Pileus Git - ~andy/linux/blob - fs/f2fs/node.c

   1 /*
   2  * fs/f2fs/node.c
   3  *
   4  * Copyright (c) 2012 Samsung Electronics Co., Ltd.
   5  *             http://www.samsung.com/
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License version 2 as
   9  * published by the Free Software Foundation.
  10  */
  11 #include <linux/fs.h>
  12 #include <linux/f2fs_fs.h>
  13 #include <linux/mpage.h>
  14 #include <linux/backing-dev.h>
  15 #include <linux/blkdev.h>
  16 #include <linux/pagevec.h>
  17 #include <linux/swap.h>
  18
  19 #include "f2fs.h"
  20 #include "node.h"
  21 #include "segment.h"
  22 #include <trace/events/f2fs.h>
  23
  24 static struct kmem_cache *nat_entry_slab;
  25 static struct kmem_cache *free_nid_slab;
  26
  27 static void clear_node_page_dirty(struct page *page)
  28 {
  29         struct address_space *mapping = page->mapping;
  30         struct f2fs_sb_info *sbi = F2FS_SB(mapping->host->i_sb);
  31         unsigned int long flags;
  32
  33         if (PageDirty(page)) {
  34                 spin_lock_irqsave(&mapping->tree_lock, flags);
  35                 radix_tree_tag_clear(&mapping->page_tree,
  36                                 page_index(page),
  37                                 PAGECACHE_TAG_DIRTY);
  38                 spin_unlock_irqrestore(&mapping->tree_lock, flags);
  39
  40                 clear_page_dirty_for_io(page);
  41                 dec_page_count(sbi, F2FS_DIRTY_NODES);
  42         }
  43         ClearPageUptodate(page);
  44 }
  45
  46 static struct page *get_current_nat_page(struct f2fs_sb_info *sbi, nid_t nid)
  47 {
  48         pgoff_t index = current_nat_addr(sbi, nid);
  49         return get_meta_page(sbi, index);
  50 }
  51
  52 static struct page *get_next_nat_page(struct f2fs_sb_info *sbi, nid_t nid)
  53 {
  54         struct page *src_page;
  55         struct page *dst_page;
  56         pgoff_t src_off;
  57         pgoff_t dst_off;
  58         void *src_addr;
  59         void *dst_addr;
  60         struct f2fs_nm_info *nm_i = NM_I(sbi);
  61
  62         src_off = current_nat_addr(sbi, nid);
  63         dst_off = next_nat_addr(sbi, src_off);
  64
  65         /* get current nat block page with lock */
  66         src_page = get_meta_page(sbi, src_off);
  67
  68         /* Dirty src_page means that it is already the new target NAT page. */
  69         if (PageDirty(src_page))
  70                 return src_page;
  71
  72         dst_page = grab_meta_page(sbi, dst_off);
  73
  74         src_addr = page_address(src_page);
  75         dst_addr = page_address(dst_page);
  76         memcpy(dst_addr, src_addr, PAGE_CACHE_SIZE);
  77         set_page_dirty(dst_page);
  78         f2fs_put_page(src_page, 1);
  79
  80         set_to_next_nat(nm_i, nid);
  81
  82         return dst_page;
  83 }
  84
  85 /*
  86  * Readahead NAT pages
  87  */
  88 static void ra_nat_pages(struct f2fs_sb_info *sbi, int nid)
  89 {
  90         struct address_space *mapping = META_MAPPING(sbi);
  91         struct f2fs_nm_info *nm_i = NM_I(sbi);
  92         struct page *page;
  93         pgoff_t index;
  94         int i;
  95         struct f2fs_io_info fio = {
  96                 .type = META,
  97                 .rw = READ_SYNC | REQ_META | REQ_PRIO
  98         };
  99
 100
 101         for (i = 0; i < FREE_NID_PAGES; i++, nid += NAT_ENTRY_PER_BLOCK) {
 102                 if (unlikely(nid >= nm_i->max_nid))
 103                         nid = 0;
 104                 index = current_nat_addr(sbi, nid);
 105
 106                 page = grab_cache_page(mapping, index);
 107                 if (!page)
 108                         continue;
 109                 if (PageUptodate(page)) {
 110                         mark_page_accessed(page);
 111                         f2fs_put_page(page, 1);
 112                         continue;
 113                 }
 114                 f2fs_submit_page_mbio(sbi, page, index, &fio);
 115                 mark_page_accessed(page);
 116                 f2fs_put_page(page, 0);
 117         }
 118         f2fs_submit_merged_bio(sbi, META, READ);
 119 }
 120
 121 static struct nat_entry *__lookup_nat_cache(struct f2fs_nm_info *nm_i, nid_t n)
 122 {
 123         return radix_tree_lookup(&nm_i->nat_root, n);
 124 }
 125
 126 static unsigned int __gang_lookup_nat_cache(struct f2fs_nm_info *nm_i,
 127                 nid_t start, unsigned int nr, struct nat_entry **ep)
 128 {
 129         return radix_tree_gang_lookup(&nm_i->nat_root, (void **)ep, start, nr);
 130 }
 131
 132 static void __del_from_nat_cache(struct f2fs_nm_info *nm_i, struct nat_entry *e)
 133 {
 134         list_del(&e->list);
 135         radix_tree_delete(&nm_i->nat_root, nat_get_nid(e));
 136         nm_i->nat_cnt--;
 137         kmem_cache_free(nat_entry_slab, e);
 138 }
 139
 140 int is_checkpointed_node(struct f2fs_sb_info *sbi, nid_t nid)
 141 {
 142         struct f2fs_nm_info *nm_i = NM_I(sbi);
 143         struct nat_entry *e;
 144         int is_cp = 1;
 145
 146         read_lock(&nm_i->nat_tree_lock);
 147         e = __lookup_nat_cache(nm_i, nid);
 148         if (e && !e->checkpointed)
 149                 is_cp = 0;
 150         read_unlock(&nm_i->nat_tree_lock);
 151         return is_cp;
 152 }
 153
 154 static struct nat_entry *grab_nat_entry(struct f2fs_nm_info *nm_i, nid_t nid)
 155 {
 156         struct nat_entry *new;
 157
 158         new = kmem_cache_alloc(nat_entry_slab, GFP_ATOMIC);
 159         if (!new)
 160                 return NULL;
 161         if (radix_tree_insert(&nm_i->nat_root, nid, new)) {
 162                 kmem_cache_free(nat_entry_slab, new);
 163                 return NULL;
 164         }
 165         memset(new, 0, sizeof(struct nat_entry));
 166         nat_set_nid(new, nid);
 167         list_add_tail(&new->list, &nm_i->nat_entries);
 168         nm_i->nat_cnt++;
 169         return new;
 170 }
 171
 172 static void cache_nat_entry(struct f2fs_nm_info *nm_i, nid_t nid,
 173                                                 struct f2fs_nat_entry *ne)
 174 {
 175         struct nat_entry *e;
 176 retry:
 177         write_lock(&nm_i->nat_tree_lock);
 178         e = __lookup_nat_cache(nm_i, nid);
 179         if (!e) {
 180                 e = grab_nat_entry(nm_i, nid);
 181                 if (!e) {
 182                         write_unlock(&nm_i->nat_tree_lock);
 183                         goto retry;
 184                 }
 185                 nat_set_blkaddr(e, le32_to_cpu(ne->block_addr));
 186                 nat_set_ino(e, le32_to_cpu(ne->ino));
 187                 nat_set_version(e, ne->version);
 188                 e->checkpointed = true;
 189         }
 190         write_unlock(&nm_i->nat_tree_lock);
 191 }
 192
 193 static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
 194                         block_t new_blkaddr)
 195 {
 196         struct f2fs_nm_info *nm_i = NM_I(sbi);
 197         struct nat_entry *e;
 198 retry:
 199         write_lock(&nm_i->nat_tree_lock);
 200         e = __lookup_nat_cache(nm_i, ni->nid);
 201         if (!e) {
 202                 e = grab_nat_entry(nm_i, ni->nid);
 203                 if (!e) {
 204                         write_unlock(&nm_i->nat_tree_lock);
 205                         goto retry;
 206                 }
 207                 e->ni = *ni;
 208                 e->checkpointed = true;
 209                 f2fs_bug_on(ni->blk_addr == NEW_ADDR);
 210         } else if (new_blkaddr == NEW_ADDR) {
 211                 /*
 212                  * when nid is reallocated,
 213                  * previous nat entry can be remained in nat cache.
 214                  * So, reinitialize it with new information.
 215                  */
 216                 e->ni = *ni;
 217                 f2fs_bug_on(ni->blk_addr != NULL_ADDR);
 218         }
 219
 220         if (new_blkaddr == NEW_ADDR)
 221                 e->checkpointed = false;
 222
 223         /* sanity check */
 224         f2fs_bug_on(nat_get_blkaddr(e) != ni->blk_addr);
 225         f2fs_bug_on(nat_get_blkaddr(e) == NULL_ADDR &&
 226                         new_blkaddr == NULL_ADDR);
 227         f2fs_bug_on(nat_get_blkaddr(e) == NEW_ADDR &&
 228                         new_blkaddr == NEW_ADDR);
 229         f2fs_bug_on(nat_get_blkaddr(e) != NEW_ADDR &&
 230                         nat_get_blkaddr(e) != NULL_ADDR &&
 231                         new_blkaddr == NEW_ADDR);
 232
 233         /* increament version no as node is removed */
 234         if (nat_get_blkaddr(e) != NEW_ADDR && new_blkaddr == NULL_ADDR) {
 235                 unsigned char version = nat_get_version(e);
 236                 nat_set_version(e, inc_node_version(version));
 237         }
 238
 239         /* change address */
 240         nat_set_blkaddr(e, new_blkaddr);
 241         __set_nat_cache_dirty(nm_i, e);
 242         write_unlock(&nm_i->nat_tree_lock);
 243 }
 244
 245 int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
 246 {
 247         struct f2fs_nm_info *nm_i = NM_I(sbi);
 248
 249         if (nm_i->nat_cnt <= NM_WOUT_THRESHOLD)
 250                 return 0;
 251
 252         write_lock(&nm_i->nat_tree_lock);
 253         while (nr_shrink && !list_empty(&nm_i->nat_entries)) {
 254                 struct nat_entry *ne;
 255                 ne = list_first_entry(&nm_i->nat_entries,
 256                                         struct nat_entry, list);
 257                 __del_from_nat_cache(nm_i, ne);
 258                 nr_shrink--;
 259         }
 260         write_unlock(&nm_i->nat_tree_lock);
 261         return nr_shrink;
 262 }
 263
 264 /*
 265  * This function returns always success
 266  */
 267 void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni)
 268 {
 269         struct f2fs_nm_info *nm_i = NM_I(sbi);
 270         struct curseg_info *curseg = CURSEG_I(sbi, CURSEG_HOT_DATA);
 271         struct f2fs_summary_block *sum = curseg->sum_blk;
 272         nid_t start_nid = START_NID(nid);
 273         struct f2fs_nat_block *nat_blk;
 274         struct page *page = NULL;
 275         struct f2fs_nat_entry ne;
 276         struct nat_entry *e;
 277         int i;
 278
 279         memset(&ne, 0, sizeof(struct f2fs_nat_entry));
 280         ni->nid = nid;
 281
 282         /* Check nat cache */
 283         read_lock(&nm_i->nat_tree_lock);
 284         e = __lookup_nat_cache(nm_i, nid);
 285         if (e) {
 286                 ni->ino = nat_get_ino(e);
 287                 ni->blk_addr = nat_get_blkaddr(e);
 288                 ni->version = nat_get_version(e);
 289         }
 290         read_unlock(&nm_i->nat_tree_lock);
 291         if (e)
 292                 return;
 293
 294         /* Check current segment summary */
 295         mutex_lock(&curseg->curseg_mutex);
 296         i = lookup_journal_in_cursum(sum, NAT_JOURNAL, nid, 0);
 297         if (i >= 0) {
 298                 ne = nat_in_journal(sum, i);
 299                 node_info_from_raw_nat(ni, &ne);
 300         }
 301         mutex_unlock(&curseg->curseg_mutex);
 302         if (i >= 0)
 303                 goto cache;
 304
 305         /* Fill node_info from nat page */
 306         page = get_current_nat_page(sbi, start_nid);
 307         nat_blk = (struct f2fs_nat_block *)page_address(page);
 308         ne = nat_blk->entries[nid - start_nid];
 309         node_info_from_raw_nat(ni, &ne);
 310         f2fs_put_page(page, 1);
 311 cache:
 312         /* cache nat entry */
 313         cache_nat_entry(NM_I(sbi), nid, &ne);
 314 }
 315
 316 /*
 317  * The maximum depth is four.
 318  * Offset[0] will have raw inode offset.
 319  */
 320 static int get_node_path(struct f2fs_inode_info *fi, long block,
 321                                 int offset[4], unsigned int noffset[4])
 322 {
 323         const long direct_index = ADDRS_PER_INODE(fi);
 324         const long direct_blks = ADDRS_PER_BLOCK;
 325         const long dptrs_per_blk = NIDS_PER_BLOCK;
 326         const long indirect_blks = ADDRS_PER_BLOCK * NIDS_PER_BLOCK;
 327         const long dindirect_blks = indirect_blks * NIDS_PER_BLOCK;
 328         int n = 0;
 329         int level = 0;
 330
 331         noffset[0] = 0;
 332
 333         if (block < direct_index) {
 334                 offset[n] = block;
 335                 goto got;
 336         }
 337         block -= direct_index;
 338         if (block < direct_blks) {
 339                 offset[n++] = NODE_DIR1_BLOCK;
 340                 noffset[n] = 1;
 341                 offset[n] = block;
 342                 level = 1;
 343                 goto got;
 344         }
 345         block -= direct_blks;
 346         if (block < direct_blks) {
 347                 offset[n++] = NODE_DIR2_BLOCK;
 348                 noffset[n] = 2;
 349                 offset[n] = block;
 350                 level = 1;
 351                 goto got;
 352         }
 353         block -= direct_blks;
 354         if (block < indirect_blks) {
 355                 offset[n++] = NODE_IND1_BLOCK;
 356                 noffset[n] = 3;
 357                 offset[n++] = block / direct_blks;
 358                 noffset[n] = 4 + offset[n - 1];
 359                 offset[n] = block % direct_blks;
 360                 level = 2;
 361                 goto got;
 362         }
 363         block -= indirect_blks;
 364         if (block < indirect_blks) {
 365                 offset[n++] = NODE_IND2_BLOCK;
 366                 noffset[n] = 4 + dptrs_per_blk;
 367                 offset[n++] = block / direct_blks;
 368                 noffset[n] = 5 + dptrs_per_blk + offset[n - 1];
 369                 offset[n] = block % direct_blks;
 370                 level = 2;
 371                 goto got;
 372         }
 373         block -= indirect_blks;
 374         if (block < dindirect_blks) {
 375                 offset[n++] = NODE_DIND_BLOCK;
 376                 noffset[n] = 5 + (dptrs_per_blk * 2);
 377                 offset[n++] = block / indirect_blks;
 378                 noffset[n] = 6 + (dptrs_per_blk * 2) +
 379                               offset[n - 1] * (dptrs_per_blk + 1);
 380                 offset[n++] = (block / direct_blks) % dptrs_per_blk;
 381                 noffset[n] = 7 + (dptrs_per_blk * 2) +
 382                               offset[n - 2] * (dptrs_per_blk + 1) +
 383                               offset[n - 1];
 384                 offset[n] = block % direct_blks;
 385                 level = 3;
 386                 goto got;
 387         } else {
 388                 BUG();
 389         }
 390 got:
 391         return level;
 392 }
 393
 394 /*
 395  * Caller should call f2fs_put_dnode(dn).
 396  * Also, it should grab and release a rwsem by calling f2fs_lock_op() and
 397  * f2fs_unlock_op() only if ro is not set RDONLY_NODE.
 398  * In the case of RDONLY_NODE, we don't need to care about mutex.
 399  */
 400 int get_dnode_of_data(struct dnode_of_data *dn, pgoff_t index, int mode)
 401 {
 402         struct f2fs_sb_info *sbi = F2FS_SB(dn->inode->i_sb);
 403         struct page *npage[4];
 404         struct page *parent;
 405         int offset[4];
 406         unsigned int noffset[4];
 407         nid_t nids[4];
 408         int level, i;
 409         int err = 0;
 410
 411         level = get_node_path(F2FS_I(dn->inode), index, offset, noffset);
 412
 413         nids[0] = dn->inode->i_ino;
 414         npage[0] = dn->inode_page;
 415
 416         if (!npage[0]) {
 417                 npage[0] = get_node_page(sbi, nids[0]);
 418                 if (IS_ERR(npage[0]))
 419                         return PTR_ERR(npage[0]);
 420         }
 421         parent = npage[0];
 422         if (level != 0)
 423                 nids[1] = get_nid(parent, offset[0], true);
 424         dn->inode_page = npage[0];
 425         dn->inode_page_locked = true;
 426
 427         /* get indirect or direct nodes */
 428         for (i = 1; i <= level; i++) {
 429                 bool done = false;
 430
 431                 if (!nids[i] && mode == ALLOC_NODE) {
 432                         /* alloc new node */
 433                         if (!alloc_nid(sbi, &(nids[i]))) {
 434                                 err = -ENOSPC;
 435                                 goto release_pages;
 436                         }
 437
 438                         dn->nid = nids[i];
 439                         npage[i] = new_node_page(dn, noffset[i], NULL);
 440                         if (IS_ERR(npage[i])) {
 441                                 alloc_nid_failed(sbi, nids[i]);
 442                                 err = PTR_ERR(npage[i]);
 443                                 goto release_pages;
 444                         }
 445
 446                         set_nid(parent, offset[i - 1], nids[i], i == 1);
 447                         alloc_nid_done(sbi, nids[i]);
 448                         done = true;
 449                 } else if (mode == LOOKUP_NODE_RA && i == level && level > 1) {
 450                         npage[i] = get_node_page_ra(parent, offset[i - 1]);
 451                         if (IS_ERR(npage[i])) {
 452                                 err = PTR_ERR(npage[i]);
 453                                 goto release_pages;
 454                         }
 455                         done = true;
 456                 }
 457                 if (i == 1) {
 458                         dn->inode_page_locked = false;
 459                         unlock_page(parent);
 460                 } else {
 461                         f2fs_put_page(parent, 1);
 462                 }
 463
 464                 if (!done) {
 465                         npage[i] = get_node_page(sbi, nids[i]);
 466                         if (IS_ERR(npage[i])) {
 467                                 err = PTR_ERR(npage[i]);
 468                                 f2fs_put_page(npage[0], 0);
 469                                 goto release_out;
 470                         }
 471                 }
 472                 if (i < level) {
 473                         parent = npage[i];
 474                         nids[i + 1] = get_nid(parent, offset[i], false);
 475                 }
 476         }
 477         dn->nid = nids[level];
 478         dn->ofs_in_node = offset[level];
 479         dn->node_page = npage[level];
 480         dn->data_blkaddr = datablock_addr(dn->node_page, dn->ofs_in_node);
 481         return 0;
 482
 483 release_pages:
 484         f2fs_put_page(parent, 1);
 485         if (i > 1)
 486                 f2fs_put_page(npage[0], 0);
 487 release_out:
 488         dn->inode_page = NULL;
 489         dn->node_page = NULL;
 490         return err;
 491 }
 492
 493 static void truncate_node(struct dnode_of_data *dn)
 494 {
 495         struct f2fs_sb_info *sbi = F2FS_SB(dn->inode->i_sb);
 496         struct node_info ni;
 497
 498         get_node_info(sbi, dn->nid, &ni);
 499         if (dn->inode->i_blocks == 0) {
 500                 f2fs_bug_on(ni.blk_addr != NULL_ADDR);
 501                 goto invalidate;
 502         }
 503         f2fs_bug_on(ni.blk_addr == NULL_ADDR);
 504
 505         /* Deallocate node address */
 506         invalidate_blocks(sbi, ni.blk_addr);
 507         dec_valid_node_count(sbi, dn->inode);
 508         set_node_addr(sbi, &ni, NULL_ADDR);
 509
 510         if (dn->nid == dn->inode->i_ino) {
 511                 remove_orphan_inode(sbi, dn->nid);
 512                 dec_valid_inode_count(sbi);
 513         } else {
 514                 sync_inode_page(dn);
 515         }
 516 invalidate:
 517         clear_node_page_dirty(dn->node_page);
 518         F2FS_SET_SB_DIRT(sbi);
 519
 520         f2fs_put_page(dn->node_page, 1);
 521
 522         invalidate_mapping_pages(NODE_MAPPING(sbi),
 523                         dn->node_page->index, dn->node_page->index);
 524
 525         dn->node_page = NULL;
 526         trace_f2fs_truncate_node(dn->inode, dn->nid, ni.blk_addr);
 527 }
 528
 529 static int truncate_dnode(struct dnode_of_data *dn)
 530 {
 531         struct f2fs_sb_info *sbi = F2FS_SB(dn->inode->i_sb);
 532         struct page *page;
 533
 534         if (dn->nid == 0)
 535                 return 1;
 536
 537         /* get direct node */
 538         page = get_node_page(sbi, dn->nid);
 539         if (IS_ERR(page) && PTR_ERR(page) == -ENOENT)
 540                 return 1;
 541         else if (IS_ERR(page))
 542                 return PTR_ERR(page);
 543
 544         /* Make dnode_of_data for parameter */
 545         dn->node_page = page;
 546         dn->ofs_in_node = 0;
 547         truncate_data_blocks(dn);
 548         truncate_node(dn);
 549         return 1;
 550 }
 551
 552 static int truncate_nodes(struct dnode_of_data *dn, unsigned int nofs,
 553                                                 int ofs, int depth)
 554 {
 555         struct f2fs_sb_info *sbi = F2FS_SB(dn->inode->i_sb);
 556         struct dnode_of_data rdn = *dn;
 557         struct page *page;
 558         struct f2fs_node *rn;
 559         nid_t child_nid;
 560         unsigned int child_nofs;
 561         int freed = 0;
 562         int i, ret;
 563
 564         if (dn->nid == 0)
 565                 return NIDS_PER_BLOCK + 1;
 566
 567         trace_f2fs_truncate_nodes_enter(dn->inode, dn->nid, dn->data_blkaddr);
 568
 569         page = get_node_page(sbi, dn->nid);
 570         if (IS_ERR(page)) {
 571                 trace_f2fs_truncate_nodes_exit(dn->inode, PTR_ERR(page));
 572                 return PTR_ERR(page);
 573         }
 574
 575         rn = F2FS_NODE(page);
 576         if (depth < 3) {
 577                 for (i = ofs; i < NIDS_PER_BLOCK; i++, freed++) {
 578                         child_nid = le32_to_cpu(rn->in.nid[i]);
 579                         if (child_nid == 0)
 580                                 continue;
 581                         rdn.nid = child_nid;
 582                         ret = truncate_dnode(&rdn);
 583                         if (ret < 0)
 584                                 goto out_err;
 585                         set_nid(page, i, 0, false);
 586                 }
 587         } else {
 588                 child_nofs = nofs + ofs * (NIDS_PER_BLOCK + 1) + 1;
 589                 for (i = ofs; i < NIDS_PER_BLOCK; i++) {
 590                         child_nid = le32_to_cpu(rn->in.nid[i]);
 591                         if (child_nid == 0) {
 592                                 child_nofs += NIDS_PER_BLOCK + 1;
 593                                 continue;
 594                         }
 595                         rdn.nid = child_nid;
 596                         ret = truncate_nodes(&rdn, child_nofs, 0, depth - 1);
 597                         if (ret == (NIDS_PER_BLOCK + 1)) {
 598                                 set_nid(page, i, 0, false);
 599                                 child_nofs += ret;
 600                         } else if (ret < 0 && ret != -ENOENT) {
 601                                 goto out_err;
 602                         }
 603                 }
 604                 freed = child_nofs;
 605         }
 606
 607         if (!ofs) {
 608                 /* remove current indirect node */
 609                 dn->node_page = page;
 610                 truncate_node(dn);
 611                 freed++;
 612         } else {
 613                 f2fs_put_page(page, 1);
 614         }
 615         trace_f2fs_truncate_nodes_exit(dn->inode, freed);
 616         return freed;
 617
 618 out_err:
 619         f2fs_put_page(page, 1);
 620         trace_f2fs_truncate_nodes_exit(dn->inode, ret);
 621         return ret;
 622 }
 623
 624 static int truncate_partial_nodes(struct dnode_of_data *dn,
 625                         struct f2fs_inode *ri, int *offset, int depth)
 626 {
 627         struct f2fs_sb_info *sbi = F2FS_SB(dn->inode->i_sb);
 628         struct page *pages[2];
 629         nid_t nid[3];
 630         nid_t child_nid;
 631         int err = 0;
 632         int i;
 633         int idx = depth - 2;
 634
 635         nid[0] = le32_to_cpu(ri->i_nid[offset[0] - NODE_DIR1_BLOCK]);
 636         if (!nid[0])
 637                 return 0;
 638
 639         /* get indirect nodes in the path */
 640         for (i = 0; i < idx + 1; i++) {
 641                 /* refernece count'll be increased */
 642                 pages[i] = get_node_page(sbi, nid[i]);
 643                 if (IS_ERR(pages[i])) {
 644                         err = PTR_ERR(pages[i]);
 645                         idx = i - 1;
 646                         goto fail;
 647                 }
 648                 nid[i + 1] = get_nid(pages[i], offset[i + 1], false);
 649         }
 650
 651         /* free direct nodes linked to a partial indirect node */
 652         for (i = offset[idx + 1]; i < NIDS_PER_BLOCK; i++) {
 653                 child_nid = get_nid(pages[idx], i, false);
 654                 if (!child_nid)
 655                         continue;
 656                 dn->nid = child_nid;
 657                 err = truncate_dnode(dn);
 658                 if (err < 0)
 659                         goto fail;
 660                 set_nid(pages[idx], i, 0, false);
 661         }
 662
 663         if (offset[idx + 1] == 0) {
 664                 dn->node_page = pages[idx];
 665                 dn->nid = nid[idx];
 666                 truncate_node(dn);
 667         } else {
 668                 f2fs_put_page(pages[idx], 1);
 669         }
 670         offset[idx]++;
 671         offset[idx + 1] = 0;
 672         idx--;
 673 fail:
 674         for (i = idx; i >= 0; i--)
 675                 f2fs_put_page(pages[i], 1);
 676
 677         trace_f2fs_truncate_partial_nodes(dn->inode, nid, depth, err);
 678
 679         return err;
 680 }
 681
 682 /*
 683  * All the block addresses of data and nodes should be nullified.
 684  */
 685 int truncate_inode_blocks(struct inode *inode, pgoff_t from)
 686 {
 687         struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 688         int err = 0, cont = 1;
 689         int level, offset[4], noffset[4];
 690         unsigned int nofs = 0;
 691         struct f2fs_inode *ri;
 692         struct dnode_of_data dn;
 693         struct page *page;
 694
 695         trace_f2fs_truncate_inode_blocks_enter(inode, from);
 696
 697         level = get_node_path(F2FS_I(inode), from, offset, noffset);
 698 restart:
 699         page = get_node_page(sbi, inode->i_ino);
 700         if (IS_ERR(page)) {
 701                 trace_f2fs_truncate_inode_blocks_exit(inode, PTR_ERR(page));
 702                 return PTR_ERR(page);
 703         }
 704
 705         set_new_dnode(&dn, inode, page, NULL, 0);
 706         unlock_page(page);
 707
 708         ri = F2FS_INODE(page);
 709         switch (level) {
 710         case 0:
 711         case 1:
 712                 nofs = noffset[1];
 713                 break;
 714         case 2:
 715                 nofs = noffset[1];
 716                 if (!offset[level - 1])
 717                         goto skip_partial;
 718                 err = truncate_partial_nodes(&dn, ri, offset, level);
 719                 if (err < 0 && err != -ENOENT)
 720                         goto fail;
 721                 nofs += 1 + NIDS_PER_BLOCK;
 722                 break;
 723         case 3:
 724                 nofs = 5 + 2 * NIDS_PER_BLOCK;
 725                 if (!offset[level - 1])
 726                         goto skip_partial;
 727                 err = truncate_partial_nodes(&dn, ri, offset, level);
 728                 if (err < 0 && err != -ENOENT)
 729                         goto fail;
 730                 break;
 731         default:
 732                 BUG();
 733         }
 734
 735 skip_partial:
 736         while (cont) {
 737                 dn.nid = le32_to_cpu(ri->i_nid[offset[0] - NODE_DIR1_BLOCK]);
 738                 switch (offset[0]) {
 739                 case NODE_DIR1_BLOCK:
 740                 case NODE_DIR2_BLOCK:
 741                         err = truncate_dnode(&dn);
 742                         break;
 743
 744                 case NODE_IND1_BLOCK:
 745                 case NODE_IND2_BLOCK:
 746                         err = truncate_nodes(&dn, nofs, offset[1], 2);
 747                         break;
 748
 749                 case NODE_DIND_BLOCK:
 750                         err = truncate_nodes(&dn, nofs, offset[1], 3);
 751                         cont = 0;
 752                         break;
 753
 754                 default:
 755                         BUG();
 756                 }
 757                 if (err < 0 && err != -ENOENT)
 758                         goto fail;
 759                 if (offset[1] == 0 &&
 760                                 ri->i_nid[offset[0] - NODE_DIR1_BLOCK]) {
 761                         lock_page(page);
 762                         if (unlikely(page->mapping != NODE_MAPPING(sbi))) {
 763                                 f2fs_put_page(page, 1);
 764                                 goto restart;
 765                         }
 766                         wait_on_page_writeback(page);
 767                         ri->i_nid[offset[0] - NODE_DIR1_BLOCK] = 0;
 768                         set_page_dirty(page);
 769                         unlock_page(page);
 770                 }
 771                 offset[1] = 0;
 772                 offset[0]++;
 773                 nofs += err;
 774         }
 775 fail:
 776         f2fs_put_page(page, 0);
 777         trace_f2fs_truncate_inode_blocks_exit(inode, err);
 778         return err > 0 ? 0 : err;
 779 }
 780
 781 int truncate_xattr_node(struct inode *inode, struct page *page)
 782 {
 783         struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 784         nid_t nid = F2FS_I(inode)->i_xattr_nid;
 785         struct dnode_of_data dn;
 786         struct page *npage;
 787
 788         if (!nid)
 789                 return 0;
 790
 791         npage = get_node_page(sbi, nid);
 792         if (IS_ERR(npage))
 793                 return PTR_ERR(npage);
 794
 795         F2FS_I(inode)->i_xattr_nid = 0;
 796
 797         /* need to do checkpoint during fsync */
 798         F2FS_I(inode)->xattr_ver = cur_cp_version(F2FS_CKPT(sbi));
 799
 800         set_new_dnode(&dn, inode, page, npage, nid);
 801
 802         if (page)
 803                 dn.inode_page_locked = true;
 804         truncate_node(&dn);
 805         return 0;
 806 }
 807
 808 /*
 809  * Caller should grab and release a rwsem by calling f2fs_lock_op() and
 810  * f2fs_unlock_op().
 811  */
 812 void remove_inode_page(struct inode *inode)
 813 {
 814         struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 815         struct page *page;
 816         nid_t ino = inode->i_ino;
 817         struct dnode_of_data dn;
 818
 819         page = get_node_page(sbi, ino);
 820         if (IS_ERR(page))
 821                 return;
 822
 823         if (truncate_xattr_node(inode, page)) {
 824                 f2fs_put_page(page, 1);
 825                 return;
 826         }
 827         /* 0 is possible, after f2fs_new_inode() is failed */
 828         f2fs_bug_on(inode->i_blocks != 0 && inode->i_blocks != 1);
 829         set_new_dnode(&dn, inode, page, page, ino);
 830         truncate_node(&dn);
 831 }
 832
 833 struct page *new_inode_page(struct inode *inode, const struct qstr *name)
 834 {
 835         struct dnode_of_data dn;
 836
 837         /* allocate inode page for new inode */
 838         set_new_dnode(&dn, inode, NULL, NULL, inode->i_ino);
 839
 840         /* caller should f2fs_put_page(page, 1); */
 841         return new_node_page(&dn, 0, NULL);
 842 }
 843
 844 struct page *new_node_page(struct dnode_of_data *dn,
 845                                 unsigned int ofs, struct page *ipage)
 846 {
 847         struct f2fs_sb_info *sbi = F2FS_SB(dn->inode->i_sb);
 848         struct node_info old_ni, new_ni;
 849         struct page *page;
 850         int err;
 851
 852         if (unlikely(is_inode_flag_set(F2FS_I(dn->inode), FI_NO_ALLOC)))
 853                 return ERR_PTR(-EPERM);
 854
 855         page = grab_cache_page(NODE_MAPPING(sbi), dn->nid);
 856         if (!page)
 857                 return ERR_PTR(-ENOMEM);
 858
 859         if (unlikely(!inc_valid_node_count(sbi, dn->inode))) {
 860                 err = -ENOSPC;
 861                 goto fail;
 862         }
 863
 864         get_node_info(sbi, dn->nid, &old_ni);
 865
 866         /* Reinitialize old_ni with new node page */
 867         f2fs_bug_on(old_ni.blk_addr != NULL_ADDR);
 868         new_ni = old_ni;
 869         new_ni.ino = dn->inode->i_ino;
 870         set_node_addr(sbi, &new_ni, NEW_ADDR);
 871
 872         fill_node_footer(page, dn->nid, dn->inode->i_ino, ofs, true);
 873         set_cold_node(dn->inode, page);
 874         SetPageUptodate(page);
 875         set_page_dirty(page);
 876
 877         if (ofs == XATTR_NODE_OFFSET)
 878                 F2FS_I(dn->inode)->i_xattr_nid = dn->nid;
 879
 880         dn->node_page = page;
 881         if (ipage)
 882                 update_inode(dn->inode, ipage);
 883         else
 884                 sync_inode_page(dn);
 885         if (ofs == 0)
 886                 inc_valid_inode_count(sbi);
 887
 888         return page;
 889
 890 fail:
 891         clear_node_page_dirty(page);
 892         f2fs_put_page(page, 1);
 893         return ERR_PTR(err);
 894 }
 895
 896 /*
 897  * Caller should do after getting the following values.
 898  * 0: f2fs_put_page(page, 0)
 899  * LOCKED_PAGE: f2fs_put_page(page, 1)
 900  * error: nothing
 901  */
 902 static int read_node_page(struct page *page, int rw)
 903 {
 904         struct f2fs_sb_info *sbi = F2FS_SB(page->mapping->host->i_sb);
 905         struct node_info ni;
 906
 907         get_node_info(sbi, page->index, &ni);
 908
 909         if (unlikely(ni.blk_addr == NULL_ADDR)) {
 910                 f2fs_put_page(page, 1);
 911                 return -ENOENT;
 912         }
 913
 914         if (PageUptodate(page))
 915                 return LOCKED_PAGE;
 916
 917         return f2fs_submit_page_bio(sbi, page, ni.blk_addr, rw);
 918 }
 919
 920 /*
 921  * Readahead a node page
 922  */
 923 void ra_node_page(struct f2fs_sb_info *sbi, nid_t nid)
 924 {
 925         struct page *apage;
 926         int err;
 927
 928         apage = find_get_page(NODE_MAPPING(sbi), nid);
 929         if (apage && PageUptodate(apage)) {
 930                 f2fs_put_page(apage, 0);
 931                 return;
 932         }
 933         f2fs_put_page(apage, 0);
 934
 935         apage = grab_cache_page(NODE_MAPPING(sbi), nid);
 936         if (!apage)
 937                 return;
 938
 939         err = read_node_page(apage, READA);
 940         if (err == 0)
 941                 f2fs_put_page(apage, 0);
 942         else if (err == LOCKED_PAGE)
 943                 f2fs_put_page(apage, 1);
 944 }
 945
 946 struct page *get_node_page(struct f2fs_sb_info *sbi, pgoff_t nid)
 947 {
 948         struct page *page;
 949         int err;
 950 repeat:
 951         page = grab_cache_page(NODE_MAPPING(sbi), nid);
 952         if (!page)
 953                 return ERR_PTR(-ENOMEM);
 954
 955         err = read_node_page(page, READ_SYNC);
 956         if (err < 0)
 957                 return ERR_PTR(err);
 958         else if (err == LOCKED_PAGE)
 959                 goto got_it;
 960
 961         lock_page(page);
 962         if (unlikely(!PageUptodate(page))) {
 963                 f2fs_put_page(page, 1);
 964                 return ERR_PTR(-EIO);
 965         }
 966         if (unlikely(page->mapping != NODE_MAPPING(sbi))) {
 967                 f2fs_put_page(page, 1);
 968                 goto repeat;
 969         }
 970 got_it:
 971         f2fs_bug_on(nid != nid_of_node(page));
 972         mark_page_accessed(page);
 973         return page;
 974 }
 975
 976 /*
 977  * Return a locked page for the desired node page.
 978  * And, readahead MAX_RA_NODE number of node pages.
 979  */
 980 struct page *get_node_page_ra(struct page *parent, int start)
 981 {
 982         struct f2fs_sb_info *sbi = F2FS_SB(parent->mapping->host->i_sb);
 983         struct blk_plug plug;
 984         struct page *page;
 985         int err, i, end;
 986         nid_t nid;
 987
 988         /* First, try getting the desired direct node. */
 989         nid = get_nid(parent, start, false);
 990         if (!nid)
 991                 return ERR_PTR(-ENOENT);
 992 repeat:
 993         page = grab_cache_page(NODE_MAPPING(sbi), nid);
 994         if (!page)
 995                 return ERR_PTR(-ENOMEM);
 996
 997         err = read_node_page(page, READ_SYNC);
 998         if (err < 0)
 999                 return ERR_PTR(err);
1000         else if (err == LOCKED_PAGE)
1001                 goto page_hit;
1002
1003         blk_start_plug(&plug);
1004
1005         /* Then, try readahead for siblings of the desired node */
1006         end = start + MAX_RA_NODE;
1007         end = min(end, NIDS_PER_BLOCK);
1008         for (i = start + 1; i < end; i++) {
1009                 nid = get_nid(parent, i, false);
1010                 if (!nid)
1011                         continue;
1012                 ra_node_page(sbi, nid);
1013         }
1014
1015         blk_finish_plug(&plug);
1016
1017         lock_page(page);
1018         if (unlikely(page->mapping != NODE_MAPPING(sbi))) {
1019                 f2fs_put_page(page, 1);
1020                 goto repeat;
1021         }
1022 page_hit:
1023         if (unlikely(!PageUptodate(page))) {
1024                 f2fs_put_page(page, 1);
1025                 return ERR_PTR(-EIO);
1026         }
1027         mark_page_accessed(page);
1028         return page;
1029 }
1030
1031 void sync_inode_page(struct dnode_of_data *dn)
1032 {
1033         if (IS_INODE(dn->node_page) || dn->inode_page == dn->node_page) {
1034                 update_inode(dn->inode, dn->node_page);
1035         } else if (dn->inode_page) {
1036                 if (!dn->inode_page_locked)
1037                         lock_page(dn->inode_page);
1038                 update_inode(dn->inode, dn->inode_page);
1039                 if (!dn->inode_page_locked)
1040                         unlock_page(dn->inode_page);
1041         } else {
1042                 update_inode_page(dn->inode);
1043         }
1044 }
1045
1046 int sync_node_pages(struct f2fs_sb_info *sbi, nid_t ino,
1047                                         struct writeback_control *wbc)
1048 {
1049         pgoff_t index, end;
1050         struct pagevec pvec;
1051         int step = ino ? 2 : 0;
1052         int nwritten = 0, wrote = 0;
1053
1054         pagevec_init(&pvec, 0);
1055
1056 next_step:
1057         index = 0;
1058         end = LONG_MAX;
1059
1060         while (index <= end) {
1061                 int i, nr_pages;
1062                 nr_pages = pagevec_lookup_tag(&pvec, NODE_MAPPING(sbi), &index,
1063                                 PAGECACHE_TAG_DIRTY,
1064                                 min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1);
1065                 if (nr_pages == 0)
1066                         break;
1067
1068                 for (i = 0; i < nr_pages; i++) {
1069                         struct page *page = pvec.pages[i];
1070
1071                         /*
1072                          * flushing sequence with step:
1073                          * 0. indirect nodes
1074                          * 1. dentry dnodes
1075                          * 2. file dnodes
1076                          */
1077                         if (step == 0 && IS_DNODE(page))
1078                                 continue;
1079                         if (step == 1 && (!IS_DNODE(page) ||
1080                                                 is_cold_node(page)))
1081                                 continue;
1082                         if (step == 2 && (!IS_DNODE(page) ||
1083                                                 !is_cold_node(page)))
1084                                 continue;
1085
1086                         /*
1087                          * If an fsync mode,
1088                          * we should not skip writing node pages.
1089                          */
1090                         if (ino && ino_of_node(page) == ino)
1091                                 lock_page(page);
1092                         else if (!trylock_page(page))
1093                                 continue;
1094
1095                         if (unlikely(page->mapping != NODE_MAPPING(sbi))) {
1096 continue_unlock:
1097                                 unlock_page(page);
1098                                 continue;
1099                         }
1100                         if (ino && ino_of_node(page) != ino)
1101                                 goto continue_unlock;
1102
1103                         if (!PageDirty(page)) {
1104                                 /* someone wrote it for us */
1105                                 goto continue_unlock;
1106                         }
1107
1108                         if (!clear_page_dirty_for_io(page))
1109                                 goto continue_unlock;
1110
1111                         /* called by fsync() */
1112                         if (ino && IS_DNODE(page)) {
1113                                 int mark = !is_checkpointed_node(sbi, ino);
1114                                 set_fsync_mark(page, 1);
1115                                 if (IS_INODE(page))
1116                                         set_dentry_mark(page, mark);
1117                                 nwritten++;
1118                         } else {
1119                                 set_fsync_mark(page, 0);
1120                                 set_dentry_mark(page, 0);
1121                         }
1122                         NODE_MAPPING(sbi)->a_ops->writepage(page, wbc);
1123                         wrote++;
1124
1125                         if (--wbc->nr_to_write == 0)
1126                                 break;
1127                 }
1128                 pagevec_release(&pvec);
1129                 cond_resched();
1130
1131                 if (wbc->nr_to_write == 0) {
1132                         step = 2;
1133                         break;
1134                 }
1135         }
1136
1137         if (step < 2) {
1138                 step++;
1139                 goto next_step;
1140         }
1141
1142         if (wrote)
1143                 f2fs_submit_merged_bio(sbi, NODE, WRITE);
1144         return nwritten;
1145 }
1146
1147 int wait_on_node_pages_writeback(struct f2fs_sb_info *sbi, nid_t ino)
1148 {
1149         pgoff_t index = 0, end = LONG_MAX;
1150         struct pagevec pvec;
1151         int ret2 = 0, ret = 0;
1152
1153         pagevec_init(&pvec, 0);
1154
1155         while (index <= end) {
1156                 int i, nr_pages;
1157                 nr_pages = pagevec_lookup_tag(&pvec, NODE_MAPPING(sbi), &index,
1158                                 PAGECACHE_TAG_WRITEBACK,
1159                                 min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1);
1160                 if (nr_pages == 0)
1161                         break;
1162
1163                 for (i = 0; i < nr_pages; i++) {
1164                         struct page *page = pvec.pages[i];
1165
1166                         /* until radix tree lookup accepts end_index */
1167                         if (unlikely(page->index > end))
1168                                 continue;
1169
1170                         if (ino && ino_of_node(page) == ino) {
1171                                 wait_on_page_writeback(page);
1172                                 if (TestClearPageError(page))
1173                                         ret = -EIO;
1174                         }
1175                 }
1176                 pagevec_release(&pvec);
1177                 cond_resched();
1178         }
1179
1180         if (unlikely(test_and_clear_bit(AS_ENOSPC, &NODE_MAPPING(sbi)->flags)))
1181                 ret2 = -ENOSPC;
1182         if (unlikely(test_and_clear_bit(AS_EIO, &NODE_MAPPING(sbi)->flags)))
1183                 ret2 = -EIO;
1184         if (!ret)
1185                 ret = ret2;
1186         return ret;
1187 }
1188
1189 static int f2fs_write_node_page(struct page *page,
1190                                 struct writeback_control *wbc)
1191 {
1192         struct f2fs_sb_info *sbi = F2FS_SB(page->mapping->host->i_sb);
1193         nid_t nid;
1194         block_t new_addr;
1195         struct node_info ni;
1196         struct f2fs_io_info fio = {
1197                 .type = NODE,
1198                 .rw = (wbc->sync_mode == WB_SYNC_ALL) ? WRITE_SYNC : WRITE,
1199         };
1200
1201         if (unlikely(sbi->por_doing))
1202                 goto redirty_out;
1203
1204         wait_on_page_writeback(page);
1205
1206         /* get old block addr of this node page */
1207         nid = nid_of_node(page);
1208         f2fs_bug_on(page->index != nid);
1209
1210         get_node_info(sbi, nid, &ni);
1211
1212         /* This page is already truncated */
1213         if (unlikely(ni.blk_addr == NULL_ADDR)) {
1214                 dec_page_count(sbi, F2FS_DIRTY_NODES);
1215                 unlock_page(page);
1216                 return 0;
1217         }
1218
1219         if (wbc->for_reclaim)
1220                 goto redirty_out;
1221
1222         mutex_lock(&sbi->node_write);
1223         set_page_writeback(page);
1224         write_node_page(sbi, page, &fio, nid, ni.blk_addr, &new_addr);
1225         set_node_addr(sbi, &ni, new_addr);
1226         dec_page_count(sbi, F2FS_DIRTY_NODES);
1227         mutex_unlock(&sbi->node_write);
1228         unlock_page(page);
1229         return 0;
1230
1231 redirty_out:
1232         dec_page_count(sbi, F2FS_DIRTY_NODES);
1233         wbc->pages_skipped++;
1234         set_page_dirty(page);
1235         return AOP_WRITEPAGE_ACTIVATE;
1236 }
1237
1238 /*
1239  * It is very important to gather dirty pages and write at once, so that we can
1240  * submit a big bio without interfering other data writes.
1241  * Be default, 512 pages (2MB) * 3 node types, is more reasonable.
1242  */
1243 #define COLLECT_DIRTY_NODES     1536
1244 static int f2fs_write_node_pages(struct address_space *mapping,
1245                             struct writeback_control *wbc)
1246 {
1247         struct f2fs_sb_info *sbi = F2FS_SB(mapping->host->i_sb);
1248         long nr_to_write = wbc->nr_to_write;
1249
1250         /* balancing f2fs's metadata in background */
1251         f2fs_balance_fs_bg(sbi);
1252
1253         /* collect a number of dirty node pages and write together */
1254         if (get_pages(sbi, F2FS_DIRTY_NODES) < COLLECT_DIRTY_NODES)
1255                 return 0;
1256
1257         /* if mounting is failed, skip writing node pages */
1258         wbc->nr_to_write = 3 * max_hw_blocks(sbi);
1259         wbc->sync_mode = WB_SYNC_NONE;
1260         sync_node_pages(sbi, 0, wbc);
1261         wbc->nr_to_write = nr_to_write - (3 * max_hw_blocks(sbi) -
1262                                                 wbc->nr_to_write);
1263         return 0;
1264 }
1265
1266 static int f2fs_set_node_page_dirty(struct page *page)
1267 {
1268         struct address_space *mapping = page->mapping;
1269         struct f2fs_sb_info *sbi = F2FS_SB(mapping->host->i_sb);
1270
1271         trace_f2fs_set_page_dirty(page, NODE);
1272
1273         SetPageUptodate(page);
1274         if (!PageDirty(page)) {
1275                 __set_page_dirty_nobuffers(page);
1276                 inc_page_count(sbi, F2FS_DIRTY_NODES);
1277                 SetPagePrivate(page);
1278                 return 1;
1279         }
1280         return 0;
1281 }
1282
1283 static void f2fs_invalidate_node_page(struct page *page, unsigned int offset,
1284                                       unsigned int length)
1285 {
1286         struct inode *inode = page->mapping->host;
1287         struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
1288         if (PageDirty(page))
1289                 dec_page_count(sbi, F2FS_DIRTY_NODES);
1290         ClearPagePrivate(page);
1291 }
1292
1293 static int f2fs_release_node_page(struct page *page, gfp_t wait)
1294 {
1295         ClearPagePrivate(page);
1296         return 1;
1297 }
1298
1299 /*
1300  * Structure of the f2fs node operations
1301  */
1302 const struct address_space_operations f2fs_node_aops = {
1303         .writepage      = f2fs_write_node_page,
1304         .writepages     = f2fs_write_node_pages,
1305         .set_page_dirty = f2fs_set_node_page_dirty,
1306         .invalidatepage = f2fs_invalidate_node_page,
1307         .releasepage    = f2fs_release_node_page,
1308 };
1309
1310 static struct free_nid *__lookup_free_nid_list(nid_t n, struct list_head *head)
1311 {
1312         struct list_head *this;
1313         struct free_nid *i;
1314         list_for_each(this, head) {
1315                 i = list_entry(this, struct free_nid, list);
1316                 if (i->nid == n)
1317                         return i;
1318         }
1319         return NULL;
1320 }
1321
1322 static void __del_from_free_nid_list(struct free_nid *i)
1323 {
1324         list_del(&i->list);
1325         kmem_cache_free(free_nid_slab, i);
1326 }
1327
1328 static int add_free_nid(struct f2fs_nm_info *nm_i, nid_t nid, bool build)
1329 {
1330         struct free_nid *i;
1331         struct nat_entry *ne;
1332         bool allocated = false;
1333
1334         if (nm_i->fcnt > 2 * MAX_FREE_NIDS)
1335                 return -1;
1336
1337         /* 0 nid should not be used */
1338         if (unlikely(nid == 0))
1339                 return 0;
1340
1341         if (build) {
1342                 /* do not add allocated nids */
1343                 read_lock(&nm_i->nat_tree_lock);
1344                 ne = __lookup_nat_cache(nm_i, nid);
1345                 if (ne && nat_get_blkaddr(ne) != NULL_ADDR)
1346                         allocated = true;
1347                 read_unlock(&nm_i->nat_tree_lock);
1348                 if (allocated)
1349                         return 0;
1350         }
1351
1352         i = f2fs_kmem_cache_alloc(free_nid_slab, GFP_NOFS);
1353         i->nid = nid;
1354         i->state = NID_NEW;
1355
1356         spin_lock(&nm_i->free_nid_list_lock);
1357         if (__lookup_free_nid_list(nid, &nm_i->free_nid_list)) {
1358                 spin_unlock(&nm_i->free_nid_list_lock);
1359                 kmem_cache_free(free_nid_slab, i);
1360                 return 0;
1361         }
1362         list_add_tail(&i->list, &nm_i->free_nid_list);
1363         nm_i->fcnt++;
1364         spin_unlock(&nm_i->free_nid_list_lock);
1365         return 1;
1366 }
1367
1368 static void remove_free_nid(struct f2fs_nm_info *nm_i, nid_t nid)
1369 {
1370         struct free_nid *i;
1371         spin_lock(&nm_i->free_nid_list_lock);
1372         i = __lookup_free_nid_list(nid, &nm_i->free_nid_list);
1373         if (i && i->state == NID_NEW) {
1374                 __del_from_free_nid_list(i);
1375                 nm_i->fcnt--;
1376         }
1377         spin_unlock(&nm_i->free_nid_list_lock);
1378 }
1379
1380 static void scan_nat_page(struct f2fs_nm_info *nm_i,
1381                         struct page *nat_page, nid_t start_nid)
1382 {
1383         struct f2fs_nat_block *nat_blk = page_address(nat_page);
1384         block_t blk_addr;
1385         int i;
1386
1387         i = start_nid % NAT_ENTRY_PER_BLOCK;
1388
1389         for (; i < NAT_ENTRY_PER_BLOCK; i++, start_nid++) {
1390
1391                 if (unlikely(start_nid >= nm_i->max_nid))
1392                         break;
1393
1394                 blk_addr = le32_to_cpu(nat_blk->entries[i].block_addr);
1395                 f2fs_bug_on(blk_addr == NEW_ADDR);
1396                 if (blk_addr == NULL_ADDR) {
1397                         if (add_free_nid(nm_i, start_nid, true) < 0)
1398                                 break;
1399                 }
1400         }
1401 }
1402
1403 static void build_free_nids(struct f2fs_sb_info *sbi)
1404 {
1405         struct f2fs_nm_info *nm_i = NM_I(sbi);
1406         struct curseg_info *curseg = CURSEG_I(sbi, CURSEG_HOT_DATA);
1407         struct f2fs_summary_block *sum = curseg->sum_blk;
1408         int i = 0;
1409         nid_t nid = nm_i->next_scan_nid;
1410
1411         /* Enough entries */
1412         if (nm_i->fcnt > NAT_ENTRY_PER_BLOCK)
1413                 return;
1414
1415         /* readahead nat pages to be scanned */
1416         ra_nat_pages(sbi, nid);
1417
1418         while (1) {
1419                 struct page *page = get_current_nat_page(sbi, nid);
1420
1421                 scan_nat_page(nm_i, page, nid);
1422                 f2fs_put_page(page, 1);
1423
1424                 nid += (NAT_ENTRY_PER_BLOCK - (nid % NAT_ENTRY_PER_BLOCK));
1425                 if (unlikely(nid >= nm_i->max_nid))
1426                         nid = 0;
1427
1428                 if (i++ == FREE_NID_PAGES)
1429                         break;
1430         }
1431
1432         /* go to the next free nat pages to find free nids abundantly */
1433         nm_i->next_scan_nid = nid;
1434
1435         /* find free nids from current sum_pages */
1436         mutex_lock(&curseg->curseg_mutex);
1437         for (i = 0; i < nats_in_cursum(sum); i++) {
1438                 block_t addr = le32_to_cpu(nat_in_journal(sum, i).block_addr);
1439                 nid = le32_to_cpu(nid_in_journal(sum, i));
1440                 if (addr == NULL_ADDR)
1441                         add_free_nid(nm_i, nid, true);
1442                 else
1443                         remove_free_nid(nm_i, nid);
1444         }
1445         mutex_unlock(&curseg->curseg_mutex);
1446 }
1447
1448 /*
1449  * If this function returns success, caller can obtain a new nid
1450  * from second parameter of this function.
1451  * The returned nid could be used ino as well as nid when inode is created.
1452  */
1453 bool alloc_nid(struct f2fs_sb_info *sbi, nid_t *nid)
1454 {
1455         struct f2fs_nm_info *nm_i = NM_I(sbi);
1456         struct free_nid *i = NULL;
1457         struct list_head *this;
1458 retry:
1459         if (unlikely(sbi->total_valid_node_count + 1 >= nm_i->max_nid))
1460                 return false;
1461
1462         spin_lock(&nm_i->free_nid_list_lock);
1463
1464         /* We should not use stale free nids created by build_free_nids */
1465         if (nm_i->fcnt && !sbi->on_build_free_nids) {
1466                 f2fs_bug_on(list_empty(&nm_i->free_nid_list));
1467                 list_for_each(this, &nm_i->free_nid_list) {
1468                         i = list_entry(this, struct free_nid, list);
1469                         if (i->state == NID_NEW)
1470                                 break;
1471                 }
1472
1473                 f2fs_bug_on(i->state != NID_NEW);
1474                 *nid = i->nid;
1475                 i->state = NID_ALLOC;
1476                 nm_i->fcnt--;
1477                 spin_unlock(&nm_i->free_nid_list_lock);
1478                 return true;
1479         }
1480         spin_unlock(&nm_i->free_nid_list_lock);
1481
1482         /* Let's scan nat pages and its caches to get free nids */
1483         mutex_lock(&nm_i->build_lock);
1484         sbi->on_build_free_nids = true;
1485         build_free_nids(sbi);
1486         sbi->on_build_free_nids = false;
1487         mutex_unlock(&nm_i->build_lock);
1488         goto retry;
1489 }
1490
1491 /*
1492  * alloc_nid() should be called prior to this function.
1493  */
1494 void alloc_nid_done(struct f2fs_sb_info *sbi, nid_t nid)
1495 {
1496         struct f2fs_nm_info *nm_i = NM_I(sbi);
1497         struct free_nid *i;
1498
1499         spin_lock(&nm_i->free_nid_list_lock);
1500         i = __lookup_free_nid_list(nid, &nm_i->free_nid_list);
1501         f2fs_bug_on(!i || i->state != NID_ALLOC);
1502         __del_from_free_nid_list(i);
1503         spin_unlock(&nm_i->free_nid_list_lock);
1504 }
1505
1506 /*
1507  * alloc_nid() should be called prior to this function.
1508  */
1509 void alloc_nid_failed(struct f2fs_sb_info *sbi, nid_t nid)
1510 {
1511         struct f2fs_nm_info *nm_i = NM_I(sbi);
1512         struct free_nid *i;
1513
1514         if (!nid)
1515                 return;
1516
1517         spin_lock(&nm_i->free_nid_list_lock);
1518         i = __lookup_free_nid_list(nid, &nm_i->free_nid_list);
1519         f2fs_bug_on(!i || i->state != NID_ALLOC);
1520         if (nm_i->fcnt > 2 * MAX_FREE_NIDS) {
1521                 __del_from_free_nid_list(i);
1522         } else {
1523                 i->state = NID_NEW;
1524                 nm_i->fcnt++;
1525         }
1526         spin_unlock(&nm_i->free_nid_list_lock);
1527 }
1528
1529 void recover_node_page(struct f2fs_sb_info *sbi, struct page *page,
1530                 struct f2fs_summary *sum, struct node_info *ni,
1531                 block_t new_blkaddr)
1532 {
1533         rewrite_node_page(sbi, page, sum, ni->blk_addr, new_blkaddr);
1534         set_node_addr(sbi, ni, new_blkaddr);
1535         clear_node_page_dirty(page);
1536 }
1537
1538 int recover_inode_page(struct f2fs_sb_info *sbi, struct page *page)
1539 {
1540         struct f2fs_inode *src, *dst;
1541         nid_t ino = ino_of_node(page);
1542         struct node_info old_ni, new_ni;
1543         struct page *ipage;
1544
1545         ipage = grab_cache_page(NODE_MAPPING(sbi), ino);
1546         if (!ipage)
1547                 return -ENOMEM;
1548
1549         /* Should not use this inode  from free nid list */
1550         remove_free_nid(NM_I(sbi), ino);
1551
1552         get_node_info(sbi, ino, &old_ni);
1553         SetPageUptodate(ipage);
1554         fill_node_footer(ipage, ino, ino, 0, true);
1555
1556         src = F2FS_INODE(page);
1557         dst = F2FS_INODE(ipage);
1558
1559         memcpy(dst, src, (unsigned long)&src->i_ext - (unsigned long)src);
1560         dst->i_size = 0;
1561         dst->i_blocks = cpu_to_le64(1);
1562         dst->i_links = cpu_to_le32(1);
1563         dst->i_xattr_nid = 0;
1564
1565         new_ni = old_ni;
1566         new_ni.ino = ino;
1567
1568         if (unlikely(!inc_valid_node_count(sbi, NULL)))
1569                 WARN_ON(1);
1570         set_node_addr(sbi, &new_ni, NEW_ADDR);
1571         inc_valid_inode_count(sbi);
1572         f2fs_put_page(ipage, 1);
1573         return 0;
1574 }
1575
1576 /*
1577  * ra_sum_pages() merge contiguous pages into one bio and submit.
1578  * these pre-readed pages are linked in pages list.
1579  */
1580 static int ra_sum_pages(struct f2fs_sb_info *sbi, struct list_head *pages,
1581                                 int start, int nrpages)
1582 {
1583         struct page *page;
1584         int page_idx = start;
1585         struct f2fs_io_info fio = {
1586                 .type = META,
1587                 .rw = READ_SYNC | REQ_META | REQ_PRIO
1588         };
1589
1590         for (; page_idx < start + nrpages; page_idx++) {
1591                 /* alloc temporal page for read node summary info*/
1592                 page = alloc_page(GFP_F2FS_ZERO);
1593                 if (!page) {
1594                         struct page *tmp;
1595                         list_for_each_entry_safe(page, tmp, pages, lru) {
1596                                 list_del(&page->lru);
1597                                 unlock_page(page);
1598                                 __free_pages(page, 0);
1599                         }
1600                         return -ENOMEM;
1601                 }
1602
1603                 lock_page(page);
1604                 page->index = page_idx;
1605                 list_add_tail(&page->lru, pages);
1606         }
1607
1608         list_for_each_entry(page, pages, lru)
1609                 f2fs_submit_page_mbio(sbi, page, page->index, &fio);
1610
1611         f2fs_submit_merged_bio(sbi, META, READ);
1612         return 0;
1613 }
1614
1615 int restore_node_summary(struct f2fs_sb_info *sbi,
1616                         unsigned int segno, struct f2fs_summary_block *sum)
1617 {
1618         struct f2fs_node *rn;
1619         struct f2fs_summary *sum_entry;
1620         struct page *page, *tmp;
1621         block_t addr;
1622         int bio_blocks = MAX_BIO_BLOCKS(max_hw_blocks(sbi));
1623         int i, last_offset, nrpages, err = 0;
1624         LIST_HEAD(page_list);
1625
1626         /* scan the node segment */
1627         last_offset = sbi->blocks_per_seg;
1628         addr = START_BLOCK(sbi, segno);
1629         sum_entry = &sum->entries[0];
1630
1631         for (i = 0; i < last_offset; i += nrpages, addr += nrpages) {
1632                 nrpages = min(last_offset - i, bio_blocks);
1633
1634                 /* read ahead node pages */
1635                 err = ra_sum_pages(sbi, &page_list, addr, nrpages);
1636                 if (err)
1637                         return err;
1638
1639                 list_for_each_entry_safe(page, tmp, &page_list, lru) {
1640
1641                         lock_page(page);
1642                         if (unlikely(!PageUptodate(page))) {
1643                                 err = -EIO;
1644                         } else {
1645                                 rn = F2FS_NODE(page);
1646                                 sum_entry->nid = rn->footer.nid;
1647                                 sum_entry->version = 0;
1648                                 sum_entry->ofs_in_node = 0;
1649                                 sum_entry++;
1650                         }
1651
1652                         list_del(&page->lru);
1653                         unlock_page(page);
1654                         __free_pages(page, 0);
1655                 }
1656         }
1657         return err;
1658 }
1659
1660 static bool flush_nats_in_journal(struct f2fs_sb_info *sbi)
1661 {
1662         struct f2fs_nm_info *nm_i = NM_I(sbi);
1663         struct curseg_info *curseg = CURSEG_I(sbi, CURSEG_HOT_DATA);
1664         struct f2fs_summary_block *sum = curseg->sum_blk;
1665         int i;
1666
1667         mutex_lock(&curseg->curseg_mutex);
1668
1669         if (nats_in_cursum(sum) < NAT_JOURNAL_ENTRIES) {
1670                 mutex_unlock(&curseg->curseg_mutex);
1671                 return false;
1672         }
1673
1674         for (i = 0; i < nats_in_cursum(sum); i++) {
1675                 struct nat_entry *ne;
1676                 struct f2fs_nat_entry raw_ne;
1677                 nid_t nid = le32_to_cpu(nid_in_journal(sum, i));
1678
1679                 raw_ne = nat_in_journal(sum, i);
1680 retry:
1681                 write_lock(&nm_i->nat_tree_lock);
1682                 ne = __lookup_nat_cache(nm_i, nid);
1683                 if (ne) {
1684                         __set_nat_cache_dirty(nm_i, ne);
1685                         write_unlock(&nm_i->nat_tree_lock);
1686                         continue;
1687                 }
1688                 ne = grab_nat_entry(nm_i, nid);
1689                 if (!ne) {
1690                         write_unlock(&nm_i->nat_tree_lock);
1691                         goto retry;
1692                 }
1693                 nat_set_blkaddr(ne, le32_to_cpu(raw_ne.block_addr));
1694                 nat_set_ino(ne, le32_to_cpu(raw_ne.ino));
1695                 nat_set_version(ne, raw_ne.version);
1696                 __set_nat_cache_dirty(nm_i, ne);
1697                 write_unlock(&nm_i->nat_tree_lock);
1698         }
1699         update_nats_in_cursum(sum, -i);
1700         mutex_unlock(&curseg->curseg_mutex);
1701         return true;
1702 }
1703
1704 /*
1705  * This function is called during the checkpointing process.
1706  */
1707 void flush_nat_entries(struct f2fs_sb_info *sbi)
1708 {
1709         struct f2fs_nm_info *nm_i = NM_I(sbi);
1710         struct curseg_info *curseg = CURSEG_I(sbi, CURSEG_HOT_DATA);
1711         struct f2fs_summary_block *sum = curseg->sum_blk;
1712         struct list_head *cur, *n;
1713         struct page *page = NULL;
1714         struct f2fs_nat_block *nat_blk = NULL;
1715         nid_t start_nid = 0, end_nid = 0;
1716         bool flushed;
1717
1718         flushed = flush_nats_in_journal(sbi);
1719
1720         if (!flushed)
1721                 mutex_lock(&curseg->curseg_mutex);
1722
1723         /* 1) flush dirty nat caches */
1724         list_for_each_safe(cur, n, &nm_i->dirty_nat_entries) {
1725                 struct nat_entry *ne;
1726                 nid_t nid;
1727                 struct f2fs_nat_entry raw_ne;
1728                 int offset = -1;
1729                 block_t new_blkaddr;
1730
1731                 ne = list_entry(cur, struct nat_entry, list);
1732                 nid = nat_get_nid(ne);
1733
1734                 if (nat_get_blkaddr(ne) == NEW_ADDR)
1735                         continue;
1736                 if (flushed)
1737                         goto to_nat_page;
1738
1739                 /* if there is room for nat enries in curseg->sumpage */
1740                 offset = lookup_journal_in_cursum(sum, NAT_JOURNAL, nid, 1);
1741                 if (offset >= 0) {
1742                         raw_ne = nat_in_journal(sum, offset);
1743                         goto flush_now;
1744                 }
1745 to_nat_page:
1746                 if (!page || (start_nid > nid || nid > end_nid)) {
1747                         if (page) {
1748                                 f2fs_put_page(page, 1);
1749                                 page = NULL;
1750                         }
1751                         start_nid = START_NID(nid);
1752                         end_nid = start_nid + NAT_ENTRY_PER_BLOCK - 1;
1753
1754                         /*
1755                          * get nat block with dirty flag, increased reference
1756                          * count, mapped and lock
1757                          */
1758                         page = get_next_nat_page(sbi, start_nid);
1759                         nat_blk = page_address(page);
1760                 }
1761
1762                 f2fs_bug_on(!nat_blk);
1763                 raw_ne = nat_blk->entries[nid - start_nid];
1764 flush_now:
1765                 new_blkaddr = nat_get_blkaddr(ne);
1766
1767                 raw_ne.ino = cpu_to_le32(nat_get_ino(ne));
1768                 raw_ne.block_addr = cpu_to_le32(new_blkaddr);
1769                 raw_ne.version = nat_get_version(ne);
1770
1771                 if (offset < 0) {
1772                         nat_blk->entries[nid - start_nid] = raw_ne;
1773                 } else {
1774                         nat_in_journal(sum, offset) = raw_ne;
1775                         nid_in_journal(sum, offset) = cpu_to_le32(nid);
1776                 }
1777
1778                 if (nat_get_blkaddr(ne) == NULL_ADDR &&
1779                                 add_free_nid(NM_I(sbi), nid, false) <= 0) {
1780                         write_lock(&nm_i->nat_tree_lock);
1781                         __del_from_nat_cache(nm_i, ne);
1782                         write_unlock(&nm_i->nat_tree_lock);
1783                 } else {
1784                         write_lock(&nm_i->nat_tree_lock);
1785                         __clear_nat_cache_dirty(nm_i, ne);
1786                         ne->checkpointed = true;
1787                         write_unlock(&nm_i->nat_tree_lock);
1788                 }
1789         }
1790         if (!flushed)
1791                 mutex_unlock(&curseg->curseg_mutex);
1792         f2fs_put_page(page, 1);
1793
1794         /* 2) shrink nat caches if necessary */
1795         try_to_free_nats(sbi, nm_i->nat_cnt - NM_WOUT_THRESHOLD);
1796 }
1797
1798 static int init_node_manager(struct f2fs_sb_info *sbi)
1799 {
1800         struct f2fs_super_block *sb_raw = F2FS_RAW_SUPER(sbi);
1801         struct f2fs_nm_info *nm_i = NM_I(sbi);
1802         unsigned char *version_bitmap;
1803         unsigned int nat_segs, nat_blocks;
1804
1805         nm_i->nat_blkaddr = le32_to_cpu(sb_raw->nat_blkaddr);
1806
1807         /* segment_count_nat includes pair segment so divide to 2. */
1808         nat_segs = le32_to_cpu(sb_raw->segment_count_nat) >> 1;
1809         nat_blocks = nat_segs << le32_to_cpu(sb_raw->log_blocks_per_seg);
1810         nm_i->max_nid = NAT_ENTRY_PER_BLOCK * nat_blocks;
1811         nm_i->fcnt = 0;
1812         nm_i->nat_cnt = 0;
1813
1814         INIT_LIST_HEAD(&nm_i->free_nid_list);
1815         INIT_RADIX_TREE(&nm_i->nat_root, GFP_ATOMIC);
1816         INIT_LIST_HEAD(&nm_i->nat_entries);
1817         INIT_LIST_HEAD(&nm_i->dirty_nat_entries);
1818
1819         mutex_init(&nm_i->build_lock);
1820         spin_lock_init(&nm_i->free_nid_list_lock);
1821         rwlock_init(&nm_i->nat_tree_lock);
1822
1823         nm_i->next_scan_nid = le32_to_cpu(sbi->ckpt->next_free_nid);
1824         nm_i->bitmap_size = __bitmap_size(sbi, NAT_BITMAP);
1825         version_bitmap = __bitmap_ptr(sbi, NAT_BITMAP);
1826         if (!version_bitmap)
1827                 return -EFAULT;
1828
1829         nm_i->nat_bitmap = kmemdup(version_bitmap, nm_i->bitmap_size,
1830                                         GFP_KERNEL);
1831         if (!nm_i->nat_bitmap)
1832                 return -ENOMEM;
1833         return 0;
1834 }
1835
1836 int build_node_manager(struct f2fs_sb_info *sbi)
1837 {
1838         int err;
1839
1840         sbi->nm_info = kzalloc(sizeof(struct f2fs_nm_info), GFP_KERNEL);
1841         if (!sbi->nm_info)
1842                 return -ENOMEM;
1843
1844         err = init_node_manager(sbi);
1845         if (err)
1846                 return err;
1847
1848         build_free_nids(sbi);
1849         return 0;
1850 }
1851
1852 void destroy_node_manager(struct f2fs_sb_info *sbi)
1853 {
1854         struct f2fs_nm_info *nm_i = NM_I(sbi);
1855         struct free_nid *i, *next_i;
1856         struct nat_entry *natvec[NATVEC_SIZE];
1857         nid_t nid = 0;
1858         unsigned int found;
1859
1860         if (!nm_i)
1861                 return;
1862
1863         /* destroy free nid list */
1864         spin_lock(&nm_i->free_nid_list_lock);
1865         list_for_each_entry_safe(i, next_i, &nm_i->free_nid_list, list) {
1866                 f2fs_bug_on(i->state == NID_ALLOC);
1867                 __del_from_free_nid_list(i);
1868                 nm_i->fcnt--;
1869         }
1870         f2fs_bug_on(nm_i->fcnt);
1871         spin_unlock(&nm_i->free_nid_list_lock);
1872
1873         /* destroy nat cache */
1874         write_lock(&nm_i->nat_tree_lock);
1875         while ((found = __gang_lookup_nat_cache(nm_i,
1876                                         nid, NATVEC_SIZE, natvec))) {
1877                 unsigned idx;
1878                 for (idx = 0; idx < found; idx++) {
1879                         struct nat_entry *e = natvec[idx];
1880                         nid = nat_get_nid(e) + 1;
1881                         __del_from_nat_cache(nm_i, e);
1882                 }
1883         }
1884         f2fs_bug_on(nm_i->nat_cnt);
1885         write_unlock(&nm_i->nat_tree_lock);
1886
1887         kfree(nm_i->nat_bitmap);
1888         sbi->nm_info = NULL;
1889         kfree(nm_i);
1890 }
1891
1892 int __init create_node_manager_caches(void)
1893 {
1894         nat_entry_slab = f2fs_kmem_cache_create("nat_entry",
1895                         sizeof(struct nat_entry), NULL);
1896         if (!nat_entry_slab)
1897                 return -ENOMEM;
1898
1899         free_nid_slab = f2fs_kmem_cache_create("free_nid",
1900                         sizeof(struct free_nid), NULL);
1901         if (!free_nid_slab) {
1902                 kmem_cache_destroy(nat_entry_slab);
1903                 return -ENOMEM;
1904         }
1905         return 0;
1906 }
1907
1908 void destroy_node_manager_caches(void)
1909 {
1910         kmem_cache_destroy(free_nid_slab);
1911         kmem_cache_destroy(nat_entry_slab);
1912 }