lustre/lvfs/upcall_cache.c

   1 /* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
   2  * vim:expandtab:shiftwidth=8:tabstop=8:
   3  *
   4  *  Supplementary groups cache.
   5  *
   6  *  Copyright (c) 2004 Cluster File Systems, Inc.
   7  *
   8  *   This file is part of Lustre, http://www.lustre.org.
   9  *
  10  *   Lustre is free software; you can redistribute it and/or
  11  *   modify it under the terms of version 2 of the GNU General Public
  12  *   License as published by the Free Software Foundation.
  13  *
  14  *   Lustre is distributed in the hope that it will be useful,
  15  *   but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  17  *   GNU General Public License for more details.
  18  *
  19  *   You should have received a copy of the GNU General Public License
  20  *   along with Lustre; if not, write to the Free Software
  21  *   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  22  */
  23
  24 #define DEBUG_SUBSYSTEM S_SEC
  25
  26 #ifndef AUTOCONF_INCLUDED
  27 #include <linux/config.h>
  28 #endif
  29 #include <linux/module.h>
  30 #include <linux/kernel.h>
  31 #include <linux/mm.h>
  32 #include <linux/kmod.h>
  33 #include <linux/string.h>
  34 #include <linux/stat.h>
  35 #include <linux/errno.h>
  36 #include <linux/version.h>
  37 #include <linux/unistd.h>
  38
  39 #include <asm/system.h>
  40 #include <asm/uaccess.h>
  41
  42 #include <linux/fs.h>
  43 #include <linux/stat.h>
  44 #include <asm/uaccess.h>
  45 #include <linux/slab.h>
  46 #ifdef HAVE_SEGMENT_H
  47 # include <asm/segment.h>
  48 #endif
  49
  50 #include <obd_support.h>
  51 #include <lustre_lib.h>
  52
  53 static struct upcall_cache_entry *alloc_entry(struct upcall_cache *cache,
  54                                               __u64 key, void *args)
  55 {
  56         struct upcall_cache_entry *entry;
  57
  58         OBD_ALLOC_PTR(entry);
  59         if (!entry)
  60                 return NULL;
  61
  62         UC_CACHE_SET_NEW(entry);
  63         INIT_LIST_HEAD(&entry->ue_hash);
  64         entry->ue_key = key;
  65         atomic_set(&entry->ue_refcount, 0);
  66         init_waitqueue_head(&entry->ue_waitq);
  67         if (cache->uc_ops->init_entry)
  68                 cache->uc_ops->init_entry(entry, args);
  69         return entry;
  70 }
  71
  72 /* protected by cache lock */
  73 static void free_entry(struct upcall_cache *cache,
  74                        struct upcall_cache_entry *entry)
  75 {
  76         if (cache->uc_ops->free_entry)
  77                 cache->uc_ops->free_entry(cache, entry);
  78
  79         list_del(&entry->ue_hash);
  80         CDEBUG(D_OTHER, "destroy cache entry %p for key "LPU64"\n",
  81                entry, entry->ue_key);
  82         OBD_FREE_PTR(entry);
  83 }
  84
  85 static inline int upcall_compare(struct upcall_cache *cache,
  86                                  struct upcall_cache_entry *entry,
  87                                  __u64 key, void *args)
  88 {
  89         if (entry->ue_key != key)
  90                 return -1;
  91
  92         if (cache->uc_ops->upcall_compare)
  93                 return cache->uc_ops->upcall_compare(cache, entry, key, args);
  94
  95         return 0;
  96 }
  97
  98 static inline int downcall_compare(struct upcall_cache *cache,
  99                                    struct upcall_cache_entry *entry,
 100                                    __u64 key, void *args)
 101 {
 102         if (entry->ue_key != key)
 103                 return -1;
 104
 105         if (cache->uc_ops->downcall_compare)
 106                 return cache->uc_ops->downcall_compare(cache, entry, key, args);
 107
 108         return 0;
 109 }
 110
 111 static inline void get_entry(struct upcall_cache_entry *entry)
 112 {
 113         atomic_inc(&entry->ue_refcount);
 114 }
 115
 116 static inline void put_entry(struct upcall_cache *cache,
 117                              struct upcall_cache_entry *entry)
 118 {
 119         if (atomic_dec_and_test(&entry->ue_refcount) &&
 120             (UC_CACHE_IS_INVALID(entry) || UC_CACHE_IS_EXPIRED(entry))) {
 121                 free_entry(cache, entry);
 122         }
 123 }
 124
 125 static int check_unlink_entry(struct upcall_cache *cache,
 126                               struct upcall_cache_entry *entry)
 127 {
 128         if (UC_CACHE_IS_VALID(entry) &&
 129             time_before(jiffies, entry->ue_expire))
 130                 return 0;
 131
 132         if (UC_CACHE_IS_ACQUIRING(entry)) {
 133                 if (time_before(jiffies, entry->ue_acquire_expire))
 134                         return 0;
 135
 136                 UC_CACHE_SET_EXPIRED(entry);
 137                 wake_up_all(&entry->ue_waitq);
 138         } else if (!UC_CACHE_IS_INVALID(entry)) {
 139                 UC_CACHE_SET_EXPIRED(entry);
 140         }
 141
 142         list_del_init(&entry->ue_hash);
 143         if (!atomic_read(&entry->ue_refcount))
 144                 free_entry(cache, entry);
 145         return 1;
 146 }
 147
 148 static inline int refresh_entry(struct upcall_cache *cache,
 149                          struct upcall_cache_entry *entry)
 150 {
 151         LASSERT(cache->uc_ops->do_upcall);
 152         return cache->uc_ops->do_upcall(cache, entry);
 153 }
 154
 155 struct upcall_cache_entry *upcall_cache_get_entry(struct upcall_cache *cache,
 156                                                   __u64 key, void *args)
 157 {
 158         struct upcall_cache_entry *entry = NULL, *new = NULL, *next;
 159         struct list_head *head;
 160         wait_queue_t wait;
 161         int rc, found;
 162         ENTRY;
 163
 164         LASSERT(cache);
 165
 166         head = &cache->uc_hashtable[UC_CACHE_HASH_INDEX(key)];
 167 find_again:
 168         found = 0;
 169         spin_lock(&cache->uc_lock);
 170         list_for_each_entry_safe(entry, next, head, ue_hash) {
 171                 /* check invalid & expired items */
 172                 if (check_unlink_entry(cache, entry))
 173                         continue;
 174                 if (upcall_compare(cache, entry, key, args) == 0) {
 175                         found = 1;
 176                         break;
 177                 }
 178         }
 179
 180         if (!found) { /* didn't find it */
 181                 if (!new) {
 182                         spin_unlock(&cache->uc_lock);
 183                         new = alloc_entry(cache, key, args);
 184                         if (!new) {
 185                                 CERROR("fail to alloc entry\n");
 186                                 RETURN(ERR_PTR(-ENOMEM));
 187                         }
 188                         goto find_again;
 189                 } else {
 190                         list_add(&new->ue_hash, head);
 191                         entry = new;
 192                 }
 193         } else {
 194                 if (new) {
 195                         free_entry(cache, new);
 196                         new = NULL;
 197                 }
 198                 list_move(&entry->ue_hash, head);
 199         }
 200         get_entry(entry);
 201
 202         /* acquire for new one */
 203         if (UC_CACHE_IS_NEW(entry)) {
 204                 UC_CACHE_SET_ACQUIRING(entry);
 205                 UC_CACHE_CLEAR_NEW(entry);
 206                 entry->ue_acquire_expire = jiffies + cache->uc_acquire_expire;
 207                 spin_unlock(&cache->uc_lock);
 208                 rc = refresh_entry(cache, entry);
 209                 spin_lock(&cache->uc_lock);
 210                 if (rc < 0) {
 211                         UC_CACHE_CLEAR_ACQUIRING(entry);
 212                         UC_CACHE_SET_INVALID(entry);
 213                 }
 214                 /* fall through */
 215         }
 216         /* someone (and only one) is doing upcall upon
 217          * this item, just wait it complete
 218          */
 219         if (UC_CACHE_IS_ACQUIRING(entry)) {
 220                 unsigned long expiry = jiffies + cache->uc_acquire_expire;
 221
 222                 init_waitqueue_entry(&wait, current);
 223                 add_wait_queue(&entry->ue_waitq, &wait);
 224                 set_current_state(TASK_INTERRUPTIBLE);
 225                 spin_unlock(&cache->uc_lock);
 226
 227                 schedule_timeout(cache->uc_acquire_expire);
 228
 229                 spin_lock(&cache->uc_lock);
 230                 remove_wait_queue(&entry->ue_waitq, &wait);
 231                 if (UC_CACHE_IS_ACQUIRING(entry)) {
 232                         /* we're interrupted or upcall failed in the middle */
 233                         rc = time_before(jiffies, expiry) ? -EINTR : -ETIMEDOUT;
 234                         put_entry(cache, entry);
 235                         CERROR("acquire timeout exceeded for key "LPU64
 236                                "\n", entry->ue_key);
 237                         GOTO(out, entry = ERR_PTR(rc));
 238                 }
 239                 /* fall through */
 240         }
 241
 242         /* invalid means error, don't need to try again */
 243         if (UC_CACHE_IS_INVALID(entry)) {
 244                 put_entry(cache, entry);
 245                 GOTO(out, entry = ERR_PTR(-EIDRM));
 246         }
 247
 248         /* check expired
 249          * We can't refresh the existing one because some
 250          * memory might be shared by multiple processes.
 251          */
 252         if (check_unlink_entry(cache, entry)) {
 253                 /* if expired, try again. but if this entry is
 254                  * created by me but too quickly turn to expired
 255                  * without any error, should at least give a
 256                  * chance to use it once.
 257                  */
 258                 if (entry != new) {
 259                         put_entry(cache, entry);
 260                         spin_unlock(&cache->uc_lock);
 261                         new = NULL;
 262                         goto find_again;
 263                 }
 264         }
 265
 266         /* Now we know it's good */
 267 out:
 268         spin_unlock(&cache->uc_lock);
 269         RETURN(entry);
 270 }
 271 EXPORT_SYMBOL(upcall_cache_get_entry);
 272
 273 void upcall_cache_put_entry(struct upcall_cache *cache,
 274                             struct upcall_cache_entry *entry)
 275 {
 276         ENTRY;
 277
 278         if (!entry) {
 279                 EXIT;
 280                 return;
 281         }
 282
 283         LASSERT(atomic_read(&entry->ue_refcount) > 0);
 284         spin_lock(&cache->uc_lock);
 285         put_entry(cache, entry);
 286         spin_unlock(&cache->uc_lock);
 287         EXIT;
 288 }
 289 EXPORT_SYMBOL(upcall_cache_put_entry);
 290
 291 int upcall_cache_downcall(struct upcall_cache *cache, __u32 err, __u64 key,
 292                           void *args)
 293 {
 294         struct upcall_cache_entry *entry = NULL;
 295         struct list_head *head;
 296         int found = 0, rc = 0;
 297         ENTRY;
 298
 299         LASSERT(cache);
 300
 301         head = &cache->uc_hashtable[UC_CACHE_HASH_INDEX(key)];
 302
 303         spin_lock(&cache->uc_lock);
 304         list_for_each_entry(entry, head, ue_hash) {
 305                 if (downcall_compare(cache, entry, key, args) == 0) {
 306                         found = 1;
 307                         get_entry(entry);
 308                         break;
 309                 }
 310         }
 311
 312         if (!found) {
 313                 CDEBUG(D_OTHER, "%s: upcall for key "LPU64" not expected\n",
 314                        cache->uc_name, key);
 315                 /* haven't found, it's possible */
 316                 spin_unlock(&cache->uc_lock);
 317                 RETURN(-EINVAL);
 318         }
 319
 320         if (err) {
 321                 CDEBUG(D_OTHER, "%s: upcall for key "LPU64" returned %d\n",
 322                        cache->uc_name, entry->ue_key, err);
 323                 GOTO(out, rc = -EINVAL);
 324         }
 325
 326         if (!UC_CACHE_IS_ACQUIRING(entry)) {
 327                 CDEBUG(D_RPCTRACE,"%s: found uptodate entry %p (key "LPU64")\n",
 328                        cache->uc_name, entry, entry->ue_key);
 329                 GOTO(out, rc = 0);
 330         }
 331
 332         if (UC_CACHE_IS_INVALID(entry) || UC_CACHE_IS_EXPIRED(entry)) {
 333                 CERROR("%s: found a stale entry %p (key "LPU64") in ioctl\n",
 334                        cache->uc_name, entry, entry->ue_key);
 335                 GOTO(out, rc = -EINVAL);
 336         }
 337
 338         spin_unlock(&cache->uc_lock);
 339         if (cache->uc_ops->parse_downcall)
 340                 rc = cache->uc_ops->parse_downcall(cache, entry, args);
 341         spin_lock(&cache->uc_lock);
 342         if (rc)
 343                 GOTO(out, rc);
 344
 345         entry->ue_expire = jiffies + cache->uc_entry_expire;
 346         UC_CACHE_SET_VALID(entry);
 347         CDEBUG(D_OTHER, "%s: created upcall cache entry %p for key "LPU64"\n",
 348                cache->uc_name, entry, entry->ue_key);
 349 out:
 350         if (rc) {
 351                 UC_CACHE_SET_INVALID(entry);
 352                 list_del_init(&entry->ue_hash);
 353         }
 354         UC_CACHE_CLEAR_ACQUIRING(entry);
 355         spin_unlock(&cache->uc_lock);
 356         wake_up_all(&entry->ue_waitq);
 357         put_entry(cache, entry);
 358
 359         RETURN(rc);
 360 }
 361 EXPORT_SYMBOL(upcall_cache_downcall);
 362
 363 static void cache_flush(struct upcall_cache *cache, int force)
 364 {
 365         struct upcall_cache_entry *entry, *next;
 366         int i;
 367         ENTRY;
 368
 369         spin_lock(&cache->uc_lock);
 370         for (i = 0; i < UC_CACHE_HASH_SIZE; i++) {
 371                 list_for_each_entry_safe(entry, next,
 372                                          &cache->uc_hashtable[i], ue_hash) {
 373                         if (!force && atomic_read(&entry->ue_refcount)) {
 374                                 UC_CACHE_SET_EXPIRED(entry);
 375                                 continue;
 376                         }
 377                         LASSERT(!atomic_read(&entry->ue_refcount));
 378                         free_entry(cache, entry);
 379                 }
 380         }
 381         spin_unlock(&cache->uc_lock);
 382         EXIT;
 383 }
 384
 385 void upcall_cache_flush_idle(struct upcall_cache *cache)
 386 {
 387         cache_flush(cache, 0);
 388 }
 389 EXPORT_SYMBOL(upcall_cache_flush_idle);
 390
 391 void upcall_cache_flush_all(struct upcall_cache *cache)
 392 {
 393         cache_flush(cache, 1);
 394 }
 395 EXPORT_SYMBOL(upcall_cache_flush_all);
 396
 397 void upcall_cache_flush_one(struct upcall_cache *cache, __u64 key, void *args)
 398 {
 399         struct list_head *head;
 400         struct upcall_cache_entry *entry;
 401         int found = 0;
 402         ENTRY;
 403
 404         head = &cache->uc_hashtable[UC_CACHE_HASH_INDEX(key)];
 405
 406         spin_lock(&cache->uc_lock);
 407         list_for_each_entry(entry, head, ue_hash) {
 408                 if (upcall_compare(cache, entry, key, args) == 0) {
 409                         found = 1;
 410                         break;
 411                 }
 412         }
 413
 414         if (found) {
 415                 CWARN("%s: flush entry %p: key "LPU64", ref %d, fl %x, "
 416                       "cur %lu, ex %ld/%ld\n",
 417                       cache->uc_name, entry, entry->ue_key,
 418                       atomic_read(&entry->ue_refcount), entry->ue_flags,
 419                       get_seconds(), entry->ue_acquire_expire,
 420                       entry->ue_expire);
 421                 UC_CACHE_SET_EXPIRED(entry);
 422                 if (!atomic_read(&entry->ue_refcount))
 423                         free_entry(cache, entry);
 424         }
 425         spin_unlock(&cache->uc_lock);
 426 }
 427 EXPORT_SYMBOL(upcall_cache_flush_one);
 428
 429 struct upcall_cache *upcall_cache_init(const char *name, const char *upcall,
 430                                        struct upcall_cache_ops *ops)
 431 {
 432         struct upcall_cache *cache;
 433         int i;
 434         ENTRY;
 435
 436         OBD_ALLOC(cache, sizeof(*cache));
 437         if (!cache)
 438                 RETURN(ERR_PTR(-ENOMEM));
 439
 440         spin_lock_init(&cache->uc_lock);
 441         for (i = 0; i < UC_CACHE_HASH_SIZE; i++)
 442                 INIT_LIST_HEAD(&cache->uc_hashtable[i]);
 443         strncpy(cache->uc_name, name, sizeof(cache->uc_name) - 1);
 444         /* upcall pathname proc tunable */
 445         strncpy(cache->uc_upcall, upcall, sizeof(cache->uc_upcall) - 1);
 446         cache->uc_entry_expire = 10 * 60 * HZ;
 447         cache->uc_acquire_expire = 15 * HZ;
 448         cache->uc_ops = ops;
 449
 450         RETURN(cache);
 451 }
 452 EXPORT_SYMBOL(upcall_cache_init);
 453
 454 void upcall_cache_cleanup(struct upcall_cache *cache)
 455 {
 456         if (!cache)
 457                 return;
 458         upcall_cache_flush_all(cache);
 459         OBD_FREE(cache, sizeof(*cache));
 460 }
 461 EXPORT_SYMBOL(upcall_cache_cleanup);