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 "qemu/osdep.h" 15 16 #include <rbd/librbd.h> 17 #include "qapi/error.h" 18 #include "qemu/error-report.h" 19 #include "qemu/module.h" 20 #include "qemu/option.h" 21 #include "block/block_int.h" 22 #include "block/qdict.h" 23 #include "crypto/secret.h" 24 #include "qemu/cutils.h" 25 #include "sysemu/replay.h" 26 #include "qapi/qmp/qstring.h" 27 #include "qapi/qmp/qdict.h" 28 #include "qapi/qmp/qjson.h" 29 #include "qapi/qmp/qlist.h" 30 #include "qapi/qobject-input-visitor.h" 31 #include "qapi/qapi-visit-block-core.h" 32 33 /* 34 * When specifying the image filename use: 35 * 36 * rbd:poolname/devicename[@snapshotname][:option1=value1[:option2=value2...]] 37 * 38 * poolname must be the name of an existing rados pool. 39 * 40 * devicename is the name of the rbd image. 41 * 42 * Each option given is used to configure rados, and may be any valid 43 * Ceph option, "id", or "conf". 44 * 45 * The "id" option indicates what user we should authenticate as to 46 * the Ceph cluster. If it is excluded we will use the Ceph default 47 * (normally 'admin'). 48 * 49 * The "conf" option specifies a Ceph configuration file to read. If 50 * it is not specified, we will read from the default Ceph locations 51 * (e.g., /etc/ceph/ceph.conf). To avoid reading _any_ configuration 52 * file, specify conf=/dev/null. 53 * 54 * Configuration values containing :, @, or = can be escaped with a 55 * leading "\". 56 */ 57 58 /* rbd_aio_discard added in 0.1.2 */ 59 #if LIBRBD_VERSION_CODE >= LIBRBD_VERSION(0, 1, 2) 60 #define LIBRBD_SUPPORTS_DISCARD 61 #else 62 #undef LIBRBD_SUPPORTS_DISCARD 63 #endif 64 65 #define OBJ_MAX_SIZE (1UL << OBJ_DEFAULT_OBJ_ORDER) 66 67 #define RBD_MAX_SNAPS 100 68 69 /* The LIBRBD_SUPPORTS_IOVEC is defined in librbd.h */ 70 #ifdef LIBRBD_SUPPORTS_IOVEC 71 #define LIBRBD_USE_IOVEC 1 72 #else 73 #define LIBRBD_USE_IOVEC 0 74 #endif 75 76 typedef enum { 77 RBD_AIO_READ, 78 RBD_AIO_WRITE, 79 RBD_AIO_DISCARD, 80 RBD_AIO_FLUSH 81 } RBDAIOCmd; 82 83 typedef struct RBDAIOCB { 84 BlockAIOCB common; 85 int64_t ret; 86 QEMUIOVector *qiov; 87 char *bounce; 88 RBDAIOCmd cmd; 89 int error; 90 struct BDRVRBDState *s; 91 } RBDAIOCB; 92 93 typedef struct RADOSCB { 94 RBDAIOCB *acb; 95 struct BDRVRBDState *s; 96 int64_t size; 97 char *buf; 98 int64_t ret; 99 } RADOSCB; 100 101 typedef struct BDRVRBDState { 102 rados_t cluster; 103 rados_ioctx_t io_ctx; 104 rbd_image_t image; 105 char *image_name; 106 char *snap; 107 char *namespace; 108 uint64_t image_size; 109 } BDRVRBDState; 110 111 static int qemu_rbd_connect(rados_t *cluster, rados_ioctx_t *io_ctx, 112 BlockdevOptionsRbd *opts, bool cache, 113 const char *keypairs, const char *secretid, 114 Error **errp); 115 116 static char *qemu_rbd_next_tok(char *src, char delim, char **p) 117 { 118 char *end; 119 120 *p = NULL; 121 122 for (end = src; *end; ++end) { 123 if (*end == delim) { 124 break; 125 } 126 if (*end == '\\' && end[1] != '\0') { 127 end++; 128 } 129 } 130 if (*end == delim) { 131 *p = end + 1; 132 *end = '\0'; 133 } 134 return src; 135 } 136 137 static void qemu_rbd_unescape(char *src) 138 { 139 char *p; 140 141 for (p = src; *src; ++src, ++p) { 142 if (*src == '\\' && src[1] != '\0') { 143 src++; 144 } 145 *p = *src; 146 } 147 *p = '\0'; 148 } 149 150 static void qemu_rbd_parse_filename(const char *filename, QDict *options, 151 Error **errp) 152 { 153 const char *start; 154 char *p, *buf; 155 QList *keypairs = NULL; 156 char *found_str, *image_name; 157 158 if (!strstart(filename, "rbd:", &start)) { 159 error_setg(errp, "File name must start with 'rbd:'"); 160 return; 161 } 162 163 buf = g_strdup(start); 164 p = buf; 165 166 found_str = qemu_rbd_next_tok(p, '/', &p); 167 if (!p) { 168 error_setg(errp, "Pool name is required"); 169 goto done; 170 } 171 qemu_rbd_unescape(found_str); 172 qdict_put_str(options, "pool", found_str); 173 174 if (strchr(p, '@')) { 175 image_name = qemu_rbd_next_tok(p, '@', &p); 176 177 found_str = qemu_rbd_next_tok(p, ':', &p); 178 qemu_rbd_unescape(found_str); 179 qdict_put_str(options, "snapshot", found_str); 180 } else { 181 image_name = qemu_rbd_next_tok(p, ':', &p); 182 } 183 /* Check for namespace in the image_name */ 184 if (strchr(image_name, '/')) { 185 found_str = qemu_rbd_next_tok(image_name, '/', &image_name); 186 qemu_rbd_unescape(found_str); 187 qdict_put_str(options, "namespace", found_str); 188 } else { 189 qdict_put_str(options, "namespace", ""); 190 } 191 qemu_rbd_unescape(image_name); 192 qdict_put_str(options, "image", image_name); 193 if (!p) { 194 goto done; 195 } 196 197 /* The following are essentially all key/value pairs, and we treat 198 * 'id' and 'conf' a bit special. Key/value pairs may be in any order. */ 199 while (p) { 200 char *name, *value; 201 name = qemu_rbd_next_tok(p, '=', &p); 202 if (!p) { 203 error_setg(errp, "conf option %s has no value", name); 204 break; 205 } 206 207 qemu_rbd_unescape(name); 208 209 value = qemu_rbd_next_tok(p, ':', &p); 210 qemu_rbd_unescape(value); 211 212 if (!strcmp(name, "conf")) { 213 qdict_put_str(options, "conf", value); 214 } else if (!strcmp(name, "id")) { 215 qdict_put_str(options, "user", value); 216 } else { 217 /* 218 * We pass these internally to qemu_rbd_set_keypairs(), so 219 * we can get away with the simpler list of [ "key1", 220 * "value1", "key2", "value2" ] rather than a raw dict 221 * { "key1": "value1", "key2": "value2" } where we can't 222 * guarantee order, or even a more correct but complex 223 * [ { "key1": "value1" }, { "key2": "value2" } ] 224 */ 225 if (!keypairs) { 226 keypairs = qlist_new(); 227 } 228 qlist_append_str(keypairs, name); 229 qlist_append_str(keypairs, value); 230 } 231 } 232 233 if (keypairs) { 234 qdict_put(options, "=keyvalue-pairs", 235 qobject_to_json(QOBJECT(keypairs))); 236 } 237 238 done: 239 g_free(buf); 240 qobject_unref(keypairs); 241 return; 242 } 243 244 245 static void qemu_rbd_refresh_limits(BlockDriverState *bs, Error **errp) 246 { 247 /* XXX Does RBD support AIO on less than 512-byte alignment? */ 248 bs->bl.request_alignment = 512; 249 } 250 251 252 static int qemu_rbd_set_auth(rados_t cluster, BlockdevOptionsRbd *opts, 253 Error **errp) 254 { 255 char *key, *acr; 256 int r; 257 GString *accu; 258 RbdAuthModeList *auth; 259 260 if (opts->key_secret) { 261 key = qcrypto_secret_lookup_as_base64(opts->key_secret, errp); 262 if (!key) { 263 return -EIO; 264 } 265 r = rados_conf_set(cluster, "key", key); 266 g_free(key); 267 if (r < 0) { 268 error_setg_errno(errp, -r, "Could not set 'key'"); 269 return r; 270 } 271 } 272 273 if (opts->has_auth_client_required) { 274 accu = g_string_new(""); 275 for (auth = opts->auth_client_required; auth; auth = auth->next) { 276 if (accu->str[0]) { 277 g_string_append_c(accu, ';'); 278 } 279 g_string_append(accu, RbdAuthMode_str(auth->value)); 280 } 281 acr = g_string_free(accu, FALSE); 282 r = rados_conf_set(cluster, "auth_client_required", acr); 283 g_free(acr); 284 if (r < 0) { 285 error_setg_errno(errp, -r, 286 "Could not set 'auth_client_required'"); 287 return r; 288 } 289 } 290 291 return 0; 292 } 293 294 static int qemu_rbd_set_keypairs(rados_t cluster, const char *keypairs_json, 295 Error **errp) 296 { 297 QList *keypairs; 298 QString *name; 299 QString *value; 300 const char *key; 301 size_t remaining; 302 int ret = 0; 303 304 if (!keypairs_json) { 305 return ret; 306 } 307 keypairs = qobject_to(QList, 308 qobject_from_json(keypairs_json, &error_abort)); 309 remaining = qlist_size(keypairs) / 2; 310 assert(remaining); 311 312 while (remaining--) { 313 name = qobject_to(QString, qlist_pop(keypairs)); 314 value = qobject_to(QString, qlist_pop(keypairs)); 315 assert(name && value); 316 key = qstring_get_str(name); 317 318 ret = rados_conf_set(cluster, key, qstring_get_str(value)); 319 qobject_unref(value); 320 if (ret < 0) { 321 error_setg_errno(errp, -ret, "invalid conf option %s", key); 322 qobject_unref(name); 323 ret = -EINVAL; 324 break; 325 } 326 qobject_unref(name); 327 } 328 329 qobject_unref(keypairs); 330 return ret; 331 } 332 333 static void qemu_rbd_memset(RADOSCB *rcb, int64_t offs) 334 { 335 if (LIBRBD_USE_IOVEC) { 336 RBDAIOCB *acb = rcb->acb; 337 iov_memset(acb->qiov->iov, acb->qiov->niov, offs, 0, 338 acb->qiov->size - offs); 339 } else { 340 memset(rcb->buf + offs, 0, rcb->size - offs); 341 } 342 } 343 344 static QemuOptsList runtime_opts = { 345 .name = "rbd", 346 .head = QTAILQ_HEAD_INITIALIZER(runtime_opts.head), 347 .desc = { 348 { 349 .name = "pool", 350 .type = QEMU_OPT_STRING, 351 .help = "Rados pool name", 352 }, 353 { 354 .name = "namespace", 355 .type = QEMU_OPT_STRING, 356 .help = "Rados namespace name in the pool", 357 }, 358 { 359 .name = "image", 360 .type = QEMU_OPT_STRING, 361 .help = "Image name in the pool", 362 }, 363 { 364 .name = "conf", 365 .type = QEMU_OPT_STRING, 366 .help = "Rados config file location", 367 }, 368 { 369 .name = "snapshot", 370 .type = QEMU_OPT_STRING, 371 .help = "Ceph snapshot name", 372 }, 373 { 374 /* maps to 'id' in rados_create() */ 375 .name = "user", 376 .type = QEMU_OPT_STRING, 377 .help = "Rados id name", 378 }, 379 /* 380 * server.* extracted manually, see qemu_rbd_mon_host() 381 */ 382 { /* end of list */ } 383 }, 384 }; 385 386 /* FIXME Deprecate and remove keypairs or make it available in QMP. */ 387 static int qemu_rbd_do_create(BlockdevCreateOptions *options, 388 const char *keypairs, const char *password_secret, 389 Error **errp) 390 { 391 BlockdevCreateOptionsRbd *opts = &options->u.rbd; 392 rados_t cluster; 393 rados_ioctx_t io_ctx; 394 int obj_order = 0; 395 int ret; 396 397 assert(options->driver == BLOCKDEV_DRIVER_RBD); 398 if (opts->location->has_snapshot) { 399 error_setg(errp, "Can't use snapshot name for image creation"); 400 return -EINVAL; 401 } 402 403 if (opts->has_cluster_size) { 404 int64_t objsize = opts->cluster_size; 405 if ((objsize - 1) & objsize) { /* not a power of 2? */ 406 error_setg(errp, "obj size needs to be power of 2"); 407 return -EINVAL; 408 } 409 if (objsize < 4096) { 410 error_setg(errp, "obj size too small"); 411 return -EINVAL; 412 } 413 obj_order = ctz32(objsize); 414 } 415 416 ret = qemu_rbd_connect(&cluster, &io_ctx, opts->location, false, keypairs, 417 password_secret, errp); 418 if (ret < 0) { 419 return ret; 420 } 421 422 ret = rbd_create(io_ctx, opts->location->image, opts->size, &obj_order); 423 if (ret < 0) { 424 error_setg_errno(errp, -ret, "error rbd create"); 425 goto out; 426 } 427 428 ret = 0; 429 out: 430 rados_ioctx_destroy(io_ctx); 431 rados_shutdown(cluster); 432 return ret; 433 } 434 435 static int qemu_rbd_co_create(BlockdevCreateOptions *options, Error **errp) 436 { 437 return qemu_rbd_do_create(options, NULL, NULL, errp); 438 } 439 440 static int coroutine_fn qemu_rbd_co_create_opts(BlockDriver *drv, 441 const char *filename, 442 QemuOpts *opts, 443 Error **errp) 444 { 445 BlockdevCreateOptions *create_options; 446 BlockdevCreateOptionsRbd *rbd_opts; 447 BlockdevOptionsRbd *loc; 448 Error *local_err = NULL; 449 const char *keypairs, *password_secret; 450 QDict *options = NULL; 451 int ret = 0; 452 453 create_options = g_new0(BlockdevCreateOptions, 1); 454 create_options->driver = BLOCKDEV_DRIVER_RBD; 455 rbd_opts = &create_options->u.rbd; 456 457 rbd_opts->location = g_new0(BlockdevOptionsRbd, 1); 458 459 password_secret = qemu_opt_get(opts, "password-secret"); 460 461 /* Read out options */ 462 rbd_opts->size = ROUND_UP(qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0), 463 BDRV_SECTOR_SIZE); 464 rbd_opts->cluster_size = qemu_opt_get_size_del(opts, 465 BLOCK_OPT_CLUSTER_SIZE, 0); 466 rbd_opts->has_cluster_size = (rbd_opts->cluster_size != 0); 467 468 options = qdict_new(); 469 qemu_rbd_parse_filename(filename, options, &local_err); 470 if (local_err) { 471 ret = -EINVAL; 472 error_propagate(errp, local_err); 473 goto exit; 474 } 475 476 /* 477 * Caution: while qdict_get_try_str() is fine, getting non-string 478 * types would require more care. When @options come from -blockdev 479 * or blockdev_add, its members are typed according to the QAPI 480 * schema, but when they come from -drive, they're all QString. 481 */ 482 loc = rbd_opts->location; 483 loc->pool = g_strdup(qdict_get_try_str(options, "pool")); 484 loc->conf = g_strdup(qdict_get_try_str(options, "conf")); 485 loc->has_conf = !!loc->conf; 486 loc->user = g_strdup(qdict_get_try_str(options, "user")); 487 loc->has_user = !!loc->user; 488 loc->q_namespace = g_strdup(qdict_get_try_str(options, "namespace")); 489 loc->image = g_strdup(qdict_get_try_str(options, "image")); 490 keypairs = qdict_get_try_str(options, "=keyvalue-pairs"); 491 492 ret = qemu_rbd_do_create(create_options, keypairs, password_secret, errp); 493 if (ret < 0) { 494 goto exit; 495 } 496 497 exit: 498 qobject_unref(options); 499 qapi_free_BlockdevCreateOptions(create_options); 500 return ret; 501 } 502 503 /* 504 * This aio completion is being called from rbd_finish_bh() and runs in qemu 505 * BH context. 506 */ 507 static void qemu_rbd_complete_aio(RADOSCB *rcb) 508 { 509 RBDAIOCB *acb = rcb->acb; 510 int64_t r; 511 512 r = rcb->ret; 513 514 if (acb->cmd != RBD_AIO_READ) { 515 if (r < 0) { 516 acb->ret = r; 517 acb->error = 1; 518 } else if (!acb->error) { 519 acb->ret = rcb->size; 520 } 521 } else { 522 if (r < 0) { 523 qemu_rbd_memset(rcb, 0); 524 acb->ret = r; 525 acb->error = 1; 526 } else if (r < rcb->size) { 527 qemu_rbd_memset(rcb, r); 528 if (!acb->error) { 529 acb->ret = rcb->size; 530 } 531 } else if (!acb->error) { 532 acb->ret = r; 533 } 534 } 535 536 g_free(rcb); 537 538 if (!LIBRBD_USE_IOVEC) { 539 if (acb->cmd == RBD_AIO_READ) { 540 qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size); 541 } 542 qemu_vfree(acb->bounce); 543 } 544 545 acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret)); 546 547 qemu_aio_unref(acb); 548 } 549 550 static char *qemu_rbd_mon_host(BlockdevOptionsRbd *opts, Error **errp) 551 { 552 const char **vals; 553 const char *host, *port; 554 char *rados_str; 555 InetSocketAddressBaseList *p; 556 int i, cnt; 557 558 if (!opts->has_server) { 559 return NULL; 560 } 561 562 for (cnt = 0, p = opts->server; p; p = p->next) { 563 cnt++; 564 } 565 566 vals = g_new(const char *, cnt + 1); 567 568 for (i = 0, p = opts->server; p; p = p->next, i++) { 569 host = p->value->host; 570 port = p->value->port; 571 572 if (strchr(host, ':')) { 573 vals[i] = g_strdup_printf("[%s]:%s", host, port); 574 } else { 575 vals[i] = g_strdup_printf("%s:%s", host, port); 576 } 577 } 578 vals[i] = NULL; 579 580 rados_str = i ? g_strjoinv(";", (char **)vals) : NULL; 581 g_strfreev((char **)vals); 582 return rados_str; 583 } 584 585 static int qemu_rbd_connect(rados_t *cluster, rados_ioctx_t *io_ctx, 586 BlockdevOptionsRbd *opts, bool cache, 587 const char *keypairs, const char *secretid, 588 Error **errp) 589 { 590 char *mon_host = NULL; 591 Error *local_err = NULL; 592 int r; 593 594 if (secretid) { 595 if (opts->key_secret) { 596 error_setg(errp, 597 "Legacy 'password-secret' clashes with 'key-secret'"); 598 return -EINVAL; 599 } 600 opts->key_secret = g_strdup(secretid); 601 opts->has_key_secret = true; 602 } 603 604 mon_host = qemu_rbd_mon_host(opts, &local_err); 605 if (local_err) { 606 error_propagate(errp, local_err); 607 r = -EINVAL; 608 goto failed_opts; 609 } 610 611 r = rados_create(cluster, opts->user); 612 if (r < 0) { 613 error_setg_errno(errp, -r, "error initializing"); 614 goto failed_opts; 615 } 616 617 /* try default location when conf=NULL, but ignore failure */ 618 r = rados_conf_read_file(*cluster, opts->conf); 619 if (opts->has_conf && r < 0) { 620 error_setg_errno(errp, -r, "error reading conf file %s", opts->conf); 621 goto failed_shutdown; 622 } 623 624 r = qemu_rbd_set_keypairs(*cluster, keypairs, errp); 625 if (r < 0) { 626 goto failed_shutdown; 627 } 628 629 if (mon_host) { 630 r = rados_conf_set(*cluster, "mon_host", mon_host); 631 if (r < 0) { 632 goto failed_shutdown; 633 } 634 } 635 636 r = qemu_rbd_set_auth(*cluster, opts, errp); 637 if (r < 0) { 638 goto failed_shutdown; 639 } 640 641 /* 642 * Fallback to more conservative semantics if setting cache 643 * options fails. Ignore errors from setting rbd_cache because the 644 * only possible error is that the option does not exist, and 645 * librbd defaults to no caching. If write through caching cannot 646 * be set up, fall back to no caching. 647 */ 648 if (cache) { 649 rados_conf_set(*cluster, "rbd_cache", "true"); 650 } else { 651 rados_conf_set(*cluster, "rbd_cache", "false"); 652 } 653 654 r = rados_connect(*cluster); 655 if (r < 0) { 656 error_setg_errno(errp, -r, "error connecting"); 657 goto failed_shutdown; 658 } 659 660 r = rados_ioctx_create(*cluster, opts->pool, io_ctx); 661 if (r < 0) { 662 error_setg_errno(errp, -r, "error opening pool %s", opts->pool); 663 goto failed_shutdown; 664 } 665 /* 666 * Set the namespace after opening the io context on the pool, 667 * if nspace == NULL or if nspace == "", it is just as we did nothing 668 */ 669 rados_ioctx_set_namespace(*io_ctx, opts->q_namespace); 670 671 return 0; 672 673 failed_shutdown: 674 rados_shutdown(*cluster); 675 failed_opts: 676 g_free(mon_host); 677 return r; 678 } 679 680 static int qemu_rbd_convert_options(QDict *options, BlockdevOptionsRbd **opts, 681 Error **errp) 682 { 683 Visitor *v; 684 685 /* Convert the remaining options into a QAPI object */ 686 v = qobject_input_visitor_new_flat_confused(options, errp); 687 if (!v) { 688 return -EINVAL; 689 } 690 691 visit_type_BlockdevOptionsRbd(v, NULL, opts, errp); 692 visit_free(v); 693 if (!opts) { 694 return -EINVAL; 695 } 696 697 return 0; 698 } 699 700 static int qemu_rbd_attempt_legacy_options(QDict *options, 701 BlockdevOptionsRbd **opts, 702 char **keypairs) 703 { 704 char *filename; 705 int r; 706 707 filename = g_strdup(qdict_get_try_str(options, "filename")); 708 if (!filename) { 709 return -EINVAL; 710 } 711 qdict_del(options, "filename"); 712 713 qemu_rbd_parse_filename(filename, options, NULL); 714 715 /* keypairs freed by caller */ 716 *keypairs = g_strdup(qdict_get_try_str(options, "=keyvalue-pairs")); 717 if (*keypairs) { 718 qdict_del(options, "=keyvalue-pairs"); 719 } 720 721 r = qemu_rbd_convert_options(options, opts, NULL); 722 723 g_free(filename); 724 return r; 725 } 726 727 static int qemu_rbd_open(BlockDriverState *bs, QDict *options, int flags, 728 Error **errp) 729 { 730 BDRVRBDState *s = bs->opaque; 731 BlockdevOptionsRbd *opts = NULL; 732 const QDictEntry *e; 733 Error *local_err = NULL; 734 char *keypairs, *secretid; 735 int r; 736 737 keypairs = g_strdup(qdict_get_try_str(options, "=keyvalue-pairs")); 738 if (keypairs) { 739 qdict_del(options, "=keyvalue-pairs"); 740 } 741 742 secretid = g_strdup(qdict_get_try_str(options, "password-secret")); 743 if (secretid) { 744 qdict_del(options, "password-secret"); 745 } 746 747 r = qemu_rbd_convert_options(options, &opts, &local_err); 748 if (local_err) { 749 /* If keypairs are present, that means some options are present in 750 * the modern option format. Don't attempt to parse legacy option 751 * formats, as we won't support mixed usage. */ 752 if (keypairs) { 753 error_propagate(errp, local_err); 754 goto out; 755 } 756 757 /* If the initial attempt to convert and process the options failed, 758 * we may be attempting to open an image file that has the rbd options 759 * specified in the older format consisting of all key/value pairs 760 * encoded in the filename. Go ahead and attempt to parse the 761 * filename, and see if we can pull out the required options. */ 762 r = qemu_rbd_attempt_legacy_options(options, &opts, &keypairs); 763 if (r < 0) { 764 /* Propagate the original error, not the legacy parsing fallback 765 * error, as the latter was just a best-effort attempt. */ 766 error_propagate(errp, local_err); 767 goto out; 768 } 769 /* Take care whenever deciding to actually deprecate; once this ability 770 * is removed, we will not be able to open any images with legacy-styled 771 * backing image strings. */ 772 warn_report("RBD options encoded in the filename as keyvalue pairs " 773 "is deprecated"); 774 } 775 776 /* Remove the processed options from the QDict (the visitor processes 777 * _all_ options in the QDict) */ 778 while ((e = qdict_first(options))) { 779 qdict_del(options, e->key); 780 } 781 782 r = qemu_rbd_connect(&s->cluster, &s->io_ctx, opts, 783 !(flags & BDRV_O_NOCACHE), keypairs, secretid, errp); 784 if (r < 0) { 785 goto out; 786 } 787 788 s->snap = g_strdup(opts->snapshot); 789 s->image_name = g_strdup(opts->image); 790 791 /* rbd_open is always r/w */ 792 r = rbd_open(s->io_ctx, s->image_name, &s->image, s->snap); 793 if (r < 0) { 794 error_setg_errno(errp, -r, "error reading header from %s", 795 s->image_name); 796 goto failed_open; 797 } 798 799 r = rbd_get_size(s->image, &s->image_size); 800 if (r < 0) { 801 error_setg_errno(errp, -r, "error getting image size from %s", 802 s->image_name); 803 rbd_close(s->image); 804 goto failed_open; 805 } 806 807 /* If we are using an rbd snapshot, we must be r/o, otherwise 808 * leave as-is */ 809 if (s->snap != NULL) { 810 r = bdrv_apply_auto_read_only(bs, "rbd snapshots are read-only", errp); 811 if (r < 0) { 812 rbd_close(s->image); 813 goto failed_open; 814 } 815 } 816 817 /* When extending regular files, we get zeros from the OS */ 818 bs->supported_truncate_flags = BDRV_REQ_ZERO_WRITE; 819 820 r = 0; 821 goto out; 822 823 failed_open: 824 rados_ioctx_destroy(s->io_ctx); 825 g_free(s->snap); 826 g_free(s->image_name); 827 rados_shutdown(s->cluster); 828 out: 829 qapi_free_BlockdevOptionsRbd(opts); 830 g_free(keypairs); 831 g_free(secretid); 832 return r; 833 } 834 835 836 /* Since RBD is currently always opened R/W via the API, 837 * we just need to check if we are using a snapshot or not, in 838 * order to determine if we will allow it to be R/W */ 839 static int qemu_rbd_reopen_prepare(BDRVReopenState *state, 840 BlockReopenQueue *queue, Error **errp) 841 { 842 BDRVRBDState *s = state->bs->opaque; 843 int ret = 0; 844 845 if (s->snap && state->flags & BDRV_O_RDWR) { 846 error_setg(errp, 847 "Cannot change node '%s' to r/w when using RBD snapshot", 848 bdrv_get_device_or_node_name(state->bs)); 849 ret = -EINVAL; 850 } 851 852 return ret; 853 } 854 855 static void qemu_rbd_close(BlockDriverState *bs) 856 { 857 BDRVRBDState *s = bs->opaque; 858 859 rbd_close(s->image); 860 rados_ioctx_destroy(s->io_ctx); 861 g_free(s->snap); 862 g_free(s->image_name); 863 rados_shutdown(s->cluster); 864 } 865 866 /* Resize the RBD image and update the 'image_size' with the current size */ 867 static int qemu_rbd_resize(BlockDriverState *bs, uint64_t size) 868 { 869 BDRVRBDState *s = bs->opaque; 870 int r; 871 872 r = rbd_resize(s->image, size); 873 if (r < 0) { 874 return r; 875 } 876 877 s->image_size = size; 878 879 return 0; 880 } 881 882 static const AIOCBInfo rbd_aiocb_info = { 883 .aiocb_size = sizeof(RBDAIOCB), 884 }; 885 886 static void rbd_finish_bh(void *opaque) 887 { 888 RADOSCB *rcb = opaque; 889 qemu_rbd_complete_aio(rcb); 890 } 891 892 /* 893 * This is the callback function for rbd_aio_read and _write 894 * 895 * Note: this function is being called from a non qemu thread so 896 * we need to be careful about what we do here. Generally we only 897 * schedule a BH, and do the rest of the io completion handling 898 * from rbd_finish_bh() which runs in a qemu context. 899 */ 900 static void rbd_finish_aiocb(rbd_completion_t c, RADOSCB *rcb) 901 { 902 RBDAIOCB *acb = rcb->acb; 903 904 rcb->ret = rbd_aio_get_return_value(c); 905 rbd_aio_release(c); 906 907 replay_bh_schedule_oneshot_event(bdrv_get_aio_context(acb->common.bs), 908 rbd_finish_bh, rcb); 909 } 910 911 static int rbd_aio_discard_wrapper(rbd_image_t image, 912 uint64_t off, 913 uint64_t len, 914 rbd_completion_t comp) 915 { 916 #ifdef LIBRBD_SUPPORTS_DISCARD 917 return rbd_aio_discard(image, off, len, comp); 918 #else 919 return -ENOTSUP; 920 #endif 921 } 922 923 static int rbd_aio_flush_wrapper(rbd_image_t image, 924 rbd_completion_t comp) 925 { 926 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH 927 return rbd_aio_flush(image, comp); 928 #else 929 return -ENOTSUP; 930 #endif 931 } 932 933 static BlockAIOCB *rbd_start_aio(BlockDriverState *bs, 934 int64_t off, 935 QEMUIOVector *qiov, 936 int64_t size, 937 BlockCompletionFunc *cb, 938 void *opaque, 939 RBDAIOCmd cmd) 940 { 941 RBDAIOCB *acb; 942 RADOSCB *rcb = NULL; 943 rbd_completion_t c; 944 int r; 945 946 BDRVRBDState *s = bs->opaque; 947 948 acb = qemu_aio_get(&rbd_aiocb_info, bs, cb, opaque); 949 acb->cmd = cmd; 950 acb->qiov = qiov; 951 assert(!qiov || qiov->size == size); 952 953 rcb = g_new(RADOSCB, 1); 954 955 if (!LIBRBD_USE_IOVEC) { 956 if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) { 957 acb->bounce = NULL; 958 } else { 959 acb->bounce = qemu_try_blockalign(bs, qiov->size); 960 if (acb->bounce == NULL) { 961 goto failed; 962 } 963 } 964 if (cmd == RBD_AIO_WRITE) { 965 qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size); 966 } 967 rcb->buf = acb->bounce; 968 } 969 970 acb->ret = 0; 971 acb->error = 0; 972 acb->s = s; 973 974 rcb->acb = acb; 975 rcb->s = acb->s; 976 rcb->size = size; 977 r = rbd_aio_create_completion(rcb, (rbd_callback_t) rbd_finish_aiocb, &c); 978 if (r < 0) { 979 goto failed; 980 } 981 982 switch (cmd) { 983 case RBD_AIO_WRITE: { 984 /* 985 * RBD APIs don't allow us to write more than actual size, so in order 986 * to support growing images, we resize the image before write 987 * operations that exceed the current size. 988 */ 989 if (off + size > s->image_size) { 990 r = qemu_rbd_resize(bs, off + size); 991 if (r < 0) { 992 goto failed_completion; 993 } 994 } 995 #ifdef LIBRBD_SUPPORTS_IOVEC 996 r = rbd_aio_writev(s->image, qiov->iov, qiov->niov, off, c); 997 #else 998 r = rbd_aio_write(s->image, off, size, rcb->buf, c); 999 #endif 1000 break; 1001 } 1002 case RBD_AIO_READ: 1003 #ifdef LIBRBD_SUPPORTS_IOVEC 1004 r = rbd_aio_readv(s->image, qiov->iov, qiov->niov, off, c); 1005 #else 1006 r = rbd_aio_read(s->image, off, size, rcb->buf, c); 1007 #endif 1008 break; 1009 case RBD_AIO_DISCARD: 1010 r = rbd_aio_discard_wrapper(s->image, off, size, c); 1011 break; 1012 case RBD_AIO_FLUSH: 1013 r = rbd_aio_flush_wrapper(s->image, c); 1014 break; 1015 default: 1016 r = -EINVAL; 1017 } 1018 1019 if (r < 0) { 1020 goto failed_completion; 1021 } 1022 return &acb->common; 1023 1024 failed_completion: 1025 rbd_aio_release(c); 1026 failed: 1027 g_free(rcb); 1028 if (!LIBRBD_USE_IOVEC) { 1029 qemu_vfree(acb->bounce); 1030 } 1031 1032 qemu_aio_unref(acb); 1033 return NULL; 1034 } 1035 1036 static BlockAIOCB *qemu_rbd_aio_preadv(BlockDriverState *bs, 1037 uint64_t offset, uint64_t bytes, 1038 QEMUIOVector *qiov, int flags, 1039 BlockCompletionFunc *cb, 1040 void *opaque) 1041 { 1042 return rbd_start_aio(bs, offset, qiov, bytes, cb, opaque, 1043 RBD_AIO_READ); 1044 } 1045 1046 static BlockAIOCB *qemu_rbd_aio_pwritev(BlockDriverState *bs, 1047 uint64_t offset, uint64_t bytes, 1048 QEMUIOVector *qiov, int flags, 1049 BlockCompletionFunc *cb, 1050 void *opaque) 1051 { 1052 return rbd_start_aio(bs, offset, qiov, bytes, cb, opaque, 1053 RBD_AIO_WRITE); 1054 } 1055 1056 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH 1057 static BlockAIOCB *qemu_rbd_aio_flush(BlockDriverState *bs, 1058 BlockCompletionFunc *cb, 1059 void *opaque) 1060 { 1061 return rbd_start_aio(bs, 0, NULL, 0, cb, opaque, RBD_AIO_FLUSH); 1062 } 1063 1064 #else 1065 1066 static int qemu_rbd_co_flush(BlockDriverState *bs) 1067 { 1068 #if LIBRBD_VERSION_CODE >= LIBRBD_VERSION(0, 1, 1) 1069 /* rbd_flush added in 0.1.1 */ 1070 BDRVRBDState *s = bs->opaque; 1071 return rbd_flush(s->image); 1072 #else 1073 return 0; 1074 #endif 1075 } 1076 #endif 1077 1078 static int qemu_rbd_getinfo(BlockDriverState *bs, BlockDriverInfo *bdi) 1079 { 1080 BDRVRBDState *s = bs->opaque; 1081 rbd_image_info_t info; 1082 int r; 1083 1084 r = rbd_stat(s->image, &info, sizeof(info)); 1085 if (r < 0) { 1086 return r; 1087 } 1088 1089 bdi->cluster_size = info.obj_size; 1090 return 0; 1091 } 1092 1093 static int64_t qemu_rbd_getlength(BlockDriverState *bs) 1094 { 1095 BDRVRBDState *s = bs->opaque; 1096 rbd_image_info_t info; 1097 int r; 1098 1099 r = rbd_stat(s->image, &info, sizeof(info)); 1100 if (r < 0) { 1101 return r; 1102 } 1103 1104 return info.size; 1105 } 1106 1107 static int coroutine_fn qemu_rbd_co_truncate(BlockDriverState *bs, 1108 int64_t offset, 1109 bool exact, 1110 PreallocMode prealloc, 1111 BdrvRequestFlags flags, 1112 Error **errp) 1113 { 1114 int r; 1115 1116 if (prealloc != PREALLOC_MODE_OFF) { 1117 error_setg(errp, "Unsupported preallocation mode '%s'", 1118 PreallocMode_str(prealloc)); 1119 return -ENOTSUP; 1120 } 1121 1122 r = qemu_rbd_resize(bs, offset); 1123 if (r < 0) { 1124 error_setg_errno(errp, -r, "Failed to resize file"); 1125 return r; 1126 } 1127 1128 return 0; 1129 } 1130 1131 static int qemu_rbd_snap_create(BlockDriverState *bs, 1132 QEMUSnapshotInfo *sn_info) 1133 { 1134 BDRVRBDState *s = bs->opaque; 1135 int r; 1136 1137 if (sn_info->name[0] == '\0') { 1138 return -EINVAL; /* we need a name for rbd snapshots */ 1139 } 1140 1141 /* 1142 * rbd snapshots are using the name as the user controlled unique identifier 1143 * we can't use the rbd snapid for that purpose, as it can't be set 1144 */ 1145 if (sn_info->id_str[0] != '\0' && 1146 strcmp(sn_info->id_str, sn_info->name) != 0) { 1147 return -EINVAL; 1148 } 1149 1150 if (strlen(sn_info->name) >= sizeof(sn_info->id_str)) { 1151 return -ERANGE; 1152 } 1153 1154 r = rbd_snap_create(s->image, sn_info->name); 1155 if (r < 0) { 1156 error_report("failed to create snap: %s", strerror(-r)); 1157 return r; 1158 } 1159 1160 return 0; 1161 } 1162 1163 static int qemu_rbd_snap_remove(BlockDriverState *bs, 1164 const char *snapshot_id, 1165 const char *snapshot_name, 1166 Error **errp) 1167 { 1168 BDRVRBDState *s = bs->opaque; 1169 int r; 1170 1171 if (!snapshot_name) { 1172 error_setg(errp, "rbd need a valid snapshot name"); 1173 return -EINVAL; 1174 } 1175 1176 /* If snapshot_id is specified, it must be equal to name, see 1177 qemu_rbd_snap_list() */ 1178 if (snapshot_id && strcmp(snapshot_id, snapshot_name)) { 1179 error_setg(errp, 1180 "rbd do not support snapshot id, it should be NULL or " 1181 "equal to snapshot name"); 1182 return -EINVAL; 1183 } 1184 1185 r = rbd_snap_remove(s->image, snapshot_name); 1186 if (r < 0) { 1187 error_setg_errno(errp, -r, "Failed to remove the snapshot"); 1188 } 1189 return r; 1190 } 1191 1192 static int qemu_rbd_snap_rollback(BlockDriverState *bs, 1193 const char *snapshot_name) 1194 { 1195 BDRVRBDState *s = bs->opaque; 1196 1197 return rbd_snap_rollback(s->image, snapshot_name); 1198 } 1199 1200 static int qemu_rbd_snap_list(BlockDriverState *bs, 1201 QEMUSnapshotInfo **psn_tab) 1202 { 1203 BDRVRBDState *s = bs->opaque; 1204 QEMUSnapshotInfo *sn_info, *sn_tab = NULL; 1205 int i, snap_count; 1206 rbd_snap_info_t *snaps; 1207 int max_snaps = RBD_MAX_SNAPS; 1208 1209 do { 1210 snaps = g_new(rbd_snap_info_t, max_snaps); 1211 snap_count = rbd_snap_list(s->image, snaps, &max_snaps); 1212 if (snap_count <= 0) { 1213 g_free(snaps); 1214 } 1215 } while (snap_count == -ERANGE); 1216 1217 if (snap_count <= 0) { 1218 goto done; 1219 } 1220 1221 sn_tab = g_new0(QEMUSnapshotInfo, snap_count); 1222 1223 for (i = 0; i < snap_count; i++) { 1224 const char *snap_name = snaps[i].name; 1225 1226 sn_info = sn_tab + i; 1227 pstrcpy(sn_info->id_str, sizeof(sn_info->id_str), snap_name); 1228 pstrcpy(sn_info->name, sizeof(sn_info->name), snap_name); 1229 1230 sn_info->vm_state_size = snaps[i].size; 1231 sn_info->date_sec = 0; 1232 sn_info->date_nsec = 0; 1233 sn_info->vm_clock_nsec = 0; 1234 } 1235 rbd_snap_list_end(snaps); 1236 g_free(snaps); 1237 1238 done: 1239 *psn_tab = sn_tab; 1240 return snap_count; 1241 } 1242 1243 #ifdef LIBRBD_SUPPORTS_DISCARD 1244 static BlockAIOCB *qemu_rbd_aio_pdiscard(BlockDriverState *bs, 1245 int64_t offset, 1246 int bytes, 1247 BlockCompletionFunc *cb, 1248 void *opaque) 1249 { 1250 return rbd_start_aio(bs, offset, NULL, bytes, cb, opaque, 1251 RBD_AIO_DISCARD); 1252 } 1253 #endif 1254 1255 #ifdef LIBRBD_SUPPORTS_INVALIDATE 1256 static void coroutine_fn qemu_rbd_co_invalidate_cache(BlockDriverState *bs, 1257 Error **errp) 1258 { 1259 BDRVRBDState *s = bs->opaque; 1260 int r = rbd_invalidate_cache(s->image); 1261 if (r < 0) { 1262 error_setg_errno(errp, -r, "Failed to invalidate the cache"); 1263 } 1264 } 1265 #endif 1266 1267 static QemuOptsList qemu_rbd_create_opts = { 1268 .name = "rbd-create-opts", 1269 .head = QTAILQ_HEAD_INITIALIZER(qemu_rbd_create_opts.head), 1270 .desc = { 1271 { 1272 .name = BLOCK_OPT_SIZE, 1273 .type = QEMU_OPT_SIZE, 1274 .help = "Virtual disk size" 1275 }, 1276 { 1277 .name = BLOCK_OPT_CLUSTER_SIZE, 1278 .type = QEMU_OPT_SIZE, 1279 .help = "RBD object size" 1280 }, 1281 { 1282 .name = "password-secret", 1283 .type = QEMU_OPT_STRING, 1284 .help = "ID of secret providing the password", 1285 }, 1286 { /* end of list */ } 1287 } 1288 }; 1289 1290 static const char *const qemu_rbd_strong_runtime_opts[] = { 1291 "pool", 1292 "image", 1293 "conf", 1294 "snapshot", 1295 "user", 1296 "server.", 1297 "password-secret", 1298 1299 NULL 1300 }; 1301 1302 static BlockDriver bdrv_rbd = { 1303 .format_name = "rbd", 1304 .instance_size = sizeof(BDRVRBDState), 1305 .bdrv_parse_filename = qemu_rbd_parse_filename, 1306 .bdrv_refresh_limits = qemu_rbd_refresh_limits, 1307 .bdrv_file_open = qemu_rbd_open, 1308 .bdrv_close = qemu_rbd_close, 1309 .bdrv_reopen_prepare = qemu_rbd_reopen_prepare, 1310 .bdrv_co_create = qemu_rbd_co_create, 1311 .bdrv_co_create_opts = qemu_rbd_co_create_opts, 1312 .bdrv_has_zero_init = bdrv_has_zero_init_1, 1313 .bdrv_get_info = qemu_rbd_getinfo, 1314 .create_opts = &qemu_rbd_create_opts, 1315 .bdrv_getlength = qemu_rbd_getlength, 1316 .bdrv_co_truncate = qemu_rbd_co_truncate, 1317 .protocol_name = "rbd", 1318 1319 .bdrv_aio_preadv = qemu_rbd_aio_preadv, 1320 .bdrv_aio_pwritev = qemu_rbd_aio_pwritev, 1321 1322 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH 1323 .bdrv_aio_flush = qemu_rbd_aio_flush, 1324 #else 1325 .bdrv_co_flush_to_disk = qemu_rbd_co_flush, 1326 #endif 1327 1328 #ifdef LIBRBD_SUPPORTS_DISCARD 1329 .bdrv_aio_pdiscard = qemu_rbd_aio_pdiscard, 1330 #endif 1331 1332 .bdrv_snapshot_create = qemu_rbd_snap_create, 1333 .bdrv_snapshot_delete = qemu_rbd_snap_remove, 1334 .bdrv_snapshot_list = qemu_rbd_snap_list, 1335 .bdrv_snapshot_goto = qemu_rbd_snap_rollback, 1336 #ifdef LIBRBD_SUPPORTS_INVALIDATE 1337 .bdrv_co_invalidate_cache = qemu_rbd_co_invalidate_cache, 1338 #endif 1339 1340 .strong_runtime_opts = qemu_rbd_strong_runtime_opts, 1341 }; 1342 1343 static void bdrv_rbd_init(void) 1344 { 1345 bdrv_register(&bdrv_rbd); 1346 } 1347 1348 block_init(bdrv_rbd_init); 1349