fs/jffs2/build.c

   1 /*
   2  * JFFS2 -- Journalling Flash File System, Version 2.
   3  *
   4  * Copyright © 2001-2007 Red Hat, Inc.
   5  *
   6  * Created by David Woodhouse <dwmw2@infradead.org>
   7  *
   8  * For licensing information, see the file 'LICENCE' in this directory.
   9  *
  10  */
  11
  12 #include <linux/kernel.h>
  13 #include <linux/sched.h>
  14 #include <linux/slab.h>
  15 #include <linux/vmalloc.h>
  16 #include <linux/mtd/mtd.h>
  17 #include "nodelist.h"
  18
  19 static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *,
  20                 struct jffs2_inode_cache *, struct jffs2_full_dirent **);
  21
  22 static inline struct jffs2_inode_cache *
  23 first_inode_chain(int *i, struct jffs2_sb_info *c)
  24 {
  25         for (; *i < INOCACHE_HASHSIZE; (*i)++) {
  26                 if (c->inocache_list[*i])
  27                         return c->inocache_list[*i];
  28         }
  29         return NULL;
  30 }
  31
  32 static inline struct jffs2_inode_cache *
  33 next_inode(int *i, struct jffs2_inode_cache *ic, struct jffs2_sb_info *c)
  34 {
  35         /* More in this chain? */
  36         if (ic->next)
  37                 return ic->next;
  38         (*i)++;
  39         return first_inode_chain(i, c);
  40 }
  41
  42 #define for_each_inode(i, c, ic)                        \
  43         for (i = 0, ic = first_inode_chain(&i, (c));    \
  44              ic;                                        \
  45              ic = next_inode(&i, ic, (c)))
  46
  47
  48 static void jffs2_build_inode_pass1(struct jffs2_sb_info *c,
  49                                     struct jffs2_inode_cache *ic)
  50 {
  51         struct jffs2_full_dirent *fd;
  52
  53         dbg_fsbuild("building directory inode #%u\n", ic->ino);
  54
  55         /* For each child, increase nlink */
  56         for(fd = ic->scan_dents; fd; fd = fd->next) {
  57                 struct jffs2_inode_cache *child_ic;
  58                 if (!fd->ino)
  59                         continue;
  60
  61                 /* we can get high latency here with huge directories */
  62
  63                 child_ic = jffs2_get_ino_cache(c, fd->ino);
  64                 if (!child_ic) {
  65                         dbg_fsbuild("child \"%s\" (ino #%u) of dir ino #%u doesn't exist!\n",
  66                                   fd->name, fd->ino, ic->ino);
  67                         jffs2_mark_node_obsolete(c, fd->raw);
  68                         continue;
  69                 }
  70
  71                 if (fd->type == DT_DIR) {
  72                         if (child_ic->pino_nlink) {
  73                                 JFFS2_ERROR("child dir \"%s\" (ino #%u) of dir ino #%u appears to be a hard link\n",
  74                                             fd->name, fd->ino, ic->ino);
  75                                 /* TODO: What do we do about it? */
  76                         } else {
  77                                 child_ic->pino_nlink = ic->ino;
  78                         }
  79                 } else
  80                         child_ic->pino_nlink++;
  81
  82                 dbg_fsbuild("increased nlink for child \"%s\" (ino #%u)\n", fd->name, fd->ino);
  83                 /* Can't free scan_dents so far. We might need them in pass 2 */
  84         }
  85 }
  86
  87 /* Scan plan:
  88  - Scan physical nodes. Build map of inodes/dirents. Allocate inocaches as we go
  89  - Scan directory tree from top down, setting nlink in inocaches
  90  - Scan inocaches for inodes with nlink==0
  91 */
  92 static int jffs2_build_filesystem(struct jffs2_sb_info *c)
  93 {
  94         int ret;
  95         int i;
  96         struct jffs2_inode_cache *ic;
  97         struct jffs2_full_dirent *fd;
  98         struct jffs2_full_dirent *dead_fds = NULL;
  99
 100         dbg_fsbuild("build FS data structures\n");
 101
 102         /* First, scan the medium and build all the inode caches with
 103            lists of physical nodes */
 104
 105         c->flags |= JFFS2_SB_FLAG_SCANNING;
 106         ret = jffs2_scan_medium(c);
 107         c->flags &= ~JFFS2_SB_FLAG_SCANNING;
 108         if (ret)
 109                 goto exit;
 110
 111         dbg_fsbuild("scanned flash completely\n");
 112         jffs2_dbg_dump_block_lists_nolock(c);
 113
 114         dbg_fsbuild("pass 1 starting\n");
 115         c->flags |= JFFS2_SB_FLAG_BUILDING;
 116         /* Now scan the directory tree, increasing nlink according to every dirent found. */
 117         for_each_inode(i, c, ic) {
 118                 if (ic->scan_dents) {
 119                         jffs2_build_inode_pass1(c, ic);
 120                         cond_resched();
 121                 }
 122         }
 123
 124         dbg_fsbuild("pass 1 complete\n");
 125
 126         /* Next, scan for inodes with nlink == 0 and remove them. If
 127            they were directories, then decrement the nlink of their
 128            children too, and repeat the scan. As that's going to be
 129            a fairly uncommon occurrence, it's not so evil to do it this
 130            way. Recursion bad. */
 131         dbg_fsbuild("pass 2 starting\n");
 132
 133         for_each_inode(i, c, ic) {
 134                 if (ic->pino_nlink)
 135                         continue;
 136
 137                 jffs2_build_remove_unlinked_inode(c, ic, &dead_fds);
 138                 cond_resched();
 139         }
 140
 141         dbg_fsbuild("pass 2a starting\n");
 142
 143         while (dead_fds) {
 144                 fd = dead_fds;
 145                 dead_fds = fd->next;
 146
 147                 ic = jffs2_get_ino_cache(c, fd->ino);
 148
 149                 if (ic)
 150                         jffs2_build_remove_unlinked_inode(c, ic, &dead_fds);
 151                 jffs2_free_full_dirent(fd);
 152         }
 153
 154         dbg_fsbuild("pass 2a complete\n");
 155         dbg_fsbuild("freeing temporary data structures\n");
 156
 157         /* Finally, we can scan again and free the dirent structs */
 158         for_each_inode(i, c, ic) {
 159                 while(ic->scan_dents) {
 160                         fd = ic->scan_dents;
 161                         ic->scan_dents = fd->next;
 162                         jffs2_free_full_dirent(fd);
 163                 }
 164                 ic->scan_dents = NULL;
 165                 cond_resched();
 166         }
 167         jffs2_build_xattr_subsystem(c);
 168         c->flags &= ~JFFS2_SB_FLAG_BUILDING;
 169
 170         dbg_fsbuild("FS build complete\n");
 171
 172         /* Rotate the lists by some number to ensure wear levelling */
 173         jffs2_rotate_lists(c);
 174
 175         ret = 0;
 176
 177 exit:
 178         if (ret) {
 179                 for_each_inode(i, c, ic) {
 180                         while(ic->scan_dents) {
 181                                 fd = ic->scan_dents;
 182                                 ic->scan_dents = fd->next;
 183                                 jffs2_free_full_dirent(fd);
 184                         }
 185                 }
 186                 jffs2_clear_xattr_subsystem(c);
 187         }
 188
 189         return ret;
 190 }
 191
 192 static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *c,
 193                                         struct jffs2_inode_cache *ic,
 194                                         struct jffs2_full_dirent **dead_fds)
 195 {
 196         struct jffs2_raw_node_ref *raw;
 197         struct jffs2_full_dirent *fd;
 198
 199         dbg_fsbuild("removing ino #%u with nlink == zero.\n", ic->ino);
 200
 201         raw = ic->nodes;
 202         while (raw != (void *)ic) {
 203                 struct jffs2_raw_node_ref *next = raw->next_in_ino;
 204                 dbg_fsbuild("obsoleting node at 0x%08x\n", ref_offset(raw));
 205                 jffs2_mark_node_obsolete(c, raw);
 206                 raw = next;
 207         }
 208
 209         if (ic->scan_dents) {
 210                 int whinged = 0;
 211                 dbg_fsbuild("inode #%u was a directory which may have children...\n", ic->ino);
 212
 213                 while(ic->scan_dents) {
 214                         struct jffs2_inode_cache *child_ic;
 215
 216                         fd = ic->scan_dents;
 217                         ic->scan_dents = fd->next;
 218
 219                         if (!fd->ino) {
 220                                 /* It's a deletion dirent. Ignore it */
 221                                 dbg_fsbuild("child \"%s\" is a deletion dirent, skipping...\n", fd->name);
 222                                 jffs2_free_full_dirent(fd);
 223                                 continue;
 224                         }
 225                         if (!whinged)
 226                                 whinged = 1;
 227
 228                         dbg_fsbuild("removing child \"%s\", ino #%u\n", fd->name, fd->ino);
 229
 230                         child_ic = jffs2_get_ino_cache(c, fd->ino);
 231                         if (!child_ic) {
 232                                 dbg_fsbuild("cannot remove child \"%s\", ino #%u, because it doesn't exist\n",
 233                                                 fd->name, fd->ino);
 234                                 jffs2_free_full_dirent(fd);
 235                                 continue;
 236                         }
 237
 238                         /* Reduce nlink of the child. If it's now zero, stick it on the
 239                            dead_fds list to be cleaned up later. Else just free the fd */
 240
 241                         if (fd->type == DT_DIR)
 242                                 child_ic->pino_nlink = 0;
 243                         else
 244                                 child_ic->pino_nlink--;
 245
 246                         if (!child_ic->pino_nlink) {
 247                                 dbg_fsbuild("inode #%u (\"%s\") now has no links; adding to dead_fds list.\n",
 248                                           fd->ino, fd->name);
 249                                 fd->next = *dead_fds;
 250                                 *dead_fds = fd;
 251                         } else {
 252                                 dbg_fsbuild("inode #%u (\"%s\") has now got nlink %d. Ignoring.\n",
 253                                           fd->ino, fd->name, child_ic->pino_nlink);
 254                                 jffs2_free_full_dirent(fd);
 255                         }
 256                 }
 257         }
 258
 259         /*
 260            We don't delete the inocache from the hash list and free it yet.
 261            The erase code will do that, when all the nodes are completely gone.
 262         */
 263 }
 264
 265 static void jffs2_calc_trigger_levels(struct jffs2_sb_info *c)
 266 {
 267         uint32_t size;
 268
 269         /* Deletion should almost _always_ be allowed. We're fairly
 270            buggered once we stop allowing people to delete stuff
 271            because there's not enough free space... */
 272         c->resv_blocks_deletion = 2;
 273
 274         /* Be conservative about how much space we need before we allow writes.
 275            On top of that which is required for deletia, require an extra 2%
 276            of the medium to be available, for overhead caused by nodes being
 277            split across blocks, etc. */
 278
 279         size = c->flash_size / 50; /* 2% of flash size */
 280         size += c->nr_blocks * 100; /* And 100 bytes per eraseblock */
 281         size += c->sector_size - 1; /* ... and round up */
 282
 283         c->resv_blocks_write = c->resv_blocks_deletion + (size / c->sector_size);
 284
 285         /* When do we let the GC thread run in the background */
 286
 287         c->resv_blocks_gctrigger = c->resv_blocks_write + 1;
 288
 289         /* When do we allow garbage collection to merge nodes to make
 290            long-term progress at the expense of short-term space exhaustion? */
 291         c->resv_blocks_gcmerge = c->resv_blocks_deletion + 1;
 292
 293         /* When do we allow garbage collection to eat from bad blocks rather
 294            than actually making progress? */
 295         c->resv_blocks_gcbad = 0;//c->resv_blocks_deletion + 2;
 296
 297         /* What number of 'very dirty' eraseblocks do we allow before we
 298            trigger the GC thread even if we don't _need_ the space. When we
 299            can't mark nodes obsolete on the medium, the old dirty nodes cause
 300            performance problems because we have to inspect and discard them. */
 301         c->vdirty_blocks_gctrigger = c->resv_blocks_gctrigger;
 302         if (jffs2_can_mark_obsolete(c))
 303                 c->vdirty_blocks_gctrigger *= 10;
 304
 305         /* If there's less than this amount of dirty space, don't bother
 306            trying to GC to make more space. It'll be a fruitless task */
 307         c->nospc_dirty_size = c->sector_size + (c->flash_size / 100);
 308
 309         dbg_fsbuild("JFFS2 trigger levels (size %d KiB, block size %d KiB, %d blocks)\n",
 310                   c->flash_size / 1024, c->sector_size / 1024, c->nr_blocks);
 311         dbg_fsbuild("Blocks required to allow deletion:    %d (%d KiB)\n",
 312                   c->resv_blocks_deletion, c->resv_blocks_deletion*c->sector_size/1024);
 313         dbg_fsbuild("Blocks required to allow writes:      %d (%d KiB)\n",
 314                   c->resv_blocks_write, c->resv_blocks_write*c->sector_size/1024);
 315         dbg_fsbuild("Blocks required to quiesce GC thread: %d (%d KiB)\n",
 316                   c->resv_blocks_gctrigger, c->resv_blocks_gctrigger*c->sector_size/1024);
 317         dbg_fsbuild("Blocks required to allow GC merges:   %d (%d KiB)\n",
 318                   c->resv_blocks_gcmerge, c->resv_blocks_gcmerge*c->sector_size/1024);
 319         dbg_fsbuild("Blocks required to GC bad blocks:     %d (%d KiB)\n",
 320                   c->resv_blocks_gcbad, c->resv_blocks_gcbad*c->sector_size/1024);
 321         dbg_fsbuild("Amount of dirty space required to GC: %d bytes\n",
 322                   c->nospc_dirty_size);
 323         dbg_fsbuild("Very dirty blocks before GC triggered: %d\n",
 324                   c->vdirty_blocks_gctrigger);
 325 }
 326
 327 int jffs2_do_mount_fs(struct jffs2_sb_info *c)
 328 {
 329         int ret;
 330         int i;
 331         int size;
 332
 333         c->free_size = c->flash_size;
 334         c->nr_blocks = c->flash_size / c->sector_size;
 335         size = sizeof(struct jffs2_eraseblock) * c->nr_blocks;
 336 #ifndef __ECOS
 337         if (jffs2_blocks_use_vmalloc(c))
 338                 c->blocks = vmalloc(size);
 339         else
 340 #endif
 341                 c->blocks = kmalloc(size, GFP_KERNEL);
 342         if (!c->blocks)
 343                 return -ENOMEM;
 344
 345         memset(c->blocks, 0, size);
 346         for (i=0; i<c->nr_blocks; i++) {
 347                 INIT_LIST_HEAD(&c->blocks[i].list);
 348                 c->blocks[i].offset = i * c->sector_size;
 349                 c->blocks[i].free_size = c->sector_size;
 350         }
 351
 352         INIT_LIST_HEAD(&c->clean_list);
 353         INIT_LIST_HEAD(&c->very_dirty_list);
 354         INIT_LIST_HEAD(&c->dirty_list);
 355         INIT_LIST_HEAD(&c->erasable_list);
 356         INIT_LIST_HEAD(&c->erasing_list);
 357         INIT_LIST_HEAD(&c->erase_checking_list);
 358         INIT_LIST_HEAD(&c->erase_pending_list);
 359         INIT_LIST_HEAD(&c->erasable_pending_wbuf_list);
 360         INIT_LIST_HEAD(&c->erase_complete_list);
 361         INIT_LIST_HEAD(&c->free_list);
 362         INIT_LIST_HEAD(&c->bad_list);
 363         INIT_LIST_HEAD(&c->bad_used_list);
 364         c->highest_ino = 1;
 365         c->summary = NULL;
 366
 367         ret = jffs2_sum_init(c);
 368         if (ret)
 369                 goto out_free;
 370
 371         if (jffs2_build_filesystem(c)) {
 372                 dbg_fsbuild("build_fs failed\n");
 373                 jffs2_free_ino_caches(c);
 374                 jffs2_free_raw_node_refs(c);
 375                 ret = -EIO;
 376                 goto out_free;
 377         }
 378
 379         jffs2_calc_trigger_levels(c);
 380
 381         return 0;
 382
 383  out_free:
 384 #ifndef __ECOS
 385         if (jffs2_blocks_use_vmalloc(c))
 386                 vfree(c->blocks);
 387         else
 388 #endif
 389                 kfree(c->blocks);
 390
 391         return ret;
 392 }