ceph/src/dpdk/lib/librte_vhost/vhost.h

   1 /*-
   2  *   BSD LICENSE
   3  *
   4  *   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
   5  *   All rights reserved.
   6  *
   7  *   Redistribution and use in source and binary forms, with or without
   8  *   modification, are permitted provided that the following conditions
   9  *   are met:
  10  *
  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
  16  *       distribution.
  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.
  20  *
  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.
  32  */
  33
  34 #ifndef _VHOST_NET_CDEV_H_
  35 #define _VHOST_NET_CDEV_H_
  36 #include <stdint.h>
  37 #include <stdio.h>
  38 #include <sys/types.h>
  39 #include <sys/queue.h>
  40 #include <unistd.h>
  41 #include <linux/vhost.h>
  42
  43 #include <rte_log.h>
  44
  45 #include "rte_virtio_net.h"
  46
  47 /* Used to indicate that the device is running on a data core */
  48 #define VIRTIO_DEV_RUNNING 1
  49
  50 /* Backend value set by guest. */
  51 #define VIRTIO_DEV_STOPPED -1
  52
  53 #define BUF_VECTOR_MAX 256
  54
  55 /**
  56  * Structure contains buffer address, length and descriptor index
  57  * from vring to do scatter RX.
  58  */
  59 struct buf_vector {
  60         uint64_t buf_addr;
  61         uint32_t buf_len;
  62         uint32_t desc_idx;
  63 };
  64
  65 /*
  66  * A structure to hold some fields needed in zero copy code path,
  67  * mainly for associating an mbuf with the right desc_idx.
  68  */
  69 struct zcopy_mbuf {
  70         struct rte_mbuf *mbuf;
  71         uint32_t desc_idx;
  72         uint16_t in_use;
  73
  74         TAILQ_ENTRY(zcopy_mbuf) next;
  75 };
  76 TAILQ_HEAD(zcopy_mbuf_list, zcopy_mbuf);
  77
  78 /**
  79  * Structure contains variables relevant to RX/TX virtqueues.
  80  */
  81 struct vhost_virtqueue {
  82         struct vring_desc       *desc;
  83         struct vring_avail      *avail;
  84         struct vring_used       *used;
  85         uint32_t                size;
  86
  87         uint16_t                last_avail_idx;
  88         uint16_t                last_used_idx;
  89 #define VIRTIO_INVALID_EVENTFD          (-1)
  90 #define VIRTIO_UNINITIALIZED_EVENTFD    (-2)
  91
  92         /* Backend value to determine if device should started/stopped */
  93         int                     backend;
  94         /* Used to notify the guest (trigger interrupt) */
  95         int                     callfd;
  96         /* Currently unused as polling mode is enabled */
  97         int                     kickfd;
  98         int                     enabled;
  99
 100         /* Physical address of used ring, for logging */
 101         uint64_t                log_guest_addr;
 102
 103         uint16_t                nr_zmbuf;
 104         uint16_t                zmbuf_size;
 105         uint16_t                last_zmbuf_idx;
 106         struct zcopy_mbuf       *zmbufs;
 107         struct zcopy_mbuf_list  zmbuf_list;
 108
 109         struct vring_used_elem  *shadow_used_ring;
 110         uint16_t                shadow_used_idx;
 111 } __rte_cache_aligned;
 112
 113 /* Old kernels have no such macro defined */
 114 #ifndef VIRTIO_NET_F_GUEST_ANNOUNCE
 115  #define VIRTIO_NET_F_GUEST_ANNOUNCE 21
 116 #endif
 117
 118
 119 /*
 120  * Make an extra wrapper for VIRTIO_NET_F_MQ and
 121  * VIRTIO_NET_CTRL_MQ_VQ_PAIRS_MAX as they are
 122  * introduced since kernel v3.8. This makes our
 123  * code buildable for older kernel.
 124  */
 125 #ifdef VIRTIO_NET_F_MQ
 126  #define VHOST_MAX_QUEUE_PAIRS  VIRTIO_NET_CTRL_MQ_VQ_PAIRS_MAX
 127  #define VHOST_SUPPORTS_MQ      (1ULL << VIRTIO_NET_F_MQ)
 128 #else
 129  #define VHOST_MAX_QUEUE_PAIRS  1
 130  #define VHOST_SUPPORTS_MQ      0
 131 #endif
 132
 133 /*
 134  * Define virtio 1.0 for older kernels
 135  */
 136 #ifndef VIRTIO_F_VERSION_1
 137  #define VIRTIO_F_VERSION_1 32
 138 #endif
 139
 140 struct guest_page {
 141         uint64_t guest_phys_addr;
 142         uint64_t host_phys_addr;
 143         uint64_t size;
 144 };
 145
 146 /**
 147  * Device structure contains all configuration information relating
 148  * to the device.
 149  */
 150 struct virtio_net {
 151         /* Frontend (QEMU) memory and memory region information */
 152         struct virtio_memory    *mem;
 153         uint64_t                features;
 154         uint64_t                protocol_features;
 155         int                     vid;
 156         uint32_t                flags;
 157         uint16_t                vhost_hlen;
 158         /* to tell if we need broadcast rarp packet */
 159         rte_atomic16_t          broadcast_rarp;
 160         uint32_t                virt_qp_nb;
 161         int                     dequeue_zero_copy;
 162         struct vhost_virtqueue  *virtqueue[VHOST_MAX_QUEUE_PAIRS * 2];
 163 #define IF_NAME_SZ (PATH_MAX > IFNAMSIZ ? PATH_MAX : IFNAMSIZ)
 164         char                    ifname[IF_NAME_SZ];
 165         uint64_t                log_size;
 166         uint64_t                log_base;
 167         uint64_t                log_addr;
 168         struct ether_addr       mac;
 169
 170         uint32_t                nr_guest_pages;
 171         uint32_t                max_guest_pages;
 172         struct guest_page       *guest_pages;
 173 } __rte_cache_aligned;
 174
 175 /**
 176  * Information relating to memory regions including offsets to
 177  * addresses in QEMUs memory file.
 178  */
 179 struct virtio_memory_region {
 180         uint64_t guest_phys_addr;
 181         uint64_t guest_user_addr;
 182         uint64_t host_user_addr;
 183         uint64_t size;
 184         void     *mmap_addr;
 185         uint64_t mmap_size;
 186         int fd;
 187 };
 188
 189
 190 /**
 191  * Memory structure includes region and mapping information.
 192  */
 193 struct virtio_memory {
 194         uint32_t nregions;
 195         struct virtio_memory_region regions[0];
 196 };
 197
 198
 199 /* Macros for printing using RTE_LOG */
 200 #define RTE_LOGTYPE_VHOST_CONFIG RTE_LOGTYPE_USER1
 201 #define RTE_LOGTYPE_VHOST_DATA   RTE_LOGTYPE_USER1
 202
 203 #ifdef RTE_LIBRTE_VHOST_DEBUG
 204 #define VHOST_MAX_PRINT_BUFF 6072
 205 #define LOG_LEVEL RTE_LOG_DEBUG
 206 #define LOG_DEBUG(log_type, fmt, args...) RTE_LOG(DEBUG, log_type, fmt, ##args)
 207 #define PRINT_PACKET(device, addr, size, header) do { \
 208         char *pkt_addr = (char *)(addr); \
 209         unsigned int index; \
 210         char packet[VHOST_MAX_PRINT_BUFF]; \
 211         \
 212         if ((header)) \
 213                 snprintf(packet, VHOST_MAX_PRINT_BUFF, "(%d) Header size %d: ", (device->vid), (size)); \
 214         else \
 215                 snprintf(packet, VHOST_MAX_PRINT_BUFF, "(%d) Packet size %d: ", (device->vid), (size)); \
 216         for (index = 0; index < (size); index++) { \
 217                 snprintf(packet + strnlen(packet, VHOST_MAX_PRINT_BUFF), VHOST_MAX_PRINT_BUFF - strnlen(packet, VHOST_MAX_PRINT_BUFF), \
 218                         "%02hhx ", pkt_addr[index]); \
 219         } \
 220         snprintf(packet + strnlen(packet, VHOST_MAX_PRINT_BUFF), VHOST_MAX_PRINT_BUFF - strnlen(packet, VHOST_MAX_PRINT_BUFF), "\n"); \
 221         \
 222         LOG_DEBUG(VHOST_DATA, "%s", packet); \
 223 } while (0)
 224 #else
 225 #define LOG_LEVEL RTE_LOG_INFO
 226 #define LOG_DEBUG(log_type, fmt, args...) do {} while (0)
 227 #define PRINT_PACKET(device, addr, size, header) do {} while (0)
 228 #endif
 229
 230 extern uint64_t VHOST_FEATURES;
 231 #define MAX_VHOST_DEVICE        1024
 232 extern struct virtio_net *vhost_devices[MAX_VHOST_DEVICE];
 233
 234 /* Convert guest physical Address to host virtual address */
 235 static inline uint64_t __attribute__((always_inline))
 236 gpa_to_vva(struct virtio_net *dev, uint64_t gpa)
 237 {
 238         struct virtio_memory_region *reg;
 239         uint32_t i;
 240
 241         for (i = 0; i < dev->mem->nregions; i++) {
 242                 reg = &dev->mem->regions[i];
 243                 if (gpa >= reg->guest_phys_addr &&
 244                     gpa <  reg->guest_phys_addr + reg->size) {
 245                         return gpa - reg->guest_phys_addr +
 246                                reg->host_user_addr;
 247                 }
 248         }
 249
 250         return 0;
 251 }
 252
 253 /* Convert guest physical address to host physical address */
 254 static inline phys_addr_t __attribute__((always_inline))
 255 gpa_to_hpa(struct virtio_net *dev, uint64_t gpa, uint64_t size)
 256 {
 257         uint32_t i;
 258         struct guest_page *page;
 259
 260         for (i = 0; i < dev->nr_guest_pages; i++) {
 261                 page = &dev->guest_pages[i];
 262
 263                 if (gpa >= page->guest_phys_addr &&
 264                     gpa + size < page->guest_phys_addr + page->size) {
 265                         return gpa - page->guest_phys_addr +
 266                                page->host_phys_addr;
 267                 }
 268         }
 269
 270         return 0;
 271 }
 272
 273 struct virtio_net_device_ops const *notify_ops;
 274 struct virtio_net *get_device(int vid);
 275
 276 int vhost_new_device(void);
 277 void cleanup_device(struct virtio_net *dev, int destroy);
 278 void reset_device(struct virtio_net *dev);
 279 void vhost_destroy_device(int);
 280
 281 int alloc_vring_queue_pair(struct virtio_net *dev, uint32_t qp_idx);
 282
 283 void vhost_set_ifname(int, const char *if_name, unsigned int if_len);
 284 void vhost_enable_dequeue_zero_copy(int vid);
 285
 286 /*
 287  * Backend-specific cleanup.
 288  *
 289  * TODO: fix it; we have one backend now
 290  */
 291 void vhost_backend_cleanup(struct virtio_net *dev);
 292
 293 #endif /* _VHOST_NET_CDEV_H_ */