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  * Contributions after 2012-01-13 are licensed under the terms of the
  11  * GNU GPL, version 2 or (at your option) any later version.
  12  */
  13
  14 #include <inttypes.h>
  15
  16 #include "qemu-common.h"
  17 #include "qemu/error-report.h"
  18 #include "block/block_int.h"
  19
  20 #include <rbd/librbd.h>
  21
  22 /*
  23  * When specifying the image filename use:
  24  *
  25  * rbd:poolname/devicename[@snapshotname][:option1=value1[:option2=value2...]]
  26  *
  27  * poolname must be the name of an existing rados pool.
  28  *
  29  * devicename is the name of the rbd image.
  30  *
  31  * Each option given is used to configure rados, and may be any valid
  32  * Ceph option, "id", or "conf".
  33  *
  34  * The "id" option indicates what user we should authenticate as to
  35  * the Ceph cluster.  If it is excluded we will use the Ceph default
  36  * (normally 'admin').
  37  *
  38  * The "conf" option specifies a Ceph configuration file to read.  If
  39  * it is not specified, we will read from the default Ceph locations
  40  * (e.g., /etc/ceph/ceph.conf).  To avoid reading _any_ configuration
  41  * file, specify conf=/dev/null.
  42  *
  43  * Configuration values containing :, @, or = can be escaped with a
  44  * leading "\".
  45  */
  46
  47 /* rbd_aio_discard added in 0.1.2 */
  48 #if LIBRBD_VERSION_CODE >= LIBRBD_VERSION(0, 1, 2)
  49 #define LIBRBD_SUPPORTS_DISCARD
  50 #else
  51 #undef LIBRBD_SUPPORTS_DISCARD
  52 #endif
  53
  54 #define OBJ_MAX_SIZE (1UL << OBJ_DEFAULT_OBJ_ORDER)
  55
  56 #define RBD_MAX_CONF_NAME_SIZE 128
  57 #define RBD_MAX_CONF_VAL_SIZE 512
  58 #define RBD_MAX_CONF_SIZE 1024
  59 #define RBD_MAX_POOL_NAME_SIZE 128
  60 #define RBD_MAX_SNAP_NAME_SIZE 128
  61 #define RBD_MAX_SNAPS 100
  62
  63 typedef enum {
  64     RBD_AIO_READ,
  65     RBD_AIO_WRITE,
  66     RBD_AIO_DISCARD,
  67     RBD_AIO_FLUSH
  68 } RBDAIOCmd;
  69
  70 typedef struct RBDAIOCB {
  71     BlockDriverAIOCB common;
  72     QEMUBH *bh;
  73     int64_t ret;
  74     QEMUIOVector *qiov;
  75     char *bounce;
  76     RBDAIOCmd cmd;
  77     int64_t sector_num;
  78     int error;
  79     struct BDRVRBDState *s;
  80     int cancelled;
  81     int status;
  82 } RBDAIOCB;
  83
  84 typedef struct RADOSCB {
  85     int rcbid;
  86     RBDAIOCB *acb;
  87     struct BDRVRBDState *s;
  88     int done;
  89     int64_t size;
  90     char *buf;
  91     int64_t ret;
  92 } RADOSCB;
  93
  94 #define RBD_FD_READ 0
  95 #define RBD_FD_WRITE 1
  96
  97 typedef struct BDRVRBDState {
  98     int fds[2];
  99     rados_t cluster;
 100     rados_ioctx_t io_ctx;
 101     rbd_image_t image;
 102     char name[RBD_MAX_IMAGE_NAME_SIZE];
 103     char *snap;
 104     int event_reader_pos;
 105     RADOSCB *event_rcb;
 106 } BDRVRBDState;
 107
 108 static void rbd_aio_bh_cb(void *opaque);
 109
 110 static int qemu_rbd_next_tok(char *dst, int dst_len,
 111                              char *src, char delim,
 112                              const char *name,
 113                              char **p)
 114 {
 115     int l;
 116     char *end;
 117
 118     *p = NULL;
 119
 120     if (delim != '\0') {
 121         for (end = src; *end; ++end) {
 122             if (*end == delim) {
 123                 break;
 124             }
 125             if (*end == '\\' && end[1] != '\0') {
 126                 end++;
 127             }
 128         }
 129         if (*end == delim) {
 130             *p = end + 1;
 131             *end = '\0';
 132         }
 133     }
 134     l = strlen(src);
 135     if (l >= dst_len) {
 136         error_report("%s too long", name);
 137         return -EINVAL;
 138     } else if (l == 0) {
 139         error_report("%s too short", name);
 140         return -EINVAL;
 141     }
 142
 143     pstrcpy(dst, dst_len, src);
 144
 145     return 0;
 146 }
 147
 148 static void qemu_rbd_unescape(char *src)
 149 {
 150     char *p;
 151
 152     for (p = src; *src; ++src, ++p) {
 153         if (*src == '\\' && src[1] != '\0') {
 154             src++;
 155         }
 156         *p = *src;
 157     }
 158     *p = '\0';
 159 }
 160
 161 static int qemu_rbd_parsename(const char *filename,
 162                               char *pool, int pool_len,
 163                               char *snap, int snap_len,
 164                               char *name, int name_len,
 165                               char *conf, int conf_len)
 166 {
 167     const char *start;
 168     char *p, *buf;
 169     int ret;
 170
 171     if (!strstart(filename, "rbd:", &start)) {
 172         return -EINVAL;
 173     }
 174
 175     buf = g_strdup(start);
 176     p = buf;
 177     *snap = '\0';
 178     *conf = '\0';
 179
 180     ret = qemu_rbd_next_tok(pool, pool_len, p, '/', "pool name", &p);
 181     if (ret < 0 || !p) {
 182         ret = -EINVAL;
 183         goto done;
 184     }
 185     qemu_rbd_unescape(pool);
 186
 187     if (strchr(p, '@')) {
 188         ret = qemu_rbd_next_tok(name, name_len, p, '@', "object name", &p);
 189         if (ret < 0) {
 190             goto done;
 191         }
 192         ret = qemu_rbd_next_tok(snap, snap_len, p, ':', "snap name", &p);
 193         qemu_rbd_unescape(snap);
 194     } else {
 195         ret = qemu_rbd_next_tok(name, name_len, p, ':', "object name", &p);
 196     }
 197     qemu_rbd_unescape(name);
 198     if (ret < 0 || !p) {
 199         goto done;
 200     }
 201
 202     ret = qemu_rbd_next_tok(conf, conf_len, p, '\0', "configuration", &p);
 203
 204 done:
 205     g_free(buf);
 206     return ret;
 207 }
 208
 209 static char *qemu_rbd_parse_clientname(const char *conf, char *clientname)
 210 {
 211     const char *p = conf;
 212
 213     while (*p) {
 214         int len;
 215         const char *end = strchr(p, ':');
 216
 217         if (end) {
 218             len = end - p;
 219         } else {
 220             len = strlen(p);
 221         }
 222
 223         if (strncmp(p, "id=", 3) == 0) {
 224             len -= 3;
 225             strncpy(clientname, p + 3, len);
 226             clientname[len] = '\0';
 227             return clientname;
 228         }
 229         if (end == NULL) {
 230             break;
 231         }
 232         p = end + 1;
 233     }
 234     return NULL;
 235 }
 236
 237 static int qemu_rbd_set_conf(rados_t cluster, const char *conf)
 238 {
 239     char *p, *buf;
 240     char name[RBD_MAX_CONF_NAME_SIZE];
 241     char value[RBD_MAX_CONF_VAL_SIZE];
 242     int ret = 0;
 243
 244     buf = g_strdup(conf);
 245     p = buf;
 246
 247     while (p) {
 248         ret = qemu_rbd_next_tok(name, sizeof(name), p,
 249                                 '=', "conf option name", &p);
 250         if (ret < 0) {
 251             break;
 252         }
 253         qemu_rbd_unescape(name);
 254
 255         if (!p) {
 256             error_report("conf option %s has no value", name);
 257             ret = -EINVAL;
 258             break;
 259         }
 260
 261         ret = qemu_rbd_next_tok(value, sizeof(value), p,
 262                                 ':', "conf option value", &p);
 263         if (ret < 0) {
 264             break;
 265         }
 266         qemu_rbd_unescape(value);
 267
 268         if (strcmp(name, "conf") == 0) {
 269             ret = rados_conf_read_file(cluster, value);
 270             if (ret < 0) {
 271                 error_report("error reading conf file %s", value);
 272                 break;
 273             }
 274         } else if (strcmp(name, "id") == 0) {
 275             /* ignore, this is parsed by qemu_rbd_parse_clientname() */
 276         } else {
 277             ret = rados_conf_set(cluster, name, value);
 278             if (ret < 0) {
 279                 error_report("invalid conf option %s", name);
 280                 ret = -EINVAL;
 281                 break;
 282             }
 283         }
 284     }
 285
 286     g_free(buf);
 287     return ret;
 288 }
 289
 290 static int qemu_rbd_create(const char *filename, QEMUOptionParameter *options)
 291 {
 292     int64_t bytes = 0;
 293     int64_t objsize;
 294     int obj_order = 0;
 295     char pool[RBD_MAX_POOL_NAME_SIZE];
 296     char name[RBD_MAX_IMAGE_NAME_SIZE];
 297     char snap_buf[RBD_MAX_SNAP_NAME_SIZE];
 298     char conf[RBD_MAX_CONF_SIZE];
 299     char clientname_buf[RBD_MAX_CONF_SIZE];
 300     char *clientname;
 301     rados_t cluster;
 302     rados_ioctx_t io_ctx;
 303     int ret;
 304
 305     if (qemu_rbd_parsename(filename, pool, sizeof(pool),
 306                            snap_buf, sizeof(snap_buf),
 307                            name, sizeof(name),
 308                            conf, sizeof(conf)) < 0) {
 309         return -EINVAL;
 310     }
 311
 312     /* Read out options */
 313     while (options && options->name) {
 314         if (!strcmp(options->name, BLOCK_OPT_SIZE)) {
 315             bytes = options->value.n;
 316         } else if (!strcmp(options->name, BLOCK_OPT_CLUSTER_SIZE)) {
 317             if (options->value.n) {
 318                 objsize = options->value.n;
 319                 if ((objsize - 1) & objsize) {    /* not a power of 2? */
 320                     error_report("obj size needs to be power of 2");
 321                     return -EINVAL;
 322                 }
 323                 if (objsize < 4096) {
 324                     error_report("obj size too small");
 325                     return -EINVAL;
 326                 }
 327                 obj_order = ffs(objsize) - 1;
 328             }
 329         }
 330         options++;
 331     }
 332
 333     clientname = qemu_rbd_parse_clientname(conf, clientname_buf);
 334     if (rados_create(&cluster, clientname) < 0) {
 335         error_report("error initializing");
 336         return -EIO;
 337     }
 338
 339     if (strstr(conf, "conf=") == NULL) {
 340         /* try default location, but ignore failure */
 341         rados_conf_read_file(cluster, NULL);
 342     }
 343
 344     if (conf[0] != '\0' &&
 345         qemu_rbd_set_conf(cluster, conf) < 0) {
 346         error_report("error setting config options");
 347         rados_shutdown(cluster);
 348         return -EIO;
 349     }
 350
 351     if (rados_connect(cluster) < 0) {
 352         error_report("error connecting");
 353         rados_shutdown(cluster);
 354         return -EIO;
 355     }
 356
 357     if (rados_ioctx_create(cluster, pool, &io_ctx) < 0) {
 358         error_report("error opening pool %s", pool);
 359         rados_shutdown(cluster);
 360         return -EIO;
 361     }
 362
 363     ret = rbd_create(io_ctx, name, bytes, &obj_order);
 364     rados_ioctx_destroy(io_ctx);
 365     rados_shutdown(cluster);
 366
 367     return ret;
 368 }
 369
 370 /*
 371  * This aio completion is being called from qemu_rbd_aio_event_reader()
 372  * and runs in qemu context. It schedules a bh, but just in case the aio
 373  * was not cancelled before.
 374  */
 375 static void qemu_rbd_complete_aio(RADOSCB *rcb)
 376 {
 377     RBDAIOCB *acb = rcb->acb;
 378     int64_t r;
 379
 380     r = rcb->ret;
 381
 382     if (acb->cmd != RBD_AIO_READ) {
 383         if (r < 0) {
 384             acb->ret = r;
 385             acb->error = 1;
 386         } else if (!acb->error) {
 387             acb->ret = rcb->size;
 388         }
 389     } else {
 390         if (r < 0) {
 391             memset(rcb->buf, 0, rcb->size);
 392             acb->ret = r;
 393             acb->error = 1;
 394         } else if (r < rcb->size) {
 395             memset(rcb->buf + r, 0, rcb->size - r);
 396             if (!acb->error) {
 397                 acb->ret = rcb->size;
 398             }
 399         } else if (!acb->error) {
 400             acb->ret = r;
 401         }
 402     }
 403     /* Note that acb->bh can be NULL in case where the aio was cancelled */
 404     acb->bh = qemu_bh_new(rbd_aio_bh_cb, acb);
 405     qemu_bh_schedule(acb->bh);
 406     g_free(rcb);
 407 }
 408
 409 /*
 410  * aio fd read handler. It runs in the qemu context and calls the
 411  * completion handling of completed rados aio operations.
 412  */
 413 static void qemu_rbd_aio_event_reader(void *opaque)
 414 {
 415     BDRVRBDState *s = opaque;
 416
 417     ssize_t ret;
 418
 419     do {
 420         char *p = (char *)&s->event_rcb;
 421
 422         /* now read the rcb pointer that was sent from a non qemu thread */
 423         ret = read(s->fds[RBD_FD_READ], p + s->event_reader_pos,
 424                    sizeof(s->event_rcb) - s->event_reader_pos);
 425         if (ret > 0) {
 426             s->event_reader_pos += ret;
 427             if (s->event_reader_pos == sizeof(s->event_rcb)) {
 428                 s->event_reader_pos = 0;
 429                 qemu_rbd_complete_aio(s->event_rcb);
 430             }
 431         }
 432     } while (ret < 0 && errno == EINTR);
 433 }
 434
 435 /* TODO Convert to fine grained options */
 436 static QemuOptsList runtime_opts = {
 437     .name = "rbd",
 438     .head = QTAILQ_HEAD_INITIALIZER(runtime_opts.head),
 439     .desc = {
 440         {
 441             .name = "filename",
 442             .type = QEMU_OPT_STRING,
 443             .help = "Specification of the rbd image",
 444         },
 445         { /* end of list */ }
 446     },
 447 };
 448
 449 static int qemu_rbd_open(BlockDriverState *bs, QDict *options, int flags)
 450 {
 451     BDRVRBDState *s = bs->opaque;
 452     char pool[RBD_MAX_POOL_NAME_SIZE];
 453     char snap_buf[RBD_MAX_SNAP_NAME_SIZE];
 454     char conf[RBD_MAX_CONF_SIZE];
 455     char clientname_buf[RBD_MAX_CONF_SIZE];
 456     char *clientname;
 457     QemuOpts *opts;
 458     Error *local_err = NULL;
 459     const char *filename;
 460     int r;
 461
 462     opts = qemu_opts_create_nofail(&runtime_opts);
 463     qemu_opts_absorb_qdict(opts, options, &local_err);
 464     if (error_is_set(&local_err)) {
 465         qerror_report_err(local_err);
 466         error_free(local_err);
 467         qemu_opts_del(opts);
 468         return -EINVAL;
 469     }
 470
 471     filename = qemu_opt_get(opts, "filename");
 472
 473     if (qemu_rbd_parsename(filename, pool, sizeof(pool),
 474                            snap_buf, sizeof(snap_buf),
 475                            s->name, sizeof(s->name),
 476                            conf, sizeof(conf)) < 0) {
 477         r = -EINVAL;
 478         goto failed_opts;
 479     }
 480
 481     clientname = qemu_rbd_parse_clientname(conf, clientname_buf);
 482     r = rados_create(&s->cluster, clientname);
 483     if (r < 0) {
 484         error_report("error initializing");
 485         goto failed_opts;
 486     }
 487
 488     s->snap = NULL;
 489     if (snap_buf[0] != '\0') {
 490         s->snap = g_strdup(snap_buf);
 491     }
 492
 493     /*
 494      * Fallback to more conservative semantics if setting cache
 495      * options fails. Ignore errors from setting rbd_cache because the
 496      * only possible error is that the option does not exist, and
 497      * librbd defaults to no caching. If write through caching cannot
 498      * be set up, fall back to no caching.
 499      */
 500     if (flags & BDRV_O_NOCACHE) {
 501         rados_conf_set(s->cluster, "rbd_cache", "false");
 502     } else {
 503         rados_conf_set(s->cluster, "rbd_cache", "true");
 504     }
 505
 506     if (strstr(conf, "conf=") == NULL) {
 507         /* try default location, but ignore failure */
 508         rados_conf_read_file(s->cluster, NULL);
 509     }
 510
 511     if (conf[0] != '\0') {
 512         r = qemu_rbd_set_conf(s->cluster, conf);
 513         if (r < 0) {
 514             error_report("error setting config options");
 515             goto failed_shutdown;
 516         }
 517     }
 518
 519     r = rados_connect(s->cluster);
 520     if (r < 0) {
 521         error_report("error connecting");
 522         goto failed_shutdown;
 523     }
 524
 525     r = rados_ioctx_create(s->cluster, pool, &s->io_ctx);
 526     if (r < 0) {
 527         error_report("error opening pool %s", pool);
 528         goto failed_shutdown;
 529     }
 530
 531     r = rbd_open(s->io_ctx, s->name, &s->image, s->snap);
 532     if (r < 0) {
 533         error_report("error reading header from %s", s->name);
 534         goto failed_open;
 535     }
 536
 537     bs->read_only = (s->snap != NULL);
 538
 539     s->event_reader_pos = 0;
 540     r = qemu_pipe(s->fds);
 541     if (r < 0) {
 542         error_report("error opening eventfd");
 543         goto failed;
 544     }
 545     fcntl(s->fds[0], F_SETFL, O_NONBLOCK);
 546     fcntl(s->fds[1], F_SETFL, O_NONBLOCK);
 547     qemu_aio_set_fd_handler(s->fds[RBD_FD_READ], qemu_rbd_aio_event_reader,
 548                             NULL, s);
 549
 550
 551     qemu_opts_del(opts);
 552     return 0;
 553
 554 failed:
 555     rbd_close(s->image);
 556 failed_open:
 557     rados_ioctx_destroy(s->io_ctx);
 558 failed_shutdown:
 559     rados_shutdown(s->cluster);
 560     g_free(s->snap);
 561 failed_opts:
 562     qemu_opts_del(opts);
 563     return r;
 564 }
 565
 566 static void qemu_rbd_close(BlockDriverState *bs)
 567 {
 568     BDRVRBDState *s = bs->opaque;
 569
 570     close(s->fds[0]);
 571     close(s->fds[1]);
 572     qemu_aio_set_fd_handler(s->fds[RBD_FD_READ], NULL, NULL, NULL);
 573
 574     rbd_close(s->image);
 575     rados_ioctx_destroy(s->io_ctx);
 576     g_free(s->snap);
 577     rados_shutdown(s->cluster);
 578 }
 579
 580 /*
 581  * Cancel aio. Since we don't reference acb in a non qemu threads,
 582  * it is safe to access it here.
 583  */
 584 static void qemu_rbd_aio_cancel(BlockDriverAIOCB *blockacb)
 585 {
 586     RBDAIOCB *acb = (RBDAIOCB *) blockacb;
 587     acb->cancelled = 1;
 588
 589     while (acb->status == -EINPROGRESS) {
 590         qemu_aio_wait();
 591     }
 592
 593     qemu_aio_release(acb);
 594 }
 595
 596 static const AIOCBInfo rbd_aiocb_info = {
 597     .aiocb_size = sizeof(RBDAIOCB),
 598     .cancel = qemu_rbd_aio_cancel,
 599 };
 600
 601 static int qemu_rbd_send_pipe(BDRVRBDState *s, RADOSCB *rcb)
 602 {
 603     int ret = 0;
 604     while (1) {
 605         fd_set wfd;
 606         int fd = s->fds[RBD_FD_WRITE];
 607
 608         /* send the op pointer to the qemu thread that is responsible
 609            for the aio/op completion. Must do it in a qemu thread context */
 610         ret = write(fd, (void *)&rcb, sizeof(rcb));
 611         if (ret >= 0) {
 612             break;
 613         }
 614         if (errno == EINTR) {
 615             continue;
 616         }
 617         if (errno != EAGAIN) {
 618             break;
 619         }
 620
 621         FD_ZERO(&wfd);
 622         FD_SET(fd, &wfd);
 623         do {
 624             ret = select(fd + 1, NULL, &wfd, NULL, NULL);
 625         } while (ret < 0 && errno == EINTR);
 626     }
 627
 628     return ret;
 629 }
 630
 631 /*
 632  * This is the callback function for rbd_aio_read and _write
 633  *
 634  * Note: this function is being called from a non qemu thread so
 635  * we need to be careful about what we do here. Generally we only
 636  * write to the block notification pipe, and do the rest of the
 637  * io completion handling from qemu_rbd_aio_event_reader() which
 638  * runs in a qemu context.
 639  */
 640 static void rbd_finish_aiocb(rbd_completion_t c, RADOSCB *rcb)
 641 {
 642     int ret;
 643     rcb->ret = rbd_aio_get_return_value(c);
 644     rbd_aio_release(c);
 645     ret = qemu_rbd_send_pipe(rcb->s, rcb);
 646     if (ret < 0) {
 647         error_report("failed writing to acb->s->fds");
 648         g_free(rcb);
 649     }
 650 }
 651
 652 /* Callback when all queued rbd_aio requests are complete */
 653
 654 static void rbd_aio_bh_cb(void *opaque)
 655 {
 656     RBDAIOCB *acb = opaque;
 657
 658     if (acb->cmd == RBD_AIO_READ) {
 659         qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
 660     }
 661     qemu_vfree(acb->bounce);
 662     acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret));
 663     qemu_bh_delete(acb->bh);
 664     acb->bh = NULL;
 665     acb->status = 0;
 666
 667     if (!acb->cancelled) {
 668         qemu_aio_release(acb);
 669     }
 670 }
 671
 672 static int rbd_aio_discard_wrapper(rbd_image_t image,
 673                                    uint64_t off,
 674                                    uint64_t len,
 675                                    rbd_completion_t comp)
 676 {
 677 #ifdef LIBRBD_SUPPORTS_DISCARD
 678     return rbd_aio_discard(image, off, len, comp);
 679 #else
 680     return -ENOTSUP;
 681 #endif
 682 }
 683
 684 static int rbd_aio_flush_wrapper(rbd_image_t image,
 685                                  rbd_completion_t comp)
 686 {
 687 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH
 688     return rbd_aio_flush(image, comp);
 689 #else
 690     return -ENOTSUP;
 691 #endif
 692 }
 693
 694 static BlockDriverAIOCB *rbd_start_aio(BlockDriverState *bs,
 695                                        int64_t sector_num,
 696                                        QEMUIOVector *qiov,
 697                                        int nb_sectors,
 698                                        BlockDriverCompletionFunc *cb,
 699                                        void *opaque,
 700                                        RBDAIOCmd cmd)
 701 {
 702     RBDAIOCB *acb;
 703     RADOSCB *rcb;
 704     rbd_completion_t c;
 705     int64_t off, size;
 706     char *buf;
 707     int r;
 708
 709     BDRVRBDState *s = bs->opaque;
 710
 711     acb = qemu_aio_get(&rbd_aiocb_info, bs, cb, opaque);
 712     acb->cmd = cmd;
 713     acb->qiov = qiov;
 714     if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
 715         acb->bounce = NULL;
 716     } else {
 717         acb->bounce = qemu_blockalign(bs, qiov->size);
 718     }
 719     acb->ret = 0;
 720     acb->error = 0;
 721     acb->s = s;
 722     acb->cancelled = 0;
 723     acb->bh = NULL;
 724     acb->status = -EINPROGRESS;
 725
 726     if (cmd == RBD_AIO_WRITE) {
 727         qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
 728     }
 729
 730     buf = acb->bounce;
 731
 732     off = sector_num * BDRV_SECTOR_SIZE;
 733     size = nb_sectors * BDRV_SECTOR_SIZE;
 734
 735     rcb = g_malloc(sizeof(RADOSCB));
 736     rcb->done = 0;
 737     rcb->acb = acb;
 738     rcb->buf = buf;
 739     rcb->s = acb->s;
 740     rcb->size = size;
 741     r = rbd_aio_create_completion(rcb, (rbd_callback_t) rbd_finish_aiocb, &c);
 742     if (r < 0) {
 743         goto failed;
 744     }
 745
 746     switch (cmd) {
 747     case RBD_AIO_WRITE:
 748         r = rbd_aio_write(s->image, off, size, buf, c);
 749         break;
 750     case RBD_AIO_READ:
 751         r = rbd_aio_read(s->image, off, size, buf, c);
 752         break;
 753     case RBD_AIO_DISCARD:
 754         r = rbd_aio_discard_wrapper(s->image, off, size, c);
 755         break;
 756     case RBD_AIO_FLUSH:
 757         r = rbd_aio_flush_wrapper(s->image, c);
 758         break;
 759     default:
 760         r = -EINVAL;
 761     }
 762
 763     if (r < 0) {
 764         goto failed;
 765     }
 766
 767     return &acb->common;
 768
 769 failed:
 770     g_free(rcb);
 771     qemu_aio_release(acb);
 772     return NULL;
 773 }
 774
 775 static BlockDriverAIOCB *qemu_rbd_aio_readv(BlockDriverState *bs,
 776                                             int64_t sector_num,
 777                                             QEMUIOVector *qiov,
 778                                             int nb_sectors,
 779                                             BlockDriverCompletionFunc *cb,
 780                                             void *opaque)
 781 {
 782     return rbd_start_aio(bs, sector_num, qiov, nb_sectors, cb, opaque,
 783                          RBD_AIO_READ);
 784 }
 785
 786 static BlockDriverAIOCB *qemu_rbd_aio_writev(BlockDriverState *bs,
 787                                              int64_t sector_num,
 788                                              QEMUIOVector *qiov,
 789                                              int nb_sectors,
 790                                              BlockDriverCompletionFunc *cb,
 791                                              void *opaque)
 792 {
 793     return rbd_start_aio(bs, sector_num, qiov, nb_sectors, cb, opaque,
 794                          RBD_AIO_WRITE);
 795 }
 796
 797 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH
 798 static BlockDriverAIOCB *qemu_rbd_aio_flush(BlockDriverState *bs,
 799                                             BlockDriverCompletionFunc *cb,
 800                                             void *opaque)
 801 {
 802     return rbd_start_aio(bs, 0, NULL, 0, cb, opaque, RBD_AIO_FLUSH);
 803 }
 804
 805 #else
 806
 807 static int qemu_rbd_co_flush(BlockDriverState *bs)
 808 {
 809 #if LIBRBD_VERSION_CODE >= LIBRBD_VERSION(0, 1, 1)
 810     /* rbd_flush added in 0.1.1 */
 811     BDRVRBDState *s = bs->opaque;
 812     return rbd_flush(s->image);
 813 #else
 814     return 0;
 815 #endif
 816 }
 817 #endif
 818
 819 static int qemu_rbd_getinfo(BlockDriverState *bs, BlockDriverInfo *bdi)
 820 {
 821     BDRVRBDState *s = bs->opaque;
 822     rbd_image_info_t info;
 823     int r;
 824
 825     r = rbd_stat(s->image, &info, sizeof(info));
 826     if (r < 0) {
 827         return r;
 828     }
 829
 830     bdi->cluster_size = info.obj_size;
 831     return 0;
 832 }
 833
 834 static int64_t qemu_rbd_getlength(BlockDriverState *bs)
 835 {
 836     BDRVRBDState *s = bs->opaque;
 837     rbd_image_info_t info;
 838     int r;
 839
 840     r = rbd_stat(s->image, &info, sizeof(info));
 841     if (r < 0) {
 842         return r;
 843     }
 844
 845     return info.size;
 846 }
 847
 848 static int qemu_rbd_truncate(BlockDriverState *bs, int64_t offset)
 849 {
 850     BDRVRBDState *s = bs->opaque;
 851     int r;
 852
 853     r = rbd_resize(s->image, offset);
 854     if (r < 0) {
 855         return r;
 856     }
 857
 858     return 0;
 859 }
 860
 861 static int qemu_rbd_snap_create(BlockDriverState *bs,
 862                                 QEMUSnapshotInfo *sn_info)
 863 {
 864     BDRVRBDState *s = bs->opaque;
 865     int r;
 866
 867     if (sn_info->name[0] == '\0') {
 868         return -EINVAL; /* we need a name for rbd snapshots */
 869     }
 870
 871     /*
 872      * rbd snapshots are using the name as the user controlled unique identifier
 873      * we can't use the rbd snapid for that purpose, as it can't be set
 874      */
 875     if (sn_info->id_str[0] != '\0' &&
 876         strcmp(sn_info->id_str, sn_info->name) != 0) {
 877         return -EINVAL;
 878     }
 879
 880     if (strlen(sn_info->name) >= sizeof(sn_info->id_str)) {
 881         return -ERANGE;
 882     }
 883
 884     r = rbd_snap_create(s->image, sn_info->name);
 885     if (r < 0) {
 886         error_report("failed to create snap: %s", strerror(-r));
 887         return r;
 888     }
 889
 890     return 0;
 891 }
 892
 893 static int qemu_rbd_snap_remove(BlockDriverState *bs,
 894                                 const char *snapshot_name)
 895 {
 896     BDRVRBDState *s = bs->opaque;
 897     int r;
 898
 899     r = rbd_snap_remove(s->image, snapshot_name);
 900     return r;
 901 }
 902
 903 static int qemu_rbd_snap_rollback(BlockDriverState *bs,
 904                                   const char *snapshot_name)
 905 {
 906     BDRVRBDState *s = bs->opaque;
 907     int r;
 908
 909     r = rbd_snap_rollback(s->image, snapshot_name);
 910     return r;
 911 }
 912
 913 static int qemu_rbd_snap_list(BlockDriverState *bs,
 914                               QEMUSnapshotInfo **psn_tab)
 915 {
 916     BDRVRBDState *s = bs->opaque;
 917     QEMUSnapshotInfo *sn_info, *sn_tab = NULL;
 918     int i, snap_count;
 919     rbd_snap_info_t *snaps;
 920     int max_snaps = RBD_MAX_SNAPS;
 921
 922     do {
 923         snaps = g_malloc(sizeof(*snaps) * max_snaps);
 924         snap_count = rbd_snap_list(s->image, snaps, &max_snaps);
 925         if (snap_count < 0) {
 926             g_free(snaps);
 927         }
 928     } while (snap_count == -ERANGE);
 929
 930     if (snap_count <= 0) {
 931         goto done;
 932     }
 933
 934     sn_tab = g_malloc0(snap_count * sizeof(QEMUSnapshotInfo));
 935
 936     for (i = 0; i < snap_count; i++) {
 937         const char *snap_name = snaps[i].name;
 938
 939         sn_info = sn_tab + i;
 940         pstrcpy(sn_info->id_str, sizeof(sn_info->id_str), snap_name);
 941         pstrcpy(sn_info->name, sizeof(sn_info->name), snap_name);
 942
 943         sn_info->vm_state_size = snaps[i].size;
 944         sn_info->date_sec = 0;
 945         sn_info->date_nsec = 0;
 946         sn_info->vm_clock_nsec = 0;
 947     }
 948     rbd_snap_list_end(snaps);
 949
 950  done:
 951     *psn_tab = sn_tab;
 952     return snap_count;
 953 }
 954
 955 #ifdef LIBRBD_SUPPORTS_DISCARD
 956 static BlockDriverAIOCB* qemu_rbd_aio_discard(BlockDriverState *bs,
 957                                               int64_t sector_num,
 958                                               int nb_sectors,
 959                                               BlockDriverCompletionFunc *cb,
 960                                               void *opaque)
 961 {
 962     return rbd_start_aio(bs, sector_num, NULL, nb_sectors, cb, opaque,
 963                          RBD_AIO_DISCARD);
 964 }
 965 #endif
 966
 967 static QEMUOptionParameter qemu_rbd_create_options[] = {
 968     {
 969      .name = BLOCK_OPT_SIZE,
 970      .type = OPT_SIZE,
 971      .help = "Virtual disk size"
 972     },
 973     {
 974      .name = BLOCK_OPT_CLUSTER_SIZE,
 975      .type = OPT_SIZE,
 976      .help = "RBD object size"
 977     },
 978     {NULL}
 979 };
 980
 981 static BlockDriver bdrv_rbd = {
 982     .format_name        = "rbd",
 983     .instance_size      = sizeof(BDRVRBDState),
 984     .bdrv_file_open     = qemu_rbd_open,
 985     .bdrv_close         = qemu_rbd_close,
 986     .bdrv_create        = qemu_rbd_create,
 987     .bdrv_has_zero_init = bdrv_has_zero_init_1,
 988     .bdrv_get_info      = qemu_rbd_getinfo,
 989     .create_options     = qemu_rbd_create_options,
 990     .bdrv_getlength     = qemu_rbd_getlength,
 991     .bdrv_truncate      = qemu_rbd_truncate,
 992     .protocol_name      = "rbd",
 993
 994     .bdrv_aio_readv         = qemu_rbd_aio_readv,
 995     .bdrv_aio_writev        = qemu_rbd_aio_writev,
 996
 997 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH
 998     .bdrv_aio_flush         = qemu_rbd_aio_flush,
 999 #else
1000     .bdrv_co_flush_to_disk  = qemu_rbd_co_flush,
1001 #endif
1002
1003 #ifdef LIBRBD_SUPPORTS_DISCARD
1004     .bdrv_aio_discard       = qemu_rbd_aio_discard,
1005 #endif
1006
1007     .bdrv_snapshot_create   = qemu_rbd_snap_create,
1008     .bdrv_snapshot_delete   = qemu_rbd_snap_remove,
1009     .bdrv_snapshot_list     = qemu_rbd_snap_list,
1010     .bdrv_snapshot_goto     = qemu_rbd_snap_rollback,
1011 };
1012
1013 static void bdrv_rbd_init(void)
1014 {
1015     bdrv_register(&bdrv_rbd);
1016 }
1017
1018 block_init(bdrv_rbd_init);