< Base QEMU userspace address of the memory file. */
167 uint64_t mapped_address; /**< Mapped address of memory file base in our applications memory space. */
168 uint64_t mapped_size; /**< Total size of memory file. */
169 uint32_t nregions; /**< Number of memory regions. */
170 struct virtio_memory_regions regions[0]; /**< Memory region information. */
174 * Device and vring operations.
176 * Make sure to set VIRTIO_DEV_RUNNING to the device flags in new_device and
177 * remove it in destroy_device.
180 struct virtio_net_device_ops {
181 int (*new_device)(struct virtio_net *); /**< Add device. */
182 void (*destroy_device)(volatile struct virtio_net *); /**< Remove device. */
184 int (*vring_state_changed)(struct virtio_net *dev, uint16_t queue_id, int enable); /**< triggered when a vring is enabled or disabled */
187 static inline uint16_t __attribute__((always_inline))
188 rte_vring_available_entries(struct virtio_net *dev, uint16_t queue_id)
190 struct vhost_virtqueue *vq = dev->virtqueue[queue_id];
195 return *(volatile uint16_t *)&vq->avail->idx - vq->last_used_idx_res;
199 * Function to convert guest physical addresses to vhost virtual addresses.
200 * This is used to convert guest virtio buffer addresses.
202 static inline uint64_t __attribute__((always_inline))
203 gpa_to_vva(struct virtio_net *dev, uint64_t guest_pa)
205 struct virtio_memory_regions *region;
207 uint64_t vhost_va = 0;
209 for (regionidx = 0; regionidx < dev->mem->nregions; regionidx++) {
210 region = &dev->mem->regions[regionidx];
211 if ((guest_pa >= region->guest_phys_address) &&
212 (guest_pa <= region->guest_phys_address_end)) {
213 vhost_va = region->address_offset + guest_pa;
222 * Disable features in feature_mask. Returns 0 on success.
224 int rte_vhost_feature_disable(uint64_t feature_mask);
227 * Enable features in feature_mask. Returns 0 on success.
229 int rte_vhost_feature_enable(uint64_t feature_mask);
231 /* Returns currently supported vhost features */
232 uint64_t rte_vhost_feature_get(void);
234 int rte_vhost_enable_guest_notification(struct virtio_net *dev, uint16_t queue_id, int enable);
236 /* Register vhost driver. dev_name could be different for multiple instance support. */
237 int rte_vhost_driver_register(const char *dev_name);
239 /* Unregister vhost driver. This is only meaningful to vhost user. */
240 int rte_vhost_driver_unregister(const char *dev_name);
242 /* Register callbacks. */
243 int rte_vhost_driver_callback_register(struct virtio_net_device_ops const * const);
244 /* Start vhost driver session blocking loop. */
245 int rte_vhost_driver_session_start(void);
248 * Get the numa node from which the virtio net device's memory
252 * virtio-net device ID
255 * The numa node, -1 on failure
257 int rte_vhost_get_numa_node(int vid);
260 * Get the number of queues the device supports.
263 * virtio-net device ID
266 * The number of queues, 0 on failure
268 uint32_t rte_vhost_get_queue_num(int vid);
271 * Get the virtio net device's ifname. For vhost-cuse, ifname is the
272 * path of the char device. For vhost-user, ifname is the vhost-user
276 * virtio-net device ID
278 * The buffer to stored the queried ifname
283 * 0 on success, -1 on failure
285 int rte_vhost_get_ifname(int vid, char *buf, size_t len);
288 * This function adds buffers to the virtio devices RX virtqueue. Buffers can
289 * be received from the physical port or from another virtual device. A packet
290 * count is returned to indicate the number of packets that were succesfully
291 * added to the RX queue.
295 * virtio queue index in mq case
297 * array to contain packets to be enqueued
299 * packets num to be enqueued
301 * num of packets enqueued
303 uint16_t rte_vhost_enqueue_burst(struct virtio_net *dev, uint16_t queue_id,
304 struct rte_mbuf **pkts, uint16_t count);
307 * This function gets guest buffers from the virtio device TX virtqueue,
308 * construct host mbufs, copies guest buffer content to host mbufs and
309 * store them in pkts to be processed.
313 * virtio queue index in mq case
315 * mbuf_pool where host mbuf is allocated.
317 * array to contain packets to be dequeued
319 * packets num to be dequeued
321 * num of packets dequeued
323 uint16_t rte_vhost_dequeue_burst(struct virtio_net *dev, uint16_t queue_id,
324 struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count);
326 #endif /* _VIRTIO_NET_H_ */