fs/ceph/mds_client.c

   1 // SPDX-License-Identifier: GPL-2.0
   2 #include <linux/ceph/ceph_debug.h>
   3
   4 #include <linux/fs.h>
   5 #include <linux/wait.h>
   6 #include <linux/slab.h>
   7 #include <linux/gfp.h>
   8 #include <linux/sched.h>
   9 #include <linux/debugfs.h>
  10 #include <linux/seq_file.h>
  11 #include <linux/ratelimit.h>
  12 #include <linux/bits.h>
  13 #include <linux/ktime.h>
  14
  15 #include "super.h"
  16 #include "mds_client.h"
  17
  18 #include <linux/ceph/ceph_features.h>
  19 #include <linux/ceph/messenger.h>
  20 #include <linux/ceph/decode.h>
  21 #include <linux/ceph/pagelist.h>
  22 #include <linux/ceph/auth.h>
  23 #include <linux/ceph/debugfs.h>
  24
  25 #define RECONNECT_MAX_SIZE (INT_MAX - PAGE_SIZE)
  26
  27 /*
  28  * A cluster of MDS (metadata server) daemons is responsible for
  29  * managing the file system namespace (the directory hierarchy and
  30  * inodes) and for coordinating shared access to storage.  Metadata is
  31  * partitioning hierarchically across a number of servers, and that
  32  * partition varies over time as the cluster adjusts the distribution
  33  * in order to balance load.
  34  *
  35  * The MDS client is primarily responsible to managing synchronous
  36  * metadata requests for operations like open, unlink, and so forth.
  37  * If there is a MDS failure, we find out about it when we (possibly
  38  * request and) receive a new MDS map, and can resubmit affected
  39  * requests.
  40  *
  41  * For the most part, though, we take advantage of a lossless
  42  * communications channel to the MDS, and do not need to worry about
  43  * timing out or resubmitting requests.
  44  *
  45  * We maintain a stateful "session" with each MDS we interact with.
  46  * Within each session, we sent periodic heartbeat messages to ensure
  47  * any capabilities or leases we have been issues remain valid.  If
  48  * the session times out and goes stale, our leases and capabilities
  49  * are no longer valid.
  50  */
  51
  52 struct ceph_reconnect_state {
  53         struct ceph_mds_session *session;
  54         int nr_caps, nr_realms;
  55         struct ceph_pagelist *pagelist;
  56         unsigned msg_version;
  57         bool allow_multi;
  58 };
  59
  60 static void __wake_requests(struct ceph_mds_client *mdsc,
  61                             struct list_head *head);
  62 static void ceph_cap_release_work(struct work_struct *work);
  63 static void ceph_cap_reclaim_work(struct work_struct *work);
  64
  65 static const struct ceph_connection_operations mds_con_ops;
  66
  67
  68 /*
  69  * mds reply parsing
  70  */
  71
  72 static int parse_reply_info_quota(void **p, void *end,
  73                                   struct ceph_mds_reply_info_in *info)
  74 {
  75         u8 struct_v, struct_compat;
  76         u32 struct_len;
  77
  78         ceph_decode_8_safe(p, end, struct_v, bad);
  79         ceph_decode_8_safe(p, end, struct_compat, bad);
  80         /* struct_v is expected to be >= 1. we only
  81          * understand encoding with struct_compat == 1. */
  82         if (!struct_v || struct_compat != 1)
  83                 goto bad;
  84         ceph_decode_32_safe(p, end, struct_len, bad);
  85         ceph_decode_need(p, end, struct_len, bad);
  86         end = *p + struct_len;
  87         ceph_decode_64_safe(p, end, info->max_bytes, bad);
  88         ceph_decode_64_safe(p, end, info->max_files, bad);
  89         *p = end;
  90         return 0;
  91 bad:
  92         return -EIO;
  93 }
  94
  95 /*
  96  * parse individual inode info
  97  */
  98 static int parse_reply_info_in(void **p, void *end,
  99                                struct ceph_mds_reply_info_in *info,
 100                                u64 features)
 101 {
 102         int err = 0;
 103         u8 struct_v = 0;
 104
 105         if (features == (u64)-1) {
 106                 u32 struct_len;
 107                 u8 struct_compat;
 108                 ceph_decode_8_safe(p, end, struct_v, bad);
 109                 ceph_decode_8_safe(p, end, struct_compat, bad);
 110                 /* struct_v is expected to be >= 1. we only understand
 111                  * encoding with struct_compat == 1. */
 112                 if (!struct_v || struct_compat != 1)
 113                         goto bad;
 114                 ceph_decode_32_safe(p, end, struct_len, bad);
 115                 ceph_decode_need(p, end, struct_len, bad);
 116                 end = *p + struct_len;
 117         }
 118
 119         ceph_decode_need(p, end, sizeof(struct ceph_mds_reply_inode), bad);
 120         info->in = *p;
 121         *p += sizeof(struct ceph_mds_reply_inode) +
 122                 sizeof(*info->in->fragtree.splits) *
 123                 le32_to_cpu(info->in->fragtree.nsplits);
 124
 125         ceph_decode_32_safe(p, end, info->symlink_len, bad);
 126         ceph_decode_need(p, end, info->symlink_len, bad);
 127         info->symlink = *p;
 128         *p += info->symlink_len;
 129
 130         ceph_decode_copy_safe(p, end, &info->dir_layout,
 131                               sizeof(info->dir_layout), bad);
 132         ceph_decode_32_safe(p, end, info->xattr_len, bad);
 133         ceph_decode_need(p, end, info->xattr_len, bad);
 134         info->xattr_data = *p;
 135         *p += info->xattr_len;
 136
 137         if (features == (u64)-1) {
 138                 /* inline data */
 139                 ceph_decode_64_safe(p, end, info->inline_version, bad);
 140                 ceph_decode_32_safe(p, end, info->inline_len, bad);
 141                 ceph_decode_need(p, end, info->inline_len, bad);
 142                 info->inline_data = *p;
 143                 *p += info->inline_len;
 144                 /* quota */
 145                 err = parse_reply_info_quota(p, end, info);
 146                 if (err < 0)
 147                         goto out_bad;
 148                 /* pool namespace */
 149                 ceph_decode_32_safe(p, end, info->pool_ns_len, bad);
 150                 if (info->pool_ns_len > 0) {
 151                         ceph_decode_need(p, end, info->pool_ns_len, bad);
 152                         info->pool_ns_data = *p;
 153                         *p += info->pool_ns_len;
 154                 }
 155
 156                 /* btime */
 157                 ceph_decode_need(p, end, sizeof(info->btime), bad);
 158                 ceph_decode_copy(p, &info->btime, sizeof(info->btime));
 159
 160                 /* change attribute */
 161                 ceph_decode_64_safe(p, end, info->change_attr, bad);
 162
 163                 /* dir pin */
 164                 if (struct_v >= 2) {
 165                         ceph_decode_32_safe(p, end, info->dir_pin, bad);
 166                 } else {
 167                         info->dir_pin = -ENODATA;
 168                 }
 169
 170                 /* snapshot birth time, remains zero for v<=2 */
 171                 if (struct_v >= 3) {
 172                         ceph_decode_need(p, end, sizeof(info->snap_btime), bad);
 173                         ceph_decode_copy(p, &info->snap_btime,
 174                                          sizeof(info->snap_btime));
 175                 } else {
 176                         memset(&info->snap_btime, 0, sizeof(info->snap_btime));
 177                 }
 178
 179                 /* snapshot count, remains zero for v<=3 */
 180                 if (struct_v >= 4) {
 181                         ceph_decode_64_safe(p, end, info->rsnaps, bad);
 182                 } else {
 183                         info->rsnaps = 0;
 184                 }
 185
 186                 *p = end;
 187         } else {
 188                 if (features & CEPH_FEATURE_MDS_INLINE_DATA) {
 189                         ceph_decode_64_safe(p, end, info->inline_version, bad);
 190                         ceph_decode_32_safe(p, end, info->inline_len, bad);
 191                         ceph_decode_need(p, end, info->inline_len, bad);
 192                         info->inline_data = *p;
 193                         *p += info->inline_len;
 194                 } else
 195                         info->inline_version = CEPH_INLINE_NONE;
 196
 197                 if (features & CEPH_FEATURE_MDS_QUOTA) {
 198                         err = parse_reply_info_quota(p, end, info);
 199                         if (err < 0)
 200                                 goto out_bad;
 201                 } else {
 202                         info->max_bytes = 0;
 203                         info->max_files = 0;
 204                 }
 205
 206                 info->pool_ns_len = 0;
 207                 info->pool_ns_data = NULL;
 208                 if (features & CEPH_FEATURE_FS_FILE_LAYOUT_V2) {
 209                         ceph_decode_32_safe(p, end, info->pool_ns_len, bad);
 210                         if (info->pool_ns_len > 0) {
 211                                 ceph_decode_need(p, end, info->pool_ns_len, bad);
 212                                 info->pool_ns_data = *p;
 213                                 *p += info->pool_ns_len;
 214                         }
 215                 }
 216
 217                 if (features & CEPH_FEATURE_FS_BTIME) {
 218                         ceph_decode_need(p, end, sizeof(info->btime), bad);
 219                         ceph_decode_copy(p, &info->btime, sizeof(info->btime));
 220                         ceph_decode_64_safe(p, end, info->change_attr, bad);
 221                 }
 222
 223                 info->dir_pin = -ENODATA;
 224                 /* info->snap_btime and info->rsnaps remain zero */
 225         }
 226         return 0;
 227 bad:
 228         err = -EIO;
 229 out_bad:
 230         return err;
 231 }
 232
 233 static int parse_reply_info_dir(void **p, void *end,
 234                                 struct ceph_mds_reply_dirfrag **dirfrag,
 235                                 u64 features)
 236 {
 237         if (features == (u64)-1) {
 238                 u8 struct_v, struct_compat;
 239                 u32 struct_len;
 240                 ceph_decode_8_safe(p, end, struct_v, bad);
 241                 ceph_decode_8_safe(p, end, struct_compat, bad);
 242                 /* struct_v is expected to be >= 1. we only understand
 243                  * encoding whose struct_compat == 1. */
 244                 if (!struct_v || struct_compat != 1)
 245                         goto bad;
 246                 ceph_decode_32_safe(p, end, struct_len, bad);
 247                 ceph_decode_need(p, end, struct_len, bad);
 248                 end = *p + struct_len;
 249         }
 250
 251         ceph_decode_need(p, end, sizeof(**dirfrag), bad);
 252         *dirfrag = *p;
 253         *p += sizeof(**dirfrag) + sizeof(u32) * le32_to_cpu((*dirfrag)->ndist);
 254         if (unlikely(*p > end))
 255                 goto bad;
 256         if (features == (u64)-1)
 257                 *p = end;
 258         return 0;
 259 bad:
 260         return -EIO;
 261 }
 262
 263 static int parse_reply_info_lease(void **p, void *end,
 264                                   struct ceph_mds_reply_lease **lease,
 265                                   u64 features)
 266 {
 267         if (features == (u64)-1) {
 268                 u8 struct_v, struct_compat;
 269                 u32 struct_len;
 270                 ceph_decode_8_safe(p, end, struct_v, bad);
 271                 ceph_decode_8_safe(p, end, struct_compat, bad);
 272                 /* struct_v is expected to be >= 1. we only understand
 273                  * encoding whose struct_compat == 1. */
 274                 if (!struct_v || struct_compat != 1)
 275                         goto bad;
 276                 ceph_decode_32_safe(p, end, struct_len, bad);
 277                 ceph_decode_need(p, end, struct_len, bad);
 278                 end = *p + struct_len;
 279         }
 280
 281         ceph_decode_need(p, end, sizeof(**lease), bad);
 282         *lease = *p;
 283         *p += sizeof(**lease);
 284         if (features == (u64)-1)
 285                 *p = end;
 286         return 0;
 287 bad:
 288         return -EIO;
 289 }
 290
 291 /*
 292  * parse a normal reply, which may contain a (dir+)dentry and/or a
 293  * target inode.
 294  */
 295 static int parse_reply_info_trace(void **p, void *end,
 296                                   struct ceph_mds_reply_info_parsed *info,
 297                                   u64 features)
 298 {
 299         int err;
 300
 301         if (info->head->is_dentry) {
 302                 err = parse_reply_info_in(p, end, &info->diri, features);
 303                 if (err < 0)
 304                         goto out_bad;
 305
 306                 err = parse_reply_info_dir(p, end, &info->dirfrag, features);
 307                 if (err < 0)
 308                         goto out_bad;
 309
 310                 ceph_decode_32_safe(p, end, info->dname_len, bad);
 311                 ceph_decode_need(p, end, info->dname_len, bad);
 312                 info->dname = *p;
 313                 *p += info->dname_len;
 314
 315                 err = parse_reply_info_lease(p, end, &info->dlease, features);
 316                 if (err < 0)
 317                         goto out_bad;
 318         }
 319
 320         if (info->head->is_target) {
 321                 err = parse_reply_info_in(p, end, &info->targeti, features);
 322                 if (err < 0)
 323                         goto out_bad;
 324         }
 325
 326         if (unlikely(*p != end))
 327                 goto bad;
 328         return 0;
 329
 330 bad:
 331         err = -EIO;
 332 out_bad:
 333         pr_err("problem parsing mds trace %d\n", err);
 334         return err;
 335 }
 336
 337 /*
 338  * parse readdir results
 339  */
 340 static int parse_reply_info_readdir(void **p, void *end,
 341                                 struct ceph_mds_reply_info_parsed *info,
 342                                 u64 features)
 343 {
 344         u32 num, i = 0;
 345         int err;
 346
 347         err = parse_reply_info_dir(p, end, &info->dir_dir, features);
 348         if (err < 0)
 349                 goto out_bad;
 350
 351         ceph_decode_need(p, end, sizeof(num) + 2, bad);
 352         num = ceph_decode_32(p);
 353         {
 354                 u16 flags = ceph_decode_16(p);
 355                 info->dir_end = !!(flags & CEPH_READDIR_FRAG_END);
 356                 info->dir_complete = !!(flags & CEPH_READDIR_FRAG_COMPLETE);
 357                 info->hash_order = !!(flags & CEPH_READDIR_HASH_ORDER);
 358                 info->offset_hash = !!(flags & CEPH_READDIR_OFFSET_HASH);
 359         }
 360         if (num == 0)
 361                 goto done;
 362
 363         BUG_ON(!info->dir_entries);
 364         if ((unsigned long)(info->dir_entries + num) >
 365             (unsigned long)info->dir_entries + info->dir_buf_size) {
 366                 pr_err("dir contents are larger than expected\n");
 367                 WARN_ON(1);
 368                 goto bad;
 369         }
 370
 371         info->dir_nr = num;
 372         while (num) {
 373                 struct ceph_mds_reply_dir_entry *rde = info->dir_entries + i;
 374                 /* dentry */
 375                 ceph_decode_32_safe(p, end, rde->name_len, bad);
 376                 ceph_decode_need(p, end, rde->name_len, bad);
 377                 rde->name = *p;
 378                 *p += rde->name_len;
 379                 dout("parsed dir dname '%.*s'\n", rde->name_len, rde->name);
 380
 381                 /* dentry lease */
 382                 err = parse_reply_info_lease(p, end, &rde->lease, features);
 383                 if (err)
 384                         goto out_bad;
 385                 /* inode */
 386                 err = parse_reply_info_in(p, end, &rde->inode, features);
 387                 if (err < 0)
 388                         goto out_bad;
 389                 /* ceph_readdir_prepopulate() will update it */
 390                 rde->offset = 0;
 391                 i++;
 392                 num--;
 393         }
 394
 395 done:
 396         /* Skip over any unrecognized fields */
 397         *p = end;
 398         return 0;
 399
 400 bad:
 401         err = -EIO;
 402 out_bad:
 403         pr_err("problem parsing dir contents %d\n", err);
 404         return err;
 405 }
 406
 407 /*
 408  * parse fcntl F_GETLK results
 409  */
 410 static int parse_reply_info_filelock(void **p, void *end,
 411                                      struct ceph_mds_reply_info_parsed *info,
 412                                      u64 features)
 413 {
 414         if (*p + sizeof(*info->filelock_reply) > end)
 415                 goto bad;
 416
 417         info->filelock_reply = *p;
 418
 419         /* Skip over any unrecognized fields */
 420         *p = end;
 421         return 0;
 422 bad:
 423         return -EIO;
 424 }
 425
 426
 427 #if BITS_PER_LONG == 64
 428
 429 #define DELEGATED_INO_AVAILABLE         xa_mk_value(1)
 430
 431 static int ceph_parse_deleg_inos(void **p, void *end,
 432                                  struct ceph_mds_session *s)
 433 {
 434         u32 sets;
 435
 436         ceph_decode_32_safe(p, end, sets, bad);
 437         dout("got %u sets of delegated inodes\n", sets);
 438         while (sets--) {
 439                 u64 start, len, ino;
 440
 441                 ceph_decode_64_safe(p, end, start, bad);
 442                 ceph_decode_64_safe(p, end, len, bad);
 443
 444                 /* Don't accept a delegation of system inodes */
 445                 if (start < CEPH_INO_SYSTEM_BASE) {
 446                         pr_warn_ratelimited("ceph: ignoring reserved inode range delegation (start=0x%llx len=0x%llx)\n",
 447                                         start, len);
 448                         continue;
 449                 }
 450                 while (len--) {
 451                         int err = xa_insert(&s->s_delegated_inos, ino = start++,
 452                                             DELEGATED_INO_AVAILABLE,
 453                                             GFP_KERNEL);
 454                         if (!err) {
 455                                 dout("added delegated inode 0x%llx\n",
 456                                      start - 1);
 457                         } else if (err == -EBUSY) {
 458                                 pr_warn("ceph: MDS delegated inode 0x%llx more than once.\n",
 459                                         start - 1);
 460                         } else {
 461                                 return err;
 462                         }
 463                 }
 464         }
 465         return 0;
 466 bad:
 467         return -EIO;
 468 }
 469
 470 u64 ceph_get_deleg_ino(struct ceph_mds_session *s)
 471 {
 472         unsigned long ino;
 473         void *val;
 474
 475         xa_for_each(&s->s_delegated_inos, ino, val) {
 476                 val = xa_erase(&s->s_delegated_inos, ino);
 477                 if (val == DELEGATED_INO_AVAILABLE)
 478                         return ino;
 479         }
 480         return 0;
 481 }
 482
 483 int ceph_restore_deleg_ino(struct ceph_mds_session *s, u64 ino)
 484 {
 485         return xa_insert(&s->s_delegated_inos, ino, DELEGATED_INO_AVAILABLE,
 486                          GFP_KERNEL);
 487 }
 488 #else /* BITS_PER_LONG == 64 */
 489 /*
 490  * FIXME: xarrays can't handle 64-bit indexes on a 32-bit arch. For now, just
 491  * ignore delegated_inos on 32 bit arch. Maybe eventually add xarrays for top
 492  * and bottom words?
 493  */
 494 static int ceph_parse_deleg_inos(void **p, void *end,
 495                                  struct ceph_mds_session *s)
 496 {
 497         u32 sets;
 498
 499         ceph_decode_32_safe(p, end, sets, bad);
 500         if (sets)
 501                 ceph_decode_skip_n(p, end, sets * 2 * sizeof(__le64), bad);
 502         return 0;
 503 bad:
 504         return -EIO;
 505 }
 506
 507 u64 ceph_get_deleg_ino(struct ceph_mds_session *s)
 508 {
 509         return 0;
 510 }
 511
 512 int ceph_restore_deleg_ino(struct ceph_mds_session *s, u64 ino)
 513 {
 514         return 0;
 515 }
 516 #endif /* BITS_PER_LONG == 64 */
 517
 518 /*
 519  * parse create results
 520  */
 521 static int parse_reply_info_create(void **p, void *end,
 522                                   struct ceph_mds_reply_info_parsed *info,
 523                                   u64 features, struct ceph_mds_session *s)
 524 {
 525         int ret;
 526
 527         if (features == (u64)-1 ||
 528             (features & CEPH_FEATURE_REPLY_CREATE_INODE)) {
 529                 if (*p == end) {
 530                         /* Malformed reply? */
 531                         info->has_create_ino = false;
 532                 } else if (test_bit(CEPHFS_FEATURE_DELEG_INO, &s->s_features)) {
 533                         info->has_create_ino = true;
 534                         /* struct_v, struct_compat, and len */
 535                         ceph_decode_skip_n(p, end, 2 + sizeof(u32), bad);
 536                         ceph_decode_64_safe(p, end, info->ino, bad);
 537                         ret = ceph_parse_deleg_inos(p, end, s);
 538                         if (ret)
 539                                 return ret;
 540                 } else {
 541                         /* legacy */
 542                         ceph_decode_64_safe(p, end, info->ino, bad);
 543                         info->has_create_ino = true;
 544                 }
 545         } else {
 546                 if (*p != end)
 547                         goto bad;
 548         }
 549
 550         /* Skip over any unrecognized fields */
 551         *p = end;
 552         return 0;
 553 bad:
 554         return -EIO;
 555 }
 556
 557 /*
 558  * parse extra results
 559  */
 560 static int parse_reply_info_extra(void **p, void *end,
 561                                   struct ceph_mds_reply_info_parsed *info,
 562                                   u64 features, struct ceph_mds_session *s)
 563 {
 564         u32 op = le32_to_cpu(info->head->op);
 565
 566         if (op == CEPH_MDS_OP_GETFILELOCK)
 567                 return parse_reply_info_filelock(p, end, info, features);
 568         else if (op == CEPH_MDS_OP_READDIR || op == CEPH_MDS_OP_LSSNAP)
 569                 return parse_reply_info_readdir(p, end, info, features);
 570         else if (op == CEPH_MDS_OP_CREATE)
 571                 return parse_reply_info_create(p, end, info, features, s);
 572         else
 573                 return -EIO;
 574 }
 575
 576 /*
 577  * parse entire mds reply
 578  */
 579 static int parse_reply_info(struct ceph_mds_session *s, struct ceph_msg *msg,
 580                             struct ceph_mds_reply_info_parsed *info,
 581                             u64 features)
 582 {
 583         void *p, *end;
 584         u32 len;
 585         int err;
 586
 587         info->head = msg->front.iov_base;
 588         p = msg->front.iov_base + sizeof(struct ceph_mds_reply_head);
 589         end = p + msg->front.iov_len - sizeof(struct ceph_mds_reply_head);
 590
 591         /* trace */
 592         ceph_decode_32_safe(&p, end, len, bad);
 593         if (len > 0) {
 594                 ceph_decode_need(&p, end, len, bad);
 595                 err = parse_reply_info_trace(&p, p+len, info, features);
 596                 if (err < 0)
 597                         goto out_bad;
 598         }
 599
 600         /* extra */
 601         ceph_decode_32_safe(&p, end, len, bad);
 602         if (len > 0) {
 603                 ceph_decode_need(&p, end, len, bad);
 604                 err = parse_reply_info_extra(&p, p+len, info, features, s);
 605                 if (err < 0)
 606                         goto out_bad;
 607         }
 608
 609         /* snap blob */
 610         ceph_decode_32_safe(&p, end, len, bad);
 611         info->snapblob_len = len;
 612         info->snapblob = p;
 613         p += len;
 614
 615         if (p != end)
 616                 goto bad;
 617         return 0;
 618
 619 bad:
 620         err = -EIO;
 621 out_bad:
 622         pr_err("mds parse_reply err %d\n", err);
 623         return err;
 624 }
 625
 626 static void destroy_reply_info(struct ceph_mds_reply_info_parsed *info)
 627 {
 628         if (!info->dir_entries)
 629                 return;
 630         free_pages((unsigned long)info->dir_entries, get_order(info->dir_buf_size));
 631 }
 632
 633
 634 /*
 635  * sessions
 636  */
 637 const char *ceph_session_state_name(int s)
 638 {
 639         switch (s) {
 640         case CEPH_MDS_SESSION_NEW: return "new";
 641         case CEPH_MDS_SESSION_OPENING: return "opening";
 642         case CEPH_MDS_SESSION_OPEN: return "open";
 643         case CEPH_MDS_SESSION_HUNG: return "hung";
 644         case CEPH_MDS_SESSION_CLOSING: return "closing";
 645         case CEPH_MDS_SESSION_CLOSED: return "closed";
 646         case CEPH_MDS_SESSION_RESTARTING: return "restarting";
 647         case CEPH_MDS_SESSION_RECONNECTING: return "reconnecting";
 648         case CEPH_MDS_SESSION_REJECTED: return "rejected";
 649         default: return "???";
 650         }
 651 }
 652
 653 struct ceph_mds_session *ceph_get_mds_session(struct ceph_mds_session *s)
 654 {
 655         if (refcount_inc_not_zero(&s->s_ref)) {
 656                 dout("mdsc get_session %p %d -> %d\n", s,
 657                      refcount_read(&s->s_ref)-1, refcount_read(&s->s_ref));
 658                 return s;
 659         } else {
 660                 dout("mdsc get_session %p 0 -- FAIL\n", s);
 661                 return NULL;
 662         }
 663 }
 664
 665 void ceph_put_mds_session(struct ceph_mds_session *s)
 666 {
 667         if (IS_ERR_OR_NULL(s))
 668                 return;
 669
 670         dout("mdsc put_session %p %d -> %d\n", s,
 671              refcount_read(&s->s_ref), refcount_read(&s->s_ref)-1);
 672         if (refcount_dec_and_test(&s->s_ref)) {
 673                 if (s->s_auth.authorizer)
 674                         ceph_auth_destroy_authorizer(s->s_auth.authorizer);
 675                 WARN_ON(mutex_is_locked(&s->s_mutex));
 676                 xa_destroy(&s->s_delegated_inos);
 677                 kfree(s);
 678         }
 679 }
 680
 681 /*
 682  * called under mdsc->mutex
 683  */
 684 struct ceph_mds_session *__ceph_lookup_mds_session(struct ceph_mds_client *mdsc,
 685                                                    int mds)
 686 {
 687         if (mds >= mdsc->max_sessions || !mdsc->sessions[mds])
 688                 return NULL;
 689         return ceph_get_mds_session(mdsc->sessions[mds]);
 690 }
 691
 692 static bool __have_session(struct ceph_mds_client *mdsc, int mds)
 693 {
 694         if (mds >= mdsc->max_sessions || !mdsc->sessions[mds])
 695                 return false;
 696         else
 697                 return true;
 698 }
 699
 700 static int __verify_registered_session(struct ceph_mds_client *mdsc,
 701                                        struct ceph_mds_session *s)
 702 {
 703         if (s->s_mds >= mdsc->max_sessions ||
 704             mdsc->sessions[s->s_mds] != s)
 705                 return -ENOENT;
 706         return 0;
 707 }
 708
 709 /*
 710  * create+register a new session for given mds.
 711  * called under mdsc->mutex.
 712  */
 713 static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc,
 714                                                  int mds)
 715 {
 716         struct ceph_mds_session *s;
 717
 718         if (mds >= mdsc->mdsmap->possible_max_rank)
 719                 return ERR_PTR(-EINVAL);
 720
 721         s = kzalloc(sizeof(*s), GFP_NOFS);
 722         if (!s)
 723                 return ERR_PTR(-ENOMEM);
 724
 725         if (mds >= mdsc->max_sessions) {
 726                 int newmax = 1 << get_count_order(mds + 1);
 727                 struct ceph_mds_session **sa;
 728
 729                 dout("%s: realloc to %d\n", __func__, newmax);
 730                 sa = kcalloc(newmax, sizeof(void *), GFP_NOFS);
 731                 if (!sa)
 732                         goto fail_realloc;
 733                 if (mdsc->sessions) {
 734                         memcpy(sa, mdsc->sessions,
 735                                mdsc->max_sessions * sizeof(void *));
 736                         kfree(mdsc->sessions);
 737                 }
 738                 mdsc->sessions = sa;
 739                 mdsc->max_sessions = newmax;
 740         }
 741
 742         dout("%s: mds%d\n", __func__, mds);
 743         s->s_mdsc = mdsc;
 744         s->s_mds = mds;
 745         s->s_state = CEPH_MDS_SESSION_NEW;
 746         s->s_ttl = 0;
 747         s->s_seq = 0;
 748         mutex_init(&s->s_mutex);
 749
 750         ceph_con_init(&s->s_con, s, &mds_con_ops, &mdsc->fsc->client->msgr);
 751
 752         atomic_set(&s->s_cap_gen, 1);
 753         s->s_cap_ttl = jiffies - 1;
 754
 755         spin_lock_init(&s->s_cap_lock);
 756         s->s_renew_requested = 0;
 757         s->s_renew_seq = 0;
 758         INIT_LIST_HEAD(&s->s_caps);
 759         s->s_nr_caps = 0;
 760         refcount_set(&s->s_ref, 1);
 761         INIT_LIST_HEAD(&s->s_waiting);
 762         INIT_LIST_HEAD(&s->s_unsafe);
 763         xa_init(&s->s_delegated_inos);
 764         s->s_num_cap_releases = 0;
 765         s->s_cap_reconnect = 0;
 766         s->s_cap_iterator = NULL;
 767         INIT_LIST_HEAD(&s->s_cap_releases);
 768         INIT_WORK(&s->s_cap_release_work, ceph_cap_release_work);
 769
 770         INIT_LIST_HEAD(&s->s_cap_dirty);
 771         INIT_LIST_HEAD(&s->s_cap_flushing);
 772
 773         mdsc->sessions[mds] = s;
 774         atomic_inc(&mdsc->num_sessions);
 775         refcount_inc(&s->s_ref);  /* one ref to sessions[], one to caller */
 776
 777         ceph_con_open(&s->s_con, CEPH_ENTITY_TYPE_MDS, mds,
 778                       ceph_mdsmap_get_addr(mdsc->mdsmap, mds));
 779
 780         return s;
 781
 782 fail_realloc:
 783         kfree(s);
 784         return ERR_PTR(-ENOMEM);
 785 }
 786
 787 /*
 788  * called under mdsc->mutex
 789  */
 790 static void __unregister_session(struct ceph_mds_client *mdsc,
 791                                struct ceph_mds_session *s)
 792 {
 793         dout("__unregister_session mds%d %p\n", s->s_mds, s);
 794         BUG_ON(mdsc->sessions[s->s_mds] != s);
 795         mdsc->sessions[s->s_mds] = NULL;
 796         ceph_con_close(&s->s_con);
 797         ceph_put_mds_session(s);
 798         atomic_dec(&mdsc->num_sessions);
 799 }
 800
 801 /*
 802  * drop session refs in request.
 803  *
 804  * should be last request ref, or hold mdsc->mutex
 805  */
 806 static void put_request_session(struct ceph_mds_request *req)
 807 {
 808         if (req->r_session) {
 809                 ceph_put_mds_session(req->r_session);
 810                 req->r_session = NULL;
 811         }
 812 }
 813
 814 void ceph_mdsc_release_request(struct kref *kref)
 815 {
 816         struct ceph_mds_request *req = container_of(kref,
 817                                                     struct ceph_mds_request,
 818                                                     r_kref);
 819         ceph_mdsc_release_dir_caps_no_check(req);
 820         destroy_reply_info(&req->r_reply_info);
 821         if (req->r_request)
 822                 ceph_msg_put(req->r_request);
 823         if (req->r_reply)
 824                 ceph_msg_put(req->r_reply);
 825         if (req->r_inode) {
 826                 ceph_put_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
 827                 iput(req->r_inode);
 828         }
 829         if (req->r_parent) {
 830                 ceph_put_cap_refs(ceph_inode(req->r_parent), CEPH_CAP_PIN);
 831                 iput(req->r_parent);
 832         }
 833         iput(req->r_target_inode);
 834         if (req->r_dentry)
 835                 dput(req->r_dentry);
 836         if (req->r_old_dentry)
 837                 dput(req->r_old_dentry);
 838         if (req->r_old_dentry_dir) {
 839                 /*
 840                  * track (and drop pins for) r_old_dentry_dir
 841                  * separately, since r_old_dentry's d_parent may have
 842                  * changed between the dir mutex being dropped and
 843                  * this request being freed.
 844                  */
 845                 ceph_put_cap_refs(ceph_inode(req->r_old_dentry_dir),
 846                                   CEPH_CAP_PIN);
 847                 iput(req->r_old_dentry_dir);
 848         }
 849         kfree(req->r_path1);
 850         kfree(req->r_path2);
 851         put_cred(req->r_cred);
 852         if (req->r_pagelist)
 853                 ceph_pagelist_release(req->r_pagelist);
 854         put_request_session(req);
 855         ceph_unreserve_caps(req->r_mdsc, &req->r_caps_reservation);
 856         WARN_ON_ONCE(!list_empty(&req->r_wait));
 857         kmem_cache_free(ceph_mds_request_cachep, req);
 858 }
 859
 860 DEFINE_RB_FUNCS(request, struct ceph_mds_request, r_tid, r_node)
 861
 862 /*
 863  * lookup session, bump ref if found.
 864  *
 865  * called under mdsc->mutex.
 866  */
 867 static struct ceph_mds_request *
 868 lookup_get_request(struct ceph_mds_client *mdsc, u64 tid)
 869 {
 870         struct ceph_mds_request *req;
 871
 872         req = lookup_request(&mdsc->request_tree, tid);
 873         if (req)
 874                 ceph_mdsc_get_request(req);
 875
 876         return req;
 877 }
 878
 879 /*
 880  * Register an in-flight request, and assign a tid.  Link to directory
 881  * are modifying (if any).
 882  *
 883  * Called under mdsc->mutex.
 884  */
 885 static void __register_request(struct ceph_mds_client *mdsc,
 886                                struct ceph_mds_request *req,
 887                                struct inode *dir)
 888 {
 889         int ret = 0;
 890
 891         req->r_tid = ++mdsc->last_tid;
 892         if (req->r_num_caps) {
 893                 ret = ceph_reserve_caps(mdsc, &req->r_caps_reservation,
 894                                         req->r_num_caps);
 895                 if (ret < 0) {
 896                         pr_err("__register_request %p "
 897                                "failed to reserve caps: %d\n", req, ret);
 898                         /* set req->r_err to fail early from __do_request */
 899                         req->r_err = ret;
 900                         return;
 901                 }
 902         }
 903         dout("__register_request %p tid %lld\n", req, req->r_tid);
 904         ceph_mdsc_get_request(req);
 905         insert_request(&mdsc->request_tree, req);
 906
 907         req->r_cred = get_current_cred();
 908
 909         if (mdsc->oldest_tid == 0 && req->r_op != CEPH_MDS_OP_SETFILELOCK)
 910                 mdsc->oldest_tid = req->r_tid;
 911
 912         if (dir) {
 913                 struct ceph_inode_info *ci = ceph_inode(dir);
 914
 915                 ihold(dir);
 916                 req->r_unsafe_dir = dir;
 917                 spin_lock(&ci->i_unsafe_lock);
 918                 list_add_tail(&req->r_unsafe_dir_item, &ci->i_unsafe_dirops);
 919                 spin_unlock(&ci->i_unsafe_lock);
 920         }
 921 }
 922
 923 static void __unregister_request(struct ceph_mds_client *mdsc,
 924                                  struct ceph_mds_request *req)
 925 {
 926         dout("__unregister_request %p tid %lld\n", req, req->r_tid);
 927
 928         /* Never leave an unregistered request on an unsafe list! */
 929         list_del_init(&req->r_unsafe_item);
 930
 931         if (req->r_tid == mdsc->oldest_tid) {
 932                 struct rb_node *p = rb_next(&req->r_node);
 933                 mdsc->oldest_tid = 0;
 934                 while (p) {
 935                         struct ceph_mds_request *next_req =
 936                                 rb_entry(p, struct ceph_mds_request, r_node);
 937                         if (next_req->r_op != CEPH_MDS_OP_SETFILELOCK) {
 938                                 mdsc->oldest_tid = next_req->r_tid;
 939                                 break;
 940                         }
 941                         p = rb_next(p);
 942                 }
 943         }
 944
 945         erase_request(&mdsc->request_tree, req);
 946
 947         if (req->r_unsafe_dir) {
 948                 struct ceph_inode_info *ci = ceph_inode(req->r_unsafe_dir);
 949                 spin_lock(&ci->i_unsafe_lock);
 950                 list_del_init(&req->r_unsafe_dir_item);
 951                 spin_unlock(&ci->i_unsafe_lock);
 952         }
 953         if (req->r_target_inode &&
 954             test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
 955                 struct ceph_inode_info *ci = ceph_inode(req->r_target_inode);
 956                 spin_lock(&ci->i_unsafe_lock);
 957                 list_del_init(&req->r_unsafe_target_item);
 958                 spin_unlock(&ci->i_unsafe_lock);
 959         }
 960
 961         if (req->r_unsafe_dir) {
 962                 iput(req->r_unsafe_dir);
 963                 req->r_unsafe_dir = NULL;
 964         }
 965
 966         complete_all(&req->r_safe_completion);
 967
 968         ceph_mdsc_put_request(req);
 969 }
 970
 971 /*
 972  * Walk back up the dentry tree until we hit a dentry representing a
 973  * non-snapshot inode. We do this using the rcu_read_lock (which must be held
 974  * when calling this) to ensure that the objects won't disappear while we're
 975  * working with them. Once we hit a candidate dentry, we attempt to take a
 976  * reference to it, and return that as the result.
 977  */
 978 static struct inode *get_nonsnap_parent(struct dentry *dentry)
 979 {
 980         struct inode *inode = NULL;
 981
 982         while (dentry && !IS_ROOT(dentry)) {
 983                 inode = d_inode_rcu(dentry);
 984                 if (!inode || ceph_snap(inode) == CEPH_NOSNAP)
 985                         break;
 986                 dentry = dentry->d_parent;
 987         }
 988         if (inode)
 989                 inode = igrab(inode);
 990         return inode;
 991 }
 992
 993 /*
 994  * Choose mds to send request to next.  If there is a hint set in the
 995  * request (e.g., due to a prior forward hint from the mds), use that.
 996  * Otherwise, consult frag tree and/or caps to identify the
 997  * appropriate mds.  If all else fails, choose randomly.
 998  *
 999  * Called under mdsc->mutex.
1000  */
1001 static int __choose_mds(struct ceph_mds_client *mdsc,
1002                         struct ceph_mds_request *req,
1003                         bool *random)
1004 {
1005         struct inode *inode;
1006         struct ceph_inode_info *ci;
1007         struct ceph_cap *cap;
1008         int mode = req->r_direct_mode;
1009         int mds = -1;
1010         u32 hash = req->r_direct_hash;
1011         bool is_hash = test_bit(CEPH_MDS_R_DIRECT_IS_HASH, &req->r_req_flags);
1012
1013         if (random)
1014                 *random = false;
1015
1016         /*
1017          * is there a specific mds we should try?  ignore hint if we have
1018          * no session and the mds is not up (active or recovering).
1019          */
1020         if (req->r_resend_mds >= 0 &&
1021             (__have_session(mdsc, req->r_resend_mds) ||
1022              ceph_mdsmap_get_state(mdsc->mdsmap, req->r_resend_mds) > 0)) {
1023                 dout("%s using resend_mds mds%d\n", __func__,
1024                      req->r_resend_mds);
1025                 return req->r_resend_mds;
1026         }
1027
1028         if (mode == USE_RANDOM_MDS)
1029                 goto random;
1030
1031         inode = NULL;
1032         if (req->r_inode) {
1033                 if (ceph_snap(req->r_inode) != CEPH_SNAPDIR) {
1034                         inode = req->r_inode;
1035                         ihold(inode);
1036                 } else {
1037                         /* req->r_dentry is non-null for LSSNAP request */
1038                         rcu_read_lock();
1039                         inode = get_nonsnap_parent(req->r_dentry);
1040                         rcu_read_unlock();
1041                         dout("%s using snapdir's parent %p\n", __func__, inode);
1042                 }
1043         } else if (req->r_dentry) {
1044                 /* ignore race with rename; old or new d_parent is okay */
1045                 struct dentry *parent;
1046                 struct inode *dir;
1047
1048                 rcu_read_lock();
1049                 parent = READ_ONCE(req->r_dentry->d_parent);
1050                 dir = req->r_parent ? : d_inode_rcu(parent);
1051
1052                 if (!dir || dir->i_sb != mdsc->fsc->sb) {
1053                         /*  not this fs or parent went negative */
1054                         inode = d_inode(req->r_dentry);
1055                         if (inode)
1056                                 ihold(inode);
1057                 } else if (ceph_snap(dir) != CEPH_NOSNAP) {
1058                         /* direct snapped/virtual snapdir requests
1059                          * based on parent dir inode */
1060                         inode = get_nonsnap_parent(parent);
1061                         dout("%s using nonsnap parent %p\n", __func__, inode);
1062                 } else {
1063                         /* dentry target */
1064                         inode = d_inode(req->r_dentry);
1065                         if (!inode || mode == USE_AUTH_MDS) {
1066                                 /* dir + name */
1067                                 inode = igrab(dir);
1068                                 hash = ceph_dentry_hash(dir, req->r_dentry);
1069                                 is_hash = true;
1070                         } else {
1071                                 ihold(inode);
1072                         }
1073                 }
1074                 rcu_read_unlock();
1075         }
1076
1077         dout("%s %p is_hash=%d (0x%x) mode %d\n", __func__, inode, (int)is_hash,
1078              hash, mode);
1079         if (!inode)
1080                 goto random;
1081         ci = ceph_inode(inode);
1082
1083         if (is_hash && S_ISDIR(inode->i_mode)) {
1084                 struct ceph_inode_frag frag;
1085                 int found;
1086
1087                 ceph_choose_frag(ci, hash, &frag, &found);
1088                 if (found) {
1089                         if (mode == USE_ANY_MDS && frag.ndist > 0) {
1090                                 u8 r;
1091
1092                                 /* choose a random replica */
1093                                 get_random_bytes(&r, 1);
1094                                 r %= frag.ndist;
1095                                 mds = frag.dist[r];
1096                                 dout("%s %p %llx.%llx frag %u mds%d (%d/%d)\n",
1097                                      __func__, inode, ceph_vinop(inode),
1098                                      frag.frag, mds, (int)r, frag.ndist);
1099                                 if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
1100                                     CEPH_MDS_STATE_ACTIVE &&
1101                                     !ceph_mdsmap_is_laggy(mdsc->mdsmap, mds))
1102                                         goto out;
1103                         }
1104
1105                         /* since this file/dir wasn't known to be
1106                          * replicated, then we want to look for the
1107                          * authoritative mds. */
1108                         if (frag.mds >= 0) {
1109                                 /* choose auth mds */
1110                                 mds = frag.mds;
1111                                 dout("%s %p %llx.%llx frag %u mds%d (auth)\n",
1112                                      __func__, inode, ceph_vinop(inode),
1113                                      frag.frag, mds);
1114                                 if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
1115                                     CEPH_MDS_STATE_ACTIVE) {
1116                                         if (!ceph_mdsmap_is_laggy(mdsc->mdsmap,
1117                                                                   mds))
1118                                                 goto out;
1119                                 }
1120                         }
1121                         mode = USE_AUTH_MDS;
1122                 }
1123         }
1124
1125         spin_lock(&ci->i_ceph_lock);
1126         cap = NULL;
1127         if (mode == USE_AUTH_MDS)
1128                 cap = ci->i_auth_cap;
1129         if (!cap && !RB_EMPTY_ROOT(&ci->i_caps))
1130                 cap = rb_entry(rb_first(&ci->i_caps), struct ceph_cap, ci_node);
1131         if (!cap) {
1132                 spin_unlock(&ci->i_ceph_lock);
1133                 iput(inode);
1134                 goto random;
1135         }
1136         mds = cap->session->s_mds;
1137         dout("%s %p %llx.%llx mds%d (%scap %p)\n", __func__,
1138              inode, ceph_vinop(inode), mds,
1139              cap == ci->i_auth_cap ? "auth " : "", cap);
1140         spin_unlock(&ci->i_ceph_lock);
1141 out:
1142         iput(inode);
1143         return mds;
1144
1145 random:
1146         if (random)
1147                 *random = true;
1148
1149         mds = ceph_mdsmap_get_random_mds(mdsc->mdsmap);
1150         dout("%s chose random mds%d\n", __func__, mds);
1151         return mds;
1152 }
1153
1154
1155 /*
1156  * session messages
1157  */
1158 static struct ceph_msg *create_session_msg(u32 op, u64 seq)
1159 {
1160         struct ceph_msg *msg;
1161         struct ceph_mds_session_head *h;
1162
1163         msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h), GFP_NOFS,
1164                            false);
1165         if (!msg) {
1166                 pr_err("create_session_msg ENOMEM creating msg\n");
1167                 return NULL;
1168         }
1169         h = msg->front.iov_base;
1170         h->op = cpu_to_le32(op);
1171         h->seq = cpu_to_le64(seq);
1172
1173         return msg;
1174 }
1175
1176 static const unsigned char feature_bits[] = CEPHFS_FEATURES_CLIENT_SUPPORTED;
1177 #define FEATURE_BYTES(c) (DIV_ROUND_UP((size_t)feature_bits[c - 1] + 1, 64) * 8)
1178 static int encode_supported_features(void **p, void *end)
1179 {
1180         static const size_t count = ARRAY_SIZE(feature_bits);
1181
1182         if (count > 0) {
1183                 size_t i;
1184                 size_t size = FEATURE_BYTES(count);
1185
1186                 if (WARN_ON_ONCE(*p + 4 + size > end))
1187                         return -ERANGE;
1188
1189                 ceph_encode_32(p, size);
1190                 memset(*p, 0, size);
1191                 for (i = 0; i < count; i++)
1192                         ((unsigned char*)(*p))[i / 8] |= BIT(feature_bits[i] % 8);
1193                 *p += size;
1194         } else {
1195                 if (WARN_ON_ONCE(*p + 4 > end))
1196                         return -ERANGE;
1197
1198                 ceph_encode_32(p, 0);
1199         }
1200
1201         return 0;
1202 }
1203
1204 static const unsigned char metric_bits[] = CEPHFS_METRIC_SPEC_CLIENT_SUPPORTED;
1205 #define METRIC_BYTES(cnt) (DIV_ROUND_UP((size_t)metric_bits[cnt - 1] + 1, 64) * 8)
1206 static int encode_metric_spec(void **p, void *end)
1207 {
1208         static const size_t count = ARRAY_SIZE(metric_bits);
1209
1210         /* header */
1211         if (WARN_ON_ONCE(*p + 2 > end))
1212                 return -ERANGE;
1213
1214         ceph_encode_8(p, 1); /* version */
1215         ceph_encode_8(p, 1); /* compat */
1216
1217         if (count > 0) {
1218                 size_t i;
1219                 size_t size = METRIC_BYTES(count);
1220
1221                 if (WARN_ON_ONCE(*p + 4 + 4 + size > end))
1222                         return -ERANGE;
1223
1224                 /* metric spec info length */
1225                 ceph_encode_32(p, 4 + size);
1226
1227                 /* metric spec */
1228                 ceph_encode_32(p, size);
1229                 memset(*p, 0, size);
1230                 for (i = 0; i < count; i++)
1231                         ((unsigned char *)(*p))[i / 8] |= BIT(metric_bits[i] % 8);
1232                 *p += size;
1233         } else {
1234                 if (WARN_ON_ONCE(*p + 4 + 4 > end))
1235                         return -ERANGE;
1236
1237                 /* metric spec info length */
1238                 ceph_encode_32(p, 4);
1239                 /* metric spec */
1240                 ceph_encode_32(p, 0);
1241         }
1242
1243         return 0;
1244 }
1245
1246 /*
1247  * session message, specialization for CEPH_SESSION_REQUEST_OPEN
1248  * to include additional client metadata fields.
1249  */
1250 static struct ceph_msg *create_session_open_msg(struct ceph_mds_client *mdsc, u64 seq)
1251 {
1252         struct ceph_msg *msg;
1253         struct ceph_mds_session_head *h;
1254         int i;
1255         int extra_bytes = 0;
1256         int metadata_key_count = 0;
1257         struct ceph_options *opt = mdsc->fsc->client->options;
1258         struct ceph_mount_options *fsopt = mdsc->fsc->mount_options;
1259         size_t size, count;
1260         void *p, *end;
1261         int ret;
1262
1263         const char* metadata[][2] = {
1264                 {"hostname", mdsc->nodename},
1265                 {"kernel_version", init_utsname()->release},
1266                 {"entity_id", opt->name ? : ""},
1267                 {"root", fsopt->server_path ? : "/"},
1268                 {NULL, NULL}
1269         };
1270
1271         /* Calculate serialized length of metadata */
1272         extra_bytes = 4;  /* map length */
1273         for (i = 0; metadata[i][0]; ++i) {
1274                 extra_bytes += 8 + strlen(metadata[i][0]) +
1275                         strlen(metadata[i][1]);
1276                 metadata_key_count++;
1277         }
1278
1279         /* supported feature */
1280         size = 0;
1281         count = ARRAY_SIZE(feature_bits);
1282         if (count > 0)
1283                 size = FEATURE_BYTES(count);
1284         extra_bytes += 4 + size;
1285
1286         /* metric spec */
1287         size = 0;
1288         count = ARRAY_SIZE(metric_bits);
1289         if (count > 0)
1290                 size = METRIC_BYTES(count);
1291         extra_bytes += 2 + 4 + 4 + size;
1292
1293         /* Allocate the message */
1294         msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h) + extra_bytes,
1295                            GFP_NOFS, false);
1296         if (!msg) {
1297                 pr_err("create_session_msg ENOMEM creating msg\n");
1298                 return ERR_PTR(-ENOMEM);
1299         }
1300         p = msg->front.iov_base;
1301         end = p + msg->front.iov_len;
1302
1303         h = p;
1304         h->op = cpu_to_le32(CEPH_SESSION_REQUEST_OPEN);
1305         h->seq = cpu_to_le64(seq);
1306
1307         /*
1308          * Serialize client metadata into waiting buffer space, using
1309          * the format that userspace expects for map<string, string>
1310          *
1311          * ClientSession messages with metadata are v4
1312          */
1313         msg->hdr.version = cpu_to_le16(4);
1314         msg->hdr.compat_version = cpu_to_le16(1);
1315
1316         /* The write pointer, following the session_head structure */
1317         p += sizeof(*h);
1318
1319         /* Number of entries in the map */
1320         ceph_encode_32(&p, metadata_key_count);
1321
1322         /* Two length-prefixed strings for each entry in the map */
1323         for (i = 0; metadata[i][0]; ++i) {
1324                 size_t const key_len = strlen(metadata[i][0]);
1325                 size_t const val_len = strlen(metadata[i][1]);
1326
1327                 ceph_encode_32(&p, key_len);
1328                 memcpy(p, metadata[i][0], key_len);
1329                 p += key_len;
1330                 ceph_encode_32(&p, val_len);
1331                 memcpy(p, metadata[i][1], val_len);
1332                 p += val_len;
1333         }
1334
1335         ret = encode_supported_features(&p, end);
1336         if (ret) {
1337                 pr_err("encode_supported_features failed!\n");
1338                 ceph_msg_put(msg);
1339                 return ERR_PTR(ret);
1340         }
1341
1342         ret = encode_metric_spec(&p, end);
1343         if (ret) {
1344                 pr_err("encode_metric_spec failed!\n");
1345                 ceph_msg_put(msg);
1346                 return ERR_PTR(ret);
1347         }
1348
1349         msg->front.iov_len = p - msg->front.iov_base;
1350         msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
1351
1352         return msg;
1353 }
1354
1355 /*
1356  * send session open request.
1357  *
1358  * called under mdsc->mutex
1359  */
1360 static int __open_session(struct ceph_mds_client *mdsc,
1361                           struct ceph_mds_session *session)
1362 {
1363         struct ceph_msg *msg;
1364         int mstate;
1365         int mds = session->s_mds;
1366
1367         /* wait for mds to go active? */
1368         mstate = ceph_mdsmap_get_state(mdsc->mdsmap, mds);
1369         dout("open_session to mds%d (%s)\n", mds,
1370              ceph_mds_state_name(mstate));
1371         session->s_state = CEPH_MDS_SESSION_OPENING;
1372         session->s_renew_requested = jiffies;
1373
1374         /* send connect message */
1375         msg = create_session_open_msg(mdsc, session->s_seq);
1376         if (IS_ERR(msg))
1377                 return PTR_ERR(msg);
1378         ceph_con_send(&session->s_con, msg);
1379         return 0;
1380 }
1381
1382 /*
1383  * open sessions for any export targets for the given mds
1384  *
1385  * called under mdsc->mutex
1386  */
1387 static struct ceph_mds_session *
1388 __open_export_target_session(struct ceph_mds_client *mdsc, int target)
1389 {
1390         struct ceph_mds_session *session;
1391         int ret;
1392
1393         session = __ceph_lookup_mds_session(mdsc, target);
1394         if (!session) {
1395                 session = register_session(mdsc, target);
1396                 if (IS_ERR(session))
1397                         return session;
1398         }
1399         if (session->s_state == CEPH_MDS_SESSION_NEW ||
1400             session->s_state == CEPH_MDS_SESSION_CLOSING) {
1401                 ret = __open_session(mdsc, session);
1402                 if (ret)
1403                         return ERR_PTR(ret);
1404         }
1405
1406         return session;
1407 }
1408
1409 struct ceph_mds_session *
1410 ceph_mdsc_open_export_target_session(struct ceph_mds_client *mdsc, int target)
1411 {
1412         struct ceph_mds_session *session;
1413
1414         dout("open_export_target_session to mds%d\n", target);
1415
1416         mutex_lock(&mdsc->mutex);
1417         session = __open_export_target_session(mdsc, target);
1418         mutex_unlock(&mdsc->mutex);
1419
1420         return session;
1421 }
1422
1423 static void __open_export_target_sessions(struct ceph_mds_client *mdsc,
1424                                           struct ceph_mds_session *session)
1425 {
1426         struct ceph_mds_info *mi;
1427         struct ceph_mds_session *ts;
1428         int i, mds = session->s_mds;
1429
1430         if (mds >= mdsc->mdsmap->possible_max_rank)
1431                 return;
1432
1433         mi = &mdsc->mdsmap->m_info[mds];
1434         dout("open_export_target_sessions for mds%d (%d targets)\n",
1435              session->s_mds, mi->num_export_targets);
1436
1437         for (i = 0; i < mi->num_export_targets; i++) {
1438                 ts = __open_export_target_session(mdsc, mi->export_targets[i]);
1439                 ceph_put_mds_session(ts);
1440         }
1441 }
1442
1443 void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc,
1444                                            struct ceph_mds_session *session)
1445 {
1446         mutex_lock(&mdsc->mutex);
1447         __open_export_target_sessions(mdsc, session);
1448         mutex_unlock(&mdsc->mutex);
1449 }
1450
1451 /*
1452  * session caps
1453  */
1454
1455 static void detach_cap_releases(struct ceph_mds_session *session,
1456                                 struct list_head *target)
1457 {
1458         lockdep_assert_held(&session->s_cap_lock);
1459
1460         list_splice_init(&session->s_cap_releases, target);
1461         session->s_num_cap_releases = 0;
1462         dout("dispose_cap_releases mds%d\n", session->s_mds);
1463 }
1464
1465 static void dispose_cap_releases(struct ceph_mds_client *mdsc,
1466                                  struct list_head *dispose)
1467 {
1468         while (!list_empty(dispose)) {
1469                 struct ceph_cap *cap;
1470                 /* zero out the in-progress message */
1471                 cap = list_first_entry(dispose, struct ceph_cap, session_caps);
1472                 list_del(&cap->session_caps);
1473                 ceph_put_cap(mdsc, cap);
1474         }
1475 }
1476
1477 static void cleanup_session_requests(struct ceph_mds_client *mdsc,
1478                                      struct ceph_mds_session *session)
1479 {
1480         struct ceph_mds_request *req;
1481         struct rb_node *p;
1482         struct ceph_inode_info *ci;
1483
1484         dout("cleanup_session_requests mds%d\n", session->s_mds);
1485         mutex_lock(&mdsc->mutex);
1486         while (!list_empty(&session->s_unsafe)) {
1487                 req = list_first_entry(&session->s_unsafe,
1488                                        struct ceph_mds_request, r_unsafe_item);
1489                 pr_warn_ratelimited(" dropping unsafe request %llu\n",
1490                                     req->r_tid);
1491                 if (req->r_target_inode) {
1492                         /* dropping unsafe change of inode's attributes */
1493                         ci = ceph_inode(req->r_target_inode);
1494                         errseq_set(&ci->i_meta_err, -EIO);
1495                 }
1496                 if (req->r_unsafe_dir) {
1497                         /* dropping unsafe directory operation */
1498                         ci = ceph_inode(req->r_unsafe_dir);
1499                         errseq_set(&ci->i_meta_err, -EIO);
1500                 }
1501                 __unregister_request(mdsc, req);
1502         }
1503         /* zero r_attempts, so kick_requests() will re-send requests */
1504         p = rb_first(&mdsc->request_tree);
1505         while (p) {
1506                 req = rb_entry(p, struct ceph_mds_request, r_node);
1507                 p = rb_next(p);
1508                 if (req->r_session &&
1509                     req->r_session->s_mds == session->s_mds)
1510                         req->r_attempts = 0;
1511         }
1512         mutex_unlock(&mdsc->mutex);
1513 }
1514
1515 /*
1516  * Helper to safely iterate over all caps associated with a session, with
1517  * special care taken to handle a racing __ceph_remove_cap().
1518  *
1519  * Caller must hold session s_mutex.
1520  */
1521 int ceph_iterate_session_caps(struct ceph_mds_session *session,
1522                               int (*cb)(struct inode *, struct ceph_cap *,
1523                                         void *), void *arg)
1524 {
1525         struct list_head *p;
1526         struct ceph_cap *cap;
1527         struct inode *inode, *last_inode = NULL;
1528         struct ceph_cap *old_cap = NULL;
1529         int ret;
1530
1531         dout("iterate_session_caps %p mds%d\n", session, session->s_mds);
1532         spin_lock(&session->s_cap_lock);
1533         p = session->s_caps.next;
1534         while (p != &session->s_caps) {
1535                 cap = list_entry(p, struct ceph_cap, session_caps);
1536                 inode = igrab(&cap->ci->vfs_inode);
1537                 if (!inode) {
1538                         p = p->next;
1539                         continue;
1540                 }
1541                 session->s_cap_iterator = cap;
1542                 spin_unlock(&session->s_cap_lock);
1543
1544                 if (last_inode) {
1545                         iput(last_inode);
1546                         last_inode = NULL;
1547                 }
1548                 if (old_cap) {
1549                         ceph_put_cap(session->s_mdsc, old_cap);
1550                         old_cap = NULL;
1551                 }
1552
1553                 ret = cb(inode, cap, arg);
1554                 last_inode = inode;
1555
1556                 spin_lock(&session->s_cap_lock);
1557                 p = p->next;
1558                 if (!cap->ci) {
1559                         dout("iterate_session_caps  finishing cap %p removal\n",
1560                              cap);
1561                         BUG_ON(cap->session != session);
1562                         cap->session = NULL;
1563                         list_del_init(&cap->session_caps);
1564                         session->s_nr_caps--;
1565                         atomic64_dec(&session->s_mdsc->metric.total_caps);
1566                         if (cap->queue_release)
1567                                 __ceph_queue_cap_release(session, cap);
1568                         else
1569                                 old_cap = cap;  /* put_cap it w/o locks held */
1570                 }
1571                 if (ret < 0)
1572                         goto out;
1573         }
1574         ret = 0;
1575 out:
1576         session->s_cap_iterator = NULL;
1577         spin_unlock(&session->s_cap_lock);
1578
1579         iput(last_inode);
1580         if (old_cap)
1581                 ceph_put_cap(session->s_mdsc, old_cap);
1582
1583         return ret;
1584 }
1585
1586 static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap,
1587                                   void *arg)
1588 {
1589         struct ceph_fs_client *fsc = (struct ceph_fs_client *)arg;
1590         struct ceph_inode_info *ci = ceph_inode(inode);
1591         LIST_HEAD(to_remove);
1592         bool dirty_dropped = false;
1593         bool invalidate = false;
1594
1595         dout("removing cap %p, ci is %p, inode is %p\n",
1596              cap, ci, &ci->vfs_inode);
1597         spin_lock(&ci->i_ceph_lock);
1598         __ceph_remove_cap(cap, false);
1599         if (!ci->i_auth_cap) {
1600                 struct ceph_cap_flush *cf;
1601                 struct ceph_mds_client *mdsc = fsc->mdsc;
1602
1603                 if (READ_ONCE(fsc->mount_state) >= CEPH_MOUNT_SHUTDOWN) {
1604                         if (inode->i_data.nrpages > 0)
1605                                 invalidate = true;
1606                         if (ci->i_wrbuffer_ref > 0)
1607                                 mapping_set_error(&inode->i_data, -EIO);
1608                 }
1609
1610                 while (!list_empty(&ci->i_cap_flush_list)) {
1611                         cf = list_first_entry(&ci->i_cap_flush_list,
1612                                               struct ceph_cap_flush, i_list);
1613                         list_move(&cf->i_list, &to_remove);
1614                 }
1615
1616                 spin_lock(&mdsc->cap_dirty_lock);
1617
1618                 list_for_each_entry(cf, &to_remove, i_list)
1619                         list_del_init(&cf->g_list);
1620
1621                 if (!list_empty(&ci->i_dirty_item)) {
1622                         pr_warn_ratelimited(
1623                                 " dropping dirty %s state for %p %lld\n",
1624                                 ceph_cap_string(ci->i_dirty_caps),
1625                                 inode, ceph_ino(inode));
1626                         ci->i_dirty_caps = 0;
1627                         list_del_init(&ci->i_dirty_item);
1628                         dirty_dropped = true;
1629                 }
1630                 if (!list_empty(&ci->i_flushing_item)) {
1631                         pr_warn_ratelimited(
1632                                 " dropping dirty+flushing %s state for %p %lld\n",
1633                                 ceph_cap_string(ci->i_flushing_caps),
1634                                 inode, ceph_ino(inode));
1635                         ci->i_flushing_caps = 0;
1636                         list_del_init(&ci->i_flushing_item);
1637                         mdsc->num_cap_flushing--;
1638                         dirty_dropped = true;
1639                 }
1640                 spin_unlock(&mdsc->cap_dirty_lock);
1641
1642                 if (dirty_dropped) {
1643                         errseq_set(&ci->i_meta_err, -EIO);
1644
1645                         if (ci->i_wrbuffer_ref_head == 0 &&
1646                             ci->i_wr_ref == 0 &&
1647                             ci->i_dirty_caps == 0 &&
1648                             ci->i_flushing_caps == 0) {
1649                                 ceph_put_snap_context(ci->i_head_snapc);
1650                                 ci->i_head_snapc = NULL;
1651                         }
1652                 }
1653
1654                 if (atomic_read(&ci->i_filelock_ref) > 0) {
1655                         /* make further file lock syscall return -EIO */
1656                         ci->i_ceph_flags |= CEPH_I_ERROR_FILELOCK;
1657                         pr_warn_ratelimited(" dropping file locks for %p %lld\n",
1658                                             inode, ceph_ino(inode));
1659                 }
1660
1661                 if (!ci->i_dirty_caps && ci->i_prealloc_cap_flush) {
1662                         list_add(&ci->i_prealloc_cap_flush->i_list, &to_remove);
1663                         ci->i_prealloc_cap_flush = NULL;
1664                 }
1665         }
1666         spin_unlock(&ci->i_ceph_lock);
1667         while (!list_empty(&to_remove)) {
1668                 struct ceph_cap_flush *cf;
1669                 cf = list_first_entry(&to_remove,
1670                                       struct ceph_cap_flush, i_list);
1671                 list_del_init(&cf->i_list);
1672                 if (!cf->is_capsnap)
1673                         ceph_free_cap_flush(cf);
1674         }
1675
1676         wake_up_all(&ci->i_cap_wq);
1677         if (invalidate)
1678                 ceph_queue_invalidate(inode);
1679         if (dirty_dropped)
1680                 iput(inode);
1681         return 0;
1682 }
1683
1684 /*
1685  * caller must hold session s_mutex
1686  */
1687 static void remove_session_caps(struct ceph_mds_session *session)
1688 {
1689         struct ceph_fs_client *fsc = session->s_mdsc->fsc;
1690         struct super_block *sb = fsc->sb;
1691         LIST_HEAD(dispose);
1692
1693         dout("remove_session_caps on %p\n", session);
1694         ceph_iterate_session_caps(session, remove_session_caps_cb, fsc);
1695
1696         wake_up_all(&fsc->mdsc->cap_flushing_wq);
1697
1698         spin_lock(&session->s_cap_lock);
1699         if (session->s_nr_caps > 0) {
1700                 struct inode *inode;
1701                 struct ceph_cap *cap, *prev = NULL;
1702                 struct ceph_vino vino;
1703                 /*
1704                  * iterate_session_caps() skips inodes that are being
1705                  * deleted, we need to wait until deletions are complete.
1706                  * __wait_on_freeing_inode() is designed for the job,
1707                  * but it is not exported, so use lookup inode function
1708                  * to access it.
1709                  */
1710                 while (!list_empty(&session->s_caps)) {
1711                         cap = list_entry(session->s_caps.next,
1712                                          struct ceph_cap, session_caps);
1713                         if (cap == prev)
1714                                 break;
1715                         prev = cap;
1716                         vino = cap->ci->i_vino;
1717                         spin_unlock(&session->s_cap_lock);
1718
1719                         inode = ceph_find_inode(sb, vino);
1720                         iput(inode);
1721
1722                         spin_lock(&session->s_cap_lock);
1723                 }
1724         }
1725
1726         // drop cap expires and unlock s_cap_lock
1727         detach_cap_releases(session, &dispose);
1728
1729         BUG_ON(session->s_nr_caps > 0);
1730         BUG_ON(!list_empty(&session->s_cap_flushing));
1731         spin_unlock(&session->s_cap_lock);
1732         dispose_cap_releases(session->s_mdsc, &dispose);
1733 }
1734
1735 enum {
1736         RECONNECT,
1737         RENEWCAPS,
1738         FORCE_RO,
1739 };
1740
1741 /*
1742  * wake up any threads waiting on this session's caps.  if the cap is
1743  * old (didn't get renewed on the client reconnect), remove it now.
1744  *
1745  * caller must hold s_mutex.
1746  */
1747 static int wake_up_session_cb(struct inode *inode, struct ceph_cap *cap,
1748                               void *arg)
1749 {
1750         struct ceph_inode_info *ci = ceph_inode(inode);
1751         unsigned long ev = (unsigned long)arg;
1752
1753         if (ev == RECONNECT) {
1754                 spin_lock(&ci->i_ceph_lock);
1755                 ci->i_wanted_max_size = 0;
1756                 ci->i_requested_max_size = 0;
1757                 spin_unlock(&ci->i_ceph_lock);
1758         } else if (ev == RENEWCAPS) {
1759                 if (cap->cap_gen < atomic_read(&cap->session->s_cap_gen)) {
1760                         /* mds did not re-issue stale cap */
1761                         spin_lock(&ci->i_ceph_lock);
1762                         cap->issued = cap->implemented = CEPH_CAP_PIN;
1763                         spin_unlock(&ci->i_ceph_lock);
1764                 }
1765         } else if (ev == FORCE_RO) {
1766         }
1767         wake_up_all(&ci->i_cap_wq);
1768         return 0;
1769 }
1770
1771 static void wake_up_session_caps(struct ceph_mds_session *session, int ev)
1772 {
1773         dout("wake_up_session_caps %p mds%d\n", session, session->s_mds);
1774         ceph_iterate_session_caps(session, wake_up_session_cb,
1775                                   (void *)(unsigned long)ev);
1776 }
1777
1778 /*
1779  * Send periodic message to MDS renewing all currently held caps.  The
1780  * ack will reset the expiration for all caps from this session.
1781  *
1782  * caller holds s_mutex
1783  */
1784 static int send_renew_caps(struct ceph_mds_client *mdsc,
1785                            struct ceph_mds_session *session)
1786 {
1787         struct ceph_msg *msg;
1788         int state;
1789
1790         if (time_after_eq(jiffies, session->s_cap_ttl) &&
1791             time_after_eq(session->s_cap_ttl, session->s_renew_requested))
1792                 pr_info("mds%d caps stale\n", session->s_mds);
1793         session->s_renew_requested = jiffies;
1794
1795         /* do not try to renew caps until a recovering mds has reconnected
1796          * with its clients. */
1797         state = ceph_mdsmap_get_state(mdsc->mdsmap, session->s_mds);
1798         if (state < CEPH_MDS_STATE_RECONNECT) {
1799                 dout("send_renew_caps ignoring mds%d (%s)\n",
1800                      session->s_mds, ceph_mds_state_name(state));
1801                 return 0;
1802         }
1803
1804         dout("send_renew_caps to mds%d (%s)\n", session->s_mds,
1805                 ceph_mds_state_name(state));
1806         msg = create_session_msg(CEPH_SESSION_REQUEST_RENEWCAPS,
1807                                  ++session->s_renew_seq);
1808         if (!msg)
1809                 return -ENOMEM;
1810         ceph_con_send(&session->s_con, msg);
1811         return 0;
1812 }
1813
1814 static int send_flushmsg_ack(struct ceph_mds_client *mdsc,
1815                              struct ceph_mds_session *session, u64 seq)
1816 {
1817         struct ceph_msg *msg;
1818
1819         dout("send_flushmsg_ack to mds%d (%s)s seq %lld\n",
1820              session->s_mds, ceph_session_state_name(session->s_state), seq);
1821         msg = create_session_msg(CEPH_SESSION_FLUSHMSG_ACK, seq);
1822         if (!msg)
1823                 return -ENOMEM;
1824         ceph_con_send(&session->s_con, msg);
1825         return 0;
1826 }
1827
1828
1829 /*
1830  * Note new cap ttl, and any transition from stale -> not stale (fresh?).
1831  *
1832  * Called under session->s_mutex
1833  */
1834 static void renewed_caps(struct ceph_mds_client *mdsc,
1835                          struct ceph_mds_session *session, int is_renew)
1836 {
1837         int was_stale;
1838         int wake = 0;
1839
1840         spin_lock(&session->s_cap_lock);
1841         was_stale = is_renew && time_after_eq(jiffies, session->s_cap_ttl);
1842
1843         session->s_cap_ttl = session->s_renew_requested +
1844                 mdsc->mdsmap->m_session_timeout*HZ;
1845
1846         if (was_stale) {
1847                 if (time_before(jiffies, session->s_cap_ttl)) {
1848                         pr_info("mds%d caps renewed\n", session->s_mds);
1849                         wake = 1;
1850                 } else {
1851                         pr_info("mds%d caps still stale\n", session->s_mds);
1852                 }
1853         }
1854         dout("renewed_caps mds%d ttl now %lu, was %s, now %s\n",
1855              session->s_mds, session->s_cap_ttl, was_stale ? "stale" : "fresh",
1856              time_before(jiffies, session->s_cap_ttl) ? "stale" : "fresh");
1857         spin_unlock(&session->s_cap_lock);
1858
1859         if (wake)
1860                 wake_up_session_caps(session, RENEWCAPS);
1861 }
1862
1863 /*
1864  * send a session close request
1865  */
1866 static int request_close_session(struct ceph_mds_session *session)
1867 {
1868         struct ceph_msg *msg;
1869
1870         dout("request_close_session mds%d state %s seq %lld\n",
1871              session->s_mds, ceph_session_state_name(session->s_state),
1872              session->s_seq);
1873         msg = create_session_msg(CEPH_SESSION_REQUEST_CLOSE, session->s_seq);
1874         if (!msg)
1875                 return -ENOMEM;
1876         ceph_con_send(&session->s_con, msg);
1877         return 1;
1878 }
1879
1880 /*
1881  * Called with s_mutex held.
1882  */
1883 static int __close_session(struct ceph_mds_client *mdsc,
1884                          struct ceph_mds_session *session)
1885 {
1886         if (session->s_state >= CEPH_MDS_SESSION_CLOSING)
1887                 return 0;
1888         session->s_state = CEPH_MDS_SESSION_CLOSING;
1889         return request_close_session(session);
1890 }
1891
1892 static bool drop_negative_children(struct dentry *dentry)
1893 {
1894         struct dentry *child;
1895         bool all_negative = true;
1896
1897         if (!d_is_dir(dentry))
1898                 goto out;
1899
1900         spin_lock(&dentry->d_lock);
1901         list_for_each_entry(child, &dentry->d_subdirs, d_child) {
1902                 if (d_really_is_positive(child)) {
1903                         all_negative = false;
1904                         break;
1905                 }
1906         }
1907         spin_unlock(&dentry->d_lock);
1908
1909         if (all_negative)
1910                 shrink_dcache_parent(dentry);
1911 out:
1912         return all_negative;
1913 }
1914
1915 /*
1916  * Trim old(er) caps.
1917  *
1918  * Because we can't cache an inode without one or more caps, we do
1919  * this indirectly: if a cap is unused, we prune its aliases, at which
1920  * point the inode will hopefully get dropped to.
1921  *
1922  * Yes, this is a bit sloppy.  Our only real goal here is to respond to
1923  * memory pressure from the MDS, though, so it needn't be perfect.
1924  */
1925 static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg)
1926 {
1927         int *remaining = arg;
1928         struct ceph_inode_info *ci = ceph_inode(inode);
1929         int used, wanted, oissued, mine;
1930
1931         if (*remaining <= 0)
1932                 return -1;
1933
1934         spin_lock(&ci->i_ceph_lock);
1935         mine = cap->issued | cap->implemented;
1936         used = __ceph_caps_used(ci);
1937         wanted = __ceph_caps_file_wanted(ci);
1938         oissued = __ceph_caps_issued_other(ci, cap);
1939
1940         dout("trim_caps_cb %p cap %p mine %s oissued %s used %s wanted %s\n",
1941              inode, cap, ceph_cap_string(mine), ceph_cap_string(oissued),
1942              ceph_cap_string(used), ceph_cap_string(wanted));
1943         if (cap == ci->i_auth_cap) {
1944                 if (ci->i_dirty_caps || ci->i_flushing_caps ||
1945                     !list_empty(&ci->i_cap_snaps))
1946                         goto out;
1947                 if ((used | wanted) & CEPH_CAP_ANY_WR)
1948                         goto out;
1949                 /* Note: it's possible that i_filelock_ref becomes non-zero
1950                  * after dropping auth caps. It doesn't hurt because reply
1951                  * of lock mds request will re-add auth caps. */
1952                 if (atomic_read(&ci->i_filelock_ref) > 0)
1953                         goto out;
1954         }
1955         /* The inode has cached pages, but it's no longer used.
1956          * we can safely drop it */
1957         if (S_ISREG(inode->i_mode) &&
1958             wanted == 0 && used == CEPH_CAP_FILE_CACHE &&
1959             !(oissued & CEPH_CAP_FILE_CACHE)) {
1960           used = 0;
1961           oissued = 0;
1962         }
1963         if ((used | wanted) & ~oissued & mine)
1964                 goto out;   /* we need these caps */
1965
1966         if (oissued) {
1967                 /* we aren't the only cap.. just remove us */
1968                 __ceph_remove_cap(cap, true);
1969                 (*remaining)--;
1970         } else {
1971                 struct dentry *dentry;
1972                 /* try dropping referring dentries */
1973                 spin_unlock(&ci->i_ceph_lock);
1974                 dentry = d_find_any_alias(inode);
1975                 if (dentry && drop_negative_children(dentry)) {
1976                         int count;
1977                         dput(dentry);
1978                         d_prune_aliases(inode);
1979                         count = atomic_read(&inode->i_count);
1980                         if (count == 1)
1981                                 (*remaining)--;
1982                         dout("trim_caps_cb %p cap %p pruned, count now %d\n",
1983                              inode, cap, count);
1984                 } else {
1985                         dput(dentry);
1986                 }
1987                 return 0;
1988         }
1989
1990 out:
1991         spin_unlock(&ci->i_ceph_lock);
1992         return 0;
1993 }
1994
1995 /*
1996  * Trim session cap count down to some max number.
1997  */
1998 int ceph_trim_caps(struct ceph_mds_client *mdsc,
1999                    struct ceph_mds_session *session,
2000                    int max_caps)
2001 {
2002         int trim_caps = session->s_nr_caps - max_caps;
2003
2004         dout("trim_caps mds%d start: %d / %d, trim %d\n",
2005              session->s_mds, session->s_nr_caps, max_caps, trim_caps);
2006         if (trim_caps > 0) {
2007                 int remaining = trim_caps;
2008
2009                 ceph_iterate_session_caps(session, trim_caps_cb, &remaining);
2010                 dout("trim_caps mds%d done: %d / %d, trimmed %d\n",
2011                      session->s_mds, session->s_nr_caps, max_caps,
2012                         trim_caps - remaining);
2013         }
2014
2015         ceph_flush_cap_releases(mdsc, session);
2016         return 0;
2017 }
2018
2019 static int check_caps_flush(struct ceph_mds_client *mdsc,
2020                             u64 want_flush_tid)
2021 {
2022         int ret = 1;
2023
2024         spin_lock(&mdsc->cap_dirty_lock);
2025         if (!list_empty(&mdsc->cap_flush_list)) {
2026                 struct ceph_cap_flush *cf =
2027                         list_first_entry(&mdsc->cap_flush_list,
2028                                          struct ceph_cap_flush, g_list);
2029                 if (cf->tid <= want_flush_tid) {
2030                         dout("check_caps_flush still flushing tid "
2031                              "%llu <= %llu\n", cf->tid, want_flush_tid);
2032                         ret = 0;
2033                 }
2034         }
2035         spin_unlock(&mdsc->cap_dirty_lock);
2036         return ret;
2037 }
2038
2039 /*
2040  * flush all dirty inode data to disk.
2041  *
2042  * returns true if we've flushed through want_flush_tid
2043  */
2044 static void wait_caps_flush(struct ceph_mds_client *mdsc,
2045                             u64 want_flush_tid)
2046 {
2047         dout("check_caps_flush want %llu\n", want_flush_tid);
2048
2049         wait_event(mdsc->cap_flushing_wq,
2050                    check_caps_flush(mdsc, want_flush_tid));
2051
2052         dout("check_caps_flush ok, flushed thru %llu\n", want_flush_tid);
2053 }
2054
2055 /*
2056  * called under s_mutex
2057  */
2058 static void ceph_send_cap_releases(struct ceph_mds_client *mdsc,
2059                                    struct ceph_mds_session *session)
2060 {
2061         struct ceph_msg *msg = NULL;
2062         struct ceph_mds_cap_release *head;
2063         struct ceph_mds_cap_item *item;
2064         struct ceph_osd_client *osdc = &mdsc->fsc->client->osdc;
2065         struct ceph_cap *cap;
2066         LIST_HEAD(tmp_list);
2067         int num_cap_releases;
2068         __le32  barrier, *cap_barrier;
2069
2070         down_read(&osdc->lock);
2071         barrier = cpu_to_le32(osdc->epoch_barrier);
2072         up_read(&osdc->lock);
2073
2074         spin_lock(&session->s_cap_lock);
2075 again:
2076         list_splice_init(&session->s_cap_releases, &tmp_list);
2077         num_cap_releases = session->s_num_cap_releases;
2078         session->s_num_cap_releases = 0;
2079         spin_unlock(&session->s_cap_lock);
2080
2081         while (!list_empty(&tmp_list)) {
2082                 if (!msg) {
2083                         msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPRELEASE,
2084                                         PAGE_SIZE, GFP_NOFS, false);
2085                         if (!msg)
2086                                 goto out_err;
2087                         head = msg->front.iov_base;
2088                         head->num = cpu_to_le32(0);
2089                         msg->front.iov_len = sizeof(*head);
2090
2091                         msg->hdr.version = cpu_to_le16(2);
2092                         msg->hdr.compat_version = cpu_to_le16(1);
2093                 }
2094
2095                 cap = list_first_entry(&tmp_list, struct ceph_cap,
2096                                         session_caps);
2097                 list_del(&cap->session_caps);
2098                 num_cap_releases--;
2099
2100                 head = msg->front.iov_base;
2101                 put_unaligned_le32(get_unaligned_le32(&head->num) + 1,
2102                                    &head->num);
2103                 item = msg->front.iov_base + msg->front.iov_len;
2104                 item->ino = cpu_to_le64(cap->cap_ino);
2105                 item->cap_id = cpu_to_le64(cap->cap_id);
2106                 item->migrate_seq = cpu_to_le32(cap->mseq);
2107                 item->seq = cpu_to_le32(cap->issue_seq);
2108                 msg->front.iov_len += sizeof(*item);
2109
2110                 ceph_put_cap(mdsc, cap);
2111
2112                 if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) {
2113                         // Append cap_barrier field
2114                         cap_barrier = msg->front.iov_base + msg->front.iov_len;
2115                         *cap_barrier = barrier;
2116                         msg->front.iov_len += sizeof(*cap_barrier);
2117
2118                         msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2119                         dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
2120                         ceph_con_send(&session->s_con, msg);
2121                         msg = NULL;
2122                 }
2123         }
2124
2125         BUG_ON(num_cap_releases != 0);
2126
2127         spin_lock(&session->s_cap_lock);
2128         if (!list_empty(&session->s_cap_releases))
2129                 goto again;
2130         spin_unlock(&session->s_cap_lock);
2131
2132         if (msg) {
2133                 // Append cap_barrier field
2134                 cap_barrier = msg->front.iov_base + msg->front.iov_len;
2135                 *cap_barrier = barrier;
2136                 msg->front.iov_len += sizeof(*cap_barrier);
2137
2138                 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2139                 dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
2140                 ceph_con_send(&session->s_con, msg);
2141         }
2142         return;
2143 out_err:
2144         pr_err("send_cap_releases mds%d, failed to allocate message\n",
2145                 session->s_mds);
2146         spin_lock(&session->s_cap_lock);
2147         list_splice(&tmp_list, &session->s_cap_releases);
2148         session->s_num_cap_releases += num_cap_releases;
2149         spin_unlock(&session->s_cap_lock);
2150 }
2151
2152 static void ceph_cap_release_work(struct work_struct *work)
2153 {
2154         struct ceph_mds_session *session =
2155                 container_of(work, struct ceph_mds_session, s_cap_release_work);
2156
2157         mutex_lock(&session->s_mutex);
2158         if (session->s_state == CEPH_MDS_SESSION_OPEN ||
2159             session->s_state == CEPH_MDS_SESSION_HUNG)
2160                 ceph_send_cap_releases(session->s_mdsc, session);
2161         mutex_unlock(&session->s_mutex);
2162         ceph_put_mds_session(session);
2163 }
2164
2165 void ceph_flush_cap_releases(struct ceph_mds_client *mdsc,
2166                              struct ceph_mds_session *session)
2167 {
2168         if (mdsc->stopping)
2169                 return;
2170
2171         ceph_get_mds_session(session);
2172         if (queue_work(mdsc->fsc->cap_wq,
2173                        &session->s_cap_release_work)) {
2174                 dout("cap release work queued\n");
2175         } else {
2176                 ceph_put_mds_session(session);
2177                 dout("failed to queue cap release work\n");
2178         }
2179 }
2180
2181 /*
2182  * caller holds session->s_cap_lock
2183  */
2184 void __ceph_queue_cap_release(struct ceph_mds_session *session,
2185                               struct ceph_cap *cap)
2186 {
2187         list_add_tail(&cap->session_caps, &session->s_cap_releases);
2188         session->s_num_cap_releases++;
2189
2190         if (!(session->s_num_cap_releases % CEPH_CAPS_PER_RELEASE))
2191                 ceph_flush_cap_releases(session->s_mdsc, session);
2192 }
2193
2194 static void ceph_cap_reclaim_work(struct work_struct *work)
2195 {
2196         struct ceph_mds_client *mdsc =
2197                 container_of(work, struct ceph_mds_client, cap_reclaim_work);
2198         int ret = ceph_trim_dentries(mdsc);
2199         if (ret == -EAGAIN)
2200                 ceph_queue_cap_reclaim_work(mdsc);
2201 }
2202
2203 void ceph_queue_cap_reclaim_work(struct ceph_mds_client *mdsc)
2204 {
2205         if (mdsc->stopping)
2206                 return;
2207
2208         if (queue_work(mdsc->fsc->cap_wq, &mdsc->cap_reclaim_work)) {
2209                 dout("caps reclaim work queued\n");
2210         } else {
2211                 dout("failed to queue caps release work\n");
2212         }
2213 }
2214
2215 void ceph_reclaim_caps_nr(struct ceph_mds_client *mdsc, int nr)
2216 {
2217         int val;
2218         if (!nr)
2219                 return;
2220         val = atomic_add_return(nr, &mdsc->cap_reclaim_pending);
2221         if ((val % CEPH_CAPS_PER_RELEASE) < nr) {
2222                 atomic_set(&mdsc->cap_reclaim_pending, 0);
2223                 ceph_queue_cap_reclaim_work(mdsc);
2224         }
2225 }
2226
2227 /*
2228  * requests
2229  */
2230
2231 int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req,
2232                                     struct inode *dir)
2233 {
2234         struct ceph_inode_info *ci = ceph_inode(dir);
2235         struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
2236         struct ceph_mount_options *opt = req->r_mdsc->fsc->mount_options;
2237         size_t size = sizeof(struct ceph_mds_reply_dir_entry);
2238         unsigned int num_entries;
2239         int order;
2240
2241         spin_lock(&ci->i_ceph_lock);
2242         num_entries = ci->i_files + ci->i_subdirs;
2243         spin_unlock(&ci->i_ceph_lock);
2244         num_entries = max(num_entries, 1U);
2245         num_entries = min(num_entries, opt->max_readdir);
2246
2247         order = get_order(size * num_entries);
2248         while (order >= 0) {
2249                 rinfo->dir_entries = (void*)__get_free_pages(GFP_KERNEL |
2250                                                              __GFP_NOWARN,
2251                                                              order);
2252                 if (rinfo->dir_entries)
2253                         break;
2254                 order--;
2255         }
2256         if (!rinfo->dir_entries)
2257                 return -ENOMEM;
2258
2259         num_entries = (PAGE_SIZE << order) / size;
2260         num_entries = min(num_entries, opt->max_readdir);
2261
2262         rinfo->dir_buf_size = PAGE_SIZE << order;
2263         req->r_num_caps = num_entries + 1;
2264         req->r_args.readdir.max_entries = cpu_to_le32(num_entries);
2265         req->r_args.readdir.max_bytes = cpu_to_le32(opt->max_readdir_bytes);
2266         return 0;
2267 }
2268
2269 /*
2270  * Create an mds request.
2271  */
2272 struct ceph_mds_request *
2273 ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode)
2274 {
2275         struct ceph_mds_request *req;
2276
2277         req = kmem_cache_zalloc(ceph_mds_request_cachep, GFP_NOFS);
2278         if (!req)
2279                 return ERR_PTR(-ENOMEM);
2280
2281         mutex_init(&req->r_fill_mutex);
2282         req->r_mdsc = mdsc;
2283         req->r_started = jiffies;
2284         req->r_start_latency = ktime_get();
2285         req->r_resend_mds = -1;
2286         INIT_LIST_HEAD(&req->r_unsafe_dir_item);
2287         INIT_LIST_HEAD(&req->r_unsafe_target_item);
2288         req->r_fmode = -1;
2289         kref_init(&req->r_kref);
2290         RB_CLEAR_NODE(&req->r_node);
2291         INIT_LIST_HEAD(&req->r_wait);
2292         init_completion(&req->r_completion);
2293         init_completion(&req->r_safe_completion);
2294         INIT_LIST_HEAD(&req->r_unsafe_item);
2295
2296         ktime_get_coarse_real_ts64(&req->r_stamp);
2297
2298         req->r_op = op;
2299         req->r_direct_mode = mode;
2300         return req;
2301 }
2302
2303 /*
2304  * return oldest (lowest) request, tid in request tree, 0 if none.
2305  *
2306  * called under mdsc->mutex.
2307  */
2308 static struct ceph_mds_request *__get_oldest_req(struct ceph_mds_client *mdsc)
2309 {
2310         if (RB_EMPTY_ROOT(&mdsc->request_tree))
2311                 return NULL;
2312         return rb_entry(rb_first(&mdsc->request_tree),
2313                         struct ceph_mds_request, r_node);
2314 }
2315
2316 static inline  u64 __get_oldest_tid(struct ceph_mds_client *mdsc)
2317 {
2318         return mdsc->oldest_tid;
2319 }
2320
2321 /*
2322  * Build a dentry's path.  Allocate on heap; caller must kfree.  Based
2323  * on build_path_from_dentry in fs/cifs/dir.c.
2324  *
2325  * If @stop_on_nosnap, generate path relative to the first non-snapped
2326  * inode.
2327  *
2328  * Encode hidden .snap dirs as a double /, i.e.
2329  *   foo/.snap/bar -> foo//bar
2330  */
2331 char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *pbase,
2332                            int stop_on_nosnap)
2333 {
2334         struct dentry *temp;
2335         char *path;
2336         int pos;
2337         unsigned seq;
2338         u64 base;
2339
2340         if (!dentry)
2341                 return ERR_PTR(-EINVAL);
2342
2343         path = __getname();
2344         if (!path)
2345                 return ERR_PTR(-ENOMEM);
2346 retry:
2347         pos = PATH_MAX - 1;
2348         path[pos] = '\0';
2349
2350         seq = read_seqbegin(&rename_lock);
2351         rcu_read_lock();
2352         temp = dentry;
2353         for (;;) {
2354                 struct inode *inode;
2355
2356                 spin_lock(&temp->d_lock);
2357                 inode = d_inode(temp);
2358                 if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
2359                         dout("build_path path+%d: %p SNAPDIR\n",
2360                              pos, temp);
2361                 } else if (stop_on_nosnap && inode && dentry != temp &&
2362                            ceph_snap(inode) == CEPH_NOSNAP) {
2363                         spin_unlock(&temp->d_lock);
2364                         pos++; /* get rid of any prepended '/' */
2365                         break;
2366                 } else {
2367                         pos -= temp->d_name.len;
2368                         if (pos < 0) {
2369                                 spin_unlock(&temp->d_lock);
2370                                 break;
2371                         }
2372                         memcpy(path + pos, temp->d_name.name, temp->d_name.len);
2373                 }
2374                 spin_unlock(&temp->d_lock);
2375                 temp = READ_ONCE(temp->d_parent);
2376
2377                 /* Are we at the root? */
2378                 if (IS_ROOT(temp))
2379                         break;
2380
2381                 /* Are we out of buffer? */
2382                 if (--pos < 0)
2383                         break;
2384
2385                 path[pos] = '/';
2386         }
2387         base = ceph_ino(d_inode(temp));
2388         rcu_read_unlock();
2389
2390         if (read_seqretry(&rename_lock, seq))
2391                 goto retry;
2392
2393         if (pos < 0) {
2394                 /*
2395                  * A rename didn't occur, but somehow we didn't end up where
2396                  * we thought we would. Throw a warning and try again.
2397                  */
2398                 pr_warn("build_path did not end path lookup where "
2399                         "expected, pos is %d\n", pos);
2400                 goto retry;
2401         }
2402
2403         *pbase = base;
2404         *plen = PATH_MAX - 1 - pos;
2405         dout("build_path on %p %d built %llx '%.*s'\n",
2406              dentry, d_count(dentry), base, *plen, path + pos);
2407         return path + pos;
2408 }
2409
2410 static int build_dentry_path(struct dentry *dentry, struct inode *dir,
2411                              const char **ppath, int *ppathlen, u64 *pino,
2412                              bool *pfreepath, bool parent_locked)
2413 {
2414         char *path;
2415
2416         rcu_read_lock();
2417         if (!dir)
2418                 dir = d_inode_rcu(dentry->d_parent);
2419         if (dir && parent_locked && ceph_snap(dir) == CEPH_NOSNAP) {
2420                 *pino = ceph_ino(dir);
2421                 rcu_read_unlock();
2422                 *ppath = dentry->d_name.name;
2423                 *ppathlen = dentry->d_name.len;
2424                 return 0;
2425         }
2426         rcu_read_unlock();
2427         path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
2428         if (IS_ERR(path))
2429                 return PTR_ERR(path);
2430         *ppath = path;
2431         *pfreepath = true;
2432         return 0;
2433 }
2434
2435 static int build_inode_path(struct inode *inode,
2436                             const char **ppath, int *ppathlen, u64 *pino,
2437                             bool *pfreepath)
2438 {
2439         struct dentry *dentry;
2440         char *path;
2441
2442         if (ceph_snap(inode) == CEPH_NOSNAP) {
2443                 *pino = ceph_ino(inode);
2444                 *ppathlen = 0;
2445                 return 0;
2446         }
2447         dentry = d_find_alias(inode);
2448         path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
2449         dput(dentry);
2450         if (IS_ERR(path))
2451                 return PTR_ERR(path);
2452         *ppath = path;
2453         *pfreepath = true;
2454         return 0;
2455 }
2456
2457 /*
2458  * request arguments may be specified via an inode *, a dentry *, or
2459  * an explicit ino+path.
2460  */
2461 static int set_request_path_attr(struct inode *rinode, struct dentry *rdentry,
2462                                   struct inode *rdiri, const char *rpath,
2463                                   u64 rino, const char **ppath, int *pathlen,
2464                                   u64 *ino, bool *freepath, bool parent_locked)
2465 {
2466         int r = 0;
2467
2468         if (rinode) {
2469                 r = build_inode_path(rinode, ppath, pathlen, ino, freepath);
2470                 dout(" inode %p %llx.%llx\n", rinode, ceph_ino(rinode),
2471                      ceph_snap(rinode));
2472         } else if (rdentry) {
2473                 r = build_dentry_path(rdentry, rdiri, ppath, pathlen, ino,
2474                                         freepath, parent_locked);
2475                 dout(" dentry %p %llx/%.*s\n", rdentry, *ino, *pathlen,
2476                      *ppath);
2477         } else if (rpath || rino) {
2478                 *ino = rino;
2479                 *ppath = rpath;
2480                 *pathlen = rpath ? strlen(rpath) : 0;
2481                 dout(" path %.*s\n", *pathlen, rpath);
2482         }
2483
2484         return r;
2485 }
2486
2487 static void encode_timestamp_and_gids(void **p,
2488                                       const struct ceph_mds_request *req)
2489 {
2490         struct ceph_timespec ts;
2491         int i;
2492
2493         ceph_encode_timespec64(&ts, &req->r_stamp);
2494         ceph_encode_copy(p, &ts, sizeof(ts));
2495
2496         /* gid_list */
2497         ceph_encode_32(p, req->r_cred->group_info->ngroups);
2498         for (i = 0; i < req->r_cred->group_info->ngroups; i++)
2499                 ceph_encode_64(p, from_kgid(&init_user_ns,
2500                                             req->r_cred->group_info->gid[i]));
2501 }
2502
2503 /*
2504  * called under mdsc->mutex
2505  */
2506 static struct ceph_msg *create_request_message(struct ceph_mds_session *session,
2507                                                struct ceph_mds_request *req,
2508                                                bool drop_cap_releases)
2509 {
2510         int mds = session->s_mds;
2511         struct ceph_mds_client *mdsc = session->s_mdsc;
2512         struct ceph_msg *msg;
2513         struct ceph_mds_request_head_old *head;
2514         const char *path1 = NULL;
2515         const char *path2 = NULL;
2516         u64 ino1 = 0, ino2 = 0;
2517         int pathlen1 = 0, pathlen2 = 0;
2518         bool freepath1 = false, freepath2 = false;
2519         int len;
2520         u16 releases;
2521         void *p, *end;
2522         int ret;
2523         bool legacy = !(session->s_con.peer_features & CEPH_FEATURE_FS_BTIME);
2524
2525         ret = set_request_path_attr(req->r_inode, req->r_dentry,
2526                               req->r_parent, req->r_path1, req->r_ino1.ino,
2527                               &path1, &pathlen1, &ino1, &freepath1,
2528                               test_bit(CEPH_MDS_R_PARENT_LOCKED,
2529                                         &req->r_req_flags));
2530         if (ret < 0) {
2531                 msg = ERR_PTR(ret);
2532                 goto out;
2533         }
2534
2535         /* If r_old_dentry is set, then assume that its parent is locked */
2536         ret = set_request_path_attr(NULL, req->r_old_dentry,
2537                               req->r_old_dentry_dir,
2538                               req->r_path2, req->r_ino2.ino,
2539                               &path2, &pathlen2, &ino2, &freepath2, true);
2540         if (ret < 0) {
2541                 msg = ERR_PTR(ret);
2542                 goto out_free1;
2543         }
2544
2545         len = legacy ? sizeof(*head) : sizeof(struct ceph_mds_request_head);
2546         len += pathlen1 + pathlen2 + 2*(1 + sizeof(u32) + sizeof(u64)) +
2547                 sizeof(struct ceph_timespec);
2548         len += sizeof(u32) + (sizeof(u64) * req->r_cred->group_info->ngroups);
2549
2550         /* calculate (max) length for cap releases */
2551         len += sizeof(struct ceph_mds_request_release) *
2552                 (!!req->r_inode_drop + !!req->r_dentry_drop +
2553                  !!req->r_old_inode_drop + !!req->r_old_dentry_drop);
2554
2555         if (req->r_dentry_drop)
2556                 len += pathlen1;
2557         if (req->r_old_dentry_drop)
2558                 len += pathlen2;
2559
2560         msg = ceph_msg_new2(CEPH_MSG_CLIENT_REQUEST, len, 1, GFP_NOFS, false);
2561         if (!msg) {
2562                 msg = ERR_PTR(-ENOMEM);
2563                 goto out_free2;
2564         }
2565
2566         msg->hdr.tid = cpu_to_le64(req->r_tid);
2567
2568         /*
2569          * The old ceph_mds_request_head didn't contain a version field, and
2570          * one was added when we moved the message version from 3->4.
2571          */
2572         if (legacy) {
2573                 msg->hdr.version = cpu_to_le16(3);
2574                 head = msg->front.iov_base;
2575                 p = msg->front.iov_base + sizeof(*head);
2576         } else {
2577                 struct ceph_mds_request_head *new_head = msg->front.iov_base;
2578
2579                 msg->hdr.version = cpu_to_le16(4);
2580                 new_head->version = cpu_to_le16(CEPH_MDS_REQUEST_HEAD_VERSION);
2581                 head = (struct ceph_mds_request_head_old *)&new_head->oldest_client_tid;
2582                 p = msg->front.iov_base + sizeof(*new_head);
2583         }
2584
2585         end = msg->front.iov_base + msg->front.iov_len;
2586
2587         head->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch);
2588         head->op = cpu_to_le32(req->r_op);
2589         head->caller_uid = cpu_to_le32(from_kuid(&init_user_ns,
2590                                                  req->r_cred->fsuid));
2591         head->caller_gid = cpu_to_le32(from_kgid(&init_user_ns,
2592                                                  req->r_cred->fsgid));
2593         head->ino = cpu_to_le64(req->r_deleg_ino);
2594         head->args = req->r_args;
2595
2596         ceph_encode_filepath(&p, end, ino1, path1);
2597         ceph_encode_filepath(&p, end, ino2, path2);
2598
2599         /* make note of release offset, in case we need to replay */
2600         req->r_request_release_offset = p - msg->front.iov_base;
2601
2602         /* cap releases */
2603         releases = 0;
2604         if (req->r_inode_drop)
2605                 releases += ceph_encode_inode_release(&p,
2606                       req->r_inode ? req->r_inode : d_inode(req->r_dentry),
2607                       mds, req->r_inode_drop, req->r_inode_unless,
2608                       req->r_op == CEPH_MDS_OP_READDIR);
2609         if (req->r_dentry_drop)
2610                 releases += ceph_encode_dentry_release(&p, req->r_dentry,
2611                                 req->r_parent, mds, req->r_dentry_drop,
2612                                 req->r_dentry_unless);
2613         if (req->r_old_dentry_drop)
2614                 releases += ceph_encode_dentry_release(&p, req->r_old_dentry,
2615                                 req->r_old_dentry_dir, mds,
2616                                 req->r_old_dentry_drop,
2617                                 req->r_old_dentry_unless);
2618         if (req->r_old_inode_drop)
2619                 releases += ceph_encode_inode_release(&p,
2620                       d_inode(req->r_old_dentry),
2621                       mds, req->r_old_inode_drop, req->r_old_inode_unless, 0);
2622
2623         if (drop_cap_releases) {
2624                 releases = 0;
2625                 p = msg->front.iov_base + req->r_request_release_offset;
2626         }
2627
2628         head->num_releases = cpu_to_le16(releases);
2629
2630         encode_timestamp_and_gids(&p, req);
2631
2632         if (WARN_ON_ONCE(p > end)) {
2633                 ceph_msg_put(msg);
2634                 msg = ERR_PTR(-ERANGE);
2635                 goto out_free2;
2636         }
2637
2638         msg->front.iov_len = p - msg->front.iov_base;
2639         msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2640
2641         if (req->r_pagelist) {
2642                 struct ceph_pagelist *pagelist = req->r_pagelist;
2643                 ceph_msg_data_add_pagelist(msg, pagelist);
2644                 msg->hdr.data_len = cpu_to_le32(pagelist->length);
2645         } else {
2646                 msg->hdr.data_len = 0;
2647         }
2648
2649         msg->hdr.data_off = cpu_to_le16(0);
2650
2651 out_free2:
2652         if (freepath2)
2653                 ceph_mdsc_free_path((char *)path2, pathlen2);
2654 out_free1:
2655         if (freepath1)
2656                 ceph_mdsc_free_path((char *)path1, pathlen1);
2657 out:
2658         return msg;
2659 }
2660
2661 /*
2662  * called under mdsc->mutex if error, under no mutex if
2663  * success.
2664  */
2665 static void complete_request(struct ceph_mds_client *mdsc,
2666                              struct ceph_mds_request *req)
2667 {
2668         req->r_end_latency = ktime_get();
2669
2670         if (req->r_callback)
2671                 req->r_callback(mdsc, req);
2672         complete_all(&req->r_completion);
2673 }
2674
2675 static struct ceph_mds_request_head_old *
2676 find_old_request_head(void *p, u64 features)
2677 {
2678         bool legacy = !(features & CEPH_FEATURE_FS_BTIME);
2679         struct ceph_mds_request_head *new_head;
2680
2681         if (legacy)
2682                 return (struct ceph_mds_request_head_old *)p;
2683         new_head = (struct ceph_mds_request_head *)p;
2684         return (struct ceph_mds_request_head_old *)&new_head->oldest_client_tid;
2685 }
2686
2687 /*
2688  * called under mdsc->mutex
2689  */
2690 static int __prepare_send_request(struct ceph_mds_session *session,
2691                                   struct ceph_mds_request *req,
2692                                   bool drop_cap_releases)
2693 {
2694         int mds = session->s_mds;
2695         struct ceph_mds_client *mdsc = session->s_mdsc;
2696         struct ceph_mds_request_head_old *rhead;
2697         struct ceph_msg *msg;
2698         int flags = 0;
2699
2700         req->r_attempts++;
2701         if (req->r_inode) {
2702                 struct ceph_cap *cap =
2703                         ceph_get_cap_for_mds(ceph_inode(req->r_inode), mds);
2704
2705                 if (cap)
2706                         req->r_sent_on_mseq = cap->mseq;
2707                 else
2708                         req->r_sent_on_mseq = -1;
2709         }
2710         dout("prepare_send_request %p tid %lld %s (attempt %d)\n", req,
2711              req->r_tid, ceph_mds_op_name(req->r_op), req->r_attempts);
2712
2713         if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
2714                 void *p;
2715
2716                 /*
2717                  * Replay.  Do not regenerate message (and rebuild
2718                  * paths, etc.); just use the original message.
2719                  * Rebuilding paths will break for renames because
2720                  * d_move mangles the src name.
2721                  */
2722                 msg = req->r_request;
2723                 rhead = find_old_request_head(msg->front.iov_base,
2724                                               session->s_con.peer_features);
2725
2726                 flags = le32_to_cpu(rhead->flags);
2727                 flags |= CEPH_MDS_FLAG_REPLAY;
2728                 rhead->flags = cpu_to_le32(flags);
2729
2730                 if (req->r_target_inode)
2731                         rhead->ino = cpu_to_le64(ceph_ino(req->r_target_inode));
2732
2733                 rhead->num_retry = req->r_attempts - 1;
2734
2735                 /* remove cap/dentry releases from message */
2736                 rhead->num_releases = 0;
2737
2738                 p = msg->front.iov_base + req->r_request_release_offset;
2739                 encode_timestamp_and_gids(&p, req);
2740
2741                 msg->front.iov_len = p - msg->front.iov_base;
2742                 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2743                 return 0;
2744         }
2745
2746         if (req->r_request) {
2747                 ceph_msg_put(req->r_request);
2748                 req->r_request = NULL;
2749         }
2750         msg = create_request_message(session, req, drop_cap_releases);
2751         if (IS_ERR(msg)) {
2752                 req->r_err = PTR_ERR(msg);
2753                 return PTR_ERR(msg);
2754         }
2755         req->r_request = msg;
2756
2757         rhead = find_old_request_head(msg->front.iov_base,
2758                                       session->s_con.peer_features);
2759         rhead->oldest_client_tid = cpu_to_le64(__get_oldest_tid(mdsc));
2760         if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
2761                 flags |= CEPH_MDS_FLAG_REPLAY;
2762         if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags))
2763                 flags |= CEPH_MDS_FLAG_ASYNC;
2764         if (req->r_parent)
2765                 flags |= CEPH_MDS_FLAG_WANT_DENTRY;
2766         rhead->flags = cpu_to_le32(flags);
2767         rhead->num_fwd = req->r_num_fwd;
2768         rhead->num_retry = req->r_attempts - 1;
2769
2770         dout(" r_parent = %p\n", req->r_parent);
2771         return 0;
2772 }
2773
2774 /*
2775  * called under mdsc->mutex
2776  */
2777 static int __send_request(struct ceph_mds_session *session,
2778                           struct ceph_mds_request *req,
2779                           bool drop_cap_releases)
2780 {
2781         int err;
2782
2783         err = __prepare_send_request(session, req, drop_cap_releases);
2784         if (!err) {
2785                 ceph_msg_get(req->r_request);
2786                 ceph_con_send(&session->s_con, req->r_request);
2787         }
2788
2789         return err;
2790 }
2791
2792 /*
2793  * send request, or put it on the appropriate wait list.
2794  */
2795 static void __do_request(struct ceph_mds_client *mdsc,
2796                         struct ceph_mds_request *req)
2797 {
2798         struct ceph_mds_session *session = NULL;
2799         int mds = -1;
2800         int err = 0;
2801         bool random;
2802
2803         if (req->r_err || test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags)) {
2804                 if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags))
2805                         __unregister_request(mdsc, req);
2806                 return;
2807         }
2808
2809         if (req->r_timeout &&
2810             time_after_eq(jiffies, req->r_started + req->r_timeout)) {
2811                 dout("do_request timed out\n");
2812                 err = -ETIMEDOUT;
2813                 goto finish;
2814         }
2815         if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN) {
2816                 dout("do_request forced umount\n");
2817                 err = -EIO;
2818                 goto finish;
2819         }
2820         if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_MOUNTING) {
2821                 if (mdsc->mdsmap_err) {
2822                         err = mdsc->mdsmap_err;
2823                         dout("do_request mdsmap err %d\n", err);
2824                         goto finish;
2825                 }
2826                 if (mdsc->mdsmap->m_epoch == 0) {
2827                         dout("do_request no mdsmap, waiting for map\n");
2828                         list_add(&req->r_wait, &mdsc->waiting_for_map);
2829                         return;
2830                 }
2831                 if (!(mdsc->fsc->mount_options->flags &
2832                       CEPH_MOUNT_OPT_MOUNTWAIT) &&
2833                     !ceph_mdsmap_is_cluster_available(mdsc->mdsmap)) {
2834                         err = -EHOSTUNREACH;
2835                         goto finish;
2836                 }
2837         }
2838
2839         put_request_session(req);
2840
2841         mds = __choose_mds(mdsc, req, &random);
2842         if (mds < 0 ||
2843             ceph_mdsmap_get_state(mdsc->mdsmap, mds) < CEPH_MDS_STATE_ACTIVE) {
2844                 if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags)) {
2845                         err = -EJUKEBOX;
2846                         goto finish;
2847                 }
2848                 dout("do_request no mds or not active, waiting for map\n");
2849                 list_add(&req->r_wait, &mdsc->waiting_for_map);
2850                 return;
2851         }
2852
2853         /* get, open session */
2854         session = __ceph_lookup_mds_session(mdsc, mds);
2855         if (!session) {
2856                 session = register_session(mdsc, mds);
2857                 if (IS_ERR(session)) {
2858                         err = PTR_ERR(session);
2859                         goto finish;
2860                 }
2861         }
2862         req->r_session = ceph_get_mds_session(session);
2863
2864         dout("do_request mds%d session %p state %s\n", mds, session,
2865              ceph_session_state_name(session->s_state));
2866         if (session->s_state != CEPH_MDS_SESSION_OPEN &&
2867             session->s_state != CEPH_MDS_SESSION_HUNG) {
2868                 /*
2869                  * We cannot queue async requests since the caps and delegated
2870                  * inodes are bound to the session. Just return -EJUKEBOX and
2871                  * let the caller retry a sync request in that case.
2872                  */
2873                 if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags)) {
2874                         err = -EJUKEBOX;
2875                         goto out_session;
2876                 }
2877
2878                 /*
2879                  * If the session has been REJECTED, then return a hard error,
2880                  * unless it's a CLEANRECOVER mount, in which case we'll queue
2881                  * it to the mdsc queue.
2882                  */
2883                 if (session->s_state == CEPH_MDS_SESSION_REJECTED) {
2884                         if (ceph_test_mount_opt(mdsc->fsc, CLEANRECOVER))
2885                                 list_add(&req->r_wait, &mdsc->waiting_for_map);
2886                         else
2887                                 err = -EACCES;
2888                         goto out_session;
2889                 }
2890
2891                 if (session->s_state == CEPH_MDS_SESSION_NEW ||
2892                     session->s_state == CEPH_MDS_SESSION_CLOSING) {
2893                         err = __open_session(mdsc, session);
2894                         if (err)
2895                                 goto out_session;
2896                         /* retry the same mds later */
2897                         if (random)
2898                                 req->r_resend_mds = mds;
2899                 }
2900                 list_add(&req->r_wait, &session->s_waiting);
2901                 goto out_session;
2902         }
2903
2904         /* send request */
2905         req->r_resend_mds = -1;   /* forget any previous mds hint */
2906
2907         if (req->r_request_started == 0)   /* note request start time */
2908                 req->r_request_started = jiffies;
2909
2910         err = __send_request(session, req, false);
2911
2912 out_session:
2913         ceph_put_mds_session(session);
2914 finish:
2915         if (err) {
2916                 dout("__do_request early error %d\n", err);
2917                 req->r_err = err;
2918                 complete_request(mdsc, req);
2919                 __unregister_request(mdsc, req);
2920         }
2921         return;
2922 }
2923
2924 /*
2925  * called under mdsc->mutex
2926  */
2927 static void __wake_requests(struct ceph_mds_client *mdsc,
2928                             struct list_head *head)
2929 {
2930         struct ceph_mds_request *req;
2931         LIST_HEAD(tmp_list);
2932
2933         list_splice_init(head, &tmp_list);
2934
2935         while (!list_empty(&tmp_list)) {
2936                 req = list_entry(tmp_list.next,
2937                                  struct ceph_mds_request, r_wait);
2938                 list_del_init(&req->r_wait);
2939                 dout(" wake request %p tid %llu\n", req, req->r_tid);
2940                 __do_request(mdsc, req);
2941         }
2942 }
2943
2944 /*
2945  * Wake up threads with requests pending for @mds, so that they can
2946  * resubmit their requests to a possibly different mds.
2947  */
2948 static void kick_requests(struct ceph_mds_client *mdsc, int mds)
2949 {
2950         struct ceph_mds_request *req;
2951         struct rb_node *p = rb_first(&mdsc->request_tree);
2952
2953         dout("kick_requests mds%d\n", mds);
2954         while (p) {
2955                 req = rb_entry(p, struct ceph_mds_request, r_node);
2956                 p = rb_next(p);
2957                 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
2958                         continue;
2959                 if (req->r_attempts > 0)
2960                         continue; /* only new requests */
2961                 if (req->r_session &&
2962                     req->r_session->s_mds == mds) {
2963                         dout(" kicking tid %llu\n", req->r_tid);
2964                         list_del_init(&req->r_wait);
2965                         __do_request(mdsc, req);
2966                 }
2967         }
2968 }
2969
2970 int ceph_mdsc_submit_request(struct ceph_mds_client *mdsc, struct inode *dir,
2971                               struct ceph_mds_request *req)
2972 {
2973         int err = 0;
2974
2975         /* take CAP_PIN refs for r_inode, r_parent, r_old_dentry */
2976         if (req->r_inode)
2977                 ceph_get_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
2978         if (req->r_parent) {
2979                 struct ceph_inode_info *ci = ceph_inode(req->r_parent);
2980                 int fmode = (req->r_op & CEPH_MDS_OP_WRITE) ?
2981                             CEPH_FILE_MODE_WR : CEPH_FILE_MODE_RD;
2982                 spin_lock(&ci->i_ceph_lock);
2983                 ceph_take_cap_refs(ci, CEPH_CAP_PIN, false);
2984                 __ceph_touch_fmode(ci, mdsc, fmode);
2985                 spin_unlock(&ci->i_ceph_lock);
2986         }
2987         if (req->r_old_dentry_dir)
2988                 ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir),
2989                                   CEPH_CAP_PIN);
2990
2991         if (req->r_inode) {
2992                 err = ceph_wait_on_async_create(req->r_inode);
2993                 if (err) {
2994                         dout("%s: wait for async create returned: %d\n",
2995                              __func__, err);
2996                         return err;
2997                 }
2998         }
2999
3000         if (!err && req->r_old_inode) {
3001                 err = ceph_wait_on_async_create(req->r_old_inode);
3002                 if (err) {
3003                         dout("%s: wait for async create returned: %d\n",
3004                              __func__, err);
3005                         return err;
3006                 }
3007         }
3008
3009         dout("submit_request on %p for inode %p\n", req, dir);
3010         mutex_lock(&mdsc->mutex);
3011         __register_request(mdsc, req, dir);
3012         __do_request(mdsc, req);
3013         err = req->r_err;
3014         mutex_unlock(&mdsc->mutex);
3015         return err;
3016 }
3017
3018 static int ceph_mdsc_wait_request(struct ceph_mds_client *mdsc,
3019                                   struct ceph_mds_request *req)
3020 {
3021         int err;
3022
3023         /* wait */
3024         dout("do_request waiting\n");
3025         if (!req->r_timeout && req->r_wait_for_completion) {
3026                 err = req->r_wait_for_completion(mdsc, req);
3027         } else {
3028                 long timeleft = wait_for_completion_killable_timeout(
3029                                         &req->r_completion,
3030                                         ceph_timeout_jiffies(req->r_timeout));
3031                 if (timeleft > 0)
3032                         err = 0;
3033                 else if (!timeleft)
3034                         err = -ETIMEDOUT;  /* timed out */
3035                 else
3036                         err = timeleft;  /* killed */
3037         }
3038         dout("do_request waited, got %d\n", err);
3039         mutex_lock(&mdsc->mutex);
3040
3041         /* only abort if we didn't race with a real reply */
3042         if (test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags)) {
3043                 err = le32_to_cpu(req->r_reply_info.head->result);
3044         } else if (err < 0) {
3045                 dout("aborted request %lld with %d\n", req->r_tid, err);
3046
3047                 /*
3048                  * ensure we aren't running concurrently with
3049                  * ceph_fill_trace or ceph_readdir_prepopulate, which
3050                  * rely on locks (dir mutex) held by our caller.
3051                  */
3052                 mutex_lock(&req->r_fill_mutex);
3053                 req->r_err = err;
3054                 set_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags);
3055                 mutex_unlock(&req->r_fill_mutex);
3056
3057                 if (req->r_parent &&
3058                     (req->r_op & CEPH_MDS_OP_WRITE))
3059                         ceph_invalidate_dir_request(req);
3060         } else {
3061                 err = req->r_err;
3062         }
3063
3064         mutex_unlock(&mdsc->mutex);
3065         return err;
3066 }
3067
3068 /*
3069  * Synchrously perform an mds request.  Take care of all of the
3070  * session setup, forwarding, retry details.
3071  */
3072 int ceph_mdsc_do_request(struct ceph_mds_client *mdsc,
3073                          struct inode *dir,
3074                          struct ceph_mds_request *req)
3075 {
3076         int err;
3077
3078         dout("do_request on %p\n", req);
3079
3080         /* issue */
3081         err = ceph_mdsc_submit_request(mdsc, dir, req);
3082         if (!err)
3083                 err = ceph_mdsc_wait_request(mdsc, req);
3084         dout("do_request %p done, result %d\n", req, err);
3085         return err;
3086 }
3087
3088 /*
3089  * Invalidate dir's completeness, dentry lease state on an aborted MDS
3090  * namespace request.
3091  */
3092 void ceph_invalidate_dir_request(struct ceph_mds_request *req)
3093 {
3094         struct inode *dir = req->r_parent;
3095         struct inode *old_dir = req->r_old_dentry_dir;
3096
3097         dout("invalidate_dir_request %p %p (complete, lease(s))\n", dir, old_dir);
3098
3099         ceph_dir_clear_complete(dir);
3100         if (old_dir)
3101                 ceph_dir_clear_complete(old_dir);
3102         if (req->r_dentry)
3103                 ceph_invalidate_dentry_lease(req->r_dentry);
3104         if (req->r_old_dentry)
3105                 ceph_invalidate_dentry_lease(req->r_old_dentry);
3106 }
3107
3108 /*
3109  * Handle mds reply.
3110  *
3111  * We take the session mutex and parse and process the reply immediately.
3112  * This preserves the logical ordering of replies, capabilities, etc., sent
3113  * by the MDS as they are applied to our local cache.
3114  */
3115 static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
3116 {
3117         struct ceph_mds_client *mdsc = session->s_mdsc;
3118         struct ceph_mds_request *req;
3119         struct ceph_mds_reply_head *head = msg->front.iov_base;
3120         struct ceph_mds_reply_info_parsed *rinfo;  /* parsed reply info */
3121         struct ceph_snap_realm *realm;
3122         u64 tid;
3123         int err, result;
3124         int mds = session->s_mds;
3125
3126         if (msg->front.iov_len < sizeof(*head)) {
3127                 pr_err("mdsc_handle_reply got corrupt (short) reply\n");
3128                 ceph_msg_dump(msg);
3129                 return;
3130         }
3131
3132         /* get request, session */
3133         tid = le64_to_cpu(msg->hdr.tid);
3134         mutex_lock(&mdsc->mutex);
3135         req = lookup_get_request(mdsc, tid);
3136         if (!req) {
3137                 dout("handle_reply on unknown tid %llu\n", tid);
3138                 mutex_unlock(&mdsc->mutex);
3139                 return;
3140         }
3141         dout("handle_reply %p\n", req);
3142
3143         /* correct session? */
3144         if (req->r_session != session) {
3145                 pr_err("mdsc_handle_reply got %llu on session mds%d"
3146                        " not mds%d\n", tid, session->s_mds,
3147                        req->r_session ? req->r_session->s_mds : -1);
3148                 mutex_unlock(&mdsc->mutex);
3149                 goto out;
3150         }
3151
3152         /* dup? */
3153         if ((test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags) && !head->safe) ||
3154             (test_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags) && head->safe)) {
3155                 pr_warn("got a dup %s reply on %llu from mds%d\n",
3156                            head->safe ? "safe" : "unsafe", tid, mds);
3157                 mutex_unlock(&mdsc->mutex);
3158                 goto out;
3159         }
3160         if (test_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags)) {
3161                 pr_warn("got unsafe after safe on %llu from mds%d\n",
3162                            tid, mds);
3163                 mutex_unlock(&mdsc->mutex);
3164                 goto out;
3165         }
3166
3167         result = le32_to_cpu(head->result);
3168
3169         /*
3170          * Handle an ESTALE
3171          * if we're not talking to the authority, send to them
3172          * if the authority has changed while we weren't looking,
3173          * send to new authority
3174          * Otherwise we just have to return an ESTALE
3175          */
3176         if (result == -ESTALE) {
3177                 dout("got ESTALE on request %llu\n", req->r_tid);
3178                 req->r_resend_mds = -1;
3179                 if (req->r_direct_mode != USE_AUTH_MDS) {
3180                         dout("not using auth, setting for that now\n");
3181                         req->r_direct_mode = USE_AUTH_MDS;
3182                         __do_request(mdsc, req);
3183                         mutex_unlock(&mdsc->mutex);
3184                         goto out;
3185                 } else  {
3186                         int mds = __choose_mds(mdsc, req, NULL);
3187                         if (mds >= 0 && mds != req->r_session->s_mds) {
3188                                 dout("but auth changed, so resending\n");
3189                                 __do_request(mdsc, req);
3190                                 mutex_unlock(&mdsc->mutex);
3191                                 goto out;
3192                         }
3193                 }
3194                 dout("have to return ESTALE on request %llu\n", req->r_tid);
3195         }
3196
3197
3198         if (head->safe) {
3199                 set_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags);
3200                 __unregister_request(mdsc, req);
3201
3202                 /* last request during umount? */
3203                 if (mdsc->stopping && !__get_oldest_req(mdsc))
3204                         complete_all(&mdsc->safe_umount_waiters);
3205
3206                 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
3207                         /*
3208                          * We already handled the unsafe response, now do the
3209                          * cleanup.  No need to examine the response; the MDS
3210                          * doesn't include any result info in the safe
3211                          * response.  And even if it did, there is nothing
3212                          * useful we could do with a revised return value.
3213                          */
3214                         dout("got safe reply %llu, mds%d\n", tid, mds);
3215
3216                         mutex_unlock(&mdsc->mutex);
3217                         goto out;
3218                 }
3219         } else {
3220                 set_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags);
3221                 list_add_tail(&req->r_unsafe_item, &req->r_session->s_unsafe);
3222         }
3223
3224         dout("handle_reply tid %lld result %d\n", tid, result);
3225         rinfo = &req->r_reply_info;
3226         if (test_bit(CEPHFS_FEATURE_REPLY_ENCODING, &session->s_features))
3227                 err = parse_reply_info(session, msg, rinfo, (u64)-1);
3228         else
3229                 err = parse_reply_info(session, msg, rinfo, session->s_con.peer_features);
3230         mutex_unlock(&mdsc->mutex);
3231
3232         /* Must find target inode outside of mutexes to avoid deadlocks */
3233         if ((err >= 0) && rinfo->head->is_target) {
3234                 struct inode *in;
3235                 struct ceph_vino tvino = {
3236                         .ino  = le64_to_cpu(rinfo->targeti.in->ino),
3237                         .snap = le64_to_cpu(rinfo->targeti.in->snapid)
3238                 };
3239
3240                 in = ceph_get_inode(mdsc->fsc->sb, tvino);
3241                 if (IS_ERR(in)) {
3242                         err = PTR_ERR(in);
3243                         mutex_lock(&session->s_mutex);
3244                         goto out_err;
3245                 }
3246                 req->r_target_inode = in;
3247         }
3248
3249         mutex_lock(&session->s_mutex);
3250         if (err < 0) {
3251                 pr_err("mdsc_handle_reply got corrupt reply mds%d(tid:%lld)\n", mds, tid);
3252                 ceph_msg_dump(msg);
3253                 goto out_err;
3254         }
3255
3256         /* snap trace */
3257         realm = NULL;
3258         if (rinfo->snapblob_len) {
3259                 down_write(&mdsc->snap_rwsem);
3260                 ceph_update_snap_trace(mdsc, rinfo->snapblob,
3261                                 rinfo->snapblob + rinfo->snapblob_len,
3262                                 le32_to_cpu(head->op) == CEPH_MDS_OP_RMSNAP,
3263                                 &realm);
3264                 downgrade_write(&mdsc->snap_rwsem);
3265         } else {
3266                 down_read(&mdsc->snap_rwsem);
3267         }
3268
3269         /* insert trace into our cache */
3270         mutex_lock(&req->r_fill_mutex);
3271         current->journal_info = req;
3272         err = ceph_fill_trace(mdsc->fsc->sb, req);
3273         if (err == 0) {
3274                 if (result == 0 && (req->r_op == CEPH_MDS_OP_READDIR ||
3275                                     req->r_op == CEPH_MDS_OP_LSSNAP))
3276                         ceph_readdir_prepopulate(req, req->r_session);
3277         }
3278         current->journal_info = NULL;
3279         mutex_unlock(&req->r_fill_mutex);
3280
3281         up_read(&mdsc->snap_rwsem);
3282         if (realm)
3283                 ceph_put_snap_realm(mdsc, realm);
3284
3285         if (err == 0) {
3286                 if (req->r_target_inode &&
3287                     test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
3288                         struct ceph_inode_info *ci =
3289                                 ceph_inode(req->r_target_inode);
3290                         spin_lock(&ci->i_unsafe_lock);
3291                         list_add_tail(&req->r_unsafe_target_item,
3292                                       &ci->i_unsafe_iops);
3293                         spin_unlock(&ci->i_unsafe_lock);
3294                 }
3295
3296                 ceph_unreserve_caps(mdsc, &req->r_caps_reservation);
3297         }
3298 out_err:
3299         mutex_lock(&mdsc->mutex);
3300         if (!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
3301                 if (err) {
3302                         req->r_err = err;
3303                 } else {
3304                         req->r_reply =  ceph_msg_get(msg);
3305                         set_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags);
3306                 }
3307         } else {
3308                 dout("reply arrived after request %lld was aborted\n", tid);
3309         }
3310         mutex_unlock(&mdsc->mutex);
3311
3312         mutex_unlock(&session->s_mutex);
3313
3314         /* kick calling process */
3315         complete_request(mdsc, req);
3316
3317         ceph_update_metadata_metrics(&mdsc->metric, req->r_start_latency,
3318                                      req->r_end_latency, err);
3319 out:
3320         ceph_mdsc_put_request(req);
3321         return;
3322 }
3323
3324
3325
3326 /*
3327  * handle mds notification that our request has been forwarded.
3328  */
3329 static void handle_forward(struct ceph_mds_client *mdsc,
3330                            struct ceph_mds_session *session,
3331                            struct ceph_msg *msg)
3332 {
3333         struct ceph_mds_request *req;
3334         u64 tid = le64_to_cpu(msg->hdr.tid);
3335         u32 next_mds;
3336         u32 fwd_seq;
3337         int err = -EINVAL;
3338         void *p = msg->front.iov_base;
3339         void *end = p + msg->front.iov_len;
3340
3341         ceph_decode_need(&p, end, 2*sizeof(u32), bad);
3342         next_mds = ceph_decode_32(&p);
3343         fwd_seq = ceph_decode_32(&p);
3344
3345         mutex_lock(&mdsc->mutex);
3346         req = lookup_get_request(mdsc, tid);
3347         if (!req) {
3348                 dout("forward tid %llu to mds%d - req dne\n", tid, next_mds);
3349                 goto out;  /* dup reply? */
3350         }
3351
3352         if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
3353                 dout("forward tid %llu aborted, unregistering\n", tid);
3354                 __unregister_request(mdsc, req);
3355         } else if (fwd_seq <= req->r_num_fwd) {
3356                 dout("forward tid %llu to mds%d - old seq %d <= %d\n",
3357                      tid, next_mds, req->r_num_fwd, fwd_seq);
3358         } else {
3359                 /* resend. forward race not possible; mds would drop */
3360                 dout("forward tid %llu to mds%d (we resend)\n", tid, next_mds);
3361                 BUG_ON(req->r_err);
3362                 BUG_ON(test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags));
3363                 req->r_attempts = 0;
3364                 req->r_num_fwd = fwd_seq;
3365                 req->r_resend_mds = next_mds;
3366                 put_request_session(req);
3367                 __do_request(mdsc, req);
3368         }
3369         ceph_mdsc_put_request(req);
3370 out:
3371         mutex_unlock(&mdsc->mutex);
3372         return;
3373
3374 bad:
3375         pr_err("mdsc_handle_forward decode error err=%d\n", err);
3376 }
3377
3378 static int __decode_session_metadata(void **p, void *end,
3379                                      bool *blocklisted)
3380 {
3381         /* map<string,string> */
3382         u32 n;
3383         bool err_str;
3384         ceph_decode_32_safe(p, end, n, bad);
3385         while (n-- > 0) {
3386                 u32 len;
3387                 ceph_decode_32_safe(p, end, len, bad);
3388                 ceph_decode_need(p, end, len, bad);
3389                 err_str = !strncmp(*p, "error_string", len);
3390                 *p += len;
3391                 ceph_decode_32_safe(p, end, len, bad);
3392                 ceph_decode_need(p, end, len, bad);
3393                 /*
3394                  * Match "blocklisted (blacklisted)" from newer MDSes,
3395                  * or "blacklisted" from older MDSes.
3396                  */
3397                 if (err_str && strnstr(*p, "blacklisted", len))
3398                         *blocklisted = true;
3399                 *p += len;
3400         }
3401         return 0;
3402 bad:
3403         return -1;
3404 }
3405
3406 /*
3407  * handle a mds session control message
3408  */
3409 static void handle_session(struct ceph_mds_session *session,
3410                            struct ceph_msg *msg)
3411 {
3412         struct ceph_mds_client *mdsc = session->s_mdsc;
3413         int mds = session->s_mds;
3414         int msg_version = le16_to_cpu(msg->hdr.version);
3415         void *p = msg->front.iov_base;
3416         void *end = p + msg->front.iov_len;
3417         struct ceph_mds_session_head *h;
3418         u32 op;
3419         u64 seq, features = 0;
3420         int wake = 0;
3421         bool blocklisted = false;
3422
3423         /* decode */
3424         ceph_decode_need(&p, end, sizeof(*h), bad);
3425         h = p;
3426         p += sizeof(*h);
3427
3428         op = le32_to_cpu(h->op);
3429         seq = le64_to_cpu(h->seq);
3430
3431         if (msg_version >= 3) {
3432                 u32 len;
3433                 /* version >= 2, metadata */
3434                 if (__decode_session_metadata(&p, end, &blocklisted) < 0)
3435                         goto bad;
3436                 /* version >= 3, feature bits */
3437                 ceph_decode_32_safe(&p, end, len, bad);
3438                 if (len) {
3439                         ceph_decode_64_safe(&p, end, features, bad);
3440                         p += len - sizeof(features);
3441                 }
3442         }
3443
3444         mutex_lock(&mdsc->mutex);
3445         if (op == CEPH_SESSION_CLOSE) {
3446                 ceph_get_mds_session(session);
3447                 __unregister_session(mdsc, session);
3448         }
3449         /* FIXME: this ttl calculation is generous */
3450         session->s_ttl = jiffies + HZ*mdsc->mdsmap->m_session_autoclose;
3451         mutex_unlock(&mdsc->mutex);
3452
3453         mutex_lock(&session->s_mutex);
3454
3455         dout("handle_session mds%d %s %p state %s seq %llu\n",
3456              mds, ceph_session_op_name(op), session,
3457              ceph_session_state_name(session->s_state), seq);
3458
3459         if (session->s_state == CEPH_MDS_SESSION_HUNG) {
3460                 session->s_state = CEPH_MDS_SESSION_OPEN;
3461                 pr_info("mds%d came back\n", session->s_mds);
3462         }
3463
3464         switch (op) {
3465         case CEPH_SESSION_OPEN:
3466                 if (session->s_state == CEPH_MDS_SESSION_RECONNECTING)
3467                         pr_info("mds%d reconnect success\n", session->s_mds);
3468                 session->s_state = CEPH_MDS_SESSION_OPEN;
3469                 session->s_features = features;
3470                 renewed_caps(mdsc, session, 0);
3471                 if (test_bit(CEPHFS_FEATURE_METRIC_COLLECT, &session->s_features))
3472                         metric_schedule_delayed(&mdsc->metric);
3473                 wake = 1;
3474                 if (mdsc->stopping)
3475                         __close_session(mdsc, session);
3476                 break;
3477
3478         case CEPH_SESSION_RENEWCAPS:
3479                 if (session->s_renew_seq == seq)
3480                         renewed_caps(mdsc, session, 1);
3481                 break;
3482
3483         case CEPH_SESSION_CLOSE:
3484                 if (session->s_state == CEPH_MDS_SESSION_RECONNECTING)
3485                         pr_info("mds%d reconnect denied\n", session->s_mds);
3486                 session->s_state = CEPH_MDS_SESSION_CLOSED;
3487                 cleanup_session_requests(mdsc, session);
3488                 remove_session_caps(session);
3489                 wake = 2; /* for good measure */
3490                 wake_up_all(&mdsc->session_close_wq);
3491                 break;
3492
3493         case CEPH_SESSION_STALE:
3494                 pr_info("mds%d caps went stale, renewing\n",
3495                         session->s_mds);
3496                 atomic_inc(&session->s_cap_gen);
3497                 session->s_cap_ttl = jiffies - 1;
3498                 send_renew_caps(mdsc, session);
3499                 break;
3500
3501         case CEPH_SESSION_RECALL_STATE:
3502                 ceph_trim_caps(mdsc, session, le32_to_cpu(h->max_caps));
3503                 break;
3504
3505         case CEPH_SESSION_FLUSHMSG:
3506                 send_flushmsg_ack(mdsc, session, seq);
3507                 break;
3508
3509         case CEPH_SESSION_FORCE_RO:
3510                 dout("force_session_readonly %p\n", session);
3511                 spin_lock(&session->s_cap_lock);
3512                 session->s_readonly = true;
3513                 spin_unlock(&session->s_cap_lock);
3514                 wake_up_session_caps(session, FORCE_RO);
3515                 break;
3516
3517         case CEPH_SESSION_REJECT:
3518                 WARN_ON(session->s_state != CEPH_MDS_SESSION_OPENING);
3519                 pr_info("mds%d rejected session\n", session->s_mds);
3520                 session->s_state = CEPH_MDS_SESSION_REJECTED;
3521                 cleanup_session_requests(mdsc, session);
3522                 remove_session_caps(session);
3523                 if (blocklisted)
3524                         mdsc->fsc->blocklisted = true;
3525                 wake = 2; /* for good measure */
3526                 break;
3527
3528         default:
3529                 pr_err("mdsc_handle_session bad op %d mds%d\n", op, mds);
3530                 WARN_ON(1);
3531         }
3532
3533         mutex_unlock(&session->s_mutex);
3534         if (wake) {
3535                 mutex_lock(&mdsc->mutex);
3536                 __wake_requests(mdsc, &session->s_waiting);
3537                 if (wake == 2)
3538                         kick_requests(mdsc, mds);
3539                 mutex_unlock(&mdsc->mutex);
3540         }
3541         if (op == CEPH_SESSION_CLOSE)
3542                 ceph_put_mds_session(session);
3543         return;
3544
3545 bad:
3546         pr_err("mdsc_handle_session corrupt message mds%d len %d\n", mds,
3547                (int)msg->front.iov_len);
3548         ceph_msg_dump(msg);
3549         return;
3550 }
3551
3552 void ceph_mdsc_release_dir_caps(struct ceph_mds_request *req)
3553 {
3554         int dcaps;
3555
3556         dcaps = xchg(&req->r_dir_caps, 0);
3557         if (dcaps) {
3558                 dout("releasing r_dir_caps=%s\n", ceph_cap_string(dcaps));
3559                 ceph_put_cap_refs(ceph_inode(req->r_parent), dcaps);
3560         }
3561 }
3562
3563 void ceph_mdsc_release_dir_caps_no_check(struct ceph_mds_request *req)
3564 {
3565         int dcaps;
3566
3567         dcaps = xchg(&req->r_dir_caps, 0);
3568         if (dcaps) {
3569                 dout("releasing r_dir_caps=%s\n", ceph_cap_string(dcaps));
3570                 ceph_put_cap_refs_no_check_caps(ceph_inode(req->r_parent),
3571                                                 dcaps);
3572         }
3573 }
3574
3575 /*
3576  * called under session->mutex.
3577  */
3578 static void replay_unsafe_requests(struct ceph_mds_client *mdsc,
3579                                    struct ceph_mds_session *session)
3580 {
3581         struct ceph_mds_request *req, *nreq;
3582         struct rb_node *p;
3583
3584         dout("replay_unsafe_requests mds%d\n", session->s_mds);
3585
3586         mutex_lock(&mdsc->mutex);
3587         list_for_each_entry_safe(req, nreq, &session->s_unsafe, r_unsafe_item)
3588                 __send_request(session, req, true);
3589
3590         /*
3591          * also re-send old requests when MDS enters reconnect stage. So that MDS
3592          * can process completed request in clientreplay stage.
3593          */
3594         p = rb_first(&mdsc->request_tree);
3595         while (p) {
3596                 req = rb_entry(p, struct ceph_mds_request, r_node);
3597                 p = rb_next(p);
3598                 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
3599                         continue;
3600                 if (req->r_attempts == 0)
3601                         continue; /* only old requests */
3602                 if (!req->r_session)
3603                         continue;
3604                 if (req->r_session->s_mds != session->s_mds)
3605                         continue;
3606
3607                 ceph_mdsc_release_dir_caps_no_check(req);
3608
3609                 __send_request(session, req, true);
3610         }
3611         mutex_unlock(&mdsc->mutex);
3612 }
3613
3614 static int send_reconnect_partial(struct ceph_reconnect_state *recon_state)
3615 {
3616         struct ceph_msg *reply;
3617         struct ceph_pagelist *_pagelist;
3618         struct page *page;
3619         __le32 *addr;
3620         int err = -ENOMEM;
3621
3622         if (!recon_state->allow_multi)
3623                 return -ENOSPC;
3624
3625         /* can't handle message that contains both caps and realm */
3626         BUG_ON(!recon_state->nr_caps == !recon_state->nr_realms);
3627
3628         /* pre-allocate new pagelist */
3629         _pagelist = ceph_pagelist_alloc(GFP_NOFS);
3630         if (!_pagelist)
3631                 return -ENOMEM;
3632
3633         reply = ceph_msg_new2(CEPH_MSG_CLIENT_RECONNECT, 0, 1, GFP_NOFS, false);
3634         if (!reply)
3635                 goto fail_msg;
3636
3637         /* placeholder for nr_caps */
3638         err = ceph_pagelist_encode_32(_pagelist, 0);
3639         if (err < 0)
3640                 goto fail;
3641
3642         if (recon_state->nr_caps) {
3643                 /* currently encoding caps */
3644                 err = ceph_pagelist_encode_32(recon_state->pagelist, 0);
3645                 if (err)
3646                         goto fail;
3647         } else {
3648                 /* placeholder for nr_realms (currently encoding relams) */
3649                 err = ceph_pagelist_encode_32(_pagelist, 0);
3650                 if (err < 0)
3651                         goto fail;
3652         }
3653
3654         err = ceph_pagelist_encode_8(recon_state->pagelist, 1);
3655         if (err)
3656                 goto fail;
3657
3658         page = list_first_entry(&recon_state->pagelist->head, struct page, lru);
3659         addr = kmap_atomic(page);
3660         if (recon_state->nr_caps) {
3661                 /* currently encoding caps */
3662                 *addr = cpu_to_le32(recon_state->nr_caps);
3663         } else {
3664                 /* currently encoding relams */
3665                 *(addr + 1) = cpu_to_le32(recon_state->nr_realms);
3666         }
3667         kunmap_atomic(addr);
3668
3669         reply->hdr.version = cpu_to_le16(5);
3670         reply->hdr.compat_version = cpu_to_le16(4);
3671
3672         reply->hdr.data_len = cpu_to_le32(recon_state->pagelist->length);
3673         ceph_msg_data_add_pagelist(reply, recon_state->pagelist);
3674
3675         ceph_con_send(&recon_state->session->s_con, reply);
3676         ceph_pagelist_release(recon_state->pagelist);
3677
3678         recon_state->pagelist = _pagelist;
3679         recon_state->nr_caps = 0;
3680         recon_state->nr_realms = 0;
3681         recon_state->msg_version = 5;
3682         return 0;
3683 fail:
3684         ceph_msg_put(reply);
3685 fail_msg:
3686         ceph_pagelist_release(_pagelist);
3687         return err;
3688 }
3689
3690 static struct dentry* d_find_primary(struct inode *inode)
3691 {
3692         struct dentry *alias, *dn = NULL;
3693
3694         if (hlist_empty(&inode->i_dentry))
3695                 return NULL;
3696
3697         spin_lock(&inode->i_lock);
3698         if (hlist_empty(&inode->i_dentry))
3699                 goto out_unlock;
3700
3701         if (S_ISDIR(inode->i_mode)) {
3702                 alias = hlist_entry(inode->i_dentry.first, struct dentry, d_u.d_alias);
3703                 if (!IS_ROOT(alias))
3704                         dn = dget(alias);
3705                 goto out_unlock;
3706         }
3707
3708         hlist_for_each_entry(alias, &inode->i_dentry, d_u.d_alias) {
3709                 spin_lock(&alias->d_lock);
3710                 if (!d_unhashed(alias) &&
3711                     (ceph_dentry(alias)->flags & CEPH_DENTRY_PRIMARY_LINK)) {
3712                         dn = dget_dlock(alias);
3713                 }
3714                 spin_unlock(&alias->d_lock);
3715                 if (dn)
3716                         break;
3717         }
3718 out_unlock:
3719         spin_unlock(&inode->i_lock);
3720         return dn;
3721 }
3722
3723 /*
3724  * Encode information about a cap for a reconnect with the MDS.
3725  */
3726 static int reconnect_caps_cb(struct inode *inode, struct ceph_cap *cap,
3727                           void *arg)
3728 {
3729         union {
3730                 struct ceph_mds_cap_reconnect v2;
3731                 struct ceph_mds_cap_reconnect_v1 v1;
3732         } rec;
3733         struct ceph_inode_info *ci = cap->ci;
3734         struct ceph_reconnect_state *recon_state = arg;
3735         struct ceph_pagelist *pagelist = recon_state->pagelist;
3736         struct dentry *dentry;
3737         char *path;
3738         int pathlen, err;
3739         u64 pathbase;
3740         u64 snap_follows;
3741
3742         dout(" adding %p ino %llx.%llx cap %p %lld %s\n",
3743              inode, ceph_vinop(inode), cap, cap->cap_id,
3744              ceph_cap_string(cap->issued));
3745
3746         dentry = d_find_primary(inode);
3747         if (dentry) {
3748                 /* set pathbase to parent dir when msg_version >= 2 */
3749                 path = ceph_mdsc_build_path(dentry, &pathlen, &pathbase,
3750                                             recon_state->msg_version >= 2);
3751                 dput(dentry);
3752                 if (IS_ERR(path)) {
3753                         err = PTR_ERR(path);
3754                         goto out_err;
3755                 }
3756         } else {
3757                 path = NULL;
3758                 pathlen = 0;
3759                 pathbase = 0;
3760         }
3761
3762         spin_lock(&ci->i_ceph_lock);
3763         cap->seq = 0;        /* reset cap seq */
3764         cap->issue_seq = 0;  /* and issue_seq */
3765         cap->mseq = 0;       /* and migrate_seq */
3766         cap->cap_gen = atomic_read(&cap->session->s_cap_gen);
3767
3768         /* These are lost when the session goes away */
3769         if (S_ISDIR(inode->i_mode)) {
3770                 if (cap->issued & CEPH_CAP_DIR_CREATE) {
3771                         ceph_put_string(rcu_dereference_raw(ci->i_cached_layout.pool_ns));
3772                         memset(&ci->i_cached_layout, 0, sizeof(ci->i_cached_layout));
3773                 }
3774                 cap->issued &= ~CEPH_CAP_ANY_DIR_OPS;
3775         }
3776
3777         if (recon_state->msg_version >= 2) {
3778                 rec.v2.cap_id = cpu_to_le64(cap->cap_id);
3779                 rec.v2.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
3780                 rec.v2.issued = cpu_to_le32(cap->issued);
3781                 rec.v2.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
3782                 rec.v2.pathbase = cpu_to_le64(pathbase);
3783                 rec.v2.flock_len = (__force __le32)
3784                         ((ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) ? 0 : 1);
3785         } else {
3786                 rec.v1.cap_id = cpu_to_le64(cap->cap_id);
3787                 rec.v1.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
3788                 rec.v1.issued = cpu_to_le32(cap->issued);
3789                 rec.v1.size = cpu_to_le64(i_size_read(inode));
3790                 ceph_encode_timespec64(&rec.v1.mtime, &inode->i_mtime);
3791                 ceph_encode_timespec64(&rec.v1.atime, &inode->i_atime);
3792                 rec.v1.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
3793                 rec.v1.pathbase = cpu_to_le64(pathbase);
3794         }
3795
3796         if (list_empty(&ci->i_cap_snaps)) {
3797                 snap_follows = ci->i_head_snapc ? ci->i_head_snapc->seq : 0;
3798         } else {
3799                 struct ceph_cap_snap *capsnap =
3800                         list_first_entry(&ci->i_cap_snaps,
3801                                          struct ceph_cap_snap, ci_item);
3802                 snap_follows = capsnap->follows;
3803         }
3804         spin_unlock(&ci->i_ceph_lock);
3805
3806         if (recon_state->msg_version >= 2) {
3807                 int num_fcntl_locks, num_flock_locks;
3808                 struct ceph_filelock *flocks = NULL;
3809                 size_t struct_len, total_len = sizeof(u64);
3810                 u8 struct_v = 0;
3811
3812 encode_again:
3813                 if (rec.v2.flock_len) {
3814                         ceph_count_locks(inode, &num_fcntl_locks, &num_flock_locks);
3815                 } else {
3816                         num_fcntl_locks = 0;
3817                         num_flock_locks = 0;
3818                 }
3819                 if (num_fcntl_locks + num_flock_locks > 0) {
3820                         flocks = kmalloc_array(num_fcntl_locks + num_flock_locks,
3821                                                sizeof(struct ceph_filelock),
3822                                                GFP_NOFS);
3823                         if (!flocks) {
3824                                 err = -ENOMEM;
3825                                 goto out_err;
3826                         }
3827                         err = ceph_encode_locks_to_buffer(inode, flocks,
3828                                                           num_fcntl_locks,
3829                                                           num_flock_locks);
3830                         if (err) {
3831                                 kfree(flocks);
3832                                 flocks = NULL;
3833                                 if (err == -ENOSPC)
3834                                         goto encode_again;
3835                                 goto out_err;
3836                         }
3837                 } else {
3838                         kfree(flocks);
3839                         flocks = NULL;
3840                 }
3841
3842                 if (recon_state->msg_version >= 3) {
3843                         /* version, compat_version and struct_len */
3844                         total_len += 2 * sizeof(u8) + sizeof(u32);
3845                         struct_v = 2;
3846                 }
3847                 /*
3848                  * number of encoded locks is stable, so copy to pagelist
3849                  */
3850                 struct_len = 2 * sizeof(u32) +
3851                             (num_fcntl_locks + num_flock_locks) *
3852                             sizeof(struct ceph_filelock);
3853                 rec.v2.flock_len = cpu_to_le32(struct_len);
3854
3855                 struct_len += sizeof(u32) + pathlen + sizeof(rec.v2);
3856
3857                 if (struct_v >= 2)
3858                         struct_len += sizeof(u64); /* snap_follows */
3859
3860                 total_len += struct_len;
3861
3862                 if (pagelist->length + total_len > RECONNECT_MAX_SIZE) {
3863                         err = send_reconnect_partial(recon_state);
3864                         if (err)
3865                                 goto out_freeflocks;
3866                         pagelist = recon_state->pagelist;
3867                 }
3868
3869                 err = ceph_pagelist_reserve(pagelist, total_len);
3870                 if (err)
3871                         goto out_freeflocks;
3872
3873                 ceph_pagelist_encode_64(pagelist, ceph_ino(inode));
3874                 if (recon_state->msg_version >= 3) {
3875                         ceph_pagelist_encode_8(pagelist, struct_v);
3876                         ceph_pagelist_encode_8(pagelist, 1);
3877                         ceph_pagelist_encode_32(pagelist, struct_len);
3878                 }
3879                 ceph_pagelist_encode_string(pagelist, path, pathlen);
3880                 ceph_pagelist_append(pagelist, &rec, sizeof(rec.v2));
3881                 ceph_locks_to_pagelist(flocks, pagelist,
3882                                        num_fcntl_locks, num_flock_locks);
3883                 if (struct_v >= 2)
3884                         ceph_pagelist_encode_64(pagelist, snap_follows);
3885 out_freeflocks:
3886                 kfree(flocks);
3887         } else {
3888                 err = ceph_pagelist_reserve(pagelist,
3889                                             sizeof(u64) + sizeof(u32) +
3890                                             pathlen + sizeof(rec.v1));
3891                 if (err)
3892                         goto out_err;
3893
3894                 ceph_pagelist_encode_64(pagelist, ceph_ino(inode));
3895                 ceph_pagelist_encode_string(pagelist, path, pathlen);
3896                 ceph_pagelist_append(pagelist, &rec, sizeof(rec.v1));
3897         }
3898
3899 out_err:
3900         ceph_mdsc_free_path(path, pathlen);
3901         if (!err)
3902                 recon_state->nr_caps++;
3903         return err;
3904 }
3905
3906 static int encode_snap_realms(struct ceph_mds_client *mdsc,
3907                               struct ceph_reconnect_state *recon_state)
3908 {
3909         struct rb_node *p;
3910         struct ceph_pagelist *pagelist = recon_state->pagelist;
3911         int err = 0;
3912
3913         if (recon_state->msg_version >= 4) {
3914                 err = ceph_pagelist_encode_32(pagelist, mdsc->num_snap_realms);
3915                 if (err < 0)
3916                         goto fail;
3917         }
3918
3919         /*
3920          * snaprealms.  we provide mds with the ino, seq (version), and
3921          * parent for all of our realms.  If the mds has any newer info,
3922          * it will tell us.
3923          */
3924         for (p = rb_first(&mdsc->snap_realms); p; p = rb_next(p)) {
3925                 struct ceph_snap_realm *realm =
3926                        rb_entry(p, struct ceph_snap_realm, node);
3927                 struct ceph_mds_snaprealm_reconnect sr_rec;
3928
3929                 if (recon_state->msg_version >= 4) {
3930                         size_t need = sizeof(u8) * 2 + sizeof(u32) +
3931                                       sizeof(sr_rec);
3932
3933                         if (pagelist->length + need > RECONNECT_MAX_SIZE) {
3934                                 err = send_reconnect_partial(recon_state);
3935                                 if (err)
3936                                         goto fail;
3937                                 pagelist = recon_state->pagelist;
3938                         }
3939
3940                         err = ceph_pagelist_reserve(pagelist, need);
3941                         if (err)
3942                                 goto fail;
3943
3944                         ceph_pagelist_encode_8(pagelist, 1);
3945                         ceph_pagelist_encode_8(pagelist, 1);
3946                         ceph_pagelist_encode_32(pagelist, sizeof(sr_rec));
3947                 }
3948
3949                 dout(" adding snap realm %llx seq %lld parent %llx\n",
3950                      realm->ino, realm->seq, realm->parent_ino);
3951                 sr_rec.ino = cpu_to_le64(realm->ino);
3952                 sr_rec.seq = cpu_to_le64(realm->seq);
3953                 sr_rec.parent = cpu_to_le64(realm->parent_ino);
3954
3955                 err = ceph_pagelist_append(pagelist, &sr_rec, sizeof(sr_rec));
3956                 if (err)
3957                         goto fail;
3958
3959                 recon_state->nr_realms++;
3960         }
3961 fail:
3962         return err;
3963 }
3964
3965
3966 /*
3967  * If an MDS fails and recovers, clients need to reconnect in order to
3968  * reestablish shared state.  This includes all caps issued through
3969  * this session _and_ the snap_realm hierarchy.  Because it's not
3970  * clear which snap realms the mds cares about, we send everything we
3971  * know about.. that ensures we'll then get any new info the
3972  * recovering MDS might have.
3973  *
3974  * This is a relatively heavyweight operation, but it's rare.
3975  */
3976 static void send_mds_reconnect(struct ceph_mds_client *mdsc,
3977                                struct ceph_mds_session *session)
3978 {
3979         struct ceph_msg *reply;
3980         int mds = session->s_mds;
3981         int err = -ENOMEM;
3982         struct ceph_reconnect_state recon_state = {
3983                 .session = session,
3984         };
3985         LIST_HEAD(dispose);
3986
3987         pr_info("mds%d reconnect start\n", mds);
3988
3989         recon_state.pagelist = ceph_pagelist_alloc(GFP_NOFS);
3990         if (!recon_state.pagelist)
3991                 goto fail_nopagelist;
3992
3993         reply = ceph_msg_new2(CEPH_MSG_CLIENT_RECONNECT, 0, 1, GFP_NOFS, false);
3994         if (!reply)
3995                 goto fail_nomsg;
3996
3997         xa_destroy(&session->s_delegated_inos);
3998
3999         mutex_lock(&session->s_mutex);
4000         session->s_state = CEPH_MDS_SESSION_RECONNECTING;
4001         session->s_seq = 0;
4002
4003         dout("session %p state %s\n", session,
4004              ceph_session_state_name(session->s_state));
4005
4006         atomic_inc(&session->s_cap_gen);
4007
4008         spin_lock(&session->s_cap_lock);
4009         /* don't know if session is readonly */
4010         session->s_readonly = 0;
4011         /*
4012          * notify __ceph_remove_cap() that we are composing cap reconnect.
4013          * If a cap get released before being added to the cap reconnect,
4014          * __ceph_remove_cap() should skip queuing cap release.
4015          */
4016         session->s_cap_reconnect = 1;
4017         /* drop old cap expires; we're about to reestablish that state */
4018         detach_cap_releases(session, &dispose);
4019         spin_unlock(&session->s_cap_lock);
4020         dispose_cap_releases(mdsc, &dispose);
4021
4022         /* trim unused caps to reduce MDS's cache rejoin time */
4023         if (mdsc->fsc->sb->s_root)
4024                 shrink_dcache_parent(mdsc->fsc->sb->s_root);
4025
4026         ceph_con_close(&session->s_con);
4027         ceph_con_open(&session->s_con,
4028                       CEPH_ENTITY_TYPE_MDS, mds,
4029                       ceph_mdsmap_get_addr(mdsc->mdsmap, mds));
4030
4031         /* replay unsafe requests */
4032         replay_unsafe_requests(mdsc, session);
4033
4034         ceph_early_kick_flushing_caps(mdsc, session);
4035
4036         down_read(&mdsc->snap_rwsem);
4037
4038         /* placeholder for nr_caps */
4039         err = ceph_pagelist_encode_32(recon_state.pagelist, 0);
4040         if (err)
4041                 goto fail;
4042
4043         if (test_bit(CEPHFS_FEATURE_MULTI_RECONNECT, &session->s_features)) {
4044                 recon_state.msg_version = 3;
4045                 recon_state.allow_multi = true;
4046         } else if (session->s_con.peer_features & CEPH_FEATURE_MDSENC) {
4047                 recon_state.msg_version = 3;
4048         } else {
4049                 recon_state.msg_version = 2;
4050         }
4051         /* trsaverse this session's caps */
4052         err = ceph_iterate_session_caps(session, reconnect_caps_cb, &recon_state);
4053
4054         spin_lock(&session->s_cap_lock);
4055         session->s_cap_reconnect = 0;
4056         spin_unlock(&session->s_cap_lock);
4057
4058         if (err < 0)
4059                 goto fail;
4060
4061         /* check if all realms can be encoded into current message */
4062         if (mdsc->num_snap_realms) {
4063                 size_t total_len =
4064                         recon_state.pagelist->length +
4065                         mdsc->num_snap_realms *
4066                         sizeof(struct ceph_mds_snaprealm_reconnect);
4067                 if (recon_state.msg_version >= 4) {
4068                         /* number of realms */
4069                         total_len += sizeof(u32);
4070                         /* version, compat_version and struct_len */
4071                         total_len += mdsc->num_snap_realms *
4072                                      (2 * sizeof(u8) + sizeof(u32));
4073                 }
4074                 if (total_len > RECONNECT_MAX_SIZE) {
4075                         if (!recon_state.allow_multi) {
4076                                 err = -ENOSPC;
4077                                 goto fail;
4078                         }
4079                         if (recon_state.nr_caps) {
4080                                 err = send_reconnect_partial(&recon_state);
4081                                 if (err)
4082                                         goto fail;
4083                         }
4084                         recon_state.msg_version = 5;
4085                 }
4086         }
4087
4088         err = encode_snap_realms(mdsc, &recon_state);
4089         if (err < 0)
4090                 goto fail;
4091
4092         if (recon_state.msg_version >= 5) {
4093                 err = ceph_pagelist_encode_8(recon_state.pagelist, 0);
4094                 if (err < 0)
4095                         goto fail;
4096         }
4097
4098         if (recon_state.nr_caps || recon_state.nr_realms) {
4099                 struct page *page =
4100                         list_first_entry(&recon_state.pagelist->head,
4101                                         struct page, lru);
4102                 __le32 *addr = kmap_atomic(page);
4103                 if (recon_state.nr_caps) {
4104                         WARN_ON(recon_state.nr_realms != mdsc->num_snap_realms);
4105                         *addr = cpu_to_le32(recon_state.nr_caps);
4106                 } else if (recon_state.msg_version >= 4) {
4107                         *(addr + 1) = cpu_to_le32(recon_state.nr_realms);
4108                 }
4109                 kunmap_atomic(addr);
4110         }
4111
4112         reply->hdr.version = cpu_to_le16(recon_state.msg_version);
4113         if (recon_state.msg_version >= 4)
4114                 reply->hdr.compat_version = cpu_to_le16(4);
4115
4116         reply->hdr.data_len = cpu_to_le32(recon_state.pagelist->length);
4117         ceph_msg_data_add_pagelist(reply, recon_state.pagelist);
4118
4119         ceph_con_send(&session->s_con, reply);
4120
4121         mutex_unlock(&session->s_mutex);
4122
4123         mutex_lock(&mdsc->mutex);
4124         __wake_requests(mdsc, &session->s_waiting);
4125         mutex_unlock(&mdsc->mutex);
4126
4127         up_read(&mdsc->snap_rwsem);
4128         ceph_pagelist_release(recon_state.pagelist);
4129         return;
4130
4131 fail:
4132         ceph_msg_put(reply);
4133         up_read(&mdsc->snap_rwsem);
4134         mutex_unlock(&session->s_mutex);
4135 fail_nomsg:
4136         ceph_pagelist_release(recon_state.pagelist);
4137 fail_nopagelist:
4138         pr_err("error %d preparing reconnect for mds%d\n", err, mds);
4139         return;
4140 }
4141
4142
4143 /*
4144  * compare old and new mdsmaps, kicking requests
4145  * and closing out old connections as necessary
4146  *
4147  * called under mdsc->mutex.
4148  */
4149 static void check_new_map(struct ceph_mds_client *mdsc,
4150                           struct ceph_mdsmap *newmap,
4151                           struct ceph_mdsmap *oldmap)
4152 {
4153         int i;
4154         int oldstate, newstate;
4155         struct ceph_mds_session *s;
4156
4157         dout("check_new_map new %u old %u\n",
4158              newmap->m_epoch, oldmap->m_epoch);
4159
4160         for (i = 0; i < oldmap->possible_max_rank && i < mdsc->max_sessions; i++) {
4161                 if (!mdsc->sessions[i])
4162                         continue;
4163                 s = mdsc->sessions[i];
4164                 oldstate = ceph_mdsmap_get_state(oldmap, i);
4165                 newstate = ceph_mdsmap_get_state(newmap, i);
4166
4167                 dout("check_new_map mds%d state %s%s -> %s%s (session %s)\n",
4168                      i, ceph_mds_state_name(oldstate),
4169                      ceph_mdsmap_is_laggy(oldmap, i) ? " (laggy)" : "",
4170                      ceph_mds_state_name(newstate),
4171                      ceph_mdsmap_is_laggy(newmap, i) ? " (laggy)" : "",
4172                      ceph_session_state_name(s->s_state));
4173
4174                 if (i >= newmap->possible_max_rank) {
4175                         /* force close session for stopped mds */
4176                         ceph_get_mds_session(s);
4177                         __unregister_session(mdsc, s);
4178                         __wake_requests(mdsc, &s->s_waiting);
4179                         mutex_unlock(&mdsc->mutex);
4180
4181                         mutex_lock(&s->s_mutex);
4182                         cleanup_session_requests(mdsc, s);
4183                         remove_session_caps(s);
4184                         mutex_unlock(&s->s_mutex);
4185
4186                         ceph_put_mds_session(s);
4187
4188                         mutex_lock(&mdsc->mutex);
4189                         kick_requests(mdsc, i);
4190                         continue;
4191                 }
4192
4193                 if (memcmp(ceph_mdsmap_get_addr(oldmap, i),
4194                            ceph_mdsmap_get_addr(newmap, i),
4195                            sizeof(struct ceph_entity_addr))) {
4196                         /* just close it */
4197                         mutex_unlock(&mdsc->mutex);
4198                         mutex_lock(&s->s_mutex);
4199                         mutex_lock(&mdsc->mutex);
4200                         ceph_con_close(&s->s_con);
4201                         mutex_unlock(&s->s_mutex);
4202                         s->s_state = CEPH_MDS_SESSION_RESTARTING;
4203                 } else if (oldstate == newstate) {
4204                         continue;  /* nothing new with this mds */
4205                 }
4206
4207                 /*
4208                  * send reconnect?
4209                  */
4210                 if (s->s_state == CEPH_MDS_SESSION_RESTARTING &&
4211                     newstate >= CEPH_MDS_STATE_RECONNECT) {
4212                         mutex_unlock(&mdsc->mutex);
4213                         send_mds_reconnect(mdsc, s);
4214                         mutex_lock(&mdsc->mutex);
4215                 }
4216
4217                 /*
4218                  * kick request on any mds that has gone active.
4219                  */
4220                 if (oldstate < CEPH_MDS_STATE_ACTIVE &&
4221                     newstate >= CEPH_MDS_STATE_ACTIVE) {
4222                         if (oldstate != CEPH_MDS_STATE_CREATING &&
4223                             oldstate != CEPH_MDS_STATE_STARTING)
4224                                 pr_info("mds%d recovery completed\n", s->s_mds);
4225                         kick_requests(mdsc, i);
4226                         mutex_unlock(&mdsc->mutex);
4227                         mutex_lock(&s->s_mutex);
4228                         mutex_lock(&mdsc->mutex);
4229                         ceph_kick_flushing_caps(mdsc, s);
4230                         mutex_unlock(&s->s_mutex);
4231                         wake_up_session_caps(s, RECONNECT);
4232                 }
4233         }
4234
4235         for (i = 0; i < newmap->possible_max_rank && i < mdsc->max_sessions; i++) {
4236                 s = mdsc->sessions[i];
4237                 if (!s)
4238                         continue;
4239                 if (!ceph_mdsmap_is_laggy(newmap, i))
4240                         continue;
4241                 if (s->s_state == CEPH_MDS_SESSION_OPEN ||
4242                     s->s_state == CEPH_MDS_SESSION_HUNG ||
4243                     s->s_state == CEPH_MDS_SESSION_CLOSING) {
4244                         dout(" connecting to export targets of laggy mds%d\n",
4245                              i);
4246                         __open_export_target_sessions(mdsc, s);
4247                 }
4248         }
4249 }
4250
4251
4252
4253 /*
4254  * leases
4255  */
4256
4257 /*
4258  * caller must hold session s_mutex, dentry->d_lock
4259  */
4260 void __ceph_mdsc_drop_dentry_lease(struct dentry *dentry)
4261 {
4262         struct ceph_dentry_info *di = ceph_dentry(dentry);
4263
4264         ceph_put_mds_session(di->lease_session);
4265         di->lease_session = NULL;
4266 }
4267
4268 static void handle_lease(struct ceph_mds_client *mdsc,
4269                          struct ceph_mds_session *session,
4270                          struct ceph_msg *msg)
4271 {
4272         struct super_block *sb = mdsc->fsc->sb;
4273         struct inode *inode;
4274         struct dentry *parent, *dentry;
4275         struct ceph_dentry_info *di;
4276         int mds = session->s_mds;
4277         struct ceph_mds_lease *h = msg->front.iov_base;
4278         u32 seq;
4279         struct ceph_vino vino;
4280         struct qstr dname;
4281         int release = 0;
4282
4283         dout("handle_lease from mds%d\n", mds);
4284
4285         /* decode */
4286         if (msg->front.iov_len < sizeof(*h) + sizeof(u32))
4287                 goto bad;
4288         vino.ino = le64_to_cpu(h->ino);
4289         vino.snap = CEPH_NOSNAP;
4290         seq = le32_to_cpu(h->seq);
4291         dname.len = get_unaligned_le32(h + 1);
4292         if (msg->front.iov_len < sizeof(*h) + sizeof(u32) + dname.len)
4293                 goto bad;
4294         dname.name = (void *)(h + 1) + sizeof(u32);
4295
4296         /* lookup inode */
4297         inode = ceph_find_inode(sb, vino);
4298         dout("handle_lease %s, ino %llx %p %.*s\n",
4299              ceph_lease_op_name(h->action), vino.ino, inode,
4300              dname.len, dname.name);
4301
4302         mutex_lock(&session->s_mutex);
4303         inc_session_sequence(session);
4304
4305         if (!inode) {
4306                 dout("handle_lease no inode %llx\n", vino.ino);
4307                 goto release;
4308         }
4309
4310         /* dentry */
4311         parent = d_find_alias(inode);
4312         if (!parent) {
4313                 dout("no parent dentry on inode %p\n", inode);
4314                 WARN_ON(1);
4315                 goto release;  /* hrm... */
4316         }
4317         dname.hash = full_name_hash(parent, dname.name, dname.len);
4318         dentry = d_lookup(parent, &dname);
4319         dput(parent);
4320         if (!dentry)
4321                 goto release;
4322
4323         spin_lock(&dentry->d_lock);
4324         di = ceph_dentry(dentry);
4325         switch (h->action) {
4326         case CEPH_MDS_LEASE_REVOKE:
4327                 if (di->lease_session == session) {
4328                         if (ceph_seq_cmp(di->lease_seq, seq) > 0)
4329                                 h->seq = cpu_to_le32(di->lease_seq);
4330                         __ceph_mdsc_drop_dentry_lease(dentry);
4331                 }
4332                 release = 1;
4333                 break;
4334
4335         case CEPH_MDS_LEASE_RENEW:
4336                 if (di->lease_session == session &&
4337                     di->lease_gen == atomic_read(&session->s_cap_gen) &&
4338                     di->lease_renew_from &&
4339                     di->lease_renew_after == 0) {
4340                         unsigned long duration =
4341                                 msecs_to_jiffies(le32_to_cpu(h->duration_ms));
4342
4343                         di->lease_seq = seq;
4344                         di->time = di->lease_renew_from + duration;
4345                         di->lease_renew_after = di->lease_renew_from +
4346                                 (duration >> 1);
4347                         di->lease_renew_from = 0;
4348                 }
4349                 break;
4350         }
4351         spin_unlock(&dentry->d_lock);
4352         dput(dentry);
4353
4354         if (!release)
4355                 goto out;
4356
4357 release:
4358         /* let's just reuse the same message */
4359         h->action = CEPH_MDS_LEASE_REVOKE_ACK;
4360         ceph_msg_get(msg);
4361         ceph_con_send(&session->s_con, msg);
4362
4363 out:
4364         mutex_unlock(&session->s_mutex);
4365         iput(inode);
4366         return;
4367
4368 bad:
4369         pr_err("corrupt lease message\n");
4370         ceph_msg_dump(msg);
4371 }
4372
4373 void ceph_mdsc_lease_send_msg(struct ceph_mds_session *session,
4374                               struct dentry *dentry, char action,
4375                               u32 seq)
4376 {
4377         struct ceph_msg *msg;
4378         struct ceph_mds_lease *lease;
4379         struct inode *dir;
4380         int len = sizeof(*lease) + sizeof(u32) + NAME_MAX;
4381
4382         dout("lease_send_msg identry %p %s to mds%d\n",
4383              dentry, ceph_lease_op_name(action), session->s_mds);
4384
4385         msg = ceph_msg_new(CEPH_MSG_CLIENT_LEASE, len, GFP_NOFS, false);
4386         if (!msg)
4387                 return;
4388         lease = msg->front.iov_base;
4389         lease->action = action;
4390         lease->seq = cpu_to_le32(seq);
4391
4392         spin_lock(&dentry->d_lock);
4393         dir = d_inode(dentry->d_parent);
4394         lease->ino = cpu_to_le64(ceph_ino(dir));
4395         lease->first = lease->last = cpu_to_le64(ceph_snap(dir));
4396
4397         put_unaligned_le32(dentry->d_name.len, lease + 1);
4398         memcpy((void *)(lease + 1) + 4,
4399                dentry->d_name.name, dentry->d_name.len);
4400         spin_unlock(&dentry->d_lock);
4401         /*
4402          * if this is a preemptive lease RELEASE, no need to
4403          * flush request stream, since the actual request will
4404          * soon follow.
4405          */
4406         msg->more_to_follow = (action == CEPH_MDS_LEASE_RELEASE);
4407
4408         ceph_con_send(&session->s_con, msg);
4409 }
4410
4411 /*
4412  * lock unlock sessions, to wait ongoing session activities
4413  */
4414 static void lock_unlock_sessions(struct ceph_mds_client *mdsc)
4415 {
4416         int i;
4417
4418         mutex_lock(&mdsc->mutex);
4419         for (i = 0; i < mdsc->max_sessions; i++) {
4420                 struct ceph_mds_session *s = __ceph_lookup_mds_session(mdsc, i);
4421                 if (!s)
4422                         continue;
4423                 mutex_unlock(&mdsc->mutex);
4424                 mutex_lock(&s->s_mutex);
4425                 mutex_unlock(&s->s_mutex);
4426                 ceph_put_mds_session(s);
4427                 mutex_lock(&mdsc->mutex);
4428         }
4429         mutex_unlock(&mdsc->mutex);
4430 }
4431
4432 static void maybe_recover_session(struct ceph_mds_client *mdsc)
4433 {
4434         struct ceph_fs_client *fsc = mdsc->fsc;
4435
4436         if (!ceph_test_mount_opt(fsc, CLEANRECOVER))
4437                 return;
4438
4439         if (READ_ONCE(fsc->mount_state) != CEPH_MOUNT_MOUNTED)
4440                 return;
4441
4442         if (!READ_ONCE(fsc->blocklisted))
4443                 return;
4444
4445         pr_info("auto reconnect after blocklisted\n");
4446         ceph_force_reconnect(fsc->sb);
4447 }
4448
4449 bool check_session_state(struct ceph_mds_session *s)
4450 {
4451         switch (s->s_state) {
4452         case CEPH_MDS_SESSION_OPEN:
4453                 if (s->s_ttl && time_after(jiffies, s->s_ttl)) {
4454                         s->s_state = CEPH_MDS_SESSION_HUNG;
4455                         pr_info("mds%d hung\n", s->s_mds);
4456                 }
4457                 break;
4458         case CEPH_MDS_SESSION_CLOSING:
4459                 /* Should never reach this when we're unmounting */
4460                 WARN_ON_ONCE(s->s_ttl);
4461                 fallthrough;
4462         case CEPH_MDS_SESSION_NEW:
4463         case CEPH_MDS_SESSION_RESTARTING:
4464         case CEPH_MDS_SESSION_CLOSED:
4465         case CEPH_MDS_SESSION_REJECTED:
4466                 return false;
4467         }
4468
4469         return true;
4470 }
4471
4472 /*
4473  * If the sequence is incremented while we're waiting on a REQUEST_CLOSE reply,
4474  * then we need to retransmit that request.
4475  */
4476 void inc_session_sequence(struct ceph_mds_session *s)
4477 {
4478         lockdep_assert_held(&s->s_mutex);
4479
4480         s->s_seq++;
4481
4482         if (s->s_state == CEPH_MDS_SESSION_CLOSING) {
4483                 int ret;
4484
4485                 dout("resending session close request for mds%d\n", s->s_mds);
4486                 ret = request_close_session(s);
4487                 if (ret < 0)
4488                         pr_err("unable to close session to mds%d: %d\n",
4489                                s->s_mds, ret);
4490         }
4491 }
4492
4493 /*
4494  * delayed work -- periodically trim expired leases, renew caps with mds.  If
4495  * the @delay parameter is set to 0 or if it's more than 5 secs, the default
4496  * workqueue delay value of 5 secs will be used.
4497  */
4498 static void schedule_delayed(struct ceph_mds_client *mdsc, unsigned long delay)
4499 {
4500         unsigned long max_delay = HZ * 5;
4501
4502         /* 5 secs default delay */
4503         if (!delay || (delay > max_delay))
4504                 delay = max_delay;
4505         schedule_delayed_work(&mdsc->delayed_work,
4506                               round_jiffies_relative(delay));
4507 }
4508
4509 static void delayed_work(struct work_struct *work)
4510 {
4511         struct ceph_mds_client *mdsc =
4512                 container_of(work, struct ceph_mds_client, delayed_work.work);
4513         unsigned long delay;
4514         int renew_interval;
4515         int renew_caps;
4516         int i;
4517
4518         dout("mdsc delayed_work\n");
4519
4520         if (mdsc->stopping)
4521                 return;
4522
4523         mutex_lock(&mdsc->mutex);
4524         renew_interval = mdsc->mdsmap->m_session_timeout >> 2;
4525         renew_caps = time_after_eq(jiffies, HZ*renew_interval +
4526                                    mdsc->last_renew_caps);
4527         if (renew_caps)
4528                 mdsc->last_renew_caps = jiffies;
4529
4530         for (i = 0; i < mdsc->max_sessions; i++) {
4531                 struct ceph_mds_session *s = __ceph_lookup_mds_session(mdsc, i);
4532                 if (!s)
4533                         continue;
4534
4535                 if (!check_session_state(s)) {
4536                         ceph_put_mds_session(s);
4537                         continue;
4538                 }
4539                 mutex_unlock(&mdsc->mutex);
4540
4541                 mutex_lock(&s->s_mutex);
4542                 if (renew_caps)
4543                         send_renew_caps(mdsc, s);
4544                 else
4545                         ceph_con_keepalive(&s->s_con);
4546                 if (s->s_state == CEPH_MDS_SESSION_OPEN ||
4547                     s->s_state == CEPH_MDS_SESSION_HUNG)
4548                         ceph_send_cap_releases(mdsc, s);
4549                 mutex_unlock(&s->s_mutex);
4550                 ceph_put_mds_session(s);
4551
4552                 mutex_lock(&mdsc->mutex);
4553         }
4554         mutex_unlock(&mdsc->mutex);
4555
4556         delay = ceph_check_delayed_caps(mdsc);
4557
4558         ceph_queue_cap_reclaim_work(mdsc);
4559
4560         ceph_trim_snapid_map(mdsc);
4561
4562         maybe_recover_session(mdsc);
4563
4564         schedule_delayed(mdsc, delay);
4565 }
4566
4567 int ceph_mdsc_init(struct ceph_fs_client *fsc)
4568
4569 {
4570         struct ceph_mds_client *mdsc;
4571         int err;
4572
4573         mdsc = kzalloc(sizeof(struct ceph_mds_client), GFP_NOFS);
4574         if (!mdsc)
4575                 return -ENOMEM;
4576         mdsc->fsc = fsc;
4577         mutex_init(&mdsc->mutex);
4578         mdsc->mdsmap = kzalloc(sizeof(*mdsc->mdsmap), GFP_NOFS);
4579         if (!mdsc->mdsmap) {
4580                 err = -ENOMEM;
4581                 goto err_mdsc;
4582         }
4583
4584         init_completion(&mdsc->safe_umount_waiters);
4585         init_waitqueue_head(&mdsc->session_close_wq);
4586         INIT_LIST_HEAD(&mdsc->waiting_for_map);
4587         mdsc->sessions = NULL;
4588         atomic_set(&mdsc->num_sessions, 0);
4589         mdsc->max_sessions = 0;
4590         mdsc->stopping = 0;
4591         atomic64_set(&mdsc->quotarealms_count, 0);
4592         mdsc->quotarealms_inodes = RB_ROOT;
4593         mutex_init(&mdsc->quotarealms_inodes_mutex);
4594         mdsc->last_snap_seq = 0;
4595         init_rwsem(&mdsc->snap_rwsem);
4596         mdsc->snap_realms = RB_ROOT;
4597         INIT_LIST_HEAD(&mdsc->snap_empty);
4598         mdsc->num_snap_realms = 0;
4599         spin_lock_init(&mdsc->snap_empty_lock);
4600         mdsc->last_tid = 0;
4601         mdsc->oldest_tid = 0;
4602         mdsc->request_tree = RB_ROOT;
4603         INIT_DELAYED_WORK(&mdsc->delayed_work, delayed_work);
4604         mdsc->last_renew_caps = jiffies;
4605         INIT_LIST_HEAD(&mdsc->cap_delay_list);
4606         INIT_LIST_HEAD(&mdsc->cap_wait_list);
4607         spin_lock_init(&mdsc->cap_delay_lock);
4608         INIT_LIST_HEAD(&mdsc->snap_flush_list);
4609         spin_lock_init(&mdsc->snap_flush_lock);
4610         mdsc->last_cap_flush_tid = 1;
4611         INIT_LIST_HEAD(&mdsc->cap_flush_list);
4612         INIT_LIST_HEAD(&mdsc->cap_dirty_migrating);
4613         mdsc->num_cap_flushing = 0;
4614         spin_lock_init(&mdsc->cap_dirty_lock);
4615         init_waitqueue_head(&mdsc->cap_flushing_wq);
4616         INIT_WORK(&mdsc->cap_reclaim_work, ceph_cap_reclaim_work);
4617         atomic_set(&mdsc->cap_reclaim_pending, 0);
4618         err = ceph_metric_init(&mdsc->metric);
4619         if (err)
4620                 goto err_mdsmap;
4621
4622         spin_lock_init(&mdsc->dentry_list_lock);
4623         INIT_LIST_HEAD(&mdsc->dentry_leases);
4624         INIT_LIST_HEAD(&mdsc->dentry_dir_leases);
4625
4626         ceph_caps_init(mdsc);
4627         ceph_adjust_caps_max_min(mdsc, fsc->mount_options);
4628
4629         spin_lock_init(&mdsc->snapid_map_lock);
4630         mdsc->snapid_map_tree = RB_ROOT;
4631         INIT_LIST_HEAD(&mdsc->snapid_map_lru);
4632
4633         init_rwsem(&mdsc->pool_perm_rwsem);
4634         mdsc->pool_perm_tree = RB_ROOT;
4635
4636         strscpy(mdsc->nodename, utsname()->nodename,
4637                 sizeof(mdsc->nodename));
4638
4639         fsc->mdsc = mdsc;
4640         return 0;
4641
4642 err_mdsmap:
4643         kfree(mdsc->mdsmap);
4644 err_mdsc:
4645         kfree(mdsc);
4646         return err;
4647 }
4648
4649 /*
4650  * Wait for safe replies on open mds requests.  If we time out, drop
4651  * all requests from the tree to avoid dangling dentry refs.
4652  */
4653 static void wait_requests(struct ceph_mds_client *mdsc)
4654 {
4655         struct ceph_options *opts = mdsc->fsc->client->options;
4656         struct ceph_mds_request *req;
4657
4658         mutex_lock(&mdsc->mutex);
4659         if (__get_oldest_req(mdsc)) {
4660                 mutex_unlock(&mdsc->mutex);
4661
4662                 dout("wait_requests waiting for requests\n");
4663                 wait_for_completion_timeout(&mdsc->safe_umount_waiters,
4664                                     ceph_timeout_jiffies(opts->mount_timeout));
4665
4666                 /* tear down remaining requests */
4667                 mutex_lock(&mdsc->mutex);
4668                 while ((req = __get_oldest_req(mdsc))) {
4669                         dout("wait_requests timed out on tid %llu\n",
4670                              req->r_tid);
4671                         list_del_init(&req->r_wait);
4672                         __unregister_request(mdsc, req);
4673                 }
4674         }
4675         mutex_unlock(&mdsc->mutex);
4676         dout("wait_requests done\n");
4677 }
4678
4679 /*
4680  * called before mount is ro, and before dentries are torn down.
4681  * (hmm, does this still race with new lookups?)
4682  */
4683 void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc)
4684 {
4685         dout("pre_umount\n");
4686         mdsc->stopping = 1;
4687
4688         lock_unlock_sessions(mdsc);
4689         ceph_flush_dirty_caps(mdsc);
4690         wait_requests(mdsc);
4691
4692         /*
4693          * wait for reply handlers to drop their request refs and
4694          * their inode/dcache refs
4695          */
4696         ceph_msgr_flush();
4697
4698         ceph_cleanup_quotarealms_inodes(mdsc);
4699 }
4700
4701 /*
4702  * wait for all write mds requests to flush.
4703  */
4704 static void wait_unsafe_requests(struct ceph_mds_client *mdsc, u64 want_tid)
4705 {
4706         struct ceph_mds_request *req = NULL, *nextreq;
4707         struct rb_node *n;
4708
4709         mutex_lock(&mdsc->mutex);
4710         dout("wait_unsafe_requests want %lld\n", want_tid);
4711 restart:
4712         req = __get_oldest_req(mdsc);
4713         while (req && req->r_tid <= want_tid) {
4714                 /* find next request */
4715                 n = rb_next(&req->r_node);
4716                 if (n)
4717                         nextreq = rb_entry(n, struct ceph_mds_request, r_node);
4718                 else
4719                         nextreq = NULL;
4720                 if (req->r_op != CEPH_MDS_OP_SETFILELOCK &&
4721                     (req->r_op & CEPH_MDS_OP_WRITE)) {
4722                         /* write op */
4723                         ceph_mdsc_get_request(req);
4724                         if (nextreq)
4725                                 ceph_mdsc_get_request(nextreq);
4726                         mutex_unlock(&mdsc->mutex);
4727                         dout("wait_unsafe_requests  wait on %llu (want %llu)\n",
4728                              req->r_tid, want_tid);
4729                         wait_for_completion(&req->r_safe_completion);
4730                         mutex_lock(&mdsc->mutex);
4731                         ceph_mdsc_put_request(req);
4732                         if (!nextreq)
4733                                 break;  /* next dne before, so we're done! */
4734                         if (RB_EMPTY_NODE(&nextreq->r_node)) {
4735                                 /* next request was removed from tree */
4736                                 ceph_mdsc_put_request(nextreq);
4737                                 goto restart;
4738                         }
4739                         ceph_mdsc_put_request(nextreq);  /* won't go away */
4740                 }
4741                 req = nextreq;
4742         }
4743         mutex_unlock(&mdsc->mutex);
4744         dout("wait_unsafe_requests done\n");
4745 }
4746
4747 void ceph_mdsc_sync(struct ceph_mds_client *mdsc)
4748 {
4749         u64 want_tid, want_flush;
4750
4751         if (READ_ONCE(mdsc->fsc->mount_state) >= CEPH_MOUNT_SHUTDOWN)
4752                 return;
4753
4754         dout("sync\n");
4755         mutex_lock(&mdsc->mutex);
4756         want_tid = mdsc->last_tid;
4757         mutex_unlock(&mdsc->mutex);
4758
4759         ceph_flush_dirty_caps(mdsc);
4760         spin_lock(&mdsc->cap_dirty_lock);
4761         want_flush = mdsc->last_cap_flush_tid;
4762         if (!list_empty(&mdsc->cap_flush_list)) {
4763                 struct ceph_cap_flush *cf =
4764                         list_last_entry(&mdsc->cap_flush_list,
4765                                         struct ceph_cap_flush, g_list);
4766                 cf->wake = true;
4767         }
4768         spin_unlock(&mdsc->cap_dirty_lock);
4769
4770         dout("sync want tid %lld flush_seq %lld\n",
4771              want_tid, want_flush);
4772
4773         wait_unsafe_requests(mdsc, want_tid);
4774         wait_caps_flush(mdsc, want_flush);
4775 }
4776
4777 /*
4778  * true if all sessions are closed, or we force unmount
4779  */
4780 static bool done_closing_sessions(struct ceph_mds_client *mdsc, int skipped)
4781 {
4782         if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN)
4783                 return true;
4784         return atomic_read(&mdsc->num_sessions) <= skipped;
4785 }
4786
4787 /*
4788  * called after sb is ro.
4789  */
4790 void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc)
4791 {
4792         struct ceph_options *opts = mdsc->fsc->client->options;
4793         struct ceph_mds_session *session;
4794         int i;
4795         int skipped = 0;
4796
4797         dout("close_sessions\n");
4798
4799         /* close sessions */
4800         mutex_lock(&mdsc->mutex);
4801         for (i = 0; i < mdsc->max_sessions; i++) {
4802                 session = __ceph_lookup_mds_session(mdsc, i);
4803                 if (!session)
4804                         continue;
4805                 mutex_unlock(&mdsc->mutex);
4806                 mutex_lock(&session->s_mutex);
4807                 if (__close_session(mdsc, session) <= 0)
4808                         skipped++;
4809                 mutex_unlock(&session->s_mutex);
4810                 ceph_put_mds_session(session);
4811                 mutex_lock(&mdsc->mutex);
4812         }
4813         mutex_unlock(&mdsc->mutex);
4814
4815         dout("waiting for sessions to close\n");
4816         wait_event_timeout(mdsc->session_close_wq,
4817                            done_closing_sessions(mdsc, skipped),
4818                            ceph_timeout_jiffies(opts->mount_timeout));
4819
4820         /* tear down remaining sessions */
4821         mutex_lock(&mdsc->mutex);
4822         for (i = 0; i < mdsc->max_sessions; i++) {
4823                 if (mdsc->sessions[i]) {
4824                         session = ceph_get_mds_session(mdsc->sessions[i]);
4825                         __unregister_session(mdsc, session);
4826                         mutex_unlock(&mdsc->mutex);
4827                         mutex_lock(&session->s_mutex);
4828                         remove_session_caps(session);
4829                         mutex_unlock(&session->s_mutex);
4830                         ceph_put_mds_session(session);
4831                         mutex_lock(&mdsc->mutex);
4832                 }
4833         }
4834         WARN_ON(!list_empty(&mdsc->cap_delay_list));
4835         mutex_unlock(&mdsc->mutex);
4836
4837         ceph_cleanup_snapid_map(mdsc);
4838         ceph_cleanup_empty_realms(mdsc);
4839
4840         cancel_work_sync(&mdsc->cap_reclaim_work);
4841         cancel_delayed_work_sync(&mdsc->delayed_work); /* cancel timer */
4842
4843         dout("stopped\n");
4844 }
4845
4846 void ceph_mdsc_force_umount(struct ceph_mds_client *mdsc)
4847 {
4848         struct ceph_mds_session *session;
4849         int mds;
4850
4851         dout("force umount\n");
4852
4853         mutex_lock(&mdsc->mutex);
4854         for (mds = 0; mds < mdsc->max_sessions; mds++) {
4855                 session = __ceph_lookup_mds_session(mdsc, mds);
4856                 if (!session)
4857                         continue;
4858
4859                 if (session->s_state == CEPH_MDS_SESSION_REJECTED)
4860                         __unregister_session(mdsc, session);
4861                 __wake_requests(mdsc, &session->s_waiting);
4862                 mutex_unlock(&mdsc->mutex);
4863
4864                 mutex_lock(&session->s_mutex);
4865                 __close_session(mdsc, session);
4866                 if (session->s_state == CEPH_MDS_SESSION_CLOSING) {
4867                         cleanup_session_requests(mdsc, session);
4868                         remove_session_caps(session);
4869                 }
4870                 mutex_unlock(&session->s_mutex);
4871                 ceph_put_mds_session(session);
4872
4873                 mutex_lock(&mdsc->mutex);
4874                 kick_requests(mdsc, mds);
4875         }
4876         __wake_requests(mdsc, &mdsc->waiting_for_map);
4877         mutex_unlock(&mdsc->mutex);
4878 }
4879
4880 static void ceph_mdsc_stop(struct ceph_mds_client *mdsc)
4881 {
4882         dout("stop\n");
4883         /*
4884          * Make sure the delayed work stopped before releasing
4885          * the resources.
4886          *
4887          * Because the cancel_delayed_work_sync() will only
4888          * guarantee that the work finishes executing. But the
4889          * delayed work will re-arm itself again after that.
4890          */
4891         flush_delayed_work(&mdsc->delayed_work);
4892
4893         if (mdsc->mdsmap)
4894                 ceph_mdsmap_destroy(mdsc->mdsmap);
4895         kfree(mdsc->sessions);
4896         ceph_caps_finalize(mdsc);
4897         ceph_pool_perm_destroy(mdsc);
4898 }
4899
4900 void ceph_mdsc_destroy(struct ceph_fs_client *fsc)
4901 {
4902         struct ceph_mds_client *mdsc = fsc->mdsc;
4903         dout("mdsc_destroy %p\n", mdsc);
4904
4905         if (!mdsc)
4906                 return;
4907
4908         /* flush out any connection work with references to us */
4909         ceph_msgr_flush();
4910
4911         ceph_mdsc_stop(mdsc);
4912
4913         ceph_metric_destroy(&mdsc->metric);
4914
4915         flush_delayed_work(&mdsc->metric.delayed_work);
4916         fsc->mdsc = NULL;
4917         kfree(mdsc);
4918         dout("mdsc_destroy %p done\n", mdsc);
4919 }
4920
4921 void ceph_mdsc_handle_fsmap(struct ceph_mds_client *mdsc, struct ceph_msg *msg)
4922 {
4923         struct ceph_fs_client *fsc = mdsc->fsc;
4924         const char *mds_namespace = fsc->mount_options->mds_namespace;
4925         void *p = msg->front.iov_base;
4926         void *end = p + msg->front.iov_len;
4927         u32 epoch;
4928         u32 num_fs;
4929         u32 mount_fscid = (u32)-1;
4930         int err = -EINVAL;
4931
4932         ceph_decode_need(&p, end, sizeof(u32), bad);
4933         epoch = ceph_decode_32(&p);
4934
4935         dout("handle_fsmap epoch %u\n", epoch);
4936
4937         /* struct_v, struct_cv, map_len, epoch, legacy_client_fscid */
4938         ceph_decode_skip_n(&p, end, 2 + sizeof(u32) * 3, bad);
4939
4940         ceph_decode_32_safe(&p, end, num_fs, bad);
4941         while (num_fs-- > 0) {
4942                 void *info_p, *info_end;
4943                 u32 info_len;
4944                 u32 fscid, namelen;
4945
4946                 ceph_decode_need(&p, end, 2 + sizeof(u32), bad);
4947                 p += 2;         // info_v, info_cv
4948                 info_len = ceph_decode_32(&p);
4949                 ceph_decode_need(&p, end, info_len, bad);
4950                 info_p = p;
4951                 info_end = p + info_len;
4952                 p = info_end;
4953
4954                 ceph_decode_need(&info_p, info_end, sizeof(u32) * 2, bad);
4955                 fscid = ceph_decode_32(&info_p);
4956                 namelen = ceph_decode_32(&info_p);
4957                 ceph_decode_need(&info_p, info_end, namelen, bad);
4958
4959                 if (mds_namespace &&
4960                     strlen(mds_namespace) == namelen &&
4961                     !strncmp(mds_namespace, (char *)info_p, namelen)) {
4962                         mount_fscid = fscid;
4963                         break;
4964                 }
4965         }
4966
4967         ceph_monc_got_map(&fsc->client->monc, CEPH_SUB_FSMAP, epoch);
4968         if (mount_fscid != (u32)-1) {
4969                 fsc->client->monc.fs_cluster_id = mount_fscid;
4970                 ceph_monc_want_map(&fsc->client->monc, CEPH_SUB_MDSMAP,
4971                                    0, true);
4972                 ceph_monc_renew_subs(&fsc->client->monc);
4973         } else {
4974                 err = -ENOENT;
4975                 goto err_out;
4976         }
4977         return;
4978
4979 bad:
4980         pr_err("error decoding fsmap\n");
4981 err_out:
4982         mutex_lock(&mdsc->mutex);
4983         mdsc->mdsmap_err = err;
4984         __wake_requests(mdsc, &mdsc->waiting_for_map);
4985         mutex_unlock(&mdsc->mutex);
4986 }
4987
4988 /*
4989  * handle mds map update.
4990  */
4991 void ceph_mdsc_handle_mdsmap(struct ceph_mds_client *mdsc, struct ceph_msg *msg)
4992 {
4993         u32 epoch;
4994         u32 maplen;
4995         void *p = msg->front.iov_base;
4996         void *end = p + msg->front.iov_len;
4997         struct ceph_mdsmap *newmap, *oldmap;
4998         struct ceph_fsid fsid;
4999         int err = -EINVAL;
5000
5001         ceph_decode_need(&p, end, sizeof(fsid)+2*sizeof(u32), bad);
5002         ceph_decode_copy(&p, &fsid, sizeof(fsid));
5003         if (ceph_check_fsid(mdsc->fsc->client, &fsid) < 0)
5004                 return;
5005         epoch = ceph_decode_32(&p);
5006         maplen = ceph_decode_32(&p);
5007         dout("handle_map epoch %u len %d\n", epoch, (int)maplen);
5008
5009         /* do we need it? */
5010         mutex_lock(&mdsc->mutex);
5011         if (mdsc->mdsmap && epoch <= mdsc->mdsmap->m_epoch) {
5012                 dout("handle_map epoch %u <= our %u\n",
5013                      epoch, mdsc->mdsmap->m_epoch);
5014                 mutex_unlock(&mdsc->mutex);
5015                 return;
5016         }
5017
5018         newmap = ceph_mdsmap_decode(&p, end, ceph_msgr2(mdsc->fsc->client));
5019         if (IS_ERR(newmap)) {
5020                 err = PTR_ERR(newmap);
5021                 goto bad_unlock;
5022         }
5023
5024         /* swap into place */
5025         if (mdsc->mdsmap) {
5026                 oldmap = mdsc->mdsmap;
5027                 mdsc->mdsmap = newmap;
5028                 check_new_map(mdsc, newmap, oldmap);
5029                 ceph_mdsmap_destroy(oldmap);
5030         } else {
5031                 mdsc->mdsmap = newmap;  /* first mds map */
5032         }
5033         mdsc->fsc->max_file_size = min((loff_t)mdsc->mdsmap->m_max_file_size,
5034                                         MAX_LFS_FILESIZE);
5035
5036         __wake_requests(mdsc, &mdsc->waiting_for_map);
5037         ceph_monc_got_map(&mdsc->fsc->client->monc, CEPH_SUB_MDSMAP,
5038                           mdsc->mdsmap->m_epoch);
5039
5040         mutex_unlock(&mdsc->mutex);
5041         schedule_delayed(mdsc, 0);
5042         return;
5043
5044 bad_unlock:
5045         mutex_unlock(&mdsc->mutex);
5046 bad:
5047         pr_err("error decoding mdsmap %d\n", err);
5048         return;
5049 }
5050
5051 static struct ceph_connection *mds_get_con(struct ceph_connection *con)
5052 {
5053         struct ceph_mds_session *s = con->private;
5054
5055         if (ceph_get_mds_session(s))
5056                 return con;
5057         return NULL;
5058 }
5059
5060 static void mds_put_con(struct ceph_connection *con)
5061 {
5062         struct ceph_mds_session *s = con->private;
5063
5064         ceph_put_mds_session(s);
5065 }
5066
5067 /*
5068  * if the client is unresponsive for long enough, the mds will kill
5069  * the session entirely.
5070  */
5071 static void mds_peer_reset(struct ceph_connection *con)
5072 {
5073         struct ceph_mds_session *s = con->private;
5074         struct ceph_mds_client *mdsc = s->s_mdsc;
5075
5076         pr_warn("mds%d closed our session\n", s->s_mds);
5077         send_mds_reconnect(mdsc, s);
5078 }
5079
5080 static void mds_dispatch(struct ceph_connection *con, struct ceph_msg *msg)
5081 {
5082         struct ceph_mds_session *s = con->private;
5083         struct ceph_mds_client *mdsc = s->s_mdsc;
5084         int type = le16_to_cpu(msg->hdr.type);
5085
5086         mutex_lock(&mdsc->mutex);
5087         if (__verify_registered_session(mdsc, s) < 0) {
5088                 mutex_unlock(&mdsc->mutex);
5089                 goto out;
5090         }
5091         mutex_unlock(&mdsc->mutex);
5092
5093         switch (type) {
5094         case CEPH_MSG_MDS_MAP:
5095                 ceph_mdsc_handle_mdsmap(mdsc, msg);
5096                 break;
5097         case CEPH_MSG_FS_MAP_USER:
5098                 ceph_mdsc_handle_fsmap(mdsc, msg);
5099                 break;
5100         case CEPH_MSG_CLIENT_SESSION:
5101                 handle_session(s, msg);
5102                 break;
5103         case CEPH_MSG_CLIENT_REPLY:
5104                 handle_reply(s, msg);
5105                 break;
5106         case CEPH_MSG_CLIENT_REQUEST_FORWARD:
5107                 handle_forward(mdsc, s, msg);
5108                 break;
5109         case CEPH_MSG_CLIENT_CAPS:
5110                 ceph_handle_caps(s, msg);
5111                 break;
5112         case CEPH_MSG_CLIENT_SNAP:
5113                 ceph_handle_snap(mdsc, s, msg);
5114                 break;
5115         case CEPH_MSG_CLIENT_LEASE:
5116                 handle_lease(mdsc, s, msg);
5117                 break;
5118         case CEPH_MSG_CLIENT_QUOTA:
5119                 ceph_handle_quota(mdsc, s, msg);
5120                 break;
5121
5122         default:
5123                 pr_err("received unknown message type %d %s\n", type,
5124                        ceph_msg_type_name(type));
5125         }
5126 out:
5127         ceph_msg_put(msg);
5128 }
5129
5130 /*
5131  * authentication
5132  */
5133
5134 /*
5135  * Note: returned pointer is the address of a structure that's
5136  * managed separately.  Caller must *not* attempt to free it.
5137  */
5138 static struct ceph_auth_handshake *
5139 mds_get_authorizer(struct ceph_connection *con, int *proto, int force_new)
5140 {
5141         struct ceph_mds_session *s = con->private;
5142         struct ceph_mds_client *mdsc = s->s_mdsc;
5143         struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
5144         struct ceph_auth_handshake *auth = &s->s_auth;
5145         int ret;
5146
5147         ret = __ceph_auth_get_authorizer(ac, auth, CEPH_ENTITY_TYPE_MDS,
5148                                          force_new, proto, NULL, NULL);
5149         if (ret)
5150                 return ERR_PTR(ret);
5151
5152         return auth;
5153 }
5154
5155 static int mds_add_authorizer_challenge(struct ceph_connection *con,
5156                                     void *challenge_buf, int challenge_buf_len)
5157 {
5158         struct ceph_mds_session *s = con->private;
5159         struct ceph_mds_client *mdsc = s->s_mdsc;
5160         struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
5161
5162         return ceph_auth_add_authorizer_challenge(ac, s->s_auth.authorizer,
5163                                             challenge_buf, challenge_buf_len);
5164 }
5165
5166 static int mds_verify_authorizer_reply(struct ceph_connection *con)
5167 {
5168         struct ceph_mds_session *s = con->private;
5169         struct ceph_mds_client *mdsc = s->s_mdsc;
5170         struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
5171         struct ceph_auth_handshake *auth = &s->s_auth;
5172
5173         return ceph_auth_verify_authorizer_reply(ac, auth->authorizer,
5174                 auth->authorizer_reply_buf, auth->authorizer_reply_buf_len,
5175                 NULL, NULL, NULL, NULL);
5176 }
5177
5178 static int mds_invalidate_authorizer(struct ceph_connection *con)
5179 {
5180         struct ceph_mds_session *s = con->private;
5181         struct ceph_mds_client *mdsc = s->s_mdsc;
5182         struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
5183
5184         ceph_auth_invalidate_authorizer(ac, CEPH_ENTITY_TYPE_MDS);
5185
5186         return ceph_monc_validate_auth(&mdsc->fsc->client->monc);
5187 }
5188
5189 static int mds_get_auth_request(struct ceph_connection *con,
5190                                 void *buf, int *buf_len,
5191                                 void **authorizer, int *authorizer_len)
5192 {
5193         struct ceph_mds_session *s = con->private;
5194         struct ceph_auth_client *ac = s->s_mdsc->fsc->client->monc.auth;
5195         struct ceph_auth_handshake *auth = &s->s_auth;
5196         int ret;
5197
5198         ret = ceph_auth_get_authorizer(ac, auth, CEPH_ENTITY_TYPE_MDS,
5199                                        buf, buf_len);
5200         if (ret)
5201                 return ret;
5202
5203         *authorizer = auth->authorizer_buf;
5204         *authorizer_len = auth->authorizer_buf_len;
5205         return 0;
5206 }
5207
5208 static int mds_handle_auth_reply_more(struct ceph_connection *con,
5209                                       void *reply, int reply_len,
5210                                       void *buf, int *buf_len,
5211                                       void **authorizer, int *authorizer_len)
5212 {
5213         struct ceph_mds_session *s = con->private;
5214         struct ceph_auth_client *ac = s->s_mdsc->fsc->client->monc.auth;
5215         struct ceph_auth_handshake *auth = &s->s_auth;
5216         int ret;
5217
5218         ret = ceph_auth_handle_svc_reply_more(ac, auth, reply, reply_len,
5219                                               buf, buf_len);
5220         if (ret)
5221                 return ret;
5222
5223         *authorizer = auth->authorizer_buf;
5224         *authorizer_len = auth->authorizer_buf_len;
5225         return 0;
5226 }
5227
5228 static int mds_handle_auth_done(struct ceph_connection *con,
5229                                 u64 global_id, void *reply, int reply_len,
5230                                 u8 *session_key, int *session_key_len,
5231                                 u8 *con_secret, int *con_secret_len)
5232 {
5233         struct ceph_mds_session *s = con->private;
5234         struct ceph_auth_client *ac = s->s_mdsc->fsc->client->monc.auth;
5235         struct ceph_auth_handshake *auth = &s->s_auth;
5236
5237         return ceph_auth_handle_svc_reply_done(ac, auth, reply, reply_len,
5238                                                session_key, session_key_len,
5239                                                con_secret, con_secret_len);
5240 }
5241
5242 static int mds_handle_auth_bad_method(struct ceph_connection *con,
5243                                       int used_proto, int result,
5244                                       const int *allowed_protos, int proto_cnt,
5245                                       const int *allowed_modes, int mode_cnt)
5246 {
5247         struct ceph_mds_session *s = con->private;
5248         struct ceph_mon_client *monc = &s->s_mdsc->fsc->client->monc;
5249         int ret;
5250
5251         if (ceph_auth_handle_bad_authorizer(monc->auth, CEPH_ENTITY_TYPE_MDS,
5252                                             used_proto, result,
5253                                             allowed_protos, proto_cnt,
5254                                             allowed_modes, mode_cnt)) {
5255                 ret = ceph_monc_validate_auth(monc);
5256                 if (ret)
5257                         return ret;
5258         }
5259
5260         return -EACCES;
5261 }
5262
5263 static struct ceph_msg *mds_alloc_msg(struct ceph_connection *con,
5264                                 struct ceph_msg_header *hdr, int *skip)
5265 {
5266         struct ceph_msg *msg;
5267         int type = (int) le16_to_cpu(hdr->type);
5268         int front_len = (int) le32_to_cpu(hdr->front_len);
5269
5270         if (con->in_msg)
5271                 return con->in_msg;
5272
5273         *skip = 0;
5274         msg = ceph_msg_new(type, front_len, GFP_NOFS, false);
5275         if (!msg) {
5276                 pr_err("unable to allocate msg type %d len %d\n",
5277                        type, front_len);
5278                 return NULL;
5279         }
5280
5281         return msg;
5282 }
5283
5284 static int mds_sign_message(struct ceph_msg *msg)
5285 {
5286        struct ceph_mds_session *s = msg->con->private;
5287        struct ceph_auth_handshake *auth = &s->s_auth;
5288
5289        return ceph_auth_sign_message(auth, msg);
5290 }
5291
5292 static int mds_check_message_signature(struct ceph_msg *msg)
5293 {
5294        struct ceph_mds_session *s = msg->con->private;
5295        struct ceph_auth_handshake *auth = &s->s_auth;
5296
5297        return ceph_auth_check_message_signature(auth, msg);
5298 }
5299
5300 static const struct ceph_connection_operations mds_con_ops = {
5301         .get = mds_get_con,
5302         .put = mds_put_con,
5303         .alloc_msg = mds_alloc_msg,
5304         .dispatch = mds_dispatch,
5305         .peer_reset = mds_peer_reset,
5306         .get_authorizer = mds_get_authorizer,
5307         .add_authorizer_challenge = mds_add_authorizer_challenge,
5308         .verify_authorizer_reply = mds_verify_authorizer_reply,
5309         .invalidate_authorizer = mds_invalidate_authorizer,
5310         .sign_message = mds_sign_message,
5311         .check_message_signature = mds_check_message_signature,
5312         .get_auth_request = mds_get_auth_request,
5313         .handle_auth_reply_more = mds_handle_auth_reply_more,
5314         .handle_auth_done = mds_handle_auth_done,
5315         .handle_auth_bad_method = mds_handle_auth_bad_method,
5316 };
5317
5318 /* eof */