qemu/migration/rdma.c
<<
>>
Prefs
   1/*
   2 * RDMA protocol and interfaces
   3 *
   4 * Copyright IBM, Corp. 2010-2013
   5 *
   6 * Authors:
   7 *  Michael R. Hines <mrhines@us.ibm.com>
   8 *  Jiuxing Liu <jl@us.ibm.com>
   9 *
  10 * This work is licensed under the terms of the GNU GPL, version 2 or
  11 * later.  See the COPYING file in the top-level directory.
  12 *
  13 */
  14#include "qemu/osdep.h"
  15#include "qapi/error.h"
  16#include "qemu-common.h"
  17#include "qemu/cutils.h"
  18#include "migration/migration.h"
  19#include "migration/qemu-file.h"
  20#include "exec/cpu-common.h"
  21#include "qemu/error-report.h"
  22#include "qemu/main-loop.h"
  23#include "qemu/sockets.h"
  24#include "qemu/bitmap.h"
  25#include "qemu/coroutine.h"
  26#include <sys/socket.h>
  27#include <netdb.h>
  28#include <arpa/inet.h>
  29#include <rdma/rdma_cma.h>
  30#include "trace.h"
  31
  32/*
  33 * Print and error on both the Monitor and the Log file.
  34 */
  35#define ERROR(errp, fmt, ...) \
  36    do { \
  37        fprintf(stderr, "RDMA ERROR: " fmt "\n", ## __VA_ARGS__); \
  38        if (errp && (*(errp) == NULL)) { \
  39            error_setg(errp, "RDMA ERROR: " fmt, ## __VA_ARGS__); \
  40        } \
  41    } while (0)
  42
  43#define RDMA_RESOLVE_TIMEOUT_MS 10000
  44
  45/* Do not merge data if larger than this. */
  46#define RDMA_MERGE_MAX (2 * 1024 * 1024)
  47#define RDMA_SIGNALED_SEND_MAX (RDMA_MERGE_MAX / 4096)
  48
  49#define RDMA_REG_CHUNK_SHIFT 20 /* 1 MB */
  50
  51/*
  52 * This is only for non-live state being migrated.
  53 * Instead of RDMA_WRITE messages, we use RDMA_SEND
  54 * messages for that state, which requires a different
  55 * delivery design than main memory.
  56 */
  57#define RDMA_SEND_INCREMENT 32768
  58
  59/*
  60 * Maximum size infiniband SEND message
  61 */
  62#define RDMA_CONTROL_MAX_BUFFER (512 * 1024)
  63#define RDMA_CONTROL_MAX_COMMANDS_PER_MESSAGE 4096
  64
  65#define RDMA_CONTROL_VERSION_CURRENT 1
  66/*
  67 * Capabilities for negotiation.
  68 */
  69#define RDMA_CAPABILITY_PIN_ALL 0x01
  70
  71/*
  72 * Add the other flags above to this list of known capabilities
  73 * as they are introduced.
  74 */
  75static uint32_t known_capabilities = RDMA_CAPABILITY_PIN_ALL;
  76
  77#define CHECK_ERROR_STATE() \
  78    do { \
  79        if (rdma->error_state) { \
  80            if (!rdma->error_reported) { \
  81                error_report("RDMA is in an error state waiting migration" \
  82                                " to abort!"); \
  83                rdma->error_reported = 1; \
  84            } \
  85            return rdma->error_state; \
  86        } \
  87    } while (0);
  88
  89/*
  90 * A work request ID is 64-bits and we split up these bits
  91 * into 3 parts:
  92 *
  93 * bits 0-15 : type of control message, 2^16
  94 * bits 16-29: ram block index, 2^14
  95 * bits 30-63: ram block chunk number, 2^34
  96 *
  97 * The last two bit ranges are only used for RDMA writes,
  98 * in order to track their completion and potentially
  99 * also track unregistration status of the message.
 100 */
 101#define RDMA_WRID_TYPE_SHIFT  0UL
 102#define RDMA_WRID_BLOCK_SHIFT 16UL
 103#define RDMA_WRID_CHUNK_SHIFT 30UL
 104
 105#define RDMA_WRID_TYPE_MASK \
 106    ((1UL << RDMA_WRID_BLOCK_SHIFT) - 1UL)
 107
 108#define RDMA_WRID_BLOCK_MASK \
 109    (~RDMA_WRID_TYPE_MASK & ((1UL << RDMA_WRID_CHUNK_SHIFT) - 1UL))
 110
 111#define RDMA_WRID_CHUNK_MASK (~RDMA_WRID_BLOCK_MASK & ~RDMA_WRID_TYPE_MASK)
 112
 113/*
 114 * RDMA migration protocol:
 115 * 1. RDMA Writes (data messages, i.e. RAM)
 116 * 2. IB Send/Recv (control channel messages)
 117 */
 118enum {
 119    RDMA_WRID_NONE = 0,
 120    RDMA_WRID_RDMA_WRITE = 1,
 121    RDMA_WRID_SEND_CONTROL = 2000,
 122    RDMA_WRID_RECV_CONTROL = 4000,
 123};
 124
 125static const char *wrid_desc[] = {
 126    [RDMA_WRID_NONE] = "NONE",
 127    [RDMA_WRID_RDMA_WRITE] = "WRITE RDMA",
 128    [RDMA_WRID_SEND_CONTROL] = "CONTROL SEND",
 129    [RDMA_WRID_RECV_CONTROL] = "CONTROL RECV",
 130};
 131
 132/*
 133 * Work request IDs for IB SEND messages only (not RDMA writes).
 134 * This is used by the migration protocol to transmit
 135 * control messages (such as device state and registration commands)
 136 *
 137 * We could use more WRs, but we have enough for now.
 138 */
 139enum {
 140    RDMA_WRID_READY = 0,
 141    RDMA_WRID_DATA,
 142    RDMA_WRID_CONTROL,
 143    RDMA_WRID_MAX,
 144};
 145
 146/*
 147 * SEND/RECV IB Control Messages.
 148 */
 149enum {
 150    RDMA_CONTROL_NONE = 0,
 151    RDMA_CONTROL_ERROR,
 152    RDMA_CONTROL_READY,               /* ready to receive */
 153    RDMA_CONTROL_QEMU_FILE,           /* QEMUFile-transmitted bytes */
 154    RDMA_CONTROL_RAM_BLOCKS_REQUEST,  /* RAMBlock synchronization */
 155    RDMA_CONTROL_RAM_BLOCKS_RESULT,   /* RAMBlock synchronization */
 156    RDMA_CONTROL_COMPRESS,            /* page contains repeat values */
 157    RDMA_CONTROL_REGISTER_REQUEST,    /* dynamic page registration */
 158    RDMA_CONTROL_REGISTER_RESULT,     /* key to use after registration */
 159    RDMA_CONTROL_REGISTER_FINISHED,   /* current iteration finished */
 160    RDMA_CONTROL_UNREGISTER_REQUEST,  /* dynamic UN-registration */
 161    RDMA_CONTROL_UNREGISTER_FINISHED, /* unpinning finished */
 162};
 163
 164static const char *control_desc[] = {
 165    [RDMA_CONTROL_NONE] = "NONE",
 166    [RDMA_CONTROL_ERROR] = "ERROR",
 167    [RDMA_CONTROL_READY] = "READY",
 168    [RDMA_CONTROL_QEMU_FILE] = "QEMU FILE",
 169    [RDMA_CONTROL_RAM_BLOCKS_REQUEST] = "RAM BLOCKS REQUEST",
 170    [RDMA_CONTROL_RAM_BLOCKS_RESULT] = "RAM BLOCKS RESULT",
 171    [RDMA_CONTROL_COMPRESS] = "COMPRESS",
 172    [RDMA_CONTROL_REGISTER_REQUEST] = "REGISTER REQUEST",
 173    [RDMA_CONTROL_REGISTER_RESULT] = "REGISTER RESULT",
 174    [RDMA_CONTROL_REGISTER_FINISHED] = "REGISTER FINISHED",
 175    [RDMA_CONTROL_UNREGISTER_REQUEST] = "UNREGISTER REQUEST",
 176    [RDMA_CONTROL_UNREGISTER_FINISHED] = "UNREGISTER FINISHED",
 177};
 178
 179/*
 180 * Memory and MR structures used to represent an IB Send/Recv work request.
 181 * This is *not* used for RDMA writes, only IB Send/Recv.
 182 */
 183typedef struct {
 184    uint8_t  control[RDMA_CONTROL_MAX_BUFFER]; /* actual buffer to register */
 185    struct   ibv_mr *control_mr;               /* registration metadata */
 186    size_t   control_len;                      /* length of the message */
 187    uint8_t *control_curr;                     /* start of unconsumed bytes */
 188} RDMAWorkRequestData;
 189
 190/*
 191 * Negotiate RDMA capabilities during connection-setup time.
 192 */
 193typedef struct {
 194    uint32_t version;
 195    uint32_t flags;
 196} RDMACapabilities;
 197
 198static void caps_to_network(RDMACapabilities *cap)
 199{
 200    cap->version = htonl(cap->version);
 201    cap->flags = htonl(cap->flags);
 202}
 203
 204static void network_to_caps(RDMACapabilities *cap)
 205{
 206    cap->version = ntohl(cap->version);
 207    cap->flags = ntohl(cap->flags);
 208}
 209
 210/*
 211 * Representation of a RAMBlock from an RDMA perspective.
 212 * This is not transmitted, only local.
 213 * This and subsequent structures cannot be linked lists
 214 * because we're using a single IB message to transmit
 215 * the information. It's small anyway, so a list is overkill.
 216 */
 217typedef struct RDMALocalBlock {
 218    char          *block_name;
 219    uint8_t       *local_host_addr; /* local virtual address */
 220    uint64_t       remote_host_addr; /* remote virtual address */
 221    uint64_t       offset;
 222    uint64_t       length;
 223    struct         ibv_mr **pmr;    /* MRs for chunk-level registration */
 224    struct         ibv_mr *mr;      /* MR for non-chunk-level registration */
 225    uint32_t      *remote_keys;     /* rkeys for chunk-level registration */
 226    uint32_t       remote_rkey;     /* rkeys for non-chunk-level registration */
 227    int            index;           /* which block are we */
 228    unsigned int   src_index;       /* (Only used on dest) */
 229    bool           is_ram_block;
 230    int            nb_chunks;
 231    unsigned long *transit_bitmap;
 232    unsigned long *unregister_bitmap;
 233} RDMALocalBlock;
 234
 235/*
 236 * Also represents a RAMblock, but only on the dest.
 237 * This gets transmitted by the dest during connection-time
 238 * to the source VM and then is used to populate the
 239 * corresponding RDMALocalBlock with
 240 * the information needed to perform the actual RDMA.
 241 */
 242typedef struct QEMU_PACKED RDMADestBlock {
 243    uint64_t remote_host_addr;
 244    uint64_t offset;
 245    uint64_t length;
 246    uint32_t remote_rkey;
 247    uint32_t padding;
 248} RDMADestBlock;
 249
 250static uint64_t htonll(uint64_t v)
 251{
 252    union { uint32_t lv[2]; uint64_t llv; } u;
 253    u.lv[0] = htonl(v >> 32);
 254    u.lv[1] = htonl(v & 0xFFFFFFFFULL);
 255    return u.llv;
 256}
 257
 258static uint64_t ntohll(uint64_t v) {
 259    union { uint32_t lv[2]; uint64_t llv; } u;
 260    u.llv = v;
 261    return ((uint64_t)ntohl(u.lv[0]) << 32) | (uint64_t) ntohl(u.lv[1]);
 262}
 263
 264static void dest_block_to_network(RDMADestBlock *db)
 265{
 266    db->remote_host_addr = htonll(db->remote_host_addr);
 267    db->offset = htonll(db->offset);
 268    db->length = htonll(db->length);
 269    db->remote_rkey = htonl(db->remote_rkey);
 270}
 271
 272static void network_to_dest_block(RDMADestBlock *db)
 273{
 274    db->remote_host_addr = ntohll(db->remote_host_addr);
 275    db->offset = ntohll(db->offset);
 276    db->length = ntohll(db->length);
 277    db->remote_rkey = ntohl(db->remote_rkey);
 278}
 279
 280/*
 281 * Virtual address of the above structures used for transmitting
 282 * the RAMBlock descriptions at connection-time.
 283 * This structure is *not* transmitted.
 284 */
 285typedef struct RDMALocalBlocks {
 286    int nb_blocks;
 287    bool     init;             /* main memory init complete */
 288    RDMALocalBlock *block;
 289} RDMALocalBlocks;
 290
 291/*
 292 * Main data structure for RDMA state.
 293 * While there is only one copy of this structure being allocated right now,
 294 * this is the place where one would start if you wanted to consider
 295 * having more than one RDMA connection open at the same time.
 296 */
 297typedef struct RDMAContext {
 298    char *host;
 299    int port;
 300
 301    RDMAWorkRequestData wr_data[RDMA_WRID_MAX];
 302
 303    /*
 304     * This is used by *_exchange_send() to figure out whether or not
 305     * the initial "READY" message has already been received or not.
 306     * This is because other functions may potentially poll() and detect
 307     * the READY message before send() does, in which case we need to
 308     * know if it completed.
 309     */
 310    int control_ready_expected;
 311
 312    /* number of outstanding writes */
 313    int nb_sent;
 314
 315    /* store info about current buffer so that we can
 316       merge it with future sends */
 317    uint64_t current_addr;
 318    uint64_t current_length;
 319    /* index of ram block the current buffer belongs to */
 320    int current_index;
 321    /* index of the chunk in the current ram block */
 322    int current_chunk;
 323
 324    bool pin_all;
 325
 326    /*
 327     * infiniband-specific variables for opening the device
 328     * and maintaining connection state and so forth.
 329     *
 330     * cm_id also has ibv_context, rdma_event_channel, and ibv_qp in
 331     * cm_id->verbs, cm_id->channel, and cm_id->qp.
 332     */
 333    struct rdma_cm_id *cm_id;               /* connection manager ID */
 334    struct rdma_cm_id *listen_id;
 335    bool connected;
 336
 337    struct ibv_context          *verbs;
 338    struct rdma_event_channel   *channel;
 339    struct ibv_qp *qp;                      /* queue pair */
 340    struct ibv_comp_channel *comp_channel;  /* completion channel */
 341    struct ibv_pd *pd;                      /* protection domain */
 342    struct ibv_cq *cq;                      /* completion queue */
 343
 344    /*
 345     * If a previous write failed (perhaps because of a failed
 346     * memory registration, then do not attempt any future work
 347     * and remember the error state.
 348     */
 349    int error_state;
 350    int error_reported;
 351
 352    /*
 353     * Description of ram blocks used throughout the code.
 354     */
 355    RDMALocalBlocks local_ram_blocks;
 356    RDMADestBlock  *dest_blocks;
 357
 358    /* Index of the next RAMBlock received during block registration */
 359    unsigned int    next_src_index;
 360
 361    /*
 362     * Migration on *destination* started.
 363     * Then use coroutine yield function.
 364     * Source runs in a thread, so we don't care.
 365     */
 366    int migration_started_on_destination;
 367
 368    int total_registrations;
 369    int total_writes;
 370
 371    int unregister_current, unregister_next;
 372    uint64_t unregistrations[RDMA_SIGNALED_SEND_MAX];
 373
 374    GHashTable *blockmap;
 375} RDMAContext;
 376
 377/*
 378 * Interface to the rest of the migration call stack.
 379 */
 380typedef struct QEMUFileRDMA {
 381    RDMAContext *rdma;
 382    size_t len;
 383    void *file;
 384} QEMUFileRDMA;
 385
 386/*
 387 * Main structure for IB Send/Recv control messages.
 388 * This gets prepended at the beginning of every Send/Recv.
 389 */
 390typedef struct QEMU_PACKED {
 391    uint32_t len;     /* Total length of data portion */
 392    uint32_t type;    /* which control command to perform */
 393    uint32_t repeat;  /* number of commands in data portion of same type */
 394    uint32_t padding;
 395} RDMAControlHeader;
 396
 397static void control_to_network(RDMAControlHeader *control)
 398{
 399    control->type = htonl(control->type);
 400    control->len = htonl(control->len);
 401    control->repeat = htonl(control->repeat);
 402}
 403
 404static void network_to_control(RDMAControlHeader *control)
 405{
 406    control->type = ntohl(control->type);
 407    control->len = ntohl(control->len);
 408    control->repeat = ntohl(control->repeat);
 409}
 410
 411/*
 412 * Register a single Chunk.
 413 * Information sent by the source VM to inform the dest
 414 * to register an single chunk of memory before we can perform
 415 * the actual RDMA operation.
 416 */
 417typedef struct QEMU_PACKED {
 418    union QEMU_PACKED {
 419        uint64_t current_addr;  /* offset into the ram_addr_t space */
 420        uint64_t chunk;         /* chunk to lookup if unregistering */
 421    } key;
 422    uint32_t current_index; /* which ramblock the chunk belongs to */
 423    uint32_t padding;
 424    uint64_t chunks;            /* how many sequential chunks to register */
 425} RDMARegister;
 426
 427static void register_to_network(RDMAContext *rdma, RDMARegister *reg)
 428{
 429    RDMALocalBlock *local_block;
 430    local_block  = &rdma->local_ram_blocks.block[reg->current_index];
 431
 432    if (local_block->is_ram_block) {
 433        /*
 434         * current_addr as passed in is an address in the local ram_addr_t
 435         * space, we need to translate this for the destination
 436         */
 437        reg->key.current_addr -= local_block->offset;
 438        reg->key.current_addr += rdma->dest_blocks[reg->current_index].offset;
 439    }
 440    reg->key.current_addr = htonll(reg->key.current_addr);
 441    reg->current_index = htonl(reg->current_index);
 442    reg->chunks = htonll(reg->chunks);
 443}
 444
 445static void network_to_register(RDMARegister *reg)
 446{
 447    reg->key.current_addr = ntohll(reg->key.current_addr);
 448    reg->current_index = ntohl(reg->current_index);
 449    reg->chunks = ntohll(reg->chunks);
 450}
 451
 452typedef struct QEMU_PACKED {
 453    uint32_t value;     /* if zero, we will madvise() */
 454    uint32_t block_idx; /* which ram block index */
 455    uint64_t offset;    /* Address in remote ram_addr_t space */
 456    uint64_t length;    /* length of the chunk */
 457} RDMACompress;
 458
 459static void compress_to_network(RDMAContext *rdma, RDMACompress *comp)
 460{
 461    comp->value = htonl(comp->value);
 462    /*
 463     * comp->offset as passed in is an address in the local ram_addr_t
 464     * space, we need to translate this for the destination
 465     */
 466    comp->offset -= rdma->local_ram_blocks.block[comp->block_idx].offset;
 467    comp->offset += rdma->dest_blocks[comp->block_idx].offset;
 468    comp->block_idx = htonl(comp->block_idx);
 469    comp->offset = htonll(comp->offset);
 470    comp->length = htonll(comp->length);
 471}
 472
 473static void network_to_compress(RDMACompress *comp)
 474{
 475    comp->value = ntohl(comp->value);
 476    comp->block_idx = ntohl(comp->block_idx);
 477    comp->offset = ntohll(comp->offset);
 478    comp->length = ntohll(comp->length);
 479}
 480
 481/*
 482 * The result of the dest's memory registration produces an "rkey"
 483 * which the source VM must reference in order to perform
 484 * the RDMA operation.
 485 */
 486typedef struct QEMU_PACKED {
 487    uint32_t rkey;
 488    uint32_t padding;
 489    uint64_t host_addr;
 490} RDMARegisterResult;
 491
 492static void result_to_network(RDMARegisterResult *result)
 493{
 494    result->rkey = htonl(result->rkey);
 495    result->host_addr = htonll(result->host_addr);
 496};
 497
 498static void network_to_result(RDMARegisterResult *result)
 499{
 500    result->rkey = ntohl(result->rkey);
 501    result->host_addr = ntohll(result->host_addr);
 502};
 503
 504const char *print_wrid(int wrid);
 505static int qemu_rdma_exchange_send(RDMAContext *rdma, RDMAControlHeader *head,
 506                                   uint8_t *data, RDMAControlHeader *resp,
 507                                   int *resp_idx,
 508                                   int (*callback)(RDMAContext *rdma));
 509
 510static inline uint64_t ram_chunk_index(const uint8_t *start,
 511                                       const uint8_t *host)
 512{
 513    return ((uintptr_t) host - (uintptr_t) start) >> RDMA_REG_CHUNK_SHIFT;
 514}
 515
 516static inline uint8_t *ram_chunk_start(const RDMALocalBlock *rdma_ram_block,
 517                                       uint64_t i)
 518{
 519    return (uint8_t *)(uintptr_t)(rdma_ram_block->local_host_addr +
 520                                  (i << RDMA_REG_CHUNK_SHIFT));
 521}
 522
 523static inline uint8_t *ram_chunk_end(const RDMALocalBlock *rdma_ram_block,
 524                                     uint64_t i)
 525{
 526    uint8_t *result = ram_chunk_start(rdma_ram_block, i) +
 527                                         (1UL << RDMA_REG_CHUNK_SHIFT);
 528
 529    if (result > (rdma_ram_block->local_host_addr + rdma_ram_block->length)) {
 530        result = rdma_ram_block->local_host_addr + rdma_ram_block->length;
 531    }
 532
 533    return result;
 534}
 535
 536static int rdma_add_block(RDMAContext *rdma, const char *block_name,
 537                         void *host_addr,
 538                         ram_addr_t block_offset, uint64_t length)
 539{
 540    RDMALocalBlocks *local = &rdma->local_ram_blocks;
 541    RDMALocalBlock *block;
 542    RDMALocalBlock *old = local->block;
 543
 544    local->block = g_new0(RDMALocalBlock, local->nb_blocks + 1);
 545
 546    if (local->nb_blocks) {
 547        int x;
 548
 549        if (rdma->blockmap) {
 550            for (x = 0; x < local->nb_blocks; x++) {
 551                g_hash_table_remove(rdma->blockmap,
 552                                    (void *)(uintptr_t)old[x].offset);
 553                g_hash_table_insert(rdma->blockmap,
 554                                    (void *)(uintptr_t)old[x].offset,
 555                                    &local->block[x]);
 556            }
 557        }
 558        memcpy(local->block, old, sizeof(RDMALocalBlock) * local->nb_blocks);
 559        g_free(old);
 560    }
 561
 562    block = &local->block[local->nb_blocks];
 563
 564    block->block_name = g_strdup(block_name);
 565    block->local_host_addr = host_addr;
 566    block->offset = block_offset;
 567    block->length = length;
 568    block->index = local->nb_blocks;
 569    block->src_index = ~0U; /* Filled in by the receipt of the block list */
 570    block->nb_chunks = ram_chunk_index(host_addr, host_addr + length) + 1UL;
 571    block->transit_bitmap = bitmap_new(block->nb_chunks);
 572    bitmap_clear(block->transit_bitmap, 0, block->nb_chunks);
 573    block->unregister_bitmap = bitmap_new(block->nb_chunks);
 574    bitmap_clear(block->unregister_bitmap, 0, block->nb_chunks);
 575    block->remote_keys = g_new0(uint32_t, block->nb_chunks);
 576
 577    block->is_ram_block = local->init ? false : true;
 578
 579    if (rdma->blockmap) {
 580        g_hash_table_insert(rdma->blockmap, (void *)(uintptr_t)block_offset, block);
 581    }
 582
 583    trace_rdma_add_block(block_name, local->nb_blocks,
 584                         (uintptr_t) block->local_host_addr,
 585                         block->offset, block->length,
 586                         (uintptr_t) (block->local_host_addr + block->length),
 587                         BITS_TO_LONGS(block->nb_chunks) *
 588                             sizeof(unsigned long) * 8,
 589                         block->nb_chunks);
 590
 591    local->nb_blocks++;
 592
 593    return 0;
 594}
 595
 596/*
 597 * Memory regions need to be registered with the device and queue pairs setup
 598 * in advanced before the migration starts. This tells us where the RAM blocks
 599 * are so that we can register them individually.
 600 */
 601static int qemu_rdma_init_one_block(const char *block_name, void *host_addr,
 602    ram_addr_t block_offset, ram_addr_t length, void *opaque)
 603{
 604    return rdma_add_block(opaque, block_name, host_addr, block_offset, length);
 605}
 606
 607/*
 608 * Identify the RAMBlocks and their quantity. They will be references to
 609 * identify chunk boundaries inside each RAMBlock and also be referenced
 610 * during dynamic page registration.
 611 */
 612static int qemu_rdma_init_ram_blocks(RDMAContext *rdma)
 613{
 614    RDMALocalBlocks *local = &rdma->local_ram_blocks;
 615
 616    assert(rdma->blockmap == NULL);
 617    memset(local, 0, sizeof *local);
 618    qemu_ram_foreach_block(qemu_rdma_init_one_block, rdma);
 619    trace_qemu_rdma_init_ram_blocks(local->nb_blocks);
 620    rdma->dest_blocks = g_new0(RDMADestBlock,
 621                               rdma->local_ram_blocks.nb_blocks);
 622    local->init = true;
 623    return 0;
 624}
 625
 626/*
 627 * Note: If used outside of cleanup, the caller must ensure that the destination
 628 * block structures are also updated
 629 */
 630static int rdma_delete_block(RDMAContext *rdma, RDMALocalBlock *block)
 631{
 632    RDMALocalBlocks *local = &rdma->local_ram_blocks;
 633    RDMALocalBlock *old = local->block;
 634    int x;
 635
 636    if (rdma->blockmap) {
 637        g_hash_table_remove(rdma->blockmap, (void *)(uintptr_t)block->offset);
 638    }
 639    if (block->pmr) {
 640        int j;
 641
 642        for (j = 0; j < block->nb_chunks; j++) {
 643            if (!block->pmr[j]) {
 644                continue;
 645            }
 646            ibv_dereg_mr(block->pmr[j]);
 647            rdma->total_registrations--;
 648        }
 649        g_free(block->pmr);
 650        block->pmr = NULL;
 651    }
 652
 653    if (block->mr) {
 654        ibv_dereg_mr(block->mr);
 655        rdma->total_registrations--;
 656        block->mr = NULL;
 657    }
 658
 659    g_free(block->transit_bitmap);
 660    block->transit_bitmap = NULL;
 661
 662    g_free(block->unregister_bitmap);
 663    block->unregister_bitmap = NULL;
 664
 665    g_free(block->remote_keys);
 666    block->remote_keys = NULL;
 667
 668    g_free(block->block_name);
 669    block->block_name = NULL;
 670
 671    if (rdma->blockmap) {
 672        for (x = 0; x < local->nb_blocks; x++) {
 673            g_hash_table_remove(rdma->blockmap,
 674                                (void *)(uintptr_t)old[x].offset);
 675        }
 676    }
 677
 678    if (local->nb_blocks > 1) {
 679
 680        local->block = g_new0(RDMALocalBlock, local->nb_blocks - 1);
 681
 682        if (block->index) {
 683            memcpy(local->block, old, sizeof(RDMALocalBlock) * block->index);
 684        }
 685
 686        if (block->index < (local->nb_blocks - 1)) {
 687            memcpy(local->block + block->index, old + (block->index + 1),
 688                sizeof(RDMALocalBlock) *
 689                    (local->nb_blocks - (block->index + 1)));
 690        }
 691    } else {
 692        assert(block == local->block);
 693        local->block = NULL;
 694    }
 695
 696    trace_rdma_delete_block(block, (uintptr_t)block->local_host_addr,
 697                           block->offset, block->length,
 698                            (uintptr_t)(block->local_host_addr + block->length),
 699                           BITS_TO_LONGS(block->nb_chunks) *
 700                               sizeof(unsigned long) * 8, block->nb_chunks);
 701
 702    g_free(old);
 703
 704    local->nb_blocks--;
 705
 706    if (local->nb_blocks && rdma->blockmap) {
 707        for (x = 0; x < local->nb_blocks; x++) {
 708            g_hash_table_insert(rdma->blockmap,
 709                                (void *)(uintptr_t)local->block[x].offset,
 710                                &local->block[x]);
 711        }
 712    }
 713
 714    return 0;
 715}
 716
 717/*
 718 * Put in the log file which RDMA device was opened and the details
 719 * associated with that device.
 720 */
 721static void qemu_rdma_dump_id(const char *who, struct ibv_context *verbs)
 722{
 723    struct ibv_port_attr port;
 724
 725    if (ibv_query_port(verbs, 1, &port)) {
 726        error_report("Failed to query port information");
 727        return;
 728    }
 729
 730    printf("%s RDMA Device opened: kernel name %s "
 731           "uverbs device name %s, "
 732           "infiniband_verbs class device path %s, "
 733           "infiniband class device path %s, "
 734           "transport: (%d) %s\n",
 735                who,
 736                verbs->device->name,
 737                verbs->device->dev_name,
 738                verbs->device->dev_path,
 739                verbs->device->ibdev_path,
 740                port.link_layer,
 741                (port.link_layer == IBV_LINK_LAYER_INFINIBAND) ? "Infiniband" :
 742                 ((port.link_layer == IBV_LINK_LAYER_ETHERNET)
 743                    ? "Ethernet" : "Unknown"));
 744}
 745
 746/*
 747 * Put in the log file the RDMA gid addressing information,
 748 * useful for folks who have trouble understanding the
 749 * RDMA device hierarchy in the kernel.
 750 */
 751static void qemu_rdma_dump_gid(const char *who, struct rdma_cm_id *id)
 752{
 753    char sgid[33];
 754    char dgid[33];
 755    inet_ntop(AF_INET6, &id->route.addr.addr.ibaddr.sgid, sgid, sizeof sgid);
 756    inet_ntop(AF_INET6, &id->route.addr.addr.ibaddr.dgid, dgid, sizeof dgid);
 757    trace_qemu_rdma_dump_gid(who, sgid, dgid);
 758}
 759
 760/*
 761 * As of now, IPv6 over RoCE / iWARP is not supported by linux.
 762 * We will try the next addrinfo struct, and fail if there are
 763 * no other valid addresses to bind against.
 764 *
 765 * If user is listening on '[::]', then we will not have a opened a device
 766 * yet and have no way of verifying if the device is RoCE or not.
 767 *
 768 * In this case, the source VM will throw an error for ALL types of
 769 * connections (both IPv4 and IPv6) if the destination machine does not have
 770 * a regular infiniband network available for use.
 771 *
 772 * The only way to guarantee that an error is thrown for broken kernels is
 773 * for the management software to choose a *specific* interface at bind time
 774 * and validate what time of hardware it is.
 775 *
 776 * Unfortunately, this puts the user in a fix:
 777 *
 778 *  If the source VM connects with an IPv4 address without knowing that the
 779 *  destination has bound to '[::]' the migration will unconditionally fail
 780 *  unless the management software is explicitly listening on the IPv4
 781 *  address while using a RoCE-based device.
 782 *
 783 *  If the source VM connects with an IPv6 address, then we're OK because we can
 784 *  throw an error on the source (and similarly on the destination).
 785 *
 786 *  But in mixed environments, this will be broken for a while until it is fixed
 787 *  inside linux.
 788 *
 789 * We do provide a *tiny* bit of help in this function: We can list all of the
 790 * devices in the system and check to see if all the devices are RoCE or
 791 * Infiniband.
 792 *
 793 * If we detect that we have a *pure* RoCE environment, then we can safely
 794 * thrown an error even if the management software has specified '[::]' as the
 795 * bind address.
 796 *
 797 * However, if there is are multiple hetergeneous devices, then we cannot make
 798 * this assumption and the user just has to be sure they know what they are
 799 * doing.
 800 *
 801 * Patches are being reviewed on linux-rdma.
 802 */
 803static int qemu_rdma_broken_ipv6_kernel(Error **errp, struct ibv_context *verbs)
 804{
 805    struct ibv_port_attr port_attr;
 806
 807    /* This bug only exists in linux, to our knowledge. */
 808#ifdef CONFIG_LINUX
 809
 810    /*
 811     * Verbs are only NULL if management has bound to '[::]'.
 812     *
 813     * Let's iterate through all the devices and see if there any pure IB
 814     * devices (non-ethernet).
 815     *
 816     * If not, then we can safely proceed with the migration.
 817     * Otherwise, there are no guarantees until the bug is fixed in linux.
 818     */
 819    if (!verbs) {
 820        int num_devices, x;
 821        struct ibv_device ** dev_list = ibv_get_device_list(&num_devices);
 822        bool roce_found = false;
 823        bool ib_found = false;
 824
 825        for (x = 0; x < num_devices; x++) {
 826            verbs = ibv_open_device(dev_list[x]);
 827            if (!verbs) {
 828                if (errno == EPERM) {
 829                    continue;
 830                } else {
 831                    return -EINVAL;
 832                }
 833            }
 834
 835            if (ibv_query_port(verbs, 1, &port_attr)) {
 836                ibv_close_device(verbs);
 837                ERROR(errp, "Could not query initial IB port");
 838                return -EINVAL;
 839            }
 840
 841            if (port_attr.link_layer == IBV_LINK_LAYER_INFINIBAND) {
 842                ib_found = true;
 843            } else if (port_attr.link_layer == IBV_LINK_LAYER_ETHERNET) {
 844                roce_found = true;
 845            }
 846
 847            ibv_close_device(verbs);
 848
 849        }
 850
 851        if (roce_found) {
 852            if (ib_found) {
 853                fprintf(stderr, "WARN: migrations may fail:"
 854                                " IPv6 over RoCE / iWARP in linux"
 855                                " is broken. But since you appear to have a"
 856                                " mixed RoCE / IB environment, be sure to only"
 857                                " migrate over the IB fabric until the kernel "
 858                                " fixes the bug.\n");
 859            } else {
 860                ERROR(errp, "You only have RoCE / iWARP devices in your systems"
 861                            " and your management software has specified '[::]'"
 862                            ", but IPv6 over RoCE / iWARP is not supported in Linux.");
 863                return -ENONET;
 864            }
 865        }
 866
 867        return 0;
 868    }
 869
 870    /*
 871     * If we have a verbs context, that means that some other than '[::]' was
 872     * used by the management software for binding. In which case we can
 873     * actually warn the user about a potentially broken kernel.
 874     */
 875
 876    /* IB ports start with 1, not 0 */
 877    if (ibv_query_port(verbs, 1, &port_attr)) {
 878        ERROR(errp, "Could not query initial IB port");
 879        return -EINVAL;
 880    }
 881
 882    if (port_attr.link_layer == IBV_LINK_LAYER_ETHERNET) {
 883        ERROR(errp, "Linux kernel's RoCE / iWARP does not support IPv6 "
 884                    "(but patches on linux-rdma in progress)");
 885        return -ENONET;
 886    }
 887
 888#endif
 889
 890    return 0;
 891}
 892
 893/*
 894 * Figure out which RDMA device corresponds to the requested IP hostname
 895 * Also create the initial connection manager identifiers for opening
 896 * the connection.
 897 */
 898static int qemu_rdma_resolve_host(RDMAContext *rdma, Error **errp)
 899{
 900    int ret;
 901    struct rdma_addrinfo *res;
 902    char port_str[16];
 903    struct rdma_cm_event *cm_event;
 904    char ip[40] = "unknown";
 905    struct rdma_addrinfo *e;
 906
 907    if (rdma->host == NULL || !strcmp(rdma->host, "")) {
 908        ERROR(errp, "RDMA hostname has not been set");
 909        return -EINVAL;
 910    }
 911
 912    /* create CM channel */
 913    rdma->channel = rdma_create_event_channel();
 914    if (!rdma->channel) {
 915        ERROR(errp, "could not create CM channel");
 916        return -EINVAL;
 917    }
 918
 919    /* create CM id */
 920    ret = rdma_create_id(rdma->channel, &rdma->cm_id, NULL, RDMA_PS_TCP);
 921    if (ret) {
 922        ERROR(errp, "could not create channel id");
 923        goto err_resolve_create_id;
 924    }
 925
 926    snprintf(port_str, 16, "%d", rdma->port);
 927    port_str[15] = '\0';
 928
 929    ret = rdma_getaddrinfo(rdma->host, port_str, NULL, &res);
 930    if (ret < 0) {
 931        ERROR(errp, "could not rdma_getaddrinfo address %s", rdma->host);
 932        goto err_resolve_get_addr;
 933    }
 934
 935    for (e = res; e != NULL; e = e->ai_next) {
 936        inet_ntop(e->ai_family,
 937            &((struct sockaddr_in *) e->ai_dst_addr)->sin_addr, ip, sizeof ip);
 938        trace_qemu_rdma_resolve_host_trying(rdma->host, ip);
 939
 940        ret = rdma_resolve_addr(rdma->cm_id, NULL, e->ai_dst_addr,
 941                RDMA_RESOLVE_TIMEOUT_MS);
 942        if (!ret) {
 943            if (e->ai_family == AF_INET6) {
 944                ret = qemu_rdma_broken_ipv6_kernel(errp, rdma->cm_id->verbs);
 945                if (ret) {
 946                    continue;
 947                }
 948            }
 949            goto route;
 950        }
 951    }
 952
 953    ERROR(errp, "could not resolve address %s", rdma->host);
 954    goto err_resolve_get_addr;
 955
 956route:
 957    qemu_rdma_dump_gid("source_resolve_addr", rdma->cm_id);
 958
 959    ret = rdma_get_cm_event(rdma->channel, &cm_event);
 960    if (ret) {
 961        ERROR(errp, "could not perform event_addr_resolved");
 962        goto err_resolve_get_addr;
 963    }
 964
 965    if (cm_event->event != RDMA_CM_EVENT_ADDR_RESOLVED) {
 966        ERROR(errp, "result not equal to event_addr_resolved %s",
 967                rdma_event_str(cm_event->event));
 968        perror("rdma_resolve_addr");
 969        rdma_ack_cm_event(cm_event);
 970        ret = -EINVAL;
 971        goto err_resolve_get_addr;
 972    }
 973    rdma_ack_cm_event(cm_event);
 974
 975    /* resolve route */
 976    ret = rdma_resolve_route(rdma->cm_id, RDMA_RESOLVE_TIMEOUT_MS);
 977    if (ret) {
 978        ERROR(errp, "could not resolve rdma route");
 979        goto err_resolve_get_addr;
 980    }
 981
 982    ret = rdma_get_cm_event(rdma->channel, &cm_event);
 983    if (ret) {
 984        ERROR(errp, "could not perform event_route_resolved");
 985        goto err_resolve_get_addr;
 986    }
 987    if (cm_event->event != RDMA_CM_EVENT_ROUTE_RESOLVED) {
 988        ERROR(errp, "result not equal to event_route_resolved: %s",
 989                        rdma_event_str(cm_event->event));
 990        rdma_ack_cm_event(cm_event);
 991        ret = -EINVAL;
 992        goto err_resolve_get_addr;
 993    }
 994    rdma_ack_cm_event(cm_event);
 995    rdma->verbs = rdma->cm_id->verbs;
 996    qemu_rdma_dump_id("source_resolve_host", rdma->cm_id->verbs);
 997    qemu_rdma_dump_gid("source_resolve_host", rdma->cm_id);
 998    return 0;
 999
1000err_resolve_get_addr:
1001    rdma_destroy_id(rdma->cm_id);
1002    rdma->cm_id = NULL;
1003err_resolve_create_id:
1004    rdma_destroy_event_channel(rdma->channel);
1005    rdma->channel = NULL;
1006    return ret;
1007}
1008
1009/*
1010 * Create protection domain and completion queues
1011 */
1012static int qemu_rdma_alloc_pd_cq(RDMAContext *rdma)
1013{
1014    /* allocate pd */
1015    rdma->pd = ibv_alloc_pd(rdma->verbs);
1016    if (!rdma->pd) {
1017        error_report("failed to allocate protection domain");
1018        return -1;
1019    }
1020
1021    /* create completion channel */
1022    rdma->comp_channel = ibv_create_comp_channel(rdma->verbs);
1023    if (!rdma->comp_channel) {
1024        error_report("failed to allocate completion channel");
1025        goto err_alloc_pd_cq;
1026    }
1027
1028    /*
1029     * Completion queue can be filled by both read and write work requests,
1030     * so must reflect the sum of both possible queue sizes.
1031     */
1032    rdma->cq = ibv_create_cq(rdma->verbs, (RDMA_SIGNALED_SEND_MAX * 3),
1033            NULL, rdma->comp_channel, 0);
1034    if (!rdma->cq) {
1035        error_report("failed to allocate completion queue");
1036        goto err_alloc_pd_cq;
1037    }
1038
1039    return 0;
1040
1041err_alloc_pd_cq:
1042    if (rdma->pd) {
1043        ibv_dealloc_pd(rdma->pd);
1044    }
1045    if (rdma->comp_channel) {
1046        ibv_destroy_comp_channel(rdma->comp_channel);
1047    }
1048    rdma->pd = NULL;
1049    rdma->comp_channel = NULL;
1050    return -1;
1051
1052}
1053
1054/*
1055 * Create queue pairs.
1056 */
1057static int qemu_rdma_alloc_qp(RDMAContext *rdma)
1058{
1059    struct ibv_qp_init_attr attr = { 0 };
1060    int ret;
1061
1062    attr.cap.max_send_wr = RDMA_SIGNALED_SEND_MAX;
1063    attr.cap.max_recv_wr = 3;
1064    attr.cap.max_send_sge = 1;
1065    attr.cap.max_recv_sge = 1;
1066    attr.send_cq = rdma->cq;
1067    attr.recv_cq = rdma->cq;
1068    attr.qp_type = IBV_QPT_RC;
1069
1070    ret = rdma_create_qp(rdma->cm_id, rdma->pd, &attr);
1071    if (ret) {
1072        return -1;
1073    }
1074
1075    rdma->qp = rdma->cm_id->qp;
1076    return 0;
1077}
1078
1079static int qemu_rdma_reg_whole_ram_blocks(RDMAContext *rdma)
1080{
1081    int i;
1082    RDMALocalBlocks *local = &rdma->local_ram_blocks;
1083
1084    for (i = 0; i < local->nb_blocks; i++) {
1085        local->block[i].mr =
1086            ibv_reg_mr(rdma->pd,
1087                    local->block[i].local_host_addr,
1088                    local->block[i].length,
1089                    IBV_ACCESS_LOCAL_WRITE |
1090                    IBV_ACCESS_REMOTE_WRITE
1091                    );
1092        if (!local->block[i].mr) {
1093            perror("Failed to register local dest ram block!\n");
1094            break;
1095        }
1096        rdma->total_registrations++;
1097    }
1098
1099    if (i >= local->nb_blocks) {
1100        return 0;
1101    }
1102
1103    for (i--; i >= 0; i--) {
1104        ibv_dereg_mr(local->block[i].mr);
1105        rdma->total_registrations--;
1106    }
1107
1108    return -1;
1109
1110}
1111
1112/*
1113 * Find the ram block that corresponds to the page requested to be
1114 * transmitted by QEMU.
1115 *
1116 * Once the block is found, also identify which 'chunk' within that
1117 * block that the page belongs to.
1118 *
1119 * This search cannot fail or the migration will fail.
1120 */
1121static int qemu_rdma_search_ram_block(RDMAContext *rdma,
1122                                      uintptr_t block_offset,
1123                                      uint64_t offset,
1124                                      uint64_t length,
1125                                      uint64_t *block_index,
1126                                      uint64_t *chunk_index)
1127{
1128    uint64_t current_addr = block_offset + offset;
1129    RDMALocalBlock *block = g_hash_table_lookup(rdma->blockmap,
1130                                                (void *) block_offset);
1131    assert(block);
1132    assert(current_addr >= block->offset);
1133    assert((current_addr + length) <= (block->offset + block->length));
1134
1135    *block_index = block->index;
1136    *chunk_index = ram_chunk_index(block->local_host_addr,
1137                block->local_host_addr + (current_addr - block->offset));
1138
1139    return 0;
1140}
1141
1142/*
1143 * Register a chunk with IB. If the chunk was already registered
1144 * previously, then skip.
1145 *
1146 * Also return the keys associated with the registration needed
1147 * to perform the actual RDMA operation.
1148 */
1149static int qemu_rdma_register_and_get_keys(RDMAContext *rdma,
1150        RDMALocalBlock *block, uintptr_t host_addr,
1151        uint32_t *lkey, uint32_t *rkey, int chunk,
1152        uint8_t *chunk_start, uint8_t *chunk_end)
1153{
1154    if (block->mr) {
1155        if (lkey) {
1156            *lkey = block->mr->lkey;
1157        }
1158        if (rkey) {
1159            *rkey = block->mr->rkey;
1160        }
1161        return 0;
1162    }
1163
1164    /* allocate memory to store chunk MRs */
1165    if (!block->pmr) {
1166        block->pmr = g_new0(struct ibv_mr *, block->nb_chunks);
1167    }
1168
1169    /*
1170     * If 'rkey', then we're the destination, so grant access to the source.
1171     *
1172     * If 'lkey', then we're the source VM, so grant access only to ourselves.
1173     */
1174    if (!block->pmr[chunk]) {
1175        uint64_t len = chunk_end - chunk_start;
1176
1177        trace_qemu_rdma_register_and_get_keys(len, chunk_start);
1178
1179        block->pmr[chunk] = ibv_reg_mr(rdma->pd,
1180                chunk_start, len,
1181                (rkey ? (IBV_ACCESS_LOCAL_WRITE |
1182                        IBV_ACCESS_REMOTE_WRITE) : 0));
1183
1184        if (!block->pmr[chunk]) {
1185            perror("Failed to register chunk!");
1186            fprintf(stderr, "Chunk details: block: %d chunk index %d"
1187                            " start %" PRIuPTR " end %" PRIuPTR
1188                            " host %" PRIuPTR
1189                            " local %" PRIuPTR " registrations: %d\n",
1190                            block->index, chunk, (uintptr_t)chunk_start,
1191                            (uintptr_t)chunk_end, host_addr,
1192                            (uintptr_t)block->local_host_addr,
1193                            rdma->total_registrations);
1194            return -1;
1195        }
1196        rdma->total_registrations++;
1197    }
1198
1199    if (lkey) {
1200        *lkey = block->pmr[chunk]->lkey;
1201    }
1202    if (rkey) {
1203        *rkey = block->pmr[chunk]->rkey;
1204    }
1205    return 0;
1206}
1207
1208/*
1209 * Register (at connection time) the memory used for control
1210 * channel messages.
1211 */
1212static int qemu_rdma_reg_control(RDMAContext *rdma, int idx)
1213{
1214    rdma->wr_data[idx].control_mr = ibv_reg_mr(rdma->pd,
1215            rdma->wr_data[idx].control, RDMA_CONTROL_MAX_BUFFER,
1216            IBV_ACCESS_LOCAL_WRITE | IBV_ACCESS_REMOTE_WRITE);
1217    if (rdma->wr_data[idx].control_mr) {
1218        rdma->total_registrations++;
1219        return 0;
1220    }
1221    error_report("qemu_rdma_reg_control failed");
1222    return -1;
1223}
1224
1225const char *print_wrid(int wrid)
1226{
1227    if (wrid >= RDMA_WRID_RECV_CONTROL) {
1228        return wrid_desc[RDMA_WRID_RECV_CONTROL];
1229    }
1230    return wrid_desc[wrid];
1231}
1232
1233/*
1234 * RDMA requires memory registration (mlock/pinning), but this is not good for
1235 * overcommitment.
1236 *
1237 * In preparation for the future where LRU information or workload-specific
1238 * writable writable working set memory access behavior is available to QEMU
1239 * it would be nice to have in place the ability to UN-register/UN-pin
1240 * particular memory regions from the RDMA hardware when it is determine that
1241 * those regions of memory will likely not be accessed again in the near future.
1242 *
1243 * While we do not yet have such information right now, the following
1244 * compile-time option allows us to perform a non-optimized version of this
1245 * behavior.
1246 *
1247 * By uncommenting this option, you will cause *all* RDMA transfers to be
1248 * unregistered immediately after the transfer completes on both sides of the
1249 * connection. This has no effect in 'rdma-pin-all' mode, only regular mode.
1250 *
1251 * This will have a terrible impact on migration performance, so until future
1252 * workload information or LRU information is available, do not attempt to use
1253 * this feature except for basic testing.
1254 */
1255//#define RDMA_UNREGISTRATION_EXAMPLE
1256
1257/*
1258 * Perform a non-optimized memory unregistration after every transfer
1259 * for demonstration purposes, only if pin-all is not requested.
1260 *
1261 * Potential optimizations:
1262 * 1. Start a new thread to run this function continuously
1263        - for bit clearing
1264        - and for receipt of unregister messages
1265 * 2. Use an LRU.
1266 * 3. Use workload hints.
1267 */
1268static int qemu_rdma_unregister_waiting(RDMAContext *rdma)
1269{
1270    while (rdma->unregistrations[rdma->unregister_current]) {
1271        int ret;
1272        uint64_t wr_id = rdma->unregistrations[rdma->unregister_current];
1273        uint64_t chunk =
1274            (wr_id & RDMA_WRID_CHUNK_MASK) >> RDMA_WRID_CHUNK_SHIFT;
1275        uint64_t index =
1276            (wr_id & RDMA_WRID_BLOCK_MASK) >> RDMA_WRID_BLOCK_SHIFT;
1277        RDMALocalBlock *block =
1278            &(rdma->local_ram_blocks.block[index]);
1279        RDMARegister reg = { .current_index = index };
1280        RDMAControlHeader resp = { .type = RDMA_CONTROL_UNREGISTER_FINISHED,
1281                                 };
1282        RDMAControlHeader head = { .len = sizeof(RDMARegister),
1283                                   .type = RDMA_CONTROL_UNREGISTER_REQUEST,
1284                                   .repeat = 1,
1285                                 };
1286
1287        trace_qemu_rdma_unregister_waiting_proc(chunk,
1288                                                rdma->unregister_current);
1289
1290        rdma->unregistrations[rdma->unregister_current] = 0;
1291        rdma->unregister_current++;
1292
1293        if (rdma->unregister_current == RDMA_SIGNALED_SEND_MAX) {
1294            rdma->unregister_current = 0;
1295        }
1296
1297
1298        /*
1299         * Unregistration is speculative (because migration is single-threaded
1300         * and we cannot break the protocol's inifinband message ordering).
1301         * Thus, if the memory is currently being used for transmission,
1302         * then abort the attempt to unregister and try again
1303         * later the next time a completion is received for this memory.
1304         */
1305        clear_bit(chunk, block->unregister_bitmap);
1306
1307        if (test_bit(chunk, block->transit_bitmap)) {
1308            trace_qemu_rdma_unregister_waiting_inflight(chunk);
1309            continue;
1310        }
1311
1312        trace_qemu_rdma_unregister_waiting_send(chunk);
1313
1314        ret = ibv_dereg_mr(block->pmr[chunk]);
1315        block->pmr[chunk] = NULL;
1316        block->remote_keys[chunk] = 0;
1317
1318        if (ret != 0) {
1319            perror("unregistration chunk failed");
1320            return -ret;
1321        }
1322        rdma->total_registrations--;
1323
1324        reg.key.chunk = chunk;
1325        register_to_network(rdma, &reg);
1326        ret = qemu_rdma_exchange_send(rdma, &head, (uint8_t *) &reg,
1327                                &resp, NULL, NULL);
1328        if (ret < 0) {
1329            return ret;
1330        }
1331
1332        trace_qemu_rdma_unregister_waiting_complete(chunk);
1333    }
1334
1335    return 0;
1336}
1337
1338static uint64_t qemu_rdma_make_wrid(uint64_t wr_id, uint64_t index,
1339                                         uint64_t chunk)
1340{
1341    uint64_t result = wr_id & RDMA_WRID_TYPE_MASK;
1342
1343    result |= (index << RDMA_WRID_BLOCK_SHIFT);
1344    result |= (chunk << RDMA_WRID_CHUNK_SHIFT);
1345
1346    return result;
1347}
1348
1349/*
1350 * Set bit for unregistration in the next iteration.
1351 * We cannot transmit right here, but will unpin later.
1352 */
1353static void qemu_rdma_signal_unregister(RDMAContext *rdma, uint64_t index,
1354                                        uint64_t chunk, uint64_t wr_id)
1355{
1356    if (rdma->unregistrations[rdma->unregister_next] != 0) {
1357        error_report("rdma migration: queue is full");
1358    } else {
1359        RDMALocalBlock *block = &(rdma->local_ram_blocks.block[index]);
1360
1361        if (!test_and_set_bit(chunk, block->unregister_bitmap)) {
1362            trace_qemu_rdma_signal_unregister_append(chunk,
1363                                                     rdma->unregister_next);
1364
1365            rdma->unregistrations[rdma->unregister_next++] =
1366                    qemu_rdma_make_wrid(wr_id, index, chunk);
1367
1368            if (rdma->unregister_next == RDMA_SIGNALED_SEND_MAX) {
1369                rdma->unregister_next = 0;
1370            }
1371        } else {
1372            trace_qemu_rdma_signal_unregister_already(chunk);
1373        }
1374    }
1375}
1376
1377/*
1378 * Consult the connection manager to see a work request
1379 * (of any kind) has completed.
1380 * Return the work request ID that completed.
1381 */
1382static uint64_t qemu_rdma_poll(RDMAContext *rdma, uint64_t *wr_id_out,
1383                               uint32_t *byte_len)
1384{
1385    int ret;
1386    struct ibv_wc wc;
1387    uint64_t wr_id;
1388
1389    ret = ibv_poll_cq(rdma->cq, 1, &wc);
1390
1391    if (!ret) {
1392        *wr_id_out = RDMA_WRID_NONE;
1393        return 0;
1394    }
1395
1396    if (ret < 0) {
1397        error_report("ibv_poll_cq return %d", ret);
1398        return ret;
1399    }
1400
1401    wr_id = wc.wr_id & RDMA_WRID_TYPE_MASK;
1402
1403    if (wc.status != IBV_WC_SUCCESS) {
1404        fprintf(stderr, "ibv_poll_cq wc.status=%d %s!\n",
1405                        wc.status, ibv_wc_status_str(wc.status));
1406        fprintf(stderr, "ibv_poll_cq wrid=%s!\n", wrid_desc[wr_id]);
1407
1408        return -1;
1409    }
1410
1411    if (rdma->control_ready_expected &&
1412        (wr_id >= RDMA_WRID_RECV_CONTROL)) {
1413        trace_qemu_rdma_poll_recv(wrid_desc[RDMA_WRID_RECV_CONTROL],
1414                  wr_id - RDMA_WRID_RECV_CONTROL, wr_id, rdma->nb_sent);
1415        rdma->control_ready_expected = 0;
1416    }
1417
1418    if (wr_id == RDMA_WRID_RDMA_WRITE) {
1419        uint64_t chunk =
1420            (wc.wr_id & RDMA_WRID_CHUNK_MASK) >> RDMA_WRID_CHUNK_SHIFT;
1421        uint64_t index =
1422            (wc.wr_id & RDMA_WRID_BLOCK_MASK) >> RDMA_WRID_BLOCK_SHIFT;
1423        RDMALocalBlock *block = &(rdma->local_ram_blocks.block[index]);
1424
1425        trace_qemu_rdma_poll_write(print_wrid(wr_id), wr_id, rdma->nb_sent,
1426                                   index, chunk, block->local_host_addr,
1427                                   (void *)(uintptr_t)block->remote_host_addr);
1428
1429        clear_bit(chunk, block->transit_bitmap);
1430
1431        if (rdma->nb_sent > 0) {
1432            rdma->nb_sent--;
1433        }
1434
1435        if (!rdma->pin_all) {
1436            /*
1437             * FYI: If one wanted to signal a specific chunk to be unregistered
1438             * using LRU or workload-specific information, this is the function
1439             * you would call to do so. That chunk would then get asynchronously
1440             * unregistered later.
1441             */
1442#ifdef RDMA_UNREGISTRATION_EXAMPLE
1443            qemu_rdma_signal_unregister(rdma, index, chunk, wc.wr_id);
1444#endif
1445        }
1446    } else {
1447        trace_qemu_rdma_poll_other(print_wrid(wr_id), wr_id, rdma->nb_sent);
1448    }
1449
1450    *wr_id_out = wc.wr_id;
1451    if (byte_len) {
1452        *byte_len = wc.byte_len;
1453    }
1454
1455    return  0;
1456}
1457
1458/*
1459 * Block until the next work request has completed.
1460 *
1461 * First poll to see if a work request has already completed,
1462 * otherwise block.
1463 *
1464 * If we encounter completed work requests for IDs other than
1465 * the one we're interested in, then that's generally an error.
1466 *
1467 * The only exception is actual RDMA Write completions. These
1468 * completions only need to be recorded, but do not actually
1469 * need further processing.
1470 */
1471static int qemu_rdma_block_for_wrid(RDMAContext *rdma, int wrid_requested,
1472                                    uint32_t *byte_len)
1473{
1474    int num_cq_events = 0, ret = 0;
1475    struct ibv_cq *cq;
1476    void *cq_ctx;
1477    uint64_t wr_id = RDMA_WRID_NONE, wr_id_in;
1478
1479    if (ibv_req_notify_cq(rdma->cq, 0)) {
1480        return -1;
1481    }
1482    /* poll cq first */
1483    while (wr_id != wrid_requested) {
1484        ret = qemu_rdma_poll(rdma, &wr_id_in, byte_len);
1485        if (ret < 0) {
1486            return ret;
1487        }
1488
1489        wr_id = wr_id_in & RDMA_WRID_TYPE_MASK;
1490
1491        if (wr_id == RDMA_WRID_NONE) {
1492            break;
1493        }
1494        if (wr_id != wrid_requested) {
1495            trace_qemu_rdma_block_for_wrid_miss(print_wrid(wrid_requested),
1496                       wrid_requested, print_wrid(wr_id), wr_id);
1497        }
1498    }
1499
1500    if (wr_id == wrid_requested) {
1501        return 0;
1502    }
1503
1504    while (1) {
1505        /*
1506         * Coroutine doesn't start until process_incoming_migration()
1507         * so don't yield unless we know we're running inside of a coroutine.
1508         */
1509        if (rdma->migration_started_on_destination) {
1510            yield_until_fd_readable(rdma->comp_channel->fd);
1511        }
1512
1513        if (ibv_get_cq_event(rdma->comp_channel, &cq, &cq_ctx)) {
1514            perror("ibv_get_cq_event");
1515            goto err_block_for_wrid;
1516        }
1517
1518        num_cq_events++;
1519
1520        if (ibv_req_notify_cq(cq, 0)) {
1521            goto err_block_for_wrid;
1522        }
1523
1524        while (wr_id != wrid_requested) {
1525            ret = qemu_rdma_poll(rdma, &wr_id_in, byte_len);
1526            if (ret < 0) {
1527                goto err_block_for_wrid;
1528            }
1529
1530            wr_id = wr_id_in & RDMA_WRID_TYPE_MASK;
1531
1532            if (wr_id == RDMA_WRID_NONE) {
1533                break;
1534            }
1535            if (wr_id != wrid_requested) {
1536                trace_qemu_rdma_block_for_wrid_miss(print_wrid(wrid_requested),
1537                                   wrid_requested, print_wrid(wr_id), wr_id);
1538            }
1539        }
1540
1541        if (wr_id == wrid_requested) {
1542            goto success_block_for_wrid;
1543        }
1544    }
1545
1546success_block_for_wrid:
1547    if (num_cq_events) {
1548        ibv_ack_cq_events(cq, num_cq_events);
1549    }
1550    return 0;
1551
1552err_block_for_wrid:
1553    if (num_cq_events) {
1554        ibv_ack_cq_events(cq, num_cq_events);
1555    }
1556    return ret;
1557}
1558
1559/*
1560 * Post a SEND message work request for the control channel
1561 * containing some data and block until the post completes.
1562 */
1563static int qemu_rdma_post_send_control(RDMAContext *rdma, uint8_t *buf,
1564                                       RDMAControlHeader *head)
1565{
1566    int ret = 0;
1567    RDMAWorkRequestData *wr = &rdma->wr_data[RDMA_WRID_CONTROL];
1568    struct ibv_send_wr *bad_wr;
1569    struct ibv_sge sge = {
1570                           .addr = (uintptr_t)(wr->control),
1571                           .length = head->len + sizeof(RDMAControlHeader),
1572                           .lkey = wr->control_mr->lkey,
1573                         };
1574    struct ibv_send_wr send_wr = {
1575                                   .wr_id = RDMA_WRID_SEND_CONTROL,
1576                                   .opcode = IBV_WR_SEND,
1577                                   .send_flags = IBV_SEND_SIGNALED,
1578                                   .sg_list = &sge,
1579                                   .num_sge = 1,
1580                                };
1581
1582    trace_qemu_rdma_post_send_control(control_desc[head->type]);
1583
1584    /*
1585     * We don't actually need to do a memcpy() in here if we used
1586     * the "sge" properly, but since we're only sending control messages
1587     * (not RAM in a performance-critical path), then its OK for now.
1588     *
1589     * The copy makes the RDMAControlHeader simpler to manipulate
1590     * for the time being.
1591     */
1592    assert(head->len <= RDMA_CONTROL_MAX_BUFFER - sizeof(*head));
1593    memcpy(wr->control, head, sizeof(RDMAControlHeader));
1594    control_to_network((void *) wr->control);
1595
1596    if (buf) {
1597        memcpy(wr->control + sizeof(RDMAControlHeader), buf, head->len);
1598    }
1599
1600
1601    ret = ibv_post_send(rdma->qp, &send_wr, &bad_wr);
1602
1603    if (ret > 0) {
1604        error_report("Failed to use post IB SEND for control");
1605        return -ret;
1606    }
1607
1608    ret = qemu_rdma_block_for_wrid(rdma, RDMA_WRID_SEND_CONTROL, NULL);
1609    if (ret < 0) {
1610        error_report("rdma migration: send polling control error");
1611    }
1612
1613    return ret;
1614}
1615
1616/*
1617 * Post a RECV work request in anticipation of some future receipt
1618 * of data on the control channel.
1619 */
1620static int qemu_rdma_post_recv_control(RDMAContext *rdma, int idx)
1621{
1622    struct ibv_recv_wr *bad_wr;
1623    struct ibv_sge sge = {
1624                            .addr = (uintptr_t)(rdma->wr_data[idx].control),
1625                            .length = RDMA_CONTROL_MAX_BUFFER,
1626                            .lkey = rdma->wr_data[idx].control_mr->lkey,
1627                         };
1628
1629    struct ibv_recv_wr recv_wr = {
1630                                    .wr_id = RDMA_WRID_RECV_CONTROL + idx,
1631                                    .sg_list = &sge,
1632                                    .num_sge = 1,
1633                                 };
1634
1635
1636    if (ibv_post_recv(rdma->qp, &recv_wr, &bad_wr)) {
1637        return -1;
1638    }
1639
1640    return 0;
1641}
1642
1643/*
1644 * Block and wait for a RECV control channel message to arrive.
1645 */
1646static int qemu_rdma_exchange_get_response(RDMAContext *rdma,
1647                RDMAControlHeader *head, int expecting, int idx)
1648{
1649    uint32_t byte_len;
1650    int ret = qemu_rdma_block_for_wrid(rdma, RDMA_WRID_RECV_CONTROL + idx,
1651                                       &byte_len);
1652
1653    if (ret < 0) {
1654        error_report("rdma migration: recv polling control error!");
1655        return ret;
1656    }
1657
1658    network_to_control((void *) rdma->wr_data[idx].control);
1659    memcpy(head, rdma->wr_data[idx].control, sizeof(RDMAControlHeader));
1660
1661    trace_qemu_rdma_exchange_get_response_start(control_desc[expecting]);
1662
1663    if (expecting == RDMA_CONTROL_NONE) {
1664        trace_qemu_rdma_exchange_get_response_none(control_desc[head->type],
1665                                             head->type);
1666    } else if (head->type != expecting || head->type == RDMA_CONTROL_ERROR) {
1667        error_report("Was expecting a %s (%d) control message"
1668                ", but got: %s (%d), length: %d",
1669                control_desc[expecting], expecting,
1670                control_desc[head->type], head->type, head->len);
1671        return -EIO;
1672    }
1673    if (head->len > RDMA_CONTROL_MAX_BUFFER - sizeof(*head)) {
1674        error_report("too long length: %d", head->len);
1675        return -EINVAL;
1676    }
1677    if (sizeof(*head) + head->len != byte_len) {
1678        error_report("Malformed length: %d byte_len %d", head->len, byte_len);
1679        return -EINVAL;
1680    }
1681
1682    return 0;
1683}
1684
1685/*
1686 * When a RECV work request has completed, the work request's
1687 * buffer is pointed at the header.
1688 *
1689 * This will advance the pointer to the data portion
1690 * of the control message of the work request's buffer that
1691 * was populated after the work request finished.
1692 */
1693static void qemu_rdma_move_header(RDMAContext *rdma, int idx,
1694                                  RDMAControlHeader *head)
1695{
1696    rdma->wr_data[idx].control_len = head->len;
1697    rdma->wr_data[idx].control_curr =
1698        rdma->wr_data[idx].control + sizeof(RDMAControlHeader);
1699}
1700
1701/*
1702 * This is an 'atomic' high-level operation to deliver a single, unified
1703 * control-channel message.
1704 *
1705 * Additionally, if the user is expecting some kind of reply to this message,
1706 * they can request a 'resp' response message be filled in by posting an
1707 * additional work request on behalf of the user and waiting for an additional
1708 * completion.
1709 *
1710 * The extra (optional) response is used during registration to us from having
1711 * to perform an *additional* exchange of message just to provide a response by
1712 * instead piggy-backing on the acknowledgement.
1713 */
1714static int qemu_rdma_exchange_send(RDMAContext *rdma, RDMAControlHeader *head,
1715                                   uint8_t *data, RDMAControlHeader *resp,
1716                                   int *resp_idx,
1717                                   int (*callback)(RDMAContext *rdma))
1718{
1719    int ret = 0;
1720
1721    /*
1722     * Wait until the dest is ready before attempting to deliver the message
1723     * by waiting for a READY message.
1724     */
1725    if (rdma->control_ready_expected) {
1726        RDMAControlHeader resp;
1727        ret = qemu_rdma_exchange_get_response(rdma,
1728                                    &resp, RDMA_CONTROL_READY, RDMA_WRID_READY);
1729        if (ret < 0) {
1730            return ret;
1731        }
1732    }
1733
1734    /*
1735     * If the user is expecting a response, post a WR in anticipation of it.
1736     */
1737    if (resp) {
1738        ret = qemu_rdma_post_recv_control(rdma, RDMA_WRID_DATA);
1739        if (ret) {
1740            error_report("rdma migration: error posting"
1741                    " extra control recv for anticipated result!");
1742            return ret;
1743        }
1744    }
1745
1746    /*
1747     * Post a WR to replace the one we just consumed for the READY message.
1748     */
1749    ret = qemu_rdma_post_recv_control(rdma, RDMA_WRID_READY);
1750    if (ret) {
1751        error_report("rdma migration: error posting first control recv!");
1752        return ret;
1753    }
1754
1755    /*
1756     * Deliver the control message that was requested.
1757     */
1758    ret = qemu_rdma_post_send_control(rdma, data, head);
1759
1760    if (ret < 0) {
1761        error_report("Failed to send control buffer!");
1762        return ret;
1763    }
1764
1765    /*
1766     * If we're expecting a response, block and wait for it.
1767     */
1768    if (resp) {
1769        if (callback) {
1770            trace_qemu_rdma_exchange_send_issue_callback();
1771            ret = callback(rdma);
1772            if (ret < 0) {
1773                return ret;
1774            }
1775        }
1776
1777        trace_qemu_rdma_exchange_send_waiting(control_desc[resp->type]);
1778        ret = qemu_rdma_exchange_get_response(rdma, resp,
1779                                              resp->type, RDMA_WRID_DATA);
1780
1781        if (ret < 0) {
1782            return ret;
1783        }
1784
1785        qemu_rdma_move_header(rdma, RDMA_WRID_DATA, resp);
1786        if (resp_idx) {
1787            *resp_idx = RDMA_WRID_DATA;
1788        }
1789        trace_qemu_rdma_exchange_send_received(control_desc[resp->type]);
1790    }
1791
1792    rdma->control_ready_expected = 1;
1793
1794    return 0;
1795}
1796
1797/*
1798 * This is an 'atomic' high-level operation to receive a single, unified
1799 * control-channel message.
1800 */
1801static int qemu_rdma_exchange_recv(RDMAContext *rdma, RDMAControlHeader *head,
1802                                int expecting)
1803{
1804    RDMAControlHeader ready = {
1805                                .len = 0,
1806                                .type = RDMA_CONTROL_READY,
1807                                .repeat = 1,
1808                              };
1809    int ret;
1810
1811    /*
1812     * Inform the source that we're ready to receive a message.
1813     */
1814    ret = qemu_rdma_post_send_control(rdma, NULL, &ready);
1815
1816    if (ret < 0) {
1817        error_report("Failed to send control buffer!");
1818        return ret;
1819    }
1820
1821    /*
1822     * Block and wait for the message.
1823     */
1824    ret = qemu_rdma_exchange_get_response(rdma, head,
1825                                          expecting, RDMA_WRID_READY);
1826
1827    if (ret < 0) {
1828        return ret;
1829    }
1830
1831    qemu_rdma_move_header(rdma, RDMA_WRID_READY, head);
1832
1833    /*
1834     * Post a new RECV work request to replace the one we just consumed.
1835     */
1836    ret = qemu_rdma_post_recv_control(rdma, RDMA_WRID_READY);
1837    if (ret) {
1838        error_report("rdma migration: error posting second control recv!");
1839        return ret;
1840    }
1841
1842    return 0;
1843}
1844
1845/*
1846 * Write an actual chunk of memory using RDMA.
1847 *
1848 * If we're using dynamic registration on the dest-side, we have to
1849 * send a registration command first.
1850 */
1851static int qemu_rdma_write_one(QEMUFile *f, RDMAContext *rdma,
1852                               int current_index, uint64_t current_addr,
1853                               uint64_t length)
1854{
1855    struct ibv_sge sge;
1856    struct ibv_send_wr send_wr = { 0 };
1857    struct ibv_send_wr *bad_wr;
1858    int reg_result_idx, ret, count = 0;
1859    uint64_t chunk, chunks;
1860    uint8_t *chunk_start, *chunk_end;
1861    RDMALocalBlock *block = &(rdma->local_ram_blocks.block[current_index]);
1862    RDMARegister reg;
1863    RDMARegisterResult *reg_result;
1864    RDMAControlHeader resp = { .type = RDMA_CONTROL_REGISTER_RESULT };
1865    RDMAControlHeader head = { .len = sizeof(RDMARegister),
1866                               .type = RDMA_CONTROL_REGISTER_REQUEST,
1867                               .repeat = 1,
1868                             };
1869
1870retry:
1871    sge.addr = (uintptr_t)(block->local_host_addr +
1872                            (current_addr - block->offset));
1873    sge.length = length;
1874
1875    chunk = ram_chunk_index(block->local_host_addr,
1876                            (uint8_t *)(uintptr_t)sge.addr);
1877    chunk_start = ram_chunk_start(block, chunk);
1878
1879    if (block->is_ram_block) {
1880        chunks = length / (1UL << RDMA_REG_CHUNK_SHIFT);
1881
1882        if (chunks && ((length % (1UL << RDMA_REG_CHUNK_SHIFT)) == 0)) {
1883            chunks--;
1884        }
1885    } else {
1886        chunks = block->length / (1UL << RDMA_REG_CHUNK_SHIFT);
1887
1888        if (chunks && ((block->length % (1UL << RDMA_REG_CHUNK_SHIFT)) == 0)) {
1889            chunks--;
1890        }
1891    }
1892
1893    trace_qemu_rdma_write_one_top(chunks + 1,
1894                                  (chunks + 1) *
1895                                  (1UL << RDMA_REG_CHUNK_SHIFT) / 1024 / 1024);
1896
1897    chunk_end = ram_chunk_end(block, chunk + chunks);
1898
1899    if (!rdma->pin_all) {
1900#ifdef RDMA_UNREGISTRATION_EXAMPLE
1901        qemu_rdma_unregister_waiting(rdma);
1902#endif
1903    }
1904
1905    while (test_bit(chunk, block->transit_bitmap)) {
1906        (void)count;
1907        trace_qemu_rdma_write_one_block(count++, current_index, chunk,
1908                sge.addr, length, rdma->nb_sent, block->nb_chunks);
1909
1910        ret = qemu_rdma_block_for_wrid(rdma, RDMA_WRID_RDMA_WRITE, NULL);
1911
1912        if (ret < 0) {
1913            error_report("Failed to Wait for previous write to complete "
1914                    "block %d chunk %" PRIu64
1915                    " current %" PRIu64 " len %" PRIu64 " %d",
1916                    current_index, chunk, sge.addr, length, rdma->nb_sent);
1917            return ret;
1918        }
1919    }
1920
1921    if (!rdma->pin_all || !block->is_ram_block) {
1922        if (!block->remote_keys[chunk]) {
1923            /*
1924             * This chunk has not yet been registered, so first check to see
1925             * if the entire chunk is zero. If so, tell the other size to
1926             * memset() + madvise() the entire chunk without RDMA.
1927             */
1928
1929            if (can_use_buffer_find_nonzero_offset((void *)(uintptr_t)sge.addr,
1930                                                   length)
1931                   && buffer_find_nonzero_offset((void *)(uintptr_t)sge.addr,
1932                                                    length) == length) {
1933                RDMACompress comp = {
1934                                        .offset = current_addr,
1935                                        .value = 0,
1936                                        .block_idx = current_index,
1937                                        .length = length,
1938                                    };
1939
1940                head.len = sizeof(comp);
1941                head.type = RDMA_CONTROL_COMPRESS;
1942
1943                trace_qemu_rdma_write_one_zero(chunk, sge.length,
1944                                               current_index, current_addr);
1945
1946                compress_to_network(rdma, &comp);
1947                ret = qemu_rdma_exchange_send(rdma, &head,
1948                                (uint8_t *) &comp, NULL, NULL, NULL);
1949
1950                if (ret < 0) {
1951                    return -EIO;
1952                }
1953
1954                acct_update_position(f, sge.length, true);
1955
1956                return 1;
1957            }
1958
1959            /*
1960             * Otherwise, tell other side to register.
1961             */
1962            reg.current_index = current_index;
1963            if (block->is_ram_block) {
1964                reg.key.current_addr = current_addr;
1965            } else {
1966                reg.key.chunk = chunk;
1967            }
1968            reg.chunks = chunks;
1969
1970            trace_qemu_rdma_write_one_sendreg(chunk, sge.length, current_index,
1971                                              current_addr);
1972
1973            register_to_network(rdma, &reg);
1974            ret = qemu_rdma_exchange_send(rdma, &head, (uint8_t *) &reg,
1975                                    &resp, &reg_result_idx, NULL);
1976            if (ret < 0) {
1977                return ret;
1978            }
1979
1980            /* try to overlap this single registration with the one we sent. */
1981            if (qemu_rdma_register_and_get_keys(rdma, block, sge.addr,
1982                                                &sge.lkey, NULL, chunk,
1983                                                chunk_start, chunk_end)) {
1984                error_report("cannot get lkey");
1985                return -EINVAL;
1986            }
1987
1988            reg_result = (RDMARegisterResult *)
1989                    rdma->wr_data[reg_result_idx].control_curr;
1990
1991            network_to_result(reg_result);
1992
1993            trace_qemu_rdma_write_one_recvregres(block->remote_keys[chunk],
1994                                                 reg_result->rkey, chunk);
1995
1996            block->remote_keys[chunk] = reg_result->rkey;
1997            block->remote_host_addr = reg_result->host_addr;
1998        } else {
1999            /* already registered before */
2000            if (qemu_rdma_register_and_get_keys(rdma, block, sge.addr,
2001                                                &sge.lkey, NULL, chunk,
2002                                                chunk_start, chunk_end)) {
2003                error_report("cannot get lkey!");
2004                return -EINVAL;
2005            }
2006        }
2007
2008        send_wr.wr.rdma.rkey = block->remote_keys[chunk];
2009    } else {
2010        send_wr.wr.rdma.rkey = block->remote_rkey;
2011
2012        if (qemu_rdma_register_and_get_keys(rdma, block, sge.addr,
2013                                                     &sge.lkey, NULL, chunk,
2014                                                     chunk_start, chunk_end)) {
2015            error_report("cannot get lkey!");
2016            return -EINVAL;
2017        }
2018    }
2019
2020    /*
2021     * Encode the ram block index and chunk within this wrid.
2022     * We will use this information at the time of completion
2023     * to figure out which bitmap to check against and then which
2024     * chunk in the bitmap to look for.
2025     */
2026    send_wr.wr_id = qemu_rdma_make_wrid(RDMA_WRID_RDMA_WRITE,
2027                                        current_index, chunk);
2028
2029    send_wr.opcode = IBV_WR_RDMA_WRITE;
2030    send_wr.send_flags = IBV_SEND_SIGNALED;
2031    send_wr.sg_list = &sge;
2032    send_wr.num_sge = 1;
2033    send_wr.wr.rdma.remote_addr = block->remote_host_addr +
2034                                (current_addr - block->offset);
2035
2036    trace_qemu_rdma_write_one_post(chunk, sge.addr, send_wr.wr.rdma.remote_addr,
2037                                   sge.length);
2038
2039    /*
2040     * ibv_post_send() does not return negative error numbers,
2041     * per the specification they are positive - no idea why.
2042     */
2043    ret = ibv_post_send(rdma->qp, &send_wr, &bad_wr);
2044
2045    if (ret == ENOMEM) {
2046        trace_qemu_rdma_write_one_queue_full();
2047        ret = qemu_rdma_block_for_wrid(rdma, RDMA_WRID_RDMA_WRITE, NULL);
2048        if (ret < 0) {
2049            error_report("rdma migration: failed to make "
2050                         "room in full send queue! %d", ret);
2051            return ret;
2052        }
2053
2054        goto retry;
2055
2056    } else if (ret > 0) {
2057        perror("rdma migration: post rdma write failed");
2058        return -ret;
2059    }
2060
2061    set_bit(chunk, block->transit_bitmap);
2062    acct_update_position(f, sge.length, false);
2063    rdma->total_writes++;
2064
2065    return 0;
2066}
2067
2068/*
2069 * Push out any unwritten RDMA operations.
2070 *
2071 * We support sending out multiple chunks at the same time.
2072 * Not all of them need to get signaled in the completion queue.
2073 */
2074static int qemu_rdma_write_flush(QEMUFile *f, RDMAContext *rdma)
2075{
2076    int ret;
2077
2078    if (!rdma->current_length) {
2079        return 0;
2080    }
2081
2082    ret = qemu_rdma_write_one(f, rdma,
2083            rdma->current_index, rdma->current_addr, rdma->current_length);
2084
2085    if (ret < 0) {
2086        return ret;
2087    }
2088
2089    if (ret == 0) {
2090        rdma->nb_sent++;
2091        trace_qemu_rdma_write_flush(rdma->nb_sent);
2092    }
2093
2094    rdma->current_length = 0;
2095    rdma->current_addr = 0;
2096
2097    return 0;
2098}
2099
2100static inline int qemu_rdma_buffer_mergable(RDMAContext *rdma,
2101                    uint64_t offset, uint64_t len)
2102{
2103    RDMALocalBlock *block;
2104    uint8_t *host_addr;
2105    uint8_t *chunk_end;
2106
2107    if (rdma->current_index < 0) {
2108        return 0;
2109    }
2110
2111    if (rdma->current_chunk < 0) {
2112        return 0;
2113    }
2114
2115    block = &(rdma->local_ram_blocks.block[rdma->current_index]);
2116    host_addr = block->local_host_addr + (offset - block->offset);
2117    chunk_end = ram_chunk_end(block, rdma->current_chunk);
2118
2119    if (rdma->current_length == 0) {
2120        return 0;
2121    }
2122
2123    /*
2124     * Only merge into chunk sequentially.
2125     */
2126    if (offset != (rdma->current_addr + rdma->current_length)) {
2127        return 0;
2128    }
2129
2130    if (offset < block->offset) {
2131        return 0;
2132    }
2133
2134    if ((offset + len) > (block->offset + block->length)) {
2135        return 0;
2136    }
2137
2138    if ((host_addr + len) > chunk_end) {
2139        return 0;
2140    }
2141
2142    return 1;
2143}
2144
2145/*
2146 * We're not actually writing here, but doing three things:
2147 *
2148 * 1. Identify the chunk the buffer belongs to.
2149 * 2. If the chunk is full or the buffer doesn't belong to the current
2150 *    chunk, then start a new chunk and flush() the old chunk.
2151 * 3. To keep the hardware busy, we also group chunks into batches
2152 *    and only require that a batch gets acknowledged in the completion
2153 *    qeueue instead of each individual chunk.
2154 */
2155static int qemu_rdma_write(QEMUFile *f, RDMAContext *rdma,
2156                           uint64_t block_offset, uint64_t offset,
2157                           uint64_t len)
2158{
2159    uint64_t current_addr = block_offset + offset;
2160    uint64_t index = rdma->current_index;
2161    uint64_t chunk = rdma->current_chunk;
2162    int ret;
2163
2164    /* If we cannot merge it, we flush the current buffer first. */
2165    if (!qemu_rdma_buffer_mergable(rdma, current_addr, len)) {
2166        ret = qemu_rdma_write_flush(f, rdma);
2167        if (ret) {
2168            return ret;
2169        }
2170        rdma->current_length = 0;
2171        rdma->current_addr = current_addr;
2172
2173        ret = qemu_rdma_search_ram_block(rdma, block_offset,
2174                                         offset, len, &index, &chunk);
2175        if (ret) {
2176            error_report("ram block search failed");
2177            return ret;
2178        }
2179        rdma->current_index = index;
2180        rdma->current_chunk = chunk;
2181    }
2182
2183    /* merge it */
2184    rdma->current_length += len;
2185
2186    /* flush it if buffer is too large */
2187    if (rdma->current_length >= RDMA_MERGE_MAX) {
2188        return qemu_rdma_write_flush(f, rdma);
2189    }
2190
2191    return 0;
2192}
2193
2194static void qemu_rdma_cleanup(RDMAContext *rdma)
2195{
2196    struct rdma_cm_event *cm_event;
2197    int ret, idx;
2198
2199    if (rdma->cm_id && rdma->connected) {
2200        if (rdma->error_state) {
2201            RDMAControlHeader head = { .len = 0,
2202                                       .type = RDMA_CONTROL_ERROR,
2203                                       .repeat = 1,
2204                                     };
2205            error_report("Early error. Sending error.");
2206            qemu_rdma_post_send_control(rdma, NULL, &head);
2207        }
2208
2209        ret = rdma_disconnect(rdma->cm_id);
2210        if (!ret) {
2211            trace_qemu_rdma_cleanup_waiting_for_disconnect();
2212            ret = rdma_get_cm_event(rdma->channel, &cm_event);
2213            if (!ret) {
2214                rdma_ack_cm_event(cm_event);
2215            }
2216        }
2217        trace_qemu_rdma_cleanup_disconnect();
2218        rdma->connected = false;
2219    }
2220
2221    g_free(rdma->dest_blocks);
2222    rdma->dest_blocks = NULL;
2223
2224    for (idx = 0; idx < RDMA_WRID_MAX; idx++) {
2225        if (rdma->wr_data[idx].control_mr) {
2226            rdma->total_registrations--;
2227            ibv_dereg_mr(rdma->wr_data[idx].control_mr);
2228        }
2229        rdma->wr_data[idx].control_mr = NULL;
2230    }
2231
2232    if (rdma->local_ram_blocks.block) {
2233        while (rdma->local_ram_blocks.nb_blocks) {
2234            rdma_delete_block(rdma, &rdma->local_ram_blocks.block[0]);
2235        }
2236    }
2237
2238    if (rdma->qp) {
2239        rdma_destroy_qp(rdma->cm_id);
2240        rdma->qp = NULL;
2241    }
2242    if (rdma->cq) {
2243        ibv_destroy_cq(rdma->cq);
2244        rdma->cq = NULL;
2245    }
2246    if (rdma->comp_channel) {
2247        ibv_destroy_comp_channel(rdma->comp_channel);
2248        rdma->comp_channel = NULL;
2249    }
2250    if (rdma->pd) {
2251        ibv_dealloc_pd(rdma->pd);
2252        rdma->pd = NULL;
2253    }
2254    if (rdma->cm_id) {
2255        rdma_destroy_id(rdma->cm_id);
2256        rdma->cm_id = NULL;
2257    }
2258    if (rdma->listen_id) {
2259        rdma_destroy_id(rdma->listen_id);
2260        rdma->listen_id = NULL;
2261    }
2262    if (rdma->channel) {
2263        rdma_destroy_event_channel(rdma->channel);
2264        rdma->channel = NULL;
2265    }
2266    g_free(rdma->host);
2267    rdma->host = NULL;
2268}
2269
2270
2271static int qemu_rdma_source_init(RDMAContext *rdma, Error **errp, bool pin_all)
2272{
2273    int ret, idx;
2274    Error *local_err = NULL, **temp = &local_err;
2275
2276    /*
2277     * Will be validated against destination's actual capabilities
2278     * after the connect() completes.
2279     */
2280    rdma->pin_all = pin_all;
2281
2282    ret = qemu_rdma_resolve_host(rdma, temp);
2283    if (ret) {
2284        goto err_rdma_source_init;
2285    }
2286
2287    ret = qemu_rdma_alloc_pd_cq(rdma);
2288    if (ret) {
2289        ERROR(temp, "rdma migration: error allocating pd and cq! Your mlock()"
2290                    " limits may be too low. Please check $ ulimit -a # and "
2291                    "search for 'ulimit -l' in the output");
2292        goto err_rdma_source_init;
2293    }
2294
2295    ret = qemu_rdma_alloc_qp(rdma);
2296    if (ret) {
2297        ERROR(temp, "rdma migration: error allocating qp!");
2298        goto err_rdma_source_init;
2299    }
2300
2301    ret = qemu_rdma_init_ram_blocks(rdma);
2302    if (ret) {
2303        ERROR(temp, "rdma migration: error initializing ram blocks!");
2304        goto err_rdma_source_init;
2305    }
2306
2307    /* Build the hash that maps from offset to RAMBlock */
2308    rdma->blockmap = g_hash_table_new(g_direct_hash, g_direct_equal);
2309    for (idx = 0; idx < rdma->local_ram_blocks.nb_blocks; idx++) {
2310        g_hash_table_insert(rdma->blockmap,
2311                (void *)(uintptr_t)rdma->local_ram_blocks.block[idx].offset,
2312                &rdma->local_ram_blocks.block[idx]);
2313    }
2314
2315    for (idx = 0; idx < RDMA_WRID_MAX; idx++) {
2316        ret = qemu_rdma_reg_control(rdma, idx);
2317        if (ret) {
2318            ERROR(temp, "rdma migration: error registering %d control!",
2319                                                            idx);
2320            goto err_rdma_source_init;
2321        }
2322    }
2323
2324    return 0;
2325
2326err_rdma_source_init:
2327    error_propagate(errp, local_err);
2328    qemu_rdma_cleanup(rdma);
2329    return -1;
2330}
2331
2332static int qemu_rdma_connect(RDMAContext *rdma, Error **errp)
2333{
2334    RDMACapabilities cap = {
2335                                .version = RDMA_CONTROL_VERSION_CURRENT,
2336                                .flags = 0,
2337                           };
2338    struct rdma_conn_param conn_param = { .initiator_depth = 2,
2339                                          .retry_count = 5,
2340                                          .private_data = &cap,
2341                                          .private_data_len = sizeof(cap),
2342                                        };
2343    struct rdma_cm_event *cm_event;
2344    int ret;
2345
2346    /*
2347     * Only negotiate the capability with destination if the user
2348     * on the source first requested the capability.
2349     */
2350    if (rdma->pin_all) {
2351        trace_qemu_rdma_connect_pin_all_requested();
2352        cap.flags |= RDMA_CAPABILITY_PIN_ALL;
2353    }
2354
2355    caps_to_network(&cap);
2356
2357    ret = rdma_connect(rdma->cm_id, &conn_param);
2358    if (ret) {
2359        perror("rdma_connect");
2360        ERROR(errp, "connecting to destination!");
2361        goto err_rdma_source_connect;
2362    }
2363
2364    ret = rdma_get_cm_event(rdma->channel, &cm_event);
2365    if (ret) {
2366        perror("rdma_get_cm_event after rdma_connect");
2367        ERROR(errp, "connecting to destination!");
2368        rdma_ack_cm_event(cm_event);
2369        goto err_rdma_source_connect;
2370    }
2371
2372    if (cm_event->event != RDMA_CM_EVENT_ESTABLISHED) {
2373        perror("rdma_get_cm_event != EVENT_ESTABLISHED after rdma_connect");
2374        ERROR(errp, "connecting to destination!");
2375        rdma_ack_cm_event(cm_event);
2376        goto err_rdma_source_connect;
2377    }
2378    rdma->connected = true;
2379
2380    memcpy(&cap, cm_event->param.conn.private_data, sizeof(cap));
2381    network_to_caps(&cap);
2382
2383    /*
2384     * Verify that the *requested* capabilities are supported by the destination
2385     * and disable them otherwise.
2386     */
2387    if (rdma->pin_all && !(cap.flags & RDMA_CAPABILITY_PIN_ALL)) {
2388        ERROR(errp, "Server cannot support pinning all memory. "
2389                        "Will register memory dynamically.");
2390        rdma->pin_all = false;
2391    }
2392
2393    trace_qemu_rdma_connect_pin_all_outcome(rdma->pin_all);
2394
2395    rdma_ack_cm_event(cm_event);
2396
2397    ret = qemu_rdma_post_recv_control(rdma, RDMA_WRID_READY);
2398    if (ret) {
2399        ERROR(errp, "posting second control recv!");
2400        goto err_rdma_source_connect;
2401    }
2402
2403    rdma->control_ready_expected = 1;
2404    rdma->nb_sent = 0;
2405    return 0;
2406
2407err_rdma_source_connect:
2408    qemu_rdma_cleanup(rdma);
2409    return -1;
2410}
2411
2412static int qemu_rdma_dest_init(RDMAContext *rdma, Error **errp)
2413{
2414    int ret, idx;
2415    struct rdma_cm_id *listen_id;
2416    char ip[40] = "unknown";
2417    struct rdma_addrinfo *res, *e;
2418    char port_str[16];
2419
2420    for (idx = 0; idx < RDMA_WRID_MAX; idx++) {
2421        rdma->wr_data[idx].control_len = 0;
2422        rdma->wr_data[idx].control_curr = NULL;
2423    }
2424
2425    if (!rdma->host || !rdma->host[0]) {
2426        ERROR(errp, "RDMA host is not set!");
2427        rdma->error_state = -EINVAL;
2428        return -1;
2429    }
2430    /* create CM channel */
2431    rdma->channel = rdma_create_event_channel();
2432    if (!rdma->channel) {
2433        ERROR(errp, "could not create rdma event channel");
2434        rdma->error_state = -EINVAL;
2435        return -1;
2436    }
2437
2438    /* create CM id */
2439    ret = rdma_create_id(rdma->channel, &listen_id, NULL, RDMA_PS_TCP);
2440    if (ret) {
2441        ERROR(errp, "could not create cm_id!");
2442        goto err_dest_init_create_listen_id;
2443    }
2444
2445    snprintf(port_str, 16, "%d", rdma->port);
2446    port_str[15] = '\0';
2447
2448    ret = rdma_getaddrinfo(rdma->host, port_str, NULL, &res);
2449    if (ret < 0) {
2450        ERROR(errp, "could not rdma_getaddrinfo address %s", rdma->host);
2451        goto err_dest_init_bind_addr;
2452    }
2453
2454    for (e = res; e != NULL; e = e->ai_next) {
2455        inet_ntop(e->ai_family,
2456            &((struct sockaddr_in *) e->ai_dst_addr)->sin_addr, ip, sizeof ip);
2457        trace_qemu_rdma_dest_init_trying(rdma->host, ip);
2458        ret = rdma_bind_addr(listen_id, e->ai_dst_addr);
2459        if (ret) {
2460            continue;
2461        }
2462        if (e->ai_family == AF_INET6) {
2463            ret = qemu_rdma_broken_ipv6_kernel(errp, listen_id->verbs);
2464            if (ret) {
2465                continue;
2466            }
2467        }
2468        break;
2469    }
2470
2471    if (!e) {
2472        ERROR(errp, "Error: could not rdma_bind_addr!");
2473        goto err_dest_init_bind_addr;
2474    }
2475
2476    rdma->listen_id = listen_id;
2477    qemu_rdma_dump_gid("dest_init", listen_id);
2478    return 0;
2479
2480err_dest_init_bind_addr:
2481    rdma_destroy_id(listen_id);
2482err_dest_init_create_listen_id:
2483    rdma_destroy_event_channel(rdma->channel);
2484    rdma->channel = NULL;
2485    rdma->error_state = ret;
2486    return ret;
2487
2488}
2489
2490static void *qemu_rdma_data_init(const char *host_port, Error **errp)
2491{
2492    RDMAContext *rdma = NULL;
2493    InetSocketAddress *addr;
2494
2495    if (host_port) {
2496        rdma = g_new0(RDMAContext, 1);
2497        rdma->current_index = -1;
2498        rdma->current_chunk = -1;
2499
2500        addr = inet_parse(host_port, NULL);
2501        if (addr != NULL) {
2502            rdma->port = atoi(addr->port);
2503            rdma->host = g_strdup(addr->host);
2504        } else {
2505            ERROR(errp, "bad RDMA migration address '%s'", host_port);
2506            g_free(rdma);
2507            rdma = NULL;
2508        }
2509
2510        qapi_free_InetSocketAddress(addr);
2511    }
2512
2513    return rdma;
2514}
2515
2516/*
2517 * QEMUFile interface to the control channel.
2518 * SEND messages for control only.
2519 * VM's ram is handled with regular RDMA messages.
2520 */
2521static ssize_t qemu_rdma_put_buffer(void *opaque, const uint8_t *buf,
2522                                    int64_t pos, size_t size)
2523{
2524    QEMUFileRDMA *r = opaque;
2525    QEMUFile *f = r->file;
2526    RDMAContext *rdma = r->rdma;
2527    size_t remaining = size;
2528    uint8_t * data = (void *) buf;
2529    int ret;
2530
2531    CHECK_ERROR_STATE();
2532
2533    /*
2534     * Push out any writes that
2535     * we're queued up for VM's ram.
2536     */
2537    ret = qemu_rdma_write_flush(f, rdma);
2538    if (ret < 0) {
2539        rdma->error_state = ret;
2540        return ret;
2541    }
2542
2543    while (remaining) {
2544        RDMAControlHeader head;
2545
2546        r->len = MIN(remaining, RDMA_SEND_INCREMENT);
2547        remaining -= r->len;
2548
2549        /* Guaranteed to fit due to RDMA_SEND_INCREMENT MIN above */
2550        head.len = (uint32_t)r->len;
2551        head.type = RDMA_CONTROL_QEMU_FILE;
2552
2553        ret = qemu_rdma_exchange_send(rdma, &head, data, NULL, NULL, NULL);
2554
2555        if (ret < 0) {
2556            rdma->error_state = ret;
2557            return ret;
2558        }
2559
2560        data += r->len;
2561    }
2562
2563    return size;
2564}
2565
2566static size_t qemu_rdma_fill(RDMAContext *rdma, uint8_t *buf,
2567                             size_t size, int idx)
2568{
2569    size_t len = 0;
2570
2571    if (rdma->wr_data[idx].control_len) {
2572        trace_qemu_rdma_fill(rdma->wr_data[idx].control_len, size);
2573
2574        len = MIN(size, rdma->wr_data[idx].control_len);
2575        memcpy(buf, rdma->wr_data[idx].control_curr, len);
2576        rdma->wr_data[idx].control_curr += len;
2577        rdma->wr_data[idx].control_len -= len;
2578    }
2579
2580    return len;
2581}
2582
2583/*
2584 * QEMUFile interface to the control channel.
2585 * RDMA links don't use bytestreams, so we have to
2586 * return bytes to QEMUFile opportunistically.
2587 */
2588static ssize_t qemu_rdma_get_buffer(void *opaque, uint8_t *buf,
2589                                    int64_t pos, size_t size)
2590{
2591    QEMUFileRDMA *r = opaque;
2592    RDMAContext *rdma = r->rdma;
2593    RDMAControlHeader head;
2594    int ret = 0;
2595
2596    CHECK_ERROR_STATE();
2597
2598    /*
2599     * First, we hold on to the last SEND message we
2600     * were given and dish out the bytes until we run
2601     * out of bytes.
2602     */
2603    r->len = qemu_rdma_fill(r->rdma, buf, size, 0);
2604    if (r->len) {
2605        return r->len;
2606    }
2607
2608    /*
2609     * Once we run out, we block and wait for another
2610     * SEND message to arrive.
2611     */
2612    ret = qemu_rdma_exchange_recv(rdma, &head, RDMA_CONTROL_QEMU_FILE);
2613
2614    if (ret < 0) {
2615        rdma->error_state = ret;
2616        return ret;
2617    }
2618
2619    /*
2620     * SEND was received with new bytes, now try again.
2621     */
2622    return qemu_rdma_fill(r->rdma, buf, size, 0);
2623}
2624
2625/*
2626 * Block until all the outstanding chunks have been delivered by the hardware.
2627 */
2628static int qemu_rdma_drain_cq(QEMUFile *f, RDMAContext *rdma)
2629{
2630    int ret;
2631
2632    if (qemu_rdma_write_flush(f, rdma) < 0) {
2633        return -EIO;
2634    }
2635
2636    while (rdma->nb_sent) {
2637        ret = qemu_rdma_block_for_wrid(rdma, RDMA_WRID_RDMA_WRITE, NULL);
2638        if (ret < 0) {
2639            error_report("rdma migration: complete polling error!");
2640            return -EIO;
2641        }
2642    }
2643
2644    qemu_rdma_unregister_waiting(rdma);
2645
2646    return 0;
2647}
2648
2649static int qemu_rdma_close(void *opaque)
2650{
2651    trace_qemu_rdma_close();
2652    QEMUFileRDMA *r = opaque;
2653    if (r->rdma) {
2654        qemu_rdma_cleanup(r->rdma);
2655        g_free(r->rdma);
2656    }
2657    g_free(r);
2658    return 0;
2659}
2660
2661/*
2662 * Parameters:
2663 *    @offset == 0 :
2664 *        This means that 'block_offset' is a full virtual address that does not
2665 *        belong to a RAMBlock of the virtual machine and instead
2666 *        represents a private malloc'd memory area that the caller wishes to
2667 *        transfer.
2668 *
2669 *    @offset != 0 :
2670 *        Offset is an offset to be added to block_offset and used
2671 *        to also lookup the corresponding RAMBlock.
2672 *
2673 *    @size > 0 :
2674 *        Initiate an transfer this size.
2675 *
2676 *    @size == 0 :
2677 *        A 'hint' or 'advice' that means that we wish to speculatively
2678 *        and asynchronously unregister this memory. In this case, there is no
2679 *        guarantee that the unregister will actually happen, for example,
2680 *        if the memory is being actively transmitted. Additionally, the memory
2681 *        may be re-registered at any future time if a write within the same
2682 *        chunk was requested again, even if you attempted to unregister it
2683 *        here.
2684 *
2685 *    @size < 0 : TODO, not yet supported
2686 *        Unregister the memory NOW. This means that the caller does not
2687 *        expect there to be any future RDMA transfers and we just want to clean
2688 *        things up. This is used in case the upper layer owns the memory and
2689 *        cannot wait for qemu_fclose() to occur.
2690 *
2691 *    @bytes_sent : User-specificed pointer to indicate how many bytes were
2692 *                  sent. Usually, this will not be more than a few bytes of
2693 *                  the protocol because most transfers are sent asynchronously.
2694 */
2695static size_t qemu_rdma_save_page(QEMUFile *f, void *opaque,
2696                                  ram_addr_t block_offset, ram_addr_t offset,
2697                                  size_t size, uint64_t *bytes_sent)
2698{
2699    QEMUFileRDMA *rfile = opaque;
2700    RDMAContext *rdma = rfile->rdma;
2701    int ret;
2702
2703    CHECK_ERROR_STATE();
2704
2705    qemu_fflush(f);
2706
2707    if (size > 0) {
2708        /*
2709         * Add this page to the current 'chunk'. If the chunk
2710         * is full, or the page doen't belong to the current chunk,
2711         * an actual RDMA write will occur and a new chunk will be formed.
2712         */
2713        ret = qemu_rdma_write(f, rdma, block_offset, offset, size);
2714        if (ret < 0) {
2715            error_report("rdma migration: write error! %d", ret);
2716            goto err;
2717        }
2718
2719        /*
2720         * We always return 1 bytes because the RDMA
2721         * protocol is completely asynchronous. We do not yet know
2722         * whether an  identified chunk is zero or not because we're
2723         * waiting for other pages to potentially be merged with
2724         * the current chunk. So, we have to call qemu_update_position()
2725         * later on when the actual write occurs.
2726         */
2727        if (bytes_sent) {
2728            *bytes_sent = 1;
2729        }
2730    } else {
2731        uint64_t index, chunk;
2732
2733        /* TODO: Change QEMUFileOps prototype to be signed: size_t => long
2734        if (size < 0) {
2735            ret = qemu_rdma_drain_cq(f, rdma);
2736            if (ret < 0) {
2737                fprintf(stderr, "rdma: failed to synchronously drain"
2738                                " completion queue before unregistration.\n");
2739                goto err;
2740            }
2741        }
2742        */
2743
2744        ret = qemu_rdma_search_ram_block(rdma, block_offset,
2745                                         offset, size, &index, &chunk);
2746
2747        if (ret) {
2748            error_report("ram block search failed");
2749            goto err;
2750        }
2751
2752        qemu_rdma_signal_unregister(rdma, index, chunk, 0);
2753
2754        /*
2755         * TODO: Synchronous, guaranteed unregistration (should not occur during
2756         * fast-path). Otherwise, unregisters will process on the next call to
2757         * qemu_rdma_drain_cq()
2758        if (size < 0) {
2759            qemu_rdma_unregister_waiting(rdma);
2760        }
2761        */
2762    }
2763
2764    /*
2765     * Drain the Completion Queue if possible, but do not block,
2766     * just poll.
2767     *
2768     * If nothing to poll, the end of the iteration will do this
2769     * again to make sure we don't overflow the request queue.
2770     */
2771    while (1) {
2772        uint64_t wr_id, wr_id_in;
2773        int ret = qemu_rdma_poll(rdma, &wr_id_in, NULL);
2774        if (ret < 0) {
2775            error_report("rdma migration: polling error! %d", ret);
2776            goto err;
2777        }
2778
2779        wr_id = wr_id_in & RDMA_WRID_TYPE_MASK;
2780
2781        if (wr_id == RDMA_WRID_NONE) {
2782            break;
2783        }
2784    }
2785
2786    return RAM_SAVE_CONTROL_DELAYED;
2787err:
2788    rdma->error_state = ret;
2789    return ret;
2790}
2791
2792static int qemu_rdma_accept(RDMAContext *rdma)
2793{
2794    RDMACapabilities cap;
2795    struct rdma_conn_param conn_param = {
2796                                            .responder_resources = 2,
2797                                            .private_data = &cap,
2798                                            .private_data_len = sizeof(cap),
2799                                         };
2800    struct rdma_cm_event *cm_event;
2801    struct ibv_context *verbs;
2802    int ret = -EINVAL;
2803    int idx;
2804
2805    ret = rdma_get_cm_event(rdma->channel, &cm_event);
2806    if (ret) {
2807        goto err_rdma_dest_wait;
2808    }
2809
2810    if (cm_event->event != RDMA_CM_EVENT_CONNECT_REQUEST) {
2811        rdma_ack_cm_event(cm_event);
2812        goto err_rdma_dest_wait;
2813    }
2814
2815    memcpy(&cap, cm_event->param.conn.private_data, sizeof(cap));
2816
2817    network_to_caps(&cap);
2818
2819    if (cap.version < 1 || cap.version > RDMA_CONTROL_VERSION_CURRENT) {
2820            error_report("Unknown source RDMA version: %d, bailing...",
2821                            cap.version);
2822            rdma_ack_cm_event(cm_event);
2823            goto err_rdma_dest_wait;
2824    }
2825
2826    /*
2827     * Respond with only the capabilities this version of QEMU knows about.
2828     */
2829    cap.flags &= known_capabilities;
2830
2831    /*
2832     * Enable the ones that we do know about.
2833     * Add other checks here as new ones are introduced.
2834     */
2835    if (cap.flags & RDMA_CAPABILITY_PIN_ALL) {
2836        rdma->pin_all = true;
2837    }
2838
2839    rdma->cm_id = cm_event->id;
2840    verbs = cm_event->id->verbs;
2841
2842    rdma_ack_cm_event(cm_event);
2843
2844    trace_qemu_rdma_accept_pin_state(rdma->pin_all);
2845
2846    caps_to_network(&cap);
2847
2848    trace_qemu_rdma_accept_pin_verbsc(verbs);
2849
2850    if (!rdma->verbs) {
2851        rdma->verbs = verbs;
2852    } else if (rdma->verbs != verbs) {
2853            error_report("ibv context not matching %p, %p!", rdma->verbs,
2854                         verbs);
2855            goto err_rdma_dest_wait;
2856    }
2857
2858    qemu_rdma_dump_id("dest_init", verbs);
2859
2860    ret = qemu_rdma_alloc_pd_cq(rdma);
2861    if (ret) {
2862        error_report("rdma migration: error allocating pd and cq!");
2863        goto err_rdma_dest_wait;
2864    }
2865
2866    ret = qemu_rdma_alloc_qp(rdma);
2867    if (ret) {
2868        error_report("rdma migration: error allocating qp!");
2869        goto err_rdma_dest_wait;
2870    }
2871
2872    ret = qemu_rdma_init_ram_blocks(rdma);
2873    if (ret) {
2874        error_report("rdma migration: error initializing ram blocks!");
2875        goto err_rdma_dest_wait;
2876    }
2877
2878    for (idx = 0; idx < RDMA_WRID_MAX; idx++) {
2879        ret = qemu_rdma_reg_control(rdma, idx);
2880        if (ret) {
2881            error_report("rdma: error registering %d control", idx);
2882            goto err_rdma_dest_wait;
2883        }
2884    }
2885
2886    qemu_set_fd_handler(rdma->channel->fd, NULL, NULL, NULL);
2887
2888    ret = rdma_accept(rdma->cm_id, &conn_param);
2889    if (ret) {
2890        error_report("rdma_accept returns %d", ret);
2891        goto err_rdma_dest_wait;
2892    }
2893
2894    ret = rdma_get_cm_event(rdma->channel, &cm_event);
2895    if (ret) {
2896        error_report("rdma_accept get_cm_event failed %d", ret);
2897        goto err_rdma_dest_wait;
2898    }
2899
2900    if (cm_event->event != RDMA_CM_EVENT_ESTABLISHED) {
2901        error_report("rdma_accept not event established");
2902        rdma_ack_cm_event(cm_event);
2903        goto err_rdma_dest_wait;
2904    }
2905
2906    rdma_ack_cm_event(cm_event);
2907    rdma->connected = true;
2908
2909    ret = qemu_rdma_post_recv_control(rdma, RDMA_WRID_READY);
2910    if (ret) {
2911        error_report("rdma migration: error posting second control recv");
2912        goto err_rdma_dest_wait;
2913    }
2914
2915    qemu_rdma_dump_gid("dest_connect", rdma->cm_id);
2916
2917    return 0;
2918
2919err_rdma_dest_wait:
2920    rdma->error_state = ret;
2921    qemu_rdma_cleanup(rdma);
2922    return ret;
2923}
2924
2925static int dest_ram_sort_func(const void *a, const void *b)
2926{
2927    unsigned int a_index = ((const RDMALocalBlock *)a)->src_index;
2928    unsigned int b_index = ((const RDMALocalBlock *)b)->src_index;
2929
2930    return (a_index < b_index) ? -1 : (a_index != b_index);
2931}
2932
2933/*
2934 * During each iteration of the migration, we listen for instructions
2935 * by the source VM to perform dynamic page registrations before they
2936 * can perform RDMA operations.
2937 *
2938 * We respond with the 'rkey'.
2939 *
2940 * Keep doing this until the source tells us to stop.
2941 */
2942static int qemu_rdma_registration_handle(QEMUFile *f, void *opaque)
2943{
2944    RDMAControlHeader reg_resp = { .len = sizeof(RDMARegisterResult),
2945                               .type = RDMA_CONTROL_REGISTER_RESULT,
2946                               .repeat = 0,
2947                             };
2948    RDMAControlHeader unreg_resp = { .len = 0,
2949                               .type = RDMA_CONTROL_UNREGISTER_FINISHED,
2950                               .repeat = 0,
2951                             };
2952    RDMAControlHeader blocks = { .type = RDMA_CONTROL_RAM_BLOCKS_RESULT,
2953                                 .repeat = 1 };
2954    QEMUFileRDMA *rfile = opaque;
2955    RDMAContext *rdma = rfile->rdma;
2956    RDMALocalBlocks *local = &rdma->local_ram_blocks;
2957    RDMAControlHeader head;
2958    RDMARegister *reg, *registers;
2959    RDMACompress *comp;
2960    RDMARegisterResult *reg_result;
2961    static RDMARegisterResult results[RDMA_CONTROL_MAX_COMMANDS_PER_MESSAGE];
2962    RDMALocalBlock *block;
2963    void *host_addr;
2964    int ret = 0;
2965    int idx = 0;
2966    int count = 0;
2967    int i = 0;
2968
2969    CHECK_ERROR_STATE();
2970
2971    do {
2972        trace_qemu_rdma_registration_handle_wait();
2973
2974        ret = qemu_rdma_exchange_recv(rdma, &head, RDMA_CONTROL_NONE);
2975
2976        if (ret < 0) {
2977            break;
2978        }
2979
2980        if (head.repeat > RDMA_CONTROL_MAX_COMMANDS_PER_MESSAGE) {
2981            error_report("rdma: Too many requests in this message (%d)."
2982                            "Bailing.", head.repeat);
2983            ret = -EIO;
2984            break;
2985        }
2986
2987        switch (head.type) {
2988        case RDMA_CONTROL_COMPRESS:
2989            comp = (RDMACompress *) rdma->wr_data[idx].control_curr;
2990            network_to_compress(comp);
2991
2992            trace_qemu_rdma_registration_handle_compress(comp->length,
2993                                                         comp->block_idx,
2994                                                         comp->offset);
2995            if (comp->block_idx >= rdma->local_ram_blocks.nb_blocks) {
2996                error_report("rdma: 'compress' bad block index %u (vs %d)",
2997                             (unsigned int)comp->block_idx,
2998                             rdma->local_ram_blocks.nb_blocks);
2999                ret = -EIO;
3000                goto out;
3001            }
3002            block = &(rdma->local_ram_blocks.block[comp->block_idx]);
3003
3004            host_addr = block->local_host_addr +
3005                            (comp->offset - block->offset);
3006
3007            ram_handle_compressed(host_addr, comp->value, comp->length);
3008            break;
3009
3010        case RDMA_CONTROL_REGISTER_FINISHED:
3011            trace_qemu_rdma_registration_handle_finished();
3012            goto out;
3013
3014        case RDMA_CONTROL_RAM_BLOCKS_REQUEST:
3015            trace_qemu_rdma_registration_handle_ram_blocks();
3016
3017            /* Sort our local RAM Block list so it's the same as the source,
3018             * we can do this since we've filled in a src_index in the list
3019             * as we received the RAMBlock list earlier.
3020             */
3021            qsort(rdma->local_ram_blocks.block,
3022                  rdma->local_ram_blocks.nb_blocks,
3023                  sizeof(RDMALocalBlock), dest_ram_sort_func);
3024            if (rdma->pin_all) {
3025                ret = qemu_rdma_reg_whole_ram_blocks(rdma);
3026                if (ret) {
3027                    error_report("rdma migration: error dest "
3028                                    "registering ram blocks");
3029                    goto out;
3030                }
3031            }
3032
3033            /*
3034             * Dest uses this to prepare to transmit the RAMBlock descriptions
3035             * to the source VM after connection setup.
3036             * Both sides use the "remote" structure to communicate and update
3037             * their "local" descriptions with what was sent.
3038             */
3039            for (i = 0; i < local->nb_blocks; i++) {
3040                rdma->dest_blocks[i].remote_host_addr =
3041                    (uintptr_t)(local->block[i].local_host_addr);
3042
3043                if (rdma->pin_all) {
3044                    rdma->dest_blocks[i].remote_rkey = local->block[i].mr->rkey;
3045                }
3046
3047                rdma->dest_blocks[i].offset = local->block[i].offset;
3048                rdma->dest_blocks[i].length = local->block[i].length;
3049
3050                dest_block_to_network(&rdma->dest_blocks[i]);
3051                trace_qemu_rdma_registration_handle_ram_blocks_loop(
3052                    local->block[i].block_name,
3053                    local->block[i].offset,
3054                    local->block[i].length,
3055                    local->block[i].local_host_addr,
3056                    local->block[i].src_index);
3057            }
3058
3059            blocks.len = rdma->local_ram_blocks.nb_blocks
3060                                                * sizeof(RDMADestBlock);
3061
3062
3063            ret = qemu_rdma_post_send_control(rdma,
3064                                        (uint8_t *) rdma->dest_blocks, &blocks);
3065
3066            if (ret < 0) {
3067                error_report("rdma migration: error sending remote info");
3068                goto out;
3069            }
3070
3071            break;
3072        case RDMA_CONTROL_REGISTER_REQUEST:
3073            trace_qemu_rdma_registration_handle_register(head.repeat);
3074
3075            reg_resp.repeat = head.repeat;
3076            registers = (RDMARegister *) rdma->wr_data[idx].control_curr;
3077
3078            for (count = 0; count < head.repeat; count++) {
3079                uint64_t chunk;
3080                uint8_t *chunk_start, *chunk_end;
3081
3082                reg = &registers[count];
3083                network_to_register(reg);
3084
3085                reg_result = &results[count];
3086
3087                trace_qemu_rdma_registration_handle_register_loop(count,
3088                         reg->current_index, reg->key.current_addr, reg->chunks);
3089
3090                if (reg->current_index >= rdma->local_ram_blocks.nb_blocks) {
3091                    error_report("rdma: 'register' bad block index %u (vs %d)",
3092                                 (unsigned int)reg->current_index,
3093                                 rdma->local_ram_blocks.nb_blocks);
3094                    ret = -ENOENT;
3095                    goto out;
3096                }
3097                block = &(rdma->local_ram_blocks.block[reg->current_index]);
3098                if (block->is_ram_block) {
3099                    if (block->offset > reg->key.current_addr) {
3100                        error_report("rdma: bad register address for block %s"
3101                            " offset: %" PRIx64 " current_addr: %" PRIx64,
3102                            block->block_name, block->offset,
3103                            reg->key.current_addr);
3104                        ret = -ERANGE;
3105                        goto out;
3106                    }
3107                    host_addr = (block->local_host_addr +
3108                                (reg->key.current_addr - block->offset));
3109                    chunk = ram_chunk_index(block->local_host_addr,
3110                                            (uint8_t *) host_addr);
3111                } else {
3112                    chunk = reg->key.chunk;
3113                    host_addr = block->local_host_addr +
3114                        (reg->key.chunk * (1UL << RDMA_REG_CHUNK_SHIFT));
3115                    /* Check for particularly bad chunk value */
3116                    if (host_addr < (void *)block->local_host_addr) {
3117                        error_report("rdma: bad chunk for block %s"
3118                            " chunk: %" PRIx64,
3119                            block->block_name, reg->key.chunk);
3120                        ret = -ERANGE;
3121                        goto out;
3122                    }
3123                }
3124                chunk_start = ram_chunk_start(block, chunk);
3125                chunk_end = ram_chunk_end(block, chunk + reg->chunks);
3126                if (qemu_rdma_register_and_get_keys(rdma, block,
3127                            (uintptr_t)host_addr, NULL, &reg_result->rkey,
3128                            chunk, chunk_start, chunk_end)) {
3129                    error_report("cannot get rkey");
3130                    ret = -EINVAL;
3131                    goto out;
3132                }
3133
3134                reg_result->host_addr = (uintptr_t)block->local_host_addr;
3135
3136                trace_qemu_rdma_registration_handle_register_rkey(
3137                                                           reg_result->rkey);
3138
3139                result_to_network(reg_result);
3140            }
3141
3142            ret = qemu_rdma_post_send_control(rdma,
3143                            (uint8_t *) results, &reg_resp);
3144
3145            if (ret < 0) {
3146                error_report("Failed to send control buffer");
3147                goto out;
3148            }
3149            break;
3150        case RDMA_CONTROL_UNREGISTER_REQUEST:
3151            trace_qemu_rdma_registration_handle_unregister(head.repeat);
3152            unreg_resp.repeat = head.repeat;
3153            registers = (RDMARegister *) rdma->wr_data[idx].control_curr;
3154
3155            for (count = 0; count < head.repeat; count++) {
3156                reg = &registers[count];
3157                network_to_register(reg);
3158
3159                trace_qemu_rdma_registration_handle_unregister_loop(count,
3160                           reg->current_index, reg->key.chunk);
3161
3162                block = &(rdma->local_ram_blocks.block[reg->current_index]);
3163
3164                ret = ibv_dereg_mr(block->pmr[reg->key.chunk]);
3165                block->pmr[reg->key.chunk] = NULL;
3166
3167                if (ret != 0) {
3168                    perror("rdma unregistration chunk failed");
3169                    ret = -ret;
3170                    goto out;
3171                }
3172
3173                rdma->total_registrations--;
3174
3175                trace_qemu_rdma_registration_handle_unregister_success(
3176                                                       reg->key.chunk);
3177            }
3178
3179            ret = qemu_rdma_post_send_control(rdma, NULL, &unreg_resp);
3180
3181            if (ret < 0) {
3182                error_report("Failed to send control buffer");
3183                goto out;
3184            }
3185            break;
3186        case RDMA_CONTROL_REGISTER_RESULT:
3187            error_report("Invalid RESULT message at dest.");
3188            ret = -EIO;
3189            goto out;
3190        default:
3191            error_report("Unknown control message %s", control_desc[head.type]);
3192            ret = -EIO;
3193            goto out;
3194        }
3195    } while (1);
3196out:
3197    if (ret < 0) {
3198        rdma->error_state = ret;
3199    }
3200    return ret;
3201}
3202
3203/* Destination:
3204 * Called via a ram_control_load_hook during the initial RAM load section which
3205 * lists the RAMBlocks by name.  This lets us know the order of the RAMBlocks
3206 * on the source.
3207 * We've already built our local RAMBlock list, but not yet sent the list to
3208 * the source.
3209 */
3210static int rdma_block_notification_handle(QEMUFileRDMA *rfile, const char *name)
3211{
3212    RDMAContext *rdma = rfile->rdma;
3213    int curr;
3214    int found = -1;
3215
3216    /* Find the matching RAMBlock in our local list */
3217    for (curr = 0; curr < rdma->local_ram_blocks.nb_blocks; curr++) {
3218        if (!strcmp(rdma->local_ram_blocks.block[curr].block_name, name)) {
3219            found = curr;
3220            break;
3221        }
3222    }
3223
3224    if (found == -1) {
3225        error_report("RAMBlock '%s' not found on destination", name);
3226        return -ENOENT;
3227    }
3228
3229    rdma->local_ram_blocks.block[curr].src_index = rdma->next_src_index;
3230    trace_rdma_block_notification_handle(name, rdma->next_src_index);
3231    rdma->next_src_index++;
3232
3233    return 0;
3234}
3235
3236static int rdma_load_hook(QEMUFile *f, void *opaque, uint64_t flags, void *data)
3237{
3238    switch (flags) {
3239    case RAM_CONTROL_BLOCK_REG:
3240        return rdma_block_notification_handle(opaque, data);
3241
3242    case RAM_CONTROL_HOOK:
3243        return qemu_rdma_registration_handle(f, opaque);
3244
3245    default:
3246        /* Shouldn't be called with any other values */
3247        abort();
3248    }
3249}
3250
3251static int qemu_rdma_registration_start(QEMUFile *f, void *opaque,
3252                                        uint64_t flags, void *data)
3253{
3254    QEMUFileRDMA *rfile = opaque;
3255    RDMAContext *rdma = rfile->rdma;
3256
3257    CHECK_ERROR_STATE();
3258
3259    trace_qemu_rdma_registration_start(flags);
3260    qemu_put_be64(f, RAM_SAVE_FLAG_HOOK);
3261    qemu_fflush(f);
3262
3263    return 0;
3264}
3265
3266/*
3267 * Inform dest that dynamic registrations are done for now.
3268 * First, flush writes, if any.
3269 */
3270static int qemu_rdma_registration_stop(QEMUFile *f, void *opaque,
3271                                       uint64_t flags, void *data)
3272{
3273    Error *local_err = NULL, **errp = &local_err;
3274    QEMUFileRDMA *rfile = opaque;
3275    RDMAContext *rdma = rfile->rdma;
3276    RDMAControlHeader head = { .len = 0, .repeat = 1 };
3277    int ret = 0;
3278
3279    CHECK_ERROR_STATE();
3280
3281    qemu_fflush(f);
3282    ret = qemu_rdma_drain_cq(f, rdma);
3283
3284    if (ret < 0) {
3285        goto err;
3286    }
3287
3288    if (flags == RAM_CONTROL_SETUP) {
3289        RDMAControlHeader resp = {.type = RDMA_CONTROL_RAM_BLOCKS_RESULT };
3290        RDMALocalBlocks *local = &rdma->local_ram_blocks;
3291        int reg_result_idx, i, nb_dest_blocks;
3292
3293        head.type = RDMA_CONTROL_RAM_BLOCKS_REQUEST;
3294        trace_qemu_rdma_registration_stop_ram();
3295
3296        /*
3297         * Make sure that we parallelize the pinning on both sides.
3298         * For very large guests, doing this serially takes a really
3299         * long time, so we have to 'interleave' the pinning locally
3300         * with the control messages by performing the pinning on this
3301         * side before we receive the control response from the other
3302         * side that the pinning has completed.
3303         */
3304        ret = qemu_rdma_exchange_send(rdma, &head, NULL, &resp,
3305                    &reg_result_idx, rdma->pin_all ?
3306                    qemu_rdma_reg_whole_ram_blocks : NULL);
3307        if (ret < 0) {
3308            ERROR(errp, "receiving remote info!");
3309            return ret;
3310        }
3311
3312        nb_dest_blocks = resp.len / sizeof(RDMADestBlock);
3313
3314        /*
3315         * The protocol uses two different sets of rkeys (mutually exclusive):
3316         * 1. One key to represent the virtual address of the entire ram block.
3317         *    (dynamic chunk registration disabled - pin everything with one rkey.)
3318         * 2. One to represent individual chunks within a ram block.
3319         *    (dynamic chunk registration enabled - pin individual chunks.)
3320         *
3321         * Once the capability is successfully negotiated, the destination transmits
3322         * the keys to use (or sends them later) including the virtual addresses
3323         * and then propagates the remote ram block descriptions to his local copy.
3324         */
3325
3326        if (local->nb_blocks != nb_dest_blocks) {
3327            ERROR(errp, "ram blocks mismatch (Number of blocks %d vs %d) "
3328                        "Your QEMU command line parameters are probably "
3329                        "not identical on both the source and destination.",
3330                        local->nb_blocks, nb_dest_blocks);
3331            rdma->error_state = -EINVAL;
3332            return -EINVAL;
3333        }
3334
3335        qemu_rdma_move_header(rdma, reg_result_idx, &resp);
3336        memcpy(rdma->dest_blocks,
3337            rdma->wr_data[reg_result_idx].control_curr, resp.len);
3338        for (i = 0; i < nb_dest_blocks; i++) {
3339            network_to_dest_block(&rdma->dest_blocks[i]);
3340
3341            /* We require that the blocks are in the same order */
3342            if (rdma->dest_blocks[i].length != local->block[i].length) {
3343                ERROR(errp, "Block %s/%d has a different length %" PRIu64
3344                            "vs %" PRIu64, local->block[i].block_name, i,
3345                            local->block[i].length,
3346                            rdma->dest_blocks[i].length);
3347                rdma->error_state = -EINVAL;
3348                return -EINVAL;
3349            }
3350            local->block[i].remote_host_addr =
3351                    rdma->dest_blocks[i].remote_host_addr;
3352            local->block[i].remote_rkey = rdma->dest_blocks[i].remote_rkey;
3353        }
3354    }
3355
3356    trace_qemu_rdma_registration_stop(flags);
3357
3358    head.type = RDMA_CONTROL_REGISTER_FINISHED;
3359    ret = qemu_rdma_exchange_send(rdma, &head, NULL, NULL, NULL, NULL);
3360
3361    if (ret < 0) {
3362        goto err;
3363    }
3364
3365    return 0;
3366err:
3367    rdma->error_state = ret;
3368    return ret;
3369}
3370
3371static int qemu_rdma_get_fd(void *opaque)
3372{
3373    QEMUFileRDMA *rfile = opaque;
3374    RDMAContext *rdma = rfile->rdma;
3375
3376    return rdma->comp_channel->fd;
3377}
3378
3379static const QEMUFileOps rdma_read_ops = {
3380    .get_buffer    = qemu_rdma_get_buffer,
3381    .get_fd        = qemu_rdma_get_fd,
3382    .close         = qemu_rdma_close,
3383    .hook_ram_load = rdma_load_hook,
3384};
3385
3386static const QEMUFileOps rdma_write_ops = {
3387    .put_buffer         = qemu_rdma_put_buffer,
3388    .close              = qemu_rdma_close,
3389    .before_ram_iterate = qemu_rdma_registration_start,
3390    .after_ram_iterate  = qemu_rdma_registration_stop,
3391    .save_page          = qemu_rdma_save_page,
3392};
3393
3394static void *qemu_fopen_rdma(RDMAContext *rdma, const char *mode)
3395{
3396    QEMUFileRDMA *r;
3397
3398    if (qemu_file_mode_is_not_valid(mode)) {
3399        return NULL;
3400    }
3401
3402    r = g_new0(QEMUFileRDMA, 1);
3403    r->rdma = rdma;
3404
3405    if (mode[0] == 'w') {
3406        r->file = qemu_fopen_ops(r, &rdma_write_ops);
3407    } else {
3408        r->file = qemu_fopen_ops(r, &rdma_read_ops);
3409    }
3410
3411    return r->file;
3412}
3413
3414static void rdma_accept_incoming_migration(void *opaque)
3415{
3416    RDMAContext *rdma = opaque;
3417    int ret;
3418    QEMUFile *f;
3419    Error *local_err = NULL, **errp = &local_err;
3420
3421    trace_qemu_rdma_accept_incoming_migration();
3422    ret = qemu_rdma_accept(rdma);
3423
3424    if (ret) {
3425        ERROR(errp, "RDMA Migration initialization failed!");
3426        return;
3427    }
3428
3429    trace_qemu_rdma_accept_incoming_migration_accepted();
3430
3431    f = qemu_fopen_rdma(rdma, "rb");
3432    if (f == NULL) {
3433        ERROR(errp, "could not qemu_fopen_rdma!");
3434        qemu_rdma_cleanup(rdma);
3435        return;
3436    }
3437
3438    rdma->migration_started_on_destination = 1;
3439    process_incoming_migration(f);
3440}
3441
3442void rdma_start_incoming_migration(const char *host_port, Error **errp)
3443{
3444    int ret;
3445    RDMAContext *rdma;
3446    Error *local_err = NULL;
3447
3448    trace_rdma_start_incoming_migration();
3449    rdma = qemu_rdma_data_init(host_port, &local_err);
3450
3451    if (rdma == NULL) {
3452        goto err;
3453    }
3454
3455    ret = qemu_rdma_dest_init(rdma, &local_err);
3456
3457    if (ret) {
3458        goto err;
3459    }
3460
3461    trace_rdma_start_incoming_migration_after_dest_init();
3462
3463    ret = rdma_listen(rdma->listen_id, 5);
3464
3465    if (ret) {
3466        ERROR(errp, "listening on socket!");
3467        goto err;
3468    }
3469
3470    trace_rdma_start_incoming_migration_after_rdma_listen();
3471
3472    qemu_set_fd_handler(rdma->channel->fd, rdma_accept_incoming_migration,
3473                        NULL, (void *)(intptr_t)rdma);
3474    return;
3475err:
3476    error_propagate(errp, local_err);
3477    g_free(rdma);
3478}
3479
3480void rdma_start_outgoing_migration(void *opaque,
3481                            const char *host_port, Error **errp)
3482{
3483    MigrationState *s = opaque;
3484    Error *local_err = NULL, **temp = &local_err;
3485    RDMAContext *rdma = qemu_rdma_data_init(host_port, &local_err);
3486    int ret = 0;
3487
3488    if (rdma == NULL) {
3489        ERROR(temp, "Failed to initialize RDMA data structures! %d", ret);
3490        goto err;
3491    }
3492
3493    ret = qemu_rdma_source_init(rdma, &local_err,
3494        s->enabled_capabilities[MIGRATION_CAPABILITY_RDMA_PIN_ALL]);
3495
3496    if (ret) {
3497        goto err;
3498    }
3499
3500    trace_rdma_start_outgoing_migration_after_rdma_source_init();
3501    ret = qemu_rdma_connect(rdma, &local_err);
3502
3503    if (ret) {
3504        goto err;
3505    }
3506
3507    trace_rdma_start_outgoing_migration_after_rdma_connect();
3508
3509    s->to_dst_file = qemu_fopen_rdma(rdma, "wb");
3510    migrate_fd_connect(s);
3511    return;
3512err:
3513    error_propagate(errp, local_err);
3514    g_free(rdma);
3515    migrate_fd_error(s);
3516}
3517