1// SPDX-License-Identifier: GPL-2.0 2#include <linux/ceph/ceph_debug.h> 3 4#include <linux/file.h> 5#include <linux/namei.h> 6#include <linux/random.h> 7 8#include "super.h" 9#include "mds_client.h" 10#include <linux/ceph/pagelist.h> 11 12static u64 lock_secret; 13static int ceph_lock_wait_for_completion(struct ceph_mds_client *mdsc, 14 struct ceph_mds_request *req); 15 16static inline u64 secure_addr(void *addr) 17{ 18 u64 v = lock_secret ^ (u64)(unsigned long)addr; 19 /* 20 * Set the most significant bit, so that MDS knows the 'owner' 21 * is sufficient to identify the owner of lock. (old code uses 22 * both 'owner' and 'pid') 23 */ 24 v |= (1ULL << 63); 25 return v; 26} 27 28void __init ceph_flock_init(void) 29{ 30 get_random_bytes(&lock_secret, sizeof(lock_secret)); 31} 32 33static void ceph_fl_copy_lock(struct file_lock *dst, struct file_lock *src) 34{ 35 struct inode *inode = file_inode(dst->fl_file); 36 atomic_inc(&ceph_inode(inode)->i_filelock_ref); 37} 38 39static void ceph_fl_release_lock(struct file_lock *fl) 40{ 41 struct inode *inode = file_inode(fl->fl_file); 42 struct ceph_inode_info *ci = ceph_inode(inode); 43 if (atomic_dec_and_test(&ci->i_filelock_ref)) { 44 /* clear error when all locks are released */ 45 spin_lock(&ci->i_ceph_lock); 46 ci->i_ceph_flags &= ~CEPH_I_ERROR_FILELOCK; 47 spin_unlock(&ci->i_ceph_lock); 48 } 49} 50 51static const struct file_lock_operations ceph_fl_lock_ops = { 52 .fl_copy_lock = ceph_fl_copy_lock, 53 .fl_release_private = ceph_fl_release_lock, 54}; 55 56/** 57 * Implement fcntl and flock locking functions. 58 */ 59static int ceph_lock_message(u8 lock_type, u16 operation, struct inode *inode, 60 int cmd, u8 wait, struct file_lock *fl) 61{ 62 struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(inode->i_sb); 63 struct ceph_mds_request *req; 64 int err; 65 u64 length = 0; 66 u64 owner; 67 68 if (operation == CEPH_MDS_OP_SETFILELOCK) { 69 /* 70 * increasing i_filelock_ref closes race window between 71 * handling request reply and adding file_lock struct to 72 * inode. Otherwise, auth caps may get trimmed in the 73 * window. Caller function will decrease the counter. 74 */ 75 fl->fl_ops = &ceph_fl_lock_ops; 76 fl->fl_ops->fl_copy_lock(fl, NULL); 77 } 78 79 if (operation != CEPH_MDS_OP_SETFILELOCK || cmd == CEPH_LOCK_UNLOCK) 80 wait = 0; 81 82 req = ceph_mdsc_create_request(mdsc, operation, USE_AUTH_MDS); 83 if (IS_ERR(req)) 84 return PTR_ERR(req); 85 req->r_inode = inode; 86 ihold(inode); 87 req->r_num_caps = 1; 88 89 /* mds requires start and length rather than start and end */ 90 if (LLONG_MAX == fl->fl_end) 91 length = 0; 92 else 93 length = fl->fl_end - fl->fl_start + 1; 94 95 owner = secure_addr(fl->fl_owner); 96 97 dout("ceph_lock_message: rule: %d, op: %d, owner: %llx, pid: %llu, " 98 "start: %llu, length: %llu, wait: %d, type: %d\n", (int)lock_type, 99 (int)operation, owner, (u64)fl->fl_pid, fl->fl_start, length, 100 wait, fl->fl_type); 101 102 req->r_args.filelock_change.rule = lock_type; 103 req->r_args.filelock_change.type = cmd; 104 req->r_args.filelock_change.owner = cpu_to_le64(owner); 105 req->r_args.filelock_change.pid = cpu_to_le64((u64)fl->fl_pid); 106 req->r_args.filelock_change.start = cpu_to_le64(fl->fl_start); 107 req->r_args.filelock_change.length = cpu_to_le64(length); 108 req->r_args.filelock_change.wait = wait; 109 110 if (wait) 111 req->r_wait_for_completion = ceph_lock_wait_for_completion; 112 113 err = ceph_mdsc_do_request(mdsc, inode, req); 114 if (!err && operation == CEPH_MDS_OP_GETFILELOCK) { 115 fl->fl_pid = -le64_to_cpu(req->r_reply_info.filelock_reply->pid); 116 if (CEPH_LOCK_SHARED == req->r_reply_info.filelock_reply->type) 117 fl->fl_type = F_RDLCK; 118 else if (CEPH_LOCK_EXCL == req->r_reply_info.filelock_reply->type) 119 fl->fl_type = F_WRLCK; 120 else 121 fl->fl_type = F_UNLCK; 122 123 fl->fl_start = le64_to_cpu(req->r_reply_info.filelock_reply->start); 124 length = le64_to_cpu(req->r_reply_info.filelock_reply->start) + 125 le64_to_cpu(req->r_reply_info.filelock_reply->length); 126 if (length >= 1) 127 fl->fl_end = length -1; 128 else 129 fl->fl_end = 0; 130 131 } 132 ceph_mdsc_put_request(req); 133 dout("ceph_lock_message: rule: %d, op: %d, pid: %llu, start: %llu, " 134 "length: %llu, wait: %d, type: %d, err code %d\n", (int)lock_type, 135 (int)operation, (u64)fl->fl_pid, fl->fl_start, 136 length, wait, fl->fl_type, err); 137 return err; 138} 139 140static int ceph_lock_wait_for_completion(struct ceph_mds_client *mdsc, 141 struct ceph_mds_request *req) 142{ 143 struct ceph_mds_request *intr_req; 144 struct inode *inode = req->r_inode; 145 int err, lock_type; 146 147 BUG_ON(req->r_op != CEPH_MDS_OP_SETFILELOCK); 148 if (req->r_args.filelock_change.rule == CEPH_LOCK_FCNTL) 149 lock_type = CEPH_LOCK_FCNTL_INTR; 150 else if (req->r_args.filelock_change.rule == CEPH_LOCK_FLOCK) 151 lock_type = CEPH_LOCK_FLOCK_INTR; 152 else 153 BUG_ON(1); 154 BUG_ON(req->r_args.filelock_change.type == CEPH_LOCK_UNLOCK); 155 156 err = wait_for_completion_interruptible(&req->r_completion); 157 if (!err) 158 return 0; 159 160 dout("ceph_lock_wait_for_completion: request %llu was interrupted\n", 161 req->r_tid); 162 163 mutex_lock(&mdsc->mutex); 164 if (test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags)) { 165 err = 0; 166 } else { 167 /* 168 * ensure we aren't running concurrently with 169 * ceph_fill_trace or ceph_readdir_prepopulate, which 170 * rely on locks (dir mutex) held by our caller. 171 */ 172 mutex_lock(&req->r_fill_mutex); 173 req->r_err = err; 174 set_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags); 175 mutex_unlock(&req->r_fill_mutex); 176 177 if (!req->r_session) { 178 // haven't sent the request 179 err = 0; 180 } 181 } 182 mutex_unlock(&mdsc->mutex); 183 if (!err) 184 return 0; 185 186 intr_req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_SETFILELOCK, 187 USE_AUTH_MDS); 188 if (IS_ERR(intr_req)) 189 return PTR_ERR(intr_req); 190 191 intr_req->r_inode = inode; 192 ihold(inode); 193 intr_req->r_num_caps = 1; 194 195 intr_req->r_args.filelock_change = req->r_args.filelock_change; 196 intr_req->r_args.filelock_change.rule = lock_type; 197 intr_req->r_args.filelock_change.type = CEPH_LOCK_UNLOCK; 198 199 err = ceph_mdsc_do_request(mdsc, inode, intr_req); 200 ceph_mdsc_put_request(intr_req); 201 202 if (err && err != -ERESTARTSYS) 203 return err; 204 205 wait_for_completion_killable(&req->r_safe_completion); 206 return 0; 207} 208 209static int try_unlock_file(struct file *file, struct file_lock *fl) 210{ 211 int err; 212 unsigned int orig_flags = fl->fl_flags; 213 fl->fl_flags |= FL_EXISTS; 214 err = locks_lock_file_wait(file, fl); 215 fl->fl_flags = orig_flags; 216 if (err == -ENOENT) { 217 if (!(orig_flags & FL_EXISTS)) 218 err = 0; 219 return err; 220 } 221 return 1; 222} 223 224/** 225 * Attempt to set an fcntl lock. 226 * For now, this just goes away to the server. Later it may be more awesome. 227 */ 228int ceph_lock(struct file *file, int cmd, struct file_lock *fl) 229{ 230 struct inode *inode = file_inode(file); 231 struct ceph_inode_info *ci = ceph_inode(inode); 232 int err = 0; 233 u16 op = CEPH_MDS_OP_SETFILELOCK; 234 u8 wait = 0; 235 u8 lock_cmd; 236 237 if (!(fl->fl_flags & FL_POSIX)) 238 return -ENOLCK; 239 /* No mandatory locks */ 240 if (__mandatory_lock(file->f_mapping->host) && fl->fl_type != F_UNLCK) 241 return -ENOLCK; 242 243 dout("ceph_lock, fl_owner: %p\n", fl->fl_owner); 244 245 /* set wait bit as appropriate, then make command as Ceph expects it*/ 246 if (IS_GETLK(cmd)) 247 op = CEPH_MDS_OP_GETFILELOCK; 248 else if (IS_SETLKW(cmd)) 249 wait = 1; 250 251 spin_lock(&ci->i_ceph_lock); 252 if (ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) { 253 err = -EIO; 254 } 255 spin_unlock(&ci->i_ceph_lock); 256 if (err < 0) { 257 if (op == CEPH_MDS_OP_SETFILELOCK && F_UNLCK == fl->fl_type) 258 posix_lock_file(file, fl, NULL); 259 return err; 260 } 261 262 if (F_RDLCK == fl->fl_type) 263 lock_cmd = CEPH_LOCK_SHARED; 264 else if (F_WRLCK == fl->fl_type) 265 lock_cmd = CEPH_LOCK_EXCL; 266 else 267 lock_cmd = CEPH_LOCK_UNLOCK; 268 269 if (op == CEPH_MDS_OP_SETFILELOCK && F_UNLCK == fl->fl_type) { 270 err = try_unlock_file(file, fl); 271 if (err <= 0) 272 return err; 273 } 274 275 err = ceph_lock_message(CEPH_LOCK_FCNTL, op, inode, lock_cmd, wait, fl); 276 if (!err) { 277 if (op == CEPH_MDS_OP_SETFILELOCK && F_UNLCK != fl->fl_type) { 278 dout("mds locked, locking locally\n"); 279 err = posix_lock_file(file, fl, NULL); 280 if (err) { 281 /* undo! This should only happen if 282 * the kernel detects local 283 * deadlock. */ 284 ceph_lock_message(CEPH_LOCK_FCNTL, op, inode, 285 CEPH_LOCK_UNLOCK, 0, fl); 286 dout("got %d on posix_lock_file, undid lock\n", 287 err); 288 } 289 } 290 } 291 return err; 292} 293 294int ceph_flock(struct file *file, int cmd, struct file_lock *fl) 295{ 296 struct inode *inode = file_inode(file); 297 struct ceph_inode_info *ci = ceph_inode(inode); 298 int err = 0; 299 u8 wait = 0; 300 u8 lock_cmd; 301 302 if (!(fl->fl_flags & FL_FLOCK)) 303 return -ENOLCK; 304 /* No mandatory locks */ 305 if (fl->fl_type & LOCK_MAND) 306 return -EOPNOTSUPP; 307 308 dout("ceph_flock, fl_file: %p\n", fl->fl_file); 309 310 spin_lock(&ci->i_ceph_lock); 311 if (ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) { 312 err = -EIO; 313 } 314 spin_unlock(&ci->i_ceph_lock); 315 if (err < 0) { 316 if (F_UNLCK == fl->fl_type) 317 locks_lock_file_wait(file, fl); 318 return err; 319 } 320 321 if (IS_SETLKW(cmd)) 322 wait = 1; 323 324 if (F_RDLCK == fl->fl_type) 325 lock_cmd = CEPH_LOCK_SHARED; 326 else if (F_WRLCK == fl->fl_type) 327 lock_cmd = CEPH_LOCK_EXCL; 328 else 329 lock_cmd = CEPH_LOCK_UNLOCK; 330 331 if (F_UNLCK == fl->fl_type) { 332 err = try_unlock_file(file, fl); 333 if (err <= 0) 334 return err; 335 } 336 337 err = ceph_lock_message(CEPH_LOCK_FLOCK, CEPH_MDS_OP_SETFILELOCK, 338 inode, lock_cmd, wait, fl); 339 if (!err && F_UNLCK != fl->fl_type) { 340 err = locks_lock_file_wait(file, fl); 341 if (err) { 342 ceph_lock_message(CEPH_LOCK_FLOCK, 343 CEPH_MDS_OP_SETFILELOCK, 344 inode, CEPH_LOCK_UNLOCK, 0, fl); 345 dout("got %d on locks_lock_file_wait, undid lock\n", err); 346 } 347 } 348 return err; 349} 350 351/* 352 * Fills in the passed counter variables, so you can prepare pagelist metadata 353 * before calling ceph_encode_locks. 354 */ 355void ceph_count_locks(struct inode *inode, int *fcntl_count, int *flock_count) 356{ 357 struct file_lock *lock; 358 struct file_lock_context *ctx; 359 360 *fcntl_count = 0; 361 *flock_count = 0; 362 363 ctx = inode->i_flctx; 364 if (ctx) { 365 spin_lock(&ctx->flc_lock); 366 list_for_each_entry(lock, &ctx->flc_posix, fl_list) 367 ++(*fcntl_count); 368 list_for_each_entry(lock, &ctx->flc_flock, fl_list) 369 ++(*flock_count); 370 spin_unlock(&ctx->flc_lock); 371 } 372 dout("counted %d flock locks and %d fcntl locks\n", 373 *flock_count, *fcntl_count); 374} 375 376/* 377 * Given a pointer to a lock, convert it to a ceph filelock 378 */ 379static int lock_to_ceph_filelock(struct file_lock *lock, 380 struct ceph_filelock *cephlock) 381{ 382 int err = 0; 383 cephlock->start = cpu_to_le64(lock->fl_start); 384 cephlock->length = cpu_to_le64(lock->fl_end - lock->fl_start + 1); 385 cephlock->client = cpu_to_le64(0); 386 cephlock->pid = cpu_to_le64((u64)lock->fl_pid); 387 cephlock->owner = cpu_to_le64(secure_addr(lock->fl_owner)); 388 389 switch (lock->fl_type) { 390 case F_RDLCK: 391 cephlock->type = CEPH_LOCK_SHARED; 392 break; 393 case F_WRLCK: 394 cephlock->type = CEPH_LOCK_EXCL; 395 break; 396 case F_UNLCK: 397 cephlock->type = CEPH_LOCK_UNLOCK; 398 break; 399 default: 400 dout("Have unknown lock type %d\n", lock->fl_type); 401 err = -EINVAL; 402 } 403 404 return err; 405} 406 407/** 408 * Encode the flock and fcntl locks for the given inode into the ceph_filelock 409 * array. Must be called with inode->i_lock already held. 410 * If we encounter more of a specific lock type than expected, return -ENOSPC. 411 */ 412int ceph_encode_locks_to_buffer(struct inode *inode, 413 struct ceph_filelock *flocks, 414 int num_fcntl_locks, int num_flock_locks) 415{ 416 struct file_lock *lock; 417 struct file_lock_context *ctx = inode->i_flctx; 418 int err = 0; 419 int seen_fcntl = 0; 420 int seen_flock = 0; 421 int l = 0; 422 423 dout("encoding %d flock and %d fcntl locks\n", num_flock_locks, 424 num_fcntl_locks); 425 426 if (!ctx) 427 return 0; 428 429 spin_lock(&ctx->flc_lock); 430 list_for_each_entry(lock, &ctx->flc_posix, fl_list) { 431 ++seen_fcntl; 432 if (seen_fcntl > num_fcntl_locks) { 433 err = -ENOSPC; 434 goto fail; 435 } 436 err = lock_to_ceph_filelock(lock, &flocks[l]); 437 if (err) 438 goto fail; 439 ++l; 440 } 441 list_for_each_entry(lock, &ctx->flc_flock, fl_list) { 442 ++seen_flock; 443 if (seen_flock > num_flock_locks) { 444 err = -ENOSPC; 445 goto fail; 446 } 447 err = lock_to_ceph_filelock(lock, &flocks[l]); 448 if (err) 449 goto fail; 450 ++l; 451 } 452fail: 453 spin_unlock(&ctx->flc_lock); 454 return err; 455} 456 457/** 458 * Copy the encoded flock and fcntl locks into the pagelist. 459 * Format is: #fcntl locks, sequential fcntl locks, #flock locks, 460 * sequential flock locks. 461 * Returns zero on success. 462 */ 463int ceph_locks_to_pagelist(struct ceph_filelock *flocks, 464 struct ceph_pagelist *pagelist, 465 int num_fcntl_locks, int num_flock_locks) 466{ 467 int err = 0; 468 __le32 nlocks; 469 470 nlocks = cpu_to_le32(num_fcntl_locks); 471 err = ceph_pagelist_append(pagelist, &nlocks, sizeof(nlocks)); 472 if (err) 473 goto out_fail; 474 475 if (num_fcntl_locks > 0) { 476 err = ceph_pagelist_append(pagelist, flocks, 477 num_fcntl_locks * sizeof(*flocks)); 478 if (err) 479 goto out_fail; 480 } 481 482 nlocks = cpu_to_le32(num_flock_locks); 483 err = ceph_pagelist_append(pagelist, &nlocks, sizeof(nlocks)); 484 if (err) 485 goto out_fail; 486 487 if (num_flock_locks > 0) { 488 err = ceph_pagelist_append(pagelist, &flocks[num_fcntl_locks], 489 num_flock_locks * sizeof(*flocks)); 490 } 491out_fail: 492 return err; 493} 494