docs/specs/vhost-user.txt

   1 Vhost-user Protocol
   2 ===================
   3
   4 Copyright (c) 2014 Virtual Open Systems Sarl.
   5
   6 This work is licensed under the terms of the GNU GPL, version 2 or later.
   7 See the COPYING file in the top-level directory.
   8 ===================
   9
  10 This protocol is aiming to complement the ioctl interface used to control the
  11 vhost implementation in the Linux kernel. It implements the control plane needed
  12 to establish virtqueue sharing with a user space process on the same host. It
  13 uses communication over a Unix domain socket to share file descriptors in the
  14 ancillary data of the message.
  15
  16 The protocol defines 2 sides of the communication, master and slave. Master is
  17 the application that shares its virtqueues, in our case QEMU. Slave is the
  18 consumer of the virtqueues.
  19
  20 In the current implementation QEMU is the Master, and the Slave is intended to
  21 be a software Ethernet switch running in user space, such as Snabbswitch.
  22
  23 Master and slave can be either a client (i.e. connecting) or server (listening)
  24 in the socket communication.
  25
  26 Message Specification
  27 ---------------------
  28
  29 Note that all numbers are in the machine native byte order. A vhost-user message
  30 consists of 3 header fields and a payload:
  31
  32 ------------------------------------
  33 | request | flags | size | payload |
  34 ------------------------------------
  35
  36  * Request: 32-bit type of the request
  37  * Flags: 32-bit bit field:
  38    - Lower 2 bits are the version (currently 0x01)
  39    - Bit 2 is the reply flag - needs to be sent on each reply from the slave
  40  * Size - 32-bit size of the payload
  41
  42
  43 Depending on the request type, payload can be:
  44
  45  * A single 64-bit integer
  46    -------
  47    | u64 |
  48    -------
  49
  50    u64: a 64-bit unsigned integer
  51
  52  * A vring state description
  53    ---------------
  54   | index | num |
  55   ---------------
  56
  57    Index: a 32-bit index
  58    Num: a 32-bit number
  59
  60  * A vring address description
  61    --------------------------------------------------------------
  62    | index | flags | size | descriptor | used | available | log |
  63    --------------------------------------------------------------
  64
  65    Index: a 32-bit vring index
  66    Flags: a 32-bit vring flags
  67    Descriptor: a 64-bit user address of the vring descriptor table
  68    Used: a 64-bit user address of the vring used ring
  69    Available: a 64-bit user address of the vring available ring
  70    Log: a 64-bit guest address for logging
  71
  72  * Memory regions description
  73    ---------------------------------------------------
  74    | num regions | padding | region0 | ... | region7 |
  75    ---------------------------------------------------
  76
  77    Num regions: a 32-bit number of regions
  78    Padding: 32-bit
  79
  80    A region is:
  81    -----------------------------------------------------
  82    | guest address | size | user address | mmap offset |
  83    -----------------------------------------------------
  84
  85    Guest address: a 64-bit guest address of the region
  86    Size: a 64-bit size
  87    User address: a 64-bit user address
  88    mmap offset: 64-bit offset where region starts in the mapped memory
  89
  90 In QEMU the vhost-user message is implemented with the following struct:
  91
  92 typedef struct VhostUserMsg {
  93     VhostUserRequest request;
  94     uint32_t flags;
  95     uint32_t size;
  96     union {
  97         uint64_t u64;
  98         struct vhost_vring_state state;
  99         struct vhost_vring_addr addr;
 100         VhostUserMemory memory;
 101     };
 102 } QEMU_PACKED VhostUserMsg;
 103
 104 Communication
 105 -------------
 106
 107 The protocol for vhost-user is based on the existing implementation of vhost
 108 for the Linux Kernel. Most messages that can be sent via the Unix domain socket
 109 implementing vhost-user have an equivalent ioctl to the kernel implementation.
 110
 111 The communication consists of master sending message requests and slave sending
 112 message replies. Most of the requests don't require replies. Here is a list of
 113 the ones that do:
 114
 115  * VHOST_GET_FEATURES
 116  * VHOST_GET_PROTOCOL_FEATURES
 117  * VHOST_GET_VRING_BASE
 118  * VHOST_SET_LOG_BASE (if VHOST_USER_PROTOCOL_F_LOG_SHMFD)
 119
 120 There are several messages that the master sends with file descriptors passed
 121 in the ancillary data:
 122
 123  * VHOST_SET_MEM_TABLE
 124  * VHOST_SET_LOG_BASE (if VHOST_USER_PROTOCOL_F_LOG_SHMFD)
 125  * VHOST_SET_LOG_FD
 126  * VHOST_SET_VRING_KICK
 127  * VHOST_SET_VRING_CALL
 128  * VHOST_SET_VRING_ERR
 129
 130 If Master is unable to send the full message or receives a wrong reply it will
 131 close the connection. An optional reconnection mechanism can be implemented.
 132
 133 Any protocol extensions are gated by protocol feature bits,
 134 which allows full backwards compatibility on both master
 135 and slave.
 136 As older slaves don't support negotiating protocol features,
 137 a feature bit was dedicated for this purpose:
 138 #define VHOST_USER_F_PROTOCOL_FEATURES 30
 139
 140 Multiple queue support
 141 ----------------------
 142
 143 Multiple queue is treated as a protocol extension, hence the slave has to
 144 implement protocol features first. The multiple queues feature is supported
 145 only when the protocol feature VHOST_USER_PROTOCOL_F_MQ (bit 0) is set.
 146
 147 The max number of queues the slave supports can be queried with message
 148 VHOST_USER_GET_PROTOCOL_FEATURES. Master should stop when the number of
 149 requested queues is bigger than that.
 150
 151 As all queues share one connection, the master uses a unique index for each
 152 queue in the sent message to identify a specified queue. One queue pair
 153 is enabled initially. More queues are enabled dynamically, by sending
 154 message VHOST_USER_SET_VRING_ENABLE.
 155
 156 Migration
 157 ---------
 158
 159 During live migration, the master may need to track the modifications
 160 the slave makes to the memory mapped regions. The client should mark
 161 the dirty pages in a log. Once it complies to this logging, it may
 162 declare the VHOST_F_LOG_ALL vhost feature.
 163
 164 All the modifications to memory pointed by vring "descriptor" should
 165 be marked. Modifications to "used" vring should be marked if
 166 VHOST_VRING_F_LOG is part of ring's features.
 167
 168 Dirty pages are of size:
 169 #define VHOST_LOG_PAGE 0x1000
 170
 171 The log memory fd is provided in the ancillary data of
 172 VHOST_USER_SET_LOG_BASE message when the slave has
 173 VHOST_USER_PROTOCOL_F_LOG_SHMFD protocol feature.
 174
 175 The size of the log may be computed by using all the known guest
 176 addresses. The log covers from address 0 to the maximum of guest
 177 regions. In pseudo-code, to mark page at "addr" as dirty:
 178
 179 page = addr / VHOST_LOG_PAGE
 180 log[page / 8] |= 1 << page % 8
 181
 182 Use atomic operations, as the log may be concurrently manipulated.
 183
 184 VHOST_USER_SET_LOG_FD is an optional message with an eventfd in
 185 ancillary data, it may be used to inform the master that the log has
 186 been modified.
 187
 188 Once the source has finished migration, VHOST_USER_RESET_OWNER message
 189 will be sent by the source. No further update must be done before the
 190 destination takes over with new regions & rings.
 191
 192 Protocol features
 193 -----------------
 194
 195 #define VHOST_USER_PROTOCOL_F_MQ             0
 196 #define VHOST_USER_PROTOCOL_F_LOG_SHMFD      1
 197
 198 Message types
 199 -------------
 200
 201  * VHOST_USER_GET_FEATURES
 202
 203       Id: 1
 204       Equivalent ioctl: VHOST_GET_FEATURES
 205       Master payload: N/A
 206       Slave payload: u64
 207
 208       Get from the underlying vhost implementation the features bitmask.
 209       Feature bit VHOST_USER_F_PROTOCOL_FEATURES signals slave support for
 210       VHOST_USER_GET_PROTOCOL_FEATURES and VHOST_USER_SET_PROTOCOL_FEATURES.
 211
 212  * VHOST_USER_SET_FEATURES
 213
 214       Id: 2
 215       Ioctl: VHOST_SET_FEATURES
 216       Master payload: u64
 217
 218       Enable features in the underlying vhost implementation using a bitmask.
 219       Feature bit VHOST_USER_F_PROTOCOL_FEATURES signals slave support for
 220       VHOST_USER_GET_PROTOCOL_FEATURES and VHOST_USER_SET_PROTOCOL_FEATURES.
 221
 222  * VHOST_USER_GET_PROTOCOL_FEATURES
 223
 224       Id: 15
 225       Equivalent ioctl: VHOST_GET_FEATURES
 226       Master payload: N/A
 227       Slave payload: u64
 228
 229       Get the protocol feature bitmask from the underlying vhost implementation.
 230       Only legal if feature bit VHOST_USER_F_PROTOCOL_FEATURES is present in
 231       VHOST_USER_GET_FEATURES.
 232       Note: slave that reported VHOST_USER_F_PROTOCOL_FEATURES must support
 233       this message even before VHOST_USER_SET_FEATURES was called.
 234
 235  * VHOST_USER_SET_PROTOCOL_FEATURES
 236
 237       Id: 16
 238       Ioctl: VHOST_SET_FEATURES
 239       Master payload: u64
 240
 241       Enable protocol features in the underlying vhost implementation.
 242       Only legal if feature bit VHOST_USER_F_PROTOCOL_FEATURES is present in
 243       VHOST_USER_GET_FEATURES.
 244       Note: slave that reported VHOST_USER_F_PROTOCOL_FEATURES must support
 245       this message even before VHOST_USER_SET_FEATURES was called.
 246
 247  * VHOST_USER_SET_OWNER
 248
 249       Id: 3
 250       Equivalent ioctl: VHOST_SET_OWNER
 251       Master payload: N/A
 252
 253       Issued when a new connection is established. It sets the current Master
 254       as an owner of the session. This can be used on the Slave as a
 255       "session start" flag.
 256
 257  * VHOST_USER_RESET_DEVICE
 258
 259       Id: 4
 260       Equivalent ioctl: VHOST_RESET_DEVICE
 261       Master payload: N/A
 262
 263       Issued when a new connection is about to be closed. The Master will no
 264       longer own this connection (and will usually close it).
 265
 266  * VHOST_USER_SET_MEM_TABLE
 267
 268       Id: 5
 269       Equivalent ioctl: VHOST_SET_MEM_TABLE
 270       Master payload: memory regions description
 271
 272       Sets the memory map regions on the slave so it can translate the vring
 273       addresses. In the ancillary data there is an array of file descriptors
 274       for each memory mapped region. The size and ordering of the fds matches
 275       the number and ordering of memory regions.
 276
 277  * VHOST_USER_SET_LOG_BASE
 278
 279       Id: 6
 280       Equivalent ioctl: VHOST_SET_LOG_BASE
 281       Master payload: u64
 282       Slave payload: N/A
 283
 284       Sets the logging base address.
 285
 286  * VHOST_USER_SET_LOG_FD
 287
 288       Id: 7
 289       Equivalent ioctl: VHOST_SET_LOG_FD
 290       Master payload: N/A
 291
 292       Sets the logging file descriptor, which is passed as ancillary data.
 293
 294  * VHOST_USER_SET_VRING_NUM
 295
 296       Id: 8
 297       Equivalent ioctl: VHOST_SET_VRING_NUM
 298       Master payload: vring state description
 299
 300       Sets the number of vrings for this owner.
 301
 302  * VHOST_USER_SET_VRING_ADDR
 303
 304       Id: 9
 305       Equivalent ioctl: VHOST_SET_VRING_ADDR
 306       Master payload: vring address description
 307       Slave payload: N/A
 308
 309       Sets the addresses of the different aspects of the vring.
 310
 311  * VHOST_USER_SET_VRING_BASE
 312
 313       Id: 10
 314       Equivalent ioctl: VHOST_SET_VRING_BASE
 315       Master payload: vring state description
 316
 317       Sets the base offset in the available vring.
 318
 319  * VHOST_USER_GET_VRING_BASE
 320
 321       Id: 11
 322       Equivalent ioctl: VHOST_USER_GET_VRING_BASE
 323       Master payload: vring state description
 324       Slave payload: vring state description
 325
 326       Get the available vring base offset.
 327
 328  * VHOST_USER_SET_VRING_KICK
 329
 330       Id: 12
 331       Equivalent ioctl: VHOST_SET_VRING_KICK
 332       Master payload: u64
 333
 334       Set the event file descriptor for adding buffers to the vring. It
 335       is passed in the ancillary data.
 336       Bits (0-7) of the payload contain the vring index. Bit 8 is the
 337       invalid FD flag. This flag is set when there is no file descriptor
 338       in the ancillary data. This signals that polling should be used
 339       instead of waiting for a kick.
 340
 341  * VHOST_USER_SET_VRING_CALL
 342
 343       Id: 13
 344       Equivalent ioctl: VHOST_SET_VRING_CALL
 345       Master payload: u64
 346
 347       Set the event file descriptor to signal when buffers are used. It
 348       is passed in the ancillary data.
 349       Bits (0-7) of the payload contain the vring index. Bit 8 is the
 350       invalid FD flag. This flag is set when there is no file descriptor
 351       in the ancillary data. This signals that polling will be used
 352       instead of waiting for the call.
 353
 354  * VHOST_USER_SET_VRING_ERR
 355
 356       Id: 14
 357       Equivalent ioctl: VHOST_SET_VRING_ERR
 358       Master payload: u64
 359
 360       Set the event file descriptor to signal when error occurs. It
 361       is passed in the ancillary data.
 362       Bits (0-7) of the payload contain the vring index. Bit 8 is the
 363       invalid FD flag. This flag is set when there is no file descriptor
 364       in the ancillary data.
 365
 366  * VHOST_USER_GET_QUEUE_NUM
 367
 368       Id: 17
 369       Equivalent ioctl: N/A
 370       Master payload: N/A
 371       Slave payload: u64
 372
 373       Query how many queues the backend supports. This request should be
 374       sent only when VHOST_USER_PROTOCOL_F_MQ is set in quried protocol
 375       features by VHOST_USER_GET_PROTOCOL_FEATURES.
 376
 377  * VHOST_USER_SET_VRING_ENABLE
 378
 379       Id: 18
 380       Equivalent ioctl: N/A
 381       Master payload: vring state description
 382
 383       Signal slave to enable or disable corresponding vring.