fs/ubifs/shrinker.c

   1 /*
   2  * This file is part of UBIFS.
   3  *
   4  * Copyright (C) 2006-2008 Nokia Corporation.
   5  *
   6  * This program is free software; you can redistribute it and/or modify it
   7  * under the terms of the GNU General Public License version 2 as published by
   8  * the Free Software Foundation.
   9  *
  10  * This program is distributed in the hope that it will be useful, but WITHOUT
  11  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
  12  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  13  * more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along with
  16  * this program; if not, write to the Free Software Foundation, Inc., 51
  17  * Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
  18  *
  19  * Authors: Artem Bityutskiy (Битюцкий Артём)
  20  *          Adrian Hunter
  21  */
  22
  23 /*
  24  * This file implements UBIFS shrinker which evicts clean znodes from the TNC
  25  * tree when Linux VM needs more RAM.
  26  *
  27  * We do not implement any LRU lists to find oldest znodes to free because it
  28  * would add additional overhead to the file system fast paths. So the shrinker
  29  * just walks the TNC tree when searching for znodes to free.
  30  *
  31  * If the root of a TNC sub-tree is clean and old enough, then the children are
  32  * also clean and old enough. So the shrinker walks the TNC in level order and
  33  * dumps entire sub-trees.
  34  *
  35  * The age of znodes is just the time-stamp when they were last looked at.
  36  * The current shrinker first tries to evict old znodes, then young ones.
  37  *
  38  * Since the shrinker is global, it has to protect against races with FS
  39  * un-mounts, which is done by the 'ubifs_infos_lock' and 'c->umount_mutex'.
  40  */
  41
  42 #include "ubifs.h"
  43
  44 /* List of all UBIFS file-system instances */
  45 LIST_HEAD(ubifs_infos);
  46
  47 /*
  48  * We number each shrinker run and record the number on the ubifs_info structure
  49  * so that we can easily work out which ubifs_info structures have already been
  50  * done by the current run.
  51  */
  52 static unsigned int shrinker_run_no;
  53
  54 /* Protects 'ubifs_infos' list */
  55 DEFINE_SPINLOCK(ubifs_infos_lock);
  56
  57 /* Global clean znode counter (for all mounted UBIFS instances) */
  58 atomic_long_t ubifs_clean_zn_cnt;
  59
  60 /**
  61  * shrink_tnc - shrink TNC tree.
  62  * @c: UBIFS file-system description object
  63  * @nr: number of znodes to free
  64  * @age: the age of znodes to free
  65  * @contention: if any contention, this is set to %1
  66  *
  67  * This function traverses TNC tree and frees clean znodes. It does not free
  68  * clean znodes which younger then @age. Returns number of freed znodes.
  69  */
  70 static int shrink_tnc(struct ubifs_info *c, int nr, int age, int *contention)
  71 {
  72         int total_freed = 0;
  73         struct ubifs_znode *znode, *zprev;
  74         int time = get_seconds();
  75
  76         ubifs_assert(mutex_is_locked(&c->umount_mutex));
  77         ubifs_assert(mutex_is_locked(&c->tnc_mutex));
  78
  79         if (!c->zroot.znode || atomic_long_read(&c->clean_zn_cnt) == 0)
  80                 return 0;
  81
  82         /*
  83          * Traverse the TNC tree in levelorder manner, so that it is possible
  84          * to destroy large sub-trees. Indeed, if a znode is old, then all its
  85          * children are older or of the same age.
  86          *
  87          * Note, we are holding 'c->tnc_mutex', so we do not have to lock the
  88          * 'c->space_lock' when _reading_ 'c->clean_zn_cnt', because it is
  89          * changed only when the 'c->tnc_mutex' is held.
  90          */
  91         zprev = NULL;
  92         znode = ubifs_tnc_levelorder_next(c->zroot.znode, NULL);
  93         while (znode && total_freed < nr &&
  94                atomic_long_read(&c->clean_zn_cnt) > 0) {
  95                 int freed;
  96
  97                 /*
  98                  * If the znode is clean, but it is in the 'c->cnext' list, this
  99                  * means that this znode has just been written to flash as a
 100                  * part of commit and was marked clean. They will be removed
 101                  * from the list at end commit. We cannot change the list,
 102                  * because it is not protected by any mutex (design decision to
 103                  * make commit really independent and parallel to main I/O). So
 104                  * we just skip these znodes.
 105                  *
 106                  * Note, the 'clean_zn_cnt' counters are not updated until
 107                  * after the commit, so the UBIFS shrinker does not report
 108                  * the znodes which are in the 'c->cnext' list as freeable.
 109                  *
 110                  * Also note, if the root of a sub-tree is not in 'c->cnext',
 111                  * then the whole sub-tree is not in 'c->cnext' as well, so it
 112                  * is safe to dump whole sub-tree.
 113                  */
 114
 115                 if (znode->cnext) {
 116                         /*
 117                          * Very soon these znodes will be removed from the list
 118                          * and become freeable.
 119                          */
 120                         *contention = 1;
 121                 } else if (!ubifs_zn_dirty(znode) &&
 122                            abs(time - znode->time) >= age) {
 123                         if (znode->parent)
 124                                 znode->parent->zbranch[znode->iip].znode = NULL;
 125                         else
 126                                 c->zroot.znode = NULL;
 127
 128                         freed = ubifs_destroy_tnc_subtree(znode);
 129                         atomic_long_sub(freed, &ubifs_clean_zn_cnt);
 130                         atomic_long_sub(freed, &c->clean_zn_cnt);
 131                         total_freed += freed;
 132                         znode = zprev;
 133                 }
 134
 135                 if (unlikely(!c->zroot.znode))
 136                         break;
 137
 138                 zprev = znode;
 139                 znode = ubifs_tnc_levelorder_next(c->zroot.znode, znode);
 140                 cond_resched();
 141         }
 142
 143         return total_freed;
 144 }
 145
 146 /**
 147  * shrink_tnc_trees - shrink UBIFS TNC trees.
 148  * @nr: number of znodes to free
 149  * @age: the age of znodes to free
 150  * @contention: if any contention, this is set to %1
 151  *
 152  * This function walks the list of mounted UBIFS file-systems and frees clean
 153  * znodes which are older than @age, until at least @nr znodes are freed.
 154  * Returns the number of freed znodes.
 155  */
 156 static int shrink_tnc_trees(int nr, int age, int *contention)
 157 {
 158         struct ubifs_info *c;
 159         struct list_head *p;
 160         unsigned int run_no;
 161         int freed = 0;
 162
 163         spin_lock(&ubifs_infos_lock);
 164         do {
 165                 run_no = ++shrinker_run_no;
 166         } while (run_no == 0);
 167         /* Iterate over all mounted UBIFS file-systems and try to shrink them */
 168         p = ubifs_infos.next;
 169         while (p != &ubifs_infos) {
 170                 c = list_entry(p, struct ubifs_info, infos_list);
 171                 /*
 172                  * We move the ones we do to the end of the list, so we stop
 173                  * when we see one we have already done.
 174                  */
 175                 if (c->shrinker_run_no == run_no)
 176                         break;
 177                 if (!mutex_trylock(&c->umount_mutex)) {
 178                         /* Some un-mount is in progress, try next FS */
 179                         *contention = 1;
 180                         p = p->next;
 181                         continue;
 182                 }
 183                 /*
 184                  * We're holding 'c->umount_mutex', so the file-system won't go
 185                  * away.
 186                  */
 187                 if (!mutex_trylock(&c->tnc_mutex)) {
 188                         mutex_unlock(&c->umount_mutex);
 189                         *contention = 1;
 190                         p = p->next;
 191                         continue;
 192                 }
 193                 spin_unlock(&ubifs_infos_lock);
 194                 /*
 195                  * OK, now we have TNC locked, the file-system cannot go away -
 196                  * it is safe to reap the cache.
 197                  */
 198                 c->shrinker_run_no = run_no;
 199                 freed += shrink_tnc(c, nr, age, contention);
 200                 mutex_unlock(&c->tnc_mutex);
 201                 spin_lock(&ubifs_infos_lock);
 202                 /* Get the next list element before we move this one */
 203                 p = p->next;
 204                 /*
 205                  * Move this one to the end of the list to provide some
 206                  * fairness.
 207                  */
 208                 list_move_tail(&c->infos_list, &ubifs_infos);
 209                 mutex_unlock(&c->umount_mutex);
 210                 if (freed >= nr)
 211                         break;
 212         }
 213         spin_unlock(&ubifs_infos_lock);
 214         return freed;
 215 }
 216
 217 /**
 218  * kick_a_thread - kick a background thread to start commit.
 219  *
 220  * This function kicks a background thread to start background commit. Returns
 221  * %-1 if a thread was kicked or there is another reason to assume the memory
 222  * will soon be freed or become freeable. If there are no dirty znodes, returns
 223  * %0.
 224  */
 225 static int kick_a_thread(void)
 226 {
 227         int i;
 228         struct ubifs_info *c;
 229
 230         /*
 231          * Iterate over all mounted UBIFS file-systems and find out if there is
 232          * already an ongoing commit operation there. If no, then iterate for
 233          * the second time and initiate background commit.
 234          */
 235         spin_lock(&ubifs_infos_lock);
 236         for (i = 0; i < 2; i++) {
 237                 list_for_each_entry(c, &ubifs_infos, infos_list) {
 238                         long dirty_zn_cnt;
 239
 240                         if (!mutex_trylock(&c->umount_mutex)) {
 241                                 /*
 242                                  * Some un-mount is in progress, it will
 243                                  * certainly free memory, so just return.
 244                                  */
 245                                 spin_unlock(&ubifs_infos_lock);
 246                                 return -1;
 247                         }
 248
 249                         dirty_zn_cnt = atomic_long_read(&c->dirty_zn_cnt);
 250
 251                         if (!dirty_zn_cnt || c->cmt_state == COMMIT_BROKEN ||
 252                             c->ro_mount || c->ro_error) {
 253                                 mutex_unlock(&c->umount_mutex);
 254                                 continue;
 255                         }
 256
 257                         if (c->cmt_state != COMMIT_RESTING) {
 258                                 spin_unlock(&ubifs_infos_lock);
 259                                 mutex_unlock(&c->umount_mutex);
 260                                 return -1;
 261                         }
 262
 263                         if (i == 1) {
 264                                 list_move_tail(&c->infos_list, &ubifs_infos);
 265                                 spin_unlock(&ubifs_infos_lock);
 266
 267                                 ubifs_request_bg_commit(c);
 268                                 mutex_unlock(&c->umount_mutex);
 269                                 return -1;
 270                         }
 271                         mutex_unlock(&c->umount_mutex);
 272                 }
 273         }
 274         spin_unlock(&ubifs_infos_lock);
 275
 276         return 0;
 277 }
 278
 279 unsigned long ubifs_shrink_count(struct shrinker *shrink,
 280                                  struct shrink_control *sc)
 281 {
 282         long clean_zn_cnt = atomic_long_read(&ubifs_clean_zn_cnt);
 283
 284         /*
 285          * Due to the way UBIFS updates the clean znode counter it may
 286          * temporarily be negative.
 287          */
 288         return clean_zn_cnt >= 0 ? clean_zn_cnt : 1;
 289 }
 290
 291 unsigned long ubifs_shrink_scan(struct shrinker *shrink,
 292                                 struct shrink_control *sc)
 293 {
 294         unsigned long nr = sc->nr_to_scan;
 295         int contention = 0;
 296         unsigned long freed;
 297         long clean_zn_cnt = atomic_long_read(&ubifs_clean_zn_cnt);
 298
 299         if (!clean_zn_cnt) {
 300                 /*
 301                  * No clean znodes, nothing to reap. All we can do in this case
 302                  * is to kick background threads to start commit, which will
 303                  * probably make clean znodes which, in turn, will be freeable.
 304                  * And we return -1 which means will make VM call us again
 305                  * later.
 306                  */
 307                 dbg_tnc("no clean znodes, kick a thread");
 308                 return kick_a_thread();
 309         }
 310
 311         freed = shrink_tnc_trees(nr, OLD_ZNODE_AGE, &contention);
 312         if (freed >= nr)
 313                 goto out;
 314
 315         dbg_tnc("not enough old znodes, try to free young ones");
 316         freed += shrink_tnc_trees(nr - freed, YOUNG_ZNODE_AGE, &contention);
 317         if (freed >= nr)
 318                 goto out;
 319
 320         dbg_tnc("not enough young znodes, free all");
 321         freed += shrink_tnc_trees(nr - freed, 0, &contention);
 322
 323         if (!freed && contention) {
 324                 dbg_tnc("freed nothing, but contention");
 325                 return SHRINK_STOP;
 326         }
 327
 328 out:
 329         dbg_tnc("%lu znodes were freed, requested %lu", freed, nr);
 330         return freed;
 331 }