SAFE public projects git trees. - safe/jmp/linux-2.6/blob - fs/ext4/balloc.c

   1 /*
   2  *  linux/fs/ext4/balloc.c
   3  *
   4  * Copyright (C) 1992, 1993, 1994, 1995
   5  * Remy Card (card@masi.ibp.fr)
   6  * Laboratoire MASI - Institut Blaise Pascal
   7  * Universite Pierre et Marie Curie (Paris VI)
   8  *
   9  *  Enhanced block allocation by Stephen Tweedie (sct@redhat.com), 1993
  10  *  Big-endian to little-endian byte-swapping/bitmaps by
  11  *        David S. Miller (davem@caip.rutgers.edu), 1995
  12  */
  13
  14 #include <linux/time.h>
  15 #include <linux/capability.h>
  16 #include <linux/fs.h>
  17 #include <linux/jbd2.h>
  18 #include <linux/quotaops.h>
  19 #include <linux/buffer_head.h>
  20 #include "ext4.h"
  21 #include "ext4_jbd2.h"
  22 #include "mballoc.h"
  23
  24 /*
  25  * balloc.c contains the blocks allocation and deallocation routines
  26  */
  27
  28 /*
  29  * Calculate the block group number and offset, given a block number
  30  */
  31 void ext4_get_group_no_and_offset(struct super_block *sb, ext4_fsblk_t blocknr,
  32                 ext4_group_t *blockgrpp, ext4_grpblk_t *offsetp)
  33 {
  34         struct ext4_super_block *es = EXT4_SB(sb)->s_es;
  35         ext4_grpblk_t offset;
  36
  37         blocknr = blocknr - le32_to_cpu(es->s_first_data_block);
  38         offset = do_div(blocknr, EXT4_BLOCKS_PER_GROUP(sb));
  39         if (offsetp)
  40                 *offsetp = offset;
  41         if (blockgrpp)
  42                 *blockgrpp = blocknr;
  43
  44 }
  45
  46 static int ext4_block_in_group(struct super_block *sb, ext4_fsblk_t block,
  47                         ext4_group_t block_group)
  48 {
  49         ext4_group_t actual_group;
  50         ext4_get_group_no_and_offset(sb, block, &actual_group, NULL);
  51         if (actual_group == block_group)
  52                 return 1;
  53         return 0;
  54 }
  55
  56 static int ext4_group_used_meta_blocks(struct super_block *sb,
  57                                        ext4_group_t block_group,
  58                                        struct ext4_group_desc *gdp)
  59 {
  60         ext4_fsblk_t tmp;
  61         struct ext4_sb_info *sbi = EXT4_SB(sb);
  62         /* block bitmap, inode bitmap, and inode table blocks */
  63         int used_blocks = sbi->s_itb_per_group + 2;
  64
  65         if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG)) {
  66                 if (!ext4_block_in_group(sb, ext4_block_bitmap(sb, gdp),
  67                                         block_group))
  68                         used_blocks--;
  69
  70                 if (!ext4_block_in_group(sb, ext4_inode_bitmap(sb, gdp),
  71                                         block_group))
  72                         used_blocks--;
  73
  74                 tmp = ext4_inode_table(sb, gdp);
  75                 for (; tmp < ext4_inode_table(sb, gdp) +
  76                                 sbi->s_itb_per_group; tmp++) {
  77                         if (!ext4_block_in_group(sb, tmp, block_group))
  78                                 used_blocks -= 1;
  79                 }
  80         }
  81         return used_blocks;
  82 }
  83
  84 /* Initializes an uninitialized block bitmap if given, and returns the
  85  * number of blocks free in the group. */
  86 unsigned ext4_init_block_bitmap(struct super_block *sb, struct buffer_head *bh,
  87                  ext4_group_t block_group, struct ext4_group_desc *gdp)
  88 {
  89         int bit, bit_max;
  90         ext4_group_t ngroups = ext4_get_groups_count(sb);
  91         unsigned free_blocks, group_blocks;
  92         struct ext4_sb_info *sbi = EXT4_SB(sb);
  93
  94         if (bh) {
  95                 J_ASSERT_BH(bh, buffer_locked(bh));
  96
  97                 /* If checksum is bad mark all blocks used to prevent allocation
  98                  * essentially implementing a per-group read-only flag. */
  99                 if (!ext4_group_desc_csum_verify(sbi, block_group, gdp)) {
 100                         ext4_error(sb, __func__,
 101                                   "Checksum bad for group %u", block_group);
 102                         ext4_free_blks_set(sb, gdp, 0);
 103                         ext4_free_inodes_set(sb, gdp, 0);
 104                         ext4_itable_unused_set(sb, gdp, 0);
 105                         memset(bh->b_data, 0xff, sb->s_blocksize);
 106                         return 0;
 107                 }
 108                 memset(bh->b_data, 0, sb->s_blocksize);
 109         }
 110
 111         /* Check for superblock and gdt backups in this group */
 112         bit_max = ext4_bg_has_super(sb, block_group);
 113
 114         if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_META_BG) ||
 115             block_group < le32_to_cpu(sbi->s_es->s_first_meta_bg) *
 116                           sbi->s_desc_per_block) {
 117                 if (bit_max) {
 118                         bit_max += ext4_bg_num_gdb(sb, block_group);
 119                         bit_max +=
 120                                 le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks);
 121                 }
 122         } else { /* For META_BG_BLOCK_GROUPS */
 123                 bit_max += ext4_bg_num_gdb(sb, block_group);
 124         }
 125
 126         if (block_group == ngroups - 1) {
 127                 /*
 128                  * Even though mke2fs always initialize first and last group
 129                  * if some other tool enabled the EXT4_BG_BLOCK_UNINIT we need
 130                  * to make sure we calculate the right free blocks
 131                  */
 132                 group_blocks = ext4_blocks_count(sbi->s_es) -
 133                         le32_to_cpu(sbi->s_es->s_first_data_block) -
 134                         (EXT4_BLOCKS_PER_GROUP(sb) * (ngroups - 1));
 135         } else {
 136                 group_blocks = EXT4_BLOCKS_PER_GROUP(sb);
 137         }
 138
 139         free_blocks = group_blocks - bit_max;
 140
 141         if (bh) {
 142                 ext4_fsblk_t start, tmp;
 143                 int flex_bg = 0;
 144
 145                 for (bit = 0; bit < bit_max; bit++)
 146                         ext4_set_bit(bit, bh->b_data);
 147
 148                 start = ext4_group_first_block_no(sb, block_group);
 149
 150                 if (EXT4_HAS_INCOMPAT_FEATURE(sb,
 151                                               EXT4_FEATURE_INCOMPAT_FLEX_BG))
 152                         flex_bg = 1;
 153
 154                 /* Set bits for block and inode bitmaps, and inode table */
 155                 tmp = ext4_block_bitmap(sb, gdp);
 156                 if (!flex_bg || ext4_block_in_group(sb, tmp, block_group))
 157                         ext4_set_bit(tmp - start, bh->b_data);
 158
 159                 tmp = ext4_inode_bitmap(sb, gdp);
 160                 if (!flex_bg || ext4_block_in_group(sb, tmp, block_group))
 161                         ext4_set_bit(tmp - start, bh->b_data);
 162
 163                 tmp = ext4_inode_table(sb, gdp);
 164                 for (; tmp < ext4_inode_table(sb, gdp) +
 165                                 sbi->s_itb_per_group; tmp++) {
 166                         if (!flex_bg ||
 167                                 ext4_block_in_group(sb, tmp, block_group))
 168                                 ext4_set_bit(tmp - start, bh->b_data);
 169                 }
 170                 /*
 171                  * Also if the number of blocks within the group is
 172                  * less than the blocksize * 8 ( which is the size
 173                  * of bitmap ), set rest of the block bitmap to 1
 174                  */
 175                 mark_bitmap_end(group_blocks, sb->s_blocksize * 8, bh->b_data);
 176         }
 177         return free_blocks - ext4_group_used_meta_blocks(sb, block_group, gdp);
 178 }
 179
 180
 181 /*
 182  * The free blocks are managed by bitmaps.  A file system contains several
 183  * blocks groups.  Each group contains 1 bitmap block for blocks, 1 bitmap
 184  * block for inodes, N blocks for the inode table and data blocks.
 185  *
 186  * The file system contains group descriptors which are located after the
 187  * super block.  Each descriptor contains the number of the bitmap block and
 188  * the free blocks count in the block.  The descriptors are loaded in memory
 189  * when a file system is mounted (see ext4_fill_super).
 190  */
 191
 192
 193 #define in_range(b, first, len) ((b) >= (first) && (b) <= (first) + (len) - 1)
 194
 195 /**
 196  * ext4_get_group_desc() -- load group descriptor from disk
 197  * @sb:                 super block
 198  * @block_group:        given block group
 199  * @bh:                 pointer to the buffer head to store the block
 200  *                      group descriptor
 201  */
 202 struct ext4_group_desc * ext4_get_group_desc(struct super_block *sb,
 203                                              ext4_group_t block_group,
 204                                              struct buffer_head **bh)
 205 {
 206         unsigned int group_desc;
 207         unsigned int offset;
 208         ext4_group_t ngroups = ext4_get_groups_count(sb);
 209         struct ext4_group_desc *desc;
 210         struct ext4_sb_info *sbi = EXT4_SB(sb);
 211
 212         if (block_group >= ngroups) {
 213                 ext4_error(sb, "ext4_get_group_desc",
 214                            "block_group >= groups_count - "
 215                            "block_group = %u, groups_count = %u",
 216                            block_group, ngroups);
 217
 218                 return NULL;
 219         }
 220
 221         group_desc = block_group >> EXT4_DESC_PER_BLOCK_BITS(sb);
 222         offset = block_group & (EXT4_DESC_PER_BLOCK(sb) - 1);
 223         if (!sbi->s_group_desc[group_desc]) {
 224                 ext4_error(sb, "ext4_get_group_desc",
 225                            "Group descriptor not loaded - "
 226                            "block_group = %u, group_desc = %u, desc = %u",
 227                            block_group, group_desc, offset);
 228                 return NULL;
 229         }
 230
 231         desc = (struct ext4_group_desc *)(
 232                 (__u8 *)sbi->s_group_desc[group_desc]->b_data +
 233                 offset * EXT4_DESC_SIZE(sb));
 234         if (bh)
 235                 *bh = sbi->s_group_desc[group_desc];
 236         return desc;
 237 }
 238
 239 static int ext4_valid_block_bitmap(struct super_block *sb,
 240                                         struct ext4_group_desc *desc,
 241                                         unsigned int block_group,
 242                                         struct buffer_head *bh)
 243 {
 244         ext4_grpblk_t offset;
 245         ext4_grpblk_t next_zero_bit;
 246         ext4_fsblk_t bitmap_blk;
 247         ext4_fsblk_t group_first_block;
 248
 249         if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG)) {
 250                 /* with FLEX_BG, the inode/block bitmaps and itable
 251                  * blocks may not be in the group at all
 252                  * so the bitmap validation will be skipped for those groups
 253                  * or it has to also read the block group where the bitmaps
 254                  * are located to verify they are set.
 255                  */
 256                 return 1;
 257         }
 258         group_first_block = ext4_group_first_block_no(sb, block_group);
 259
 260         /* check whether block bitmap block number is set */
 261         bitmap_blk = ext4_block_bitmap(sb, desc);
 262         offset = bitmap_blk - group_first_block;
 263         if (!ext4_test_bit(offset, bh->b_data))
 264                 /* bad block bitmap */
 265                 goto err_out;
 266
 267         /* check whether the inode bitmap block number is set */
 268         bitmap_blk = ext4_inode_bitmap(sb, desc);
 269         offset = bitmap_blk - group_first_block;
 270         if (!ext4_test_bit(offset, bh->b_data))
 271                 /* bad block bitmap */
 272                 goto err_out;
 273
 274         /* check whether the inode table block number is set */
 275         bitmap_blk = ext4_inode_table(sb, desc);
 276         offset = bitmap_blk - group_first_block;
 277         next_zero_bit = ext4_find_next_zero_bit(bh->b_data,
 278                                 offset + EXT4_SB(sb)->s_itb_per_group,
 279                                 offset);
 280         if (next_zero_bit >= offset + EXT4_SB(sb)->s_itb_per_group)
 281                 /* good bitmap for inode tables */
 282                 return 1;
 283
 284 err_out:
 285         ext4_error(sb, __func__,
 286                         "Invalid block bitmap - "
 287                         "block_group = %d, block = %llu",
 288                         block_group, bitmap_blk);
 289         return 0;
 290 }
 291 /**
 292  * ext4_read_block_bitmap()
 293  * @sb:                 super block
 294  * @block_group:        given block group
 295  *
 296  * Read the bitmap for a given block_group,and validate the
 297  * bits for block/inode/inode tables are set in the bitmaps
 298  *
 299  * Return buffer_head on success or NULL in case of failure.
 300  */
 301 struct buffer_head *
 302 ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group)
 303 {
 304         struct ext4_group_desc *desc;
 305         struct buffer_head *bh = NULL;
 306         ext4_fsblk_t bitmap_blk;
 307
 308         desc = ext4_get_group_desc(sb, block_group, NULL);
 309         if (!desc)
 310                 return NULL;
 311         bitmap_blk = ext4_block_bitmap(sb, desc);
 312         bh = sb_getblk(sb, bitmap_blk);
 313         if (unlikely(!bh)) {
 314                 ext4_error(sb, __func__,
 315                             "Cannot read block bitmap - "
 316                             "block_group = %u, block_bitmap = %llu",
 317                             block_group, bitmap_blk);
 318                 return NULL;
 319         }
 320
 321         if (bitmap_uptodate(bh))
 322                 return bh;
 323
 324         lock_buffer(bh);
 325         if (bitmap_uptodate(bh)) {
 326                 unlock_buffer(bh);
 327                 return bh;
 328         }
 329         ext4_lock_group(sb, block_group);
 330         if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
 331                 ext4_init_block_bitmap(sb, bh, block_group, desc);
 332                 set_bitmap_uptodate(bh);
 333                 set_buffer_uptodate(bh);
 334                 ext4_unlock_group(sb, block_group);
 335                 unlock_buffer(bh);
 336                 return bh;
 337         }
 338         ext4_unlock_group(sb, block_group);
 339         if (buffer_uptodate(bh)) {
 340                 /*
 341                  * if not uninit if bh is uptodate,
 342                  * bitmap is also uptodate
 343                  */
 344                 set_bitmap_uptodate(bh);
 345                 unlock_buffer(bh);
 346                 return bh;
 347         }
 348         /*
 349          * submit the buffer_head for read. We can
 350          * safely mark the bitmap as uptodate now.
 351          * We do it here so the bitmap uptodate bit
 352          * get set with buffer lock held.
 353          */
 354         set_bitmap_uptodate(bh);
 355         if (bh_submit_read(bh) < 0) {
 356                 put_bh(bh);
 357                 ext4_error(sb, __func__,
 358                             "Cannot read block bitmap - "
 359                             "block_group = %u, block_bitmap = %llu",
 360                             block_group, bitmap_blk);
 361                 return NULL;
 362         }
 363         ext4_valid_block_bitmap(sb, desc, block_group, bh);
 364         /*
 365          * file system mounted not to panic on error,
 366          * continue with corrupt bitmap
 367          */
 368         return bh;
 369 }
 370
 371 /**
 372  * ext4_add_groupblocks() -- Add given blocks to an existing group
 373  * @handle:                     handle to this transaction
 374  * @sb:                         super block
 375  * @block:                      start physcial block to add to the block group
 376  * @count:                      number of blocks to free
 377  *
 378  * This marks the blocks as free in the bitmap. We ask the
 379  * mballoc to reload the buddy after this by setting group
 380  * EXT4_GROUP_INFO_NEED_INIT_BIT flag
 381  */
 382 void ext4_add_groupblocks(handle_t *handle, struct super_block *sb,
 383                          ext4_fsblk_t block, unsigned long count)
 384 {
 385         struct buffer_head *bitmap_bh = NULL;
 386         struct buffer_head *gd_bh;
 387         ext4_group_t block_group;
 388         ext4_grpblk_t bit;
 389         unsigned int i;
 390         struct ext4_group_desc *desc;
 391         struct ext4_super_block *es;
 392         struct ext4_sb_info *sbi;
 393         int err = 0, ret, blk_free_count;
 394         ext4_grpblk_t blocks_freed;
 395         struct ext4_group_info *grp;
 396
 397         sbi = EXT4_SB(sb);
 398         es = sbi->s_es;
 399         ext4_debug("Adding block(s) %llu-%llu\n", block, block + count - 1);
 400
 401         ext4_get_group_no_and_offset(sb, block, &block_group, &bit);
 402         grp = ext4_get_group_info(sb, block_group);
 403         /*
 404          * Check to see if we are freeing blocks across a group
 405          * boundary.
 406          */
 407         if (bit + count > EXT4_BLOCKS_PER_GROUP(sb)) {
 408                 goto error_return;
 409         }
 410         bitmap_bh = ext4_read_block_bitmap(sb, block_group);
 411         if (!bitmap_bh)
 412                 goto error_return;
 413         desc = ext4_get_group_desc(sb, block_group, &gd_bh);
 414         if (!desc)
 415                 goto error_return;
 416
 417         if (in_range(ext4_block_bitmap(sb, desc), block, count) ||
 418             in_range(ext4_inode_bitmap(sb, desc), block, count) ||
 419             in_range(block, ext4_inode_table(sb, desc), sbi->s_itb_per_group) ||
 420             in_range(block + count - 1, ext4_inode_table(sb, desc),
 421                      sbi->s_itb_per_group)) {
 422                 ext4_error(sb, __func__,
 423                            "Adding blocks in system zones - "
 424                            "Block = %llu, count = %lu",
 425                            block, count);
 426                 goto error_return;
 427         }
 428
 429         /*
 430          * We are about to add blocks to the bitmap,
 431          * so we need undo access.
 432          */
 433         BUFFER_TRACE(bitmap_bh, "getting undo access");
 434         err = ext4_journal_get_undo_access(handle, bitmap_bh);
 435         if (err)
 436                 goto error_return;
 437
 438         /*
 439          * We are about to modify some metadata.  Call the journal APIs
 440          * to unshare ->b_data if a currently-committing transaction is
 441          * using it
 442          */
 443         BUFFER_TRACE(gd_bh, "get_write_access");
 444         err = ext4_journal_get_write_access(handle, gd_bh);
 445         if (err)
 446                 goto error_return;
 447         /*
 448          * make sure we don't allow a parallel init on other groups in the
 449          * same buddy cache
 450          */
 451         down_write(&grp->alloc_sem);
 452         for (i = 0, blocks_freed = 0; i < count; i++) {
 453                 BUFFER_TRACE(bitmap_bh, "clear bit");
 454                 if (!ext4_clear_bit_atomic(ext4_group_lock_ptr(sb, block_group),
 455                                                 bit + i, bitmap_bh->b_data)) {
 456                         ext4_error(sb, __func__,
 457                                    "bit already cleared for block %llu",
 458                                    (ext4_fsblk_t)(block + i));
 459                         BUFFER_TRACE(bitmap_bh, "bit already cleared");
 460                 } else {
 461                         blocks_freed++;
 462                 }
 463         }
 464         ext4_lock_group(sb, block_group);
 465         blk_free_count = blocks_freed + ext4_free_blks_count(sb, desc);
 466         ext4_free_blks_set(sb, desc, blk_free_count);
 467         desc->bg_checksum = ext4_group_desc_csum(sbi, block_group, desc);
 468         ext4_unlock_group(sb, block_group);
 469         percpu_counter_add(&sbi->s_freeblocks_counter, blocks_freed);
 470
 471         if (sbi->s_log_groups_per_flex) {
 472                 ext4_group_t flex_group = ext4_flex_group(sbi, block_group);
 473                 atomic_add(blocks_freed,
 474                            &sbi->s_flex_groups[flex_group].free_blocks);
 475         }
 476         /*
 477          * request to reload the buddy with the
 478          * new bitmap information
 479          */
 480         set_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &(grp->bb_state));
 481         grp->bb_free += blocks_freed;
 482         up_write(&grp->alloc_sem);
 483
 484         /* We dirtied the bitmap block */
 485         BUFFER_TRACE(bitmap_bh, "dirtied bitmap block");
 486         err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh);
 487
 488         /* And the group descriptor block */
 489         BUFFER_TRACE(gd_bh, "dirtied group descriptor block");
 490         ret = ext4_handle_dirty_metadata(handle, NULL, gd_bh);
 491         if (!err)
 492                 err = ret;
 493         sb->s_dirt = 1;
 494
 495 error_return:
 496         brelse(bitmap_bh);
 497         ext4_std_error(sb, err);
 498         return;
 499 }
 500
 501 /**
 502  * ext4_free_blocks() -- Free given blocks and update quota
 503  * @handle:             handle for this transaction
 504  * @inode:              inode
 505  * @block:              start physical block to free
 506  * @count:              number of blocks to count
 507  * @metadata:           Are these metadata blocks
 508  */
 509 void ext4_free_blocks(handle_t *handle, struct inode *inode,
 510                         ext4_fsblk_t block, unsigned long count,
 511                         int metadata)
 512 {
 513         struct super_block *sb;
 514         unsigned long dquot_freed_blocks;
 515
 516         /* this isn't the right place to decide whether block is metadata
 517          * inode.c/extents.c knows better, but for safety ... */
 518         if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
 519                 metadata = 1;
 520
 521         /* We need to make sure we don't reuse
 522          * block released untill the transaction commit.
 523          * writeback mode have weak data consistency so
 524          * don't force data as metadata when freeing block
 525          * for writeback mode.
 526          */
 527         if (metadata == 0 && !ext4_should_writeback_data(inode))
 528                 metadata = 1;
 529
 530         sb = inode->i_sb;
 531
 532         ext4_mb_free_blocks(handle, inode, block, count,
 533                             metadata, &dquot_freed_blocks);
 534         if (dquot_freed_blocks)
 535                 vfs_dq_free_block(inode, dquot_freed_blocks);
 536         return;
 537 }
 538
 539 /**
 540  * ext4_has_free_blocks()
 541  * @sbi:        in-core super block structure.
 542  * @nblocks:    number of needed blocks
 543  *
 544  * Check if filesystem has nblocks free & available for allocation.
 545  * On success return 1, return 0 on failure.
 546  */
 547 int ext4_has_free_blocks(struct ext4_sb_info *sbi, s64 nblocks)
 548 {
 549         s64 free_blocks, dirty_blocks, root_blocks;
 550         struct percpu_counter *fbc = &sbi->s_freeblocks_counter;
 551         struct percpu_counter *dbc = &sbi->s_dirtyblocks_counter;
 552
 553         free_blocks  = percpu_counter_read_positive(fbc);
 554         dirty_blocks = percpu_counter_read_positive(dbc);
 555         root_blocks = ext4_r_blocks_count(sbi->s_es);
 556
 557         if (free_blocks - (nblocks + root_blocks + dirty_blocks) <
 558                                                 EXT4_FREEBLOCKS_WATERMARK) {
 559                 free_blocks  = percpu_counter_sum_positive(fbc);
 560                 dirty_blocks = percpu_counter_sum_positive(dbc);
 561                 if (dirty_blocks < 0) {
 562                         printk(KERN_CRIT "Dirty block accounting "
 563                                         "went wrong %lld\n",
 564                                         (long long)dirty_blocks);
 565                 }
 566         }
 567         /* Check whether we have space after
 568          * accounting for current dirty blocks & root reserved blocks.
 569          */
 570         if (free_blocks >= ((root_blocks + nblocks) + dirty_blocks))
 571                 return 1;
 572
 573         /* Hm, nope.  Are (enough) root reserved blocks available? */
 574         if (sbi->s_resuid == current_fsuid() ||
 575             ((sbi->s_resgid != 0) && in_group_p(sbi->s_resgid)) ||
 576             capable(CAP_SYS_RESOURCE)) {
 577                 if (free_blocks >= (nblocks + dirty_blocks))
 578                         return 1;
 579         }
 580
 581         return 0;
 582 }
 583
 584 int ext4_claim_free_blocks(struct ext4_sb_info *sbi,
 585                                                 s64 nblocks)
 586 {
 587         if (ext4_has_free_blocks(sbi, nblocks)) {
 588                 percpu_counter_add(&sbi->s_dirtyblocks_counter, nblocks);
 589                 return 0;
 590         } else
 591                 return -ENOSPC;
 592 }
 593
 594 /**
 595  * ext4_should_retry_alloc()
 596  * @sb:                 super block
 597  * @retries             number of attemps has been made
 598  *
 599  * ext4_should_retry_alloc() is called when ENOSPC is returned, and if
 600  * it is profitable to retry the operation, this function will wait
 601  * for the current or commiting transaction to complete, and then
 602  * return TRUE.
 603  *
 604  * if the total number of retries exceed three times, return FALSE.
 605  */
 606 int ext4_should_retry_alloc(struct super_block *sb, int *retries)
 607 {
 608         if (!ext4_has_free_blocks(EXT4_SB(sb), 1) ||
 609             (*retries)++ > 3 ||
 610             !EXT4_SB(sb)->s_journal)
 611                 return 0;
 612
 613         jbd_debug(1, "%s: retrying operation after ENOSPC\n", sb->s_id);
 614
 615         return jbd2_journal_force_commit_nested(EXT4_SB(sb)->s_journal);
 616 }
 617
 618 /*
 619  * ext4_new_meta_blocks() -- allocate block for meta data (indexing) blocks
 620  *
 621  * @handle:             handle to this transaction
 622  * @inode:              file inode
 623  * @goal:               given target block(filesystem wide)
 624  * @count:              pointer to total number of blocks needed
 625  * @errp:               error code
 626  *
 627  * Return 1st allocated block number on success, *count stores total account
 628  * error stores in errp pointer
 629  */
 630 ext4_fsblk_t ext4_new_meta_blocks(handle_t *handle, struct inode *inode,
 631                 ext4_fsblk_t goal, unsigned long *count, int *errp)
 632 {
 633         struct ext4_allocation_request ar;
 634         ext4_fsblk_t ret;
 635
 636         memset(&ar, 0, sizeof(ar));
 637         /* Fill with neighbour allocated blocks */
 638         ar.inode = inode;
 639         ar.goal = goal;
 640         ar.len = count ? *count : 1;
 641
 642         ret = ext4_mb_new_blocks(handle, &ar, errp);
 643         if (count)
 644                 *count = ar.len;
 645
 646         /*
 647          * Account for the allocated meta blocks
 648          */
 649         if (!(*errp) && EXT4_I(inode)->i_delalloc_reserved_flag) {
 650                 spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
 651                 EXT4_I(inode)->i_allocated_meta_blocks += ar.len;
 652                 spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
 653         }
 654         return ret;
 655 }
 656
 657 /**
 658  * ext4_count_free_blocks() -- count filesystem free blocks
 659  * @sb:         superblock
 660  *
 661  * Adds up the number of free blocks from each block group.
 662  */
 663 ext4_fsblk_t ext4_count_free_blocks(struct super_block *sb)
 664 {
 665         ext4_fsblk_t desc_count;
 666         struct ext4_group_desc *gdp;
 667         ext4_group_t i;
 668         ext4_group_t ngroups = ext4_get_groups_count(sb);
 669 #ifdef EXT4FS_DEBUG
 670         struct ext4_super_block *es;
 671         ext4_fsblk_t bitmap_count;
 672         unsigned int x;
 673         struct buffer_head *bitmap_bh = NULL;
 674
 675         es = EXT4_SB(sb)->s_es;
 676         desc_count = 0;
 677         bitmap_count = 0;
 678         gdp = NULL;
 679
 680         for (i = 0; i < ngroups; i++) {
 681                 gdp = ext4_get_group_desc(sb, i, NULL);
 682                 if (!gdp)
 683                         continue;
 684                 desc_count += ext4_free_blks_count(sb, gdp);
 685                 brelse(bitmap_bh);
 686                 bitmap_bh = ext4_read_block_bitmap(sb, i);
 687                 if (bitmap_bh == NULL)
 688                         continue;
 689
 690                 x = ext4_count_free(bitmap_bh, sb->s_blocksize);
 691                 printk(KERN_DEBUG "group %u: stored = %d, counted = %u\n",
 692                         i, ext4_free_blks_count(sb, gdp), x);
 693                 bitmap_count += x;
 694         }
 695         brelse(bitmap_bh);
 696         printk(KERN_DEBUG "ext4_count_free_blocks: stored = %llu"
 697                 ", computed = %llu, %llu\n", ext4_free_blocks_count(es),
 698                desc_count, bitmap_count);
 699         return bitmap_count;
 700 #else
 701         desc_count = 0;
 702         for (i = 0; i < ngroups; i++) {
 703                 gdp = ext4_get_group_desc(sb, i, NULL);
 704                 if (!gdp)
 705                         continue;
 706                 desc_count += ext4_free_blks_count(sb, gdp);
 707         }
 708
 709         return desc_count;
 710 #endif
 711 }
 712
 713 static inline int test_root(ext4_group_t a, int b)
 714 {
 715         int num = b;
 716
 717         while (a > num)
 718                 num *= b;
 719         return num == a;
 720 }
 721
 722 static int ext4_group_sparse(ext4_group_t group)
 723 {
 724         if (group <= 1)
 725                 return 1;
 726         if (!(group & 1))
 727                 return 0;
 728         return (test_root(group, 7) || test_root(group, 5) ||
 729                 test_root(group, 3));
 730 }
 731
 732 /**
 733  *      ext4_bg_has_super - number of blocks used by the superblock in group
 734  *      @sb: superblock for filesystem
 735  *      @group: group number to check
 736  *
 737  *      Return the number of blocks used by the superblock (primary or backup)
 738  *      in this group.  Currently this will be only 0 or 1.
 739  */
 740 int ext4_bg_has_super(struct super_block *sb, ext4_group_t group)
 741 {
 742         if (EXT4_HAS_RO_COMPAT_FEATURE(sb,
 743                                 EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER) &&
 744                         !ext4_group_sparse(group))
 745                 return 0;
 746         return 1;
 747 }
 748
 749 static unsigned long ext4_bg_num_gdb_meta(struct super_block *sb,
 750                                         ext4_group_t group)
 751 {
 752         unsigned long metagroup = group / EXT4_DESC_PER_BLOCK(sb);
 753         ext4_group_t first = metagroup * EXT4_DESC_PER_BLOCK(sb);
 754         ext4_group_t last = first + EXT4_DESC_PER_BLOCK(sb) - 1;
 755
 756         if (group == first || group == first + 1 || group == last)
 757                 return 1;
 758         return 0;
 759 }
 760
 761 static unsigned long ext4_bg_num_gdb_nometa(struct super_block *sb,
 762                                         ext4_group_t group)
 763 {
 764         return ext4_bg_has_super(sb, group) ? EXT4_SB(sb)->s_gdb_count : 0;
 765 }
 766
 767 /**
 768  *      ext4_bg_num_gdb - number of blocks used by the group table in group
 769  *      @sb: superblock for filesystem
 770  *      @group: group number to check
 771  *
 772  *      Return the number of blocks used by the group descriptor table
 773  *      (primary or backup) in this group.  In the future there may be a
 774  *      different number of descriptor blocks in each group.
 775  */
 776 unsigned long ext4_bg_num_gdb(struct super_block *sb, ext4_group_t group)
 777 {
 778         unsigned long first_meta_bg =
 779                         le32_to_cpu(EXT4_SB(sb)->s_es->s_first_meta_bg);
 780         unsigned long metagroup = group / EXT4_DESC_PER_BLOCK(sb);
 781
 782         if (!EXT4_HAS_INCOMPAT_FEATURE(sb,EXT4_FEATURE_INCOMPAT_META_BG) ||
 783                         metagroup < first_meta_bg)
 784                 return ext4_bg_num_gdb_nometa(sb, group);
 785
 786         return ext4_bg_num_gdb_meta(sb,group);
 787
 788 }
 789