vhost-user.c 43.8 KB
Newer Older
1 2 3 4 5 6 7 8 9 10
/*
 * vhost-user
 *
 * Copyright (c) 2013 Virtual Open Systems Sarl.
 *
 * This work is licensed under the terms of the GNU GPL, version 2 or later.
 * See the COPYING file in the top-level directory.
 *
 */

P
Peter Maydell 已提交
11
#include "qemu/osdep.h"
12
#include "qapi/error.h"
13 14
#include "hw/virtio/vhost.h"
#include "hw/virtio/vhost-backend.h"
15
#include "hw/virtio/virtio-net.h"
16
#include "chardev/char-fe.h"
17 18 19
#include "sysemu/kvm.h"
#include "qemu/error-report.h"
#include "qemu/sockets.h"
20
#include "sysemu/cryptodev.h"
21 22
#include "migration/migration.h"
#include "migration/postcopy-ram.h"
23
#include "trace.h"
24 25 26 27 28 29 30

#include <sys/ioctl.h>
#include <sys/socket.h>
#include <sys/un.h>
#include <linux/vhost.h>

#define VHOST_MEMORY_MAX_NREGIONS    8
31
#define VHOST_USER_F_PROTOCOL_FEATURES 30
32

33 34 35 36 37
/*
 * Maximum size of virtio device config space
 */
#define VHOST_USER_MAX_CONFIG_SIZE 256

38 39 40 41
enum VhostUserProtocolFeature {
    VHOST_USER_PROTOCOL_F_MQ = 0,
    VHOST_USER_PROTOCOL_F_LOG_SHMFD = 1,
    VHOST_USER_PROTOCOL_F_RARP = 2,
42
    VHOST_USER_PROTOCOL_F_REPLY_ACK = 3,
43
    VHOST_USER_PROTOCOL_F_NET_MTU = 4,
44
    VHOST_USER_PROTOCOL_F_SLAVE_REQ = 5,
45
    VHOST_USER_PROTOCOL_F_CROSS_ENDIAN = 6,
46
    VHOST_USER_PROTOCOL_F_CRYPTO_SESSION = 7,
47
    VHOST_USER_PROTOCOL_F_PAGEFAULT = 8,
48 49 50 51
    VHOST_USER_PROTOCOL_F_MAX
};

#define VHOST_USER_PROTOCOL_FEATURE_MASK ((1 << VHOST_USER_PROTOCOL_F_MAX) - 1)
52 53 54 55 56 57

typedef enum VhostUserRequest {
    VHOST_USER_NONE = 0,
    VHOST_USER_GET_FEATURES = 1,
    VHOST_USER_SET_FEATURES = 2,
    VHOST_USER_SET_OWNER = 3,
58
    VHOST_USER_RESET_OWNER = 4,
59 60 61 62 63 64 65 66 67 68
    VHOST_USER_SET_MEM_TABLE = 5,
    VHOST_USER_SET_LOG_BASE = 6,
    VHOST_USER_SET_LOG_FD = 7,
    VHOST_USER_SET_VRING_NUM = 8,
    VHOST_USER_SET_VRING_ADDR = 9,
    VHOST_USER_SET_VRING_BASE = 10,
    VHOST_USER_GET_VRING_BASE = 11,
    VHOST_USER_SET_VRING_KICK = 12,
    VHOST_USER_SET_VRING_CALL = 13,
    VHOST_USER_SET_VRING_ERR = 14,
69 70
    VHOST_USER_GET_PROTOCOL_FEATURES = 15,
    VHOST_USER_SET_PROTOCOL_FEATURES = 16,
71
    VHOST_USER_GET_QUEUE_NUM = 17,
72
    VHOST_USER_SET_VRING_ENABLE = 18,
73
    VHOST_USER_SEND_RARP = 19,
74
    VHOST_USER_NET_SET_MTU = 20,
75
    VHOST_USER_SET_SLAVE_REQ_FD = 21,
76
    VHOST_USER_IOTLB_MSG = 22,
77
    VHOST_USER_SET_VRING_ENDIAN = 23,
78 79
    VHOST_USER_GET_CONFIG = 24,
    VHOST_USER_SET_CONFIG = 25,
80 81
    VHOST_USER_CREATE_CRYPTO_SESSION = 26,
    VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
82
    VHOST_USER_POSTCOPY_ADVISE  = 28,
83
    VHOST_USER_POSTCOPY_LISTEN  = 29,
84 85 86
    VHOST_USER_MAX
} VhostUserRequest;

87 88
typedef enum VhostUserSlaveRequest {
    VHOST_USER_SLAVE_NONE = 0,
89
    VHOST_USER_SLAVE_IOTLB_MSG = 1,
90
    VHOST_USER_SLAVE_CONFIG_CHANGE_MSG = 2,
91 92 93
    VHOST_USER_SLAVE_MAX
}  VhostUserSlaveRequest;

94 95 96 97
typedef struct VhostUserMemoryRegion {
    uint64_t guest_phys_addr;
    uint64_t memory_size;
    uint64_t userspace_addr;
98
    uint64_t mmap_offset;
99 100 101 102 103 104 105 106
} VhostUserMemoryRegion;

typedef struct VhostUserMemory {
    uint32_t nregions;
    uint32_t padding;
    VhostUserMemoryRegion regions[VHOST_MEMORY_MAX_NREGIONS];
} VhostUserMemory;

107 108 109 110 111
typedef struct VhostUserLog {
    uint64_t mmap_size;
    uint64_t mmap_offset;
} VhostUserLog;

112 113 114 115 116 117 118
typedef struct VhostUserConfig {
    uint32_t offset;
    uint32_t size;
    uint32_t flags;
    uint8_t region[VHOST_USER_MAX_CONFIG_SIZE];
} VhostUserConfig;

119 120 121 122 123 124 125 126 127 128 129
#define VHOST_CRYPTO_SYM_HMAC_MAX_KEY_LEN    512
#define VHOST_CRYPTO_SYM_CIPHER_MAX_KEY_LEN  64

typedef struct VhostUserCryptoSession {
    /* session id for success, -1 on errors */
    int64_t session_id;
    CryptoDevBackendSymSessionInfo session_setup_data;
    uint8_t key[VHOST_CRYPTO_SYM_CIPHER_MAX_KEY_LEN];
    uint8_t auth_key[VHOST_CRYPTO_SYM_HMAC_MAX_KEY_LEN];
} VhostUserCryptoSession;

130 131 132 133 134
static VhostUserConfig c __attribute__ ((unused));
#define VHOST_USER_CONFIG_HDR_SIZE (sizeof(c.offset) \
                                   + sizeof(c.size) \
                                   + sizeof(c.flags))

135
typedef struct {
136 137 138 139
    VhostUserRequest request;

#define VHOST_USER_VERSION_MASK     (0x3)
#define VHOST_USER_REPLY_MASK       (0x1<<2)
140
#define VHOST_USER_NEED_REPLY_MASK  (0x1 << 3)
141 142
    uint32_t flags;
    uint32_t size; /* the following payload size */
143 144 145
} QEMU_PACKED VhostUserHeader;

typedef union {
146 147 148 149 150 151
#define VHOST_USER_VRING_IDX_MASK   (0xff)
#define VHOST_USER_VRING_NOFD_MASK  (0x1<<8)
        uint64_t u64;
        struct vhost_vring_state state;
        struct vhost_vring_addr addr;
        VhostUserMemory memory;
152
        VhostUserLog log;
153
        struct vhost_iotlb_msg iotlb;
154
        VhostUserConfig config;
155
        VhostUserCryptoSession session;
156 157 158 159 160
} VhostUserPayload;

typedef struct VhostUserMsg {
    VhostUserHeader hdr;
    VhostUserPayload payload;
161 162 163
} QEMU_PACKED VhostUserMsg;

static VhostUserMsg m __attribute__ ((unused));
164
#define VHOST_USER_HDR_SIZE (sizeof(VhostUserHeader))
165

166
#define VHOST_USER_PAYLOAD_SIZE (sizeof(VhostUserPayload))
167 168 169 170

/* The version of the protocol we support */
#define VHOST_USER_VERSION    (0x1)

171
struct vhost_user {
172
    struct vhost_dev *dev;
173
    CharBackend *chr;
174
    int slave_fd;
175
    NotifierWithReturn postcopy_notifier;
176
    struct PostCopyFD  postcopy_fd;
177
    uint64_t           postcopy_client_bases[VHOST_MEMORY_MAX_NREGIONS];
178 179 180 181 182 183 184 185 186
    /* Length of the region_rb and region_rb_offset arrays */
    size_t             region_rb_len;
    /* RAMBlock associated with a given region */
    RAMBlock         **region_rb;
    /* The offset from the start of the RAMBlock to the start of the
     * vhost region.
     */
    ram_addr_t        *region_rb_offset;

187 188
    /* True once we've entered postcopy_listen */
    bool               postcopy_listen;
189 190
};

191 192 193 194 195 196 197
static bool ioeventfd_enabled(void)
{
    return kvm_enabled() && kvm_eventfds_enabled();
}

static int vhost_user_read(struct vhost_dev *dev, VhostUserMsg *msg)
{
198 199
    struct vhost_user *u = dev->opaque;
    CharBackend *chr = u->chr;
200 201 202 203 204
    uint8_t *p = (uint8_t *) msg;
    int r, size = VHOST_USER_HDR_SIZE;

    r = qemu_chr_fe_read_all(chr, p, size);
    if (r != size) {
205
        error_report("Failed to read msg header. Read %d instead of %d."
206
                     " Original request %d.", r, size, msg->hdr.request);
207 208 209 210
        goto fail;
    }

    /* validate received flags */
211
    if (msg->hdr.flags != (VHOST_USER_REPLY_MASK | VHOST_USER_VERSION)) {
212
        error_report("Failed to read msg header."
213
                " Flags 0x%x instead of 0x%x.", msg->hdr.flags,
214 215 216 217 218
                VHOST_USER_REPLY_MASK | VHOST_USER_VERSION);
        goto fail;
    }

    /* validate message size is sane */
219
    if (msg->hdr.size > VHOST_USER_PAYLOAD_SIZE) {
220
        error_report("Failed to read msg header."
221
                " Size %d exceeds the maximum %zu.", msg->hdr.size,
222 223 224 225
                VHOST_USER_PAYLOAD_SIZE);
        goto fail;
    }

226
    if (msg->hdr.size) {
227
        p += VHOST_USER_HDR_SIZE;
228
        size = msg->hdr.size;
229 230 231
        r = qemu_chr_fe_read_all(chr, p, size);
        if (r != size) {
            error_report("Failed to read msg payload."
232
                         " Read %d instead of %d.", r, msg->hdr.size);
233 234 235 236 237 238 239 240 241 242
            goto fail;
        }
    }

    return 0;

fail:
    return -1;
}

243
static int process_message_reply(struct vhost_dev *dev,
244
                                 const VhostUserMsg *msg)
245
{
246
    VhostUserMsg msg_reply;
247

248
    if ((msg->hdr.flags & VHOST_USER_NEED_REPLY_MASK) == 0) {
249 250 251 252
        return 0;
    }

    if (vhost_user_read(dev, &msg_reply) < 0) {
253 254 255
        return -1;
    }

256
    if (msg_reply.hdr.request != msg->hdr.request) {
257 258
        error_report("Received unexpected msg type."
                     "Expected %d received %d",
259
                     msg->hdr.request, msg_reply.hdr.request);
260 261 262
        return -1;
    }

263
    return msg_reply.payload.u64 ? -1 : 0;
264 265
}

266 267 268 269
static bool vhost_user_one_time_request(VhostUserRequest request)
{
    switch (request) {
    case VHOST_USER_SET_OWNER:
270
    case VHOST_USER_RESET_OWNER:
271 272
    case VHOST_USER_SET_MEM_TABLE:
    case VHOST_USER_GET_QUEUE_NUM:
273
    case VHOST_USER_NET_SET_MTU:
274 275 276 277 278 279 280
        return true;
    default:
        return false;
    }
}

/* most non-init callers ignore the error */
281 282 283
static int vhost_user_write(struct vhost_dev *dev, VhostUserMsg *msg,
                            int *fds, int fd_num)
{
284 285
    struct vhost_user *u = dev->opaque;
    CharBackend *chr = u->chr;
286
    int ret, size = VHOST_USER_HDR_SIZE + msg->hdr.size;
287

288 289 290 291 292
    /*
     * For non-vring specific requests, like VHOST_USER_SET_MEM_TABLE,
     * we just need send it once in the first time. For later such
     * request, we just ignore it.
     */
293 294
    if (vhost_user_one_time_request(msg->hdr.request) && dev->vq_index != 0) {
        msg->hdr.flags &= ~VHOST_USER_NEED_REPLY_MASK;
295 296 297
        return 0;
    }

298
    if (qemu_chr_fe_set_msgfds(chr, fds, fd_num) < 0) {
299
        error_report("Failed to set msg fds.");
300 301
        return -1;
    }
302

303 304 305 306 307 308 309 310
    ret = qemu_chr_fe_write_all(chr, (const uint8_t *) msg, size);
    if (ret != size) {
        error_report("Failed to write msg."
                     " Wrote %d instead of %d.", ret, size);
        return -1;
    }

    return 0;
311 312
}

313 314
static int vhost_user_set_log_base(struct vhost_dev *dev, uint64_t base,
                                   struct vhost_log *log)
315
{
316 317 318 319 320
    int fds[VHOST_MEMORY_MAX_NREGIONS];
    size_t fd_num = 0;
    bool shmfd = virtio_has_feature(dev->protocol_features,
                                    VHOST_USER_PROTOCOL_F_LOG_SHMFD);
    VhostUserMsg msg = {
321 322
        .hdr.request = VHOST_USER_SET_LOG_BASE,
        .hdr.flags = VHOST_USER_VERSION,
M
Michael S. Tsirkin 已提交
323
        .payload.log.mmap_size = log->size * sizeof(*(log->log)),
324
        .payload.log.mmap_offset = 0,
325
        .hdr.size = sizeof(msg.payload.log),
326 327 328 329 330 331
    };

    if (shmfd && log->fd != -1) {
        fds[fd_num++] = log->fd;
    }

332 333 334
    if (vhost_user_write(dev, &msg, fds, fd_num) < 0) {
        return -1;
    }
335 336

    if (shmfd) {
337
        msg.hdr.size = 0;
338
        if (vhost_user_read(dev, &msg) < 0) {
339
            return -1;
340 341
        }

342
        if (msg.hdr.request != VHOST_USER_SET_LOG_BASE) {
343 344
            error_report("Received unexpected msg type. "
                         "Expected %d received %d",
345
                         VHOST_USER_SET_LOG_BASE, msg.hdr.request);
346 347
            return -1;
        }
348
    }
349 350

    return 0;
351 352
}

353 354 355
static int vhost_user_set_mem_table_postcopy(struct vhost_dev *dev,
                                             struct vhost_memory *mem)
{
356
    struct vhost_user *u = dev->opaque;
357 358 359 360 361
    int fds[VHOST_MEMORY_MAX_NREGIONS];
    int i, fd;
    size_t fd_num = 0;
    bool reply_supported = virtio_has_feature(dev->protocol_features,
                                              VHOST_USER_PROTOCOL_F_REPLY_ACK);
362 363 364
    VhostUserMsg msg_reply;
    int region_i, msg_i;

365 366 367 368 369 370 371 372 373
    VhostUserMsg msg = {
        .hdr.request = VHOST_USER_SET_MEM_TABLE,
        .hdr.flags = VHOST_USER_VERSION,
    };

    if (reply_supported) {
        msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK;
    }

374 375 376 377 378 379 380 381 382 383 384
    if (u->region_rb_len < dev->mem->nregions) {
        u->region_rb = g_renew(RAMBlock*, u->region_rb, dev->mem->nregions);
        u->region_rb_offset = g_renew(ram_addr_t, u->region_rb_offset,
                                      dev->mem->nregions);
        memset(&(u->region_rb[u->region_rb_len]), '\0',
               sizeof(RAMBlock *) * (dev->mem->nregions - u->region_rb_len));
        memset(&(u->region_rb_offset[u->region_rb_len]), '\0',
               sizeof(ram_addr_t) * (dev->mem->nregions - u->region_rb_len));
        u->region_rb_len = dev->mem->nregions;
    }

385 386 387 388 389 390 391 392 393 394
    for (i = 0; i < dev->mem->nregions; ++i) {
        struct vhost_memory_region *reg = dev->mem->regions + i;
        ram_addr_t offset;
        MemoryRegion *mr;

        assert((uintptr_t)reg->userspace_addr == reg->userspace_addr);
        mr = memory_region_from_host((void *)(uintptr_t)reg->userspace_addr,
                                     &offset);
        fd = memory_region_get_fd(mr);
        if (fd > 0) {
395 396 397 398 399 400
            trace_vhost_user_set_mem_table_withfd(fd_num, mr->name,
                                                  reg->memory_size,
                                                  reg->guest_phys_addr,
                                                  reg->userspace_addr, offset);
            u->region_rb_offset[i] = offset;
            u->region_rb[i] = mr->ram_block;
401 402 403 404 405 406 407 408
            msg.payload.memory.regions[fd_num].userspace_addr =
                reg->userspace_addr;
            msg.payload.memory.regions[fd_num].memory_size  = reg->memory_size;
            msg.payload.memory.regions[fd_num].guest_phys_addr =
                reg->guest_phys_addr;
            msg.payload.memory.regions[fd_num].mmap_offset = offset;
            assert(fd_num < VHOST_MEMORY_MAX_NREGIONS);
            fds[fd_num++] = fd;
409 410 411
        } else {
            u->region_rb_offset[i] = 0;
            u->region_rb[i] = NULL;
412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430
        }
    }

    msg.payload.memory.nregions = fd_num;

    if (!fd_num) {
        error_report("Failed initializing vhost-user memory map, "
                     "consider using -object memory-backend-file share=on");
        return -1;
    }

    msg.hdr.size = sizeof(msg.payload.memory.nregions);
    msg.hdr.size += sizeof(msg.payload.memory.padding);
    msg.hdr.size += fd_num * sizeof(VhostUserMemoryRegion);

    if (vhost_user_write(dev, &msg, fds, fd_num) < 0) {
        return -1;
    }

431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488
    if (vhost_user_read(dev, &msg_reply) < 0) {
        return -1;
    }

    if (msg_reply.hdr.request != VHOST_USER_SET_MEM_TABLE) {
        error_report("%s: Received unexpected msg type."
                     "Expected %d received %d", __func__,
                     VHOST_USER_SET_MEM_TABLE, msg_reply.hdr.request);
        return -1;
    }
    /* We're using the same structure, just reusing one of the
     * fields, so it should be the same size.
     */
    if (msg_reply.hdr.size != msg.hdr.size) {
        error_report("%s: Unexpected size for postcopy reply "
                     "%d vs %d", __func__, msg_reply.hdr.size, msg.hdr.size);
        return -1;
    }

    memset(u->postcopy_client_bases, 0,
           sizeof(uint64_t) * VHOST_MEMORY_MAX_NREGIONS);

    /* They're in the same order as the regions that were sent
     * but some of the regions were skipped (above) if they
     * didn't have fd's
    */
    for (msg_i = 0, region_i = 0;
         region_i < dev->mem->nregions;
        region_i++) {
        if (msg_i < fd_num &&
            msg_reply.payload.memory.regions[msg_i].guest_phys_addr ==
            dev->mem->regions[region_i].guest_phys_addr) {
            u->postcopy_client_bases[region_i] =
                msg_reply.payload.memory.regions[msg_i].userspace_addr;
            trace_vhost_user_set_mem_table_postcopy(
                msg_reply.payload.memory.regions[msg_i].userspace_addr,
                msg.payload.memory.regions[msg_i].userspace_addr,
                msg_i, region_i);
            msg_i++;
        }
    }
    if (msg_i != fd_num) {
        error_report("%s: postcopy reply not fully consumed "
                     "%d vs %zd",
                     __func__, msg_i, fd_num);
        return -1;
    }
    /* Now we've registered this with the postcopy code, we ack to the client,
     * because now we're in the position to be able to deal with any faults
     * it generates.
     */
    /* TODO: Use this for failure cases as well with a bad value */
    msg.hdr.size = sizeof(msg.payload.u64);
    msg.payload.u64 = 0; /* OK */
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }

489 490 491 492 493 494 495
    if (reply_supported) {
        return process_message_reply(dev, &msg);
    }

    return 0;
}

496 497 498
static int vhost_user_set_mem_table(struct vhost_dev *dev,
                                    struct vhost_memory *mem)
{
499
    struct vhost_user *u = dev->opaque;
500 501 502
    int fds[VHOST_MEMORY_MAX_NREGIONS];
    int i, fd;
    size_t fd_num = 0;
503
    bool do_postcopy = u->postcopy_listen && u->postcopy_fd.handler;
504
    bool reply_supported = virtio_has_feature(dev->protocol_features,
505 506
                                          VHOST_USER_PROTOCOL_F_REPLY_ACK) &&
                                          !do_postcopy;
507

508 509 510 511 512 513 514
    if (do_postcopy) {
        /* Postcopy has enough differences that it's best done in it's own
         * version
         */
        return vhost_user_set_mem_table_postcopy(dev, mem);
    }

515
    VhostUserMsg msg = {
516 517
        .hdr.request = VHOST_USER_SET_MEM_TABLE,
        .hdr.flags = VHOST_USER_VERSION,
518 519 520
    };

    if (reply_supported) {
521
        msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK;
522 523 524 525 526 527 528 529 530 531 532 533
    }

    for (i = 0; i < dev->mem->nregions; ++i) {
        struct vhost_memory_region *reg = dev->mem->regions + i;
        ram_addr_t offset;
        MemoryRegion *mr;

        assert((uintptr_t)reg->userspace_addr == reg->userspace_addr);
        mr = memory_region_from_host((void *)(uintptr_t)reg->userspace_addr,
                                     &offset);
        fd = memory_region_get_fd(mr);
        if (fd > 0) {
534 535 536 537
            if (fd_num == VHOST_MEMORY_MAX_NREGIONS) {
                error_report("Failed preparing vhost-user memory table msg");
                return -1;
            }
538 539
            msg.payload.memory.regions[fd_num].userspace_addr =
                reg->userspace_addr;
540
            msg.payload.memory.regions[fd_num].memory_size  = reg->memory_size;
541 542
            msg.payload.memory.regions[fd_num].guest_phys_addr =
                reg->guest_phys_addr;
543 544 545 546 547 548 549 550 551 552 553 554 555
            msg.payload.memory.regions[fd_num].mmap_offset = offset;
            fds[fd_num++] = fd;
        }
    }

    msg.payload.memory.nregions = fd_num;

    if (!fd_num) {
        error_report("Failed initializing vhost-user memory map, "
                     "consider using -object memory-backend-file share=on");
        return -1;
    }

556 557 558
    msg.hdr.size = sizeof(msg.payload.memory.nregions);
    msg.hdr.size += sizeof(msg.payload.memory.padding);
    msg.hdr.size += fd_num * sizeof(VhostUserMemoryRegion);
559 560 561 562 563 564

    if (vhost_user_write(dev, &msg, fds, fd_num) < 0) {
        return -1;
    }

    if (reply_supported) {
565
        return process_message_reply(dev, &msg);
566 567 568 569 570
    }

    return 0;
}

571 572 573 574
static int vhost_user_set_vring_addr(struct vhost_dev *dev,
                                     struct vhost_vring_addr *addr)
{
    VhostUserMsg msg = {
575 576
        .hdr.request = VHOST_USER_SET_VRING_ADDR,
        .hdr.flags = VHOST_USER_VERSION,
577
        .payload.addr = *addr,
578
        .hdr.size = sizeof(msg.payload.addr),
579
    };
580

581 582 583
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }
584

585 586
    return 0;
}
587

588 589 590
static int vhost_user_set_vring_endian(struct vhost_dev *dev,
                                       struct vhost_vring_state *ring)
{
591 592 593
    bool cross_endian = virtio_has_feature(dev->protocol_features,
                                           VHOST_USER_PROTOCOL_F_CROSS_ENDIAN);
    VhostUserMsg msg = {
594 595
        .hdr.request = VHOST_USER_SET_VRING_ENDIAN,
        .hdr.flags = VHOST_USER_VERSION,
596
        .payload.state = *ring,
597
        .hdr.size = sizeof(msg.payload.state),
598 599 600 601 602 603 604 605 606 607 608 609
    };

    if (!cross_endian) {
        error_report("vhost-user trying to send unhandled ioctl");
        return -1;
    }

    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }

    return 0;
610
}
611

612 613 614 615 616
static int vhost_set_vring(struct vhost_dev *dev,
                           unsigned long int request,
                           struct vhost_vring_state *ring)
{
    VhostUserMsg msg = {
617 618
        .hdr.request = request,
        .hdr.flags = VHOST_USER_VERSION,
619
        .payload.state = *ring,
620
        .hdr.size = sizeof(msg.payload.state),
621 622
    };

623 624 625
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }
626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643

    return 0;
}

static int vhost_user_set_vring_num(struct vhost_dev *dev,
                                    struct vhost_vring_state *ring)
{
    return vhost_set_vring(dev, VHOST_USER_SET_VRING_NUM, ring);
}

static int vhost_user_set_vring_base(struct vhost_dev *dev,
                                     struct vhost_vring_state *ring)
{
    return vhost_set_vring(dev, VHOST_USER_SET_VRING_BASE, ring);
}

static int vhost_user_set_vring_enable(struct vhost_dev *dev, int enable)
{
644
    int i;
645

646
    if (!virtio_has_feature(dev->features, VHOST_USER_F_PROTOCOL_FEATURES)) {
647 648 649
        return -1;
    }

650 651 652 653 654 655 656 657
    for (i = 0; i < dev->nvqs; ++i) {
        struct vhost_vring_state state = {
            .index = dev->vq_index + i,
            .num   = enable,
        };

        vhost_set_vring(dev, VHOST_USER_SET_VRING_ENABLE, &state);
    }
658

659 660
    return 0;
}
661 662 663 664 665

static int vhost_user_get_vring_base(struct vhost_dev *dev,
                                     struct vhost_vring_state *ring)
{
    VhostUserMsg msg = {
666 667
        .hdr.request = VHOST_USER_GET_VRING_BASE,
        .hdr.flags = VHOST_USER_VERSION,
668
        .payload.state = *ring,
669
        .hdr.size = sizeof(msg.payload.state),
670 671
    };

672 673 674
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }
675 676

    if (vhost_user_read(dev, &msg) < 0) {
677
        return -1;
678 679
    }

680
    if (msg.hdr.request != VHOST_USER_GET_VRING_BASE) {
681
        error_report("Received unexpected msg type. Expected %d received %d",
682
                     VHOST_USER_GET_VRING_BASE, msg.hdr.request);
683 684
        return -1;
    }
685

686
    if (msg.hdr.size != sizeof(msg.payload.state)) {
687 688
        error_report("Received bad msg size.");
        return -1;
689 690
    }

691
    *ring = msg.payload.state;
692

693 694 695
    return 0;
}

696 697 698
static int vhost_set_vring_file(struct vhost_dev *dev,
                                VhostUserRequest request,
                                struct vhost_vring_file *file)
699
{
700 701
    int fds[VHOST_MEMORY_MAX_NREGIONS];
    size_t fd_num = 0;
702
    VhostUserMsg msg = {
703 704
        .hdr.request = request,
        .hdr.flags = VHOST_USER_VERSION,
705
        .payload.u64 = file->index & VHOST_USER_VRING_IDX_MASK,
706
        .hdr.size = sizeof(msg.payload.u64),
707 708
    };

709 710 711
    if (ioeventfd_enabled() && file->fd > 0) {
        fds[fd_num++] = file->fd;
    } else {
712
        msg.payload.u64 |= VHOST_USER_VRING_NOFD_MASK;
713 714
    }

715 716 717
    if (vhost_user_write(dev, &msg, fds, fd_num) < 0) {
        return -1;
    }
718

719 720
    return 0;
}
721

722 723 724 725 726 727 728 729 730 731 732 733 734 735 736
static int vhost_user_set_vring_kick(struct vhost_dev *dev,
                                     struct vhost_vring_file *file)
{
    return vhost_set_vring_file(dev, VHOST_USER_SET_VRING_KICK, file);
}

static int vhost_user_set_vring_call(struct vhost_dev *dev,
                                     struct vhost_vring_file *file)
{
    return vhost_set_vring_file(dev, VHOST_USER_SET_VRING_CALL, file);
}

static int vhost_user_set_u64(struct vhost_dev *dev, int request, uint64_t u64)
{
    VhostUserMsg msg = {
737 738
        .hdr.request = request,
        .hdr.flags = VHOST_USER_VERSION,
739
        .payload.u64 = u64,
740
        .hdr.size = sizeof(msg.payload.u64),
741 742
    };

743 744 745
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }
746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764

    return 0;
}

static int vhost_user_set_features(struct vhost_dev *dev,
                                   uint64_t features)
{
    return vhost_user_set_u64(dev, VHOST_USER_SET_FEATURES, features);
}

static int vhost_user_set_protocol_features(struct vhost_dev *dev,
                                            uint64_t features)
{
    return vhost_user_set_u64(dev, VHOST_USER_SET_PROTOCOL_FEATURES, features);
}

static int vhost_user_get_u64(struct vhost_dev *dev, int request, uint64_t *u64)
{
    VhostUserMsg msg = {
765 766
        .hdr.request = request,
        .hdr.flags = VHOST_USER_VERSION,
767 768 769 770
    };

    if (vhost_user_one_time_request(request) && dev->vq_index != 0) {
        return 0;
771
    }
772

773 774 775
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }
776 777

    if (vhost_user_read(dev, &msg) < 0) {
778
        return -1;
779 780
    }

781
    if (msg.hdr.request != request) {
782
        error_report("Received unexpected msg type. Expected %d received %d",
783
                     request, msg.hdr.request);
784 785 786
        return -1;
    }

787
    if (msg.hdr.size != sizeof(msg.payload.u64)) {
788 789 790 791
        error_report("Received bad msg size.");
        return -1;
    }

792
    *u64 = msg.payload.u64;
793 794 795 796 797 798 799 800 801 802 803 804

    return 0;
}

static int vhost_user_get_features(struct vhost_dev *dev, uint64_t *features)
{
    return vhost_user_get_u64(dev, VHOST_USER_GET_FEATURES, features);
}

static int vhost_user_set_owner(struct vhost_dev *dev)
{
    VhostUserMsg msg = {
805 806
        .hdr.request = VHOST_USER_SET_OWNER,
        .hdr.flags = VHOST_USER_VERSION,
807 808
    };

809 810 811
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }
812 813 814 815 816 817 818

    return 0;
}

static int vhost_user_reset_device(struct vhost_dev *dev)
{
    VhostUserMsg msg = {
819 820
        .hdr.request = VHOST_USER_RESET_OWNER,
        .hdr.flags = VHOST_USER_VERSION,
821 822
    };

823 824 825
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }
826

827 828 829
    return 0;
}

830 831 832 833 834 835 836 837 838 839 840 841 842 843 844
static int vhost_user_slave_handle_config_change(struct vhost_dev *dev)
{
    int ret = -1;

    if (!dev->config_ops) {
        return -1;
    }

    if (dev->config_ops->vhost_dev_config_notifier) {
        ret = dev->config_ops->vhost_dev_config_notifier(dev);
    }

    return ret;
}

845 846 847 848
static void slave_read(void *opaque)
{
    struct vhost_dev *dev = opaque;
    struct vhost_user *u = dev->opaque;
849 850
    VhostUserHeader hdr = { 0, };
    VhostUserPayload payload = { 0, };
851 852 853
    int size, ret = 0;

    /* Read header */
854
    size = read(u->slave_fd, &hdr, VHOST_USER_HDR_SIZE);
855 856 857 858 859
    if (size != VHOST_USER_HDR_SIZE) {
        error_report("Failed to read from slave.");
        goto err;
    }

860
    if (hdr.size > VHOST_USER_PAYLOAD_SIZE) {
861
        error_report("Failed to read msg header."
862
                " Size %d exceeds the maximum %zu.", hdr.size,
863 864 865 866 867
                VHOST_USER_PAYLOAD_SIZE);
        goto err;
    }

    /* Read payload */
868 869
    size = read(u->slave_fd, &payload, hdr.size);
    if (size != hdr.size) {
870 871 872 873
        error_report("Failed to read payload from slave.");
        goto err;
    }

874
    switch (hdr.request) {
875
    case VHOST_USER_SLAVE_IOTLB_MSG:
876
        ret = vhost_backend_handle_iotlb_msg(dev, &payload.iotlb);
877
        break;
878 879 880
    case VHOST_USER_SLAVE_CONFIG_CHANGE_MSG :
        ret = vhost_user_slave_handle_config_change(dev);
        break;
881 882 883 884 885 886 887 888 889
    default:
        error_report("Received unexpected msg type.");
        ret = -EINVAL;
    }

    /*
     * REPLY_ACK feature handling. Other reply types has to be managed
     * directly in their request handlers.
     */
890 891
    if (hdr.flags & VHOST_USER_NEED_REPLY_MASK) {
        struct iovec iovec[2];
892 893


894 895 896 897 898 899 900 901 902 903 904 905 906
        hdr.flags &= ~VHOST_USER_NEED_REPLY_MASK;
        hdr.flags |= VHOST_USER_REPLY_MASK;

        payload.u64 = !!ret;
        hdr.size = sizeof(payload.u64);

        iovec[0].iov_base = &hdr;
        iovec[0].iov_len = VHOST_USER_HDR_SIZE;
        iovec[1].iov_base = &payload;
        iovec[1].iov_len = hdr.size;

        size = writev(u->slave_fd, iovec, ARRAY_SIZE(iovec));
        if (size != VHOST_USER_HDR_SIZE + hdr.size) {
907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923
            error_report("Failed to send msg reply to slave.");
            goto err;
        }
    }

    return;

err:
    qemu_set_fd_handler(u->slave_fd, NULL, NULL, NULL);
    close(u->slave_fd);
    u->slave_fd = -1;
    return;
}

static int vhost_setup_slave_channel(struct vhost_dev *dev)
{
    VhostUserMsg msg = {
924 925
        .hdr.request = VHOST_USER_SET_SLAVE_REQ_FD,
        .hdr.flags = VHOST_USER_VERSION,
926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945
    };
    struct vhost_user *u = dev->opaque;
    int sv[2], ret = 0;
    bool reply_supported = virtio_has_feature(dev->protocol_features,
                                              VHOST_USER_PROTOCOL_F_REPLY_ACK);

    if (!virtio_has_feature(dev->protocol_features,
                            VHOST_USER_PROTOCOL_F_SLAVE_REQ)) {
        return 0;
    }

    if (socketpair(PF_UNIX, SOCK_STREAM, 0, sv) == -1) {
        error_report("socketpair() failed");
        return -1;
    }

    u->slave_fd = sv[0];
    qemu_set_fd_handler(u->slave_fd, slave_read, NULL, dev);

    if (reply_supported) {
946
        msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK;
947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968
    }

    ret = vhost_user_write(dev, &msg, &sv[1], 1);
    if (ret) {
        goto out;
    }

    if (reply_supported) {
        ret = process_message_reply(dev, &msg);
    }

out:
    close(sv[1]);
    if (ret) {
        qemu_set_fd_handler(u->slave_fd, NULL, NULL, NULL);
        close(u->slave_fd);
        u->slave_fd = -1;
    }

    return ret;
}

969 970 971 972 973 974 975 976 977 978 979
/*
 * Called back from the postcopy fault thread when a fault is received on our
 * ufd.
 * TODO: This is Linux specific
 */
static int vhost_user_postcopy_fault_handler(struct PostCopyFD *pcfd,
                                             void *ufd)
{
    return 0;
}

980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018
/*
 * Called at the start of an inbound postcopy on reception of the
 * 'advise' command.
 */
static int vhost_user_postcopy_advise(struct vhost_dev *dev, Error **errp)
{
    struct vhost_user *u = dev->opaque;
    CharBackend *chr = u->chr;
    int ufd;
    VhostUserMsg msg = {
        .hdr.request = VHOST_USER_POSTCOPY_ADVISE,
        .hdr.flags = VHOST_USER_VERSION,
    };

    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        error_setg(errp, "Failed to send postcopy_advise to vhost");
        return -1;
    }

    if (vhost_user_read(dev, &msg) < 0) {
        error_setg(errp, "Failed to get postcopy_advise reply from vhost");
        return -1;
    }

    if (msg.hdr.request != VHOST_USER_POSTCOPY_ADVISE) {
        error_setg(errp, "Unexpected msg type. Expected %d received %d",
                     VHOST_USER_POSTCOPY_ADVISE, msg.hdr.request);
        return -1;
    }

    if (msg.hdr.size) {
        error_setg(errp, "Received bad msg size.");
        return -1;
    }
    ufd = qemu_chr_fe_get_msgfd(chr);
    if (ufd < 0) {
        error_setg(errp, "%s: Failed to get ufd", __func__);
        return -1;
    }
1019
    fcntl(ufd, F_SETFL, O_NONBLOCK);
1020

1021 1022 1023 1024 1025 1026
    /* register ufd with userfault thread */
    u->postcopy_fd.fd = ufd;
    u->postcopy_fd.data = dev;
    u->postcopy_fd.handler = vhost_user_postcopy_fault_handler;
    u->postcopy_fd.idstr = "vhost-user"; /* Need to find unique name */
    postcopy_register_shared_ufd(&u->postcopy_fd);
1027 1028 1029
    return 0;
}

1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056
/*
 * Called at the switch to postcopy on reception of the 'listen' command.
 */
static int vhost_user_postcopy_listen(struct vhost_dev *dev, Error **errp)
{
    struct vhost_user *u = dev->opaque;
    int ret;
    VhostUserMsg msg = {
        .hdr.request = VHOST_USER_POSTCOPY_LISTEN,
        .hdr.flags = VHOST_USER_VERSION | VHOST_USER_NEED_REPLY_MASK,
    };
    u->postcopy_listen = true;
    trace_vhost_user_postcopy_listen();
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        error_setg(errp, "Failed to send postcopy_listen to vhost");
        return -1;
    }

    ret = process_message_reply(dev, &msg);
    if (ret) {
        error_setg(errp, "Failed to receive reply to postcopy_listen");
        return ret;
    }

    return 0;
}

1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
                                        void *opaque)
{
    struct PostcopyNotifyData *pnd = opaque;
    struct vhost_user *u = container_of(notifier, struct vhost_user,
                                         postcopy_notifier);
    struct vhost_dev *dev = u->dev;

    switch (pnd->reason) {
    case POSTCOPY_NOTIFY_PROBE:
        if (!virtio_has_feature(dev->protocol_features,
                                VHOST_USER_PROTOCOL_F_PAGEFAULT)) {
            /* TODO: Get the device name into this error somehow */
            error_setg(pnd->errp,
                       "vhost-user backend not capable of postcopy");
            return -ENOENT;
        }
        break;

1076 1077 1078
    case POSTCOPY_NOTIFY_INBOUND_ADVISE:
        return vhost_user_postcopy_advise(dev, pnd->errp);

1079 1080 1081
    case POSTCOPY_NOTIFY_INBOUND_LISTEN:
        return vhost_user_postcopy_listen(dev, pnd->errp);

1082 1083 1084 1085 1086 1087 1088 1089
    default:
        /* We ignore notifications we don't know */
        break;
    }

    return 0;
}

1090 1091
static int vhost_user_init(struct vhost_dev *dev, void *opaque)
{
1092
    uint64_t features, protocol_features;
1093
    struct vhost_user *u;
1094 1095
    int err;

1096 1097
    assert(dev->vhost_ops->backend_type == VHOST_BACKEND_TYPE_USER);

1098 1099
    u = g_new0(struct vhost_user, 1);
    u->chr = opaque;
1100
    u->slave_fd = -1;
1101
    u->dev = dev;
1102
    dev->opaque = u;
1103

1104
    err = vhost_user_get_features(dev, &features);
1105 1106 1107 1108 1109 1110 1111
    if (err < 0) {
        return err;
    }

    if (virtio_has_feature(features, VHOST_USER_F_PROTOCOL_FEATURES)) {
        dev->backend_features |= 1ULL << VHOST_USER_F_PROTOCOL_FEATURES;

1112
        err = vhost_user_get_u64(dev, VHOST_USER_GET_PROTOCOL_FEATURES,
1113
                                 &protocol_features);
1114 1115 1116 1117
        if (err < 0) {
            return err;
        }

1118 1119
        dev->protocol_features =
            protocol_features & VHOST_USER_PROTOCOL_FEATURE_MASK;
1120
        err = vhost_user_set_protocol_features(dev, dev->protocol_features);
1121 1122 1123
        if (err < 0) {
            return err;
        }
1124 1125 1126

        /* query the max queues we support if backend supports Multiple Queue */
        if (dev->protocol_features & (1ULL << VHOST_USER_PROTOCOL_F_MQ)) {
1127 1128
            err = vhost_user_get_u64(dev, VHOST_USER_GET_QUEUE_NUM,
                                     &dev->max_queues);
1129 1130 1131 1132
            if (err < 0) {
                return err;
            }
        }
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142

        if (virtio_has_feature(features, VIRTIO_F_IOMMU_PLATFORM) &&
                !(virtio_has_feature(dev->protocol_features,
                    VHOST_USER_PROTOCOL_F_SLAVE_REQ) &&
                 virtio_has_feature(dev->protocol_features,
                    VHOST_USER_PROTOCOL_F_REPLY_ACK))) {
            error_report("IOMMU support requires reply-ack and "
                         "slave-req protocol features.");
            return -1;
        }
1143 1144
    }

1145 1146 1147 1148 1149 1150 1151 1152
    if (dev->migration_blocker == NULL &&
        !virtio_has_feature(dev->protocol_features,
                            VHOST_USER_PROTOCOL_F_LOG_SHMFD)) {
        error_setg(&dev->migration_blocker,
                   "Migration disabled: vhost-user backend lacks "
                   "VHOST_USER_PROTOCOL_F_LOG_SHMFD feature.");
    }

1153 1154 1155 1156 1157
    err = vhost_setup_slave_channel(dev);
    if (err < 0) {
        return err;
    }

1158 1159 1160
    u->postcopy_notifier.notify = vhost_user_postcopy_notifier;
    postcopy_add_notifier(&u->postcopy_notifier);

1161 1162 1163 1164 1165
    return 0;
}

static int vhost_user_cleanup(struct vhost_dev *dev)
{
1166 1167
    struct vhost_user *u;

1168 1169
    assert(dev->vhost_ops->backend_type == VHOST_BACKEND_TYPE_USER);

1170
    u = dev->opaque;
1171 1172 1173 1174
    if (u->postcopy_notifier.notify) {
        postcopy_remove_notifier(&u->postcopy_notifier);
        u->postcopy_notifier.notify = NULL;
    }
1175
    if (u->slave_fd >= 0) {
1176
        qemu_set_fd_handler(u->slave_fd, NULL, NULL, NULL);
1177 1178 1179
        close(u->slave_fd);
        u->slave_fd = -1;
    }
1180 1181 1182 1183 1184
    g_free(u->region_rb);
    u->region_rb = NULL;
    g_free(u->region_rb_offset);
    u->region_rb_offset = NULL;
    u->region_rb_len = 0;
1185
    g_free(u);
1186 1187 1188 1189 1190
    dev->opaque = 0;

    return 0;
}

1191 1192 1193 1194 1195 1196 1197
static int vhost_user_get_vq_index(struct vhost_dev *dev, int idx)
{
    assert(idx >= dev->vq_index && idx < dev->vq_index + dev->nvqs);

    return idx;
}

1198 1199 1200 1201 1202
static int vhost_user_memslots_limit(struct vhost_dev *dev)
{
    return VHOST_MEMORY_MAX_NREGIONS;
}

1203 1204 1205 1206 1207 1208 1209 1210
static bool vhost_user_requires_shm_log(struct vhost_dev *dev)
{
    assert(dev->vhost_ops->backend_type == VHOST_BACKEND_TYPE_USER);

    return virtio_has_feature(dev->protocol_features,
                              VHOST_USER_PROTOCOL_F_LOG_SHMFD);
}

1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224
static int vhost_user_migration_done(struct vhost_dev *dev, char* mac_addr)
{
    VhostUserMsg msg = { 0 };

    assert(dev->vhost_ops->backend_type == VHOST_BACKEND_TYPE_USER);

    /* If guest supports GUEST_ANNOUNCE do nothing */
    if (virtio_has_feature(dev->acked_features, VIRTIO_NET_F_GUEST_ANNOUNCE)) {
        return 0;
    }

    /* if backend supports VHOST_USER_PROTOCOL_F_RARP ask it to send the RARP */
    if (virtio_has_feature(dev->protocol_features,
                           VHOST_USER_PROTOCOL_F_RARP)) {
1225 1226
        msg.hdr.request = VHOST_USER_SEND_RARP;
        msg.hdr.flags = VHOST_USER_VERSION;
1227
        memcpy((char *)&msg.payload.u64, mac_addr, 6);
1228
        msg.hdr.size = sizeof(msg.payload.u64);
1229

1230
        return vhost_user_write(dev, &msg, NULL, 0);
1231 1232 1233 1234
    }
    return -1;
}

1235 1236 1237 1238
static bool vhost_user_can_merge(struct vhost_dev *dev,
                                 uint64_t start1, uint64_t size1,
                                 uint64_t start2, uint64_t size2)
{
1239
    ram_addr_t offset;
1240 1241 1242
    int mfd, rfd;
    MemoryRegion *mr;

1243
    mr = memory_region_from_host((void *)(uintptr_t)start1, &offset);
1244
    mfd = memory_region_get_fd(mr);
1245

1246
    mr = memory_region_from_host((void *)(uintptr_t)start2, &offset);
1247
    rfd = memory_region_get_fd(mr);
1248 1249 1250 1251

    return mfd == rfd;
}

1252 1253 1254 1255 1256 1257 1258 1259 1260 1261
static int vhost_user_net_set_mtu(struct vhost_dev *dev, uint16_t mtu)
{
    VhostUserMsg msg;
    bool reply_supported = virtio_has_feature(dev->protocol_features,
                                              VHOST_USER_PROTOCOL_F_REPLY_ACK);

    if (!(dev->protocol_features & (1ULL << VHOST_USER_PROTOCOL_F_NET_MTU))) {
        return 0;
    }

1262
    msg.hdr.request = VHOST_USER_NET_SET_MTU;
1263
    msg.payload.u64 = mtu;
1264 1265
    msg.hdr.size = sizeof(msg.payload.u64);
    msg.hdr.flags = VHOST_USER_VERSION;
1266
    if (reply_supported) {
1267
        msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK;
1268 1269 1270 1271 1272 1273 1274 1275
    }

    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }

    /* If reply_ack supported, slave has to ack specified MTU is valid */
    if (reply_supported) {
1276
        return process_message_reply(dev, &msg);
1277 1278 1279 1280 1281
    }

    return 0;
}

1282 1283 1284 1285
static int vhost_user_send_device_iotlb_msg(struct vhost_dev *dev,
                                            struct vhost_iotlb_msg *imsg)
{
    VhostUserMsg msg = {
1286 1287 1288
        .hdr.request = VHOST_USER_IOTLB_MSG,
        .hdr.size = sizeof(msg.payload.iotlb),
        .hdr.flags = VHOST_USER_VERSION | VHOST_USER_NEED_REPLY_MASK,
1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304
        .payload.iotlb = *imsg,
    };

    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -EFAULT;
    }

    return process_message_reply(dev, &msg);
}


static void vhost_user_set_iotlb_callback(struct vhost_dev *dev, int enabled)
{
    /* No-op as the receive channel is not dedicated to IOTLB messages. */
}

1305 1306 1307 1308
static int vhost_user_get_config(struct vhost_dev *dev, uint8_t *config,
                                 uint32_t config_len)
{
    VhostUserMsg msg = {
1309 1310 1311
        .hdr.request = VHOST_USER_GET_CONFIG,
        .hdr.flags = VHOST_USER_VERSION,
        .hdr.size = VHOST_USER_CONFIG_HDR_SIZE + config_len,
1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327
    };

    if (config_len > VHOST_USER_MAX_CONFIG_SIZE) {
        return -1;
    }

    msg.payload.config.offset = 0;
    msg.payload.config.size = config_len;
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }

    if (vhost_user_read(dev, &msg) < 0) {
        return -1;
    }

1328
    if (msg.hdr.request != VHOST_USER_GET_CONFIG) {
1329
        error_report("Received unexpected msg type. Expected %d received %d",
1330
                     VHOST_USER_GET_CONFIG, msg.hdr.request);
1331 1332 1333
        return -1;
    }

1334
    if (msg.hdr.size != VHOST_USER_CONFIG_HDR_SIZE + config_len) {
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
        error_report("Received bad msg size.");
        return -1;
    }

    memcpy(config, msg.payload.config.region, config_len);

    return 0;
}

static int vhost_user_set_config(struct vhost_dev *dev, const uint8_t *data,
                                 uint32_t offset, uint32_t size, uint32_t flags)
{
    uint8_t *p;
    bool reply_supported = virtio_has_feature(dev->protocol_features,
                                              VHOST_USER_PROTOCOL_F_REPLY_ACK);

    VhostUserMsg msg = {
1352 1353 1354
        .hdr.request = VHOST_USER_SET_CONFIG,
        .hdr.flags = VHOST_USER_VERSION,
        .hdr.size = VHOST_USER_CONFIG_HDR_SIZE + size,
1355 1356 1357
    };

    if (reply_supported) {
1358
        msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK;
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381
    }

    if (size > VHOST_USER_MAX_CONFIG_SIZE) {
        return -1;
    }

    msg.payload.config.offset = offset,
    msg.payload.config.size = size,
    msg.payload.config.flags = flags,
    p = msg.payload.config.region;
    memcpy(p, data, size);

    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        return -1;
    }

    if (reply_supported) {
        return process_message_reply(dev, &msg);
    }

    return 0;
}

1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467
static int vhost_user_crypto_create_session(struct vhost_dev *dev,
                                            void *session_info,
                                            uint64_t *session_id)
{
    bool crypto_session = virtio_has_feature(dev->protocol_features,
                                       VHOST_USER_PROTOCOL_F_CRYPTO_SESSION);
    CryptoDevBackendSymSessionInfo *sess_info = session_info;
    VhostUserMsg msg = {
        .hdr.request = VHOST_USER_CREATE_CRYPTO_SESSION,
        .hdr.flags = VHOST_USER_VERSION,
        .hdr.size = sizeof(msg.payload.session),
    };

    assert(dev->vhost_ops->backend_type == VHOST_BACKEND_TYPE_USER);

    if (!crypto_session) {
        error_report("vhost-user trying to send unhandled ioctl");
        return -1;
    }

    memcpy(&msg.payload.session.session_setup_data, sess_info,
              sizeof(CryptoDevBackendSymSessionInfo));
    if (sess_info->key_len) {
        memcpy(&msg.payload.session.key, sess_info->cipher_key,
               sess_info->key_len);
    }
    if (sess_info->auth_key_len > 0) {
        memcpy(&msg.payload.session.auth_key, sess_info->auth_key,
               sess_info->auth_key_len);
    }
    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        error_report("vhost_user_write() return -1, create session failed");
        return -1;
    }

    if (vhost_user_read(dev, &msg) < 0) {
        error_report("vhost_user_read() return -1, create session failed");
        return -1;
    }

    if (msg.hdr.request != VHOST_USER_CREATE_CRYPTO_SESSION) {
        error_report("Received unexpected msg type. Expected %d received %d",
                     VHOST_USER_CREATE_CRYPTO_SESSION, msg.hdr.request);
        return -1;
    }

    if (msg.hdr.size != sizeof(msg.payload.session)) {
        error_report("Received bad msg size.");
        return -1;
    }

    if (msg.payload.session.session_id < 0) {
        error_report("Bad session id: %" PRId64 "",
                              msg.payload.session.session_id);
        return -1;
    }
    *session_id = msg.payload.session.session_id;

    return 0;
}

static int
vhost_user_crypto_close_session(struct vhost_dev *dev, uint64_t session_id)
{
    bool crypto_session = virtio_has_feature(dev->protocol_features,
                                       VHOST_USER_PROTOCOL_F_CRYPTO_SESSION);
    VhostUserMsg msg = {
        .hdr.request = VHOST_USER_CLOSE_CRYPTO_SESSION,
        .hdr.flags = VHOST_USER_VERSION,
        .hdr.size = sizeof(msg.payload.u64),
    };
    msg.payload.u64 = session_id;

    if (!crypto_session) {
        error_report("vhost-user trying to send unhandled ioctl");
        return -1;
    }

    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
        error_report("vhost_user_write() return -1, close session failed");
        return -1;
    }

    return 0;
}

1468 1469 1470
const VhostOps user_ops = {
        .backend_type = VHOST_BACKEND_TYPE_USER,
        .vhost_backend_init = vhost_user_init,
1471
        .vhost_backend_cleanup = vhost_user_cleanup,
1472
        .vhost_backend_memslots_limit = vhost_user_memslots_limit,
1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487
        .vhost_set_log_base = vhost_user_set_log_base,
        .vhost_set_mem_table = vhost_user_set_mem_table,
        .vhost_set_vring_addr = vhost_user_set_vring_addr,
        .vhost_set_vring_endian = vhost_user_set_vring_endian,
        .vhost_set_vring_num = vhost_user_set_vring_num,
        .vhost_set_vring_base = vhost_user_set_vring_base,
        .vhost_get_vring_base = vhost_user_get_vring_base,
        .vhost_set_vring_kick = vhost_user_set_vring_kick,
        .vhost_set_vring_call = vhost_user_set_vring_call,
        .vhost_set_features = vhost_user_set_features,
        .vhost_get_features = vhost_user_get_features,
        .vhost_set_owner = vhost_user_set_owner,
        .vhost_reset_device = vhost_user_reset_device,
        .vhost_get_vq_index = vhost_user_get_vq_index,
        .vhost_set_vring_enable = vhost_user_set_vring_enable,
1488
        .vhost_requires_shm_log = vhost_user_requires_shm_log,
1489
        .vhost_migration_done = vhost_user_migration_done,
1490
        .vhost_backend_can_merge = vhost_user_can_merge,
1491
        .vhost_net_set_mtu = vhost_user_net_set_mtu,
1492 1493
        .vhost_set_iotlb_callback = vhost_user_set_iotlb_callback,
        .vhost_send_device_iotlb_msg = vhost_user_send_device_iotlb_msg,
1494 1495
        .vhost_get_config = vhost_user_get_config,
        .vhost_set_config = vhost_user_set_config,
1496 1497
        .vhost_crypto_create_session = vhost_user_crypto_create_session,
        .vhost_crypto_close_session = vhost_user_crypto_close_session,
1498
};