linux/fs/sysv/itree.c
<<
>>
Prefs
   1/*
   2 *  linux/fs/sysv/itree.c
   3 *
   4 *  Handling of indirect blocks' trees.
   5 *  AV, Sep--Dec 2000
   6 */
   7
   8#include <linux/buffer_head.h>
   9#include <linux/mount.h>
  10#include <linux/string.h>
  11#include "sysv.h"
  12
  13enum {DIRECT = 10, DEPTH = 4};  /* Have triple indirect */
  14
  15static inline void dirty_indirect(struct buffer_head *bh, struct inode *inode)
  16{
  17        mark_buffer_dirty_inode(bh, inode);
  18        if (IS_SYNC(inode))
  19                sync_dirty_buffer(bh);
  20}
  21
  22static int block_to_path(struct inode *inode, long block, int offsets[DEPTH])
  23{
  24        struct super_block *sb = inode->i_sb;
  25        struct sysv_sb_info *sbi = SYSV_SB(sb);
  26        int ptrs_bits = sbi->s_ind_per_block_bits;
  27        unsigned long   indirect_blocks = sbi->s_ind_per_block,
  28                        double_blocks = sbi->s_ind_per_block_2;
  29        int n = 0;
  30
  31        if (block < 0) {
  32                printk("sysv_block_map: block < 0\n");
  33        } else if (block < DIRECT) {
  34                offsets[n++] = block;
  35        } else if ( (block -= DIRECT) < indirect_blocks) {
  36                offsets[n++] = DIRECT;
  37                offsets[n++] = block;
  38        } else if ((block -= indirect_blocks) < double_blocks) {
  39                offsets[n++] = DIRECT+1;
  40                offsets[n++] = block >> ptrs_bits;
  41                offsets[n++] = block & (indirect_blocks - 1);
  42        } else if (((block -= double_blocks) >> (ptrs_bits * 2)) < indirect_blocks) {
  43                offsets[n++] = DIRECT+2;
  44                offsets[n++] = block >> (ptrs_bits * 2);
  45                offsets[n++] = (block >> ptrs_bits) & (indirect_blocks - 1);
  46                offsets[n++] = block & (indirect_blocks - 1);
  47        } else {
  48                /* nothing */;
  49        }
  50        return n;
  51}
  52
  53static inline int block_to_cpu(struct sysv_sb_info *sbi, sysv_zone_t nr)
  54{
  55        return sbi->s_block_base + fs32_to_cpu(sbi, nr);
  56}
  57
  58typedef struct {
  59        sysv_zone_t     *p;
  60        sysv_zone_t     key;
  61        struct buffer_head *bh;
  62} Indirect;
  63
  64static DEFINE_RWLOCK(pointers_lock);
  65
  66static inline void add_chain(Indirect *p, struct buffer_head *bh, sysv_zone_t *v)
  67{
  68        p->key = *(p->p = v);
  69        p->bh = bh;
  70}
  71
  72static inline int verify_chain(Indirect *from, Indirect *to)
  73{
  74        while (from <= to && from->key == *from->p)
  75                from++;
  76        return (from > to);
  77}
  78
  79static inline sysv_zone_t *block_end(struct buffer_head *bh)
  80{
  81        return (sysv_zone_t*)((char*)bh->b_data + bh->b_size);
  82}
  83
  84/*
  85 * Requires read_lock(&pointers_lock) or write_lock(&pointers_lock)
  86 */
  87static Indirect *get_branch(struct inode *inode,
  88                            int depth,
  89                            int offsets[],
  90                            Indirect chain[],
  91                            int *err)
  92{
  93        struct super_block *sb = inode->i_sb;
  94        Indirect *p = chain;
  95        struct buffer_head *bh;
  96
  97        *err = 0;
  98        add_chain(chain, NULL, SYSV_I(inode)->i_data + *offsets);
  99        if (!p->key)
 100                goto no_block;
 101        while (--depth) {
 102                int block = block_to_cpu(SYSV_SB(sb), p->key);
 103                bh = sb_bread(sb, block);
 104                if (!bh)
 105                        goto failure;
 106                if (!verify_chain(chain, p))
 107                        goto changed;
 108                add_chain(++p, bh, (sysv_zone_t*)bh->b_data + *++offsets);
 109                if (!p->key)
 110                        goto no_block;
 111        }
 112        return NULL;
 113
 114changed:
 115        brelse(bh);
 116        *err = -EAGAIN;
 117        goto no_block;
 118failure:
 119        *err = -EIO;
 120no_block:
 121        return p;
 122}
 123
 124static int alloc_branch(struct inode *inode,
 125                        int num,
 126                        int *offsets,
 127                        Indirect *branch)
 128{
 129        int blocksize = inode->i_sb->s_blocksize;
 130        int n = 0;
 131        int i;
 132
 133        branch[0].key = sysv_new_block(inode->i_sb);
 134        if (branch[0].key) for (n = 1; n < num; n++) {
 135                struct buffer_head *bh;
 136                int parent;
 137                /* Allocate the next block */
 138                branch[n].key = sysv_new_block(inode->i_sb);
 139                if (!branch[n].key)
 140                        break;
 141                /*
 142                 * Get buffer_head for parent block, zero it out and set 
 143                 * the pointer to new one, then send parent to disk.
 144                 */
 145                parent = block_to_cpu(SYSV_SB(inode->i_sb), branch[n-1].key);
 146                bh = sb_getblk(inode->i_sb, parent);
 147                lock_buffer(bh);
 148                memset(bh->b_data, 0, blocksize);
 149                branch[n].bh = bh;
 150                branch[n].p = (sysv_zone_t*) bh->b_data + offsets[n];
 151                *branch[n].p = branch[n].key;
 152                set_buffer_uptodate(bh);
 153                unlock_buffer(bh);
 154                dirty_indirect(bh, inode);
 155        }
 156        if (n == num)
 157                return 0;
 158
 159        /* Allocation failed, free what we already allocated */
 160        for (i = 1; i < n; i++)
 161                bforget(branch[i].bh);
 162        for (i = 0; i < n; i++)
 163                sysv_free_block(inode->i_sb, branch[i].key);
 164        return -ENOSPC;
 165}
 166
 167static inline int splice_branch(struct inode *inode,
 168                                Indirect chain[],
 169                                Indirect *where,
 170                                int num)
 171{
 172        int i;
 173
 174        /* Verify that place we are splicing to is still there and vacant */
 175        write_lock(&pointers_lock);
 176        if (!verify_chain(chain, where-1) || *where->p)
 177                goto changed;
 178        *where->p = where->key;
 179        write_unlock(&pointers_lock);
 180
 181        inode->i_ctime = CURRENT_TIME_SEC;
 182
 183        /* had we spliced it onto indirect block? */
 184        if (where->bh)
 185                dirty_indirect(where->bh, inode);
 186
 187        if (IS_SYNC(inode))
 188                sysv_sync_inode(inode);
 189        else
 190                mark_inode_dirty(inode);
 191        return 0;
 192
 193changed:
 194        write_unlock(&pointers_lock);
 195        for (i = 1; i < num; i++)
 196                bforget(where[i].bh);
 197        for (i = 0; i < num; i++)
 198                sysv_free_block(inode->i_sb, where[i].key);
 199        return -EAGAIN;
 200}
 201
 202static int get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create)
 203{
 204        int err = -EIO;
 205        int offsets[DEPTH];
 206        Indirect chain[DEPTH];
 207        struct super_block *sb = inode->i_sb;
 208        Indirect *partial;
 209        int left;
 210        int depth = block_to_path(inode, iblock, offsets);
 211
 212        if (depth == 0)
 213                goto out;
 214
 215reread:
 216        read_lock(&pointers_lock);
 217        partial = get_branch(inode, depth, offsets, chain, &err);
 218        read_unlock(&pointers_lock);
 219
 220        /* Simplest case - block found, no allocation needed */
 221        if (!partial) {
 222got_it:
 223                map_bh(bh_result, sb, block_to_cpu(SYSV_SB(sb),
 224                                        chain[depth-1].key));
 225                /* Clean up and exit */
 226                partial = chain+depth-1; /* the whole chain */
 227                goto cleanup;
 228        }
 229
 230        /* Next simple case - plain lookup or failed read of indirect block */
 231        if (!create || err == -EIO) {
 232cleanup:
 233                while (partial > chain) {
 234                        brelse(partial->bh);
 235                        partial--;
 236                }
 237out:
 238                return err;
 239        }
 240
 241        /*
 242         * Indirect block might be removed by truncate while we were
 243         * reading it. Handling of that case (forget what we've got and
 244         * reread) is taken out of the main path.
 245         */
 246        if (err == -EAGAIN)
 247                goto changed;
 248
 249        left = (chain + depth) - partial;
 250        err = alloc_branch(inode, left, offsets+(partial-chain), partial);
 251        if (err)
 252                goto cleanup;
 253
 254        if (splice_branch(inode, chain, partial, left) < 0)
 255                goto changed;
 256
 257        set_buffer_new(bh_result);
 258        goto got_it;
 259
 260changed:
 261        while (partial > chain) {
 262                brelse(partial->bh);
 263                partial--;
 264        }
 265        goto reread;
 266}
 267
 268static inline int all_zeroes(sysv_zone_t *p, sysv_zone_t *q)
 269{
 270        while (p < q)
 271                if (*p++)
 272                        return 0;
 273        return 1;
 274}
 275
 276static Indirect *find_shared(struct inode *inode,
 277                                int depth,
 278                                int offsets[],
 279                                Indirect chain[],
 280                                sysv_zone_t *top)
 281{
 282        Indirect *partial, *p;
 283        int k, err;
 284
 285        *top = 0;
 286        for (k = depth; k > 1 && !offsets[k-1]; k--)
 287                ;
 288
 289        write_lock(&pointers_lock);
 290        partial = get_branch(inode, k, offsets, chain, &err);
 291        if (!partial)
 292                partial = chain + k-1;
 293        /*
 294         * If the branch acquired continuation since we've looked at it -
 295         * fine, it should all survive and (new) top doesn't belong to us.
 296         */
 297        if (!partial->key && *partial->p) {
 298                write_unlock(&pointers_lock);
 299                goto no_top;
 300        }
 301        for (p=partial; p>chain && all_zeroes((sysv_zone_t*)p->bh->b_data,p->p); p--)
 302                ;
 303        /*
 304         * OK, we've found the last block that must survive. The rest of our
 305         * branch should be detached before unlocking. However, if that rest
 306         * of branch is all ours and does not grow immediately from the inode
 307         * it's easier to cheat and just decrement partial->p.
 308         */
 309        if (p == chain + k - 1 && p > chain) {
 310                p->p--;
 311        } else {
 312                *top = *p->p;
 313                *p->p = 0;
 314        }
 315        write_unlock(&pointers_lock);
 316
 317        while (partial > p) {
 318                brelse(partial->bh);
 319                partial--;
 320        }
 321no_top:
 322        return partial;
 323}
 324
 325static inline void free_data(struct inode *inode, sysv_zone_t *p, sysv_zone_t *q)
 326{
 327        for ( ; p < q ; p++) {
 328                sysv_zone_t nr = *p;
 329                if (nr) {
 330                        *p = 0;
 331                        sysv_free_block(inode->i_sb, nr);
 332                        mark_inode_dirty(inode);
 333                }
 334        }
 335}
 336
 337static void free_branches(struct inode *inode, sysv_zone_t *p, sysv_zone_t *q, int depth)
 338{
 339        struct buffer_head * bh;
 340        struct super_block *sb = inode->i_sb;
 341
 342        if (depth--) {
 343                for ( ; p < q ; p++) {
 344                        int block;
 345                        sysv_zone_t nr = *p;
 346                        if (!nr)
 347                                continue;
 348                        *p = 0;
 349                        block = block_to_cpu(SYSV_SB(sb), nr);
 350                        bh = sb_bread(sb, block);
 351                        if (!bh)
 352                                continue;
 353                        free_branches(inode, (sysv_zone_t*)bh->b_data,
 354                                        block_end(bh), depth);
 355                        bforget(bh);
 356                        sysv_free_block(sb, nr);
 357                        mark_inode_dirty(inode);
 358                }
 359        } else
 360                free_data(inode, p, q);
 361}
 362
 363void sysv_truncate (struct inode * inode)
 364{
 365        sysv_zone_t *i_data = SYSV_I(inode)->i_data;
 366        int offsets[DEPTH];
 367        Indirect chain[DEPTH];
 368        Indirect *partial;
 369        sysv_zone_t nr = 0;
 370        int n;
 371        long iblock;
 372        unsigned blocksize;
 373
 374        if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) ||
 375            S_ISLNK(inode->i_mode)))
 376                return;
 377
 378        blocksize = inode->i_sb->s_blocksize;
 379        iblock = (inode->i_size + blocksize-1)
 380                                        >> inode->i_sb->s_blocksize_bits;
 381
 382        block_truncate_page(inode->i_mapping, inode->i_size, get_block);
 383
 384        n = block_to_path(inode, iblock, offsets);
 385        if (n == 0)
 386                return;
 387
 388        if (n == 1) {
 389                free_data(inode, i_data+offsets[0], i_data + DIRECT);
 390                goto do_indirects;
 391        }
 392
 393        partial = find_shared(inode, n, offsets, chain, &nr);
 394        /* Kill the top of shared branch (already detached) */
 395        if (nr) {
 396                if (partial == chain)
 397                        mark_inode_dirty(inode);
 398                else
 399                        dirty_indirect(partial->bh, inode);
 400                free_branches(inode, &nr, &nr+1, (chain+n-1) - partial);
 401        }
 402        /* Clear the ends of indirect blocks on the shared branch */
 403        while (partial > chain) {
 404                free_branches(inode, partial->p + 1, block_end(partial->bh),
 405                                (chain+n-1) - partial);
 406                dirty_indirect(partial->bh, inode);
 407                brelse (partial->bh);
 408                partial--;
 409        }
 410do_indirects:
 411        /* Kill the remaining (whole) subtrees (== subtrees deeper than...) */
 412        while (n < DEPTH) {
 413                nr = i_data[DIRECT + n - 1];
 414                if (nr) {
 415                        i_data[DIRECT + n - 1] = 0;
 416                        mark_inode_dirty(inode);
 417                        free_branches(inode, &nr, &nr+1, n);
 418                }
 419                n++;
 420        }
 421        inode->i_mtime = inode->i_ctime = CURRENT_TIME_SEC;
 422        if (IS_SYNC(inode))
 423                sysv_sync_inode (inode);
 424        else
 425                mark_inode_dirty(inode);
 426}
 427
 428static unsigned sysv_nblocks(struct super_block *s, loff_t size)
 429{
 430        struct sysv_sb_info *sbi = SYSV_SB(s);
 431        int ptrs_bits = sbi->s_ind_per_block_bits;
 432        unsigned blocks, res, direct = DIRECT, i = DEPTH;
 433        blocks = (size + s->s_blocksize - 1) >> s->s_blocksize_bits;
 434        res = blocks;
 435        while (--i && blocks > direct) {
 436                blocks = ((blocks - direct - 1) >> ptrs_bits) + 1;
 437                res += blocks;
 438                direct = 1;
 439        }
 440        return blocks;
 441}
 442
 443int sysv_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 444{
 445        struct super_block *s = dentry->d_sb;
 446        generic_fillattr(dentry->d_inode, stat);
 447        stat->blocks = (s->s_blocksize / 512) * sysv_nblocks(s, stat->size);
 448        stat->blksize = s->s_blocksize;
 449        return 0;
 450}
 451
 452static int sysv_writepage(struct page *page, struct writeback_control *wbc)
 453{
 454        return block_write_full_page(page,get_block,wbc);
 455}
 456
 457static int sysv_readpage(struct file *file, struct page *page)
 458{
 459        return block_read_full_page(page,get_block);
 460}
 461
 462int sysv_prepare_chunk(struct page *page, loff_t pos, unsigned len)
 463{
 464        return __block_write_begin(page, pos, len, get_block);
 465}
 466
 467static void sysv_write_failed(struct address_space *mapping, loff_t to)
 468{
 469        struct inode *inode = mapping->host;
 470
 471        if (to > inode->i_size) {
 472                truncate_pagecache(inode, inode->i_size);
 473                sysv_truncate(inode);
 474        }
 475}
 476
 477static int sysv_write_begin(struct file *file, struct address_space *mapping,
 478                        loff_t pos, unsigned len, unsigned flags,
 479                        struct page **pagep, void **fsdata)
 480{
 481        int ret;
 482
 483        ret = block_write_begin(mapping, pos, len, flags, pagep, get_block);
 484        if (unlikely(ret))
 485                sysv_write_failed(mapping, pos + len);
 486
 487        return ret;
 488}
 489
 490static sector_t sysv_bmap(struct address_space *mapping, sector_t block)
 491{
 492        return generic_block_bmap(mapping,block,get_block);
 493}
 494
 495const struct address_space_operations sysv_aops = {
 496        .readpage = sysv_readpage,
 497        .writepage = sysv_writepage,
 498        .write_begin = sysv_write_begin,
 499        .write_end = generic_write_end,
 500        .bmap = sysv_bmap
 501};
 502