1 2 /* 3 rbd.c -- Export ceph rados objects as a Linux block device 4 5 6 based on drivers/block/osdblk.c: 7 8 Copyright 2009 Red Hat, Inc. 9 10 This program is free software; you can redistribute it and/or modify 11 it under the terms of the GNU General Public License as published by 12 the Free Software Foundation. 13 14 This program is distributed in the hope that it will be useful, 15 but WITHOUT ANY WARRANTY; without even the implied warranty of 16 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 17 GNU General Public License for more details. 18 19 You should have received a copy of the GNU General Public License 20 along with this program; see the file COPYING. If not, write to 21 the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. 22 23 24 25 For usage instructions, please refer to: 26 27 Documentation/ABI/testing/sysfs-bus-rbd 28 29 */ 30 31 #include <linux/ceph/libceph.h> 32 #include <linux/ceph/osd_client.h> 33 #include <linux/ceph/mon_client.h> 34 #include <linux/ceph/cls_lock_client.h> 35 #include <linux/ceph/striper.h> 36 #include <linux/ceph/decode.h> 37 #include <linux/fs_parser.h> 38 #include <linux/bsearch.h> 39 40 #include <linux/kernel.h> 41 #include <linux/device.h> 42 #include <linux/module.h> 43 #include <linux/blk-mq.h> 44 #include <linux/fs.h> 45 #include <linux/blkdev.h> 46 #include <linux/slab.h> 47 #include <linux/idr.h> 48 #include <linux/workqueue.h> 49 50 #include "rbd_types.h" 51 52 #define RBD_DEBUG /* Activate rbd_assert() calls */ 53 54 /* 55 * Increment the given counter and return its updated value. 56 * If the counter is already 0 it will not be incremented. 57 * If the counter is already at its maximum value returns 58 * -EINVAL without updating it. 59 */ 60 static int atomic_inc_return_safe(atomic_t *v) 61 { 62 unsigned int counter; 63 64 counter = (unsigned int)atomic_fetch_add_unless(v, 1, 0); 65 if (counter <= (unsigned int)INT_MAX) 66 return (int)counter; 67 68 atomic_dec(v); 69 70 return -EINVAL; 71 } 72 73 /* Decrement the counter. Return the resulting value, or -EINVAL */ 74 static int atomic_dec_return_safe(atomic_t *v) 75 { 76 int counter; 77 78 counter = atomic_dec_return(v); 79 if (counter >= 0) 80 return counter; 81 82 atomic_inc(v); 83 84 return -EINVAL; 85 } 86 87 #define RBD_DRV_NAME "rbd" 88 89 #define RBD_MINORS_PER_MAJOR 256 90 #define RBD_SINGLE_MAJOR_PART_SHIFT 4 91 92 #define RBD_MAX_PARENT_CHAIN_LEN 16 93 94 #define RBD_SNAP_DEV_NAME_PREFIX "snap_" 95 #define RBD_MAX_SNAP_NAME_LEN \ 96 (NAME_MAX - (sizeof (RBD_SNAP_DEV_NAME_PREFIX) - 1)) 97 98 #define RBD_MAX_SNAP_COUNT 510 /* allows max snapc to fit in 4KB */ 99 100 #define RBD_SNAP_HEAD_NAME "-" 101 102 #define BAD_SNAP_INDEX U32_MAX /* invalid index into snap array */ 103 104 /* This allows a single page to hold an image name sent by OSD */ 105 #define RBD_IMAGE_NAME_LEN_MAX (PAGE_SIZE - sizeof (__le32) - 1) 106 #define RBD_IMAGE_ID_LEN_MAX 64 107 108 #define RBD_OBJ_PREFIX_LEN_MAX 64 109 110 #define RBD_NOTIFY_TIMEOUT 5 /* seconds */ 111 #define RBD_RETRY_DELAY msecs_to_jiffies(1000) 112 113 /* Feature bits */ 114 115 #define RBD_FEATURE_LAYERING (1ULL<<0) 116 #define RBD_FEATURE_STRIPINGV2 (1ULL<<1) 117 #define RBD_FEATURE_EXCLUSIVE_LOCK (1ULL<<2) 118 #define RBD_FEATURE_OBJECT_MAP (1ULL<<3) 119 #define RBD_FEATURE_FAST_DIFF (1ULL<<4) 120 #define RBD_FEATURE_DEEP_FLATTEN (1ULL<<5) 121 #define RBD_FEATURE_DATA_POOL (1ULL<<7) 122 #define RBD_FEATURE_OPERATIONS (1ULL<<8) 123 124 #define RBD_FEATURES_ALL (RBD_FEATURE_LAYERING | \ 125 RBD_FEATURE_STRIPINGV2 | \ 126 RBD_FEATURE_EXCLUSIVE_LOCK | \ 127 RBD_FEATURE_OBJECT_MAP | \ 128 RBD_FEATURE_FAST_DIFF | \ 129 RBD_FEATURE_DEEP_FLATTEN | \ 130 RBD_FEATURE_DATA_POOL | \ 131 RBD_FEATURE_OPERATIONS) 132 133 /* Features supported by this (client software) implementation. */ 134 135 #define RBD_FEATURES_SUPPORTED (RBD_FEATURES_ALL) 136 137 /* 138 * An RBD device name will be "rbd#", where the "rbd" comes from 139 * RBD_DRV_NAME above, and # is a unique integer identifier. 140 */ 141 #define DEV_NAME_LEN 32 142 143 /* 144 * block device image metadata (in-memory version) 145 */ 146 struct rbd_image_header { 147 /* These six fields never change for a given rbd image */ 148 char *object_prefix; 149 __u8 obj_order; 150 u64 stripe_unit; 151 u64 stripe_count; 152 s64 data_pool_id; 153 u64 features; /* Might be changeable someday? */ 154 155 /* The remaining fields need to be updated occasionally */ 156 u64 image_size; 157 struct ceph_snap_context *snapc; 158 char *snap_names; /* format 1 only */ 159 u64 *snap_sizes; /* format 1 only */ 160 }; 161 162 /* 163 * An rbd image specification. 164 * 165 * The tuple (pool_id, image_id, snap_id) is sufficient to uniquely 166 * identify an image. Each rbd_dev structure includes a pointer to 167 * an rbd_spec structure that encapsulates this identity. 168 * 169 * Each of the id's in an rbd_spec has an associated name. For a 170 * user-mapped image, the names are supplied and the id's associated 171 * with them are looked up. For a layered image, a parent image is 172 * defined by the tuple, and the names are looked up. 173 * 174 * An rbd_dev structure contains a parent_spec pointer which is 175 * non-null if the image it represents is a child in a layered 176 * image. This pointer will refer to the rbd_spec structure used 177 * by the parent rbd_dev for its own identity (i.e., the structure 178 * is shared between the parent and child). 179 * 180 * Since these structures are populated once, during the discovery 181 * phase of image construction, they are effectively immutable so 182 * we make no effort to synchronize access to them. 183 * 184 * Note that code herein does not assume the image name is known (it 185 * could be a null pointer). 186 */ 187 struct rbd_spec { 188 u64 pool_id; 189 const char *pool_name; 190 const char *pool_ns; /* NULL if default, never "" */ 191 192 const char *image_id; 193 const char *image_name; 194 195 u64 snap_id; 196 const char *snap_name; 197 198 struct kref kref; 199 }; 200 201 /* 202 * an instance of the client. multiple devices may share an rbd client. 203 */ 204 struct rbd_client { 205 struct ceph_client *client; 206 struct kref kref; 207 struct list_head node; 208 }; 209 210 struct pending_result { 211 int result; /* first nonzero result */ 212 int num_pending; 213 }; 214 215 struct rbd_img_request; 216 217 enum obj_request_type { 218 OBJ_REQUEST_NODATA = 1, 219 OBJ_REQUEST_BIO, /* pointer into provided bio (list) */ 220 OBJ_REQUEST_BVECS, /* pointer into provided bio_vec array */ 221 OBJ_REQUEST_OWN_BVECS, /* private bio_vec array, doesn't own pages */ 222 }; 223 224 enum obj_operation_type { 225 OBJ_OP_READ = 1, 226 OBJ_OP_WRITE, 227 OBJ_OP_DISCARD, 228 OBJ_OP_ZEROOUT, 229 }; 230 231 #define RBD_OBJ_FLAG_DELETION (1U << 0) 232 #define RBD_OBJ_FLAG_COPYUP_ENABLED (1U << 1) 233 #define RBD_OBJ_FLAG_COPYUP_ZEROS (1U << 2) 234 #define RBD_OBJ_FLAG_MAY_EXIST (1U << 3) 235 #define RBD_OBJ_FLAG_NOOP_FOR_NONEXISTENT (1U << 4) 236 237 enum rbd_obj_read_state { 238 RBD_OBJ_READ_START = 1, 239 RBD_OBJ_READ_OBJECT, 240 RBD_OBJ_READ_PARENT, 241 }; 242 243 /* 244 * Writes go through the following state machine to deal with 245 * layering: 246 * 247 * . . . . . RBD_OBJ_WRITE_GUARD. . . . . . . . . . . . . . 248 * . | . 249 * . v . 250 * . RBD_OBJ_WRITE_READ_FROM_PARENT. . . . 251 * . | . . 252 * . v v (deep-copyup . 253 * (image . RBD_OBJ_WRITE_COPYUP_EMPTY_SNAPC . not needed) . 254 * flattened) v | . . 255 * . v . . 256 * . . . .RBD_OBJ_WRITE_COPYUP_OPS. . . . . (copyup . 257 * | not needed) v 258 * v . 259 * done . . . . . . . . . . . . . . . . . . 260 * ^ 261 * | 262 * RBD_OBJ_WRITE_FLAT 263 * 264 * Writes start in RBD_OBJ_WRITE_GUARD or _FLAT, depending on whether 265 * assert_exists guard is needed or not (in some cases it's not needed 266 * even if there is a parent). 267 */ 268 enum rbd_obj_write_state { 269 RBD_OBJ_WRITE_START = 1, 270 RBD_OBJ_WRITE_PRE_OBJECT_MAP, 271 RBD_OBJ_WRITE_OBJECT, 272 __RBD_OBJ_WRITE_COPYUP, 273 RBD_OBJ_WRITE_COPYUP, 274 RBD_OBJ_WRITE_POST_OBJECT_MAP, 275 }; 276 277 enum rbd_obj_copyup_state { 278 RBD_OBJ_COPYUP_START = 1, 279 RBD_OBJ_COPYUP_READ_PARENT, 280 __RBD_OBJ_COPYUP_OBJECT_MAPS, 281 RBD_OBJ_COPYUP_OBJECT_MAPS, 282 __RBD_OBJ_COPYUP_WRITE_OBJECT, 283 RBD_OBJ_COPYUP_WRITE_OBJECT, 284 }; 285 286 struct rbd_obj_request { 287 struct ceph_object_extent ex; 288 unsigned int flags; /* RBD_OBJ_FLAG_* */ 289 union { 290 enum rbd_obj_read_state read_state; /* for reads */ 291 enum rbd_obj_write_state write_state; /* for writes */ 292 }; 293 294 struct rbd_img_request *img_request; 295 struct ceph_file_extent *img_extents; 296 u32 num_img_extents; 297 298 union { 299 struct ceph_bio_iter bio_pos; 300 struct { 301 struct ceph_bvec_iter bvec_pos; 302 u32 bvec_count; 303 u32 bvec_idx; 304 }; 305 }; 306 307 enum rbd_obj_copyup_state copyup_state; 308 struct bio_vec *copyup_bvecs; 309 u32 copyup_bvec_count; 310 311 struct list_head osd_reqs; /* w/ r_private_item */ 312 313 struct mutex state_mutex; 314 struct pending_result pending; 315 struct kref kref; 316 }; 317 318 enum img_req_flags { 319 IMG_REQ_CHILD, /* initiator: block = 0, child image = 1 */ 320 IMG_REQ_LAYERED, /* ENOENT handling: normal = 0, layered = 1 */ 321 }; 322 323 enum rbd_img_state { 324 RBD_IMG_START = 1, 325 RBD_IMG_EXCLUSIVE_LOCK, 326 __RBD_IMG_OBJECT_REQUESTS, 327 RBD_IMG_OBJECT_REQUESTS, 328 }; 329 330 struct rbd_img_request { 331 struct rbd_device *rbd_dev; 332 enum obj_operation_type op_type; 333 enum obj_request_type data_type; 334 unsigned long flags; 335 enum rbd_img_state state; 336 union { 337 u64 snap_id; /* for reads */ 338 struct ceph_snap_context *snapc; /* for writes */ 339 }; 340 struct rbd_obj_request *obj_request; /* obj req initiator */ 341 342 struct list_head lock_item; 343 struct list_head object_extents; /* obj_req.ex structs */ 344 345 struct mutex state_mutex; 346 struct pending_result pending; 347 struct work_struct work; 348 int work_result; 349 }; 350 351 #define for_each_obj_request(ireq, oreq) \ 352 list_for_each_entry(oreq, &(ireq)->object_extents, ex.oe_item) 353 #define for_each_obj_request_safe(ireq, oreq, n) \ 354 list_for_each_entry_safe(oreq, n, &(ireq)->object_extents, ex.oe_item) 355 356 enum rbd_watch_state { 357 RBD_WATCH_STATE_UNREGISTERED, 358 RBD_WATCH_STATE_REGISTERED, 359 RBD_WATCH_STATE_ERROR, 360 }; 361 362 enum rbd_lock_state { 363 RBD_LOCK_STATE_UNLOCKED, 364 RBD_LOCK_STATE_LOCKED, 365 RBD_LOCK_STATE_RELEASING, 366 }; 367 368 /* WatchNotify::ClientId */ 369 struct rbd_client_id { 370 u64 gid; 371 u64 handle; 372 }; 373 374 struct rbd_mapping { 375 u64 size; 376 }; 377 378 /* 379 * a single device 380 */ 381 struct rbd_device { 382 int dev_id; /* blkdev unique id */ 383 384 int major; /* blkdev assigned major */ 385 int minor; 386 struct gendisk *disk; /* blkdev's gendisk and rq */ 387 388 u32 image_format; /* Either 1 or 2 */ 389 struct rbd_client *rbd_client; 390 391 char name[DEV_NAME_LEN]; /* blkdev name, e.g. rbd3 */ 392 393 spinlock_t lock; /* queue, flags, open_count */ 394 395 struct rbd_image_header header; 396 unsigned long flags; /* possibly lock protected */ 397 struct rbd_spec *spec; 398 struct rbd_options *opts; 399 char *config_info; /* add{,_single_major} string */ 400 401 struct ceph_object_id header_oid; 402 struct ceph_object_locator header_oloc; 403 404 struct ceph_file_layout layout; /* used for all rbd requests */ 405 406 struct mutex watch_mutex; 407 enum rbd_watch_state watch_state; 408 struct ceph_osd_linger_request *watch_handle; 409 u64 watch_cookie; 410 struct delayed_work watch_dwork; 411 412 struct rw_semaphore lock_rwsem; 413 enum rbd_lock_state lock_state; 414 char lock_cookie[32]; 415 struct rbd_client_id owner_cid; 416 struct work_struct acquired_lock_work; 417 struct work_struct released_lock_work; 418 struct delayed_work lock_dwork; 419 struct work_struct unlock_work; 420 spinlock_t lock_lists_lock; 421 struct list_head acquiring_list; 422 struct list_head running_list; 423 struct completion acquire_wait; 424 int acquire_err; 425 struct completion releasing_wait; 426 427 spinlock_t object_map_lock; 428 u8 *object_map; 429 u64 object_map_size; /* in objects */ 430 u64 object_map_flags; 431 432 struct workqueue_struct *task_wq; 433 434 struct rbd_spec *parent_spec; 435 u64 parent_overlap; 436 atomic_t parent_ref; 437 struct rbd_device *parent; 438 439 /* Block layer tags. */ 440 struct blk_mq_tag_set tag_set; 441 442 /* protects updating the header */ 443 struct rw_semaphore header_rwsem; 444 445 struct rbd_mapping mapping; 446 447 struct list_head node; 448 449 /* sysfs related */ 450 struct device dev; 451 unsigned long open_count; /* protected by lock */ 452 }; 453 454 /* 455 * Flag bits for rbd_dev->flags: 456 * - REMOVING (which is coupled with rbd_dev->open_count) is protected 457 * by rbd_dev->lock 458 */ 459 enum rbd_dev_flags { 460 RBD_DEV_FLAG_EXISTS, /* rbd_dev_device_setup() ran */ 461 RBD_DEV_FLAG_REMOVING, /* this mapping is being removed */ 462 RBD_DEV_FLAG_READONLY, /* -o ro or snapshot */ 463 }; 464 465 static DEFINE_MUTEX(client_mutex); /* Serialize client creation */ 466 467 static LIST_HEAD(rbd_dev_list); /* devices */ 468 static DEFINE_SPINLOCK(rbd_dev_list_lock); 469 470 static LIST_HEAD(rbd_client_list); /* clients */ 471 static DEFINE_SPINLOCK(rbd_client_list_lock); 472 473 /* Slab caches for frequently-allocated structures */ 474 475 static struct kmem_cache *rbd_img_request_cache; 476 static struct kmem_cache *rbd_obj_request_cache; 477 478 static int rbd_major; 479 static DEFINE_IDA(rbd_dev_id_ida); 480 481 static struct workqueue_struct *rbd_wq; 482 483 static struct ceph_snap_context rbd_empty_snapc = { 484 .nref = REFCOUNT_INIT(1), 485 }; 486 487 /* 488 * single-major requires >= 0.75 version of userspace rbd utility. 489 */ 490 static bool single_major = true; 491 module_param(single_major, bool, 0444); 492 MODULE_PARM_DESC(single_major, "Use a single major number for all rbd devices (default: true)"); 493 494 static ssize_t add_store(struct bus_type *bus, const char *buf, size_t count); 495 static ssize_t remove_store(struct bus_type *bus, const char *buf, 496 size_t count); 497 static ssize_t add_single_major_store(struct bus_type *bus, const char *buf, 498 size_t count); 499 static ssize_t remove_single_major_store(struct bus_type *bus, const char *buf, 500 size_t count); 501 static int rbd_dev_image_probe(struct rbd_device *rbd_dev, int depth); 502 503 static int rbd_dev_id_to_minor(int dev_id) 504 { 505 return dev_id << RBD_SINGLE_MAJOR_PART_SHIFT; 506 } 507 508 static int minor_to_rbd_dev_id(int minor) 509 { 510 return minor >> RBD_SINGLE_MAJOR_PART_SHIFT; 511 } 512 513 static bool rbd_is_ro(struct rbd_device *rbd_dev) 514 { 515 return test_bit(RBD_DEV_FLAG_READONLY, &rbd_dev->flags); 516 } 517 518 static bool rbd_is_snap(struct rbd_device *rbd_dev) 519 { 520 return rbd_dev->spec->snap_id != CEPH_NOSNAP; 521 } 522 523 static bool __rbd_is_lock_owner(struct rbd_device *rbd_dev) 524 { 525 lockdep_assert_held(&rbd_dev->lock_rwsem); 526 527 return rbd_dev->lock_state == RBD_LOCK_STATE_LOCKED || 528 rbd_dev->lock_state == RBD_LOCK_STATE_RELEASING; 529 } 530 531 static bool rbd_is_lock_owner(struct rbd_device *rbd_dev) 532 { 533 bool is_lock_owner; 534 535 down_read(&rbd_dev->lock_rwsem); 536 is_lock_owner = __rbd_is_lock_owner(rbd_dev); 537 up_read(&rbd_dev->lock_rwsem); 538 return is_lock_owner; 539 } 540 541 static ssize_t supported_features_show(struct bus_type *bus, char *buf) 542 { 543 return sprintf(buf, "0x%llx\n", RBD_FEATURES_SUPPORTED); 544 } 545 546 static BUS_ATTR_WO(add); 547 static BUS_ATTR_WO(remove); 548 static BUS_ATTR_WO(add_single_major); 549 static BUS_ATTR_WO(remove_single_major); 550 static BUS_ATTR_RO(supported_features); 551 552 static struct attribute *rbd_bus_attrs[] = { 553 &bus_attr_add.attr, 554 &bus_attr_remove.attr, 555 &bus_attr_add_single_major.attr, 556 &bus_attr_remove_single_major.attr, 557 &bus_attr_supported_features.attr, 558 NULL, 559 }; 560 561 static umode_t rbd_bus_is_visible(struct kobject *kobj, 562 struct attribute *attr, int index) 563 { 564 if (!single_major && 565 (attr == &bus_attr_add_single_major.attr || 566 attr == &bus_attr_remove_single_major.attr)) 567 return 0; 568 569 return attr->mode; 570 } 571 572 static const struct attribute_group rbd_bus_group = { 573 .attrs = rbd_bus_attrs, 574 .is_visible = rbd_bus_is_visible, 575 }; 576 __ATTRIBUTE_GROUPS(rbd_bus); 577 578 static struct bus_type rbd_bus_type = { 579 .name = "rbd", 580 .bus_groups = rbd_bus_groups, 581 }; 582 583 static void rbd_root_dev_release(struct device *dev) 584 { 585 } 586 587 static struct device rbd_root_dev = { 588 .init_name = "rbd", 589 .release = rbd_root_dev_release, 590 }; 591 592 static __printf(2, 3) 593 void rbd_warn(struct rbd_device *rbd_dev, const char *fmt, ...) 594 { 595 struct va_format vaf; 596 va_list args; 597 598 va_start(args, fmt); 599 vaf.fmt = fmt; 600 vaf.va = &args; 601 602 if (!rbd_dev) 603 printk(KERN_WARNING "%s: %pV\n", RBD_DRV_NAME, &vaf); 604 else if (rbd_dev->disk) 605 printk(KERN_WARNING "%s: %s: %pV\n", 606 RBD_DRV_NAME, rbd_dev->disk->disk_name, &vaf); 607 else if (rbd_dev->spec && rbd_dev->spec->image_name) 608 printk(KERN_WARNING "%s: image %s: %pV\n", 609 RBD_DRV_NAME, rbd_dev->spec->image_name, &vaf); 610 else if (rbd_dev->spec && rbd_dev->spec->image_id) 611 printk(KERN_WARNING "%s: id %s: %pV\n", 612 RBD_DRV_NAME, rbd_dev->spec->image_id, &vaf); 613 else /* punt */ 614 printk(KERN_WARNING "%s: rbd_dev %p: %pV\n", 615 RBD_DRV_NAME, rbd_dev, &vaf); 616 va_end(args); 617 } 618 619 #ifdef RBD_DEBUG 620 #define rbd_assert(expr) \ 621 if (unlikely(!(expr))) { \ 622 printk(KERN_ERR "\nAssertion failure in %s() " \ 623 "at line %d:\n\n" \ 624 "\trbd_assert(%s);\n\n", \ 625 __func__, __LINE__, #expr); \ 626 BUG(); \ 627 } 628 #else /* !RBD_DEBUG */ 629 # define rbd_assert(expr) ((void) 0) 630 #endif /* !RBD_DEBUG */ 631 632 static void rbd_dev_remove_parent(struct rbd_device *rbd_dev); 633 634 static int rbd_dev_refresh(struct rbd_device *rbd_dev); 635 static int rbd_dev_v2_header_onetime(struct rbd_device *rbd_dev); 636 static int rbd_dev_header_info(struct rbd_device *rbd_dev); 637 static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev); 638 static const char *rbd_dev_v2_snap_name(struct rbd_device *rbd_dev, 639 u64 snap_id); 640 static int _rbd_dev_v2_snap_size(struct rbd_device *rbd_dev, u64 snap_id, 641 u8 *order, u64 *snap_size); 642 static int rbd_dev_v2_get_flags(struct rbd_device *rbd_dev); 643 644 static void rbd_obj_handle_request(struct rbd_obj_request *obj_req, int result); 645 static void rbd_img_handle_request(struct rbd_img_request *img_req, int result); 646 647 /* 648 * Return true if nothing else is pending. 649 */ 650 static bool pending_result_dec(struct pending_result *pending, int *result) 651 { 652 rbd_assert(pending->num_pending > 0); 653 654 if (*result && !pending->result) 655 pending->result = *result; 656 if (--pending->num_pending) 657 return false; 658 659 *result = pending->result; 660 return true; 661 } 662 663 static int rbd_open(struct block_device *bdev, fmode_t mode) 664 { 665 struct rbd_device *rbd_dev = bdev->bd_disk->private_data; 666 bool removing = false; 667 668 spin_lock_irq(&rbd_dev->lock); 669 if (test_bit(RBD_DEV_FLAG_REMOVING, &rbd_dev->flags)) 670 removing = true; 671 else 672 rbd_dev->open_count++; 673 spin_unlock_irq(&rbd_dev->lock); 674 if (removing) 675 return -ENOENT; 676 677 (void) get_device(&rbd_dev->dev); 678 679 return 0; 680 } 681 682 static void rbd_release(struct gendisk *disk, fmode_t mode) 683 { 684 struct rbd_device *rbd_dev = disk->private_data; 685 unsigned long open_count_before; 686 687 spin_lock_irq(&rbd_dev->lock); 688 open_count_before = rbd_dev->open_count--; 689 spin_unlock_irq(&rbd_dev->lock); 690 rbd_assert(open_count_before > 0); 691 692 put_device(&rbd_dev->dev); 693 } 694 695 static const struct block_device_operations rbd_bd_ops = { 696 .owner = THIS_MODULE, 697 .open = rbd_open, 698 .release = rbd_release, 699 }; 700 701 /* 702 * Initialize an rbd client instance. Success or not, this function 703 * consumes ceph_opts. Caller holds client_mutex. 704 */ 705 static struct rbd_client *rbd_client_create(struct ceph_options *ceph_opts) 706 { 707 struct rbd_client *rbdc; 708 int ret = -ENOMEM; 709 710 dout("%s:\n", __func__); 711 rbdc = kmalloc(sizeof(struct rbd_client), GFP_KERNEL); 712 if (!rbdc) 713 goto out_opt; 714 715 kref_init(&rbdc->kref); 716 INIT_LIST_HEAD(&rbdc->node); 717 718 rbdc->client = ceph_create_client(ceph_opts, rbdc); 719 if (IS_ERR(rbdc->client)) 720 goto out_rbdc; 721 ceph_opts = NULL; /* Now rbdc->client is responsible for ceph_opts */ 722 723 ret = ceph_open_session(rbdc->client); 724 if (ret < 0) 725 goto out_client; 726 727 spin_lock(&rbd_client_list_lock); 728 list_add_tail(&rbdc->node, &rbd_client_list); 729 spin_unlock(&rbd_client_list_lock); 730 731 dout("%s: rbdc %p\n", __func__, rbdc); 732 733 return rbdc; 734 out_client: 735 ceph_destroy_client(rbdc->client); 736 out_rbdc: 737 kfree(rbdc); 738 out_opt: 739 if (ceph_opts) 740 ceph_destroy_options(ceph_opts); 741 dout("%s: error %d\n", __func__, ret); 742 743 return ERR_PTR(ret); 744 } 745 746 static struct rbd_client *__rbd_get_client(struct rbd_client *rbdc) 747 { 748 kref_get(&rbdc->kref); 749 750 return rbdc; 751 } 752 753 /* 754 * Find a ceph client with specific addr and configuration. If 755 * found, bump its reference count. 756 */ 757 static struct rbd_client *rbd_client_find(struct ceph_options *ceph_opts) 758 { 759 struct rbd_client *client_node; 760 bool found = false; 761 762 if (ceph_opts->flags & CEPH_OPT_NOSHARE) 763 return NULL; 764 765 spin_lock(&rbd_client_list_lock); 766 list_for_each_entry(client_node, &rbd_client_list, node) { 767 if (!ceph_compare_options(ceph_opts, client_node->client)) { 768 __rbd_get_client(client_node); 769 770 found = true; 771 break; 772 } 773 } 774 spin_unlock(&rbd_client_list_lock); 775 776 return found ? client_node : NULL; 777 } 778 779 /* 780 * (Per device) rbd map options 781 */ 782 enum { 783 Opt_queue_depth, 784 Opt_alloc_size, 785 Opt_lock_timeout, 786 /* int args above */ 787 Opt_pool_ns, 788 Opt_compression_hint, 789 /* string args above */ 790 Opt_read_only, 791 Opt_read_write, 792 Opt_lock_on_read, 793 Opt_exclusive, 794 Opt_notrim, 795 }; 796 797 enum { 798 Opt_compression_hint_none, 799 Opt_compression_hint_compressible, 800 Opt_compression_hint_incompressible, 801 }; 802 803 static const struct constant_table rbd_param_compression_hint[] = { 804 {"none", Opt_compression_hint_none}, 805 {"compressible", Opt_compression_hint_compressible}, 806 {"incompressible", Opt_compression_hint_incompressible}, 807 {} 808 }; 809 810 static const struct fs_parameter_spec rbd_parameters[] = { 811 fsparam_u32 ("alloc_size", Opt_alloc_size), 812 fsparam_enum ("compression_hint", Opt_compression_hint, 813 rbd_param_compression_hint), 814 fsparam_flag ("exclusive", Opt_exclusive), 815 fsparam_flag ("lock_on_read", Opt_lock_on_read), 816 fsparam_u32 ("lock_timeout", Opt_lock_timeout), 817 fsparam_flag ("notrim", Opt_notrim), 818 fsparam_string ("_pool_ns", Opt_pool_ns), 819 fsparam_u32 ("queue_depth", Opt_queue_depth), 820 fsparam_flag ("read_only", Opt_read_only), 821 fsparam_flag ("read_write", Opt_read_write), 822 fsparam_flag ("ro", Opt_read_only), 823 fsparam_flag ("rw", Opt_read_write), 824 {} 825 }; 826 827 struct rbd_options { 828 int queue_depth; 829 int alloc_size; 830 unsigned long lock_timeout; 831 bool read_only; 832 bool lock_on_read; 833 bool exclusive; 834 bool trim; 835 836 u32 alloc_hint_flags; /* CEPH_OSD_OP_ALLOC_HINT_FLAG_* */ 837 }; 838 839 #define RBD_QUEUE_DEPTH_DEFAULT BLKDEV_MAX_RQ 840 #define RBD_ALLOC_SIZE_DEFAULT (64 * 1024) 841 #define RBD_LOCK_TIMEOUT_DEFAULT 0 /* no timeout */ 842 #define RBD_READ_ONLY_DEFAULT false 843 #define RBD_LOCK_ON_READ_DEFAULT false 844 #define RBD_EXCLUSIVE_DEFAULT false 845 #define RBD_TRIM_DEFAULT true 846 847 struct rbd_parse_opts_ctx { 848 struct rbd_spec *spec; 849 struct ceph_options *copts; 850 struct rbd_options *opts; 851 }; 852 853 static char* obj_op_name(enum obj_operation_type op_type) 854 { 855 switch (op_type) { 856 case OBJ_OP_READ: 857 return "read"; 858 case OBJ_OP_WRITE: 859 return "write"; 860 case OBJ_OP_DISCARD: 861 return "discard"; 862 case OBJ_OP_ZEROOUT: 863 return "zeroout"; 864 default: 865 return "???"; 866 } 867 } 868 869 /* 870 * Destroy ceph client 871 * 872 * Caller must hold rbd_client_list_lock. 873 */ 874 static void rbd_client_release(struct kref *kref) 875 { 876 struct rbd_client *rbdc = container_of(kref, struct rbd_client, kref); 877 878 dout("%s: rbdc %p\n", __func__, rbdc); 879 spin_lock(&rbd_client_list_lock); 880 list_del(&rbdc->node); 881 spin_unlock(&rbd_client_list_lock); 882 883 ceph_destroy_client(rbdc->client); 884 kfree(rbdc); 885 } 886 887 /* 888 * Drop reference to ceph client node. If it's not referenced anymore, release 889 * it. 890 */ 891 static void rbd_put_client(struct rbd_client *rbdc) 892 { 893 if (rbdc) 894 kref_put(&rbdc->kref, rbd_client_release); 895 } 896 897 /* 898 * Get a ceph client with specific addr and configuration, if one does 899 * not exist create it. Either way, ceph_opts is consumed by this 900 * function. 901 */ 902 static struct rbd_client *rbd_get_client(struct ceph_options *ceph_opts) 903 { 904 struct rbd_client *rbdc; 905 int ret; 906 907 mutex_lock(&client_mutex); 908 rbdc = rbd_client_find(ceph_opts); 909 if (rbdc) { 910 ceph_destroy_options(ceph_opts); 911 912 /* 913 * Using an existing client. Make sure ->pg_pools is up to 914 * date before we look up the pool id in do_rbd_add(). 915 */ 916 ret = ceph_wait_for_latest_osdmap(rbdc->client, 917 rbdc->client->options->mount_timeout); 918 if (ret) { 919 rbd_warn(NULL, "failed to get latest osdmap: %d", ret); 920 rbd_put_client(rbdc); 921 rbdc = ERR_PTR(ret); 922 } 923 } else { 924 rbdc = rbd_client_create(ceph_opts); 925 } 926 mutex_unlock(&client_mutex); 927 928 return rbdc; 929 } 930 931 static bool rbd_image_format_valid(u32 image_format) 932 { 933 return image_format == 1 || image_format == 2; 934 } 935 936 static bool rbd_dev_ondisk_valid(struct rbd_image_header_ondisk *ondisk) 937 { 938 size_t size; 939 u32 snap_count; 940 941 /* The header has to start with the magic rbd header text */ 942 if (memcmp(&ondisk->text, RBD_HEADER_TEXT, sizeof (RBD_HEADER_TEXT))) 943 return false; 944 945 /* The bio layer requires at least sector-sized I/O */ 946 947 if (ondisk->options.order < SECTOR_SHIFT) 948 return false; 949 950 /* If we use u64 in a few spots we may be able to loosen this */ 951 952 if (ondisk->options.order > 8 * sizeof (int) - 1) 953 return false; 954 955 /* 956 * The size of a snapshot header has to fit in a size_t, and 957 * that limits the number of snapshots. 958 */ 959 snap_count = le32_to_cpu(ondisk->snap_count); 960 size = SIZE_MAX - sizeof (struct ceph_snap_context); 961 if (snap_count > size / sizeof (__le64)) 962 return false; 963 964 /* 965 * Not only that, but the size of the entire the snapshot 966 * header must also be representable in a size_t. 967 */ 968 size -= snap_count * sizeof (__le64); 969 if ((u64) size < le64_to_cpu(ondisk->snap_names_len)) 970 return false; 971 972 return true; 973 } 974 975 /* 976 * returns the size of an object in the image 977 */ 978 static u32 rbd_obj_bytes(struct rbd_image_header *header) 979 { 980 return 1U << header->obj_order; 981 } 982 983 static void rbd_init_layout(struct rbd_device *rbd_dev) 984 { 985 if (rbd_dev->header.stripe_unit == 0 || 986 rbd_dev->header.stripe_count == 0) { 987 rbd_dev->header.stripe_unit = rbd_obj_bytes(&rbd_dev->header); 988 rbd_dev->header.stripe_count = 1; 989 } 990 991 rbd_dev->layout.stripe_unit = rbd_dev->header.stripe_unit; 992 rbd_dev->layout.stripe_count = rbd_dev->header.stripe_count; 993 rbd_dev->layout.object_size = rbd_obj_bytes(&rbd_dev->header); 994 rbd_dev->layout.pool_id = rbd_dev->header.data_pool_id == CEPH_NOPOOL ? 995 rbd_dev->spec->pool_id : rbd_dev->header.data_pool_id; 996 RCU_INIT_POINTER(rbd_dev->layout.pool_ns, NULL); 997 } 998 999 /* 1000 * Fill an rbd image header with information from the given format 1 1001 * on-disk header. 1002 */ 1003 static int rbd_header_from_disk(struct rbd_device *rbd_dev, 1004 struct rbd_image_header_ondisk *ondisk) 1005 { 1006 struct rbd_image_header *header = &rbd_dev->header; 1007 bool first_time = header->object_prefix == NULL; 1008 struct ceph_snap_context *snapc; 1009 char *object_prefix = NULL; 1010 char *snap_names = NULL; 1011 u64 *snap_sizes = NULL; 1012 u32 snap_count; 1013 int ret = -ENOMEM; 1014 u32 i; 1015 1016 /* Allocate this now to avoid having to handle failure below */ 1017 1018 if (first_time) { 1019 object_prefix = kstrndup(ondisk->object_prefix, 1020 sizeof(ondisk->object_prefix), 1021 GFP_KERNEL); 1022 if (!object_prefix) 1023 return -ENOMEM; 1024 } 1025 1026 /* Allocate the snapshot context and fill it in */ 1027 1028 snap_count = le32_to_cpu(ondisk->snap_count); 1029 snapc = ceph_create_snap_context(snap_count, GFP_KERNEL); 1030 if (!snapc) 1031 goto out_err; 1032 snapc->seq = le64_to_cpu(ondisk->snap_seq); 1033 if (snap_count) { 1034 struct rbd_image_snap_ondisk *snaps; 1035 u64 snap_names_len = le64_to_cpu(ondisk->snap_names_len); 1036 1037 /* We'll keep a copy of the snapshot names... */ 1038 1039 if (snap_names_len > (u64)SIZE_MAX) 1040 goto out_2big; 1041 snap_names = kmalloc(snap_names_len, GFP_KERNEL); 1042 if (!snap_names) 1043 goto out_err; 1044 1045 /* ...as well as the array of their sizes. */ 1046 snap_sizes = kmalloc_array(snap_count, 1047 sizeof(*header->snap_sizes), 1048 GFP_KERNEL); 1049 if (!snap_sizes) 1050 goto out_err; 1051 1052 /* 1053 * Copy the names, and fill in each snapshot's id 1054 * and size. 1055 * 1056 * Note that rbd_dev_v1_header_info() guarantees the 1057 * ondisk buffer we're working with has 1058 * snap_names_len bytes beyond the end of the 1059 * snapshot id array, this memcpy() is safe. 1060 */ 1061 memcpy(snap_names, &ondisk->snaps[snap_count], snap_names_len); 1062 snaps = ondisk->snaps; 1063 for (i = 0; i < snap_count; i++) { 1064 snapc->snaps[i] = le64_to_cpu(snaps[i].id); 1065 snap_sizes[i] = le64_to_cpu(snaps[i].image_size); 1066 } 1067 } 1068 1069 /* We won't fail any more, fill in the header */ 1070 1071 if (first_time) { 1072 header->object_prefix = object_prefix; 1073 header->obj_order = ondisk->options.order; 1074 rbd_init_layout(rbd_dev); 1075 } else { 1076 ceph_put_snap_context(header->snapc); 1077 kfree(header->snap_names); 1078 kfree(header->snap_sizes); 1079 } 1080 1081 /* The remaining fields always get updated (when we refresh) */ 1082 1083 header->image_size = le64_to_cpu(ondisk->image_size); 1084 header->snapc = snapc; 1085 header->snap_names = snap_names; 1086 header->snap_sizes = snap_sizes; 1087 1088 return 0; 1089 out_2big: 1090 ret = -EIO; 1091 out_err: 1092 kfree(snap_sizes); 1093 kfree(snap_names); 1094 ceph_put_snap_context(snapc); 1095 kfree(object_prefix); 1096 1097 return ret; 1098 } 1099 1100 static const char *_rbd_dev_v1_snap_name(struct rbd_device *rbd_dev, u32 which) 1101 { 1102 const char *snap_name; 1103 1104 rbd_assert(which < rbd_dev->header.snapc->num_snaps); 1105 1106 /* Skip over names until we find the one we are looking for */ 1107 1108 snap_name = rbd_dev->header.snap_names; 1109 while (which--) 1110 snap_name += strlen(snap_name) + 1; 1111 1112 return kstrdup(snap_name, GFP_KERNEL); 1113 } 1114 1115 /* 1116 * Snapshot id comparison function for use with qsort()/bsearch(). 1117 * Note that result is for snapshots in *descending* order. 1118 */ 1119 static int snapid_compare_reverse(const void *s1, const void *s2) 1120 { 1121 u64 snap_id1 = *(u64 *)s1; 1122 u64 snap_id2 = *(u64 *)s2; 1123 1124 if (snap_id1 < snap_id2) 1125 return 1; 1126 return snap_id1 == snap_id2 ? 0 : -1; 1127 } 1128 1129 /* 1130 * Search a snapshot context to see if the given snapshot id is 1131 * present. 1132 * 1133 * Returns the position of the snapshot id in the array if it's found, 1134 * or BAD_SNAP_INDEX otherwise. 1135 * 1136 * Note: The snapshot array is in kept sorted (by the osd) in 1137 * reverse order, highest snapshot id first. 1138 */ 1139 static u32 rbd_dev_snap_index(struct rbd_device *rbd_dev, u64 snap_id) 1140 { 1141 struct ceph_snap_context *snapc = rbd_dev->header.snapc; 1142 u64 *found; 1143 1144 found = bsearch(&snap_id, &snapc->snaps, snapc->num_snaps, 1145 sizeof (snap_id), snapid_compare_reverse); 1146 1147 return found ? (u32)(found - &snapc->snaps[0]) : BAD_SNAP_INDEX; 1148 } 1149 1150 static const char *rbd_dev_v1_snap_name(struct rbd_device *rbd_dev, 1151 u64 snap_id) 1152 { 1153 u32 which; 1154 const char *snap_name; 1155 1156 which = rbd_dev_snap_index(rbd_dev, snap_id); 1157 if (which == BAD_SNAP_INDEX) 1158 return ERR_PTR(-ENOENT); 1159 1160 snap_name = _rbd_dev_v1_snap_name(rbd_dev, which); 1161 return snap_name ? snap_name : ERR_PTR(-ENOMEM); 1162 } 1163 1164 static const char *rbd_snap_name(struct rbd_device *rbd_dev, u64 snap_id) 1165 { 1166 if (snap_id == CEPH_NOSNAP) 1167 return RBD_SNAP_HEAD_NAME; 1168 1169 rbd_assert(rbd_image_format_valid(rbd_dev->image_format)); 1170 if (rbd_dev->image_format == 1) 1171 return rbd_dev_v1_snap_name(rbd_dev, snap_id); 1172 1173 return rbd_dev_v2_snap_name(rbd_dev, snap_id); 1174 } 1175 1176 static int rbd_snap_size(struct rbd_device *rbd_dev, u64 snap_id, 1177 u64 *snap_size) 1178 { 1179 rbd_assert(rbd_image_format_valid(rbd_dev->image_format)); 1180 if (snap_id == CEPH_NOSNAP) { 1181 *snap_size = rbd_dev->header.image_size; 1182 } else if (rbd_dev->image_format == 1) { 1183 u32 which; 1184 1185 which = rbd_dev_snap_index(rbd_dev, snap_id); 1186 if (which == BAD_SNAP_INDEX) 1187 return -ENOENT; 1188 1189 *snap_size = rbd_dev->header.snap_sizes[which]; 1190 } else { 1191 u64 size = 0; 1192 int ret; 1193 1194 ret = _rbd_dev_v2_snap_size(rbd_dev, snap_id, NULL, &size); 1195 if (ret) 1196 return ret; 1197 1198 *snap_size = size; 1199 } 1200 return 0; 1201 } 1202 1203 static int rbd_dev_mapping_set(struct rbd_device *rbd_dev) 1204 { 1205 u64 snap_id = rbd_dev->spec->snap_id; 1206 u64 size = 0; 1207 int ret; 1208 1209 ret = rbd_snap_size(rbd_dev, snap_id, &size); 1210 if (ret) 1211 return ret; 1212 1213 rbd_dev->mapping.size = size; 1214 return 0; 1215 } 1216 1217 static void rbd_dev_mapping_clear(struct rbd_device *rbd_dev) 1218 { 1219 rbd_dev->mapping.size = 0; 1220 } 1221 1222 static void zero_bvec(struct bio_vec *bv) 1223 { 1224 void *buf; 1225 unsigned long flags; 1226 1227 buf = bvec_kmap_irq(bv, &flags); 1228 memset(buf, 0, bv->bv_len); 1229 flush_dcache_page(bv->bv_page); 1230 bvec_kunmap_irq(buf, &flags); 1231 } 1232 1233 static void zero_bios(struct ceph_bio_iter *bio_pos, u32 off, u32 bytes) 1234 { 1235 struct ceph_bio_iter it = *bio_pos; 1236 1237 ceph_bio_iter_advance(&it, off); 1238 ceph_bio_iter_advance_step(&it, bytes, ({ 1239 zero_bvec(&bv); 1240 })); 1241 } 1242 1243 static void zero_bvecs(struct ceph_bvec_iter *bvec_pos, u32 off, u32 bytes) 1244 { 1245 struct ceph_bvec_iter it = *bvec_pos; 1246 1247 ceph_bvec_iter_advance(&it, off); 1248 ceph_bvec_iter_advance_step(&it, bytes, ({ 1249 zero_bvec(&bv); 1250 })); 1251 } 1252 1253 /* 1254 * Zero a range in @obj_req data buffer defined by a bio (list) or 1255 * (private) bio_vec array. 1256 * 1257 * @off is relative to the start of the data buffer. 1258 */ 1259 static void rbd_obj_zero_range(struct rbd_obj_request *obj_req, u32 off, 1260 u32 bytes) 1261 { 1262 dout("%s %p data buf %u~%u\n", __func__, obj_req, off, bytes); 1263 1264 switch (obj_req->img_request->data_type) { 1265 case OBJ_REQUEST_BIO: 1266 zero_bios(&obj_req->bio_pos, off, bytes); 1267 break; 1268 case OBJ_REQUEST_BVECS: 1269 case OBJ_REQUEST_OWN_BVECS: 1270 zero_bvecs(&obj_req->bvec_pos, off, bytes); 1271 break; 1272 default: 1273 BUG(); 1274 } 1275 } 1276 1277 static void rbd_obj_request_destroy(struct kref *kref); 1278 static void rbd_obj_request_put(struct rbd_obj_request *obj_request) 1279 { 1280 rbd_assert(obj_request != NULL); 1281 dout("%s: obj %p (was %d)\n", __func__, obj_request, 1282 kref_read(&obj_request->kref)); 1283 kref_put(&obj_request->kref, rbd_obj_request_destroy); 1284 } 1285 1286 static inline void rbd_img_obj_request_add(struct rbd_img_request *img_request, 1287 struct rbd_obj_request *obj_request) 1288 { 1289 rbd_assert(obj_request->img_request == NULL); 1290 1291 /* Image request now owns object's original reference */ 1292 obj_request->img_request = img_request; 1293 dout("%s: img %p obj %p\n", __func__, img_request, obj_request); 1294 } 1295 1296 static inline void rbd_img_obj_request_del(struct rbd_img_request *img_request, 1297 struct rbd_obj_request *obj_request) 1298 { 1299 dout("%s: img %p obj %p\n", __func__, img_request, obj_request); 1300 list_del(&obj_request->ex.oe_item); 1301 rbd_assert(obj_request->img_request == img_request); 1302 rbd_obj_request_put(obj_request); 1303 } 1304 1305 static void rbd_osd_submit(struct ceph_osd_request *osd_req) 1306 { 1307 struct rbd_obj_request *obj_req = osd_req->r_priv; 1308 1309 dout("%s osd_req %p for obj_req %p objno %llu %llu~%llu\n", 1310 __func__, osd_req, obj_req, obj_req->ex.oe_objno, 1311 obj_req->ex.oe_off, obj_req->ex.oe_len); 1312 ceph_osdc_start_request(osd_req->r_osdc, osd_req, false); 1313 } 1314 1315 /* 1316 * The default/initial value for all image request flags is 0. Each 1317 * is conditionally set to 1 at image request initialization time 1318 * and currently never change thereafter. 1319 */ 1320 static void img_request_layered_set(struct rbd_img_request *img_request) 1321 { 1322 set_bit(IMG_REQ_LAYERED, &img_request->flags); 1323 } 1324 1325 static bool img_request_layered_test(struct rbd_img_request *img_request) 1326 { 1327 return test_bit(IMG_REQ_LAYERED, &img_request->flags) != 0; 1328 } 1329 1330 static bool rbd_obj_is_entire(struct rbd_obj_request *obj_req) 1331 { 1332 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 1333 1334 return !obj_req->ex.oe_off && 1335 obj_req->ex.oe_len == rbd_dev->layout.object_size; 1336 } 1337 1338 static bool rbd_obj_is_tail(struct rbd_obj_request *obj_req) 1339 { 1340 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 1341 1342 return obj_req->ex.oe_off + obj_req->ex.oe_len == 1343 rbd_dev->layout.object_size; 1344 } 1345 1346 /* 1347 * Must be called after rbd_obj_calc_img_extents(). 1348 */ 1349 static bool rbd_obj_copyup_enabled(struct rbd_obj_request *obj_req) 1350 { 1351 if (!obj_req->num_img_extents || 1352 (rbd_obj_is_entire(obj_req) && 1353 !obj_req->img_request->snapc->num_snaps)) 1354 return false; 1355 1356 return true; 1357 } 1358 1359 static u64 rbd_obj_img_extents_bytes(struct rbd_obj_request *obj_req) 1360 { 1361 return ceph_file_extents_bytes(obj_req->img_extents, 1362 obj_req->num_img_extents); 1363 } 1364 1365 static bool rbd_img_is_write(struct rbd_img_request *img_req) 1366 { 1367 switch (img_req->op_type) { 1368 case OBJ_OP_READ: 1369 return false; 1370 case OBJ_OP_WRITE: 1371 case OBJ_OP_DISCARD: 1372 case OBJ_OP_ZEROOUT: 1373 return true; 1374 default: 1375 BUG(); 1376 } 1377 } 1378 1379 static void rbd_osd_req_callback(struct ceph_osd_request *osd_req) 1380 { 1381 struct rbd_obj_request *obj_req = osd_req->r_priv; 1382 int result; 1383 1384 dout("%s osd_req %p result %d for obj_req %p\n", __func__, osd_req, 1385 osd_req->r_result, obj_req); 1386 1387 /* 1388 * Writes aren't allowed to return a data payload. In some 1389 * guarded write cases (e.g. stat + zero on an empty object) 1390 * a stat response makes it through, but we don't care. 1391 */ 1392 if (osd_req->r_result > 0 && rbd_img_is_write(obj_req->img_request)) 1393 result = 0; 1394 else 1395 result = osd_req->r_result; 1396 1397 rbd_obj_handle_request(obj_req, result); 1398 } 1399 1400 static void rbd_osd_format_read(struct ceph_osd_request *osd_req) 1401 { 1402 struct rbd_obj_request *obj_request = osd_req->r_priv; 1403 struct rbd_device *rbd_dev = obj_request->img_request->rbd_dev; 1404 struct ceph_options *opt = rbd_dev->rbd_client->client->options; 1405 1406 osd_req->r_flags = CEPH_OSD_FLAG_READ | opt->read_from_replica; 1407 osd_req->r_snapid = obj_request->img_request->snap_id; 1408 } 1409 1410 static void rbd_osd_format_write(struct ceph_osd_request *osd_req) 1411 { 1412 struct rbd_obj_request *obj_request = osd_req->r_priv; 1413 1414 osd_req->r_flags = CEPH_OSD_FLAG_WRITE; 1415 ktime_get_real_ts64(&osd_req->r_mtime); 1416 osd_req->r_data_offset = obj_request->ex.oe_off; 1417 } 1418 1419 static struct ceph_osd_request * 1420 __rbd_obj_add_osd_request(struct rbd_obj_request *obj_req, 1421 struct ceph_snap_context *snapc, int num_ops) 1422 { 1423 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 1424 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 1425 struct ceph_osd_request *req; 1426 const char *name_format = rbd_dev->image_format == 1 ? 1427 RBD_V1_DATA_FORMAT : RBD_V2_DATA_FORMAT; 1428 int ret; 1429 1430 req = ceph_osdc_alloc_request(osdc, snapc, num_ops, false, GFP_NOIO); 1431 if (!req) 1432 return ERR_PTR(-ENOMEM); 1433 1434 list_add_tail(&req->r_private_item, &obj_req->osd_reqs); 1435 req->r_callback = rbd_osd_req_callback; 1436 req->r_priv = obj_req; 1437 1438 /* 1439 * Data objects may be stored in a separate pool, but always in 1440 * the same namespace in that pool as the header in its pool. 1441 */ 1442 ceph_oloc_copy(&req->r_base_oloc, &rbd_dev->header_oloc); 1443 req->r_base_oloc.pool = rbd_dev->layout.pool_id; 1444 1445 ret = ceph_oid_aprintf(&req->r_base_oid, GFP_NOIO, name_format, 1446 rbd_dev->header.object_prefix, 1447 obj_req->ex.oe_objno); 1448 if (ret) 1449 return ERR_PTR(ret); 1450 1451 return req; 1452 } 1453 1454 static struct ceph_osd_request * 1455 rbd_obj_add_osd_request(struct rbd_obj_request *obj_req, int num_ops) 1456 { 1457 return __rbd_obj_add_osd_request(obj_req, obj_req->img_request->snapc, 1458 num_ops); 1459 } 1460 1461 static struct rbd_obj_request *rbd_obj_request_create(void) 1462 { 1463 struct rbd_obj_request *obj_request; 1464 1465 obj_request = kmem_cache_zalloc(rbd_obj_request_cache, GFP_NOIO); 1466 if (!obj_request) 1467 return NULL; 1468 1469 ceph_object_extent_init(&obj_request->ex); 1470 INIT_LIST_HEAD(&obj_request->osd_reqs); 1471 mutex_init(&obj_request->state_mutex); 1472 kref_init(&obj_request->kref); 1473 1474 dout("%s %p\n", __func__, obj_request); 1475 return obj_request; 1476 } 1477 1478 static void rbd_obj_request_destroy(struct kref *kref) 1479 { 1480 struct rbd_obj_request *obj_request; 1481 struct ceph_osd_request *osd_req; 1482 u32 i; 1483 1484 obj_request = container_of(kref, struct rbd_obj_request, kref); 1485 1486 dout("%s: obj %p\n", __func__, obj_request); 1487 1488 while (!list_empty(&obj_request->osd_reqs)) { 1489 osd_req = list_first_entry(&obj_request->osd_reqs, 1490 struct ceph_osd_request, r_private_item); 1491 list_del_init(&osd_req->r_private_item); 1492 ceph_osdc_put_request(osd_req); 1493 } 1494 1495 switch (obj_request->img_request->data_type) { 1496 case OBJ_REQUEST_NODATA: 1497 case OBJ_REQUEST_BIO: 1498 case OBJ_REQUEST_BVECS: 1499 break; /* Nothing to do */ 1500 case OBJ_REQUEST_OWN_BVECS: 1501 kfree(obj_request->bvec_pos.bvecs); 1502 break; 1503 default: 1504 BUG(); 1505 } 1506 1507 kfree(obj_request->img_extents); 1508 if (obj_request->copyup_bvecs) { 1509 for (i = 0; i < obj_request->copyup_bvec_count; i++) { 1510 if (obj_request->copyup_bvecs[i].bv_page) 1511 __free_page(obj_request->copyup_bvecs[i].bv_page); 1512 } 1513 kfree(obj_request->copyup_bvecs); 1514 } 1515 1516 kmem_cache_free(rbd_obj_request_cache, obj_request); 1517 } 1518 1519 /* It's OK to call this for a device with no parent */ 1520 1521 static void rbd_spec_put(struct rbd_spec *spec); 1522 static void rbd_dev_unparent(struct rbd_device *rbd_dev) 1523 { 1524 rbd_dev_remove_parent(rbd_dev); 1525 rbd_spec_put(rbd_dev->parent_spec); 1526 rbd_dev->parent_spec = NULL; 1527 rbd_dev->parent_overlap = 0; 1528 } 1529 1530 /* 1531 * Parent image reference counting is used to determine when an 1532 * image's parent fields can be safely torn down--after there are no 1533 * more in-flight requests to the parent image. When the last 1534 * reference is dropped, cleaning them up is safe. 1535 */ 1536 static void rbd_dev_parent_put(struct rbd_device *rbd_dev) 1537 { 1538 int counter; 1539 1540 if (!rbd_dev->parent_spec) 1541 return; 1542 1543 counter = atomic_dec_return_safe(&rbd_dev->parent_ref); 1544 if (counter > 0) 1545 return; 1546 1547 /* Last reference; clean up parent data structures */ 1548 1549 if (!counter) 1550 rbd_dev_unparent(rbd_dev); 1551 else 1552 rbd_warn(rbd_dev, "parent reference underflow"); 1553 } 1554 1555 /* 1556 * If an image has a non-zero parent overlap, get a reference to its 1557 * parent. 1558 * 1559 * Returns true if the rbd device has a parent with a non-zero 1560 * overlap and a reference for it was successfully taken, or 1561 * false otherwise. 1562 */ 1563 static bool rbd_dev_parent_get(struct rbd_device *rbd_dev) 1564 { 1565 int counter = 0; 1566 1567 if (!rbd_dev->parent_spec) 1568 return false; 1569 1570 if (rbd_dev->parent_overlap) 1571 counter = atomic_inc_return_safe(&rbd_dev->parent_ref); 1572 1573 if (counter < 0) 1574 rbd_warn(rbd_dev, "parent reference overflow"); 1575 1576 return counter > 0; 1577 } 1578 1579 static void rbd_img_request_init(struct rbd_img_request *img_request, 1580 struct rbd_device *rbd_dev, 1581 enum obj_operation_type op_type) 1582 { 1583 memset(img_request, 0, sizeof(*img_request)); 1584 1585 img_request->rbd_dev = rbd_dev; 1586 img_request->op_type = op_type; 1587 1588 INIT_LIST_HEAD(&img_request->lock_item); 1589 INIT_LIST_HEAD(&img_request->object_extents); 1590 mutex_init(&img_request->state_mutex); 1591 } 1592 1593 static void rbd_img_capture_header(struct rbd_img_request *img_req) 1594 { 1595 struct rbd_device *rbd_dev = img_req->rbd_dev; 1596 1597 lockdep_assert_held(&rbd_dev->header_rwsem); 1598 1599 if (rbd_img_is_write(img_req)) 1600 img_req->snapc = ceph_get_snap_context(rbd_dev->header.snapc); 1601 else 1602 img_req->snap_id = rbd_dev->spec->snap_id; 1603 1604 if (rbd_dev_parent_get(rbd_dev)) 1605 img_request_layered_set(img_req); 1606 } 1607 1608 static void rbd_img_request_destroy(struct rbd_img_request *img_request) 1609 { 1610 struct rbd_obj_request *obj_request; 1611 struct rbd_obj_request *next_obj_request; 1612 1613 dout("%s: img %p\n", __func__, img_request); 1614 1615 WARN_ON(!list_empty(&img_request->lock_item)); 1616 for_each_obj_request_safe(img_request, obj_request, next_obj_request) 1617 rbd_img_obj_request_del(img_request, obj_request); 1618 1619 if (img_request_layered_test(img_request)) 1620 rbd_dev_parent_put(img_request->rbd_dev); 1621 1622 if (rbd_img_is_write(img_request)) 1623 ceph_put_snap_context(img_request->snapc); 1624 1625 if (test_bit(IMG_REQ_CHILD, &img_request->flags)) 1626 kmem_cache_free(rbd_img_request_cache, img_request); 1627 } 1628 1629 #define BITS_PER_OBJ 2 1630 #define OBJS_PER_BYTE (BITS_PER_BYTE / BITS_PER_OBJ) 1631 #define OBJ_MASK ((1 << BITS_PER_OBJ) - 1) 1632 1633 static void __rbd_object_map_index(struct rbd_device *rbd_dev, u64 objno, 1634 u64 *index, u8 *shift) 1635 { 1636 u32 off; 1637 1638 rbd_assert(objno < rbd_dev->object_map_size); 1639 *index = div_u64_rem(objno, OBJS_PER_BYTE, &off); 1640 *shift = (OBJS_PER_BYTE - off - 1) * BITS_PER_OBJ; 1641 } 1642 1643 static u8 __rbd_object_map_get(struct rbd_device *rbd_dev, u64 objno) 1644 { 1645 u64 index; 1646 u8 shift; 1647 1648 lockdep_assert_held(&rbd_dev->object_map_lock); 1649 __rbd_object_map_index(rbd_dev, objno, &index, &shift); 1650 return (rbd_dev->object_map[index] >> shift) & OBJ_MASK; 1651 } 1652 1653 static void __rbd_object_map_set(struct rbd_device *rbd_dev, u64 objno, u8 val) 1654 { 1655 u64 index; 1656 u8 shift; 1657 u8 *p; 1658 1659 lockdep_assert_held(&rbd_dev->object_map_lock); 1660 rbd_assert(!(val & ~OBJ_MASK)); 1661 1662 __rbd_object_map_index(rbd_dev, objno, &index, &shift); 1663 p = &rbd_dev->object_map[index]; 1664 *p = (*p & ~(OBJ_MASK << shift)) | (val << shift); 1665 } 1666 1667 static u8 rbd_object_map_get(struct rbd_device *rbd_dev, u64 objno) 1668 { 1669 u8 state; 1670 1671 spin_lock(&rbd_dev->object_map_lock); 1672 state = __rbd_object_map_get(rbd_dev, objno); 1673 spin_unlock(&rbd_dev->object_map_lock); 1674 return state; 1675 } 1676 1677 static bool use_object_map(struct rbd_device *rbd_dev) 1678 { 1679 /* 1680 * An image mapped read-only can't use the object map -- it isn't 1681 * loaded because the header lock isn't acquired. Someone else can 1682 * write to the image and update the object map behind our back. 1683 * 1684 * A snapshot can't be written to, so using the object map is always 1685 * safe. 1686 */ 1687 if (!rbd_is_snap(rbd_dev) && rbd_is_ro(rbd_dev)) 1688 return false; 1689 1690 return ((rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP) && 1691 !(rbd_dev->object_map_flags & RBD_FLAG_OBJECT_MAP_INVALID)); 1692 } 1693 1694 static bool rbd_object_map_may_exist(struct rbd_device *rbd_dev, u64 objno) 1695 { 1696 u8 state; 1697 1698 /* fall back to default logic if object map is disabled or invalid */ 1699 if (!use_object_map(rbd_dev)) 1700 return true; 1701 1702 state = rbd_object_map_get(rbd_dev, objno); 1703 return state != OBJECT_NONEXISTENT; 1704 } 1705 1706 static void rbd_object_map_name(struct rbd_device *rbd_dev, u64 snap_id, 1707 struct ceph_object_id *oid) 1708 { 1709 if (snap_id == CEPH_NOSNAP) 1710 ceph_oid_printf(oid, "%s%s", RBD_OBJECT_MAP_PREFIX, 1711 rbd_dev->spec->image_id); 1712 else 1713 ceph_oid_printf(oid, "%s%s.%016llx", RBD_OBJECT_MAP_PREFIX, 1714 rbd_dev->spec->image_id, snap_id); 1715 } 1716 1717 static int rbd_object_map_lock(struct rbd_device *rbd_dev) 1718 { 1719 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 1720 CEPH_DEFINE_OID_ONSTACK(oid); 1721 u8 lock_type; 1722 char *lock_tag; 1723 struct ceph_locker *lockers; 1724 u32 num_lockers; 1725 bool broke_lock = false; 1726 int ret; 1727 1728 rbd_object_map_name(rbd_dev, CEPH_NOSNAP, &oid); 1729 1730 again: 1731 ret = ceph_cls_lock(osdc, &oid, &rbd_dev->header_oloc, RBD_LOCK_NAME, 1732 CEPH_CLS_LOCK_EXCLUSIVE, "", "", "", 0); 1733 if (ret != -EBUSY || broke_lock) { 1734 if (ret == -EEXIST) 1735 ret = 0; /* already locked by myself */ 1736 if (ret) 1737 rbd_warn(rbd_dev, "failed to lock object map: %d", ret); 1738 return ret; 1739 } 1740 1741 ret = ceph_cls_lock_info(osdc, &oid, &rbd_dev->header_oloc, 1742 RBD_LOCK_NAME, &lock_type, &lock_tag, 1743 &lockers, &num_lockers); 1744 if (ret) { 1745 if (ret == -ENOENT) 1746 goto again; 1747 1748 rbd_warn(rbd_dev, "failed to get object map lockers: %d", ret); 1749 return ret; 1750 } 1751 1752 kfree(lock_tag); 1753 if (num_lockers == 0) 1754 goto again; 1755 1756 rbd_warn(rbd_dev, "breaking object map lock owned by %s%llu", 1757 ENTITY_NAME(lockers[0].id.name)); 1758 1759 ret = ceph_cls_break_lock(osdc, &oid, &rbd_dev->header_oloc, 1760 RBD_LOCK_NAME, lockers[0].id.cookie, 1761 &lockers[0].id.name); 1762 ceph_free_lockers(lockers, num_lockers); 1763 if (ret) { 1764 if (ret == -ENOENT) 1765 goto again; 1766 1767 rbd_warn(rbd_dev, "failed to break object map lock: %d", ret); 1768 return ret; 1769 } 1770 1771 broke_lock = true; 1772 goto again; 1773 } 1774 1775 static void rbd_object_map_unlock(struct rbd_device *rbd_dev) 1776 { 1777 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 1778 CEPH_DEFINE_OID_ONSTACK(oid); 1779 int ret; 1780 1781 rbd_object_map_name(rbd_dev, CEPH_NOSNAP, &oid); 1782 1783 ret = ceph_cls_unlock(osdc, &oid, &rbd_dev->header_oloc, RBD_LOCK_NAME, 1784 ""); 1785 if (ret && ret != -ENOENT) 1786 rbd_warn(rbd_dev, "failed to unlock object map: %d", ret); 1787 } 1788 1789 static int decode_object_map_header(void **p, void *end, u64 *object_map_size) 1790 { 1791 u8 struct_v; 1792 u32 struct_len; 1793 u32 header_len; 1794 void *header_end; 1795 int ret; 1796 1797 ceph_decode_32_safe(p, end, header_len, e_inval); 1798 header_end = *p + header_len; 1799 1800 ret = ceph_start_decoding(p, end, 1, "BitVector header", &struct_v, 1801 &struct_len); 1802 if (ret) 1803 return ret; 1804 1805 ceph_decode_64_safe(p, end, *object_map_size, e_inval); 1806 1807 *p = header_end; 1808 return 0; 1809 1810 e_inval: 1811 return -EINVAL; 1812 } 1813 1814 static int __rbd_object_map_load(struct rbd_device *rbd_dev) 1815 { 1816 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 1817 CEPH_DEFINE_OID_ONSTACK(oid); 1818 struct page **pages; 1819 void *p, *end; 1820 size_t reply_len; 1821 u64 num_objects; 1822 u64 object_map_bytes; 1823 u64 object_map_size; 1824 int num_pages; 1825 int ret; 1826 1827 rbd_assert(!rbd_dev->object_map && !rbd_dev->object_map_size); 1828 1829 num_objects = ceph_get_num_objects(&rbd_dev->layout, 1830 rbd_dev->mapping.size); 1831 object_map_bytes = DIV_ROUND_UP_ULL(num_objects * BITS_PER_OBJ, 1832 BITS_PER_BYTE); 1833 num_pages = calc_pages_for(0, object_map_bytes) + 1; 1834 pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL); 1835 if (IS_ERR(pages)) 1836 return PTR_ERR(pages); 1837 1838 reply_len = num_pages * PAGE_SIZE; 1839 rbd_object_map_name(rbd_dev, rbd_dev->spec->snap_id, &oid); 1840 ret = ceph_osdc_call(osdc, &oid, &rbd_dev->header_oloc, 1841 "rbd", "object_map_load", CEPH_OSD_FLAG_READ, 1842 NULL, 0, pages, &reply_len); 1843 if (ret) 1844 goto out; 1845 1846 p = page_address(pages[0]); 1847 end = p + min(reply_len, (size_t)PAGE_SIZE); 1848 ret = decode_object_map_header(&p, end, &object_map_size); 1849 if (ret) 1850 goto out; 1851 1852 if (object_map_size != num_objects) { 1853 rbd_warn(rbd_dev, "object map size mismatch: %llu vs %llu", 1854 object_map_size, num_objects); 1855 ret = -EINVAL; 1856 goto out; 1857 } 1858 1859 if (offset_in_page(p) + object_map_bytes > reply_len) { 1860 ret = -EINVAL; 1861 goto out; 1862 } 1863 1864 rbd_dev->object_map = kvmalloc(object_map_bytes, GFP_KERNEL); 1865 if (!rbd_dev->object_map) { 1866 ret = -ENOMEM; 1867 goto out; 1868 } 1869 1870 rbd_dev->object_map_size = object_map_size; 1871 ceph_copy_from_page_vector(pages, rbd_dev->object_map, 1872 offset_in_page(p), object_map_bytes); 1873 1874 out: 1875 ceph_release_page_vector(pages, num_pages); 1876 return ret; 1877 } 1878 1879 static void rbd_object_map_free(struct rbd_device *rbd_dev) 1880 { 1881 kvfree(rbd_dev->object_map); 1882 rbd_dev->object_map = NULL; 1883 rbd_dev->object_map_size = 0; 1884 } 1885 1886 static int rbd_object_map_load(struct rbd_device *rbd_dev) 1887 { 1888 int ret; 1889 1890 ret = __rbd_object_map_load(rbd_dev); 1891 if (ret) 1892 return ret; 1893 1894 ret = rbd_dev_v2_get_flags(rbd_dev); 1895 if (ret) { 1896 rbd_object_map_free(rbd_dev); 1897 return ret; 1898 } 1899 1900 if (rbd_dev->object_map_flags & RBD_FLAG_OBJECT_MAP_INVALID) 1901 rbd_warn(rbd_dev, "object map is invalid"); 1902 1903 return 0; 1904 } 1905 1906 static int rbd_object_map_open(struct rbd_device *rbd_dev) 1907 { 1908 int ret; 1909 1910 ret = rbd_object_map_lock(rbd_dev); 1911 if (ret) 1912 return ret; 1913 1914 ret = rbd_object_map_load(rbd_dev); 1915 if (ret) { 1916 rbd_object_map_unlock(rbd_dev); 1917 return ret; 1918 } 1919 1920 return 0; 1921 } 1922 1923 static void rbd_object_map_close(struct rbd_device *rbd_dev) 1924 { 1925 rbd_object_map_free(rbd_dev); 1926 rbd_object_map_unlock(rbd_dev); 1927 } 1928 1929 /* 1930 * This function needs snap_id (or more precisely just something to 1931 * distinguish between HEAD and snapshot object maps), new_state and 1932 * current_state that were passed to rbd_object_map_update(). 1933 * 1934 * To avoid allocating and stashing a context we piggyback on the OSD 1935 * request. A HEAD update has two ops (assert_locked). For new_state 1936 * and current_state we decode our own object_map_update op, encoded in 1937 * rbd_cls_object_map_update(). 1938 */ 1939 static int rbd_object_map_update_finish(struct rbd_obj_request *obj_req, 1940 struct ceph_osd_request *osd_req) 1941 { 1942 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 1943 struct ceph_osd_data *osd_data; 1944 u64 objno; 1945 u8 state, new_state, current_state; 1946 bool has_current_state; 1947 void *p; 1948 1949 if (osd_req->r_result) 1950 return osd_req->r_result; 1951 1952 /* 1953 * Nothing to do for a snapshot object map. 1954 */ 1955 if (osd_req->r_num_ops == 1) 1956 return 0; 1957 1958 /* 1959 * Update in-memory HEAD object map. 1960 */ 1961 rbd_assert(osd_req->r_num_ops == 2); 1962 osd_data = osd_req_op_data(osd_req, 1, cls, request_data); 1963 rbd_assert(osd_data->type == CEPH_OSD_DATA_TYPE_PAGES); 1964 1965 p = page_address(osd_data->pages[0]); 1966 objno = ceph_decode_64(&p); 1967 rbd_assert(objno == obj_req->ex.oe_objno); 1968 rbd_assert(ceph_decode_64(&p) == objno + 1); 1969 new_state = ceph_decode_8(&p); 1970 has_current_state = ceph_decode_8(&p); 1971 if (has_current_state) 1972 current_state = ceph_decode_8(&p); 1973 1974 spin_lock(&rbd_dev->object_map_lock); 1975 state = __rbd_object_map_get(rbd_dev, objno); 1976 if (!has_current_state || current_state == state || 1977 (current_state == OBJECT_EXISTS && state == OBJECT_EXISTS_CLEAN)) 1978 __rbd_object_map_set(rbd_dev, objno, new_state); 1979 spin_unlock(&rbd_dev->object_map_lock); 1980 1981 return 0; 1982 } 1983 1984 static void rbd_object_map_callback(struct ceph_osd_request *osd_req) 1985 { 1986 struct rbd_obj_request *obj_req = osd_req->r_priv; 1987 int result; 1988 1989 dout("%s osd_req %p result %d for obj_req %p\n", __func__, osd_req, 1990 osd_req->r_result, obj_req); 1991 1992 result = rbd_object_map_update_finish(obj_req, osd_req); 1993 rbd_obj_handle_request(obj_req, result); 1994 } 1995 1996 static bool update_needed(struct rbd_device *rbd_dev, u64 objno, u8 new_state) 1997 { 1998 u8 state = rbd_object_map_get(rbd_dev, objno); 1999 2000 if (state == new_state || 2001 (new_state == OBJECT_PENDING && state == OBJECT_NONEXISTENT) || 2002 (new_state == OBJECT_NONEXISTENT && state != OBJECT_PENDING)) 2003 return false; 2004 2005 return true; 2006 } 2007 2008 static int rbd_cls_object_map_update(struct ceph_osd_request *req, 2009 int which, u64 objno, u8 new_state, 2010 const u8 *current_state) 2011 { 2012 struct page **pages; 2013 void *p, *start; 2014 int ret; 2015 2016 ret = osd_req_op_cls_init(req, which, "rbd", "object_map_update"); 2017 if (ret) 2018 return ret; 2019 2020 pages = ceph_alloc_page_vector(1, GFP_NOIO); 2021 if (IS_ERR(pages)) 2022 return PTR_ERR(pages); 2023 2024 p = start = page_address(pages[0]); 2025 ceph_encode_64(&p, objno); 2026 ceph_encode_64(&p, objno + 1); 2027 ceph_encode_8(&p, new_state); 2028 if (current_state) { 2029 ceph_encode_8(&p, 1); 2030 ceph_encode_8(&p, *current_state); 2031 } else { 2032 ceph_encode_8(&p, 0); 2033 } 2034 2035 osd_req_op_cls_request_data_pages(req, which, pages, p - start, 0, 2036 false, true); 2037 return 0; 2038 } 2039 2040 /* 2041 * Return: 2042 * 0 - object map update sent 2043 * 1 - object map update isn't needed 2044 * <0 - error 2045 */ 2046 static int rbd_object_map_update(struct rbd_obj_request *obj_req, u64 snap_id, 2047 u8 new_state, const u8 *current_state) 2048 { 2049 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 2050 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 2051 struct ceph_osd_request *req; 2052 int num_ops = 1; 2053 int which = 0; 2054 int ret; 2055 2056 if (snap_id == CEPH_NOSNAP) { 2057 if (!update_needed(rbd_dev, obj_req->ex.oe_objno, new_state)) 2058 return 1; 2059 2060 num_ops++; /* assert_locked */ 2061 } 2062 2063 req = ceph_osdc_alloc_request(osdc, NULL, num_ops, false, GFP_NOIO); 2064 if (!req) 2065 return -ENOMEM; 2066 2067 list_add_tail(&req->r_private_item, &obj_req->osd_reqs); 2068 req->r_callback = rbd_object_map_callback; 2069 req->r_priv = obj_req; 2070 2071 rbd_object_map_name(rbd_dev, snap_id, &req->r_base_oid); 2072 ceph_oloc_copy(&req->r_base_oloc, &rbd_dev->header_oloc); 2073 req->r_flags = CEPH_OSD_FLAG_WRITE; 2074 ktime_get_real_ts64(&req->r_mtime); 2075 2076 if (snap_id == CEPH_NOSNAP) { 2077 /* 2078 * Protect against possible race conditions during lock 2079 * ownership transitions. 2080 */ 2081 ret = ceph_cls_assert_locked(req, which++, RBD_LOCK_NAME, 2082 CEPH_CLS_LOCK_EXCLUSIVE, "", ""); 2083 if (ret) 2084 return ret; 2085 } 2086 2087 ret = rbd_cls_object_map_update(req, which, obj_req->ex.oe_objno, 2088 new_state, current_state); 2089 if (ret) 2090 return ret; 2091 2092 ret = ceph_osdc_alloc_messages(req, GFP_NOIO); 2093 if (ret) 2094 return ret; 2095 2096 ceph_osdc_start_request(osdc, req, false); 2097 return 0; 2098 } 2099 2100 static void prune_extents(struct ceph_file_extent *img_extents, 2101 u32 *num_img_extents, u64 overlap) 2102 { 2103 u32 cnt = *num_img_extents; 2104 2105 /* drop extents completely beyond the overlap */ 2106 while (cnt && img_extents[cnt - 1].fe_off >= overlap) 2107 cnt--; 2108 2109 if (cnt) { 2110 struct ceph_file_extent *ex = &img_extents[cnt - 1]; 2111 2112 /* trim final overlapping extent */ 2113 if (ex->fe_off + ex->fe_len > overlap) 2114 ex->fe_len = overlap - ex->fe_off; 2115 } 2116 2117 *num_img_extents = cnt; 2118 } 2119 2120 /* 2121 * Determine the byte range(s) covered by either just the object extent 2122 * or the entire object in the parent image. 2123 */ 2124 static int rbd_obj_calc_img_extents(struct rbd_obj_request *obj_req, 2125 bool entire) 2126 { 2127 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 2128 int ret; 2129 2130 if (!rbd_dev->parent_overlap) 2131 return 0; 2132 2133 ret = ceph_extent_to_file(&rbd_dev->layout, obj_req->ex.oe_objno, 2134 entire ? 0 : obj_req->ex.oe_off, 2135 entire ? rbd_dev->layout.object_size : 2136 obj_req->ex.oe_len, 2137 &obj_req->img_extents, 2138 &obj_req->num_img_extents); 2139 if (ret) 2140 return ret; 2141 2142 prune_extents(obj_req->img_extents, &obj_req->num_img_extents, 2143 rbd_dev->parent_overlap); 2144 return 0; 2145 } 2146 2147 static void rbd_osd_setup_data(struct ceph_osd_request *osd_req, int which) 2148 { 2149 struct rbd_obj_request *obj_req = osd_req->r_priv; 2150 2151 switch (obj_req->img_request->data_type) { 2152 case OBJ_REQUEST_BIO: 2153 osd_req_op_extent_osd_data_bio(osd_req, which, 2154 &obj_req->bio_pos, 2155 obj_req->ex.oe_len); 2156 break; 2157 case OBJ_REQUEST_BVECS: 2158 case OBJ_REQUEST_OWN_BVECS: 2159 rbd_assert(obj_req->bvec_pos.iter.bi_size == 2160 obj_req->ex.oe_len); 2161 rbd_assert(obj_req->bvec_idx == obj_req->bvec_count); 2162 osd_req_op_extent_osd_data_bvec_pos(osd_req, which, 2163 &obj_req->bvec_pos); 2164 break; 2165 default: 2166 BUG(); 2167 } 2168 } 2169 2170 static int rbd_osd_setup_stat(struct ceph_osd_request *osd_req, int which) 2171 { 2172 struct page **pages; 2173 2174 /* 2175 * The response data for a STAT call consists of: 2176 * le64 length; 2177 * struct { 2178 * le32 tv_sec; 2179 * le32 tv_nsec; 2180 * } mtime; 2181 */ 2182 pages = ceph_alloc_page_vector(1, GFP_NOIO); 2183 if (IS_ERR(pages)) 2184 return PTR_ERR(pages); 2185 2186 osd_req_op_init(osd_req, which, CEPH_OSD_OP_STAT, 0); 2187 osd_req_op_raw_data_in_pages(osd_req, which, pages, 2188 8 + sizeof(struct ceph_timespec), 2189 0, false, true); 2190 return 0; 2191 } 2192 2193 static int rbd_osd_setup_copyup(struct ceph_osd_request *osd_req, int which, 2194 u32 bytes) 2195 { 2196 struct rbd_obj_request *obj_req = osd_req->r_priv; 2197 int ret; 2198 2199 ret = osd_req_op_cls_init(osd_req, which, "rbd", "copyup"); 2200 if (ret) 2201 return ret; 2202 2203 osd_req_op_cls_request_data_bvecs(osd_req, which, obj_req->copyup_bvecs, 2204 obj_req->copyup_bvec_count, bytes); 2205 return 0; 2206 } 2207 2208 static int rbd_obj_init_read(struct rbd_obj_request *obj_req) 2209 { 2210 obj_req->read_state = RBD_OBJ_READ_START; 2211 return 0; 2212 } 2213 2214 static void __rbd_osd_setup_write_ops(struct ceph_osd_request *osd_req, 2215 int which) 2216 { 2217 struct rbd_obj_request *obj_req = osd_req->r_priv; 2218 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 2219 u16 opcode; 2220 2221 if (!use_object_map(rbd_dev) || 2222 !(obj_req->flags & RBD_OBJ_FLAG_MAY_EXIST)) { 2223 osd_req_op_alloc_hint_init(osd_req, which++, 2224 rbd_dev->layout.object_size, 2225 rbd_dev->layout.object_size, 2226 rbd_dev->opts->alloc_hint_flags); 2227 } 2228 2229 if (rbd_obj_is_entire(obj_req)) 2230 opcode = CEPH_OSD_OP_WRITEFULL; 2231 else 2232 opcode = CEPH_OSD_OP_WRITE; 2233 2234 osd_req_op_extent_init(osd_req, which, opcode, 2235 obj_req->ex.oe_off, obj_req->ex.oe_len, 0, 0); 2236 rbd_osd_setup_data(osd_req, which); 2237 } 2238 2239 static int rbd_obj_init_write(struct rbd_obj_request *obj_req) 2240 { 2241 int ret; 2242 2243 /* reverse map the entire object onto the parent */ 2244 ret = rbd_obj_calc_img_extents(obj_req, true); 2245 if (ret) 2246 return ret; 2247 2248 if (rbd_obj_copyup_enabled(obj_req)) 2249 obj_req->flags |= RBD_OBJ_FLAG_COPYUP_ENABLED; 2250 2251 obj_req->write_state = RBD_OBJ_WRITE_START; 2252 return 0; 2253 } 2254 2255 static u16 truncate_or_zero_opcode(struct rbd_obj_request *obj_req) 2256 { 2257 return rbd_obj_is_tail(obj_req) ? CEPH_OSD_OP_TRUNCATE : 2258 CEPH_OSD_OP_ZERO; 2259 } 2260 2261 static void __rbd_osd_setup_discard_ops(struct ceph_osd_request *osd_req, 2262 int which) 2263 { 2264 struct rbd_obj_request *obj_req = osd_req->r_priv; 2265 2266 if (rbd_obj_is_entire(obj_req) && !obj_req->num_img_extents) { 2267 rbd_assert(obj_req->flags & RBD_OBJ_FLAG_DELETION); 2268 osd_req_op_init(osd_req, which, CEPH_OSD_OP_DELETE, 0); 2269 } else { 2270 osd_req_op_extent_init(osd_req, which, 2271 truncate_or_zero_opcode(obj_req), 2272 obj_req->ex.oe_off, obj_req->ex.oe_len, 2273 0, 0); 2274 } 2275 } 2276 2277 static int rbd_obj_init_discard(struct rbd_obj_request *obj_req) 2278 { 2279 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 2280 u64 off, next_off; 2281 int ret; 2282 2283 /* 2284 * Align the range to alloc_size boundary and punt on discards 2285 * that are too small to free up any space. 2286 * 2287 * alloc_size == object_size && is_tail() is a special case for 2288 * filestore with filestore_punch_hole = false, needed to allow 2289 * truncate (in addition to delete). 2290 */ 2291 if (rbd_dev->opts->alloc_size != rbd_dev->layout.object_size || 2292 !rbd_obj_is_tail(obj_req)) { 2293 off = round_up(obj_req->ex.oe_off, rbd_dev->opts->alloc_size); 2294 next_off = round_down(obj_req->ex.oe_off + obj_req->ex.oe_len, 2295 rbd_dev->opts->alloc_size); 2296 if (off >= next_off) 2297 return 1; 2298 2299 dout("%s %p %llu~%llu -> %llu~%llu\n", __func__, 2300 obj_req, obj_req->ex.oe_off, obj_req->ex.oe_len, 2301 off, next_off - off); 2302 obj_req->ex.oe_off = off; 2303 obj_req->ex.oe_len = next_off - off; 2304 } 2305 2306 /* reverse map the entire object onto the parent */ 2307 ret = rbd_obj_calc_img_extents(obj_req, true); 2308 if (ret) 2309 return ret; 2310 2311 obj_req->flags |= RBD_OBJ_FLAG_NOOP_FOR_NONEXISTENT; 2312 if (rbd_obj_is_entire(obj_req) && !obj_req->num_img_extents) 2313 obj_req->flags |= RBD_OBJ_FLAG_DELETION; 2314 2315 obj_req->write_state = RBD_OBJ_WRITE_START; 2316 return 0; 2317 } 2318 2319 static void __rbd_osd_setup_zeroout_ops(struct ceph_osd_request *osd_req, 2320 int which) 2321 { 2322 struct rbd_obj_request *obj_req = osd_req->r_priv; 2323 u16 opcode; 2324 2325 if (rbd_obj_is_entire(obj_req)) { 2326 if (obj_req->num_img_extents) { 2327 if (!(obj_req->flags & RBD_OBJ_FLAG_COPYUP_ENABLED)) 2328 osd_req_op_init(osd_req, which++, 2329 CEPH_OSD_OP_CREATE, 0); 2330 opcode = CEPH_OSD_OP_TRUNCATE; 2331 } else { 2332 rbd_assert(obj_req->flags & RBD_OBJ_FLAG_DELETION); 2333 osd_req_op_init(osd_req, which++, 2334 CEPH_OSD_OP_DELETE, 0); 2335 opcode = 0; 2336 } 2337 } else { 2338 opcode = truncate_or_zero_opcode(obj_req); 2339 } 2340 2341 if (opcode) 2342 osd_req_op_extent_init(osd_req, which, opcode, 2343 obj_req->ex.oe_off, obj_req->ex.oe_len, 2344 0, 0); 2345 } 2346 2347 static int rbd_obj_init_zeroout(struct rbd_obj_request *obj_req) 2348 { 2349 int ret; 2350 2351 /* reverse map the entire object onto the parent */ 2352 ret = rbd_obj_calc_img_extents(obj_req, true); 2353 if (ret) 2354 return ret; 2355 2356 if (rbd_obj_copyup_enabled(obj_req)) 2357 obj_req->flags |= RBD_OBJ_FLAG_COPYUP_ENABLED; 2358 if (!obj_req->num_img_extents) { 2359 obj_req->flags |= RBD_OBJ_FLAG_NOOP_FOR_NONEXISTENT; 2360 if (rbd_obj_is_entire(obj_req)) 2361 obj_req->flags |= RBD_OBJ_FLAG_DELETION; 2362 } 2363 2364 obj_req->write_state = RBD_OBJ_WRITE_START; 2365 return 0; 2366 } 2367 2368 static int count_write_ops(struct rbd_obj_request *obj_req) 2369 { 2370 struct rbd_img_request *img_req = obj_req->img_request; 2371 2372 switch (img_req->op_type) { 2373 case OBJ_OP_WRITE: 2374 if (!use_object_map(img_req->rbd_dev) || 2375 !(obj_req->flags & RBD_OBJ_FLAG_MAY_EXIST)) 2376 return 2; /* setallochint + write/writefull */ 2377 2378 return 1; /* write/writefull */ 2379 case OBJ_OP_DISCARD: 2380 return 1; /* delete/truncate/zero */ 2381 case OBJ_OP_ZEROOUT: 2382 if (rbd_obj_is_entire(obj_req) && obj_req->num_img_extents && 2383 !(obj_req->flags & RBD_OBJ_FLAG_COPYUP_ENABLED)) 2384 return 2; /* create + truncate */ 2385 2386 return 1; /* delete/truncate/zero */ 2387 default: 2388 BUG(); 2389 } 2390 } 2391 2392 static void rbd_osd_setup_write_ops(struct ceph_osd_request *osd_req, 2393 int which) 2394 { 2395 struct rbd_obj_request *obj_req = osd_req->r_priv; 2396 2397 switch (obj_req->img_request->op_type) { 2398 case OBJ_OP_WRITE: 2399 __rbd_osd_setup_write_ops(osd_req, which); 2400 break; 2401 case OBJ_OP_DISCARD: 2402 __rbd_osd_setup_discard_ops(osd_req, which); 2403 break; 2404 case OBJ_OP_ZEROOUT: 2405 __rbd_osd_setup_zeroout_ops(osd_req, which); 2406 break; 2407 default: 2408 BUG(); 2409 } 2410 } 2411 2412 /* 2413 * Prune the list of object requests (adjust offset and/or length, drop 2414 * redundant requests). Prepare object request state machines and image 2415 * request state machine for execution. 2416 */ 2417 static int __rbd_img_fill_request(struct rbd_img_request *img_req) 2418 { 2419 struct rbd_obj_request *obj_req, *next_obj_req; 2420 int ret; 2421 2422 for_each_obj_request_safe(img_req, obj_req, next_obj_req) { 2423 switch (img_req->op_type) { 2424 case OBJ_OP_READ: 2425 ret = rbd_obj_init_read(obj_req); 2426 break; 2427 case OBJ_OP_WRITE: 2428 ret = rbd_obj_init_write(obj_req); 2429 break; 2430 case OBJ_OP_DISCARD: 2431 ret = rbd_obj_init_discard(obj_req); 2432 break; 2433 case OBJ_OP_ZEROOUT: 2434 ret = rbd_obj_init_zeroout(obj_req); 2435 break; 2436 default: 2437 BUG(); 2438 } 2439 if (ret < 0) 2440 return ret; 2441 if (ret > 0) { 2442 rbd_img_obj_request_del(img_req, obj_req); 2443 continue; 2444 } 2445 } 2446 2447 img_req->state = RBD_IMG_START; 2448 return 0; 2449 } 2450 2451 union rbd_img_fill_iter { 2452 struct ceph_bio_iter bio_iter; 2453 struct ceph_bvec_iter bvec_iter; 2454 }; 2455 2456 struct rbd_img_fill_ctx { 2457 enum obj_request_type pos_type; 2458 union rbd_img_fill_iter *pos; 2459 union rbd_img_fill_iter iter; 2460 ceph_object_extent_fn_t set_pos_fn; 2461 ceph_object_extent_fn_t count_fn; 2462 ceph_object_extent_fn_t copy_fn; 2463 }; 2464 2465 static struct ceph_object_extent *alloc_object_extent(void *arg) 2466 { 2467 struct rbd_img_request *img_req = arg; 2468 struct rbd_obj_request *obj_req; 2469 2470 obj_req = rbd_obj_request_create(); 2471 if (!obj_req) 2472 return NULL; 2473 2474 rbd_img_obj_request_add(img_req, obj_req); 2475 return &obj_req->ex; 2476 } 2477 2478 /* 2479 * While su != os && sc == 1 is technically not fancy (it's the same 2480 * layout as su == os && sc == 1), we can't use the nocopy path for it 2481 * because ->set_pos_fn() should be called only once per object. 2482 * ceph_file_to_extents() invokes action_fn once per stripe unit, so 2483 * treat su != os && sc == 1 as fancy. 2484 */ 2485 static bool rbd_layout_is_fancy(struct ceph_file_layout *l) 2486 { 2487 return l->stripe_unit != l->object_size; 2488 } 2489 2490 static int rbd_img_fill_request_nocopy(struct rbd_img_request *img_req, 2491 struct ceph_file_extent *img_extents, 2492 u32 num_img_extents, 2493 struct rbd_img_fill_ctx *fctx) 2494 { 2495 u32 i; 2496 int ret; 2497 2498 img_req->data_type = fctx->pos_type; 2499 2500 /* 2501 * Create object requests and set each object request's starting 2502 * position in the provided bio (list) or bio_vec array. 2503 */ 2504 fctx->iter = *fctx->pos; 2505 for (i = 0; i < num_img_extents; i++) { 2506 ret = ceph_file_to_extents(&img_req->rbd_dev->layout, 2507 img_extents[i].fe_off, 2508 img_extents[i].fe_len, 2509 &img_req->object_extents, 2510 alloc_object_extent, img_req, 2511 fctx->set_pos_fn, &fctx->iter); 2512 if (ret) 2513 return ret; 2514 } 2515 2516 return __rbd_img_fill_request(img_req); 2517 } 2518 2519 /* 2520 * Map a list of image extents to a list of object extents, create the 2521 * corresponding object requests (normally each to a different object, 2522 * but not always) and add them to @img_req. For each object request, 2523 * set up its data descriptor to point to the corresponding chunk(s) of 2524 * @fctx->pos data buffer. 2525 * 2526 * Because ceph_file_to_extents() will merge adjacent object extents 2527 * together, each object request's data descriptor may point to multiple 2528 * different chunks of @fctx->pos data buffer. 2529 * 2530 * @fctx->pos data buffer is assumed to be large enough. 2531 */ 2532 static int rbd_img_fill_request(struct rbd_img_request *img_req, 2533 struct ceph_file_extent *img_extents, 2534 u32 num_img_extents, 2535 struct rbd_img_fill_ctx *fctx) 2536 { 2537 struct rbd_device *rbd_dev = img_req->rbd_dev; 2538 struct rbd_obj_request *obj_req; 2539 u32 i; 2540 int ret; 2541 2542 if (fctx->pos_type == OBJ_REQUEST_NODATA || 2543 !rbd_layout_is_fancy(&rbd_dev->layout)) 2544 return rbd_img_fill_request_nocopy(img_req, img_extents, 2545 num_img_extents, fctx); 2546 2547 img_req->data_type = OBJ_REQUEST_OWN_BVECS; 2548 2549 /* 2550 * Create object requests and determine ->bvec_count for each object 2551 * request. Note that ->bvec_count sum over all object requests may 2552 * be greater than the number of bio_vecs in the provided bio (list) 2553 * or bio_vec array because when mapped, those bio_vecs can straddle 2554 * stripe unit boundaries. 2555 */ 2556 fctx->iter = *fctx->pos; 2557 for (i = 0; i < num_img_extents; i++) { 2558 ret = ceph_file_to_extents(&rbd_dev->layout, 2559 img_extents[i].fe_off, 2560 img_extents[i].fe_len, 2561 &img_req->object_extents, 2562 alloc_object_extent, img_req, 2563 fctx->count_fn, &fctx->iter); 2564 if (ret) 2565 return ret; 2566 } 2567 2568 for_each_obj_request(img_req, obj_req) { 2569 obj_req->bvec_pos.bvecs = kmalloc_array(obj_req->bvec_count, 2570 sizeof(*obj_req->bvec_pos.bvecs), 2571 GFP_NOIO); 2572 if (!obj_req->bvec_pos.bvecs) 2573 return -ENOMEM; 2574 } 2575 2576 /* 2577 * Fill in each object request's private bio_vec array, splitting and 2578 * rearranging the provided bio_vecs in stripe unit chunks as needed. 2579 */ 2580 fctx->iter = *fctx->pos; 2581 for (i = 0; i < num_img_extents; i++) { 2582 ret = ceph_iterate_extents(&rbd_dev->layout, 2583 img_extents[i].fe_off, 2584 img_extents[i].fe_len, 2585 &img_req->object_extents, 2586 fctx->copy_fn, &fctx->iter); 2587 if (ret) 2588 return ret; 2589 } 2590 2591 return __rbd_img_fill_request(img_req); 2592 } 2593 2594 static int rbd_img_fill_nodata(struct rbd_img_request *img_req, 2595 u64 off, u64 len) 2596 { 2597 struct ceph_file_extent ex = { off, len }; 2598 union rbd_img_fill_iter dummy = {}; 2599 struct rbd_img_fill_ctx fctx = { 2600 .pos_type = OBJ_REQUEST_NODATA, 2601 .pos = &dummy, 2602 }; 2603 2604 return rbd_img_fill_request(img_req, &ex, 1, &fctx); 2605 } 2606 2607 static void set_bio_pos(struct ceph_object_extent *ex, u32 bytes, void *arg) 2608 { 2609 struct rbd_obj_request *obj_req = 2610 container_of(ex, struct rbd_obj_request, ex); 2611 struct ceph_bio_iter *it = arg; 2612 2613 dout("%s objno %llu bytes %u\n", __func__, ex->oe_objno, bytes); 2614 obj_req->bio_pos = *it; 2615 ceph_bio_iter_advance(it, bytes); 2616 } 2617 2618 static void count_bio_bvecs(struct ceph_object_extent *ex, u32 bytes, void *arg) 2619 { 2620 struct rbd_obj_request *obj_req = 2621 container_of(ex, struct rbd_obj_request, ex); 2622 struct ceph_bio_iter *it = arg; 2623 2624 dout("%s objno %llu bytes %u\n", __func__, ex->oe_objno, bytes); 2625 ceph_bio_iter_advance_step(it, bytes, ({ 2626 obj_req->bvec_count++; 2627 })); 2628 2629 } 2630 2631 static void copy_bio_bvecs(struct ceph_object_extent *ex, u32 bytes, void *arg) 2632 { 2633 struct rbd_obj_request *obj_req = 2634 container_of(ex, struct rbd_obj_request, ex); 2635 struct ceph_bio_iter *it = arg; 2636 2637 dout("%s objno %llu bytes %u\n", __func__, ex->oe_objno, bytes); 2638 ceph_bio_iter_advance_step(it, bytes, ({ 2639 obj_req->bvec_pos.bvecs[obj_req->bvec_idx++] = bv; 2640 obj_req->bvec_pos.iter.bi_size += bv.bv_len; 2641 })); 2642 } 2643 2644 static int __rbd_img_fill_from_bio(struct rbd_img_request *img_req, 2645 struct ceph_file_extent *img_extents, 2646 u32 num_img_extents, 2647 struct ceph_bio_iter *bio_pos) 2648 { 2649 struct rbd_img_fill_ctx fctx = { 2650 .pos_type = OBJ_REQUEST_BIO, 2651 .pos = (union rbd_img_fill_iter *)bio_pos, 2652 .set_pos_fn = set_bio_pos, 2653 .count_fn = count_bio_bvecs, 2654 .copy_fn = copy_bio_bvecs, 2655 }; 2656 2657 return rbd_img_fill_request(img_req, img_extents, num_img_extents, 2658 &fctx); 2659 } 2660 2661 static int rbd_img_fill_from_bio(struct rbd_img_request *img_req, 2662 u64 off, u64 len, struct bio *bio) 2663 { 2664 struct ceph_file_extent ex = { off, len }; 2665 struct ceph_bio_iter it = { .bio = bio, .iter = bio->bi_iter }; 2666 2667 return __rbd_img_fill_from_bio(img_req, &ex, 1, &it); 2668 } 2669 2670 static void set_bvec_pos(struct ceph_object_extent *ex, u32 bytes, void *arg) 2671 { 2672 struct rbd_obj_request *obj_req = 2673 container_of(ex, struct rbd_obj_request, ex); 2674 struct ceph_bvec_iter *it = arg; 2675 2676 obj_req->bvec_pos = *it; 2677 ceph_bvec_iter_shorten(&obj_req->bvec_pos, bytes); 2678 ceph_bvec_iter_advance(it, bytes); 2679 } 2680 2681 static void count_bvecs(struct ceph_object_extent *ex, u32 bytes, void *arg) 2682 { 2683 struct rbd_obj_request *obj_req = 2684 container_of(ex, struct rbd_obj_request, ex); 2685 struct ceph_bvec_iter *it = arg; 2686 2687 ceph_bvec_iter_advance_step(it, bytes, ({ 2688 obj_req->bvec_count++; 2689 })); 2690 } 2691 2692 static void copy_bvecs(struct ceph_object_extent *ex, u32 bytes, void *arg) 2693 { 2694 struct rbd_obj_request *obj_req = 2695 container_of(ex, struct rbd_obj_request, ex); 2696 struct ceph_bvec_iter *it = arg; 2697 2698 ceph_bvec_iter_advance_step(it, bytes, ({ 2699 obj_req->bvec_pos.bvecs[obj_req->bvec_idx++] = bv; 2700 obj_req->bvec_pos.iter.bi_size += bv.bv_len; 2701 })); 2702 } 2703 2704 static int __rbd_img_fill_from_bvecs(struct rbd_img_request *img_req, 2705 struct ceph_file_extent *img_extents, 2706 u32 num_img_extents, 2707 struct ceph_bvec_iter *bvec_pos) 2708 { 2709 struct rbd_img_fill_ctx fctx = { 2710 .pos_type = OBJ_REQUEST_BVECS, 2711 .pos = (union rbd_img_fill_iter *)bvec_pos, 2712 .set_pos_fn = set_bvec_pos, 2713 .count_fn = count_bvecs, 2714 .copy_fn = copy_bvecs, 2715 }; 2716 2717 return rbd_img_fill_request(img_req, img_extents, num_img_extents, 2718 &fctx); 2719 } 2720 2721 static int rbd_img_fill_from_bvecs(struct rbd_img_request *img_req, 2722 struct ceph_file_extent *img_extents, 2723 u32 num_img_extents, 2724 struct bio_vec *bvecs) 2725 { 2726 struct ceph_bvec_iter it = { 2727 .bvecs = bvecs, 2728 .iter = { .bi_size = ceph_file_extents_bytes(img_extents, 2729 num_img_extents) }, 2730 }; 2731 2732 return __rbd_img_fill_from_bvecs(img_req, img_extents, num_img_extents, 2733 &it); 2734 } 2735 2736 static void rbd_img_handle_request_work(struct work_struct *work) 2737 { 2738 struct rbd_img_request *img_req = 2739 container_of(work, struct rbd_img_request, work); 2740 2741 rbd_img_handle_request(img_req, img_req->work_result); 2742 } 2743 2744 static void rbd_img_schedule(struct rbd_img_request *img_req, int result) 2745 { 2746 INIT_WORK(&img_req->work, rbd_img_handle_request_work); 2747 img_req->work_result = result; 2748 queue_work(rbd_wq, &img_req->work); 2749 } 2750 2751 static bool rbd_obj_may_exist(struct rbd_obj_request *obj_req) 2752 { 2753 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 2754 2755 if (rbd_object_map_may_exist(rbd_dev, obj_req->ex.oe_objno)) { 2756 obj_req->flags |= RBD_OBJ_FLAG_MAY_EXIST; 2757 return true; 2758 } 2759 2760 dout("%s %p objno %llu assuming dne\n", __func__, obj_req, 2761 obj_req->ex.oe_objno); 2762 return false; 2763 } 2764 2765 static int rbd_obj_read_object(struct rbd_obj_request *obj_req) 2766 { 2767 struct ceph_osd_request *osd_req; 2768 int ret; 2769 2770 osd_req = __rbd_obj_add_osd_request(obj_req, NULL, 1); 2771 if (IS_ERR(osd_req)) 2772 return PTR_ERR(osd_req); 2773 2774 osd_req_op_extent_init(osd_req, 0, CEPH_OSD_OP_READ, 2775 obj_req->ex.oe_off, obj_req->ex.oe_len, 0, 0); 2776 rbd_osd_setup_data(osd_req, 0); 2777 rbd_osd_format_read(osd_req); 2778 2779 ret = ceph_osdc_alloc_messages(osd_req, GFP_NOIO); 2780 if (ret) 2781 return ret; 2782 2783 rbd_osd_submit(osd_req); 2784 return 0; 2785 } 2786 2787 static int rbd_obj_read_from_parent(struct rbd_obj_request *obj_req) 2788 { 2789 struct rbd_img_request *img_req = obj_req->img_request; 2790 struct rbd_device *parent = img_req->rbd_dev->parent; 2791 struct rbd_img_request *child_img_req; 2792 int ret; 2793 2794 child_img_req = kmem_cache_alloc(rbd_img_request_cache, GFP_NOIO); 2795 if (!child_img_req) 2796 return -ENOMEM; 2797 2798 rbd_img_request_init(child_img_req, parent, OBJ_OP_READ); 2799 __set_bit(IMG_REQ_CHILD, &child_img_req->flags); 2800 child_img_req->obj_request = obj_req; 2801 2802 down_read(&parent->header_rwsem); 2803 rbd_img_capture_header(child_img_req); 2804 up_read(&parent->header_rwsem); 2805 2806 dout("%s child_img_req %p for obj_req %p\n", __func__, child_img_req, 2807 obj_req); 2808 2809 if (!rbd_img_is_write(img_req)) { 2810 switch (img_req->data_type) { 2811 case OBJ_REQUEST_BIO: 2812 ret = __rbd_img_fill_from_bio(child_img_req, 2813 obj_req->img_extents, 2814 obj_req->num_img_extents, 2815 &obj_req->bio_pos); 2816 break; 2817 case OBJ_REQUEST_BVECS: 2818 case OBJ_REQUEST_OWN_BVECS: 2819 ret = __rbd_img_fill_from_bvecs(child_img_req, 2820 obj_req->img_extents, 2821 obj_req->num_img_extents, 2822 &obj_req->bvec_pos); 2823 break; 2824 default: 2825 BUG(); 2826 } 2827 } else { 2828 ret = rbd_img_fill_from_bvecs(child_img_req, 2829 obj_req->img_extents, 2830 obj_req->num_img_extents, 2831 obj_req->copyup_bvecs); 2832 } 2833 if (ret) { 2834 rbd_img_request_destroy(child_img_req); 2835 return ret; 2836 } 2837 2838 /* avoid parent chain recursion */ 2839 rbd_img_schedule(child_img_req, 0); 2840 return 0; 2841 } 2842 2843 static bool rbd_obj_advance_read(struct rbd_obj_request *obj_req, int *result) 2844 { 2845 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 2846 int ret; 2847 2848 again: 2849 switch (obj_req->read_state) { 2850 case RBD_OBJ_READ_START: 2851 rbd_assert(!*result); 2852 2853 if (!rbd_obj_may_exist(obj_req)) { 2854 *result = -ENOENT; 2855 obj_req->read_state = RBD_OBJ_READ_OBJECT; 2856 goto again; 2857 } 2858 2859 ret = rbd_obj_read_object(obj_req); 2860 if (ret) { 2861 *result = ret; 2862 return true; 2863 } 2864 obj_req->read_state = RBD_OBJ_READ_OBJECT; 2865 return false; 2866 case RBD_OBJ_READ_OBJECT: 2867 if (*result == -ENOENT && rbd_dev->parent_overlap) { 2868 /* reverse map this object extent onto the parent */ 2869 ret = rbd_obj_calc_img_extents(obj_req, false); 2870 if (ret) { 2871 *result = ret; 2872 return true; 2873 } 2874 if (obj_req->num_img_extents) { 2875 ret = rbd_obj_read_from_parent(obj_req); 2876 if (ret) { 2877 *result = ret; 2878 return true; 2879 } 2880 obj_req->read_state = RBD_OBJ_READ_PARENT; 2881 return false; 2882 } 2883 } 2884 2885 /* 2886 * -ENOENT means a hole in the image -- zero-fill the entire 2887 * length of the request. A short read also implies zero-fill 2888 * to the end of the request. 2889 */ 2890 if (*result == -ENOENT) { 2891 rbd_obj_zero_range(obj_req, 0, obj_req->ex.oe_len); 2892 *result = 0; 2893 } else if (*result >= 0) { 2894 if (*result < obj_req->ex.oe_len) 2895 rbd_obj_zero_range(obj_req, *result, 2896 obj_req->ex.oe_len - *result); 2897 else 2898 rbd_assert(*result == obj_req->ex.oe_len); 2899 *result = 0; 2900 } 2901 return true; 2902 case RBD_OBJ_READ_PARENT: 2903 /* 2904 * The parent image is read only up to the overlap -- zero-fill 2905 * from the overlap to the end of the request. 2906 */ 2907 if (!*result) { 2908 u32 obj_overlap = rbd_obj_img_extents_bytes(obj_req); 2909 2910 if (obj_overlap < obj_req->ex.oe_len) 2911 rbd_obj_zero_range(obj_req, obj_overlap, 2912 obj_req->ex.oe_len - obj_overlap); 2913 } 2914 return true; 2915 default: 2916 BUG(); 2917 } 2918 } 2919 2920 static bool rbd_obj_write_is_noop(struct rbd_obj_request *obj_req) 2921 { 2922 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 2923 2924 if (rbd_object_map_may_exist(rbd_dev, obj_req->ex.oe_objno)) 2925 obj_req->flags |= RBD_OBJ_FLAG_MAY_EXIST; 2926 2927 if (!(obj_req->flags & RBD_OBJ_FLAG_MAY_EXIST) && 2928 (obj_req->flags & RBD_OBJ_FLAG_NOOP_FOR_NONEXISTENT)) { 2929 dout("%s %p noop for nonexistent\n", __func__, obj_req); 2930 return true; 2931 } 2932 2933 return false; 2934 } 2935 2936 /* 2937 * Return: 2938 * 0 - object map update sent 2939 * 1 - object map update isn't needed 2940 * <0 - error 2941 */ 2942 static int rbd_obj_write_pre_object_map(struct rbd_obj_request *obj_req) 2943 { 2944 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 2945 u8 new_state; 2946 2947 if (!(rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP)) 2948 return 1; 2949 2950 if (obj_req->flags & RBD_OBJ_FLAG_DELETION) 2951 new_state = OBJECT_PENDING; 2952 else 2953 new_state = OBJECT_EXISTS; 2954 2955 return rbd_object_map_update(obj_req, CEPH_NOSNAP, new_state, NULL); 2956 } 2957 2958 static int rbd_obj_write_object(struct rbd_obj_request *obj_req) 2959 { 2960 struct ceph_osd_request *osd_req; 2961 int num_ops = count_write_ops(obj_req); 2962 int which = 0; 2963 int ret; 2964 2965 if (obj_req->flags & RBD_OBJ_FLAG_COPYUP_ENABLED) 2966 num_ops++; /* stat */ 2967 2968 osd_req = rbd_obj_add_osd_request(obj_req, num_ops); 2969 if (IS_ERR(osd_req)) 2970 return PTR_ERR(osd_req); 2971 2972 if (obj_req->flags & RBD_OBJ_FLAG_COPYUP_ENABLED) { 2973 ret = rbd_osd_setup_stat(osd_req, which++); 2974 if (ret) 2975 return ret; 2976 } 2977 2978 rbd_osd_setup_write_ops(osd_req, which); 2979 rbd_osd_format_write(osd_req); 2980 2981 ret = ceph_osdc_alloc_messages(osd_req, GFP_NOIO); 2982 if (ret) 2983 return ret; 2984 2985 rbd_osd_submit(osd_req); 2986 return 0; 2987 } 2988 2989 /* 2990 * copyup_bvecs pages are never highmem pages 2991 */ 2992 static bool is_zero_bvecs(struct bio_vec *bvecs, u32 bytes) 2993 { 2994 struct ceph_bvec_iter it = { 2995 .bvecs = bvecs, 2996 .iter = { .bi_size = bytes }, 2997 }; 2998 2999 ceph_bvec_iter_advance_step(&it, bytes, ({ 3000 if (memchr_inv(page_address(bv.bv_page) + bv.bv_offset, 0, 3001 bv.bv_len)) 3002 return false; 3003 })); 3004 return true; 3005 } 3006 3007 #define MODS_ONLY U32_MAX 3008 3009 static int rbd_obj_copyup_empty_snapc(struct rbd_obj_request *obj_req, 3010 u32 bytes) 3011 { 3012 struct ceph_osd_request *osd_req; 3013 int ret; 3014 3015 dout("%s obj_req %p bytes %u\n", __func__, obj_req, bytes); 3016 rbd_assert(bytes > 0 && bytes != MODS_ONLY); 3017 3018 osd_req = __rbd_obj_add_osd_request(obj_req, &rbd_empty_snapc, 1); 3019 if (IS_ERR(osd_req)) 3020 return PTR_ERR(osd_req); 3021 3022 ret = rbd_osd_setup_copyup(osd_req, 0, bytes); 3023 if (ret) 3024 return ret; 3025 3026 rbd_osd_format_write(osd_req); 3027 3028 ret = ceph_osdc_alloc_messages(osd_req, GFP_NOIO); 3029 if (ret) 3030 return ret; 3031 3032 rbd_osd_submit(osd_req); 3033 return 0; 3034 } 3035 3036 static int rbd_obj_copyup_current_snapc(struct rbd_obj_request *obj_req, 3037 u32 bytes) 3038 { 3039 struct ceph_osd_request *osd_req; 3040 int num_ops = count_write_ops(obj_req); 3041 int which = 0; 3042 int ret; 3043 3044 dout("%s obj_req %p bytes %u\n", __func__, obj_req, bytes); 3045 3046 if (bytes != MODS_ONLY) 3047 num_ops++; /* copyup */ 3048 3049 osd_req = rbd_obj_add_osd_request(obj_req, num_ops); 3050 if (IS_ERR(osd_req)) 3051 return PTR_ERR(osd_req); 3052 3053 if (bytes != MODS_ONLY) { 3054 ret = rbd_osd_setup_copyup(osd_req, which++, bytes); 3055 if (ret) 3056 return ret; 3057 } 3058 3059 rbd_osd_setup_write_ops(osd_req, which); 3060 rbd_osd_format_write(osd_req); 3061 3062 ret = ceph_osdc_alloc_messages(osd_req, GFP_NOIO); 3063 if (ret) 3064 return ret; 3065 3066 rbd_osd_submit(osd_req); 3067 return 0; 3068 } 3069 3070 static int setup_copyup_bvecs(struct rbd_obj_request *obj_req, u64 obj_overlap) 3071 { 3072 u32 i; 3073 3074 rbd_assert(!obj_req->copyup_bvecs); 3075 obj_req->copyup_bvec_count = calc_pages_for(0, obj_overlap); 3076 obj_req->copyup_bvecs = kcalloc(obj_req->copyup_bvec_count, 3077 sizeof(*obj_req->copyup_bvecs), 3078 GFP_NOIO); 3079 if (!obj_req->copyup_bvecs) 3080 return -ENOMEM; 3081 3082 for (i = 0; i < obj_req->copyup_bvec_count; i++) { 3083 unsigned int len = min(obj_overlap, (u64)PAGE_SIZE); 3084 3085 obj_req->copyup_bvecs[i].bv_page = alloc_page(GFP_NOIO); 3086 if (!obj_req->copyup_bvecs[i].bv_page) 3087 return -ENOMEM; 3088 3089 obj_req->copyup_bvecs[i].bv_offset = 0; 3090 obj_req->copyup_bvecs[i].bv_len = len; 3091 obj_overlap -= len; 3092 } 3093 3094 rbd_assert(!obj_overlap); 3095 return 0; 3096 } 3097 3098 /* 3099 * The target object doesn't exist. Read the data for the entire 3100 * target object up to the overlap point (if any) from the parent, 3101 * so we can use it for a copyup. 3102 */ 3103 static int rbd_obj_copyup_read_parent(struct rbd_obj_request *obj_req) 3104 { 3105 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 3106 int ret; 3107 3108 rbd_assert(obj_req->num_img_extents); 3109 prune_extents(obj_req->img_extents, &obj_req->num_img_extents, 3110 rbd_dev->parent_overlap); 3111 if (!obj_req->num_img_extents) { 3112 /* 3113 * The overlap has become 0 (most likely because the 3114 * image has been flattened). Re-submit the original write 3115 * request -- pass MODS_ONLY since the copyup isn't needed 3116 * anymore. 3117 */ 3118 return rbd_obj_copyup_current_snapc(obj_req, MODS_ONLY); 3119 } 3120 3121 ret = setup_copyup_bvecs(obj_req, rbd_obj_img_extents_bytes(obj_req)); 3122 if (ret) 3123 return ret; 3124 3125 return rbd_obj_read_from_parent(obj_req); 3126 } 3127 3128 static void rbd_obj_copyup_object_maps(struct rbd_obj_request *obj_req) 3129 { 3130 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 3131 struct ceph_snap_context *snapc = obj_req->img_request->snapc; 3132 u8 new_state; 3133 u32 i; 3134 int ret; 3135 3136 rbd_assert(!obj_req->pending.result && !obj_req->pending.num_pending); 3137 3138 if (!(rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP)) 3139 return; 3140 3141 if (obj_req->flags & RBD_OBJ_FLAG_COPYUP_ZEROS) 3142 return; 3143 3144 for (i = 0; i < snapc->num_snaps; i++) { 3145 if ((rbd_dev->header.features & RBD_FEATURE_FAST_DIFF) && 3146 i + 1 < snapc->num_snaps) 3147 new_state = OBJECT_EXISTS_CLEAN; 3148 else 3149 new_state = OBJECT_EXISTS; 3150 3151 ret = rbd_object_map_update(obj_req, snapc->snaps[i], 3152 new_state, NULL); 3153 if (ret < 0) { 3154 obj_req->pending.result = ret; 3155 return; 3156 } 3157 3158 rbd_assert(!ret); 3159 obj_req->pending.num_pending++; 3160 } 3161 } 3162 3163 static void rbd_obj_copyup_write_object(struct rbd_obj_request *obj_req) 3164 { 3165 u32 bytes = rbd_obj_img_extents_bytes(obj_req); 3166 int ret; 3167 3168 rbd_assert(!obj_req->pending.result && !obj_req->pending.num_pending); 3169 3170 /* 3171 * Only send non-zero copyup data to save some I/O and network 3172 * bandwidth -- zero copyup data is equivalent to the object not 3173 * existing. 3174 */ 3175 if (obj_req->flags & RBD_OBJ_FLAG_COPYUP_ZEROS) 3176 bytes = 0; 3177 3178 if (obj_req->img_request->snapc->num_snaps && bytes > 0) { 3179 /* 3180 * Send a copyup request with an empty snapshot context to 3181 * deep-copyup the object through all existing snapshots. 3182 * A second request with the current snapshot context will be 3183 * sent for the actual modification. 3184 */ 3185 ret = rbd_obj_copyup_empty_snapc(obj_req, bytes); 3186 if (ret) { 3187 obj_req->pending.result = ret; 3188 return; 3189 } 3190 3191 obj_req->pending.num_pending++; 3192 bytes = MODS_ONLY; 3193 } 3194 3195 ret = rbd_obj_copyup_current_snapc(obj_req, bytes); 3196 if (ret) { 3197 obj_req->pending.result = ret; 3198 return; 3199 } 3200 3201 obj_req->pending.num_pending++; 3202 } 3203 3204 static bool rbd_obj_advance_copyup(struct rbd_obj_request *obj_req, int *result) 3205 { 3206 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 3207 int ret; 3208 3209 again: 3210 switch (obj_req->copyup_state) { 3211 case RBD_OBJ_COPYUP_START: 3212 rbd_assert(!*result); 3213 3214 ret = rbd_obj_copyup_read_parent(obj_req); 3215 if (ret) { 3216 *result = ret; 3217 return true; 3218 } 3219 if (obj_req->num_img_extents) 3220 obj_req->copyup_state = RBD_OBJ_COPYUP_READ_PARENT; 3221 else 3222 obj_req->copyup_state = RBD_OBJ_COPYUP_WRITE_OBJECT; 3223 return false; 3224 case RBD_OBJ_COPYUP_READ_PARENT: 3225 if (*result) 3226 return true; 3227 3228 if (is_zero_bvecs(obj_req->copyup_bvecs, 3229 rbd_obj_img_extents_bytes(obj_req))) { 3230 dout("%s %p detected zeros\n", __func__, obj_req); 3231 obj_req->flags |= RBD_OBJ_FLAG_COPYUP_ZEROS; 3232 } 3233 3234 rbd_obj_copyup_object_maps(obj_req); 3235 if (!obj_req->pending.num_pending) { 3236 *result = obj_req->pending.result; 3237 obj_req->copyup_state = RBD_OBJ_COPYUP_OBJECT_MAPS; 3238 goto again; 3239 } 3240 obj_req->copyup_state = __RBD_OBJ_COPYUP_OBJECT_MAPS; 3241 return false; 3242 case __RBD_OBJ_COPYUP_OBJECT_MAPS: 3243 if (!pending_result_dec(&obj_req->pending, result)) 3244 return false; 3245 fallthrough; 3246 case RBD_OBJ_COPYUP_OBJECT_MAPS: 3247 if (*result) { 3248 rbd_warn(rbd_dev, "snap object map update failed: %d", 3249 *result); 3250 return true; 3251 } 3252 3253 rbd_obj_copyup_write_object(obj_req); 3254 if (!obj_req->pending.num_pending) { 3255 *result = obj_req->pending.result; 3256 obj_req->copyup_state = RBD_OBJ_COPYUP_WRITE_OBJECT; 3257 goto again; 3258 } 3259 obj_req->copyup_state = __RBD_OBJ_COPYUP_WRITE_OBJECT; 3260 return false; 3261 case __RBD_OBJ_COPYUP_WRITE_OBJECT: 3262 if (!pending_result_dec(&obj_req->pending, result)) 3263 return false; 3264 fallthrough; 3265 case RBD_OBJ_COPYUP_WRITE_OBJECT: 3266 return true; 3267 default: 3268 BUG(); 3269 } 3270 } 3271 3272 /* 3273 * Return: 3274 * 0 - object map update sent 3275 * 1 - object map update isn't needed 3276 * <0 - error 3277 */ 3278 static int rbd_obj_write_post_object_map(struct rbd_obj_request *obj_req) 3279 { 3280 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 3281 u8 current_state = OBJECT_PENDING; 3282 3283 if (!(rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP)) 3284 return 1; 3285 3286 if (!(obj_req->flags & RBD_OBJ_FLAG_DELETION)) 3287 return 1; 3288 3289 return rbd_object_map_update(obj_req, CEPH_NOSNAP, OBJECT_NONEXISTENT, 3290 ¤t_state); 3291 } 3292 3293 static bool rbd_obj_advance_write(struct rbd_obj_request *obj_req, int *result) 3294 { 3295 struct rbd_device *rbd_dev = obj_req->img_request->rbd_dev; 3296 int ret; 3297 3298 again: 3299 switch (obj_req->write_state) { 3300 case RBD_OBJ_WRITE_START: 3301 rbd_assert(!*result); 3302 3303 if (rbd_obj_write_is_noop(obj_req)) 3304 return true; 3305 3306 ret = rbd_obj_write_pre_object_map(obj_req); 3307 if (ret < 0) { 3308 *result = ret; 3309 return true; 3310 } 3311 obj_req->write_state = RBD_OBJ_WRITE_PRE_OBJECT_MAP; 3312 if (ret > 0) 3313 goto again; 3314 return false; 3315 case RBD_OBJ_WRITE_PRE_OBJECT_MAP: 3316 if (*result) { 3317 rbd_warn(rbd_dev, "pre object map update failed: %d", 3318 *result); 3319 return true; 3320 } 3321 ret = rbd_obj_write_object(obj_req); 3322 if (ret) { 3323 *result = ret; 3324 return true; 3325 } 3326 obj_req->write_state = RBD_OBJ_WRITE_OBJECT; 3327 return false; 3328 case RBD_OBJ_WRITE_OBJECT: 3329 if (*result == -ENOENT) { 3330 if (obj_req->flags & RBD_OBJ_FLAG_COPYUP_ENABLED) { 3331 *result = 0; 3332 obj_req->copyup_state = RBD_OBJ_COPYUP_START; 3333 obj_req->write_state = __RBD_OBJ_WRITE_COPYUP; 3334 goto again; 3335 } 3336 /* 3337 * On a non-existent object: 3338 * delete - -ENOENT, truncate/zero - 0 3339 */ 3340 if (obj_req->flags & RBD_OBJ_FLAG_DELETION) 3341 *result = 0; 3342 } 3343 if (*result) 3344 return true; 3345 3346 obj_req->write_state = RBD_OBJ_WRITE_COPYUP; 3347 goto again; 3348 case __RBD_OBJ_WRITE_COPYUP: 3349 if (!rbd_obj_advance_copyup(obj_req, result)) 3350 return false; 3351 fallthrough; 3352 case RBD_OBJ_WRITE_COPYUP: 3353 if (*result) { 3354 rbd_warn(rbd_dev, "copyup failed: %d", *result); 3355 return true; 3356 } 3357 ret = rbd_obj_write_post_object_map(obj_req); 3358 if (ret < 0) { 3359 *result = ret; 3360 return true; 3361 } 3362 obj_req->write_state = RBD_OBJ_WRITE_POST_OBJECT_MAP; 3363 if (ret > 0) 3364 goto again; 3365 return false; 3366 case RBD_OBJ_WRITE_POST_OBJECT_MAP: 3367 if (*result) 3368 rbd_warn(rbd_dev, "post object map update failed: %d", 3369 *result); 3370 return true; 3371 default: 3372 BUG(); 3373 } 3374 } 3375 3376 /* 3377 * Return true if @obj_req is completed. 3378 */ 3379 static bool __rbd_obj_handle_request(struct rbd_obj_request *obj_req, 3380 int *result) 3381 { 3382 struct rbd_img_request *img_req = obj_req->img_request; 3383 struct rbd_device *rbd_dev = img_req->rbd_dev; 3384 bool done; 3385 3386 mutex_lock(&obj_req->state_mutex); 3387 if (!rbd_img_is_write(img_req)) 3388 done = rbd_obj_advance_read(obj_req, result); 3389 else 3390 done = rbd_obj_advance_write(obj_req, result); 3391 mutex_unlock(&obj_req->state_mutex); 3392 3393 if (done && *result) { 3394 rbd_assert(*result < 0); 3395 rbd_warn(rbd_dev, "%s at objno %llu %llu~%llu result %d", 3396 obj_op_name(img_req->op_type), obj_req->ex.oe_objno, 3397 obj_req->ex.oe_off, obj_req->ex.oe_len, *result); 3398 } 3399 return done; 3400 } 3401 3402 /* 3403 * This is open-coded in rbd_img_handle_request() to avoid parent chain 3404 * recursion. 3405 */ 3406 static void rbd_obj_handle_request(struct rbd_obj_request *obj_req, int result) 3407 { 3408 if (__rbd_obj_handle_request(obj_req, &result)) 3409 rbd_img_handle_request(obj_req->img_request, result); 3410 } 3411 3412 static bool need_exclusive_lock(struct rbd_img_request *img_req) 3413 { 3414 struct rbd_device *rbd_dev = img_req->rbd_dev; 3415 3416 if (!(rbd_dev->header.features & RBD_FEATURE_EXCLUSIVE_LOCK)) 3417 return false; 3418 3419 if (rbd_is_ro(rbd_dev)) 3420 return false; 3421 3422 rbd_assert(!test_bit(IMG_REQ_CHILD, &img_req->flags)); 3423 if (rbd_dev->opts->lock_on_read || 3424 (rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP)) 3425 return true; 3426 3427 return rbd_img_is_write(img_req); 3428 } 3429 3430 static bool rbd_lock_add_request(struct rbd_img_request *img_req) 3431 { 3432 struct rbd_device *rbd_dev = img_req->rbd_dev; 3433 bool locked; 3434 3435 lockdep_assert_held(&rbd_dev->lock_rwsem); 3436 locked = rbd_dev->lock_state == RBD_LOCK_STATE_LOCKED; 3437 spin_lock(&rbd_dev->lock_lists_lock); 3438 rbd_assert(list_empty(&img_req->lock_item)); 3439 if (!locked) 3440 list_add_tail(&img_req->lock_item, &rbd_dev->acquiring_list); 3441 else 3442 list_add_tail(&img_req->lock_item, &rbd_dev->running_list); 3443 spin_unlock(&rbd_dev->lock_lists_lock); 3444 return locked; 3445 } 3446 3447 static void rbd_lock_del_request(struct rbd_img_request *img_req) 3448 { 3449 struct rbd_device *rbd_dev = img_req->rbd_dev; 3450 bool need_wakeup; 3451 3452 lockdep_assert_held(&rbd_dev->lock_rwsem); 3453 spin_lock(&rbd_dev->lock_lists_lock); 3454 rbd_assert(!list_empty(&img_req->lock_item)); 3455 list_del_init(&img_req->lock_item); 3456 need_wakeup = (rbd_dev->lock_state == RBD_LOCK_STATE_RELEASING && 3457 list_empty(&rbd_dev->running_list)); 3458 spin_unlock(&rbd_dev->lock_lists_lock); 3459 if (need_wakeup) 3460 complete(&rbd_dev->releasing_wait); 3461 } 3462 3463 static int rbd_img_exclusive_lock(struct rbd_img_request *img_req) 3464 { 3465 struct rbd_device *rbd_dev = img_req->rbd_dev; 3466 3467 if (!need_exclusive_lock(img_req)) 3468 return 1; 3469 3470 if (rbd_lock_add_request(img_req)) 3471 return 1; 3472 3473 if (rbd_dev->opts->exclusive) { 3474 WARN_ON(1); /* lock got released? */ 3475 return -EROFS; 3476 } 3477 3478 /* 3479 * Note the use of mod_delayed_work() in rbd_acquire_lock() 3480 * and cancel_delayed_work() in wake_lock_waiters(). 3481 */ 3482 dout("%s rbd_dev %p queueing lock_dwork\n", __func__, rbd_dev); 3483 queue_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork, 0); 3484 return 0; 3485 } 3486 3487 static void rbd_img_object_requests(struct rbd_img_request *img_req) 3488 { 3489 struct rbd_obj_request *obj_req; 3490 3491 rbd_assert(!img_req->pending.result && !img_req->pending.num_pending); 3492 3493 for_each_obj_request(img_req, obj_req) { 3494 int result = 0; 3495 3496 if (__rbd_obj_handle_request(obj_req, &result)) { 3497 if (result) { 3498 img_req->pending.result = result; 3499 return; 3500 } 3501 } else { 3502 img_req->pending.num_pending++; 3503 } 3504 } 3505 } 3506 3507 static bool rbd_img_advance(struct rbd_img_request *img_req, int *result) 3508 { 3509 struct rbd_device *rbd_dev = img_req->rbd_dev; 3510 int ret; 3511 3512 again: 3513 switch (img_req->state) { 3514 case RBD_IMG_START: 3515 rbd_assert(!*result); 3516 3517 ret = rbd_img_exclusive_lock(img_req); 3518 if (ret < 0) { 3519 *result = ret; 3520 return true; 3521 } 3522 img_req->state = RBD_IMG_EXCLUSIVE_LOCK; 3523 if (ret > 0) 3524 goto again; 3525 return false; 3526 case RBD_IMG_EXCLUSIVE_LOCK: 3527 if (*result) 3528 return true; 3529 3530 rbd_assert(!need_exclusive_lock(img_req) || 3531 __rbd_is_lock_owner(rbd_dev)); 3532 3533 rbd_img_object_requests(img_req); 3534 if (!img_req->pending.num_pending) { 3535 *result = img_req->pending.result; 3536 img_req->state = RBD_IMG_OBJECT_REQUESTS; 3537 goto again; 3538 } 3539 img_req->state = __RBD_IMG_OBJECT_REQUESTS; 3540 return false; 3541 case __RBD_IMG_OBJECT_REQUESTS: 3542 if (!pending_result_dec(&img_req->pending, result)) 3543 return false; 3544 fallthrough; 3545 case RBD_IMG_OBJECT_REQUESTS: 3546 return true; 3547 default: 3548 BUG(); 3549 } 3550 } 3551 3552 /* 3553 * Return true if @img_req is completed. 3554 */ 3555 static bool __rbd_img_handle_request(struct rbd_img_request *img_req, 3556 int *result) 3557 { 3558 struct rbd_device *rbd_dev = img_req->rbd_dev; 3559 bool done; 3560 3561 if (need_exclusive_lock(img_req)) { 3562 down_read(&rbd_dev->lock_rwsem); 3563 mutex_lock(&img_req->state_mutex); 3564 done = rbd_img_advance(img_req, result); 3565 if (done) 3566 rbd_lock_del_request(img_req); 3567 mutex_unlock(&img_req->state_mutex); 3568 up_read(&rbd_dev->lock_rwsem); 3569 } else { 3570 mutex_lock(&img_req->state_mutex); 3571 done = rbd_img_advance(img_req, result); 3572 mutex_unlock(&img_req->state_mutex); 3573 } 3574 3575 if (done && *result) { 3576 rbd_assert(*result < 0); 3577 rbd_warn(rbd_dev, "%s%s result %d", 3578 test_bit(IMG_REQ_CHILD, &img_req->flags) ? "child " : "", 3579 obj_op_name(img_req->op_type), *result); 3580 } 3581 return done; 3582 } 3583 3584 static void rbd_img_handle_request(struct rbd_img_request *img_req, int result) 3585 { 3586 again: 3587 if (!__rbd_img_handle_request(img_req, &result)) 3588 return; 3589 3590 if (test_bit(IMG_REQ_CHILD, &img_req->flags)) { 3591 struct rbd_obj_request *obj_req = img_req->obj_request; 3592 3593 rbd_img_request_destroy(img_req); 3594 if (__rbd_obj_handle_request(obj_req, &result)) { 3595 img_req = obj_req->img_request; 3596 goto again; 3597 } 3598 } else { 3599 struct request *rq = blk_mq_rq_from_pdu(img_req); 3600 3601 rbd_img_request_destroy(img_req); 3602 blk_mq_end_request(rq, errno_to_blk_status(result)); 3603 } 3604 } 3605 3606 static const struct rbd_client_id rbd_empty_cid; 3607 3608 static bool rbd_cid_equal(const struct rbd_client_id *lhs, 3609 const struct rbd_client_id *rhs) 3610 { 3611 return lhs->gid == rhs->gid && lhs->handle == rhs->handle; 3612 } 3613 3614 static struct rbd_client_id rbd_get_cid(struct rbd_device *rbd_dev) 3615 { 3616 struct rbd_client_id cid; 3617 3618 mutex_lock(&rbd_dev->watch_mutex); 3619 cid.gid = ceph_client_gid(rbd_dev->rbd_client->client); 3620 cid.handle = rbd_dev->watch_cookie; 3621 mutex_unlock(&rbd_dev->watch_mutex); 3622 return cid; 3623 } 3624 3625 /* 3626 * lock_rwsem must be held for write 3627 */ 3628 static void rbd_set_owner_cid(struct rbd_device *rbd_dev, 3629 const struct rbd_client_id *cid) 3630 { 3631 dout("%s rbd_dev %p %llu-%llu -> %llu-%llu\n", __func__, rbd_dev, 3632 rbd_dev->owner_cid.gid, rbd_dev->owner_cid.handle, 3633 cid->gid, cid->handle); 3634 rbd_dev->owner_cid = *cid; /* struct */ 3635 } 3636 3637 static void format_lock_cookie(struct rbd_device *rbd_dev, char *buf) 3638 { 3639 mutex_lock(&rbd_dev->watch_mutex); 3640 sprintf(buf, "%s %llu", RBD_LOCK_COOKIE_PREFIX, rbd_dev->watch_cookie); 3641 mutex_unlock(&rbd_dev->watch_mutex); 3642 } 3643 3644 static void __rbd_lock(struct rbd_device *rbd_dev, const char *cookie) 3645 { 3646 struct rbd_client_id cid = rbd_get_cid(rbd_dev); 3647 3648 rbd_dev->lock_state = RBD_LOCK_STATE_LOCKED; 3649 strcpy(rbd_dev->lock_cookie, cookie); 3650 rbd_set_owner_cid(rbd_dev, &cid); 3651 queue_work(rbd_dev->task_wq, &rbd_dev->acquired_lock_work); 3652 } 3653 3654 /* 3655 * lock_rwsem must be held for write 3656 */ 3657 static int rbd_lock(struct rbd_device *rbd_dev) 3658 { 3659 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 3660 char cookie[32]; 3661 int ret; 3662 3663 WARN_ON(__rbd_is_lock_owner(rbd_dev) || 3664 rbd_dev->lock_cookie[0] != '\0'); 3665 3666 format_lock_cookie(rbd_dev, cookie); 3667 ret = ceph_cls_lock(osdc, &rbd_dev->header_oid, &rbd_dev->header_oloc, 3668 RBD_LOCK_NAME, CEPH_CLS_LOCK_EXCLUSIVE, cookie, 3669 RBD_LOCK_TAG, "", 0); 3670 if (ret) 3671 return ret; 3672 3673 __rbd_lock(rbd_dev, cookie); 3674 return 0; 3675 } 3676 3677 /* 3678 * lock_rwsem must be held for write 3679 */ 3680 static void rbd_unlock(struct rbd_device *rbd_dev) 3681 { 3682 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 3683 int ret; 3684 3685 WARN_ON(!__rbd_is_lock_owner(rbd_dev) || 3686 rbd_dev->lock_cookie[0] == '\0'); 3687 3688 ret = ceph_cls_unlock(osdc, &rbd_dev->header_oid, &rbd_dev->header_oloc, 3689 RBD_LOCK_NAME, rbd_dev->lock_cookie); 3690 if (ret && ret != -ENOENT) 3691 rbd_warn(rbd_dev, "failed to unlock header: %d", ret); 3692 3693 /* treat errors as the image is unlocked */ 3694 rbd_dev->lock_state = RBD_LOCK_STATE_UNLOCKED; 3695 rbd_dev->lock_cookie[0] = '\0'; 3696 rbd_set_owner_cid(rbd_dev, &rbd_empty_cid); 3697 queue_work(rbd_dev->task_wq, &rbd_dev->released_lock_work); 3698 } 3699 3700 static int __rbd_notify_op_lock(struct rbd_device *rbd_dev, 3701 enum rbd_notify_op notify_op, 3702 struct page ***preply_pages, 3703 size_t *preply_len) 3704 { 3705 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 3706 struct rbd_client_id cid = rbd_get_cid(rbd_dev); 3707 char buf[4 + 8 + 8 + CEPH_ENCODING_START_BLK_LEN]; 3708 int buf_size = sizeof(buf); 3709 void *p = buf; 3710 3711 dout("%s rbd_dev %p notify_op %d\n", __func__, rbd_dev, notify_op); 3712 3713 /* encode *LockPayload NotifyMessage (op + ClientId) */ 3714 ceph_start_encoding(&p, 2, 1, buf_size - CEPH_ENCODING_START_BLK_LEN); 3715 ceph_encode_32(&p, notify_op); 3716 ceph_encode_64(&p, cid.gid); 3717 ceph_encode_64(&p, cid.handle); 3718 3719 return ceph_osdc_notify(osdc, &rbd_dev->header_oid, 3720 &rbd_dev->header_oloc, buf, buf_size, 3721 RBD_NOTIFY_TIMEOUT, preply_pages, preply_len); 3722 } 3723 3724 static void rbd_notify_op_lock(struct rbd_device *rbd_dev, 3725 enum rbd_notify_op notify_op) 3726 { 3727 __rbd_notify_op_lock(rbd_dev, notify_op, NULL, NULL); 3728 } 3729 3730 static void rbd_notify_acquired_lock(struct work_struct *work) 3731 { 3732 struct rbd_device *rbd_dev = container_of(work, struct rbd_device, 3733 acquired_lock_work); 3734 3735 rbd_notify_op_lock(rbd_dev, RBD_NOTIFY_OP_ACQUIRED_LOCK); 3736 } 3737 3738 static void rbd_notify_released_lock(struct work_struct *work) 3739 { 3740 struct rbd_device *rbd_dev = container_of(work, struct rbd_device, 3741 released_lock_work); 3742 3743 rbd_notify_op_lock(rbd_dev, RBD_NOTIFY_OP_RELEASED_LOCK); 3744 } 3745 3746 static int rbd_request_lock(struct rbd_device *rbd_dev) 3747 { 3748 struct page **reply_pages; 3749 size_t reply_len; 3750 bool lock_owner_responded = false; 3751 int ret; 3752 3753 dout("%s rbd_dev %p\n", __func__, rbd_dev); 3754 3755 ret = __rbd_notify_op_lock(rbd_dev, RBD_NOTIFY_OP_REQUEST_LOCK, 3756 &reply_pages, &reply_len); 3757 if (ret && ret != -ETIMEDOUT) { 3758 rbd_warn(rbd_dev, "failed to request lock: %d", ret); 3759 goto out; 3760 } 3761 3762 if (reply_len > 0 && reply_len <= PAGE_SIZE) { 3763 void *p = page_address(reply_pages[0]); 3764 void *const end = p + reply_len; 3765 u32 n; 3766 3767 ceph_decode_32_safe(&p, end, n, e_inval); /* num_acks */ 3768 while (n--) { 3769 u8 struct_v; 3770 u32 len; 3771 3772 ceph_decode_need(&p, end, 8 + 8, e_inval); 3773 p += 8 + 8; /* skip gid and cookie */ 3774 3775 ceph_decode_32_safe(&p, end, len, e_inval); 3776 if (!len) 3777 continue; 3778 3779 if (lock_owner_responded) { 3780 rbd_warn(rbd_dev, 3781 "duplicate lock owners detected"); 3782 ret = -EIO; 3783 goto out; 3784 } 3785 3786 lock_owner_responded = true; 3787 ret = ceph_start_decoding(&p, end, 1, "ResponseMessage", 3788 &struct_v, &len); 3789 if (ret) { 3790 rbd_warn(rbd_dev, 3791 "failed to decode ResponseMessage: %d", 3792 ret); 3793 goto e_inval; 3794 } 3795 3796 ret = ceph_decode_32(&p); 3797 } 3798 } 3799 3800 if (!lock_owner_responded) { 3801 rbd_warn(rbd_dev, "no lock owners detected"); 3802 ret = -ETIMEDOUT; 3803 } 3804 3805 out: 3806 ceph_release_page_vector(reply_pages, calc_pages_for(0, reply_len)); 3807 return ret; 3808 3809 e_inval: 3810 ret = -EINVAL; 3811 goto out; 3812 } 3813 3814 /* 3815 * Either image request state machine(s) or rbd_add_acquire_lock() 3816 * (i.e. "rbd map"). 3817 */ 3818 static void wake_lock_waiters(struct rbd_device *rbd_dev, int result) 3819 { 3820 struct rbd_img_request *img_req; 3821 3822 dout("%s rbd_dev %p result %d\n", __func__, rbd_dev, result); 3823 lockdep_assert_held_write(&rbd_dev->lock_rwsem); 3824 3825 cancel_delayed_work(&rbd_dev->lock_dwork); 3826 if (!completion_done(&rbd_dev->acquire_wait)) { 3827 rbd_assert(list_empty(&rbd_dev->acquiring_list) && 3828 list_empty(&rbd_dev->running_list)); 3829 rbd_dev->acquire_err = result; 3830 complete_all(&rbd_dev->acquire_wait); 3831 return; 3832 } 3833 3834 list_for_each_entry(img_req, &rbd_dev->acquiring_list, lock_item) { 3835 mutex_lock(&img_req->state_mutex); 3836 rbd_assert(img_req->state == RBD_IMG_EXCLUSIVE_LOCK); 3837 rbd_img_schedule(img_req, result); 3838 mutex_unlock(&img_req->state_mutex); 3839 } 3840 3841 list_splice_tail_init(&rbd_dev->acquiring_list, &rbd_dev->running_list); 3842 } 3843 3844 static int get_lock_owner_info(struct rbd_device *rbd_dev, 3845 struct ceph_locker **lockers, u32 *num_lockers) 3846 { 3847 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 3848 u8 lock_type; 3849 char *lock_tag; 3850 int ret; 3851 3852 dout("%s rbd_dev %p\n", __func__, rbd_dev); 3853 3854 ret = ceph_cls_lock_info(osdc, &rbd_dev->header_oid, 3855 &rbd_dev->header_oloc, RBD_LOCK_NAME, 3856 &lock_type, &lock_tag, lockers, num_lockers); 3857 if (ret) 3858 return ret; 3859 3860 if (*num_lockers == 0) { 3861 dout("%s rbd_dev %p no lockers detected\n", __func__, rbd_dev); 3862 goto out; 3863 } 3864 3865 if (strcmp(lock_tag, RBD_LOCK_TAG)) { 3866 rbd_warn(rbd_dev, "locked by external mechanism, tag %s", 3867 lock_tag); 3868 ret = -EBUSY; 3869 goto out; 3870 } 3871 3872 if (lock_type == CEPH_CLS_LOCK_SHARED) { 3873 rbd_warn(rbd_dev, "shared lock type detected"); 3874 ret = -EBUSY; 3875 goto out; 3876 } 3877 3878 if (strncmp((*lockers)[0].id.cookie, RBD_LOCK_COOKIE_PREFIX, 3879 strlen(RBD_LOCK_COOKIE_PREFIX))) { 3880 rbd_warn(rbd_dev, "locked by external mechanism, cookie %s", 3881 (*lockers)[0].id.cookie); 3882 ret = -EBUSY; 3883 goto out; 3884 } 3885 3886 out: 3887 kfree(lock_tag); 3888 return ret; 3889 } 3890 3891 static int find_watcher(struct rbd_device *rbd_dev, 3892 const struct ceph_locker *locker) 3893 { 3894 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 3895 struct ceph_watch_item *watchers; 3896 u32 num_watchers; 3897 u64 cookie; 3898 int i; 3899 int ret; 3900 3901 ret = ceph_osdc_list_watchers(osdc, &rbd_dev->header_oid, 3902 &rbd_dev->header_oloc, &watchers, 3903 &num_watchers); 3904 if (ret) 3905 return ret; 3906 3907 sscanf(locker->id.cookie, RBD_LOCK_COOKIE_PREFIX " %llu", &cookie); 3908 for (i = 0; i < num_watchers; i++) { 3909 /* 3910 * Ignore addr->type while comparing. This mimics 3911 * entity_addr_t::get_legacy_str() + strcmp(). 3912 */ 3913 if (ceph_addr_equal_no_type(&watchers[i].addr, 3914 &locker->info.addr) && 3915 watchers[i].cookie == cookie) { 3916 struct rbd_client_id cid = { 3917 .gid = le64_to_cpu(watchers[i].name.num), 3918 .handle = cookie, 3919 }; 3920 3921 dout("%s rbd_dev %p found cid %llu-%llu\n", __func__, 3922 rbd_dev, cid.gid, cid.handle); 3923 rbd_set_owner_cid(rbd_dev, &cid); 3924 ret = 1; 3925 goto out; 3926 } 3927 } 3928 3929 dout("%s rbd_dev %p no watchers\n", __func__, rbd_dev); 3930 ret = 0; 3931 out: 3932 kfree(watchers); 3933 return ret; 3934 } 3935 3936 /* 3937 * lock_rwsem must be held for write 3938 */ 3939 static int rbd_try_lock(struct rbd_device *rbd_dev) 3940 { 3941 struct ceph_client *client = rbd_dev->rbd_client->client; 3942 struct ceph_locker *lockers; 3943 u32 num_lockers; 3944 int ret; 3945 3946 for (;;) { 3947 ret = rbd_lock(rbd_dev); 3948 if (ret != -EBUSY) 3949 return ret; 3950 3951 /* determine if the current lock holder is still alive */ 3952 ret = get_lock_owner_info(rbd_dev, &lockers, &num_lockers); 3953 if (ret) 3954 return ret; 3955 3956 if (num_lockers == 0) 3957 goto again; 3958 3959 ret = find_watcher(rbd_dev, lockers); 3960 if (ret) 3961 goto out; /* request lock or error */ 3962 3963 rbd_warn(rbd_dev, "breaking header lock owned by %s%llu", 3964 ENTITY_NAME(lockers[0].id.name)); 3965 3966 ret = ceph_monc_blocklist_add(&client->monc, 3967 &lockers[0].info.addr); 3968 if (ret) { 3969 rbd_warn(rbd_dev, "blocklist of %s%llu failed: %d", 3970 ENTITY_NAME(lockers[0].id.name), ret); 3971 goto out; 3972 } 3973 3974 ret = ceph_cls_break_lock(&client->osdc, &rbd_dev->header_oid, 3975 &rbd_dev->header_oloc, RBD_LOCK_NAME, 3976 lockers[0].id.cookie, 3977 &lockers[0].id.name); 3978 if (ret && ret != -ENOENT) 3979 goto out; 3980 3981 again: 3982 ceph_free_lockers(lockers, num_lockers); 3983 } 3984 3985 out: 3986 ceph_free_lockers(lockers, num_lockers); 3987 return ret; 3988 } 3989 3990 static int rbd_post_acquire_action(struct rbd_device *rbd_dev) 3991 { 3992 int ret; 3993 3994 if (rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP) { 3995 ret = rbd_object_map_open(rbd_dev); 3996 if (ret) 3997 return ret; 3998 } 3999 4000 return 0; 4001 } 4002 4003 /* 4004 * Return: 4005 * 0 - lock acquired 4006 * 1 - caller should call rbd_request_lock() 4007 * <0 - error 4008 */ 4009 static int rbd_try_acquire_lock(struct rbd_device *rbd_dev) 4010 { 4011 int ret; 4012 4013 down_read(&rbd_dev->lock_rwsem); 4014 dout("%s rbd_dev %p read lock_state %d\n", __func__, rbd_dev, 4015 rbd_dev->lock_state); 4016 if (__rbd_is_lock_owner(rbd_dev)) { 4017 up_read(&rbd_dev->lock_rwsem); 4018 return 0; 4019 } 4020 4021 up_read(&rbd_dev->lock_rwsem); 4022 down_write(&rbd_dev->lock_rwsem); 4023 dout("%s rbd_dev %p write lock_state %d\n", __func__, rbd_dev, 4024 rbd_dev->lock_state); 4025 if (__rbd_is_lock_owner(rbd_dev)) { 4026 up_write(&rbd_dev->lock_rwsem); 4027 return 0; 4028 } 4029 4030 ret = rbd_try_lock(rbd_dev); 4031 if (ret < 0) { 4032 rbd_warn(rbd_dev, "failed to lock header: %d", ret); 4033 if (ret == -EBLOCKLISTED) 4034 goto out; 4035 4036 ret = 1; /* request lock anyway */ 4037 } 4038 if (ret > 0) { 4039 up_write(&rbd_dev->lock_rwsem); 4040 return ret; 4041 } 4042 4043 rbd_assert(rbd_dev->lock_state == RBD_LOCK_STATE_LOCKED); 4044 rbd_assert(list_empty(&rbd_dev->running_list)); 4045 4046 ret = rbd_post_acquire_action(rbd_dev); 4047 if (ret) { 4048 rbd_warn(rbd_dev, "post-acquire action failed: %d", ret); 4049 /* 4050 * Can't stay in RBD_LOCK_STATE_LOCKED because 4051 * rbd_lock_add_request() would let the request through, 4052 * assuming that e.g. object map is locked and loaded. 4053 */ 4054 rbd_unlock(rbd_dev); 4055 } 4056 4057 out: 4058 wake_lock_waiters(rbd_dev, ret); 4059 up_write(&rbd_dev->lock_rwsem); 4060 return ret; 4061 } 4062 4063 static void rbd_acquire_lock(struct work_struct *work) 4064 { 4065 struct rbd_device *rbd_dev = container_of(to_delayed_work(work), 4066 struct rbd_device, lock_dwork); 4067 int ret; 4068 4069 dout("%s rbd_dev %p\n", __func__, rbd_dev); 4070 again: 4071 ret = rbd_try_acquire_lock(rbd_dev); 4072 if (ret <= 0) { 4073 dout("%s rbd_dev %p ret %d - done\n", __func__, rbd_dev, ret); 4074 return; 4075 } 4076 4077 ret = rbd_request_lock(rbd_dev); 4078 if (ret == -ETIMEDOUT) { 4079 goto again; /* treat this as a dead client */ 4080 } else if (ret == -EROFS) { 4081 rbd_warn(rbd_dev, "peer will not release lock"); 4082 down_write(&rbd_dev->lock_rwsem); 4083 wake_lock_waiters(rbd_dev, ret); 4084 up_write(&rbd_dev->lock_rwsem); 4085 } else if (ret < 0) { 4086 rbd_warn(rbd_dev, "error requesting lock: %d", ret); 4087 mod_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork, 4088 RBD_RETRY_DELAY); 4089 } else { 4090 /* 4091 * lock owner acked, but resend if we don't see them 4092 * release the lock 4093 */ 4094 dout("%s rbd_dev %p requeuing lock_dwork\n", __func__, 4095 rbd_dev); 4096 mod_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork, 4097 msecs_to_jiffies(2 * RBD_NOTIFY_TIMEOUT * MSEC_PER_SEC)); 4098 } 4099 } 4100 4101 static bool rbd_quiesce_lock(struct rbd_device *rbd_dev) 4102 { 4103 bool need_wait; 4104 4105 dout("%s rbd_dev %p\n", __func__, rbd_dev); 4106 lockdep_assert_held_write(&rbd_dev->lock_rwsem); 4107 4108 if (rbd_dev->lock_state != RBD_LOCK_STATE_LOCKED) 4109 return false; 4110 4111 /* 4112 * Ensure that all in-flight IO is flushed. 4113 */ 4114 rbd_dev->lock_state = RBD_LOCK_STATE_RELEASING; 4115 rbd_assert(!completion_done(&rbd_dev->releasing_wait)); 4116 need_wait = !list_empty(&rbd_dev->running_list); 4117 downgrade_write(&rbd_dev->lock_rwsem); 4118 if (need_wait) 4119 wait_for_completion(&rbd_dev->releasing_wait); 4120 up_read(&rbd_dev->lock_rwsem); 4121 4122 down_write(&rbd_dev->lock_rwsem); 4123 if (rbd_dev->lock_state != RBD_LOCK_STATE_RELEASING) 4124 return false; 4125 4126 rbd_assert(list_empty(&rbd_dev->running_list)); 4127 return true; 4128 } 4129 4130 static void rbd_pre_release_action(struct rbd_device *rbd_dev) 4131 { 4132 if (rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP) 4133 rbd_object_map_close(rbd_dev); 4134 } 4135 4136 static void __rbd_release_lock(struct rbd_device *rbd_dev) 4137 { 4138 rbd_assert(list_empty(&rbd_dev->running_list)); 4139 4140 rbd_pre_release_action(rbd_dev); 4141 rbd_unlock(rbd_dev); 4142 } 4143 4144 /* 4145 * lock_rwsem must be held for write 4146 */ 4147 static void rbd_release_lock(struct rbd_device *rbd_dev) 4148 { 4149 if (!rbd_quiesce_lock(rbd_dev)) 4150 return; 4151 4152 __rbd_release_lock(rbd_dev); 4153 4154 /* 4155 * Give others a chance to grab the lock - we would re-acquire 4156 * almost immediately if we got new IO while draining the running 4157 * list otherwise. We need to ack our own notifications, so this 4158 * lock_dwork will be requeued from rbd_handle_released_lock() by 4159 * way of maybe_kick_acquire(). 4160 */ 4161 cancel_delayed_work(&rbd_dev->lock_dwork); 4162 } 4163 4164 static void rbd_release_lock_work(struct work_struct *work) 4165 { 4166 struct rbd_device *rbd_dev = container_of(work, struct rbd_device, 4167 unlock_work); 4168 4169 down_write(&rbd_dev->lock_rwsem); 4170 rbd_release_lock(rbd_dev); 4171 up_write(&rbd_dev->lock_rwsem); 4172 } 4173 4174 static void maybe_kick_acquire(struct rbd_device *rbd_dev) 4175 { 4176 bool have_requests; 4177 4178 dout("%s rbd_dev %p\n", __func__, rbd_dev); 4179 if (__rbd_is_lock_owner(rbd_dev)) 4180 return; 4181 4182 spin_lock(&rbd_dev->lock_lists_lock); 4183 have_requests = !list_empty(&rbd_dev->acquiring_list); 4184 spin_unlock(&rbd_dev->lock_lists_lock); 4185 if (have_requests || delayed_work_pending(&rbd_dev->lock_dwork)) { 4186 dout("%s rbd_dev %p kicking lock_dwork\n", __func__, rbd_dev); 4187 mod_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork, 0); 4188 } 4189 } 4190 4191 static void rbd_handle_acquired_lock(struct rbd_device *rbd_dev, u8 struct_v, 4192 void **p) 4193 { 4194 struct rbd_client_id cid = { 0 }; 4195 4196 if (struct_v >= 2) { 4197 cid.gid = ceph_decode_64(p); 4198 cid.handle = ceph_decode_64(p); 4199 } 4200 4201 dout("%s rbd_dev %p cid %llu-%llu\n", __func__, rbd_dev, cid.gid, 4202 cid.handle); 4203 if (!rbd_cid_equal(&cid, &rbd_empty_cid)) { 4204 down_write(&rbd_dev->lock_rwsem); 4205 if (rbd_cid_equal(&cid, &rbd_dev->owner_cid)) { 4206 /* 4207 * we already know that the remote client is 4208 * the owner 4209 */ 4210 up_write(&rbd_dev->lock_rwsem); 4211 return; 4212 } 4213 4214 rbd_set_owner_cid(rbd_dev, &cid); 4215 downgrade_write(&rbd_dev->lock_rwsem); 4216 } else { 4217 down_read(&rbd_dev->lock_rwsem); 4218 } 4219 4220 maybe_kick_acquire(rbd_dev); 4221 up_read(&rbd_dev->lock_rwsem); 4222 } 4223 4224 static void rbd_handle_released_lock(struct rbd_device *rbd_dev, u8 struct_v, 4225 void **p) 4226 { 4227 struct rbd_client_id cid = { 0 }; 4228 4229 if (struct_v >= 2) { 4230 cid.gid = ceph_decode_64(p); 4231 cid.handle = ceph_decode_64(p); 4232 } 4233 4234 dout("%s rbd_dev %p cid %llu-%llu\n", __func__, rbd_dev, cid.gid, 4235 cid.handle); 4236 if (!rbd_cid_equal(&cid, &rbd_empty_cid)) { 4237 down_write(&rbd_dev->lock_rwsem); 4238 if (!rbd_cid_equal(&cid, &rbd_dev->owner_cid)) { 4239 dout("%s rbd_dev %p unexpected owner, cid %llu-%llu != owner_cid %llu-%llu\n", 4240 __func__, rbd_dev, cid.gid, cid.handle, 4241 rbd_dev->owner_cid.gid, rbd_dev->owner_cid.handle); 4242 up_write(&rbd_dev->lock_rwsem); 4243 return; 4244 } 4245 4246 rbd_set_owner_cid(rbd_dev, &rbd_empty_cid); 4247 downgrade_write(&rbd_dev->lock_rwsem); 4248 } else { 4249 down_read(&rbd_dev->lock_rwsem); 4250 } 4251 4252 maybe_kick_acquire(rbd_dev); 4253 up_read(&rbd_dev->lock_rwsem); 4254 } 4255 4256 /* 4257 * Returns result for ResponseMessage to be encoded (<= 0), or 1 if no 4258 * ResponseMessage is needed. 4259 */ 4260 static int rbd_handle_request_lock(struct rbd_device *rbd_dev, u8 struct_v, 4261 void **p) 4262 { 4263 struct rbd_client_id my_cid = rbd_get_cid(rbd_dev); 4264 struct rbd_client_id cid = { 0 }; 4265 int result = 1; 4266 4267 if (struct_v >= 2) { 4268 cid.gid = ceph_decode_64(p); 4269 cid.handle = ceph_decode_64(p); 4270 } 4271 4272 dout("%s rbd_dev %p cid %llu-%llu\n", __func__, rbd_dev, cid.gid, 4273 cid.handle); 4274 if (rbd_cid_equal(&cid, &my_cid)) 4275 return result; 4276 4277 down_read(&rbd_dev->lock_rwsem); 4278 if (__rbd_is_lock_owner(rbd_dev)) { 4279 if (rbd_dev->lock_state == RBD_LOCK_STATE_LOCKED && 4280 rbd_cid_equal(&rbd_dev->owner_cid, &rbd_empty_cid)) 4281 goto out_unlock; 4282 4283 /* 4284 * encode ResponseMessage(0) so the peer can detect 4285 * a missing owner 4286 */ 4287 result = 0; 4288 4289 if (rbd_dev->lock_state == RBD_LOCK_STATE_LOCKED) { 4290 if (!rbd_dev->opts->exclusive) { 4291 dout("%s rbd_dev %p queueing unlock_work\n", 4292 __func__, rbd_dev); 4293 queue_work(rbd_dev->task_wq, 4294 &rbd_dev->unlock_work); 4295 } else { 4296 /* refuse to release the lock */ 4297 result = -EROFS; 4298 } 4299 } 4300 } 4301 4302 out_unlock: 4303 up_read(&rbd_dev->lock_rwsem); 4304 return result; 4305 } 4306 4307 static void __rbd_acknowledge_notify(struct rbd_device *rbd_dev, 4308 u64 notify_id, u64 cookie, s32 *result) 4309 { 4310 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 4311 char buf[4 + CEPH_ENCODING_START_BLK_LEN]; 4312 int buf_size = sizeof(buf); 4313 int ret; 4314 4315 if (result) { 4316 void *p = buf; 4317 4318 /* encode ResponseMessage */ 4319 ceph_start_encoding(&p, 1, 1, 4320 buf_size - CEPH_ENCODING_START_BLK_LEN); 4321 ceph_encode_32(&p, *result); 4322 } else { 4323 buf_size = 0; 4324 } 4325 4326 ret = ceph_osdc_notify_ack(osdc, &rbd_dev->header_oid, 4327 &rbd_dev->header_oloc, notify_id, cookie, 4328 buf, buf_size); 4329 if (ret) 4330 rbd_warn(rbd_dev, "acknowledge_notify failed: %d", ret); 4331 } 4332 4333 static void rbd_acknowledge_notify(struct rbd_device *rbd_dev, u64 notify_id, 4334 u64 cookie) 4335 { 4336 dout("%s rbd_dev %p\n", __func__, rbd_dev); 4337 __rbd_acknowledge_notify(rbd_dev, notify_id, cookie, NULL); 4338 } 4339 4340 static void rbd_acknowledge_notify_result(struct rbd_device *rbd_dev, 4341 u64 notify_id, u64 cookie, s32 result) 4342 { 4343 dout("%s rbd_dev %p result %d\n", __func__, rbd_dev, result); 4344 __rbd_acknowledge_notify(rbd_dev, notify_id, cookie, &result); 4345 } 4346 4347 static void rbd_watch_cb(void *arg, u64 notify_id, u64 cookie, 4348 u64 notifier_id, void *data, size_t data_len) 4349 { 4350 struct rbd_device *rbd_dev = arg; 4351 void *p = data; 4352 void *const end = p + data_len; 4353 u8 struct_v = 0; 4354 u32 len; 4355 u32 notify_op; 4356 int ret; 4357 4358 dout("%s rbd_dev %p cookie %llu notify_id %llu data_len %zu\n", 4359 __func__, rbd_dev, cookie, notify_id, data_len); 4360 if (data_len) { 4361 ret = ceph_start_decoding(&p, end, 1, "NotifyMessage", 4362 &struct_v, &len); 4363 if (ret) { 4364 rbd_warn(rbd_dev, "failed to decode NotifyMessage: %d", 4365 ret); 4366 return; 4367 } 4368 4369 notify_op = ceph_decode_32(&p); 4370 } else { 4371 /* legacy notification for header updates */ 4372 notify_op = RBD_NOTIFY_OP_HEADER_UPDATE; 4373 len = 0; 4374 } 4375 4376 dout("%s rbd_dev %p notify_op %u\n", __func__, rbd_dev, notify_op); 4377 switch (notify_op) { 4378 case RBD_NOTIFY_OP_ACQUIRED_LOCK: 4379 rbd_handle_acquired_lock(rbd_dev, struct_v, &p); 4380 rbd_acknowledge_notify(rbd_dev, notify_id, cookie); 4381 break; 4382 case RBD_NOTIFY_OP_RELEASED_LOCK: 4383 rbd_handle_released_lock(rbd_dev, struct_v, &p); 4384 rbd_acknowledge_notify(rbd_dev, notify_id, cookie); 4385 break; 4386 case RBD_NOTIFY_OP_REQUEST_LOCK: 4387 ret = rbd_handle_request_lock(rbd_dev, struct_v, &p); 4388 if (ret <= 0) 4389 rbd_acknowledge_notify_result(rbd_dev, notify_id, 4390 cookie, ret); 4391 else 4392 rbd_acknowledge_notify(rbd_dev, notify_id, cookie); 4393 break; 4394 case RBD_NOTIFY_OP_HEADER_UPDATE: 4395 ret = rbd_dev_refresh(rbd_dev); 4396 if (ret) 4397 rbd_warn(rbd_dev, "refresh failed: %d", ret); 4398 4399 rbd_acknowledge_notify(rbd_dev, notify_id, cookie); 4400 break; 4401 default: 4402 if (rbd_is_lock_owner(rbd_dev)) 4403 rbd_acknowledge_notify_result(rbd_dev, notify_id, 4404 cookie, -EOPNOTSUPP); 4405 else 4406 rbd_acknowledge_notify(rbd_dev, notify_id, cookie); 4407 break; 4408 } 4409 } 4410 4411 static void __rbd_unregister_watch(struct rbd_device *rbd_dev); 4412 4413 static void rbd_watch_errcb(void *arg, u64 cookie, int err) 4414 { 4415 struct rbd_device *rbd_dev = arg; 4416 4417 rbd_warn(rbd_dev, "encountered watch error: %d", err); 4418 4419 down_write(&rbd_dev->lock_rwsem); 4420 rbd_set_owner_cid(rbd_dev, &rbd_empty_cid); 4421 up_write(&rbd_dev->lock_rwsem); 4422 4423 mutex_lock(&rbd_dev->watch_mutex); 4424 if (rbd_dev->watch_state == RBD_WATCH_STATE_REGISTERED) { 4425 __rbd_unregister_watch(rbd_dev); 4426 rbd_dev->watch_state = RBD_WATCH_STATE_ERROR; 4427 4428 queue_delayed_work(rbd_dev->task_wq, &rbd_dev->watch_dwork, 0); 4429 } 4430 mutex_unlock(&rbd_dev->watch_mutex); 4431 } 4432 4433 /* 4434 * watch_mutex must be locked 4435 */ 4436 static int __rbd_register_watch(struct rbd_device *rbd_dev) 4437 { 4438 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 4439 struct ceph_osd_linger_request *handle; 4440 4441 rbd_assert(!rbd_dev->watch_handle); 4442 dout("%s rbd_dev %p\n", __func__, rbd_dev); 4443 4444 handle = ceph_osdc_watch(osdc, &rbd_dev->header_oid, 4445 &rbd_dev->header_oloc, rbd_watch_cb, 4446 rbd_watch_errcb, rbd_dev); 4447 if (IS_ERR(handle)) 4448 return PTR_ERR(handle); 4449 4450 rbd_dev->watch_handle = handle; 4451 return 0; 4452 } 4453 4454 /* 4455 * watch_mutex must be locked 4456 */ 4457 static void __rbd_unregister_watch(struct rbd_device *rbd_dev) 4458 { 4459 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 4460 int ret; 4461 4462 rbd_assert(rbd_dev->watch_handle); 4463 dout("%s rbd_dev %p\n", __func__, rbd_dev); 4464 4465 ret = ceph_osdc_unwatch(osdc, rbd_dev->watch_handle); 4466 if (ret) 4467 rbd_warn(rbd_dev, "failed to unwatch: %d", ret); 4468 4469 rbd_dev->watch_handle = NULL; 4470 } 4471 4472 static int rbd_register_watch(struct rbd_device *rbd_dev) 4473 { 4474 int ret; 4475 4476 mutex_lock(&rbd_dev->watch_mutex); 4477 rbd_assert(rbd_dev->watch_state == RBD_WATCH_STATE_UNREGISTERED); 4478 ret = __rbd_register_watch(rbd_dev); 4479 if (ret) 4480 goto out; 4481 4482 rbd_dev->watch_state = RBD_WATCH_STATE_REGISTERED; 4483 rbd_dev->watch_cookie = rbd_dev->watch_handle->linger_id; 4484 4485 out: 4486 mutex_unlock(&rbd_dev->watch_mutex); 4487 return ret; 4488 } 4489 4490 static void cancel_tasks_sync(struct rbd_device *rbd_dev) 4491 { 4492 dout("%s rbd_dev %p\n", __func__, rbd_dev); 4493 4494 cancel_work_sync(&rbd_dev->acquired_lock_work); 4495 cancel_work_sync(&rbd_dev->released_lock_work); 4496 cancel_delayed_work_sync(&rbd_dev->lock_dwork); 4497 cancel_work_sync(&rbd_dev->unlock_work); 4498 } 4499 4500 /* 4501 * header_rwsem must not be held to avoid a deadlock with 4502 * rbd_dev_refresh() when flushing notifies. 4503 */ 4504 static void rbd_unregister_watch(struct rbd_device *rbd_dev) 4505 { 4506 cancel_tasks_sync(rbd_dev); 4507 4508 mutex_lock(&rbd_dev->watch_mutex); 4509 if (rbd_dev->watch_state == RBD_WATCH_STATE_REGISTERED) 4510 __rbd_unregister_watch(rbd_dev); 4511 rbd_dev->watch_state = RBD_WATCH_STATE_UNREGISTERED; 4512 mutex_unlock(&rbd_dev->watch_mutex); 4513 4514 cancel_delayed_work_sync(&rbd_dev->watch_dwork); 4515 ceph_osdc_flush_notifies(&rbd_dev->rbd_client->client->osdc); 4516 } 4517 4518 /* 4519 * lock_rwsem must be held for write 4520 */ 4521 static void rbd_reacquire_lock(struct rbd_device *rbd_dev) 4522 { 4523 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 4524 char cookie[32]; 4525 int ret; 4526 4527 if (!rbd_quiesce_lock(rbd_dev)) 4528 return; 4529 4530 format_lock_cookie(rbd_dev, cookie); 4531 ret = ceph_cls_set_cookie(osdc, &rbd_dev->header_oid, 4532 &rbd_dev->header_oloc, RBD_LOCK_NAME, 4533 CEPH_CLS_LOCK_EXCLUSIVE, rbd_dev->lock_cookie, 4534 RBD_LOCK_TAG, cookie); 4535 if (ret) { 4536 if (ret != -EOPNOTSUPP) 4537 rbd_warn(rbd_dev, "failed to update lock cookie: %d", 4538 ret); 4539 4540 /* 4541 * Lock cookie cannot be updated on older OSDs, so do 4542 * a manual release and queue an acquire. 4543 */ 4544 __rbd_release_lock(rbd_dev); 4545 queue_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork, 0); 4546 } else { 4547 __rbd_lock(rbd_dev, cookie); 4548 wake_lock_waiters(rbd_dev, 0); 4549 } 4550 } 4551 4552 static void rbd_reregister_watch(struct work_struct *work) 4553 { 4554 struct rbd_device *rbd_dev = container_of(to_delayed_work(work), 4555 struct rbd_device, watch_dwork); 4556 int ret; 4557 4558 dout("%s rbd_dev %p\n", __func__, rbd_dev); 4559 4560 mutex_lock(&rbd_dev->watch_mutex); 4561 if (rbd_dev->watch_state != RBD_WATCH_STATE_ERROR) { 4562 mutex_unlock(&rbd_dev->watch_mutex); 4563 return; 4564 } 4565 4566 ret = __rbd_register_watch(rbd_dev); 4567 if (ret) { 4568 rbd_warn(rbd_dev, "failed to reregister watch: %d", ret); 4569 if (ret != -EBLOCKLISTED && ret != -ENOENT) { 4570 queue_delayed_work(rbd_dev->task_wq, 4571 &rbd_dev->watch_dwork, 4572 RBD_RETRY_DELAY); 4573 mutex_unlock(&rbd_dev->watch_mutex); 4574 return; 4575 } 4576 4577 mutex_unlock(&rbd_dev->watch_mutex); 4578 down_write(&rbd_dev->lock_rwsem); 4579 wake_lock_waiters(rbd_dev, ret); 4580 up_write(&rbd_dev->lock_rwsem); 4581 return; 4582 } 4583 4584 rbd_dev->watch_state = RBD_WATCH_STATE_REGISTERED; 4585 rbd_dev->watch_cookie = rbd_dev->watch_handle->linger_id; 4586 mutex_unlock(&rbd_dev->watch_mutex); 4587 4588 down_write(&rbd_dev->lock_rwsem); 4589 if (rbd_dev->lock_state == RBD_LOCK_STATE_LOCKED) 4590 rbd_reacquire_lock(rbd_dev); 4591 up_write(&rbd_dev->lock_rwsem); 4592 4593 ret = rbd_dev_refresh(rbd_dev); 4594 if (ret) 4595 rbd_warn(rbd_dev, "reregistration refresh failed: %d", ret); 4596 } 4597 4598 /* 4599 * Synchronous osd object method call. Returns the number of bytes 4600 * returned in the outbound buffer, or a negative error code. 4601 */ 4602 static int rbd_obj_method_sync(struct rbd_device *rbd_dev, 4603 struct ceph_object_id *oid, 4604 struct ceph_object_locator *oloc, 4605 const char *method_name, 4606 const void *outbound, 4607 size_t outbound_size, 4608 void *inbound, 4609 size_t inbound_size) 4610 { 4611 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 4612 struct page *req_page = NULL; 4613 struct page *reply_page; 4614 int ret; 4615 4616 /* 4617 * Method calls are ultimately read operations. The result 4618 * should placed into the inbound buffer provided. They 4619 * also supply outbound data--parameters for the object 4620 * method. Currently if this is present it will be a 4621 * snapshot id. 4622 */ 4623 if (outbound) { 4624 if (outbound_size > PAGE_SIZE) 4625 return -E2BIG; 4626 4627 req_page = alloc_page(GFP_KERNEL); 4628 if (!req_page) 4629 return -ENOMEM; 4630 4631 memcpy(page_address(req_page), outbound, outbound_size); 4632 } 4633 4634 reply_page = alloc_page(GFP_KERNEL); 4635 if (!reply_page) { 4636 if (req_page) 4637 __free_page(req_page); 4638 return -ENOMEM; 4639 } 4640 4641 ret = ceph_osdc_call(osdc, oid, oloc, RBD_DRV_NAME, method_name, 4642 CEPH_OSD_FLAG_READ, req_page, outbound_size, 4643 &reply_page, &inbound_size); 4644 if (!ret) { 4645 memcpy(inbound, page_address(reply_page), inbound_size); 4646 ret = inbound_size; 4647 } 4648 4649 if (req_page) 4650 __free_page(req_page); 4651 __free_page(reply_page); 4652 return ret; 4653 } 4654 4655 static void rbd_queue_workfn(struct work_struct *work) 4656 { 4657 struct rbd_img_request *img_request = 4658 container_of(work, struct rbd_img_request, work); 4659 struct rbd_device *rbd_dev = img_request->rbd_dev; 4660 enum obj_operation_type op_type = img_request->op_type; 4661 struct request *rq = blk_mq_rq_from_pdu(img_request); 4662 u64 offset = (u64)blk_rq_pos(rq) << SECTOR_SHIFT; 4663 u64 length = blk_rq_bytes(rq); 4664 u64 mapping_size; 4665 int result; 4666 4667 /* Ignore/skip any zero-length requests */ 4668 if (!length) { 4669 dout("%s: zero-length request\n", __func__); 4670 result = 0; 4671 goto err_img_request; 4672 } 4673 4674 blk_mq_start_request(rq); 4675 4676 down_read(&rbd_dev->header_rwsem); 4677 mapping_size = rbd_dev->mapping.size; 4678 rbd_img_capture_header(img_request); 4679 up_read(&rbd_dev->header_rwsem); 4680 4681 if (offset + length > mapping_size) { 4682 rbd_warn(rbd_dev, "beyond EOD (%llu~%llu > %llu)", offset, 4683 length, mapping_size); 4684 result = -EIO; 4685 goto err_img_request; 4686 } 4687 4688 dout("%s rbd_dev %p img_req %p %s %llu~%llu\n", __func__, rbd_dev, 4689 img_request, obj_op_name(op_type), offset, length); 4690 4691 if (op_type == OBJ_OP_DISCARD || op_type == OBJ_OP_ZEROOUT) 4692 result = rbd_img_fill_nodata(img_request, offset, length); 4693 else 4694 result = rbd_img_fill_from_bio(img_request, offset, length, 4695 rq->bio); 4696 if (result) 4697 goto err_img_request; 4698 4699 rbd_img_handle_request(img_request, 0); 4700 return; 4701 4702 err_img_request: 4703 rbd_img_request_destroy(img_request); 4704 if (result) 4705 rbd_warn(rbd_dev, "%s %llx at %llx result %d", 4706 obj_op_name(op_type), length, offset, result); 4707 blk_mq_end_request(rq, errno_to_blk_status(result)); 4708 } 4709 4710 static blk_status_t rbd_queue_rq(struct blk_mq_hw_ctx *hctx, 4711 const struct blk_mq_queue_data *bd) 4712 { 4713 struct rbd_device *rbd_dev = hctx->queue->queuedata; 4714 struct rbd_img_request *img_req = blk_mq_rq_to_pdu(bd->rq); 4715 enum obj_operation_type op_type; 4716 4717 switch (req_op(bd->rq)) { 4718 case REQ_OP_DISCARD: 4719 op_type = OBJ_OP_DISCARD; 4720 break; 4721 case REQ_OP_WRITE_ZEROES: 4722 op_type = OBJ_OP_ZEROOUT; 4723 break; 4724 case REQ_OP_WRITE: 4725 op_type = OBJ_OP_WRITE; 4726 break; 4727 case REQ_OP_READ: 4728 op_type = OBJ_OP_READ; 4729 break; 4730 default: 4731 rbd_warn(rbd_dev, "unknown req_op %d", req_op(bd->rq)); 4732 return BLK_STS_IOERR; 4733 } 4734 4735 rbd_img_request_init(img_req, rbd_dev, op_type); 4736 4737 if (rbd_img_is_write(img_req)) { 4738 if (rbd_is_ro(rbd_dev)) { 4739 rbd_warn(rbd_dev, "%s on read-only mapping", 4740 obj_op_name(img_req->op_type)); 4741 return BLK_STS_IOERR; 4742 } 4743 rbd_assert(!rbd_is_snap(rbd_dev)); 4744 } 4745 4746 INIT_WORK(&img_req->work, rbd_queue_workfn); 4747 queue_work(rbd_wq, &img_req->work); 4748 return BLK_STS_OK; 4749 } 4750 4751 static void rbd_free_disk(struct rbd_device *rbd_dev) 4752 { 4753 blk_cleanup_disk(rbd_dev->disk); 4754 blk_mq_free_tag_set(&rbd_dev->tag_set); 4755 rbd_dev->disk = NULL; 4756 } 4757 4758 static int rbd_obj_read_sync(struct rbd_device *rbd_dev, 4759 struct ceph_object_id *oid, 4760 struct ceph_object_locator *oloc, 4761 void *buf, int buf_len) 4762 4763 { 4764 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 4765 struct ceph_osd_request *req; 4766 struct page **pages; 4767 int num_pages = calc_pages_for(0, buf_len); 4768 int ret; 4769 4770 req = ceph_osdc_alloc_request(osdc, NULL, 1, false, GFP_KERNEL); 4771 if (!req) 4772 return -ENOMEM; 4773 4774 ceph_oid_copy(&req->r_base_oid, oid); 4775 ceph_oloc_copy(&req->r_base_oloc, oloc); 4776 req->r_flags = CEPH_OSD_FLAG_READ; 4777 4778 pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL); 4779 if (IS_ERR(pages)) { 4780 ret = PTR_ERR(pages); 4781 goto out_req; 4782 } 4783 4784 osd_req_op_extent_init(req, 0, CEPH_OSD_OP_READ, 0, buf_len, 0, 0); 4785 osd_req_op_extent_osd_data_pages(req, 0, pages, buf_len, 0, false, 4786 true); 4787 4788 ret = ceph_osdc_alloc_messages(req, GFP_KERNEL); 4789 if (ret) 4790 goto out_req; 4791 4792 ceph_osdc_start_request(osdc, req, false); 4793 ret = ceph_osdc_wait_request(osdc, req); 4794 if (ret >= 0) 4795 ceph_copy_from_page_vector(pages, buf, 0, ret); 4796 4797 out_req: 4798 ceph_osdc_put_request(req); 4799 return ret; 4800 } 4801 4802 /* 4803 * Read the complete header for the given rbd device. On successful 4804 * return, the rbd_dev->header field will contain up-to-date 4805 * information about the image. 4806 */ 4807 static int rbd_dev_v1_header_info(struct rbd_device *rbd_dev) 4808 { 4809 struct rbd_image_header_ondisk *ondisk = NULL; 4810 u32 snap_count = 0; 4811 u64 names_size = 0; 4812 u32 want_count; 4813 int ret; 4814 4815 /* 4816 * The complete header will include an array of its 64-bit 4817 * snapshot ids, followed by the names of those snapshots as 4818 * a contiguous block of NUL-terminated strings. Note that 4819 * the number of snapshots could change by the time we read 4820 * it in, in which case we re-read it. 4821 */ 4822 do { 4823 size_t size; 4824 4825 kfree(ondisk); 4826 4827 size = sizeof (*ondisk); 4828 size += snap_count * sizeof (struct rbd_image_snap_ondisk); 4829 size += names_size; 4830 ondisk = kmalloc(size, GFP_KERNEL); 4831 if (!ondisk) 4832 return -ENOMEM; 4833 4834 ret = rbd_obj_read_sync(rbd_dev, &rbd_dev->header_oid, 4835 &rbd_dev->header_oloc, ondisk, size); 4836 if (ret < 0) 4837 goto out; 4838 if ((size_t)ret < size) { 4839 ret = -ENXIO; 4840 rbd_warn(rbd_dev, "short header read (want %zd got %d)", 4841 size, ret); 4842 goto out; 4843 } 4844 if (!rbd_dev_ondisk_valid(ondisk)) { 4845 ret = -ENXIO; 4846 rbd_warn(rbd_dev, "invalid header"); 4847 goto out; 4848 } 4849 4850 names_size = le64_to_cpu(ondisk->snap_names_len); 4851 want_count = snap_count; 4852 snap_count = le32_to_cpu(ondisk->snap_count); 4853 } while (snap_count != want_count); 4854 4855 ret = rbd_header_from_disk(rbd_dev, ondisk); 4856 out: 4857 kfree(ondisk); 4858 4859 return ret; 4860 } 4861 4862 static void rbd_dev_update_size(struct rbd_device *rbd_dev) 4863 { 4864 sector_t size; 4865 4866 /* 4867 * If EXISTS is not set, rbd_dev->disk may be NULL, so don't 4868 * try to update its size. If REMOVING is set, updating size 4869 * is just useless work since the device can't be opened. 4870 */ 4871 if (test_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags) && 4872 !test_bit(RBD_DEV_FLAG_REMOVING, &rbd_dev->flags)) { 4873 size = (sector_t)rbd_dev->mapping.size / SECTOR_SIZE; 4874 dout("setting size to %llu sectors", (unsigned long long)size); 4875 set_capacity_and_notify(rbd_dev->disk, size); 4876 } 4877 } 4878 4879 static int rbd_dev_refresh(struct rbd_device *rbd_dev) 4880 { 4881 u64 mapping_size; 4882 int ret; 4883 4884 down_write(&rbd_dev->header_rwsem); 4885 mapping_size = rbd_dev->mapping.size; 4886 4887 ret = rbd_dev_header_info(rbd_dev); 4888 if (ret) 4889 goto out; 4890 4891 /* 4892 * If there is a parent, see if it has disappeared due to the 4893 * mapped image getting flattened. 4894 */ 4895 if (rbd_dev->parent) { 4896 ret = rbd_dev_v2_parent_info(rbd_dev); 4897 if (ret) 4898 goto out; 4899 } 4900 4901 rbd_assert(!rbd_is_snap(rbd_dev)); 4902 rbd_dev->mapping.size = rbd_dev->header.image_size; 4903 4904 out: 4905 up_write(&rbd_dev->header_rwsem); 4906 if (!ret && mapping_size != rbd_dev->mapping.size) 4907 rbd_dev_update_size(rbd_dev); 4908 4909 return ret; 4910 } 4911 4912 static const struct blk_mq_ops rbd_mq_ops = { 4913 .queue_rq = rbd_queue_rq, 4914 }; 4915 4916 static int rbd_init_disk(struct rbd_device *rbd_dev) 4917 { 4918 struct gendisk *disk; 4919 struct request_queue *q; 4920 unsigned int objset_bytes = 4921 rbd_dev->layout.object_size * rbd_dev->layout.stripe_count; 4922 int err; 4923 4924 memset(&rbd_dev->tag_set, 0, sizeof(rbd_dev->tag_set)); 4925 rbd_dev->tag_set.ops = &rbd_mq_ops; 4926 rbd_dev->tag_set.queue_depth = rbd_dev->opts->queue_depth; 4927 rbd_dev->tag_set.numa_node = NUMA_NO_NODE; 4928 rbd_dev->tag_set.flags = BLK_MQ_F_SHOULD_MERGE; 4929 rbd_dev->tag_set.nr_hw_queues = num_present_cpus(); 4930 rbd_dev->tag_set.cmd_size = sizeof(struct rbd_img_request); 4931 4932 err = blk_mq_alloc_tag_set(&rbd_dev->tag_set); 4933 if (err) 4934 return err; 4935 4936 disk = blk_mq_alloc_disk(&rbd_dev->tag_set, rbd_dev); 4937 if (IS_ERR(disk)) { 4938 err = PTR_ERR(disk); 4939 goto out_tag_set; 4940 } 4941 q = disk->queue; 4942 4943 snprintf(disk->disk_name, sizeof(disk->disk_name), RBD_DRV_NAME "%d", 4944 rbd_dev->dev_id); 4945 disk->major = rbd_dev->major; 4946 disk->first_minor = rbd_dev->minor; 4947 if (single_major) { 4948 disk->minors = (1 << RBD_SINGLE_MAJOR_PART_SHIFT); 4949 disk->flags |= GENHD_FL_EXT_DEVT; 4950 } else { 4951 disk->minors = RBD_MINORS_PER_MAJOR; 4952 } 4953 disk->fops = &rbd_bd_ops; 4954 4955 blk_queue_flag_set(QUEUE_FLAG_NONROT, q); 4956 /* QUEUE_FLAG_ADD_RANDOM is off by default for blk-mq */ 4957 4958 blk_queue_max_hw_sectors(q, objset_bytes >> SECTOR_SHIFT); 4959 q->limits.max_sectors = queue_max_hw_sectors(q); 4960 blk_queue_max_segments(q, USHRT_MAX); 4961 blk_queue_max_segment_size(q, UINT_MAX); 4962 blk_queue_io_min(q, rbd_dev->opts->alloc_size); 4963 blk_queue_io_opt(q, rbd_dev->opts->alloc_size); 4964 4965 if (rbd_dev->opts->trim) { 4966 blk_queue_flag_set(QUEUE_FLAG_DISCARD, q); 4967 q->limits.discard_granularity = rbd_dev->opts->alloc_size; 4968 blk_queue_max_discard_sectors(q, objset_bytes >> SECTOR_SHIFT); 4969 blk_queue_max_write_zeroes_sectors(q, objset_bytes >> SECTOR_SHIFT); 4970 } 4971 4972 if (!ceph_test_opt(rbd_dev->rbd_client->client, NOCRC)) 4973 blk_queue_flag_set(QUEUE_FLAG_STABLE_WRITES, q); 4974 4975 rbd_dev->disk = disk; 4976 4977 return 0; 4978 out_tag_set: 4979 blk_mq_free_tag_set(&rbd_dev->tag_set); 4980 return err; 4981 } 4982 4983 /* 4984 sysfs 4985 */ 4986 4987 static struct rbd_device *dev_to_rbd_dev(struct device *dev) 4988 { 4989 return container_of(dev, struct rbd_device, dev); 4990 } 4991 4992 static ssize_t rbd_size_show(struct device *dev, 4993 struct device_attribute *attr, char *buf) 4994 { 4995 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 4996 4997 return sprintf(buf, "%llu\n", 4998 (unsigned long long)rbd_dev->mapping.size); 4999 } 5000 5001 static ssize_t rbd_features_show(struct device *dev, 5002 struct device_attribute *attr, char *buf) 5003 { 5004 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5005 5006 return sprintf(buf, "0x%016llx\n", rbd_dev->header.features); 5007 } 5008 5009 static ssize_t rbd_major_show(struct device *dev, 5010 struct device_attribute *attr, char *buf) 5011 { 5012 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5013 5014 if (rbd_dev->major) 5015 return sprintf(buf, "%d\n", rbd_dev->major); 5016 5017 return sprintf(buf, "(none)\n"); 5018 } 5019 5020 static ssize_t rbd_minor_show(struct device *dev, 5021 struct device_attribute *attr, char *buf) 5022 { 5023 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5024 5025 return sprintf(buf, "%d\n", rbd_dev->minor); 5026 } 5027 5028 static ssize_t rbd_client_addr_show(struct device *dev, 5029 struct device_attribute *attr, char *buf) 5030 { 5031 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5032 struct ceph_entity_addr *client_addr = 5033 ceph_client_addr(rbd_dev->rbd_client->client); 5034 5035 return sprintf(buf, "%pISpc/%u\n", &client_addr->in_addr, 5036 le32_to_cpu(client_addr->nonce)); 5037 } 5038 5039 static ssize_t rbd_client_id_show(struct device *dev, 5040 struct device_attribute *attr, char *buf) 5041 { 5042 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5043 5044 return sprintf(buf, "client%lld\n", 5045 ceph_client_gid(rbd_dev->rbd_client->client)); 5046 } 5047 5048 static ssize_t rbd_cluster_fsid_show(struct device *dev, 5049 struct device_attribute *attr, char *buf) 5050 { 5051 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5052 5053 return sprintf(buf, "%pU\n", &rbd_dev->rbd_client->client->fsid); 5054 } 5055 5056 static ssize_t rbd_config_info_show(struct device *dev, 5057 struct device_attribute *attr, char *buf) 5058 { 5059 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5060 5061 if (!capable(CAP_SYS_ADMIN)) 5062 return -EPERM; 5063 5064 return sprintf(buf, "%s\n", rbd_dev->config_info); 5065 } 5066 5067 static ssize_t rbd_pool_show(struct device *dev, 5068 struct device_attribute *attr, char *buf) 5069 { 5070 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5071 5072 return sprintf(buf, "%s\n", rbd_dev->spec->pool_name); 5073 } 5074 5075 static ssize_t rbd_pool_id_show(struct device *dev, 5076 struct device_attribute *attr, char *buf) 5077 { 5078 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5079 5080 return sprintf(buf, "%llu\n", 5081 (unsigned long long) rbd_dev->spec->pool_id); 5082 } 5083 5084 static ssize_t rbd_pool_ns_show(struct device *dev, 5085 struct device_attribute *attr, char *buf) 5086 { 5087 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5088 5089 return sprintf(buf, "%s\n", rbd_dev->spec->pool_ns ?: ""); 5090 } 5091 5092 static ssize_t rbd_name_show(struct device *dev, 5093 struct device_attribute *attr, char *buf) 5094 { 5095 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5096 5097 if (rbd_dev->spec->image_name) 5098 return sprintf(buf, "%s\n", rbd_dev->spec->image_name); 5099 5100 return sprintf(buf, "(unknown)\n"); 5101 } 5102 5103 static ssize_t rbd_image_id_show(struct device *dev, 5104 struct device_attribute *attr, char *buf) 5105 { 5106 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5107 5108 return sprintf(buf, "%s\n", rbd_dev->spec->image_id); 5109 } 5110 5111 /* 5112 * Shows the name of the currently-mapped snapshot (or 5113 * RBD_SNAP_HEAD_NAME for the base image). 5114 */ 5115 static ssize_t rbd_snap_show(struct device *dev, 5116 struct device_attribute *attr, 5117 char *buf) 5118 { 5119 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5120 5121 return sprintf(buf, "%s\n", rbd_dev->spec->snap_name); 5122 } 5123 5124 static ssize_t rbd_snap_id_show(struct device *dev, 5125 struct device_attribute *attr, char *buf) 5126 { 5127 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5128 5129 return sprintf(buf, "%llu\n", rbd_dev->spec->snap_id); 5130 } 5131 5132 /* 5133 * For a v2 image, shows the chain of parent images, separated by empty 5134 * lines. For v1 images or if there is no parent, shows "(no parent 5135 * image)". 5136 */ 5137 static ssize_t rbd_parent_show(struct device *dev, 5138 struct device_attribute *attr, 5139 char *buf) 5140 { 5141 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5142 ssize_t count = 0; 5143 5144 if (!rbd_dev->parent) 5145 return sprintf(buf, "(no parent image)\n"); 5146 5147 for ( ; rbd_dev->parent; rbd_dev = rbd_dev->parent) { 5148 struct rbd_spec *spec = rbd_dev->parent_spec; 5149 5150 count += sprintf(&buf[count], "%s" 5151 "pool_id %llu\npool_name %s\n" 5152 "pool_ns %s\n" 5153 "image_id %s\nimage_name %s\n" 5154 "snap_id %llu\nsnap_name %s\n" 5155 "overlap %llu\n", 5156 !count ? "" : "\n", /* first? */ 5157 spec->pool_id, spec->pool_name, 5158 spec->pool_ns ?: "", 5159 spec->image_id, spec->image_name ?: "(unknown)", 5160 spec->snap_id, spec->snap_name, 5161 rbd_dev->parent_overlap); 5162 } 5163 5164 return count; 5165 } 5166 5167 static ssize_t rbd_image_refresh(struct device *dev, 5168 struct device_attribute *attr, 5169 const char *buf, 5170 size_t size) 5171 { 5172 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5173 int ret; 5174 5175 if (!capable(CAP_SYS_ADMIN)) 5176 return -EPERM; 5177 5178 ret = rbd_dev_refresh(rbd_dev); 5179 if (ret) 5180 return ret; 5181 5182 return size; 5183 } 5184 5185 static DEVICE_ATTR(size, 0444, rbd_size_show, NULL); 5186 static DEVICE_ATTR(features, 0444, rbd_features_show, NULL); 5187 static DEVICE_ATTR(major, 0444, rbd_major_show, NULL); 5188 static DEVICE_ATTR(minor, 0444, rbd_minor_show, NULL); 5189 static DEVICE_ATTR(client_addr, 0444, rbd_client_addr_show, NULL); 5190 static DEVICE_ATTR(client_id, 0444, rbd_client_id_show, NULL); 5191 static DEVICE_ATTR(cluster_fsid, 0444, rbd_cluster_fsid_show, NULL); 5192 static DEVICE_ATTR(config_info, 0400, rbd_config_info_show, NULL); 5193 static DEVICE_ATTR(pool, 0444, rbd_pool_show, NULL); 5194 static DEVICE_ATTR(pool_id, 0444, rbd_pool_id_show, NULL); 5195 static DEVICE_ATTR(pool_ns, 0444, rbd_pool_ns_show, NULL); 5196 static DEVICE_ATTR(name, 0444, rbd_name_show, NULL); 5197 static DEVICE_ATTR(image_id, 0444, rbd_image_id_show, NULL); 5198 static DEVICE_ATTR(refresh, 0200, NULL, rbd_image_refresh); 5199 static DEVICE_ATTR(current_snap, 0444, rbd_snap_show, NULL); 5200 static DEVICE_ATTR(snap_id, 0444, rbd_snap_id_show, NULL); 5201 static DEVICE_ATTR(parent, 0444, rbd_parent_show, NULL); 5202 5203 static struct attribute *rbd_attrs[] = { 5204 &dev_attr_size.attr, 5205 &dev_attr_features.attr, 5206 &dev_attr_major.attr, 5207 &dev_attr_minor.attr, 5208 &dev_attr_client_addr.attr, 5209 &dev_attr_client_id.attr, 5210 &dev_attr_cluster_fsid.attr, 5211 &dev_attr_config_info.attr, 5212 &dev_attr_pool.attr, 5213 &dev_attr_pool_id.attr, 5214 &dev_attr_pool_ns.attr, 5215 &dev_attr_name.attr, 5216 &dev_attr_image_id.attr, 5217 &dev_attr_current_snap.attr, 5218 &dev_attr_snap_id.attr, 5219 &dev_attr_parent.attr, 5220 &dev_attr_refresh.attr, 5221 NULL 5222 }; 5223 5224 static struct attribute_group rbd_attr_group = { 5225 .attrs = rbd_attrs, 5226 }; 5227 5228 static const struct attribute_group *rbd_attr_groups[] = { 5229 &rbd_attr_group, 5230 NULL 5231 }; 5232 5233 static void rbd_dev_release(struct device *dev); 5234 5235 static const struct device_type rbd_device_type = { 5236 .name = "rbd", 5237 .groups = rbd_attr_groups, 5238 .release = rbd_dev_release, 5239 }; 5240 5241 static struct rbd_spec *rbd_spec_get(struct rbd_spec *spec) 5242 { 5243 kref_get(&spec->kref); 5244 5245 return spec; 5246 } 5247 5248 static void rbd_spec_free(struct kref *kref); 5249 static void rbd_spec_put(struct rbd_spec *spec) 5250 { 5251 if (spec) 5252 kref_put(&spec->kref, rbd_spec_free); 5253 } 5254 5255 static struct rbd_spec *rbd_spec_alloc(void) 5256 { 5257 struct rbd_spec *spec; 5258 5259 spec = kzalloc(sizeof (*spec), GFP_KERNEL); 5260 if (!spec) 5261 return NULL; 5262 5263 spec->pool_id = CEPH_NOPOOL; 5264 spec->snap_id = CEPH_NOSNAP; 5265 kref_init(&spec->kref); 5266 5267 return spec; 5268 } 5269 5270 static void rbd_spec_free(struct kref *kref) 5271 { 5272 struct rbd_spec *spec = container_of(kref, struct rbd_spec, kref); 5273 5274 kfree(spec->pool_name); 5275 kfree(spec->pool_ns); 5276 kfree(spec->image_id); 5277 kfree(spec->image_name); 5278 kfree(spec->snap_name); 5279 kfree(spec); 5280 } 5281 5282 static void rbd_dev_free(struct rbd_device *rbd_dev) 5283 { 5284 WARN_ON(rbd_dev->watch_state != RBD_WATCH_STATE_UNREGISTERED); 5285 WARN_ON(rbd_dev->lock_state != RBD_LOCK_STATE_UNLOCKED); 5286 5287 ceph_oid_destroy(&rbd_dev->header_oid); 5288 ceph_oloc_destroy(&rbd_dev->header_oloc); 5289 kfree(rbd_dev->config_info); 5290 5291 rbd_put_client(rbd_dev->rbd_client); 5292 rbd_spec_put(rbd_dev->spec); 5293 kfree(rbd_dev->opts); 5294 kfree(rbd_dev); 5295 } 5296 5297 static void rbd_dev_release(struct device *dev) 5298 { 5299 struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); 5300 bool need_put = !!rbd_dev->opts; 5301 5302 if (need_put) { 5303 destroy_workqueue(rbd_dev->task_wq); 5304 ida_simple_remove(&rbd_dev_id_ida, rbd_dev->dev_id); 5305 } 5306 5307 rbd_dev_free(rbd_dev); 5308 5309 /* 5310 * This is racy, but way better than putting module outside of 5311 * the release callback. The race window is pretty small, so 5312 * doing something similar to dm (dm-builtin.c) is overkill. 5313 */ 5314 if (need_put) 5315 module_put(THIS_MODULE); 5316 } 5317 5318 static struct rbd_device *__rbd_dev_create(struct rbd_client *rbdc, 5319 struct rbd_spec *spec) 5320 { 5321 struct rbd_device *rbd_dev; 5322 5323 rbd_dev = kzalloc(sizeof(*rbd_dev), GFP_KERNEL); 5324 if (!rbd_dev) 5325 return NULL; 5326 5327 spin_lock_init(&rbd_dev->lock); 5328 INIT_LIST_HEAD(&rbd_dev->node); 5329 init_rwsem(&rbd_dev->header_rwsem); 5330 5331 rbd_dev->header.data_pool_id = CEPH_NOPOOL; 5332 ceph_oid_init(&rbd_dev->header_oid); 5333 rbd_dev->header_oloc.pool = spec->pool_id; 5334 if (spec->pool_ns) { 5335 WARN_ON(!*spec->pool_ns); 5336 rbd_dev->header_oloc.pool_ns = 5337 ceph_find_or_create_string(spec->pool_ns, 5338 strlen(spec->pool_ns)); 5339 } 5340 5341 mutex_init(&rbd_dev->watch_mutex); 5342 rbd_dev->watch_state = RBD_WATCH_STATE_UNREGISTERED; 5343 INIT_DELAYED_WORK(&rbd_dev->watch_dwork, rbd_reregister_watch); 5344 5345 init_rwsem(&rbd_dev->lock_rwsem); 5346 rbd_dev->lock_state = RBD_LOCK_STATE_UNLOCKED; 5347 INIT_WORK(&rbd_dev->acquired_lock_work, rbd_notify_acquired_lock); 5348 INIT_WORK(&rbd_dev->released_lock_work, rbd_notify_released_lock); 5349 INIT_DELAYED_WORK(&rbd_dev->lock_dwork, rbd_acquire_lock); 5350 INIT_WORK(&rbd_dev->unlock_work, rbd_release_lock_work); 5351 spin_lock_init(&rbd_dev->lock_lists_lock); 5352 INIT_LIST_HEAD(&rbd_dev->acquiring_list); 5353 INIT_LIST_HEAD(&rbd_dev->running_list); 5354 init_completion(&rbd_dev->acquire_wait); 5355 init_completion(&rbd_dev->releasing_wait); 5356 5357 spin_lock_init(&rbd_dev->object_map_lock); 5358 5359 rbd_dev->dev.bus = &rbd_bus_type; 5360 rbd_dev->dev.type = &rbd_device_type; 5361 rbd_dev->dev.parent = &rbd_root_dev; 5362 device_initialize(&rbd_dev->dev); 5363 5364 rbd_dev->rbd_client = rbdc; 5365 rbd_dev->spec = spec; 5366 5367 return rbd_dev; 5368 } 5369 5370 /* 5371 * Create a mapping rbd_dev. 5372 */ 5373 static struct rbd_device *rbd_dev_create(struct rbd_client *rbdc, 5374 struct rbd_spec *spec, 5375 struct rbd_options *opts) 5376 { 5377 struct rbd_device *rbd_dev; 5378 5379 rbd_dev = __rbd_dev_create(rbdc, spec); 5380 if (!rbd_dev) 5381 return NULL; 5382 5383 rbd_dev->opts = opts; 5384 5385 /* get an id and fill in device name */ 5386 rbd_dev->dev_id = ida_simple_get(&rbd_dev_id_ida, 0, 5387 minor_to_rbd_dev_id(1 << MINORBITS), 5388 GFP_KERNEL); 5389 if (rbd_dev->dev_id < 0) 5390 goto fail_rbd_dev; 5391 5392 sprintf(rbd_dev->name, RBD_DRV_NAME "%d", rbd_dev->dev_id); 5393 rbd_dev->task_wq = alloc_ordered_workqueue("%s-tasks", WQ_MEM_RECLAIM, 5394 rbd_dev->name); 5395 if (!rbd_dev->task_wq) 5396 goto fail_dev_id; 5397 5398 /* we have a ref from do_rbd_add() */ 5399 __module_get(THIS_MODULE); 5400 5401 dout("%s rbd_dev %p dev_id %d\n", __func__, rbd_dev, rbd_dev->dev_id); 5402 return rbd_dev; 5403 5404 fail_dev_id: 5405 ida_simple_remove(&rbd_dev_id_ida, rbd_dev->dev_id); 5406 fail_rbd_dev: 5407 rbd_dev_free(rbd_dev); 5408 return NULL; 5409 } 5410 5411 static void rbd_dev_destroy(struct rbd_device *rbd_dev) 5412 { 5413 if (rbd_dev) 5414 put_device(&rbd_dev->dev); 5415 } 5416 5417 /* 5418 * Get the size and object order for an image snapshot, or if 5419 * snap_id is CEPH_NOSNAP, gets this information for the base 5420 * image. 5421 */ 5422 static int _rbd_dev_v2_snap_size(struct rbd_device *rbd_dev, u64 snap_id, 5423 u8 *order, u64 *snap_size) 5424 { 5425 __le64 snapid = cpu_to_le64(snap_id); 5426 int ret; 5427 struct { 5428 u8 order; 5429 __le64 size; 5430 } __attribute__ ((packed)) size_buf = { 0 }; 5431 5432 ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid, 5433 &rbd_dev->header_oloc, "get_size", 5434 &snapid, sizeof(snapid), 5435 &size_buf, sizeof(size_buf)); 5436 dout("%s: rbd_obj_method_sync returned %d\n", __func__, ret); 5437 if (ret < 0) 5438 return ret; 5439 if (ret < sizeof (size_buf)) 5440 return -ERANGE; 5441 5442 if (order) { 5443 *order = size_buf.order; 5444 dout(" order %u", (unsigned int)*order); 5445 } 5446 *snap_size = le64_to_cpu(size_buf.size); 5447 5448 dout(" snap_id 0x%016llx snap_size = %llu\n", 5449 (unsigned long long)snap_id, 5450 (unsigned long long)*snap_size); 5451 5452 return 0; 5453 } 5454 5455 static int rbd_dev_v2_image_size(struct rbd_device *rbd_dev) 5456 { 5457 return _rbd_dev_v2_snap_size(rbd_dev, CEPH_NOSNAP, 5458 &rbd_dev->header.obj_order, 5459 &rbd_dev->header.image_size); 5460 } 5461 5462 static int rbd_dev_v2_object_prefix(struct rbd_device *rbd_dev) 5463 { 5464 size_t size; 5465 void *reply_buf; 5466 int ret; 5467 void *p; 5468 5469 /* Response will be an encoded string, which includes a length */ 5470 size = sizeof(__le32) + RBD_OBJ_PREFIX_LEN_MAX; 5471 reply_buf = kzalloc(size, GFP_KERNEL); 5472 if (!reply_buf) 5473 return -ENOMEM; 5474 5475 ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid, 5476 &rbd_dev->header_oloc, "get_object_prefix", 5477 NULL, 0, reply_buf, size); 5478 dout("%s: rbd_obj_method_sync returned %d\n", __func__, ret); 5479 if (ret < 0) 5480 goto out; 5481 5482 p = reply_buf; 5483 rbd_dev->header.object_prefix = ceph_extract_encoded_string(&p, 5484 p + ret, NULL, GFP_NOIO); 5485 ret = 0; 5486 5487 if (IS_ERR(rbd_dev->header.object_prefix)) { 5488 ret = PTR_ERR(rbd_dev->header.object_prefix); 5489 rbd_dev->header.object_prefix = NULL; 5490 } else { 5491 dout(" object_prefix = %s\n", rbd_dev->header.object_prefix); 5492 } 5493 out: 5494 kfree(reply_buf); 5495 5496 return ret; 5497 } 5498 5499 static int _rbd_dev_v2_snap_features(struct rbd_device *rbd_dev, u64 snap_id, 5500 bool read_only, u64 *snap_features) 5501 { 5502 struct { 5503 __le64 snap_id; 5504 u8 read_only; 5505 } features_in; 5506 struct { 5507 __le64 features; 5508 __le64 incompat; 5509 } __attribute__ ((packed)) features_buf = { 0 }; 5510 u64 unsup; 5511 int ret; 5512 5513 features_in.snap_id = cpu_to_le64(snap_id); 5514 features_in.read_only = read_only; 5515 5516 ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid, 5517 &rbd_dev->header_oloc, "get_features", 5518 &features_in, sizeof(features_in), 5519 &features_buf, sizeof(features_buf)); 5520 dout("%s: rbd_obj_method_sync returned %d\n", __func__, ret); 5521 if (ret < 0) 5522 return ret; 5523 if (ret < sizeof (features_buf)) 5524 return -ERANGE; 5525 5526 unsup = le64_to_cpu(features_buf.incompat) & ~RBD_FEATURES_SUPPORTED; 5527 if (unsup) { 5528 rbd_warn(rbd_dev, "image uses unsupported features: 0x%llx", 5529 unsup); 5530 return -ENXIO; 5531 } 5532 5533 *snap_features = le64_to_cpu(features_buf.features); 5534 5535 dout(" snap_id 0x%016llx features = 0x%016llx incompat = 0x%016llx\n", 5536 (unsigned long long)snap_id, 5537 (unsigned long long)*snap_features, 5538 (unsigned long long)le64_to_cpu(features_buf.incompat)); 5539 5540 return 0; 5541 } 5542 5543 static int rbd_dev_v2_features(struct rbd_device *rbd_dev) 5544 { 5545 return _rbd_dev_v2_snap_features(rbd_dev, CEPH_NOSNAP, 5546 rbd_is_ro(rbd_dev), 5547 &rbd_dev->header.features); 5548 } 5549 5550 /* 5551 * These are generic image flags, but since they are used only for 5552 * object map, store them in rbd_dev->object_map_flags. 5553 * 5554 * For the same reason, this function is called only on object map 5555 * (re)load and not on header refresh. 5556 */ 5557 static int rbd_dev_v2_get_flags(struct rbd_device *rbd_dev) 5558 { 5559 __le64 snapid = cpu_to_le64(rbd_dev->spec->snap_id); 5560 __le64 flags; 5561 int ret; 5562 5563 ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid, 5564 &rbd_dev->header_oloc, "get_flags", 5565 &snapid, sizeof(snapid), 5566 &flags, sizeof(flags)); 5567 if (ret < 0) 5568 return ret; 5569 if (ret < sizeof(flags)) 5570 return -EBADMSG; 5571 5572 rbd_dev->object_map_flags = le64_to_cpu(flags); 5573 return 0; 5574 } 5575 5576 struct parent_image_info { 5577 u64 pool_id; 5578 const char *pool_ns; 5579 const char *image_id; 5580 u64 snap_id; 5581 5582 bool has_overlap; 5583 u64 overlap; 5584 }; 5585 5586 /* 5587 * The caller is responsible for @pii. 5588 */ 5589 static int decode_parent_image_spec(void **p, void *end, 5590 struct parent_image_info *pii) 5591 { 5592 u8 struct_v; 5593 u32 struct_len; 5594 int ret; 5595 5596 ret = ceph_start_decoding(p, end, 1, "ParentImageSpec", 5597 &struct_v, &struct_len); 5598 if (ret) 5599 return ret; 5600 5601 ceph_decode_64_safe(p, end, pii->pool_id, e_inval); 5602 pii->pool_ns = ceph_extract_encoded_string(p, end, NULL, GFP_KERNEL); 5603 if (IS_ERR(pii->pool_ns)) { 5604 ret = PTR_ERR(pii->pool_ns); 5605 pii->pool_ns = NULL; 5606 return ret; 5607 } 5608 pii->image_id = ceph_extract_encoded_string(p, end, NULL, GFP_KERNEL); 5609 if (IS_ERR(pii->image_id)) { 5610 ret = PTR_ERR(pii->image_id); 5611 pii->image_id = NULL; 5612 return ret; 5613 } 5614 ceph_decode_64_safe(p, end, pii->snap_id, e_inval); 5615 return 0; 5616 5617 e_inval: 5618 return -EINVAL; 5619 } 5620 5621 static int __get_parent_info(struct rbd_device *rbd_dev, 5622 struct page *req_page, 5623 struct page *reply_page, 5624 struct parent_image_info *pii) 5625 { 5626 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 5627 size_t reply_len = PAGE_SIZE; 5628 void *p, *end; 5629 int ret; 5630 5631 ret = ceph_osdc_call(osdc, &rbd_dev->header_oid, &rbd_dev->header_oloc, 5632 "rbd", "parent_get", CEPH_OSD_FLAG_READ, 5633 req_page, sizeof(u64), &reply_page, &reply_len); 5634 if (ret) 5635 return ret == -EOPNOTSUPP ? 1 : ret; 5636 5637 p = page_address(reply_page); 5638 end = p + reply_len; 5639 ret = decode_parent_image_spec(&p, end, pii); 5640 if (ret) 5641 return ret; 5642 5643 ret = ceph_osdc_call(osdc, &rbd_dev->header_oid, &rbd_dev->header_oloc, 5644 "rbd", "parent_overlap_get", CEPH_OSD_FLAG_READ, 5645 req_page, sizeof(u64), &reply_page, &reply_len); 5646 if (ret) 5647 return ret; 5648 5649 p = page_address(reply_page); 5650 end = p + reply_len; 5651 ceph_decode_8_safe(&p, end, pii->has_overlap, e_inval); 5652 if (pii->has_overlap) 5653 ceph_decode_64_safe(&p, end, pii->overlap, e_inval); 5654 5655 return 0; 5656 5657 e_inval: 5658 return -EINVAL; 5659 } 5660 5661 /* 5662 * The caller is responsible for @pii. 5663 */ 5664 static int __get_parent_info_legacy(struct rbd_device *rbd_dev, 5665 struct page *req_page, 5666 struct page *reply_page, 5667 struct parent_image_info *pii) 5668 { 5669 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 5670 size_t reply_len = PAGE_SIZE; 5671 void *p, *end; 5672 int ret; 5673 5674 ret = ceph_osdc_call(osdc, &rbd_dev->header_oid, &rbd_dev->header_oloc, 5675 "rbd", "get_parent", CEPH_OSD_FLAG_READ, 5676 req_page, sizeof(u64), &reply_page, &reply_len); 5677 if (ret) 5678 return ret; 5679 5680 p = page_address(reply_page); 5681 end = p + reply_len; 5682 ceph_decode_64_safe(&p, end, pii->pool_id, e_inval); 5683 pii->image_id = ceph_extract_encoded_string(&p, end, NULL, GFP_KERNEL); 5684 if (IS_ERR(pii->image_id)) { 5685 ret = PTR_ERR(pii->image_id); 5686 pii->image_id = NULL; 5687 return ret; 5688 } 5689 ceph_decode_64_safe(&p, end, pii->snap_id, e_inval); 5690 pii->has_overlap = true; 5691 ceph_decode_64_safe(&p, end, pii->overlap, e_inval); 5692 5693 return 0; 5694 5695 e_inval: 5696 return -EINVAL; 5697 } 5698 5699 static int get_parent_info(struct rbd_device *rbd_dev, 5700 struct parent_image_info *pii) 5701 { 5702 struct page *req_page, *reply_page; 5703 void *p; 5704 int ret; 5705 5706 req_page = alloc_page(GFP_KERNEL); 5707 if (!req_page) 5708 return -ENOMEM; 5709 5710 reply_page = alloc_page(GFP_KERNEL); 5711 if (!reply_page) { 5712 __free_page(req_page); 5713 return -ENOMEM; 5714 } 5715 5716 p = page_address(req_page); 5717 ceph_encode_64(&p, rbd_dev->spec->snap_id); 5718 ret = __get_parent_info(rbd_dev, req_page, reply_page, pii); 5719 if (ret > 0) 5720 ret = __get_parent_info_legacy(rbd_dev, req_page, reply_page, 5721 pii); 5722 5723 __free_page(req_page); 5724 __free_page(reply_page); 5725 return ret; 5726 } 5727 5728 static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev) 5729 { 5730 struct rbd_spec *parent_spec; 5731 struct parent_image_info pii = { 0 }; 5732 int ret; 5733 5734 parent_spec = rbd_spec_alloc(); 5735 if (!parent_spec) 5736 return -ENOMEM; 5737 5738 ret = get_parent_info(rbd_dev, &pii); 5739 if (ret) 5740 goto out_err; 5741 5742 dout("%s pool_id %llu pool_ns %s image_id %s snap_id %llu has_overlap %d overlap %llu\n", 5743 __func__, pii.pool_id, pii.pool_ns, pii.image_id, pii.snap_id, 5744 pii.has_overlap, pii.overlap); 5745 5746 if (pii.pool_id == CEPH_NOPOOL || !pii.has_overlap) { 5747 /* 5748 * Either the parent never existed, or we have 5749 * record of it but the image got flattened so it no 5750 * longer has a parent. When the parent of a 5751 * layered image disappears we immediately set the 5752 * overlap to 0. The effect of this is that all new 5753 * requests will be treated as if the image had no 5754 * parent. 5755 * 5756 * If !pii.has_overlap, the parent image spec is not 5757 * applicable. It's there to avoid duplication in each 5758 * snapshot record. 5759 */ 5760 if (rbd_dev->parent_overlap) { 5761 rbd_dev->parent_overlap = 0; 5762 rbd_dev_parent_put(rbd_dev); 5763 pr_info("%s: clone image has been flattened\n", 5764 rbd_dev->disk->disk_name); 5765 } 5766 5767 goto out; /* No parent? No problem. */ 5768 } 5769 5770 /* The ceph file layout needs to fit pool id in 32 bits */ 5771 5772 ret = -EIO; 5773 if (pii.pool_id > (u64)U32_MAX) { 5774 rbd_warn(NULL, "parent pool id too large (%llu > %u)", 5775 (unsigned long long)pii.pool_id, U32_MAX); 5776 goto out_err; 5777 } 5778 5779 /* 5780 * The parent won't change (except when the clone is 5781 * flattened, already handled that). So we only need to 5782 * record the parent spec we have not already done so. 5783 */ 5784 if (!rbd_dev->parent_spec) { 5785 parent_spec->pool_id = pii.pool_id; 5786 if (pii.pool_ns && *pii.pool_ns) { 5787 parent_spec->pool_ns = pii.pool_ns; 5788 pii.pool_ns = NULL; 5789 } 5790 parent_spec->image_id = pii.image_id; 5791 pii.image_id = NULL; 5792 parent_spec->snap_id = pii.snap_id; 5793 5794 rbd_dev->parent_spec = parent_spec; 5795 parent_spec = NULL; /* rbd_dev now owns this */ 5796 } 5797 5798 /* 5799 * We always update the parent overlap. If it's zero we issue 5800 * a warning, as we will proceed as if there was no parent. 5801 */ 5802 if (!pii.overlap) { 5803 if (parent_spec) { 5804 /* refresh, careful to warn just once */ 5805 if (rbd_dev->parent_overlap) 5806 rbd_warn(rbd_dev, 5807 "clone now standalone (overlap became 0)"); 5808 } else { 5809 /* initial probe */ 5810 rbd_warn(rbd_dev, "clone is standalone (overlap 0)"); 5811 } 5812 } 5813 rbd_dev->parent_overlap = pii.overlap; 5814 5815 out: 5816 ret = 0; 5817 out_err: 5818 kfree(pii.pool_ns); 5819 kfree(pii.image_id); 5820 rbd_spec_put(parent_spec); 5821 return ret; 5822 } 5823 5824 static int rbd_dev_v2_striping_info(struct rbd_device *rbd_dev) 5825 { 5826 struct { 5827 __le64 stripe_unit; 5828 __le64 stripe_count; 5829 } __attribute__ ((packed)) striping_info_buf = { 0 }; 5830 size_t size = sizeof (striping_info_buf); 5831 void *p; 5832 int ret; 5833 5834 ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid, 5835 &rbd_dev->header_oloc, "get_stripe_unit_count", 5836 NULL, 0, &striping_info_buf, size); 5837 dout("%s: rbd_obj_method_sync returned %d\n", __func__, ret); 5838 if (ret < 0) 5839 return ret; 5840 if (ret < size) 5841 return -ERANGE; 5842 5843 p = &striping_info_buf; 5844 rbd_dev->header.stripe_unit = ceph_decode_64(&p); 5845 rbd_dev->header.stripe_count = ceph_decode_64(&p); 5846 return 0; 5847 } 5848 5849 static int rbd_dev_v2_data_pool(struct rbd_device *rbd_dev) 5850 { 5851 __le64 data_pool_id; 5852 int ret; 5853 5854 ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid, 5855 &rbd_dev->header_oloc, "get_data_pool", 5856 NULL, 0, &data_pool_id, sizeof(data_pool_id)); 5857 if (ret < 0) 5858 return ret; 5859 if (ret < sizeof(data_pool_id)) 5860 return -EBADMSG; 5861 5862 rbd_dev->header.data_pool_id = le64_to_cpu(data_pool_id); 5863 WARN_ON(rbd_dev->header.data_pool_id == CEPH_NOPOOL); 5864 return 0; 5865 } 5866 5867 static char *rbd_dev_image_name(struct rbd_device *rbd_dev) 5868 { 5869 CEPH_DEFINE_OID_ONSTACK(oid); 5870 size_t image_id_size; 5871 char *image_id; 5872 void *p; 5873 void *end; 5874 size_t size; 5875 void *reply_buf = NULL; 5876 size_t len = 0; 5877 char *image_name = NULL; 5878 int ret; 5879 5880 rbd_assert(!rbd_dev->spec->image_name); 5881 5882 len = strlen(rbd_dev->spec->image_id); 5883 image_id_size = sizeof (__le32) + len; 5884 image_id = kmalloc(image_id_size, GFP_KERNEL); 5885 if (!image_id) 5886 return NULL; 5887 5888 p = image_id; 5889 end = image_id + image_id_size; 5890 ceph_encode_string(&p, end, rbd_dev->spec->image_id, (u32)len); 5891 5892 size = sizeof (__le32) + RBD_IMAGE_NAME_LEN_MAX; 5893 reply_buf = kmalloc(size, GFP_KERNEL); 5894 if (!reply_buf) 5895 goto out; 5896 5897 ceph_oid_printf(&oid, "%s", RBD_DIRECTORY); 5898 ret = rbd_obj_method_sync(rbd_dev, &oid, &rbd_dev->header_oloc, 5899 "dir_get_name", image_id, image_id_size, 5900 reply_buf, size); 5901 if (ret < 0) 5902 goto out; 5903 p = reply_buf; 5904 end = reply_buf + ret; 5905 5906 image_name = ceph_extract_encoded_string(&p, end, &len, GFP_KERNEL); 5907 if (IS_ERR(image_name)) 5908 image_name = NULL; 5909 else 5910 dout("%s: name is %s len is %zd\n", __func__, image_name, len); 5911 out: 5912 kfree(reply_buf); 5913 kfree(image_id); 5914 5915 return image_name; 5916 } 5917 5918 static u64 rbd_v1_snap_id_by_name(struct rbd_device *rbd_dev, const char *name) 5919 { 5920 struct ceph_snap_context *snapc = rbd_dev->header.snapc; 5921 const char *snap_name; 5922 u32 which = 0; 5923 5924 /* Skip over names until we find the one we are looking for */ 5925 5926 snap_name = rbd_dev->header.snap_names; 5927 while (which < snapc->num_snaps) { 5928 if (!strcmp(name, snap_name)) 5929 return snapc->snaps[which]; 5930 snap_name += strlen(snap_name) + 1; 5931 which++; 5932 } 5933 return CEPH_NOSNAP; 5934 } 5935 5936 static u64 rbd_v2_snap_id_by_name(struct rbd_device *rbd_dev, const char *name) 5937 { 5938 struct ceph_snap_context *snapc = rbd_dev->header.snapc; 5939 u32 which; 5940 bool found = false; 5941 u64 snap_id; 5942 5943 for (which = 0; !found && which < snapc->num_snaps; which++) { 5944 const char *snap_name; 5945 5946 snap_id = snapc->snaps[which]; 5947 snap_name = rbd_dev_v2_snap_name(rbd_dev, snap_id); 5948 if (IS_ERR(snap_name)) { 5949 /* ignore no-longer existing snapshots */ 5950 if (PTR_ERR(snap_name) == -ENOENT) 5951 continue; 5952 else 5953 break; 5954 } 5955 found = !strcmp(name, snap_name); 5956 kfree(snap_name); 5957 } 5958 return found ? snap_id : CEPH_NOSNAP; 5959 } 5960 5961 /* 5962 * Assumes name is never RBD_SNAP_HEAD_NAME; returns CEPH_NOSNAP if 5963 * no snapshot by that name is found, or if an error occurs. 5964 */ 5965 static u64 rbd_snap_id_by_name(struct rbd_device *rbd_dev, const char *name) 5966 { 5967 if (rbd_dev->image_format == 1) 5968 return rbd_v1_snap_id_by_name(rbd_dev, name); 5969 5970 return rbd_v2_snap_id_by_name(rbd_dev, name); 5971 } 5972 5973 /* 5974 * An image being mapped will have everything but the snap id. 5975 */ 5976 static int rbd_spec_fill_snap_id(struct rbd_device *rbd_dev) 5977 { 5978 struct rbd_spec *spec = rbd_dev->spec; 5979 5980 rbd_assert(spec->pool_id != CEPH_NOPOOL && spec->pool_name); 5981 rbd_assert(spec->image_id && spec->image_name); 5982 rbd_assert(spec->snap_name); 5983 5984 if (strcmp(spec->snap_name, RBD_SNAP_HEAD_NAME)) { 5985 u64 snap_id; 5986 5987 snap_id = rbd_snap_id_by_name(rbd_dev, spec->snap_name); 5988 if (snap_id == CEPH_NOSNAP) 5989 return -ENOENT; 5990 5991 spec->snap_id = snap_id; 5992 } else { 5993 spec->snap_id = CEPH_NOSNAP; 5994 } 5995 5996 return 0; 5997 } 5998 5999 /* 6000 * A parent image will have all ids but none of the names. 6001 * 6002 * All names in an rbd spec are dynamically allocated. It's OK if we 6003 * can't figure out the name for an image id. 6004 */ 6005 static int rbd_spec_fill_names(struct rbd_device *rbd_dev) 6006 { 6007 struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; 6008 struct rbd_spec *spec = rbd_dev->spec; 6009 const char *pool_name; 6010 const char *image_name; 6011 const char *snap_name; 6012 int ret; 6013 6014 rbd_assert(spec->pool_id != CEPH_NOPOOL); 6015 rbd_assert(spec->image_id); 6016 rbd_assert(spec->snap_id != CEPH_NOSNAP); 6017 6018 /* Get the pool name; we have to make our own copy of this */ 6019 6020 pool_name = ceph_pg_pool_name_by_id(osdc->osdmap, spec->pool_id); 6021 if (!pool_name) { 6022 rbd_warn(rbd_dev, "no pool with id %llu", spec->pool_id); 6023 return -EIO; 6024 } 6025 pool_name = kstrdup(pool_name, GFP_KERNEL); 6026 if (!pool_name) 6027 return -ENOMEM; 6028 6029 /* Fetch the image name; tolerate failure here */ 6030 6031 image_name = rbd_dev_image_name(rbd_dev); 6032 if (!image_name) 6033 rbd_warn(rbd_dev, "unable to get image name"); 6034 6035 /* Fetch the snapshot name */ 6036 6037 snap_name = rbd_snap_name(rbd_dev, spec->snap_id); 6038 if (IS_ERR(snap_name)) { 6039 ret = PTR_ERR(snap_name); 6040 goto out_err; 6041 } 6042 6043 spec->pool_name = pool_name; 6044 spec->image_name = image_name; 6045 spec->snap_name = snap_name; 6046 6047 return 0; 6048 6049 out_err: 6050 kfree(image_name); 6051 kfree(pool_name); 6052 return ret; 6053 } 6054 6055 static int rbd_dev_v2_snap_context(struct rbd_device *rbd_dev) 6056 { 6057 size_t size; 6058 int ret; 6059 void *reply_buf; 6060 void *p; 6061 void *end; 6062 u64 seq; 6063 u32 snap_count; 6064 struct ceph_snap_context *snapc; 6065 u32 i; 6066 6067 /* 6068 * We'll need room for the seq value (maximum snapshot id), 6069 * snapshot count, and array of that many snapshot ids. 6070 * For now we have a fixed upper limit on the number we're 6071 * prepared to receive. 6072 */ 6073 size = sizeof (__le64) + sizeof (__le32) + 6074 RBD_MAX_SNAP_COUNT * sizeof (__le64); 6075 reply_buf = kzalloc(size, GFP_KERNEL); 6076 if (!reply_buf) 6077 return -ENOMEM; 6078 6079 ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid, 6080 &rbd_dev->header_oloc, "get_snapcontext", 6081 NULL, 0, reply_buf, size); 6082 dout("%s: rbd_obj_method_sync returned %d\n", __func__, ret); 6083 if (ret < 0) 6084 goto out; 6085 6086 p = reply_buf; 6087 end = reply_buf + ret; 6088 ret = -ERANGE; 6089 ceph_decode_64_safe(&p, end, seq, out); 6090 ceph_decode_32_safe(&p, end, snap_count, out); 6091 6092 /* 6093 * Make sure the reported number of snapshot ids wouldn't go 6094 * beyond the end of our buffer. But before checking that, 6095 * make sure the computed size of the snapshot context we 6096 * allocate is representable in a size_t. 6097 */ 6098 if (snap_count > (SIZE_MAX - sizeof (struct ceph_snap_context)) 6099 / sizeof (u64)) { 6100 ret = -EINVAL; 6101 goto out; 6102 } 6103 if (!ceph_has_room(&p, end, snap_count * sizeof (__le64))) 6104 goto out; 6105 ret = 0; 6106 6107 snapc = ceph_create_snap_context(snap_count, GFP_KERNEL); 6108 if (!snapc) { 6109 ret = -ENOMEM; 6110 goto out; 6111 } 6112 snapc->seq = seq; 6113 for (i = 0; i < snap_count; i++) 6114 snapc->snaps[i] = ceph_decode_64(&p); 6115 6116 ceph_put_snap_context(rbd_dev->header.snapc); 6117 rbd_dev->header.snapc = snapc; 6118 6119 dout(" snap context seq = %llu, snap_count = %u\n", 6120 (unsigned long long)seq, (unsigned int)snap_count); 6121 out: 6122 kfree(reply_buf); 6123 6124 return ret; 6125 } 6126 6127 static const char *rbd_dev_v2_snap_name(struct rbd_device *rbd_dev, 6128 u64 snap_id) 6129 { 6130 size_t size; 6131 void *reply_buf; 6132 __le64 snapid; 6133 int ret; 6134 void *p; 6135 void *end; 6136 char *snap_name; 6137 6138 size = sizeof (__le32) + RBD_MAX_SNAP_NAME_LEN; 6139 reply_buf = kmalloc(size, GFP_KERNEL); 6140 if (!reply_buf) 6141 return ERR_PTR(-ENOMEM); 6142 6143 snapid = cpu_to_le64(snap_id); 6144 ret = rbd_obj_method_sync(rbd_dev, &rbd_dev->header_oid, 6145 &rbd_dev->header_oloc, "get_snapshot_name", 6146 &snapid, sizeof(snapid), reply_buf, size); 6147 dout("%s: rbd_obj_method_sync returned %d\n", __func__, ret); 6148 if (ret < 0) { 6149 snap_name = ERR_PTR(ret); 6150 goto out; 6151 } 6152 6153 p = reply_buf; 6154 end = reply_buf + ret; 6155 snap_name = ceph_extract_encoded_string(&p, end, NULL, GFP_KERNEL); 6156 if (IS_ERR(snap_name)) 6157 goto out; 6158 6159 dout(" snap_id 0x%016llx snap_name = %s\n", 6160 (unsigned long long)snap_id, snap_name); 6161 out: 6162 kfree(reply_buf); 6163 6164 return snap_name; 6165 } 6166 6167 static int rbd_dev_v2_header_info(struct rbd_device *rbd_dev) 6168 { 6169 bool first_time = rbd_dev->header.object_prefix == NULL; 6170 int ret; 6171 6172 ret = rbd_dev_v2_image_size(rbd_dev); 6173 if (ret) 6174 return ret; 6175 6176 if (first_time) { 6177 ret = rbd_dev_v2_header_onetime(rbd_dev); 6178 if (ret) 6179 return ret; 6180 } 6181 6182 ret = rbd_dev_v2_snap_context(rbd_dev); 6183 if (ret && first_time) { 6184 kfree(rbd_dev->header.object_prefix); 6185 rbd_dev->header.object_prefix = NULL; 6186 } 6187 6188 return ret; 6189 } 6190 6191 static int rbd_dev_header_info(struct rbd_device *rbd_dev) 6192 { 6193 rbd_assert(rbd_image_format_valid(rbd_dev->image_format)); 6194 6195 if (rbd_dev->image_format == 1) 6196 return rbd_dev_v1_header_info(rbd_dev); 6197 6198 return rbd_dev_v2_header_info(rbd_dev); 6199 } 6200 6201 /* 6202 * Skips over white space at *buf, and updates *buf to point to the 6203 * first found non-space character (if any). Returns the length of 6204 * the token (string of non-white space characters) found. Note 6205 * that *buf must be terminated with '\0'. 6206 */ 6207 static inline size_t next_token(const char **buf) 6208 { 6209 /* 6210 * These are the characters that produce nonzero for 6211 * isspace() in the "C" and "POSIX" locales. 6212 */ 6213 const char *spaces = " \f\n\r\t\v"; 6214 6215 *buf += strspn(*buf, spaces); /* Find start of token */ 6216 6217 return strcspn(*buf, spaces); /* Return token length */ 6218 } 6219 6220 /* 6221 * Finds the next token in *buf, dynamically allocates a buffer big 6222 * enough to hold a copy of it, and copies the token into the new 6223 * buffer. The copy is guaranteed to be terminated with '\0'. Note 6224 * that a duplicate buffer is created even for a zero-length token. 6225 * 6226 * Returns a pointer to the newly-allocated duplicate, or a null 6227 * pointer if memory for the duplicate was not available. If 6228 * the lenp argument is a non-null pointer, the length of the token 6229 * (not including the '\0') is returned in *lenp. 6230 * 6231 * If successful, the *buf pointer will be updated to point beyond 6232 * the end of the found token. 6233 * 6234 * Note: uses GFP_KERNEL for allocation. 6235 */ 6236 static inline char *dup_token(const char **buf, size_t *lenp) 6237 { 6238 char *dup; 6239 size_t len; 6240 6241 len = next_token(buf); 6242 dup = kmemdup(*buf, len + 1, GFP_KERNEL); 6243 if (!dup) 6244 return NULL; 6245 *(dup + len) = '\0'; 6246 *buf += len; 6247 6248 if (lenp) 6249 *lenp = len; 6250 6251 return dup; 6252 } 6253 6254 static int rbd_parse_param(struct fs_parameter *param, 6255 struct rbd_parse_opts_ctx *pctx) 6256 { 6257 struct rbd_options *opt = pctx->opts; 6258 struct fs_parse_result result; 6259 struct p_log log = {.prefix = "rbd"}; 6260 int token, ret; 6261 6262 ret = ceph_parse_param(param, pctx->copts, NULL); 6263 if (ret != -ENOPARAM) 6264 return ret; 6265 6266 token = __fs_parse(&log, rbd_parameters, param, &result); 6267 dout("%s fs_parse '%s' token %d\n", __func__, param->key, token); 6268 if (token < 0) { 6269 if (token == -ENOPARAM) 6270 return inval_plog(&log, "Unknown parameter '%s'", 6271 param->key); 6272 return token; 6273 } 6274 6275 switch (token) { 6276 case Opt_queue_depth: 6277 if (result.uint_32 < 1) 6278 goto out_of_range; 6279 opt->queue_depth = result.uint_32; 6280 break; 6281 case Opt_alloc_size: 6282 if (result.uint_32 < SECTOR_SIZE) 6283 goto out_of_range; 6284 if (!is_power_of_2(result.uint_32)) 6285 return inval_plog(&log, "alloc_size must be a power of 2"); 6286 opt->alloc_size = result.uint_32; 6287 break; 6288 case Opt_lock_timeout: 6289 /* 0 is "wait forever" (i.e. infinite timeout) */ 6290 if (result.uint_32 > INT_MAX / 1000) 6291 goto out_of_range; 6292 opt->lock_timeout = msecs_to_jiffies(result.uint_32 * 1000); 6293 break; 6294 case Opt_pool_ns: 6295 kfree(pctx->spec->pool_ns); 6296 pctx->spec->pool_ns = param->string; 6297 param->string = NULL; 6298 break; 6299 case Opt_compression_hint: 6300 switch (result.uint_32) { 6301 case Opt_compression_hint_none: 6302 opt->alloc_hint_flags &= 6303 ~(CEPH_OSD_ALLOC_HINT_FLAG_COMPRESSIBLE | 6304 CEPH_OSD_ALLOC_HINT_FLAG_INCOMPRESSIBLE); 6305 break; 6306 case Opt_compression_hint_compressible: 6307 opt->alloc_hint_flags |= 6308 CEPH_OSD_ALLOC_HINT_FLAG_COMPRESSIBLE; 6309 opt->alloc_hint_flags &= 6310 ~CEPH_OSD_ALLOC_HINT_FLAG_INCOMPRESSIBLE; 6311 break; 6312 case Opt_compression_hint_incompressible: 6313 opt->alloc_hint_flags |= 6314 CEPH_OSD_ALLOC_HINT_FLAG_INCOMPRESSIBLE; 6315 opt->alloc_hint_flags &= 6316 ~CEPH_OSD_ALLOC_HINT_FLAG_COMPRESSIBLE; 6317 break; 6318 default: 6319 BUG(); 6320 } 6321 break; 6322 case Opt_read_only: 6323 opt->read_only = true; 6324 break; 6325 case Opt_read_write: 6326 opt->read_only = false; 6327 break; 6328 case Opt_lock_on_read: 6329 opt->lock_on_read = true; 6330 break; 6331 case Opt_exclusive: 6332 opt->exclusive = true; 6333 break; 6334 case Opt_notrim: 6335 opt->trim = false; 6336 break; 6337 default: 6338 BUG(); 6339 } 6340 6341 return 0; 6342 6343 out_of_range: 6344 return inval_plog(&log, "%s out of range", param->key); 6345 } 6346 6347 /* 6348 * This duplicates most of generic_parse_monolithic(), untying it from 6349 * fs_context and skipping standard superblock and security options. 6350 */ 6351 static int rbd_parse_options(char *options, struct rbd_parse_opts_ctx *pctx) 6352 { 6353 char *key; 6354 int ret = 0; 6355 6356 dout("%s '%s'\n", __func__, options); 6357 while ((key = strsep(&options, ",")) != NULL) { 6358 if (*key) { 6359 struct fs_parameter param = { 6360 .key = key, 6361 .type = fs_value_is_flag, 6362 }; 6363 char *value = strchr(key, '='); 6364 size_t v_len = 0; 6365 6366 if (value) { 6367 if (value == key) 6368 continue; 6369 *value++ = 0; 6370 v_len = strlen(value); 6371 param.string = kmemdup_nul(value, v_len, 6372 GFP_KERNEL); 6373 if (!param.string) 6374 return -ENOMEM; 6375 param.type = fs_value_is_string; 6376 } 6377 param.size = v_len; 6378 6379 ret = rbd_parse_param(¶m, pctx); 6380 kfree(param.string); 6381 if (ret) 6382 break; 6383 } 6384 } 6385 6386 return ret; 6387 } 6388 6389 /* 6390 * Parse the options provided for an "rbd add" (i.e., rbd image 6391 * mapping) request. These arrive via a write to /sys/bus/rbd/add, 6392 * and the data written is passed here via a NUL-terminated buffer. 6393 * Returns 0 if successful or an error code otherwise. 6394 * 6395 * The information extracted from these options is recorded in 6396 * the other parameters which return dynamically-allocated 6397 * structures: 6398 * ceph_opts 6399 * The address of a pointer that will refer to a ceph options 6400 * structure. Caller must release the returned pointer using 6401 * ceph_destroy_options() when it is no longer needed. 6402 * rbd_opts 6403 * Address of an rbd options pointer. Fully initialized by 6404 * this function; caller must release with kfree(). 6405 * spec 6406 * Address of an rbd image specification pointer. Fully 6407 * initialized by this function based on parsed options. 6408 * Caller must release with rbd_spec_put(). 6409 * 6410 * The options passed take this form: 6411 * <mon_addrs> <options> <pool_name> <image_name> [<snap_id>] 6412 * where: 6413 * <mon_addrs> 6414 * A comma-separated list of one or more monitor addresses. 6415 * A monitor address is an ip address, optionally followed 6416 * by a port number (separated by a colon). 6417 * I.e.: ip1[:port1][,ip2[:port2]...] 6418 * <options> 6419 * A comma-separated list of ceph and/or rbd options. 6420 * <pool_name> 6421 * The name of the rados pool containing the rbd image. 6422 * <image_name> 6423 * The name of the image in that pool to map. 6424 * <snap_id> 6425 * An optional snapshot id. If provided, the mapping will 6426 * present data from the image at the time that snapshot was 6427 * created. The image head is used if no snapshot id is 6428 * provided. Snapshot mappings are always read-only. 6429 */ 6430 static int rbd_add_parse_args(const char *buf, 6431 struct ceph_options **ceph_opts, 6432 struct rbd_options **opts, 6433 struct rbd_spec **rbd_spec) 6434 { 6435 size_t len; 6436 char *options; 6437 const char *mon_addrs; 6438 char *snap_name; 6439 size_t mon_addrs_size; 6440 struct rbd_parse_opts_ctx pctx = { 0 }; 6441 int ret; 6442 6443 /* The first four tokens are required */ 6444 6445 len = next_token(&buf); 6446 if (!len) { 6447 rbd_warn(NULL, "no monitor address(es) provided"); 6448 return -EINVAL; 6449 } 6450 mon_addrs = buf; 6451 mon_addrs_size = len; 6452 buf += len; 6453 6454 ret = -EINVAL; 6455 options = dup_token(&buf, NULL); 6456 if (!options) 6457 return -ENOMEM; 6458 if (!*options) { 6459 rbd_warn(NULL, "no options provided"); 6460 goto out_err; 6461 } 6462 6463 pctx.spec = rbd_spec_alloc(); 6464 if (!pctx.spec) 6465 goto out_mem; 6466 6467 pctx.spec->pool_name = dup_token(&buf, NULL); 6468 if (!pctx.spec->pool_name) 6469 goto out_mem; 6470 if (!*pctx.spec->pool_name) { 6471 rbd_warn(NULL, "no pool name provided"); 6472 goto out_err; 6473 } 6474 6475 pctx.spec->image_name = dup_token(&buf, NULL); 6476 if (!pctx.spec->image_name) 6477 goto out_mem; 6478 if (!*pctx.spec->image_name) { 6479 rbd_warn(NULL, "no image name provided"); 6480 goto out_err; 6481 } 6482 6483 /* 6484 * Snapshot name is optional; default is to use "-" 6485 * (indicating the head/no snapshot). 6486 */ 6487 len = next_token(&buf); 6488 if (!len) { 6489 buf = RBD_SNAP_HEAD_NAME; /* No snapshot supplied */ 6490 len = sizeof (RBD_SNAP_HEAD_NAME) - 1; 6491 } else if (len > RBD_MAX_SNAP_NAME_LEN) { 6492 ret = -ENAMETOOLONG; 6493 goto out_err; 6494 } 6495 snap_name = kmemdup(buf, len + 1, GFP_KERNEL); 6496 if (!snap_name) 6497 goto out_mem; 6498 *(snap_name + len) = '\0'; 6499 pctx.spec->snap_name = snap_name; 6500 6501 pctx.copts = ceph_alloc_options(); 6502 if (!pctx.copts) 6503 goto out_mem; 6504 6505 /* Initialize all rbd options to the defaults */ 6506 6507 pctx.opts = kzalloc(sizeof(*pctx.opts), GFP_KERNEL); 6508 if (!pctx.opts) 6509 goto out_mem; 6510 6511 pctx.opts->read_only = RBD_READ_ONLY_DEFAULT; 6512 pctx.opts->queue_depth = RBD_QUEUE_DEPTH_DEFAULT; 6513 pctx.opts->alloc_size = RBD_ALLOC_SIZE_DEFAULT; 6514 pctx.opts->lock_timeout = RBD_LOCK_TIMEOUT_DEFAULT; 6515 pctx.opts->lock_on_read = RBD_LOCK_ON_READ_DEFAULT; 6516 pctx.opts->exclusive = RBD_EXCLUSIVE_DEFAULT; 6517 pctx.opts->trim = RBD_TRIM_DEFAULT; 6518 6519 ret = ceph_parse_mon_ips(mon_addrs, mon_addrs_size, pctx.copts, NULL); 6520 if (ret) 6521 goto out_err; 6522 6523 ret = rbd_parse_options(options, &pctx); 6524 if (ret) 6525 goto out_err; 6526 6527 *ceph_opts = pctx.copts; 6528 *opts = pctx.opts; 6529 *rbd_spec = pctx.spec; 6530 kfree(options); 6531 return 0; 6532 6533 out_mem: 6534 ret = -ENOMEM; 6535 out_err: 6536 kfree(pctx.opts); 6537 ceph_destroy_options(pctx.copts); 6538 rbd_spec_put(pctx.spec); 6539 kfree(options); 6540 return ret; 6541 } 6542 6543 static void rbd_dev_image_unlock(struct rbd_device *rbd_dev) 6544 { 6545 down_write(&rbd_dev->lock_rwsem); 6546 if (__rbd_is_lock_owner(rbd_dev)) 6547 __rbd_release_lock(rbd_dev); 6548 up_write(&rbd_dev->lock_rwsem); 6549 } 6550 6551 /* 6552 * If the wait is interrupted, an error is returned even if the lock 6553 * was successfully acquired. rbd_dev_image_unlock() will release it 6554 * if needed. 6555 */ 6556 static int rbd_add_acquire_lock(struct rbd_device *rbd_dev) 6557 { 6558 long ret; 6559 6560 if (!(rbd_dev->header.features & RBD_FEATURE_EXCLUSIVE_LOCK)) { 6561 if (!rbd_dev->opts->exclusive && !rbd_dev->opts->lock_on_read) 6562 return 0; 6563 6564 rbd_warn(rbd_dev, "exclusive-lock feature is not enabled"); 6565 return -EINVAL; 6566 } 6567 6568 if (rbd_is_ro(rbd_dev)) 6569 return 0; 6570 6571 rbd_assert(!rbd_is_lock_owner(rbd_dev)); 6572 queue_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork, 0); 6573 ret = wait_for_completion_killable_timeout(&rbd_dev->acquire_wait, 6574 ceph_timeout_jiffies(rbd_dev->opts->lock_timeout)); 6575 if (ret > 0) { 6576 ret = rbd_dev->acquire_err; 6577 } else { 6578 cancel_delayed_work_sync(&rbd_dev->lock_dwork); 6579 if (!ret) 6580 ret = -ETIMEDOUT; 6581 } 6582 6583 if (ret) { 6584 rbd_warn(rbd_dev, "failed to acquire exclusive lock: %ld", ret); 6585 return ret; 6586 } 6587 6588 /* 6589 * The lock may have been released by now, unless automatic lock 6590 * transitions are disabled. 6591 */ 6592 rbd_assert(!rbd_dev->opts->exclusive || rbd_is_lock_owner(rbd_dev)); 6593 return 0; 6594 } 6595 6596 /* 6597 * An rbd format 2 image has a unique identifier, distinct from the 6598 * name given to it by the user. Internally, that identifier is 6599 * what's used to specify the names of objects related to the image. 6600 * 6601 * A special "rbd id" object is used to map an rbd image name to its 6602 * id. If that object doesn't exist, then there is no v2 rbd image 6603 * with the supplied name. 6604 * 6605 * This function will record the given rbd_dev's image_id field if 6606 * it can be determined, and in that case will return 0. If any 6607 * errors occur a negative errno will be returned and the rbd_dev's 6608 * image_id field will be unchanged (and should be NULL). 6609 */ 6610 static int rbd_dev_image_id(struct rbd_device *rbd_dev) 6611 { 6612 int ret; 6613 size_t size; 6614 CEPH_DEFINE_OID_ONSTACK(oid); 6615 void *response; 6616 char *image_id; 6617 6618 /* 6619 * When probing a parent image, the image id is already 6620 * known (and the image name likely is not). There's no 6621 * need to fetch the image id again in this case. We 6622 * do still need to set the image format though. 6623 */ 6624 if (rbd_dev->spec->image_id) { 6625 rbd_dev->image_format = *rbd_dev->spec->image_id ? 2 : 1; 6626 6627 return 0; 6628 } 6629 6630 /* 6631 * First, see if the format 2 image id file exists, and if 6632 * so, get the image's persistent id from it. 6633 */ 6634 ret = ceph_oid_aprintf(&oid, GFP_KERNEL, "%s%s", RBD_ID_PREFIX, 6635 rbd_dev->spec->image_name); 6636 if (ret) 6637 return ret; 6638 6639 dout("rbd id object name is %s\n", oid.name); 6640 6641 /* Response will be an encoded string, which includes a length */ 6642 size = sizeof (__le32) + RBD_IMAGE_ID_LEN_MAX; 6643 response = kzalloc(size, GFP_NOIO); 6644 if (!response) { 6645 ret = -ENOMEM; 6646 goto out; 6647 } 6648 6649 /* If it doesn't exist we'll assume it's a format 1 image */ 6650 6651 ret = rbd_obj_method_sync(rbd_dev, &oid, &rbd_dev->header_oloc, 6652 "get_id", NULL, 0, 6653 response, size); 6654 dout("%s: rbd_obj_method_sync returned %d\n", __func__, ret); 6655 if (ret == -ENOENT) { 6656 image_id = kstrdup("", GFP_KERNEL); 6657 ret = image_id ? 0 : -ENOMEM; 6658 if (!ret) 6659 rbd_dev->image_format = 1; 6660 } else if (ret >= 0) { 6661 void *p = response; 6662 6663 image_id = ceph_extract_encoded_string(&p, p + ret, 6664 NULL, GFP_NOIO); 6665 ret = PTR_ERR_OR_ZERO(image_id); 6666 if (!ret) 6667 rbd_dev->image_format = 2; 6668 } 6669 6670 if (!ret) { 6671 rbd_dev->spec->image_id = image_id; 6672 dout("image_id is %s\n", image_id); 6673 } 6674 out: 6675 kfree(response); 6676 ceph_oid_destroy(&oid); 6677 return ret; 6678 } 6679 6680 /* 6681 * Undo whatever state changes are made by v1 or v2 header info 6682 * call. 6683 */ 6684 static void rbd_dev_unprobe(struct rbd_device *rbd_dev) 6685 { 6686 struct rbd_image_header *header; 6687 6688 rbd_dev_parent_put(rbd_dev); 6689 rbd_object_map_free(rbd_dev); 6690 rbd_dev_mapping_clear(rbd_dev); 6691 6692 /* Free dynamic fields from the header, then zero it out */ 6693 6694 header = &rbd_dev->header; 6695 ceph_put_snap_context(header->snapc); 6696 kfree(header->snap_sizes); 6697 kfree(header->snap_names); 6698 kfree(header->object_prefix); 6699 memset(header, 0, sizeof (*header)); 6700 } 6701 6702 static int rbd_dev_v2_header_onetime(struct rbd_device *rbd_dev) 6703 { 6704 int ret; 6705 6706 ret = rbd_dev_v2_object_prefix(rbd_dev); 6707 if (ret) 6708 goto out_err; 6709 6710 /* 6711 * Get the and check features for the image. Currently the 6712 * features are assumed to never change. 6713 */ 6714 ret = rbd_dev_v2_features(rbd_dev); 6715 if (ret) 6716 goto out_err; 6717 6718 /* If the image supports fancy striping, get its parameters */ 6719 6720 if (rbd_dev->header.features & RBD_FEATURE_STRIPINGV2) { 6721 ret = rbd_dev_v2_striping_info(rbd_dev); 6722 if (ret < 0) 6723 goto out_err; 6724 } 6725 6726 if (rbd_dev->header.features & RBD_FEATURE_DATA_POOL) { 6727 ret = rbd_dev_v2_data_pool(rbd_dev); 6728 if (ret) 6729 goto out_err; 6730 } 6731 6732 rbd_init_layout(rbd_dev); 6733 return 0; 6734 6735 out_err: 6736 rbd_dev->header.features = 0; 6737 kfree(rbd_dev->header.object_prefix); 6738 rbd_dev->header.object_prefix = NULL; 6739 return ret; 6740 } 6741 6742 /* 6743 * @depth is rbd_dev_image_probe() -> rbd_dev_probe_parent() -> 6744 * rbd_dev_image_probe() recursion depth, which means it's also the 6745 * length of the already discovered part of the parent chain. 6746 */ 6747 static int rbd_dev_probe_parent(struct rbd_device *rbd_dev, int depth) 6748 { 6749 struct rbd_device *parent = NULL; 6750 int ret; 6751 6752 if (!rbd_dev->parent_spec) 6753 return 0; 6754 6755 if (++depth > RBD_MAX_PARENT_CHAIN_LEN) { 6756 pr_info("parent chain is too long (%d)\n", depth); 6757 ret = -EINVAL; 6758 goto out_err; 6759 } 6760 6761 parent = __rbd_dev_create(rbd_dev->rbd_client, rbd_dev->parent_spec); 6762 if (!parent) { 6763 ret = -ENOMEM; 6764 goto out_err; 6765 } 6766 6767 /* 6768 * Images related by parent/child relationships always share 6769 * rbd_client and spec/parent_spec, so bump their refcounts. 6770 */ 6771 __rbd_get_client(rbd_dev->rbd_client); 6772 rbd_spec_get(rbd_dev->parent_spec); 6773 6774 __set_bit(RBD_DEV_FLAG_READONLY, &parent->flags); 6775 6776 ret = rbd_dev_image_probe(parent, depth); 6777 if (ret < 0) 6778 goto out_err; 6779 6780 rbd_dev->parent = parent; 6781 atomic_set(&rbd_dev->parent_ref, 1); 6782 return 0; 6783 6784 out_err: 6785 rbd_dev_unparent(rbd_dev); 6786 rbd_dev_destroy(parent); 6787 return ret; 6788 } 6789 6790 static void rbd_dev_device_release(struct rbd_device *rbd_dev) 6791 { 6792 clear_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags); 6793 rbd_free_disk(rbd_dev); 6794 if (!single_major) 6795 unregister_blkdev(rbd_dev->major, rbd_dev->name); 6796 } 6797 6798 /* 6799 * rbd_dev->header_rwsem must be locked for write and will be unlocked 6800 * upon return. 6801 */ 6802 static int rbd_dev_device_setup(struct rbd_device *rbd_dev) 6803 { 6804 int ret; 6805 6806 /* Record our major and minor device numbers. */ 6807 6808 if (!single_major) { 6809 ret = register_blkdev(0, rbd_dev->name); 6810 if (ret < 0) 6811 goto err_out_unlock; 6812 6813 rbd_dev->major = ret; 6814 rbd_dev->minor = 0; 6815 } else { 6816 rbd_dev->major = rbd_major; 6817 rbd_dev->minor = rbd_dev_id_to_minor(rbd_dev->dev_id); 6818 } 6819 6820 /* Set up the blkdev mapping. */ 6821 6822 ret = rbd_init_disk(rbd_dev); 6823 if (ret) 6824 goto err_out_blkdev; 6825 6826 set_capacity(rbd_dev->disk, rbd_dev->mapping.size / SECTOR_SIZE); 6827 set_disk_ro(rbd_dev->disk, rbd_is_ro(rbd_dev)); 6828 6829 ret = dev_set_name(&rbd_dev->dev, "%d", rbd_dev->dev_id); 6830 if (ret) 6831 goto err_out_disk; 6832 6833 set_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags); 6834 up_write(&rbd_dev->header_rwsem); 6835 return 0; 6836 6837 err_out_disk: 6838 rbd_free_disk(rbd_dev); 6839 err_out_blkdev: 6840 if (!single_major) 6841 unregister_blkdev(rbd_dev->major, rbd_dev->name); 6842 err_out_unlock: 6843 up_write(&rbd_dev->header_rwsem); 6844 return ret; 6845 } 6846 6847 static int rbd_dev_header_name(struct rbd_device *rbd_dev) 6848 { 6849 struct rbd_spec *spec = rbd_dev->spec; 6850 int ret; 6851 6852 /* Record the header object name for this rbd image. */ 6853 6854 rbd_assert(rbd_image_format_valid(rbd_dev->image_format)); 6855 if (rbd_dev->image_format == 1) 6856 ret = ceph_oid_aprintf(&rbd_dev->header_oid, GFP_KERNEL, "%s%s", 6857 spec->image_name, RBD_SUFFIX); 6858 else 6859 ret = ceph_oid_aprintf(&rbd_dev->header_oid, GFP_KERNEL, "%s%s", 6860 RBD_HEADER_PREFIX, spec->image_id); 6861 6862 return ret; 6863 } 6864 6865 static void rbd_print_dne(struct rbd_device *rbd_dev, bool is_snap) 6866 { 6867 if (!is_snap) { 6868 pr_info("image %s/%s%s%s does not exist\n", 6869 rbd_dev->spec->pool_name, 6870 rbd_dev->spec->pool_ns ?: "", 6871 rbd_dev->spec->pool_ns ? "/" : "", 6872 rbd_dev->spec->image_name); 6873 } else { 6874 pr_info("snap %s/%s%s%s@%s does not exist\n", 6875 rbd_dev->spec->pool_name, 6876 rbd_dev->spec->pool_ns ?: "", 6877 rbd_dev->spec->pool_ns ? "/" : "", 6878 rbd_dev->spec->image_name, 6879 rbd_dev->spec->snap_name); 6880 } 6881 } 6882 6883 static void rbd_dev_image_release(struct rbd_device *rbd_dev) 6884 { 6885 if (!rbd_is_ro(rbd_dev)) 6886 rbd_unregister_watch(rbd_dev); 6887 6888 rbd_dev_unprobe(rbd_dev); 6889 rbd_dev->image_format = 0; 6890 kfree(rbd_dev->spec->image_id); 6891 rbd_dev->spec->image_id = NULL; 6892 } 6893 6894 /* 6895 * Probe for the existence of the header object for the given rbd 6896 * device. If this image is the one being mapped (i.e., not a 6897 * parent), initiate a watch on its header object before using that 6898 * object to get detailed information about the rbd image. 6899 * 6900 * On success, returns with header_rwsem held for write if called 6901 * with @depth == 0. 6902 */ 6903 static int rbd_dev_image_probe(struct rbd_device *rbd_dev, int depth) 6904 { 6905 bool need_watch = !rbd_is_ro(rbd_dev); 6906 int ret; 6907 6908 /* 6909 * Get the id from the image id object. Unless there's an 6910 * error, rbd_dev->spec->image_id will be filled in with 6911 * a dynamically-allocated string, and rbd_dev->image_format 6912 * will be set to either 1 or 2. 6913 */ 6914 ret = rbd_dev_image_id(rbd_dev); 6915 if (ret) 6916 return ret; 6917 6918 ret = rbd_dev_header_name(rbd_dev); 6919 if (ret) 6920 goto err_out_format; 6921 6922 if (need_watch) { 6923 ret = rbd_register_watch(rbd_dev); 6924 if (ret) { 6925 if (ret == -ENOENT) 6926 rbd_print_dne(rbd_dev, false); 6927 goto err_out_format; 6928 } 6929 } 6930 6931 if (!depth) 6932 down_write(&rbd_dev->header_rwsem); 6933 6934 ret = rbd_dev_header_info(rbd_dev); 6935 if (ret) { 6936 if (ret == -ENOENT && !need_watch) 6937 rbd_print_dne(rbd_dev, false); 6938 goto err_out_probe; 6939 } 6940 6941 /* 6942 * If this image is the one being mapped, we have pool name and 6943 * id, image name and id, and snap name - need to fill snap id. 6944 * Otherwise this is a parent image, identified by pool, image 6945 * and snap ids - need to fill in names for those ids. 6946 */ 6947 if (!depth) 6948 ret = rbd_spec_fill_snap_id(rbd_dev); 6949 else 6950 ret = rbd_spec_fill_names(rbd_dev); 6951 if (ret) { 6952 if (ret == -ENOENT) 6953 rbd_print_dne(rbd_dev, true); 6954 goto err_out_probe; 6955 } 6956 6957 ret = rbd_dev_mapping_set(rbd_dev); 6958 if (ret) 6959 goto err_out_probe; 6960 6961 if (rbd_is_snap(rbd_dev) && 6962 (rbd_dev->header.features & RBD_FEATURE_OBJECT_MAP)) { 6963 ret = rbd_object_map_load(rbd_dev); 6964 if (ret) 6965 goto err_out_probe; 6966 } 6967 6968 if (rbd_dev->header.features & RBD_FEATURE_LAYERING) { 6969 ret = rbd_dev_v2_parent_info(rbd_dev); 6970 if (ret) 6971 goto err_out_probe; 6972 } 6973 6974 ret = rbd_dev_probe_parent(rbd_dev, depth); 6975 if (ret) 6976 goto err_out_probe; 6977 6978 dout("discovered format %u image, header name is %s\n", 6979 rbd_dev->image_format, rbd_dev->header_oid.name); 6980 return 0; 6981 6982 err_out_probe: 6983 if (!depth) 6984 up_write(&rbd_dev->header_rwsem); 6985 if (need_watch) 6986 rbd_unregister_watch(rbd_dev); 6987 rbd_dev_unprobe(rbd_dev); 6988 err_out_format: 6989 rbd_dev->image_format = 0; 6990 kfree(rbd_dev->spec->image_id); 6991 rbd_dev->spec->image_id = NULL; 6992 return ret; 6993 } 6994 6995 static ssize_t do_rbd_add(struct bus_type *bus, 6996 const char *buf, 6997 size_t count) 6998 { 6999 struct rbd_device *rbd_dev = NULL; 7000 struct ceph_options *ceph_opts = NULL; 7001 struct rbd_options *rbd_opts = NULL; 7002 struct rbd_spec *spec = NULL; 7003 struct rbd_client *rbdc; 7004 int rc; 7005 7006 if (!capable(CAP_SYS_ADMIN)) 7007 return -EPERM; 7008 7009 if (!try_module_get(THIS_MODULE)) 7010 return -ENODEV; 7011 7012 /* parse add command */ 7013 rc = rbd_add_parse_args(buf, &ceph_opts, &rbd_opts, &spec); 7014 if (rc < 0) 7015 goto out; 7016 7017 rbdc = rbd_get_client(ceph_opts); 7018 if (IS_ERR(rbdc)) { 7019 rc = PTR_ERR(rbdc); 7020 goto err_out_args; 7021 } 7022 7023 /* pick the pool */ 7024 rc = ceph_pg_poolid_by_name(rbdc->client->osdc.osdmap, spec->pool_name); 7025 if (rc < 0) { 7026 if (rc == -ENOENT) 7027 pr_info("pool %s does not exist\n", spec->pool_name); 7028 goto err_out_client; 7029 } 7030 spec->pool_id = (u64)rc; 7031 7032 rbd_dev = rbd_dev_create(rbdc, spec, rbd_opts); 7033 if (!rbd_dev) { 7034 rc = -ENOMEM; 7035 goto err_out_client; 7036 } 7037 rbdc = NULL; /* rbd_dev now owns this */ 7038 spec = NULL; /* rbd_dev now owns this */ 7039 rbd_opts = NULL; /* rbd_dev now owns this */ 7040 7041 /* if we are mapping a snapshot it will be a read-only mapping */ 7042 if (rbd_dev->opts->read_only || 7043 strcmp(rbd_dev->spec->snap_name, RBD_SNAP_HEAD_NAME)) 7044 __set_bit(RBD_DEV_FLAG_READONLY, &rbd_dev->flags); 7045 7046 rbd_dev->config_info = kstrdup(buf, GFP_KERNEL); 7047 if (!rbd_dev->config_info) { 7048 rc = -ENOMEM; 7049 goto err_out_rbd_dev; 7050 } 7051 7052 rc = rbd_dev_image_probe(rbd_dev, 0); 7053 if (rc < 0) 7054 goto err_out_rbd_dev; 7055 7056 if (rbd_dev->opts->alloc_size > rbd_dev->layout.object_size) { 7057 rbd_warn(rbd_dev, "alloc_size adjusted to %u", 7058 rbd_dev->layout.object_size); 7059 rbd_dev->opts->alloc_size = rbd_dev->layout.object_size; 7060 } 7061 7062 rc = rbd_dev_device_setup(rbd_dev); 7063 if (rc) 7064 goto err_out_image_probe; 7065 7066 rc = rbd_add_acquire_lock(rbd_dev); 7067 if (rc) 7068 goto err_out_image_lock; 7069 7070 /* Everything's ready. Announce the disk to the world. */ 7071 7072 rc = device_add(&rbd_dev->dev); 7073 if (rc) 7074 goto err_out_image_lock; 7075 7076 device_add_disk(&rbd_dev->dev, rbd_dev->disk, NULL); 7077 7078 spin_lock(&rbd_dev_list_lock); 7079 list_add_tail(&rbd_dev->node, &rbd_dev_list); 7080 spin_unlock(&rbd_dev_list_lock); 7081 7082 pr_info("%s: capacity %llu features 0x%llx\n", rbd_dev->disk->disk_name, 7083 (unsigned long long)get_capacity(rbd_dev->disk) << SECTOR_SHIFT, 7084 rbd_dev->header.features); 7085 rc = count; 7086 out: 7087 module_put(THIS_MODULE); 7088 return rc; 7089 7090 err_out_image_lock: 7091 rbd_dev_image_unlock(rbd_dev); 7092 rbd_dev_device_release(rbd_dev); 7093 err_out_image_probe: 7094 rbd_dev_image_release(rbd_dev); 7095 err_out_rbd_dev: 7096 rbd_dev_destroy(rbd_dev); 7097 err_out_client: 7098 rbd_put_client(rbdc); 7099 err_out_args: 7100 rbd_spec_put(spec); 7101 kfree(rbd_opts); 7102 goto out; 7103 } 7104 7105 static ssize_t add_store(struct bus_type *bus, const char *buf, size_t count) 7106 { 7107 if (single_major) 7108 return -EINVAL; 7109 7110 return do_rbd_add(bus, buf, count); 7111 } 7112 7113 static ssize_t add_single_major_store(struct bus_type *bus, const char *buf, 7114 size_t count) 7115 { 7116 return do_rbd_add(bus, buf, count); 7117 } 7118 7119 static void rbd_dev_remove_parent(struct rbd_device *rbd_dev) 7120 { 7121 while (rbd_dev->parent) { 7122 struct rbd_device *first = rbd_dev; 7123 struct rbd_device *second = first->parent; 7124 struct rbd_device *third; 7125 7126 /* 7127 * Follow to the parent with no grandparent and 7128 * remove it. 7129 */ 7130 while (second && (third = second->parent)) { 7131 first = second; 7132 second = third; 7133 } 7134 rbd_assert(second); 7135 rbd_dev_image_release(second); 7136 rbd_dev_destroy(second); 7137 first->parent = NULL; 7138 first->parent_overlap = 0; 7139 7140 rbd_assert(first->parent_spec); 7141 rbd_spec_put(first->parent_spec); 7142 first->parent_spec = NULL; 7143 } 7144 } 7145 7146 static ssize_t do_rbd_remove(struct bus_type *bus, 7147 const char *buf, 7148 size_t count) 7149 { 7150 struct rbd_device *rbd_dev = NULL; 7151 struct list_head *tmp; 7152 int dev_id; 7153 char opt_buf[6]; 7154 bool force = false; 7155 int ret; 7156 7157 if (!capable(CAP_SYS_ADMIN)) 7158 return -EPERM; 7159 7160 dev_id = -1; 7161 opt_buf[0] = '\0'; 7162 sscanf(buf, "%d %5s", &dev_id, opt_buf); 7163 if (dev_id < 0) { 7164 pr_err("dev_id out of range\n"); 7165 return -EINVAL; 7166 } 7167 if (opt_buf[0] != '\0') { 7168 if (!strcmp(opt_buf, "force")) { 7169 force = true; 7170 } else { 7171 pr_err("bad remove option at '%s'\n", opt_buf); 7172 return -EINVAL; 7173 } 7174 } 7175 7176 ret = -ENOENT; 7177 spin_lock(&rbd_dev_list_lock); 7178 list_for_each(tmp, &rbd_dev_list) { 7179 rbd_dev = list_entry(tmp, struct rbd_device, node); 7180 if (rbd_dev->dev_id == dev_id) { 7181 ret = 0; 7182 break; 7183 } 7184 } 7185 if (!ret) { 7186 spin_lock_irq(&rbd_dev->lock); 7187 if (rbd_dev->open_count && !force) 7188 ret = -EBUSY; 7189 else if (test_and_set_bit(RBD_DEV_FLAG_REMOVING, 7190 &rbd_dev->flags)) 7191 ret = -EINPROGRESS; 7192 spin_unlock_irq(&rbd_dev->lock); 7193 } 7194 spin_unlock(&rbd_dev_list_lock); 7195 if (ret) 7196 return ret; 7197 7198 if (force) { 7199 /* 7200 * Prevent new IO from being queued and wait for existing 7201 * IO to complete/fail. 7202 */ 7203 blk_mq_freeze_queue(rbd_dev->disk->queue); 7204 blk_set_queue_dying(rbd_dev->disk->queue); 7205 } 7206 7207 del_gendisk(rbd_dev->disk); 7208 spin_lock(&rbd_dev_list_lock); 7209 list_del_init(&rbd_dev->node); 7210 spin_unlock(&rbd_dev_list_lock); 7211 device_del(&rbd_dev->dev); 7212 7213 rbd_dev_image_unlock(rbd_dev); 7214 rbd_dev_device_release(rbd_dev); 7215 rbd_dev_image_release(rbd_dev); 7216 rbd_dev_destroy(rbd_dev); 7217 return count; 7218 } 7219 7220 static ssize_t remove_store(struct bus_type *bus, const char *buf, size_t count) 7221 { 7222 if (single_major) 7223 return -EINVAL; 7224 7225 return do_rbd_remove(bus, buf, count); 7226 } 7227 7228 static ssize_t remove_single_major_store(struct bus_type *bus, const char *buf, 7229 size_t count) 7230 { 7231 return do_rbd_remove(bus, buf, count); 7232 } 7233 7234 /* 7235 * create control files in sysfs 7236 * /sys/bus/rbd/... 7237 */ 7238 static int __init rbd_sysfs_init(void) 7239 { 7240 int ret; 7241 7242 ret = device_register(&rbd_root_dev); 7243 if (ret < 0) 7244 return ret; 7245 7246 ret = bus_register(&rbd_bus_type); 7247 if (ret < 0) 7248 device_unregister(&rbd_root_dev); 7249 7250 return ret; 7251 } 7252 7253 static void __exit rbd_sysfs_cleanup(void) 7254 { 7255 bus_unregister(&rbd_bus_type); 7256 device_unregister(&rbd_root_dev); 7257 } 7258 7259 static int __init rbd_slab_init(void) 7260 { 7261 rbd_assert(!rbd_img_request_cache); 7262 rbd_img_request_cache = KMEM_CACHE(rbd_img_request, 0); 7263 if (!rbd_img_request_cache) 7264 return -ENOMEM; 7265 7266 rbd_assert(!rbd_obj_request_cache); 7267 rbd_obj_request_cache = KMEM_CACHE(rbd_obj_request, 0); 7268 if (!rbd_obj_request_cache) 7269 goto out_err; 7270 7271 return 0; 7272 7273 out_err: 7274 kmem_cache_destroy(rbd_img_request_cache); 7275 rbd_img_request_cache = NULL; 7276 return -ENOMEM; 7277 } 7278 7279 static void rbd_slab_exit(void) 7280 { 7281 rbd_assert(rbd_obj_request_cache); 7282 kmem_cache_destroy(rbd_obj_request_cache); 7283 rbd_obj_request_cache = NULL; 7284 7285 rbd_assert(rbd_img_request_cache); 7286 kmem_cache_destroy(rbd_img_request_cache); 7287 rbd_img_request_cache = NULL; 7288 } 7289 7290 static int __init rbd_init(void) 7291 { 7292 int rc; 7293 7294 if (!libceph_compatible(NULL)) { 7295 rbd_warn(NULL, "libceph incompatibility (quitting)"); 7296 return -EINVAL; 7297 } 7298 7299 rc = rbd_slab_init(); 7300 if (rc) 7301 return rc; 7302 7303 /* 7304 * The number of active work items is limited by the number of 7305 * rbd devices * queue depth, so leave @max_active at default. 7306 */ 7307 rbd_wq = alloc_workqueue(RBD_DRV_NAME, WQ_MEM_RECLAIM, 0); 7308 if (!rbd_wq) { 7309 rc = -ENOMEM; 7310 goto err_out_slab; 7311 } 7312 7313 if (single_major) { 7314 rbd_major = register_blkdev(0, RBD_DRV_NAME); 7315 if (rbd_major < 0) { 7316 rc = rbd_major; 7317 goto err_out_wq; 7318 } 7319 } 7320 7321 rc = rbd_sysfs_init(); 7322 if (rc) 7323 goto err_out_blkdev; 7324 7325 if (single_major) 7326 pr_info("loaded (major %d)\n", rbd_major); 7327 else 7328 pr_info("loaded\n"); 7329 7330 return 0; 7331 7332 err_out_blkdev: 7333 if (single_major) 7334 unregister_blkdev(rbd_major, RBD_DRV_NAME); 7335 err_out_wq: 7336 destroy_workqueue(rbd_wq); 7337 err_out_slab: 7338 rbd_slab_exit(); 7339 return rc; 7340 } 7341 7342 static void __exit rbd_exit(void) 7343 { 7344 ida_destroy(&rbd_dev_id_ida); 7345 rbd_sysfs_cleanup(); 7346 if (single_major) 7347 unregister_blkdev(rbd_major, RBD_DRV_NAME); 7348 destroy_workqueue(rbd_wq); 7349 rbd_slab_exit(); 7350 } 7351 7352 module_init(rbd_init); 7353 module_exit(rbd_exit); 7354 7355 MODULE_AUTHOR("Alex Elder <elder@inktank.com>"); 7356 MODULE_AUTHOR("Sage Weil <sage@newdream.net>"); 7357 MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>"); 7358 /* following authorship retained from original osdblk.c */ 7359 MODULE_AUTHOR("Jeff Garzik <jeff@garzik.org>"); 7360 7361 MODULE_DESCRIPTION("RADOS Block Device (RBD) driver"); 7362 MODULE_LICENSE("GPL"); 7363