block/rbd.c

   1 /*
   2  * QEMU Block driver for RADOS (Ceph)
   3  *
   4  * Copyright (C) 2010-2011 Christian Brunner <chb@muc.de>,
   5  *                         Josh Durgin <josh.durgin@dreamhost.com>
   6  *
   7  * This work is licensed under the terms of the GNU GPL, version 2.  See
   8  * the COPYING file in the top-level directory.
   9  *
  10  */
  11
  12 #include <inttypes.h>
  13
  14 #include "qemu-common.h"
  15 #include "qemu-error.h"
  16
  17 #include "block_int.h"
  18
  19 #include <rbd/librbd.h>
  20
  21
  22
  23 /*
  24  * When specifying the image filename use:
  25  *
  26  * rbd:poolname/devicename[@snapshotname][:option1=value1[:option2=value2...]]
  27  *
  28  * poolname must be the name of an existing rados pool
  29  *
  30  * devicename is the basename for all objects used to
  31  * emulate the raw device.
  32  *
  33  * Each option given is used to configure rados, and may be
  34  * any Ceph option, or "conf". The "conf" option specifies
  35  * a Ceph configuration file to read.
  36  *
  37  * Metadata information (image size, ...) is stored in an
  38  * object with the name "devicename.rbd".
  39  *
  40  * The raw device is split into 4MB sized objects by default.
  41  * The sequencenumber is encoded in a 12 byte long hex-string,
  42  * and is attached to the devicename, separated by a dot.
  43  * e.g. "devicename.1234567890ab"
  44  *
  45  */
  46
  47 #define OBJ_MAX_SIZE (1UL << OBJ_DEFAULT_OBJ_ORDER)
  48
  49 #define RBD_MAX_CONF_NAME_SIZE 128
  50 #define RBD_MAX_CONF_VAL_SIZE 512
  51 #define RBD_MAX_CONF_SIZE 1024
  52 #define RBD_MAX_POOL_NAME_SIZE 128
  53 #define RBD_MAX_SNAP_NAME_SIZE 128
  54 #define RBD_MAX_SNAPS 100
  55
  56 typedef struct RBDAIOCB {
  57     BlockDriverAIOCB common;
  58     QEMUBH *bh;
  59     int ret;
  60     QEMUIOVector *qiov;
  61     char *bounce;
  62     int write;
  63     int64_t sector_num;
  64     int error;
  65     struct BDRVRBDState *s;
  66     int cancelled;
  67 } RBDAIOCB;
  68
  69 typedef struct RADOSCB {
  70     int rcbid;
  71     RBDAIOCB *acb;
  72     struct BDRVRBDState *s;
  73     int done;
  74     int64_t size;
  75     char *buf;
  76     int ret;
  77 } RADOSCB;
  78
  79 #define RBD_FD_READ 0
  80 #define RBD_FD_WRITE 1
  81
  82 typedef struct BDRVRBDState {
  83     int fds[2];
  84     rados_t cluster;
  85     rados_ioctx_t io_ctx;
  86     rbd_image_t image;
  87     char name[RBD_MAX_IMAGE_NAME_SIZE];
  88     int qemu_aio_count;
  89     char *snap;
  90     int event_reader_pos;
  91     RADOSCB *event_rcb;
  92 } BDRVRBDState;
  93
  94 static void rbd_aio_bh_cb(void *opaque);
  95
  96 static int qemu_rbd_next_tok(char *dst, int dst_len,
  97                              char *src, char delim,
  98                              const char *name,
  99                              char **p)
 100 {
 101     int l;
 102     char *end;
 103
 104     *p = NULL;
 105
 106     if (delim != '\0') {
 107         end = strchr(src, delim);
 108         if (end) {
 109             *p = end + 1;
 110             *end = '\0';
 111         }
 112     }
 113     l = strlen(src);
 114     if (l >= dst_len) {
 115         error_report("%s too long", name);
 116         return -EINVAL;
 117     } else if (l == 0) {
 118         error_report("%s too short", name);
 119         return -EINVAL;
 120     }
 121
 122     pstrcpy(dst, dst_len, src);
 123
 124     return 0;
 125 }
 126
 127 static int qemu_rbd_parsename(const char *filename,
 128                               char *pool, int pool_len,
 129                               char *snap, int snap_len,
 130                               char *name, int name_len,
 131                               char *conf, int conf_len)
 132 {
 133     const char *start;
 134     char *p, *buf;
 135     int ret;
 136
 137     if (!strstart(filename, "rbd:", &start)) {
 138         return -EINVAL;
 139     }
 140
 141     buf = qemu_strdup(start);
 142     p = buf;
 143     *snap = '\0';
 144     *conf = '\0';
 145
 146     ret = qemu_rbd_next_tok(pool, pool_len, p, '/', "pool name", &p);
 147     if (ret < 0 || !p) {
 148         ret = -EINVAL;
 149         goto done;
 150     }
 151
 152     if (strchr(p, '@')) {
 153         ret = qemu_rbd_next_tok(name, name_len, p, '@', "object name", &p);
 154         if (ret < 0) {
 155             goto done;
 156         }
 157         ret = qemu_rbd_next_tok(snap, snap_len, p, ':', "snap name", &p);
 158     } else {
 159         ret = qemu_rbd_next_tok(name, name_len, p, ':', "object name", &p);
 160     }
 161     if (ret < 0 || !p) {
 162         goto done;
 163     }
 164
 165     ret = qemu_rbd_next_tok(conf, conf_len, p, '\0', "configuration", &p);
 166
 167 done:
 168     qemu_free(buf);
 169     return ret;
 170 }
 171
 172 static int qemu_rbd_set_conf(rados_t cluster, const char *conf)
 173 {
 174     char *p, *buf;
 175     char name[RBD_MAX_CONF_NAME_SIZE];
 176     char value[RBD_MAX_CONF_VAL_SIZE];
 177     int ret = 0;
 178
 179     buf = qemu_strdup(conf);
 180     p = buf;
 181
 182     while (p) {
 183         ret = qemu_rbd_next_tok(name, sizeof(name), p,
 184                                 '=', "conf option name", &p);
 185         if (ret < 0) {
 186             break;
 187         }
 188
 189         if (!p) {
 190             error_report("conf option %s has no value", name);
 191             ret = -EINVAL;
 192             break;
 193         }
 194
 195         ret = qemu_rbd_next_tok(value, sizeof(value), p,
 196                                 ':', "conf option value", &p);
 197         if (ret < 0) {
 198             break;
 199         }
 200
 201         if (strcmp(name, "conf")) {
 202             ret = rados_conf_set(cluster, name, value);
 203             if (ret < 0) {
 204                 error_report("invalid conf option %s", name);
 205                 ret = -EINVAL;
 206                 break;
 207             }
 208         } else {
 209             ret = rados_conf_read_file(cluster, value);
 210             if (ret < 0) {
 211                 error_report("error reading conf file %s", value);
 212                 break;
 213             }
 214         }
 215     }
 216
 217     qemu_free(buf);
 218     return ret;
 219 }
 220
 221 static int qemu_rbd_create(const char *filename, QEMUOptionParameter *options)
 222 {
 223     int64_t bytes = 0;
 224     int64_t objsize;
 225     int obj_order = 0;
 226     char pool[RBD_MAX_POOL_NAME_SIZE];
 227     char name[RBD_MAX_IMAGE_NAME_SIZE];
 228     char snap_buf[RBD_MAX_SNAP_NAME_SIZE];
 229     char conf[RBD_MAX_CONF_SIZE];
 230     rados_t cluster;
 231     rados_ioctx_t io_ctx;
 232     int ret;
 233
 234     if (qemu_rbd_parsename(filename, pool, sizeof(pool),
 235                            snap_buf, sizeof(snap_buf),
 236                            name, sizeof(name),
 237                            conf, sizeof(conf)) < 0) {
 238         return -EINVAL;
 239     }
 240
 241     /* Read out options */
 242     while (options && options->name) {
 243         if (!strcmp(options->name, BLOCK_OPT_SIZE)) {
 244             bytes = options->value.n;
 245         } else if (!strcmp(options->name, BLOCK_OPT_CLUSTER_SIZE)) {
 246             if (options->value.n) {
 247                 objsize = options->value.n;
 248                 if ((objsize - 1) & objsize) {    /* not a power of 2? */
 249                     error_report("obj size needs to be power of 2");
 250                     return -EINVAL;
 251                 }
 252                 if (objsize < 4096) {
 253                     error_report("obj size too small");
 254                     return -EINVAL;
 255                 }
 256                 obj_order = ffs(objsize) - 1;
 257             }
 258         }
 259         options++;
 260     }
 261
 262     if (rados_create(&cluster, NULL) < 0) {
 263         error_report("error initializing");
 264         return -EIO;
 265     }
 266
 267     if (strstr(conf, "conf=") == NULL) {
 268         if (rados_conf_read_file(cluster, NULL) < 0) {
 269             error_report("error reading config file");
 270             rados_shutdown(cluster);
 271             return -EIO;
 272         }
 273     }
 274
 275     if (conf[0] != '\0' &&
 276         qemu_rbd_set_conf(cluster, conf) < 0) {
 277         error_report("error setting config options");
 278         rados_shutdown(cluster);
 279         return -EIO;
 280     }
 281
 282     if (rados_connect(cluster) < 0) {
 283         error_report("error connecting");
 284         rados_shutdown(cluster);
 285         return -EIO;
 286     }
 287
 288     if (rados_ioctx_create(cluster, pool, &io_ctx) < 0) {
 289         error_report("error opening pool %s", pool);
 290         rados_shutdown(cluster);
 291         return -EIO;
 292     }
 293
 294     ret = rbd_create(io_ctx, name, bytes, &obj_order);
 295     rados_ioctx_destroy(io_ctx);
 296     rados_shutdown(cluster);
 297
 298     return ret;
 299 }
 300
 301 /*
 302  * This aio completion is being called from qemu_rbd_aio_event_reader()
 303  * and runs in qemu context. It schedules a bh, but just in case the aio
 304  * was not cancelled before.
 305  */
 306 static void qemu_rbd_complete_aio(RADOSCB *rcb)
 307 {
 308     RBDAIOCB *acb = rcb->acb;
 309     int64_t r;
 310
 311     if (acb->cancelled) {
 312         qemu_vfree(acb->bounce);
 313         qemu_aio_release(acb);
 314         goto done;
 315     }
 316
 317     r = rcb->ret;
 318
 319     if (acb->write) {
 320         if (r < 0) {
 321             acb->ret = r;
 322             acb->error = 1;
 323         } else if (!acb->error) {
 324             acb->ret = rcb->size;
 325         }
 326     } else {
 327         if (r < 0) {
 328             memset(rcb->buf, 0, rcb->size);
 329             acb->ret = r;
 330             acb->error = 1;
 331         } else if (r < rcb->size) {
 332             memset(rcb->buf + r, 0, rcb->size - r);
 333             if (!acb->error) {
 334                 acb->ret = rcb->size;
 335             }
 336         } else if (!acb->error) {
 337             acb->ret = r;
 338         }
 339     }
 340     /* Note that acb->bh can be NULL in case where the aio was cancelled */
 341     acb->bh = qemu_bh_new(rbd_aio_bh_cb, acb);
 342     qemu_bh_schedule(acb->bh);
 343 done:
 344     qemu_free(rcb);
 345 }
 346
 347 /*
 348  * aio fd read handler. It runs in the qemu context and calls the
 349  * completion handling of completed rados aio operations.
 350  */
 351 static void qemu_rbd_aio_event_reader(void *opaque)
 352 {
 353     BDRVRBDState *s = opaque;
 354
 355     ssize_t ret;
 356
 357     do {
 358         char *p = (char *)&s->event_rcb;
 359
 360         /* now read the rcb pointer that was sent from a non qemu thread */
 361         if ((ret = read(s->fds[RBD_FD_READ], p + s->event_reader_pos,
 362                         sizeof(s->event_rcb) - s->event_reader_pos)) > 0) {
 363             if (ret > 0) {
 364                 s->event_reader_pos += ret;
 365                 if (s->event_reader_pos == sizeof(s->event_rcb)) {
 366                     s->event_reader_pos = 0;
 367                     qemu_rbd_complete_aio(s->event_rcb);
 368                     s->qemu_aio_count--;
 369                 }
 370             }
 371         }
 372     } while (ret < 0 && errno == EINTR);
 373 }
 374
 375 static int qemu_rbd_aio_flush_cb(void *opaque)
 376 {
 377     BDRVRBDState *s = opaque;
 378
 379     return (s->qemu_aio_count > 0);
 380 }
 381
 382 static int qemu_rbd_open(BlockDriverState *bs, const char *filename, int flags)
 383 {
 384     BDRVRBDState *s = bs->opaque;
 385     char pool[RBD_MAX_POOL_NAME_SIZE];
 386     char snap_buf[RBD_MAX_SNAP_NAME_SIZE];
 387     char conf[RBD_MAX_CONF_SIZE];
 388     int r;
 389
 390     if (qemu_rbd_parsename(filename, pool, sizeof(pool),
 391                            snap_buf, sizeof(snap_buf),
 392                            s->name, sizeof(s->name),
 393                            conf, sizeof(conf)) < 0) {
 394         return -EINVAL;
 395     }
 396     s->snap = NULL;
 397     if (snap_buf[0] != '\0') {
 398         s->snap = qemu_strdup(snap_buf);
 399     }
 400
 401     r = rados_create(&s->cluster, NULL);
 402     if (r < 0) {
 403         error_report("error initializing");
 404         return r;
 405     }
 406
 407     if (strstr(conf, "conf=") == NULL) {
 408         r = rados_conf_read_file(s->cluster, NULL);
 409         if (r < 0) {
 410             error_report("error reading config file");
 411             rados_shutdown(s->cluster);
 412             return r;
 413         }
 414     }
 415
 416     if (conf[0] != '\0') {
 417         r = qemu_rbd_set_conf(s->cluster, conf);
 418         if (r < 0) {
 419             error_report("error setting config options");
 420             rados_shutdown(s->cluster);
 421             return r;
 422         }
 423     }
 424
 425     r = rados_connect(s->cluster);
 426     if (r < 0) {
 427         error_report("error connecting");
 428         rados_shutdown(s->cluster);
 429         return r;
 430     }
 431
 432     r = rados_ioctx_create(s->cluster, pool, &s->io_ctx);
 433     if (r < 0) {
 434         error_report("error opening pool %s", pool);
 435         rados_shutdown(s->cluster);
 436         return r;
 437     }
 438
 439     r = rbd_open(s->io_ctx, s->name, &s->image, s->snap);
 440     if (r < 0) {
 441         error_report("error reading header from %s", s->name);
 442         rados_ioctx_destroy(s->io_ctx);
 443         rados_shutdown(s->cluster);
 444         return r;
 445     }
 446
 447     bs->read_only = (s->snap != NULL);
 448
 449     s->event_reader_pos = 0;
 450     r = qemu_pipe(s->fds);
 451     if (r < 0) {
 452         error_report("error opening eventfd");
 453         goto failed;
 454     }
 455     fcntl(s->fds[0], F_SETFL, O_NONBLOCK);
 456     fcntl(s->fds[1], F_SETFL, O_NONBLOCK);
 457     qemu_aio_set_fd_handler(s->fds[RBD_FD_READ], qemu_rbd_aio_event_reader,
 458                             NULL, qemu_rbd_aio_flush_cb, NULL, s);
 459
 460
 461     return 0;
 462
 463 failed:
 464     rbd_close(s->image);
 465     rados_ioctx_destroy(s->io_ctx);
 466     rados_shutdown(s->cluster);
 467     return r;
 468 }
 469
 470 static void qemu_rbd_close(BlockDriverState *bs)
 471 {
 472     BDRVRBDState *s = bs->opaque;
 473
 474     close(s->fds[0]);
 475     close(s->fds[1]);
 476     qemu_aio_set_fd_handler(s->fds[RBD_FD_READ], NULL , NULL, NULL, NULL,
 477         NULL);
 478
 479     rbd_close(s->image);
 480     rados_ioctx_destroy(s->io_ctx);
 481     qemu_free(s->snap);
 482     rados_shutdown(s->cluster);
 483 }
 484
 485 /*
 486  * Cancel aio. Since we don't reference acb in a non qemu threads,
 487  * it is safe to access it here.
 488  */
 489 static void qemu_rbd_aio_cancel(BlockDriverAIOCB *blockacb)
 490 {
 491     RBDAIOCB *acb = (RBDAIOCB *) blockacb;
 492     acb->cancelled = 1;
 493 }
 494
 495 static AIOPool rbd_aio_pool = {
 496     .aiocb_size = sizeof(RBDAIOCB),
 497     .cancel = qemu_rbd_aio_cancel,
 498 };
 499
 500 static int qemu_rbd_send_pipe(BDRVRBDState *s, RADOSCB *rcb)
 501 {
 502     int ret = 0;
 503     while (1) {
 504         fd_set wfd;
 505         int fd = s->fds[RBD_FD_WRITE];
 506
 507         /* send the op pointer to the qemu thread that is responsible
 508            for the aio/op completion. Must do it in a qemu thread context */
 509         ret = write(fd, (void *)&rcb, sizeof(rcb));
 510         if (ret >= 0) {
 511             break;
 512         }
 513         if (errno == EINTR) {
 514             continue;
 515         }
 516         if (errno != EAGAIN) {
 517             break;
 518         }
 519
 520         FD_ZERO(&wfd);
 521         FD_SET(fd, &wfd);
 522         do {
 523             ret = select(fd + 1, NULL, &wfd, NULL, NULL);
 524         } while (ret < 0 && errno == EINTR);
 525     }
 526
 527     return ret;
 528 }
 529
 530 /*
 531  * This is the callback function for rbd_aio_read and _write
 532  *
 533  * Note: this function is being called from a non qemu thread so
 534  * we need to be careful about what we do here. Generally we only
 535  * write to the block notification pipe, and do the rest of the
 536  * io completion handling from qemu_rbd_aio_event_reader() which
 537  * runs in a qemu context.
 538  */
 539 static void rbd_finish_aiocb(rbd_completion_t c, RADOSCB *rcb)
 540 {
 541     int ret;
 542     rcb->ret = rbd_aio_get_return_value(c);
 543     rbd_aio_release(c);
 544     ret = qemu_rbd_send_pipe(rcb->s, rcb);
 545     if (ret < 0) {
 546         error_report("failed writing to acb->s->fds");
 547         qemu_free(rcb);
 548     }
 549 }
 550
 551 /* Callback when all queued rbd_aio requests are complete */
 552
 553 static void rbd_aio_bh_cb(void *opaque)
 554 {
 555     RBDAIOCB *acb = opaque;
 556
 557     if (!acb->write) {
 558         qemu_iovec_from_buffer(acb->qiov, acb->bounce, acb->qiov->size);
 559     }
 560     qemu_vfree(acb->bounce);
 561     acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret));
 562     qemu_bh_delete(acb->bh);
 563     acb->bh = NULL;
 564
 565     qemu_aio_release(acb);
 566 }
 567
 568 static BlockDriverAIOCB *rbd_aio_rw_vector(BlockDriverState *bs,
 569                                            int64_t sector_num,
 570                                            QEMUIOVector *qiov,
 571                                            int nb_sectors,
 572                                            BlockDriverCompletionFunc *cb,
 573                                            void *opaque, int write)
 574 {
 575     RBDAIOCB *acb;
 576     RADOSCB *rcb;
 577     rbd_completion_t c;
 578     int64_t off, size;
 579     char *buf;
 580     int r;
 581
 582     BDRVRBDState *s = bs->opaque;
 583
 584     acb = qemu_aio_get(&rbd_aio_pool, bs, cb, opaque);
 585     if (!acb) {
 586         return NULL;
 587     }
 588     acb->write = write;
 589     acb->qiov = qiov;
 590     acb->bounce = qemu_blockalign(bs, qiov->size);
 591     acb->ret = 0;
 592     acb->error = 0;
 593     acb->s = s;
 594     acb->cancelled = 0;
 595     acb->bh = NULL;
 596
 597     if (write) {
 598         qemu_iovec_to_buffer(acb->qiov, acb->bounce);
 599     }
 600
 601     buf = acb->bounce;
 602
 603     off = sector_num * BDRV_SECTOR_SIZE;
 604     size = nb_sectors * BDRV_SECTOR_SIZE;
 605
 606     s->qemu_aio_count++; /* All the RADOSCB */
 607
 608     rcb = qemu_malloc(sizeof(RADOSCB));
 609     rcb->done = 0;
 610     rcb->acb = acb;
 611     rcb->buf = buf;
 612     rcb->s = acb->s;
 613     rcb->size = size;
 614     r = rbd_aio_create_completion(rcb, (rbd_callback_t) rbd_finish_aiocb, &c);
 615     if (r < 0) {
 616         goto failed;
 617     }
 618
 619     if (write) {
 620         r = rbd_aio_write(s->image, off, size, buf, c);
 621     } else {
 622         r = rbd_aio_read(s->image, off, size, buf, c);
 623     }
 624
 625     if (r < 0) {
 626         goto failed;
 627     }
 628
 629     return &acb->common;
 630
 631 failed:
 632     qemu_free(rcb);
 633     s->qemu_aio_count--;
 634     qemu_aio_release(acb);
 635     return NULL;
 636 }
 637
 638 static BlockDriverAIOCB *qemu_rbd_aio_readv(BlockDriverState *bs,
 639                                             int64_t sector_num,
 640                                             QEMUIOVector *qiov,
 641                                             int nb_sectors,
 642                                             BlockDriverCompletionFunc *cb,
 643                                             void *opaque)
 644 {
 645     return rbd_aio_rw_vector(bs, sector_num, qiov, nb_sectors, cb, opaque, 0);
 646 }
 647
 648 static BlockDriverAIOCB *qemu_rbd_aio_writev(BlockDriverState *bs,
 649                                              int64_t sector_num,
 650                                              QEMUIOVector *qiov,
 651                                              int nb_sectors,
 652                                              BlockDriverCompletionFunc *cb,
 653                                              void *opaque)
 654 {
 655     return rbd_aio_rw_vector(bs, sector_num, qiov, nb_sectors, cb, opaque, 1);
 656 }
 657
 658 static int qemu_rbd_getinfo(BlockDriverState *bs, BlockDriverInfo *bdi)
 659 {
 660     BDRVRBDState *s = bs->opaque;
 661     rbd_image_info_t info;
 662     int r;
 663
 664     r = rbd_stat(s->image, &info, sizeof(info));
 665     if (r < 0) {
 666         return r;
 667     }
 668
 669     bdi->cluster_size = info.obj_size;
 670     return 0;
 671 }
 672
 673 static int64_t qemu_rbd_getlength(BlockDriverState *bs)
 674 {
 675     BDRVRBDState *s = bs->opaque;
 676     rbd_image_info_t info;
 677     int r;
 678
 679     r = rbd_stat(s->image, &info, sizeof(info));
 680     if (r < 0) {
 681         return r;
 682     }
 683
 684     return info.size;
 685 }
 686
 687 static int qemu_rbd_truncate(BlockDriverState *bs, int64_t offset)
 688 {
 689     BDRVRBDState *s = bs->opaque;
 690     int r;
 691
 692     r = rbd_resize(s->image, offset);
 693     if (r < 0) {
 694         return r;
 695     }
 696
 697     return 0;
 698 }
 699
 700 static int qemu_rbd_snap_create(BlockDriverState *bs,
 701                                 QEMUSnapshotInfo *sn_info)
 702 {
 703     BDRVRBDState *s = bs->opaque;
 704     int r;
 705
 706     if (sn_info->name[0] == '\0') {
 707         return -EINVAL; /* we need a name for rbd snapshots */
 708     }
 709
 710     /*
 711      * rbd snapshots are using the name as the user controlled unique identifier
 712      * we can't use the rbd snapid for that purpose, as it can't be set
 713      */
 714     if (sn_info->id_str[0] != '\0' &&
 715         strcmp(sn_info->id_str, sn_info->name) != 0) {
 716         return -EINVAL;
 717     }
 718
 719     if (strlen(sn_info->name) >= sizeof(sn_info->id_str)) {
 720         return -ERANGE;
 721     }
 722
 723     r = rbd_snap_create(s->image, sn_info->name);
 724     if (r < 0) {
 725         error_report("failed to create snap: %s", strerror(-r));
 726         return r;
 727     }
 728
 729     return 0;
 730 }
 731
 732 static int qemu_rbd_snap_list(BlockDriverState *bs,
 733                               QEMUSnapshotInfo **psn_tab)
 734 {
 735     BDRVRBDState *s = bs->opaque;
 736     QEMUSnapshotInfo *sn_info, *sn_tab = NULL;
 737     int i, snap_count;
 738     rbd_snap_info_t *snaps;
 739     int max_snaps = RBD_MAX_SNAPS;
 740
 741     do {
 742         snaps = qemu_malloc(sizeof(*snaps) * max_snaps);
 743         snap_count = rbd_snap_list(s->image, snaps, &max_snaps);
 744         if (snap_count < 0) {
 745             qemu_free(snaps);
 746         }
 747     } while (snap_count == -ERANGE);
 748
 749     if (snap_count <= 0) {
 750         return snap_count;
 751     }
 752
 753     sn_tab = qemu_mallocz(snap_count * sizeof(QEMUSnapshotInfo));
 754
 755     for (i = 0; i < snap_count; i++) {
 756         const char *snap_name = snaps[i].name;
 757
 758         sn_info = sn_tab + i;
 759         pstrcpy(sn_info->id_str, sizeof(sn_info->id_str), snap_name);
 760         pstrcpy(sn_info->name, sizeof(sn_info->name), snap_name);
 761
 762         sn_info->vm_state_size = snaps[i].size;
 763         sn_info->date_sec = 0;
 764         sn_info->date_nsec = 0;
 765         sn_info->vm_clock_nsec = 0;
 766     }
 767     rbd_snap_list_end(snaps);
 768
 769     *psn_tab = sn_tab;
 770     return snap_count;
 771 }
 772
 773 static QEMUOptionParameter qemu_rbd_create_options[] = {
 774     {
 775      .name = BLOCK_OPT_SIZE,
 776      .type = OPT_SIZE,
 777      .help = "Virtual disk size"
 778     },
 779     {
 780      .name = BLOCK_OPT_CLUSTER_SIZE,
 781      .type = OPT_SIZE,
 782      .help = "RBD object size"
 783     },
 784     {NULL}
 785 };
 786
 787 static BlockDriver bdrv_rbd = {
 788     .format_name        = "rbd",
 789     .instance_size      = sizeof(BDRVRBDState),
 790     .bdrv_file_open     = qemu_rbd_open,
 791     .bdrv_close         = qemu_rbd_close,
 792     .bdrv_create        = qemu_rbd_create,
 793     .bdrv_get_info      = qemu_rbd_getinfo,
 794     .create_options     = qemu_rbd_create_options,
 795     .bdrv_getlength     = qemu_rbd_getlength,
 796     .bdrv_truncate      = qemu_rbd_truncate,
 797     .protocol_name      = "rbd",
 798
 799     .bdrv_aio_readv     = qemu_rbd_aio_readv,
 800     .bdrv_aio_writev    = qemu_rbd_aio_writev,
 801
 802     .bdrv_snapshot_create = qemu_rbd_snap_create,
 803     .bdrv_snapshot_list = qemu_rbd_snap_list,
 804 };
 805
 806 static void bdrv_rbd_init(void)
 807 {
 808     bdrv_register(&bdrv_rbd);
 809 }
 810
 811 block_init(bdrv_rbd_init);