Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Fetching contributors…

Octocat-spinner-32-eaf2f5

Cannot retrieve contributors at this time

file 860 lines (701 sloc) 23.337 kb
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859
/*
* QEMU live block migration
*
* Copyright IBM, Corp. 2009
*
* Authors:
* Liran Schour <lirans@il.ibm.com>
*
* This work is licensed under the terms of the GNU GPL, version 2. See
* the COPYING file in the top-level directory.
*
* Contributions after 2012-01-13 are licensed under the terms of the
* GNU GPL, version 2 or (at your option) any later version.
*/

#include "qemu-common.h"
#include "block/block_int.h"
#include "hw/hw.h"
#include "qemu/queue.h"
#include "qemu/timer.h"
#include "migration/block.h"
#include "migration/migration.h"
#include "sysemu/blockdev.h"
#include <assert.h>

#define BLOCK_SIZE (1 << 20)
#define BDRV_SECTORS_PER_DIRTY_CHUNK (BLOCK_SIZE >> BDRV_SECTOR_BITS)

#define BLK_MIG_FLAG_DEVICE_BLOCK 0x01
#define BLK_MIG_FLAG_EOS 0x02
#define BLK_MIG_FLAG_PROGRESS 0x04
#define BLK_MIG_FLAG_ZERO_BLOCK 0x08

#define MAX_IS_ALLOCATED_SEARCH 65536

//#define DEBUG_BLK_MIGRATION

#ifdef DEBUG_BLK_MIGRATION
#define DPRINTF(fmt, ...) \
do { printf("blk_migration: " fmt, ## __VA_ARGS__); } while (0)
#else
#define DPRINTF(fmt, ...) \
do { } while (0)
#endif

typedef struct BlkMigDevState {
    /* Written during setup phase. Can be read without a lock. */
    BlockDriverState *bs;
    int shared_base;
    int64_t total_sectors;
    QSIMPLEQ_ENTRY(BlkMigDevState) entry;

    /* Only used by migration thread. Does not need a lock. */
    int bulk_completed;
    int64_t cur_sector;
    int64_t cur_dirty;

    /* Protected by block migration lock. */
    unsigned long *aio_bitmap;
    int64_t completed_sectors;
    BdrvDirtyBitmap *dirty_bitmap;
} BlkMigDevState;

typedef struct BlkMigBlock {
    /* Only used by migration thread. */
    uint8_t *buf;
    BlkMigDevState *bmds;
    int64_t sector;
    int nr_sectors;
    struct iovec iov;
    QEMUIOVector qiov;
    BlockDriverAIOCB *aiocb;

    /* Protected by block migration lock. */
    int ret;
    QSIMPLEQ_ENTRY(BlkMigBlock) entry;
} BlkMigBlock;

typedef struct BlkMigState {
    /* Written during setup phase. Can be read without a lock. */
    int blk_enable;
    int shared_base;
    QSIMPLEQ_HEAD(bmds_list, BlkMigDevState) bmds_list;
    int64_t total_sector_sum;
    bool zero_blocks;

    /* Protected by lock. */
    QSIMPLEQ_HEAD(blk_list, BlkMigBlock) blk_list;
    int submitted;
    int read_done;

    /* Only used by migration thread. Does not need a lock. */
    int transferred;
    int prev_progress;
    int bulk_completed;

    /* Lock must be taken _inside_ the iothread lock. */
    QemuMutex lock;
} BlkMigState;

static BlkMigState block_mig_state;

static void blk_mig_lock(void)
{
    qemu_mutex_lock(&block_mig_state.lock);
}

static void blk_mig_unlock(void)
{
    qemu_mutex_unlock(&block_mig_state.lock);
}

/* Must run outside of the iothread lock during the bulk phase,
* or the VM will stall.
*/

static void blk_send(QEMUFile *f, BlkMigBlock * blk)
{
    int len;
    uint64_t flags = BLK_MIG_FLAG_DEVICE_BLOCK;

    if (block_mig_state.zero_blocks &&
        buffer_is_zero(blk->buf, BLOCK_SIZE)) {
        flags |= BLK_MIG_FLAG_ZERO_BLOCK;
    }

    /* sector number and flags */
    qemu_put_be64(f, (blk->sector << BDRV_SECTOR_BITS)
                     | flags);

    /* device name */
    len = strlen(blk->bmds->bs->device_name);
    qemu_put_byte(f, len);
    qemu_put_buffer(f, (uint8_t *)blk->bmds->bs->device_name, len);

    /* if a block is zero we need to flush here since the network
* bandwidth is now a lot higher than the storage device bandwidth.
* thus if we queue zero blocks we slow down the migration */
    if (flags & BLK_MIG_FLAG_ZERO_BLOCK) {
        qemu_fflush(f);
        return;
    }

    qemu_put_buffer(f, blk->buf, BLOCK_SIZE);
}

int blk_mig_active(void)
{
    return !QSIMPLEQ_EMPTY(&block_mig_state.bmds_list);
}

uint64_t blk_mig_bytes_transferred(void)
{
    BlkMigDevState *bmds;
    uint64_t sum = 0;

    blk_mig_lock();
    QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
        sum += bmds->completed_sectors;
    }
    blk_mig_unlock();
    return sum << BDRV_SECTOR_BITS;
}

uint64_t blk_mig_bytes_remaining(void)
{
    return blk_mig_bytes_total() - blk_mig_bytes_transferred();
}

uint64_t blk_mig_bytes_total(void)
{
    BlkMigDevState *bmds;
    uint64_t sum = 0;

    QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
        sum += bmds->total_sectors;
    }
    return sum << BDRV_SECTOR_BITS;
}


/* Called with migration lock held. */

static int bmds_aio_inflight(BlkMigDevState *bmds, int64_t sector)
{
    int64_t chunk = sector / (int64_t)BDRV_SECTORS_PER_DIRTY_CHUNK;

    if ((sector << BDRV_SECTOR_BITS) < bdrv_getlength(bmds->bs)) {
        return !!(bmds->aio_bitmap[chunk / (sizeof(unsigned long) * 8)] &
            (1UL << (chunk % (sizeof(unsigned long) * 8))));
    } else {
        return 0;
    }
}

/* Called with migration lock held. */

static void bmds_set_aio_inflight(BlkMigDevState *bmds, int64_t sector_num,
                             int nb_sectors, int set)
{
    int64_t start, end;
    unsigned long val, idx, bit;

    start = sector_num / BDRV_SECTORS_PER_DIRTY_CHUNK;
    end = (sector_num + nb_sectors - 1) / BDRV_SECTORS_PER_DIRTY_CHUNK;

    for (; start <= end; start++) {
        idx = start / (sizeof(unsigned long) * 8);
        bit = start % (sizeof(unsigned long) * 8);
        val = bmds->aio_bitmap[idx];
        if (set) {
            val |= 1UL << bit;
        } else {
            val &= ~(1UL << bit);
        }
        bmds->aio_bitmap[idx] = val;
    }
}

static void alloc_aio_bitmap(BlkMigDevState *bmds)
{
    BlockDriverState *bs = bmds->bs;
    int64_t bitmap_size;

    bitmap_size = (bdrv_getlength(bs) >> BDRV_SECTOR_BITS) +
            BDRV_SECTORS_PER_DIRTY_CHUNK * 8 - 1;
    bitmap_size /= BDRV_SECTORS_PER_DIRTY_CHUNK * 8;

    bmds->aio_bitmap = g_malloc0(bitmap_size);
}

/* Never hold migration lock when yielding to the main loop! */

static void blk_mig_read_cb(void *opaque, int ret)
{
    BlkMigBlock *blk = opaque;

    blk_mig_lock();
    blk->ret = ret;

    QSIMPLEQ_INSERT_TAIL(&block_mig_state.blk_list, blk, entry);
    bmds_set_aio_inflight(blk->bmds, blk->sector, blk->nr_sectors, 0);

    block_mig_state.submitted--;
    block_mig_state.read_done++;
    assert(block_mig_state.submitted >= 0);
    blk_mig_unlock();
}

/* Called with no lock taken. */

static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
{
    int64_t total_sectors = bmds->total_sectors;
    int64_t cur_sector = bmds->cur_sector;
    BlockDriverState *bs = bmds->bs;
    BlkMigBlock *blk;
    int nr_sectors;

    if (bmds->shared_base) {
        qemu_mutex_lock_iothread();
        while (cur_sector < total_sectors &&
               !bdrv_is_allocated(bs, cur_sector, MAX_IS_ALLOCATED_SEARCH,
                                  &nr_sectors)) {
            cur_sector += nr_sectors;
        }
        qemu_mutex_unlock_iothread();
    }

    if (cur_sector >= total_sectors) {
        bmds->cur_sector = bmds->completed_sectors = total_sectors;
        return 1;
    }

    bmds->completed_sectors = cur_sector;

    cur_sector &= ~((int64_t)BDRV_SECTORS_PER_DIRTY_CHUNK - 1);

    /* we are going to transfer a full block even if it is not allocated */
    nr_sectors = BDRV_SECTORS_PER_DIRTY_CHUNK;

    if (total_sectors - cur_sector < BDRV_SECTORS_PER_DIRTY_CHUNK) {
        nr_sectors = total_sectors - cur_sector;
    }

    blk = g_malloc(sizeof(BlkMigBlock));
    blk->buf = g_malloc(BLOCK_SIZE);
    blk->bmds = bmds;
    blk->sector = cur_sector;
    blk->nr_sectors = nr_sectors;

    blk->iov.iov_base = blk->buf;
    blk->iov.iov_len = nr_sectors * BDRV_SECTOR_SIZE;
    qemu_iovec_init_external(&blk->qiov, &blk->iov, 1);

    blk_mig_lock();
    block_mig_state.submitted++;
    blk_mig_unlock();

    qemu_mutex_lock_iothread();
    blk->aiocb = bdrv_aio_readv(bs, cur_sector, &blk->qiov,
                                nr_sectors, blk_mig_read_cb, blk);

    bdrv_reset_dirty(bs, cur_sector, nr_sectors);
    qemu_mutex_unlock_iothread();

    bmds->cur_sector = cur_sector + nr_sectors;
    return (bmds->cur_sector >= total_sectors);
}

/* Called with iothread lock taken. */

static void set_dirty_tracking(void)
{
    BlkMigDevState *bmds;

    QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
        bmds->dirty_bitmap = bdrv_create_dirty_bitmap(bmds->bs, BLOCK_SIZE);
    }
}

static void unset_dirty_tracking(void)
{
    BlkMigDevState *bmds;

    QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
        bdrv_release_dirty_bitmap(bmds->bs, bmds->dirty_bitmap);
    }
}

static void init_blk_migration_it(void *opaque, BlockDriverState *bs)
{
    BlkMigDevState *bmds;
    int64_t sectors;

    if (!bdrv_is_read_only(bs)) {
        sectors = bdrv_getlength(bs) >> BDRV_SECTOR_BITS;
        if (sectors <= 0) {
            return;
        }

        bmds = g_malloc0(sizeof(BlkMigDevState));
        bmds->bs = bs;
        bmds->bulk_completed = 0;
        bmds->total_sectors = sectors;
        bmds->completed_sectors = 0;
        bmds->shared_base = block_mig_state.shared_base;
        alloc_aio_bitmap(bmds);
        bdrv_set_in_use(bs, 1);
        bdrv_ref(bs);

        block_mig_state.total_sector_sum += sectors;

        if (bmds->shared_base) {
            DPRINTF("Start migration for %s with shared base image\n",
                    bs->device_name);
        } else {
            DPRINTF("Start full migration for %s\n", bs->device_name);
        }

        QSIMPLEQ_INSERT_TAIL(&block_mig_state.bmds_list, bmds, entry);
    }
}

static void init_blk_migration(QEMUFile *f)
{
    block_mig_state.submitted = 0;
    block_mig_state.read_done = 0;
    block_mig_state.transferred = 0;
    block_mig_state.total_sector_sum = 0;
    block_mig_state.prev_progress = -1;
    block_mig_state.bulk_completed = 0;
    block_mig_state.zero_blocks = migrate_zero_blocks();

    bdrv_iterate(init_blk_migration_it, NULL);
}

/* Called with no lock taken. */

static int blk_mig_save_bulked_block(QEMUFile *f)
{
    int64_t completed_sector_sum = 0;
    BlkMigDevState *bmds;
    int progress;
    int ret = 0;

    QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
        if (bmds->bulk_completed == 0) {
            if (mig_save_device_bulk(f, bmds) == 1) {
                /* completed bulk section for this device */
                bmds->bulk_completed = 1;
            }
            completed_sector_sum += bmds->completed_sectors;
            ret = 1;
            break;
        } else {
            completed_sector_sum += bmds->completed_sectors;
        }
    }

    if (block_mig_state.total_sector_sum != 0) {
        progress = completed_sector_sum * 100 /
                   block_mig_state.total_sector_sum;
    } else {
        progress = 100;
    }
    if (progress != block_mig_state.prev_progress) {
        block_mig_state.prev_progress = progress;
        qemu_put_be64(f, (progress << BDRV_SECTOR_BITS)
                         | BLK_MIG_FLAG_PROGRESS);
        DPRINTF("Completed %d %%\r", progress);
    }

    return ret;
}

static void blk_mig_reset_dirty_cursor(void)
{
    BlkMigDevState *bmds;

    QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
        bmds->cur_dirty = 0;
    }
}

/* Called with iothread lock taken. */

static int mig_save_device_dirty(QEMUFile *f, BlkMigDevState *bmds,
                                 int is_async)
{
    BlkMigBlock *blk;
    int64_t total_sectors = bmds->total_sectors;
    int64_t sector;
    int nr_sectors;
    int ret = -EIO;

    for (sector = bmds->cur_dirty; sector < bmds->total_sectors;) {
        blk_mig_lock();
        if (bmds_aio_inflight(bmds, sector)) {
            blk_mig_unlock();
            bdrv_drain_all();
        } else {
            blk_mig_unlock();
        }
        if (bdrv_get_dirty(bmds->bs, bmds->dirty_bitmap, sector)) {

            if (total_sectors - sector < BDRV_SECTORS_PER_DIRTY_CHUNK) {
                nr_sectors = total_sectors - sector;
            } else {
                nr_sectors = BDRV_SECTORS_PER_DIRTY_CHUNK;
            }
            blk = g_malloc(sizeof(BlkMigBlock));
            blk->buf = g_malloc(BLOCK_SIZE);
            blk->bmds = bmds;
            blk->sector = sector;
            blk->nr_sectors = nr_sectors;

            if (is_async) {
                blk->iov.iov_base = blk->buf;
                blk->iov.iov_len = nr_sectors * BDRV_SECTOR_SIZE;
                qemu_iovec_init_external(&blk->qiov, &blk->iov, 1);

                blk->aiocb = bdrv_aio_readv(bmds->bs, sector, &blk->qiov,
                                            nr_sectors, blk_mig_read_cb, blk);

                blk_mig_lock();
                block_mig_state.submitted++;
                bmds_set_aio_inflight(bmds, sector, nr_sectors, 1);
                blk_mig_unlock();
            } else {
                ret = bdrv_read(bmds->bs, sector, blk->buf, nr_sectors);
                if (ret < 0) {
                    goto error;
                }
                blk_send(f, blk);

                g_free(blk->buf);
                g_free(blk);
            }

            bdrv_reset_dirty(bmds->bs, sector, nr_sectors);
            break;
        }
        sector += BDRV_SECTORS_PER_DIRTY_CHUNK;
        bmds->cur_dirty = sector;
    }

    return (bmds->cur_dirty >= bmds->total_sectors);

error:
    DPRINTF("Error reading sector %" PRId64 "\n", sector);
    g_free(blk->buf);
    g_free(blk);
    return ret;
}

/* Called with iothread lock taken.
*
* return value:
* 0: too much data for max_downtime
* 1: few enough data for max_downtime
*/
static int blk_mig_save_dirty_block(QEMUFile *f, int is_async)
{
    BlkMigDevState *bmds;
    int ret = 1;

    QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
        ret = mig_save_device_dirty(f, bmds, is_async);
        if (ret <= 0) {
            break;
        }
    }

    return ret;
}

/* Called with no locks taken. */

static int flush_blks(QEMUFile *f)
{
    BlkMigBlock *blk;
    int ret = 0;

    DPRINTF("%s Enter submitted %d read_done %d transferred %d\n",
            __FUNCTION__, block_mig_state.submitted, block_mig_state.read_done,
            block_mig_state.transferred);

    blk_mig_lock();
    while ((blk = QSIMPLEQ_FIRST(&block_mig_state.blk_list)) != NULL) {
        if (qemu_file_rate_limit(f)) {
            break;
        }
        if (blk->ret < 0) {
            ret = blk->ret;
            break;
        }

        QSIMPLEQ_REMOVE_HEAD(&block_mig_state.blk_list, entry);
        blk_mig_unlock();
        blk_send(f, blk);
        blk_mig_lock();

        g_free(blk->buf);
        g_free(blk);

        block_mig_state.read_done--;
        block_mig_state.transferred++;
        assert(block_mig_state.read_done >= 0);
    }
    blk_mig_unlock();

    DPRINTF("%s Exit submitted %d read_done %d transferred %d\n", __FUNCTION__,
            block_mig_state.submitted, block_mig_state.read_done,
            block_mig_state.transferred);
    return ret;
}

/* Called with iothread lock taken. */

static int64_t get_remaining_dirty(void)
{
    BlkMigDevState *bmds;
    int64_t dirty = 0;

    QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
        dirty += bdrv_get_dirty_count(bmds->bs, bmds->dirty_bitmap);
    }

    return dirty << BDRV_SECTOR_BITS;
}

/* Called with iothread lock taken. */

static void blk_mig_cleanup(void)
{
    BlkMigDevState *bmds;
    BlkMigBlock *blk;

    bdrv_drain_all();

    unset_dirty_tracking();

    blk_mig_lock();
    while ((bmds = QSIMPLEQ_FIRST(&block_mig_state.bmds_list)) != NULL) {
        QSIMPLEQ_REMOVE_HEAD(&block_mig_state.bmds_list, entry);
        bdrv_set_in_use(bmds->bs, 0);
        bdrv_unref(bmds->bs);
        g_free(bmds->aio_bitmap);
        g_free(bmds);
    }

    while ((blk = QSIMPLEQ_FIRST(&block_mig_state.blk_list)) != NULL) {
        QSIMPLEQ_REMOVE_HEAD(&block_mig_state.blk_list, entry);
        g_free(blk->buf);
        g_free(blk);
    }
    blk_mig_unlock();
}

static void block_migration_cancel(void *opaque)
{
    blk_mig_cleanup();
}

static int block_save_setup(QEMUFile *f, void *opaque)
{
    int ret;

    DPRINTF("Enter save live setup submitted %d transferred %d\n",
            block_mig_state.submitted, block_mig_state.transferred);

    qemu_mutex_lock_iothread();
    init_blk_migration(f);

    /* start track dirty blocks */
    set_dirty_tracking();
    qemu_mutex_unlock_iothread();

    ret = flush_blks(f);
    blk_mig_reset_dirty_cursor();
    qemu_put_be64(f, BLK_MIG_FLAG_EOS);

    return ret;
}

static int block_save_iterate(QEMUFile *f, void *opaque)
{
    int ret;
    int64_t last_ftell = qemu_ftell(f);

    DPRINTF("Enter save live iterate submitted %d transferred %d\n",
            block_mig_state.submitted, block_mig_state.transferred);

    ret = flush_blks(f);
    if (ret) {
        return ret;
    }

    blk_mig_reset_dirty_cursor();

    /* control the rate of transfer */
    blk_mig_lock();
    while ((block_mig_state.submitted +
            block_mig_state.read_done) * BLOCK_SIZE <
           qemu_file_get_rate_limit(f)) {
        blk_mig_unlock();
        if (block_mig_state.bulk_completed == 0) {
            /* first finish the bulk phase */
            if (blk_mig_save_bulked_block(f) == 0) {
                /* finished saving bulk on all devices */
                block_mig_state.bulk_completed = 1;
            }
            ret = 0;
        } else {
            /* Always called with iothread lock taken for
* simplicity, block_save_complete also calls it.
*/
            qemu_mutex_lock_iothread();
            ret = blk_mig_save_dirty_block(f, 1);
            qemu_mutex_unlock_iothread();
        }
        if (ret < 0) {
            return ret;
        }
        blk_mig_lock();
        if (ret != 0) {
            /* no more dirty blocks */
            break;
        }
    }
    blk_mig_unlock();

    ret = flush_blks(f);
    if (ret) {
        return ret;
    }

    qemu_put_be64(f, BLK_MIG_FLAG_EOS);
    return qemu_ftell(f) - last_ftell;
}

/* Called with iothread lock taken. */

static int block_save_complete(QEMUFile *f, void *opaque)
{
    int ret;

    DPRINTF("Enter save live complete submitted %d transferred %d\n",
            block_mig_state.submitted, block_mig_state.transferred);

    ret = flush_blks(f);
    if (ret) {
        return ret;
    }

    blk_mig_reset_dirty_cursor();

    /* we know for sure that save bulk is completed and
all async read completed */
    blk_mig_lock();
    assert(block_mig_state.submitted == 0);
    blk_mig_unlock();

    do {
        ret = blk_mig_save_dirty_block(f, 0);
        if (ret < 0) {
            return ret;
        }
    } while (ret == 0);

    /* report completion */
    qemu_put_be64(f, (100 << BDRV_SECTOR_BITS) | BLK_MIG_FLAG_PROGRESS);

    DPRINTF("Block migration completed\n");

    qemu_put_be64(f, BLK_MIG_FLAG_EOS);

    blk_mig_cleanup();
    return 0;
}

static uint64_t block_save_pending(QEMUFile *f, void *opaque, uint64_t max_size)
{
    /* Estimate pending number of bytes to send */
    uint64_t pending;

    qemu_mutex_lock_iothread();
    blk_mig_lock();
    pending = get_remaining_dirty() +
                       block_mig_state.submitted * BLOCK_SIZE +
                       block_mig_state.read_done * BLOCK_SIZE;

    /* Report at least one block pending during bulk phase */
    if (pending == 0 && !block_mig_state.bulk_completed) {
        pending = BLOCK_SIZE;
    }
    blk_mig_unlock();
    qemu_mutex_unlock_iothread();

    DPRINTF("Enter save live pending %" PRIu64 "\n", pending);
    return pending;
}

static int block_load(QEMUFile *f, void *opaque, int version_id)
{
    static int banner_printed;
    int len, flags;
    char device_name[256];
    int64_t addr;
    BlockDriverState *bs, *bs_prev = NULL;
    uint8_t *buf;
    int64_t total_sectors = 0;
    int nr_sectors;
    int ret;

    do {
        addr = qemu_get_be64(f);

        flags = addr & ~BDRV_SECTOR_MASK;
        addr >>= BDRV_SECTOR_BITS;

        if (flags & BLK_MIG_FLAG_DEVICE_BLOCK) {
            /* get device name */
            len = qemu_get_byte(f);
            qemu_get_buffer(f, (uint8_t *)device_name, len);
            device_name[len] = '\0';

            bs = bdrv_find(device_name);
            if (!bs) {
                fprintf(stderr, "Error unknown block device %s\n",
                        device_name);
                return -EINVAL;
            }

            if (bs != bs_prev) {
                bs_prev = bs;
                total_sectors = bdrv_getlength(bs) >> BDRV_SECTOR_BITS;
                if (total_sectors <= 0) {
                    error_report("Error getting length of block device %s",
                                 device_name);
                    return -EINVAL;
                }
            }

            if (total_sectors - addr < BDRV_SECTORS_PER_DIRTY_CHUNK) {
                nr_sectors = total_sectors - addr;
            } else {
                nr_sectors = BDRV_SECTORS_PER_DIRTY_CHUNK;
            }

            if (flags & BLK_MIG_FLAG_ZERO_BLOCK) {
                ret = bdrv_write_zeroes(bs, addr, nr_sectors,
                                        BDRV_REQ_MAY_UNMAP);
            } else {
                buf = g_malloc(BLOCK_SIZE);
                qemu_get_buffer(f, buf, BLOCK_SIZE);
                ret = bdrv_write(bs, addr, buf, nr_sectors);
                g_free(buf);
            }

            if (ret < 0) {
                return ret;
            }
        } else if (flags & BLK_MIG_FLAG_PROGRESS) {
            if (!banner_printed) {
                printf("Receiving block device images\n");
                banner_printed = 1;
            }
            printf("Completed %d %%%c", (int)addr,
                   (addr == 100) ? '\n' : '\r');
            fflush(stdout);
        } else if (!(flags & BLK_MIG_FLAG_EOS)) {
            fprintf(stderr, "Unknown block migration flags: %#x\n", flags);
            return -EINVAL;
        }
        ret = qemu_file_get_error(f);
        if (ret != 0) {
            return ret;
        }
    } while (!(flags & BLK_MIG_FLAG_EOS));

    return 0;
}

static void block_set_params(const MigrationParams *params, void *opaque)
{
    block_mig_state.blk_enable = params->blk;
    block_mig_state.shared_base = params->shared;

    /* shared base means that blk_enable = 1 */
    block_mig_state.blk_enable |= params->shared;
}

static bool block_is_active(void *opaque)
{
    return block_mig_state.blk_enable == 1;
}

SaveVMHandlers savevm_block_handlers = {
    .set_params = block_set_params,
    .save_live_setup = block_save_setup,
    .save_live_iterate = block_save_iterate,
    .save_live_complete = block_save_complete,
    .save_live_pending = block_save_pending,
    .load_state = block_load,
    .cancel = block_migration_cancel,
    .is_active = block_is_active,
};

void blk_mig_init(void)
{
    QSIMPLEQ_INIT(&block_mig_state.bmds_list);
    QSIMPLEQ_INIT(&block_mig_state.blk_list);
    qemu_mutex_init(&block_mig_state.lock);

    register_savevm_live(NULL, "block", 0, 1, &savevm_block_handlers,
                         &block_mig_state);
}
Something went wrong with that request. Please try again.