The Design and Implementation of the FreeBSD Operating System, Second Edition
Now available: The Design and Implementation of the FreeBSD Operating System (Second Edition)


[ source navigation ] [ diff markup ] [ identifier search ] [ freetext search ] [ file search ] [ list types ] [ track identifier ]

FreeBSD/Linux Kernel Cross Reference
sys/fs/reiserfs/stree.c

Version: -  FREEBSD  -  FREEBSD-13-STABLE  -  FREEBSD-13-0  -  FREEBSD-12-STABLE  -  FREEBSD-12-0  -  FREEBSD-11-STABLE  -  FREEBSD-11-0  -  FREEBSD-10-STABLE  -  FREEBSD-10-0  -  FREEBSD-9-STABLE  -  FREEBSD-9-0  -  FREEBSD-8-STABLE  -  FREEBSD-8-0  -  FREEBSD-7-STABLE  -  FREEBSD-7-0  -  FREEBSD-6-STABLE  -  FREEBSD-6-0  -  FREEBSD-5-STABLE  -  FREEBSD-5-0  -  FREEBSD-4-STABLE  -  FREEBSD-3-STABLE  -  FREEBSD22  -  l41  -  OPENBSD  -  linux-2.6  -  MK84  -  PLAN9  -  xnu-8792 
SearchContext: -  none  -  3  -  10 

    1 /*
    2  *  Copyright 2000-2002 by Hans Reiser, licensing governed by reiserfs/README
    3  */
    4 
    5 /*
    6  *  Written by Anatoly P. Pinchuk pap@namesys.botik.ru
    7  *  Programm System Institute
    8  *  Pereslavl-Zalessky Russia
    9  */
   10 
   11 /*
   12  *  This file contains functions dealing with S+tree
   13  *
   14  * B_IS_IN_TREE
   15  * copy_short_key
   16  * copy_item_head
   17  * comp_short_keys
   18  * comp_keys
   19  * comp_cpu_keys
   20  * comp_short_le_keys
   21  * comp_short_cpu_keys
   22  * cpu_key2cpu_key
   23  * le_key2cpu_key
   24  * comp_le_keys
   25  * bin_search
   26  * get_lkey
   27  * get_rkey
   28  * key_in_buffer
   29  * decrement_bcount
   30  * decrement_counters_in_path
   31  * reiserfs_check_path
   32  * pathrelse_and_restore
   33  * pathrelse
   34  * search_by_key_reada
   35  * search_by_key
   36  * search_for_position_by_key
   37  * comp_items
   38  * prepare_for_direct_item
   39  * prepare_for_direntry_item
   40  * prepare_for_delete_or_cut
   41  * calc_deleted_bytes_number
   42  * init_tb_struct
   43  * padd_item
   44  * reiserfs_delete_item
   45  * reiserfs_delete_solid_item
   46  * reiserfs_delete_object
   47  * maybe_indirect_to_direct
   48  * indirect_to_direct_roll_back
   49  * reiserfs_cut_from_item
   50  * truncate_directory
   51  * reiserfs_do_truncate
   52  * reiserfs_paste_into_item
   53  * reiserfs_insert_item
   54  */
   55 
   56 #include <linux/config.h>
   57 #include <linux/sched.h>
   58 #include <linux/string.h>
   59 #include <linux/locks.h>
   60 #include <linux/pagemap.h>
   61 #include <linux/reiserfs_fs.h>
   62 #include <linux/smp_lock.h>
   63 
   64 /* Does the buffer contain a disk block which is in the tree. */
   65 inline int B_IS_IN_TREE (const struct buffer_head * p_s_bh)
   66 {
   67 
   68   RFALSE( B_LEVEL (p_s_bh) > MAX_HEIGHT,
   69           "PAP-1010: block (%b) has too big level (%z)", p_s_bh, p_s_bh);
   70 
   71   return ( B_LEVEL (p_s_bh) != FREE_LEVEL );
   72 }
   73 
   74 
   75 
   76 
   77 inline void copy_short_key (void * to, const void * from)
   78 {
   79     memcpy (to, from, SHORT_KEY_SIZE);
   80 }
   81 
   82 //
   83 // to gets item head in le form
   84 //
   85 inline void copy_item_head(struct item_head * p_v_to, 
   86                            const struct item_head * p_v_from)
   87 {
   88   memcpy (p_v_to, p_v_from, IH_SIZE);
   89 }
   90 
   91 
   92 /* k1 is pointer to on-disk structure which is stored in little-endian
   93    form. k2 is pointer to cpu variable. For key of items of the same
   94    object this returns 0.
   95    Returns: -1 if key1 < key2 
   96    0 if key1 == key2
   97    1 if key1 > key2 */
   98 inline int  comp_short_keys (const struct key * le_key, 
   99                              const struct cpu_key * cpu_key)
  100 {
  101   __u32 * p_s_le_u32, * p_s_cpu_u32;
  102   int n_key_length = REISERFS_SHORT_KEY_LEN;
  103 
  104   p_s_le_u32 = (__u32 *)le_key;
  105   p_s_cpu_u32 = (__u32 *)cpu_key;
  106   for( ; n_key_length--; ++p_s_le_u32, ++p_s_cpu_u32 ) {
  107     if ( le32_to_cpu (*p_s_le_u32) < *p_s_cpu_u32 )
  108       return -1;
  109     if ( le32_to_cpu (*p_s_le_u32) > *p_s_cpu_u32 )
  110       return 1;
  111   }
  112 
  113   return 0;
  114 }
  115 
  116 
  117 /* k1 is pointer to on-disk structure which is stored in little-endian
  118    form. k2 is pointer to cpu variable.
  119    Compare keys using all 4 key fields.
  120    Returns: -1 if key1 < key2 0
  121    if key1 = key2 1 if key1 > key2 */
  122 inline int  comp_keys (const struct key * le_key, const struct cpu_key * cpu_key)
  123 {
  124   int retval;
  125 
  126   retval = comp_short_keys (le_key, cpu_key);
  127   if (retval)
  128       return retval;
  129   if (le_key_k_offset (le_key_version(le_key), le_key) < cpu_key_k_offset (cpu_key))
  130       return -1;
  131   if (le_key_k_offset (le_key_version(le_key), le_key) > cpu_key_k_offset (cpu_key))
  132       return 1;
  133 
  134   if (cpu_key->key_length == 3)
  135       return 0;
  136 
  137   /* this part is needed only when tail conversion is in progress */
  138   if (le_key_k_type (le_key_version(le_key), le_key) < cpu_key_k_type (cpu_key))
  139     return -1;
  140 
  141   if (le_key_k_type (le_key_version(le_key), le_key) > cpu_key_k_type (cpu_key))
  142     return 1;
  143 
  144   return 0;
  145 }
  146 
  147 
  148 //
  149 // FIXME: not used yet
  150 //
  151 inline int comp_cpu_keys (const struct cpu_key * key1, 
  152                           const struct cpu_key * key2)
  153 {
  154     if (key1->on_disk_key.k_dir_id < key2->on_disk_key.k_dir_id)
  155         return -1;
  156     if (key1->on_disk_key.k_dir_id > key2->on_disk_key.k_dir_id)
  157         return 1;
  158 
  159     if (key1->on_disk_key.k_objectid < key2->on_disk_key.k_objectid)
  160         return -1;
  161     if (key1->on_disk_key.k_objectid > key2->on_disk_key.k_objectid)
  162         return 1;
  163 
  164     if (cpu_key_k_offset (key1) < cpu_key_k_offset (key2))
  165         return -1;
  166     if (cpu_key_k_offset (key1) > cpu_key_k_offset (key2))
  167         return 1;
  168 
  169     reiserfs_warning (NULL, "comp_cpu_keys: type are compared for %K and %K\n",
  170                       key1, key2);
  171 
  172     if (cpu_key_k_type (key1) < cpu_key_k_type (key2))
  173         return -1;
  174     if (cpu_key_k_type (key1) > cpu_key_k_type (key2))
  175         return 1;
  176     return 0;
  177 }
  178 
  179 inline int comp_short_le_keys (const struct key * key1, const struct key * key2)
  180 {
  181   __u32 * p_s_1_u32, * p_s_2_u32;
  182   int n_key_length = REISERFS_SHORT_KEY_LEN;
  183 
  184   p_s_1_u32 = (__u32 *)key1;
  185   p_s_2_u32 = (__u32 *)key2;
  186   for( ; n_key_length--; ++p_s_1_u32, ++p_s_2_u32 ) {
  187     if ( le32_to_cpu (*p_s_1_u32) < le32_to_cpu (*p_s_2_u32) )
  188       return -1;
  189     if ( le32_to_cpu (*p_s_1_u32) > le32_to_cpu (*p_s_2_u32) )
  190       return 1;
  191   }
  192   return 0;
  193 }
  194 
  195 inline int comp_short_cpu_keys (const struct cpu_key * key1, 
  196                                 const struct cpu_key * key2)
  197 {
  198   __u32 * p_s_1_u32, * p_s_2_u32;
  199   int n_key_length = REISERFS_SHORT_KEY_LEN;
  200 
  201   p_s_1_u32 = (__u32 *)key1;
  202   p_s_2_u32 = (__u32 *)key2;
  203 
  204   for( ; n_key_length--; ++p_s_1_u32, ++p_s_2_u32 ) {
  205     if ( *p_s_1_u32 < *p_s_2_u32 )
  206       return -1;
  207     if ( *p_s_1_u32 > *p_s_2_u32 )
  208       return 1;
  209   }
  210   return 0;
  211 }
  212 
  213 
  214 
  215 inline void cpu_key2cpu_key (struct cpu_key * to, const struct cpu_key * from)
  216 {
  217     memcpy (to, from, sizeof (struct cpu_key));
  218 }
  219 
  220 
  221 inline void le_key2cpu_key (struct cpu_key * to, const struct key * from)
  222 {
  223     to->on_disk_key.k_dir_id = le32_to_cpu (from->k_dir_id);
  224     to->on_disk_key.k_objectid = le32_to_cpu (from->k_objectid);
  225     
  226     // find out version of the key
  227     to->version = le_key_version (from);
  228     if (to->version == KEY_FORMAT_3_5) {
  229         to->on_disk_key.u.k_offset_v1.k_offset = le32_to_cpu (from->u.k_offset_v1.k_offset);
  230         to->on_disk_key.u.k_offset_v1.k_uniqueness = le32_to_cpu (from->u.k_offset_v1.k_uniqueness);
  231     } else {
  232         to->on_disk_key.u.k_offset_v2.k_offset = offset_v2_k_offset(&from->u.k_offset_v2);
  233         to->on_disk_key.u.k_offset_v2.k_type = offset_v2_k_type(&from->u.k_offset_v2);
  234     } 
  235 }
  236 
  237 
  238 
  239 // this does not say which one is bigger, it only returns 1 if keys
  240 // are not equal, 0 otherwise
  241 inline int comp_le_keys (const struct key * k1, const struct key * k2)
  242 {
  243     return memcmp (k1, k2, sizeof (struct key));
  244 }
  245 
  246 /**************************************************************************
  247  *  Binary search toolkit function                                        *
  248  *  Search for an item in the array by the item key                       *
  249  *  Returns:    1 if found,  0 if not found;                              *
  250  *        *p_n_pos = number of the searched element if found, else the    *
  251  *        number of the first element that is larger than p_v_key.        *
  252  **************************************************************************/
  253 /* For those not familiar with binary search: n_lbound is the leftmost item that it
  254  could be, n_rbound the rightmost item that it could be.  We examine the item
  255  halfway between n_lbound and n_rbound, and that tells us either that we can increase
  256  n_lbound, or decrease n_rbound, or that we have found it, or if n_lbound <= n_rbound that
  257  there are no possible items, and we have not found it. With each examination we
  258  cut the number of possible items it could be by one more than half rounded down,
  259  or we find it. */
  260 inline  int bin_search (
  261               const void * p_v_key, /* Key to search for.                   */
  262               const void * p_v_base,/* First item in the array.             */
  263               int       p_n_num,    /* Number of items in the array.        */
  264               int       p_n_width,  /* Item size in the array.
  265                                        searched. Lest the reader be
  266                                        confused, note that this is crafted
  267                                        as a general function, and when it
  268                                        is applied specifically to the array
  269                                        of item headers in a node, p_n_width
  270                                        is actually the item header size not
  271                                        the item size.                      */
  272               int     * p_n_pos     /* Number of the searched for element. */
  273             ) {
  274     int   n_rbound, n_lbound, n_j;
  275 
  276    for ( n_j = ((n_rbound = p_n_num - 1) + (n_lbound = 0))/2; n_lbound <= n_rbound; n_j = (n_rbound + n_lbound)/2 )
  277      switch( COMP_KEYS((struct key *)((char * )p_v_base + n_j * p_n_width), (struct cpu_key *)p_v_key) )  {
  278      case -1: n_lbound = n_j + 1; continue;
  279      case  1: n_rbound = n_j - 1; continue;
  280      case  0: *p_n_pos = n_j;     return ITEM_FOUND; /* Key found in the array.  */
  281         }
  282 
  283     /* bin_search did not find given key, it returns position of key,
  284         that is minimal and greater than the given one. */
  285     *p_n_pos = n_lbound;
  286     return ITEM_NOT_FOUND;
  287 }
  288 
  289 #ifdef CONFIG_REISERFS_CHECK
  290 extern struct tree_balance * cur_tb;
  291 #endif
  292 
  293 
  294 
  295 /* Minimal possible key. It is never in the tree. */
  296 const struct key  MIN_KEY = {0, 0, {{0, 0},}};
  297 
  298 /* Maximal possible key. It is never in the tree. */
  299 const struct key  MAX_KEY = {0xffffffff, 0xffffffff, {{0xffffffff, 0xffffffff},}};
  300 
  301 
  302 /* Get delimiting key of the buffer by looking for it in the buffers in the path, starting from the bottom
  303    of the path, and going upwards.  We must check the path's validity at each step.  If the key is not in
  304    the path, there is no delimiting key in the tree (buffer is first or last buffer in tree), and in this
  305    case we return a special key, either MIN_KEY or MAX_KEY. */
  306 inline  const struct  key * get_lkey  (
  307                         const struct path         * p_s_chk_path,
  308                         const struct super_block  * p_s_sb
  309                       ) {
  310   int                   n_position, n_path_offset = p_s_chk_path->path_length;
  311   struct buffer_head  * p_s_parent;
  312   
  313   RFALSE( n_path_offset < FIRST_PATH_ELEMENT_OFFSET, 
  314           "PAP-5010: illegal offset in the path");
  315 
  316   /* While not higher in path than first element. */
  317   while ( n_path_offset-- > FIRST_PATH_ELEMENT_OFFSET ) {
  318 
  319     RFALSE( ! buffer_uptodate(PATH_OFFSET_PBUFFER(p_s_chk_path, n_path_offset)),
  320             "PAP-5020: parent is not uptodate");
  321 
  322     /* Parent at the path is not in the tree now. */
  323     if ( ! B_IS_IN_TREE(p_s_parent = PATH_OFFSET_PBUFFER(p_s_chk_path, n_path_offset)) )
  324       return &MAX_KEY;
  325     /* Check whether position in the parent is correct. */
  326     if ( (n_position = PATH_OFFSET_POSITION(p_s_chk_path, n_path_offset)) > B_NR_ITEMS(p_s_parent) )
  327        return &MAX_KEY;
  328     /* Check whether parent at the path really points to the child. */
  329     if ( B_N_CHILD_NUM(p_s_parent, n_position) !=
  330          PATH_OFFSET_PBUFFER(p_s_chk_path, n_path_offset + 1)->b_blocknr )
  331       return &MAX_KEY;
  332     /* Return delimiting key if position in the parent is not equal to zero. */
  333     if ( n_position )
  334       return B_N_PDELIM_KEY(p_s_parent, n_position - 1);
  335   }
  336   /* Return MIN_KEY if we are in the root of the buffer tree. */
  337   if ( PATH_OFFSET_PBUFFER(p_s_chk_path, FIRST_PATH_ELEMENT_OFFSET)->b_blocknr ==
  338        SB_ROOT_BLOCK (p_s_sb) )
  339     return &MIN_KEY;
  340   return  &MAX_KEY;
  341 }
  342 
  343 
  344 /* Get delimiting key of the buffer at the path and its right neighbor. */
  345 inline  const struct  key * get_rkey  (
  346                         const struct path         * p_s_chk_path,
  347                         const struct super_block  * p_s_sb
  348                       ) {
  349   int                   n_position,
  350                         n_path_offset = p_s_chk_path->path_length;
  351   struct buffer_head  * p_s_parent;
  352 
  353   RFALSE( n_path_offset < FIRST_PATH_ELEMENT_OFFSET,
  354           "PAP-5030: illegal offset in the path");
  355 
  356   while ( n_path_offset-- > FIRST_PATH_ELEMENT_OFFSET ) {
  357 
  358     RFALSE( ! buffer_uptodate(PATH_OFFSET_PBUFFER(p_s_chk_path, n_path_offset)),
  359             "PAP-5040: parent is not uptodate");
  360 
  361     /* Parent at the path is not in the tree now. */
  362     if ( ! B_IS_IN_TREE(p_s_parent = PATH_OFFSET_PBUFFER(p_s_chk_path, n_path_offset)) )
  363       return &MIN_KEY;
  364     /* Check whether position in the parent is correct. */
  365     if ( (n_position = PATH_OFFSET_POSITION(p_s_chk_path, n_path_offset)) > B_NR_ITEMS(p_s_parent) )
  366       return &MIN_KEY;
  367     /* Check whether parent at the path really points to the child. */
  368     if ( B_N_CHILD_NUM(p_s_parent, n_position) !=
  369                                         PATH_OFFSET_PBUFFER(p_s_chk_path, n_path_offset + 1)->b_blocknr )
  370       return &MIN_KEY;
  371     /* Return delimiting key if position in the parent is not the last one. */
  372     if ( n_position != B_NR_ITEMS(p_s_parent) )
  373       return B_N_PDELIM_KEY(p_s_parent, n_position);
  374   }
  375   /* Return MAX_KEY if we are in the root of the buffer tree. */
  376   if ( PATH_OFFSET_PBUFFER(p_s_chk_path, FIRST_PATH_ELEMENT_OFFSET)->b_blocknr ==
  377        SB_ROOT_BLOCK (p_s_sb) )
  378     return &MAX_KEY;
  379   return  &MIN_KEY;
  380 }
  381 
  382 
  383 /* Check whether a key is contained in the tree rooted from a buffer at a path. */
  384 /* This works by looking at the left and right delimiting keys for the buffer in the last path_element in
  385    the path.  These delimiting keys are stored at least one level above that buffer in the tree. If the
  386    buffer is the first or last node in the tree order then one of the delimiting keys may be absent, and in
  387    this case get_lkey and get_rkey return a special key which is MIN_KEY or MAX_KEY. */
  388 static  inline  int key_in_buffer (
  389                       struct path         * p_s_chk_path, /* Path which should be checked.  */
  390                       const struct cpu_key      * p_s_key,      /* Key which should be checked.   */
  391                       struct super_block  * p_s_sb        /* Super block pointer.           */
  392                       ) {
  393 
  394   RFALSE( ! p_s_key || p_s_chk_path->path_length < FIRST_PATH_ELEMENT_OFFSET ||
  395           p_s_chk_path->path_length > MAX_HEIGHT,
  396           "PAP-5050: pointer to the key(%p) is NULL or illegal path length(%d)",
  397           p_s_key, p_s_chk_path->path_length);
  398   RFALSE( PATH_PLAST_BUFFER(p_s_chk_path)->b_dev == NODEV,
  399           "PAP-5060: device must not be NODEV");
  400 
  401   if ( COMP_KEYS(get_lkey(p_s_chk_path, p_s_sb), p_s_key) == 1 )
  402     /* left delimiting key is bigger, that the key we look for */
  403     return 0;
  404   //  if ( COMP_KEYS(p_s_key, get_rkey(p_s_chk_path, p_s_sb)) != -1 )
  405   if ( COMP_KEYS(get_rkey(p_s_chk_path, p_s_sb), p_s_key) != 1 )
  406     /* p_s_key must be less than right delimitiing key */
  407     return 0;
  408   return 1;
  409 }
  410 
  411 
  412 inline void decrement_bcount(
  413               struct buffer_head  * p_s_bh
  414             ) { 
  415   if ( p_s_bh ) {
  416     if ( atomic_read (&(p_s_bh->b_count)) ) {
  417       put_bh(p_s_bh) ;
  418       return;
  419     }
  420     reiserfs_panic(NULL, "PAP-5070: decrement_bcount: trying to free free buffer %b", p_s_bh);
  421   }
  422 }
  423 
  424 
  425 /* Decrement b_count field of the all buffers in the path. */
  426 void decrement_counters_in_path (
  427               struct path * p_s_search_path
  428             ) {
  429   int n_path_offset = p_s_search_path->path_length;
  430 
  431   RFALSE( n_path_offset < ILLEGAL_PATH_ELEMENT_OFFSET ||
  432           n_path_offset > EXTENDED_MAX_HEIGHT - 1,
  433           "PAP-5080: illegal path offset of %d", n_path_offset);
  434 
  435   while ( n_path_offset > ILLEGAL_PATH_ELEMENT_OFFSET ) {
  436     struct buffer_head * bh;
  437 
  438     bh = PATH_OFFSET_PBUFFER(p_s_search_path, n_path_offset--);
  439     decrement_bcount (bh);
  440   }
  441   p_s_search_path->path_length = ILLEGAL_PATH_ELEMENT_OFFSET;
  442 }
  443 
  444 
  445 int reiserfs_check_path(struct path *p) {
  446   RFALSE( p->path_length != ILLEGAL_PATH_ELEMENT_OFFSET,
  447           "path not properly relsed") ;
  448   return 0 ;
  449 }
  450 
  451 
  452 /* Release all buffers in the path. Restore dirty bits clean
  453 ** when preparing the buffer for the log
  454 **
  455 ** only called from fix_nodes()
  456 */
  457 void  pathrelse_and_restore (
  458         struct super_block *s, 
  459         struct path * p_s_search_path
  460       ) {
  461   int n_path_offset = p_s_search_path->path_length;
  462 
  463   RFALSE( n_path_offset < ILLEGAL_PATH_ELEMENT_OFFSET, 
  464           "clm-4000: illegal path offset");
  465   
  466   while ( n_path_offset > ILLEGAL_PATH_ELEMENT_OFFSET )  {
  467     reiserfs_restore_prepared_buffer(s, PATH_OFFSET_PBUFFER(p_s_search_path, 
  468                                      n_path_offset));
  469     brelse(PATH_OFFSET_PBUFFER(p_s_search_path, n_path_offset--));
  470   }
  471   p_s_search_path->path_length = ILLEGAL_PATH_ELEMENT_OFFSET;
  472 }
  473 
  474 /* Release all buffers in the path. */
  475 void  pathrelse (
  476         struct path * p_s_search_path
  477       ) {
  478   int n_path_offset = p_s_search_path->path_length;
  479 
  480   RFALSE( n_path_offset < ILLEGAL_PATH_ELEMENT_OFFSET,
  481           "PAP-5090: illegal path offset");
  482   
  483   while ( n_path_offset > ILLEGAL_PATH_ELEMENT_OFFSET )  
  484     brelse(PATH_OFFSET_PBUFFER(p_s_search_path, n_path_offset--));
  485 
  486   p_s_search_path->path_length = ILLEGAL_PATH_ELEMENT_OFFSET;
  487 }
  488 
  489 
  490 
  491 static int is_leaf (char * buf, int blocksize, struct buffer_head * bh)
  492 {
  493     struct block_head * blkh;
  494     struct item_head * ih;
  495     int used_space;
  496     int prev_location;
  497     int i;
  498     int nr;
  499 
  500     blkh = (struct block_head *)buf;
  501     if ( blkh_level(blkh) != DISK_LEAF_NODE_LEVEL) {
  502         printk ("is_leaf: this should be caught earlier\n");
  503         return 0;
  504     }
  505 
  506     nr = blkh_nr_item(blkh);
  507     if (nr < 1 || nr > ((blocksize - BLKH_SIZE) / (IH_SIZE + MIN_ITEM_LEN))) {
  508         /* item number is too big or too small */
  509         reiserfs_warning (NULL, "is_leaf: nr_item seems wrong: %z\n", bh);
  510         return 0;
  511     }
  512     ih = (struct item_head *)(buf + BLKH_SIZE) + nr - 1;
  513     used_space = BLKH_SIZE + IH_SIZE * nr + (blocksize - ih_location (ih));
  514     if (used_space != blocksize - blkh_free_space(blkh)) {
  515         /* free space does not match to calculated amount of use space */
  516         reiserfs_warning (NULL, "is_leaf: free space seems wrong: %z\n", bh);
  517         return 0;
  518     }
  519 
  520     // FIXME: it is_leaf will hit performance too much - we may have
  521     // return 1 here
  522 
  523     /* check tables of item heads */
  524     ih = (struct item_head *)(buf + BLKH_SIZE);
  525     prev_location = blocksize;
  526     for (i = 0; i < nr; i ++, ih ++) {
  527         if ( le_ih_k_type(ih) == TYPE_ANY) {
  528             reiserfs_warning (NULL, "is_leaf: wrong item type for item %h\n",ih);
  529             return 0;
  530         }
  531         if (ih_location (ih) >= blocksize || ih_location (ih) < IH_SIZE * nr) {
  532             reiserfs_warning (NULL, "is_leaf: item location seems wrong: %h\n", ih);
  533             return 0;
  534         }
  535         if (ih_item_len (ih) < 1 || ih_item_len (ih) > MAX_ITEM_LEN (blocksize)) {
  536             reiserfs_warning (NULL, "is_leaf: item length seems wrong: %h\n", ih);
  537             return 0;
  538         }
  539         if (prev_location - ih_location (ih) != ih_item_len (ih)) {
  540             reiserfs_warning (NULL, "is_leaf: item location seems wrong (second one): %h\n", ih);
  541             return 0;
  542         }
  543         prev_location = ih_location (ih);
  544     }
  545 
  546     // one may imagine much more checks
  547     return 1;
  548 }
  549 
  550 
  551 /* returns 1 if buf looks like an internal node, 0 otherwise */
  552 static int is_internal (char * buf, int blocksize, struct buffer_head * bh)
  553 {
  554     struct block_head * blkh;
  555     int nr;
  556     int used_space;
  557 
  558     blkh = (struct block_head *)buf;
  559     nr = blkh_level(blkh);
  560     if (nr <= DISK_LEAF_NODE_LEVEL || nr > MAX_HEIGHT) {
  561         /* this level is not possible for internal nodes */
  562         printk ("is_internal: this should be caught earlier\n");
  563         return 0;
  564     }
  565     
  566     nr = blkh_nr_item(blkh);
  567     if (nr > (blocksize - BLKH_SIZE - DC_SIZE) / (KEY_SIZE + DC_SIZE)) {
  568         /* for internal which is not root we might check min number of keys */
  569         reiserfs_warning (NULL, "is_internal: number of key seems wrong: %z\n", bh);
  570         return 0;
  571     }
  572 
  573     used_space = BLKH_SIZE + KEY_SIZE * nr + DC_SIZE * (nr + 1);
  574     if (used_space != blocksize - blkh_free_space(blkh)) {
  575         reiserfs_warning (NULL, "is_internal: free space seems wrong: %z\n", bh);
  576         return 0;
  577     }
  578 
  579     // one may imagine much more checks
  580     return 1;
  581 }
  582 
  583 
  584 // make sure that bh contains formatted node of reiserfs tree of
  585 // 'level'-th level
  586 static int is_tree_node (struct buffer_head * bh, int level)
  587 {
  588     if (B_LEVEL (bh) != level) {
  589         printk ("is_tree_node: node level %d does not match to the expected one %d\n",
  590                 B_LEVEL (bh), level);
  591         return 0;
  592     }
  593     if (level == DISK_LEAF_NODE_LEVEL)
  594         return is_leaf (bh->b_data, bh->b_size, bh);
  595 
  596     return is_internal (bh->b_data, bh->b_size, bh);
  597 }
  598 
  599 
  600 
  601 #ifdef SEARCH_BY_KEY_READA
  602 
  603 /* The function is NOT SCHEDULE-SAFE! */
  604 static void search_by_key_reada (struct super_block * s, int blocknr)
  605 {
  606     struct buffer_head * bh;
  607   
  608     if (blocknr == 0)
  609         return;
  610 
  611     bh = getblk (s->s_dev, blocknr, s->s_blocksize);
  612   
  613     if (!buffer_uptodate (bh)) {
  614         ll_rw_block (READA, 1, &bh);
  615     }
  616     bh->b_count --;
  617 }
  618 
  619 #endif
  620 
  621 /**************************************************************************
  622  * Algorithm   SearchByKey                                                *
  623  *             look for item in the Disk S+Tree by its key                *
  624  * Input:  p_s_sb   -  super block                                        *
  625  *         p_s_key  - pointer to the key to search                        *
  626  * Output: ITEM_FOUND, ITEM_NOT_FOUND or IO_ERROR                         *
  627  *         p_s_search_path - path from the root to the needed leaf        *
  628  **************************************************************************/
  629 
  630 /* This function fills up the path from the root to the leaf as it
  631    descends the tree looking for the key.  It uses reiserfs_bread to
  632    try to find buffers in the cache given their block number.  If it
  633    does not find them in the cache it reads them from disk.  For each
  634    node search_by_key finds using reiserfs_bread it then uses
  635    bin_search to look through that node.  bin_search will find the
  636    position of the block_number of the next node if it is looking
  637    through an internal node.  If it is looking through a leaf node
  638    bin_search will find the position of the item which has key either
  639    equal to given key, or which is the maximal key less than the given
  640    key.  search_by_key returns a path that must be checked for the
  641    correctness of the top of the path but need not be checked for the
  642    correctness of the bottom of the path */
  643 /* The function is NOT SCHEDULE-SAFE! */
  644 int search_by_key (struct super_block * p_s_sb,
  645                    const struct cpu_key * p_s_key, /* Key to search. */
  646                    struct path * p_s_search_path, /* This structure was
  647                                                      allocated and initialized
  648                                                      by the calling
  649                                                      function. It is filled up
  650                                                      by this function.  */
  651                    int n_stop_level /* How far down the tree to search. To
  652                                        stop at leaf level - set to
  653                                        DISK_LEAF_NODE_LEVEL */
  654     ) {
  655     int  n_block_number = SB_ROOT_BLOCK (p_s_sb),
  656       expected_level = SB_TREE_HEIGHT (p_s_sb),
  657       n_block_size    = p_s_sb->s_blocksize;
  658     struct buffer_head  *       p_s_bh;
  659     struct path_element *       p_s_last_element;
  660     int                         n_node_level, n_retval;
  661     int                         right_neighbor_of_leaf_node;
  662     int                         fs_gen;
  663 
  664 #ifdef CONFIG_REISERFS_CHECK
  665     int n_repeat_counter = 0;
  666 #endif
  667     
  668     PROC_INFO_INC( p_s_sb, search_by_key );
  669     
  670     /* As we add each node to a path we increase its count.  This means that
  671        we must be careful to release all nodes in a path before we either
  672        discard the path struct or re-use the path struct, as we do here. */
  673 
  674     decrement_counters_in_path(p_s_search_path);
  675 
  676     right_neighbor_of_leaf_node = 0;
  677 
  678     /* With each iteration of this loop we search through the items in the
  679        current node, and calculate the next current node(next path element)
  680        for the next iteration of this loop.. */
  681     while ( 1 ) {
  682 
  683 #ifdef CONFIG_REISERFS_CHECK
  684         if ( !(++n_repeat_counter % 50000) )
  685             reiserfs_warning (p_s_sb, "PAP-5100: search_by_key: %s:"
  686                               "there were %d iterations of while loop "
  687                               "looking for key %K\n",
  688                               current->comm, n_repeat_counter, p_s_key);
  689 #endif
  690 
  691         /* prep path to have another element added to it. */
  692         p_s_last_element = PATH_OFFSET_PELEMENT(p_s_search_path, ++p_s_search_path->path_length);
  693         fs_gen = get_generation (p_s_sb);
  694         expected_level --;
  695 
  696 #ifdef SEARCH_BY_KEY_READA
  697         /* schedule read of right neighbor */
  698         search_by_key_reada (p_s_sb, right_neighbor_of_leaf_node);
  699 #endif
  700 
  701         /* Read the next tree node, and set the last element in the path to
  702            have a pointer to it. */
  703         if ( ! (p_s_bh = p_s_last_element->pe_buffer =
  704                 reiserfs_bread(p_s_sb, n_block_number, n_block_size)) ) {
  705             p_s_search_path->path_length --;
  706             pathrelse(p_s_search_path);
  707             return IO_ERROR;
  708         }
  709 
  710         if( fs_changed (fs_gen, p_s_sb) ) {
  711                 PROC_INFO_INC( p_s_sb, search_by_key_fs_changed );
  712                 PROC_INFO_INC( p_s_sb, sbk_fs_changed[ expected_level - 1 ] );
  713         }
  714 
  715         /* It is possible that schedule occurred. We must check whether the key
  716            to search is still in the tree rooted from the current buffer. If
  717            not then repeat search from the root. */
  718         if ( fs_changed (fs_gen, p_s_sb) && 
  719              (!B_IS_IN_TREE (p_s_bh) || !key_in_buffer(p_s_search_path, p_s_key, p_s_sb)) ) {
  720             PROC_INFO_INC( p_s_sb, search_by_key_restarted );
  721             PROC_INFO_INC( p_s_sb, sbk_restarted[ expected_level - 1 ] );
  722             decrement_counters_in_path(p_s_search_path);
  723             
  724             /* Get the root block number so that we can repeat the search
  725                starting from the root. */
  726             n_block_number = SB_ROOT_BLOCK (p_s_sb);
  727             expected_level = SB_TREE_HEIGHT (p_s_sb);
  728             right_neighbor_of_leaf_node = 0;
  729             
  730             /* repeat search from the root */
  731             continue;
  732         }
  733 
  734         /* only check that the key is in the buffer if p_s_key is not
  735            equal to the MAX_KEY. Latter case is only possible in
  736            "finish_unfinished()" processing during mount. */
  737         RFALSE( COMP_KEYS( &MAX_KEY, p_s_key ) && 
  738                 ! key_in_buffer(p_s_search_path, p_s_key, p_s_sb),
  739                 "PAP-5130: key is not in the buffer");
  740 #ifdef CONFIG_REISERFS_CHECK
  741         if ( cur_tb ) {
  742             print_cur_tb ("5140");
  743             reiserfs_panic(p_s_sb, "PAP-5140: search_by_key: schedule occurred in do_balance!");
  744         }
  745 #endif
  746 
  747         // make sure, that the node contents look like a node of
  748         // certain level
  749         if (!is_tree_node (p_s_bh, expected_level)) {
  750             reiserfs_warning (p_s_sb, "vs-5150: search_by_key: "
  751                               "invalid format found in block %ld. Fsck?\n", 
  752                               p_s_bh->b_blocknr);
  753             pathrelse (p_s_search_path);
  754             return IO_ERROR;
  755         }
  756         
  757         /* ok, we have acquired next formatted node in the tree */
  758         n_node_level = B_LEVEL (p_s_bh);
  759 
  760         PROC_INFO_BH_STAT( p_s_sb, p_s_bh, n_node_level - 1 );
  761 
  762         RFALSE( n_node_level < n_stop_level,
  763                 "vs-5152: tree level (%d) is less than stop level (%d)",
  764                 n_node_level, n_stop_level);
  765 
  766         n_retval = bin_search( p_s_key, B_N_PITEM_HEAD(p_s_bh, 0),
  767                 B_NR_ITEMS(p_s_bh),
  768                 ( n_node_level == DISK_LEAF_NODE_LEVEL ) ? IH_SIZE : KEY_SIZE,
  769                 &(p_s_last_element->pe_position));
  770         if (n_node_level == n_stop_level) {
  771             return n_retval;
  772         }
  773 
  774         /* we are not in the stop level */
  775         if (n_retval == ITEM_FOUND)
  776             /* item has been found, so we choose the pointer which is to the right of the found one */
  777             p_s_last_element->pe_position++;
  778 
  779         /* if item was not found we choose the position which is to
  780            the left of the found item. This requires no code,
  781            bin_search did it already.*/
  782 
  783         /* So we have chosen a position in the current node which is
  784            an internal node.  Now we calculate child block number by
  785            position in the node. */
  786         n_block_number = B_N_CHILD_NUM(p_s_bh, p_s_last_element->pe_position);
  787 
  788 #ifdef SEARCH_BY_KEY_READA
  789         /* if we are going to read leaf node, then calculate its right neighbor if possible */
  790         if (n_node_level == DISK_LEAF_NODE_LEVEL + 1 && p_s_last_element->pe_position < B_NR_ITEMS (p_s_bh))
  791             right_neighbor_of_leaf_node = B_N_CHILD_NUM(p_s_bh, p_s_last_element->pe_position + 1);
  792 #endif
  793     }
  794 }
  795 
  796 
  797 /* Form the path to an item and position in this item which contains
  798    file byte defined by p_s_key. If there is no such item
  799    corresponding to the key, we point the path to the item with
  800    maximal key less than p_s_key, and *p_n_pos_in_item is set to one
  801    past the last entry/byte in the item.  If searching for entry in a
  802    directory item, and it is not found, *p_n_pos_in_item is set to one
  803    entry more than the entry with maximal key which is less than the
  804    sought key.
  805 
  806    Note that if there is no entry in this same node which is one more,
  807    then we point to an imaginary entry.  for direct items, the
  808    position is in units of bytes, for indirect items the position is
  809    in units of blocknr entries, for directory items the position is in
  810    units of directory entries.  */
  811 
  812 /* The function is NOT SCHEDULE-SAFE! */
  813 int search_for_position_by_key (struct super_block  * p_s_sb,         /* Pointer to the super block.          */
  814                                 const struct cpu_key  * p_cpu_key,      /* Key to search (cpu variable)         */
  815                                 struct path         * p_s_search_path /* Filled up by this function.          */
  816     ) {
  817     struct item_head    * p_le_ih; /* pointer to on-disk structure */
  818     int                   n_blk_size;
  819     loff_t item_offset, offset;
  820     struct reiserfs_dir_entry de;
  821     int retval;
  822 
  823     /* If searching for directory entry. */
  824     if ( is_direntry_cpu_key (p_cpu_key) )
  825         return  search_by_entry_key (p_s_sb, p_cpu_key, p_s_search_path, &de);
  826 
  827     /* If not searching for directory entry. */
  828     
  829     /* If item is found. */
  830     retval = search_item (p_s_sb, p_cpu_key, p_s_search_path);
  831     if (retval == IO_ERROR)
  832         return retval;
  833     if ( retval == ITEM_FOUND )  {
  834 
  835         RFALSE( ! ih_item_len(
  836                 B_N_PITEM_HEAD(PATH_PLAST_BUFFER(p_s_search_path),
  837                                PATH_LAST_POSITION(p_s_search_path))),
  838                 "PAP-5165: item length equals zero");
  839 
  840         pos_in_item(p_s_search_path) = 0;
  841         return POSITION_FOUND;
  842     }
  843 
  844     RFALSE( ! PATH_LAST_POSITION(p_s_search_path),
  845             "PAP-5170: position equals zero");
  846 
  847     /* Item is not found. Set path to the previous item. */
  848     p_le_ih = B_N_PITEM_HEAD(PATH_PLAST_BUFFER(p_s_search_path), --PATH_LAST_POSITION(p_s_search_path));
  849     n_blk_size = p_s_sb->s_blocksize;
  850 
  851     if (comp_short_keys (&(p_le_ih->ih_key), p_cpu_key)) {
  852         return FILE_NOT_FOUND;
  853     }
  854 
  855     // FIXME: quite ugly this far
  856 
  857     item_offset = le_ih_k_offset (p_le_ih);
  858     offset = cpu_key_k_offset (p_cpu_key);
  859 
  860     /* Needed byte is contained in the item pointed to by the path.*/
  861     if (item_offset <= offset &&
  862         item_offset + op_bytes_number (p_le_ih, n_blk_size) > offset) {
  863         pos_in_item (p_s_search_path) = offset - item_offset;
  864         if ( is_indirect_le_ih(p_le_ih) ) {
  865             pos_in_item (p_s_search_path) /= n_blk_size;
  866         }
  867         return POSITION_FOUND;
  868     }
  869 
  870     /* Needed byte is not contained in the item pointed to by the
  871      path. Set pos_in_item out of the item. */
  872     if ( is_indirect_le_ih (p_le_ih) )
  873         pos_in_item (p_s_search_path) = ih_item_len(p_le_ih) / UNFM_P_SIZE;
  874     else
  875         pos_in_item (p_s_search_path) = ih_item_len( p_le_ih );
  876   
  877     return POSITION_NOT_FOUND;
  878 }
  879 
  880 
  881 /* Compare given item and item pointed to by the path. */
  882 int comp_items (const struct item_head * stored_ih, const struct path * p_s_path)
  883 {
  884     struct buffer_head  * p_s_bh;
  885     struct item_head    * ih;
  886 
  887     /* Last buffer at the path is not in the tree. */
  888     if ( ! B_IS_IN_TREE(p_s_bh = PATH_PLAST_BUFFER(p_s_path)) )
  889         return 1;
  890 
  891     /* Last path position is invalid. */
  892     if ( PATH_LAST_POSITION(p_s_path) >= B_NR_ITEMS(p_s_bh) )
  893         return 1;
  894 
  895     /* we need only to know, whether it is the same item */
  896     ih = get_ih (p_s_path);
  897     return memcmp (stored_ih, ih, IH_SIZE);
  898 }
  899 
  900 
  901 /* unformatted nodes are not logged anymore, ever.  This is safe
  902 ** now
  903 */
  904 #define held_by_others(bh) (atomic_read(&(bh)->b_count) > 1)
  905 
  906 // block can not be forgotten as it is in I/O or held by someone
  907 #define block_in_use(bh) (buffer_locked(bh) || (held_by_others(bh)))
  908 
  909 
  910 
  911 // prepare for delete or cut of direct item
  912 static inline int prepare_for_direct_item (struct path * path,
  913                                            struct item_head * le_ih,
  914                                            struct inode * inode,
  915                                            loff_t new_file_length,
  916                                            int * cut_size)
  917 {
  918     loff_t round_len;
  919 
  920 
  921     if ( new_file_length == max_reiserfs_offset (inode) ) {
  922         /* item has to be deleted */
  923         *cut_size = -(IH_SIZE + ih_item_len(le_ih));
  924         return M_DELETE;
  925     }
  926         
  927     // new file gets truncated
  928     if (get_inode_item_key_version (inode) == KEY_FORMAT_3_6) {
  929         // 
  930         round_len = ROUND_UP (new_file_length); 
  931         /* this was n_new_file_length < le_ih ... */
  932         if ( round_len < le_ih_k_offset (le_ih) )  {
  933             *cut_size = -(IH_SIZE + ih_item_len(le_ih));
  934             return M_DELETE; /* Delete this item. */
  935         }
  936         /* Calculate first position and size for cutting from item. */
  937         pos_in_item (path) = round_len - (le_ih_k_offset (le_ih) - 1);
  938         *cut_size = -(ih_item_len(le_ih) - pos_in_item(path));
  939         
  940         return M_CUT; /* Cut from this item. */
  941     }
  942 
  943 
  944     // old file: items may have any length
  945 
  946     if ( new_file_length < le_ih_k_offset (le_ih) )  {
  947         *cut_size = -(IH_SIZE + ih_item_len(le_ih));
  948         return M_DELETE; /* Delete this item. */
  949     }
  950     /* Calculate first position and size for cutting from item. */
  951     *cut_size = -(ih_item_len(le_ih) -
  952                       (pos_in_item (path) = new_file_length + 1 - le_ih_k_offset (le_ih)));
  953     return M_CUT; /* Cut from this item. */
  954 }
  955 
  956 
  957 static inline int prepare_for_direntry_item (struct path * path,
  958                                              struct item_head * le_ih,
  959                                              struct inode * inode,
  960                                              loff_t new_file_length,
  961                                              int * cut_size)
  962 {
  963     if (le_ih_k_offset (le_ih) == DOT_OFFSET && 
  964         new_file_length == max_reiserfs_offset (inode)) {
  965         RFALSE( ih_entry_count (le_ih) != 2,
  966                 "PAP-5220: incorrect empty directory item (%h)", le_ih);
  967         *cut_size = -(IH_SIZE + ih_item_len(le_ih));
  968         return M_DELETE; /* Delete the directory item containing "." and ".." entry. */
  969     }
  970     
  971     if ( ih_entry_count (le_ih) == 1 )  {
  972         /* Delete the directory item such as there is one record only
  973            in this item*/
  974         *cut_size = -(IH_SIZE + ih_item_len(le_ih));
  975         return M_DELETE;
  976     }
  977     
  978     /* Cut one record from the directory item. */
  979     *cut_size = -(DEH_SIZE + entry_length (get_last_bh (path), le_ih, pos_in_item (path)));
  980     return M_CUT; 
  981 }
  982 
  983 
  984 /*  If the path points to a directory or direct item, calculate mode and the size cut, for balance.
  985     If the path points to an indirect item, remove some number of its unformatted nodes.
  986     In case of file truncate calculate whether this item must be deleted/truncated or last
  987     unformatted node of this item will be converted to a direct item.
  988     This function returns a determination of what balance mode the calling function should employ. */
  989 static char  prepare_for_delete_or_cut(
  990                                        struct reiserfs_transaction_handle *th, 
  991                                        struct inode * inode,
  992                                        struct path         * p_s_path,
  993                                        const struct cpu_key      * p_s_item_key,
  994                                        int                 * p_n_removed,      /* Number of unformatted nodes which were removed
  995                                                                                   from end of the file. */
  996                                        int                 * p_n_cut_size,
  997                                        unsigned long long    n_new_file_length /* MAX_KEY_OFFSET in case of delete. */
  998     ) {
  999     struct super_block  * p_s_sb = inode->i_sb;
 1000     struct item_head    * p_le_ih = PATH_PITEM_HEAD(p_s_path);
 1001     struct buffer_head  * p_s_bh = PATH_PLAST_BUFFER(p_s_path);
 1002 
 1003     /* Stat_data item. */
 1004     if ( is_statdata_le_ih (p_le_ih) ) {
 1005 
 1006         RFALSE( n_new_file_length != max_reiserfs_offset (inode),
 1007                 "PAP-5210: mode must be M_DELETE");
 1008 
 1009         *p_n_cut_size = -(IH_SIZE + ih_item_len(p_le_ih));
 1010         return M_DELETE;
 1011     }
 1012 
 1013 
 1014     /* Directory item. */
 1015     if ( is_direntry_le_ih (p_le_ih) )
 1016         return prepare_for_direntry_item (p_s_path, p_le_ih, inode, n_new_file_length, p_n_cut_size);
 1017 
 1018     /* Direct item. */
 1019     if ( is_direct_le_ih (p_le_ih) )
 1020         return prepare_for_direct_item (p_s_path, p_le_ih, inode, n_new_file_length, p_n_cut_size);
 1021 
 1022 
 1023     /* Case of an indirect item. */
 1024     {
 1025         int                   n_unfm_number,    /* Number of the item unformatted nodes. */
 1026             n_counter,
 1027             n_blk_size;
 1028         __u32               * p_n_unfm_pointer; /* Pointer to the unformatted node number. */
 1029         __u32 tmp;
 1030         struct item_head      s_ih;           /* Item header. */
 1031         char                  c_mode;           /* Returned mode of the balance. */
 1032         int need_research;
 1033 
 1034 
 1035         n_blk_size = p_s_sb->s_blocksize;
 1036 
 1037         /* Search for the needed object indirect item until there are no unformatted nodes to be removed. */
 1038         do  {
 1039             need_research = 0;
 1040             p_s_bh = PATH_PLAST_BUFFER(p_s_path);
 1041             /* Copy indirect item header to a temp variable. */
 1042             copy_item_head(&s_ih, PATH_PITEM_HEAD(p_s_path));
 1043             /* Calculate number of unformatted nodes in this item. */
 1044             n_unfm_number = I_UNFM_NUM(&s_ih);
 1045 
 1046             RFALSE( ! is_indirect_le_ih(&s_ih) || ! n_unfm_number ||
 1047                     pos_in_item (p_s_path) + 1 !=  n_unfm_number,
 1048                     "PAP-5240: illegal item %h "
 1049                     "n_unfm_number = %d *p_n_pos_in_item = %d", 
 1050                     &s_ih, n_unfm_number, pos_in_item (p_s_path));
 1051 
 1052             /* Calculate balance mode and position in the item to remove unformatted nodes. */
 1053             if ( n_new_file_length == max_reiserfs_offset (inode) ) {/* Case of delete. */
 1054                 pos_in_item (p_s_path) = 0;
 1055                 *p_n_cut_size = -(IH_SIZE + ih_item_len(&s_ih));
 1056                 c_mode = M_DELETE;
 1057             }
 1058             else  { /* Case of truncate. */
 1059                 if ( n_new_file_length < le_ih_k_offset (&s_ih) )  {
 1060                     pos_in_item (p_s_path) = 0;
 1061                     *p_n_cut_size = -(IH_SIZE + ih_item_len(&s_ih));
 1062                     c_mode = M_DELETE; /* Delete this item. */
 1063                 }
 1064                 else  {
 1065                     /* indirect item must be truncated starting from *p_n_pos_in_item-th position */
 1066                     pos_in_item (p_s_path) = (n_new_file_length + n_blk_size - le_ih_k_offset (&s_ih) ) >> p_s_sb->s_blocksize_bits;
 1067 
 1068                     RFALSE( pos_in_item (p_s_path) > n_unfm_number,
 1069                             "PAP-5250: illegal position in the item");
 1070 
 1071                     /* Either convert last unformatted node of indirect item to direct item or increase
 1072                        its free space.  */
 1073                     if ( pos_in_item (p_s_path) == n_unfm_number )  {
 1074                         *p_n_cut_size = 0; /* Nothing to cut. */
 1075                         return M_CONVERT; /* Maybe convert last unformatted node to the direct item. */
 1076                     }
 1077                     /* Calculate size to cut. */
 1078                     *p_n_cut_size = -(ih_item_len(&s_ih) - pos_in_item(p_s_path) * UNFM_P_SIZE);
 1079 
 1080                     c_mode = M_CUT;     /* Cut from this indirect item. */
 1081                 }
 1082             }
 1083 
 1084             RFALSE( n_unfm_number <= pos_in_item (p_s_path),
 1085                     "PAP-5260: illegal position in the indirect item");
 1086 
 1087             /* pointers to be cut */
 1088             n_unfm_number -= pos_in_item (p_s_path);
 1089             /* Set pointer to the last unformatted node pointer that is to be cut. */
 1090             p_n_unfm_pointer = (__u32 *)B_I_PITEM(p_s_bh, &s_ih) + I_UNFM_NUM(&s_ih) - 1 - *p_n_removed;
 1091 
 1092 
 1093             /* We go through the unformatted nodes pointers of the indirect
 1094                item and look for the unformatted nodes in the cache. If we
 1095                found some of them we free it, zero corresponding indirect item
 1096                entry and log buffer containing that indirect item. For this we
 1097                need to prepare last path element for logging. If some
 1098                unformatted node has b_count > 1 we must not free this
 1099                unformatted node since it is in use. */
 1100             reiserfs_prepare_for_journal(p_s_sb, p_s_bh, 1);
 1101             // note: path could be changed, first line in for loop takes care
 1102             // of it
 1103 
 1104             for (n_counter = *p_n_removed;
 1105                  n_counter < n_unfm_number; n_counter++, p_n_unfm_pointer-- ) {
 1106 
 1107                 if (item_moved (&s_ih, p_s_path)) {
 1108                     need_research = 1 ;
 1109                     break;
 1110                 }
 1111                 RFALSE( p_n_unfm_pointer < (__u32 *)B_I_PITEM(p_s_bh, &s_ih) ||
 1112                         p_n_unfm_pointer > (__u32 *)B_I_PITEM(p_s_bh, &s_ih) + I_UNFM_NUM(&s_ih) - 1,
 1113                         "vs-5265: pointer out of range");
 1114 
 1115                 /* Hole, nothing to remove. */
 1116                 if ( ! get_block_num(p_n_unfm_pointer,0) )  { 
 1117                         (*p_n_removed)++;
 1118                         continue;
 1119                 }
 1120 
 1121                 (*p_n_removed)++;
 1122 
 1123                 tmp = get_block_num(p_n_unfm_pointer,0);
 1124                 put_block_num(p_n_unfm_pointer, 0, 0);
 1125                 journal_mark_dirty (th, p_s_sb, p_s_bh);
 1126                 inode->i_blocks -= p_s_sb->s_blocksize / 512;
 1127                 reiserfs_free_block(th, tmp);
 1128                 /* In case of big fragmentation it is possible that each block
 1129                    freed will cause dirtying of one more bitmap and then we will
 1130                    quickly overflow our transaction space. This is a
 1131                    counter-measure against that scenario */
 1132                 if (journal_transaction_should_end(th, th->t_blocks_allocated)) {
 1133                     int orig_len_alloc = th->t_blocks_allocated ;
 1134                     pathrelse(p_s_path) ;
 1135 
 1136                     journal_end(th, p_s_sb, orig_len_alloc) ;
 1137                     journal_begin(th, p_s_sb, orig_len_alloc) ;
 1138                     reiserfs_update_inode_transaction(inode) ;
 1139                     need_research = 1;
 1140                     break;
 1141                 }
 1142 
 1143                 if ( item_moved (&s_ih, p_s_path) )  {
 1144                         need_research = 1;
 1145                         break ;
 1146                 }
 1147             }
 1148 
 1149             /* a trick.  If the buffer has been logged, this
 1150             ** will do nothing.  If we've broken the loop without
 1151             ** logging it, it will restore the buffer
 1152             **
 1153             */
 1154             reiserfs_restore_prepared_buffer(p_s_sb, p_s_bh);
 1155 
 1156             /* This loop can be optimized. */
 1157         } while ( (*p_n_removed < n_unfm_number || need_research) &&
 1158                   search_for_position_by_key(p_s_sb, p_s_item_key, p_s_path) == POSITION_FOUND );
 1159 
 1160         RFALSE( *p_n_removed < n_unfm_number, 
 1161                 "PAP-5310: indirect item is not found");
 1162         RFALSE( item_moved (&s_ih, p_s_path), 
 1163                 "after while, comp failed, retry") ;
 1164 
 1165         if (c_mode == M_CUT)
 1166             pos_in_item (p_s_path) *= UNFM_P_SIZE;
 1167         return c_mode;
 1168     }
 1169 }
 1170 
 1171 
 1172 /* Calculate bytes number which will be deleted or cutted in the balance. */
 1173 int calc_deleted_bytes_number(
 1174     struct  tree_balance  * p_s_tb,
 1175     char                    c_mode
 1176     ) {
 1177     int                     n_del_size;
 1178     struct  item_head     * p_le_ih = PATH_PITEM_HEAD(p_s_tb->tb_path);
 1179 
 1180     if ( is_statdata_le_ih (p_le_ih) )
 1181         return 0;
 1182 
 1183     if ( is_direntry_le_ih (p_le_ih) ) {
 1184         // return EMPTY_DIR_SIZE; /* We delete emty directoris only. */
 1185         // we can't use EMPTY_DIR_SIZE, as old format dirs have a different
 1186         // empty size.  ick. FIXME, is this right?
 1187         //
 1188         return ih_item_len(p_le_ih);
 1189     }
 1190     n_del_size = ( c_mode == M_DELETE ) ? ih_item_len(p_le_ih) : -p_s_tb->insert_size[0];
 1191 
 1192     if ( is_indirect_le_ih (p_le_ih) )
 1193         n_del_size = (n_del_size/UNFM_P_SIZE)*
 1194           (PATH_PLAST_BUFFER(p_s_tb->tb_path)->b_size);// - get_ih_free_space (p_le_ih);
 1195     return n_del_size;
 1196 }
 1197 
 1198 static void init_tb_struct(
 1199     struct reiserfs_transaction_handle *th,
 1200     struct tree_balance * p_s_tb,
 1201     struct super_block  * p_s_sb,
 1202     struct path         * p_s_path,
 1203     int                   n_size
 1204     ) {
 1205     memset (p_s_tb,'\0',sizeof(struct tree_balance));
 1206     p_s_tb->transaction_handle = th ;
 1207     p_s_tb->tb_sb = p_s_sb;
 1208     p_s_tb->tb_path = p_s_path;
 1209     PATH_OFFSET_PBUFFER(p_s_path, ILLEGAL_PATH_ELEMENT_OFFSET) = NULL;
 1210     PATH_OFFSET_POSITION(p_s_path, ILLEGAL_PATH_ELEMENT_OFFSET) = 0;
 1211     p_s_tb->insert_size[0] = n_size;
 1212 }
 1213 
 1214 
 1215 
 1216 void padd_item (char * item, int total_length, int length)
 1217 {
 1218     int i;
 1219 
 1220     for (i = total_length; i > length; )
 1221         item [--i] = 0;
 1222 }
 1223 
 1224 
 1225 /* Delete object item. */
 1226 int reiserfs_delete_item (struct reiserfs_transaction_handle *th, 
 1227                           struct path * p_s_path, /* Path to the deleted item. */
 1228                           const struct cpu_key * p_s_item_key, /* Key to search for the deleted item.  */
 1229                           struct inode * p_s_inode,/* inode is here just to update i_blocks */
 1230                           struct buffer_head  * p_s_un_bh)    /* NULL or unformatted node pointer.    */
 1231 {
 1232     struct super_block * p_s_sb = p_s_inode->i_sb;
 1233     struct tree_balance   s_del_balance;
 1234     struct item_head      s_ih;
 1235     int                   n_ret_value,
 1236         n_del_size,
 1237         n_removed;
 1238 
 1239 #ifdef CONFIG_REISERFS_CHECK
 1240     char                  c_mode;
 1241     int                 n_iter = 0;
 1242 #endif
 1243 
 1244     init_tb_struct(th, &s_del_balance, p_s_sb, p_s_path, 0/*size is unknown*/);
 1245 
 1246     while ( 1 ) {
 1247         n_removed = 0;
 1248 
 1249 #ifdef CONFIG_REISERFS_CHECK
 1250         n_iter++;
 1251         c_mode =
 1252 #endif
 1253             prepare_for_delete_or_cut(th, p_s_inode, p_s_path, p_s_item_key, &n_removed, &n_del_size, max_reiserfs_offset (p_s_inode));
 1254 
 1255         RFALSE( c_mode != M_DELETE, "PAP-5320: mode must be M_DELETE");
 1256 
 1257         copy_item_head(&s_ih, PATH_PITEM_HEAD(p_s_path));
 1258         s_del_balance.insert_size[0] = n_del_size;
 1259 
 1260         n_ret_value = fix_nodes(M_DELETE, &s_del_balance, NULL, 0);
 1261         if ( n_ret_value != REPEAT_SEARCH )
 1262             break;
 1263 
 1264         PROC_INFO_INC( p_s_sb, delete_item_restarted );
 1265 
 1266         // file system changed, repeat search
 1267         n_ret_value = search_for_position_by_key(p_s_sb, p_s_item_key, p_s_path);
 1268         if (n_ret_value == IO_ERROR)
 1269             break;
 1270         if (n_ret_value == FILE_NOT_FOUND) {
 1271             reiserfs_warning (p_s_sb, "vs-5340: reiserfs_delete_item: "
 1272                               "no items of the file %K found\n", p_s_item_key);
 1273             break;
 1274         }
 1275     } /* while (1) */
 1276 
 1277     if ( n_ret_value != CARRY_ON ) {
 1278         unfix_nodes(&s_del_balance);
 1279         return 0;
 1280     }
 1281 
 1282     // reiserfs_delete_item returns item length when success
 1283     n_ret_value = calc_deleted_bytes_number(&s_del_balance, M_DELETE);
 1284 
 1285     if ( p_s_un_bh )  {
 1286         int off;
 1287         char *data ;
 1288 
 1289         /* We are in direct2indirect conversion, so move tail contents
 1290            to the unformatted node */
 1291         /* note, we do the copy before preparing the buffer because we
 1292         ** don't care about the contents of the unformatted node yet.
 1293         ** the only thing we really care about is the direct item's data
 1294         ** is in the unformatted node.
 1295         **
 1296         ** Otherwise, we would have to call reiserfs_prepare_for_journal on
 1297         ** the unformatted node, which might schedule, meaning we'd have to
 1298         ** loop all the way back up to the start of the while loop.
 1299         **
 1300         ** The unformatted node must be dirtied later on.  We can't be
 1301         ** sure here if the entire tail has been deleted yet.
 1302         **
 1303         ** p_s_un_bh is from the page cache (all unformatted nodes are
 1304         ** from the page cache) and might be a highmem page.  So, we
 1305         ** can't use p_s_un_bh->b_data.  But, the page has already been
 1306         ** kmapped, so we can use page_address()
 1307         ** -clm
 1308         */
 1309 
 1310         data = page_address(p_s_un_bh->b_page) ;
 1311         off = ((le_ih_k_offset (&s_ih) - 1) & (PAGE_CACHE_SIZE - 1));
 1312         memcpy(data + off,
 1313                B_I_PITEM(PATH_PLAST_BUFFER(p_s_path), &s_ih), n_ret_value);
 1314     }
 1315 
 1316     /* Perform balancing after all resources have been collected at once. */ 
 1317     do_balance(&s_del_balance, NULL, NULL, M_DELETE);
 1318 
 1319     /* Return deleted body length */
 1320     return n_ret_value;
 1321 }
 1322 
 1323 
 1324 /* Summary Of Mechanisms For Handling Collisions Between Processes:
 1325 
 1326  deletion of the body of the object is performed by iput(), with the
 1327  result that if multiple processes are operating on a file, the
 1328  deletion of the body of the file is deferred until the last process
 1329  that has an open inode performs its iput().
 1330 
 1331  writes and truncates are protected from collisions by use of
 1332  semaphores.
 1333 
 1334  creates, linking, and mknod are protected from collisions with other
 1335  processes by making the reiserfs_add_entry() the last step in the
 1336  creation, and then rolling back all changes if there was a collision.
 1337  - Hans
 1338 */
 1339 
 1340 
 1341 /* this deletes item which never gets split */
 1342 void reiserfs_delete_solid_item (struct reiserfs_transaction_handle *th,
 1343                                  struct key * key)
 1344 {
 1345     struct tree_balance tb;
 1346     INITIALIZE_PATH (path);
 1347     int item_len;
 1348     int tb_init = 0 ;
 1349     struct cpu_key cpu_key;
 1350     int retval;
 1351     
 1352     le_key2cpu_key (&cpu_key, key);
 1353     
 1354     while (1) {
 1355         retval = search_item (th->t_super, &cpu_key, &path);
 1356         if (retval == IO_ERROR) {
 1357             reiserfs_warning (th->t_super, "vs-5350: reiserfs_delete_solid_item: "
 1358                               "i/o failure occurred trying to delete %K\n", &cpu_key);
 1359             break;
 1360         }
 1361         if (retval != ITEM_FOUND) {
 1362             pathrelse (&path);
 1363             // No need for a warning, if there is just no free space to insert '..' item into the newly-created subdir
 1364             if ( !( (unsigned long long) GET_HASH_VALUE (le_key_k_offset (le_key_version (key), key)) == 0 && \
 1365                  GET_GENERATION_NUMBER (le_key_k_offset (le_key_version (key), key)) == 1 ) )
 1366                 reiserfs_warning (th->t_super, "vs-5355: reiserfs_delete_solid_item: %k not found\n", key);
 1367             break;
 1368         }
 1369         if (!tb_init) {
 1370             tb_init = 1 ;
 1371             item_len = ih_item_len( PATH_PITEM_HEAD(&path) );
 1372             init_tb_struct (th, &tb, th->t_super, &path, - (IH_SIZE + item_len));
 1373         }
 1374 
 1375         retval = fix_nodes (M_DELETE, &tb, NULL, 0);
 1376         if (retval == REPEAT_SEARCH) {
 1377             PROC_INFO_INC( th -> t_super, delete_solid_item_restarted );
 1378             continue;
 1379         }
 1380 
 1381         if (retval == CARRY_ON) {
 1382             do_balance (&tb, 0, 0, M_DELETE);
 1383             break;
 1384         }
 1385 
 1386         // IO_ERROR, NO_DISK_SPACE, etc
 1387         reiserfs_warning (th->t_super, "vs-5360: reiserfs_delete_solid_item: "
 1388                           "could not delete %K due to fix_nodes failure\n", &cpu_key);
 1389         unfix_nodes (&tb);
 1390         break;
 1391     }
 1392 
 1393     reiserfs_check_path(&path) ;
 1394 }
 1395 
 1396 
 1397 void reiserfs_delete_object (struct reiserfs_transaction_handle *th, struct inode * inode)
 1398 {
 1399     inode->i_size = 0;
 1400 
 1401     /* for directory this deletes item containing "." and ".." */
 1402     reiserfs_do_truncate (th, inode, NULL, 0/*no timestamp updates*/);
 1403     
 1404 #if defined( USE_INODE_GENERATION_COUNTER )
 1405     if( !old_format_only ( th -> t_super ) )
 1406       {
 1407        __u32 *inode_generation;
 1408        
 1409        inode_generation = 
 1410          &th -> t_super -> u.reiserfs_sb.s_rs -> s_inode_generation;
 1411        *inode_generation = cpu_to_le32( le32_to_cpu( *inode_generation ) + 1 );
 1412       }
 1413 /* USE_INODE_GENERATION_COUNTER */
 1414 #endif
 1415     reiserfs_delete_solid_item (th, INODE_PKEY (inode));
 1416 }
 1417 
 1418 
 1419 static int maybe_indirect_to_direct (struct reiserfs_transaction_handle *th, 
 1420                               struct inode * p_s_inode,
 1421                               struct page *page, 
 1422                               struct path         * p_s_path,
 1423                               const struct cpu_key      * p_s_item_key,
 1424                               loff_t         n_new_file_size,
 1425                               char                * p_c_mode
 1426                               ) {
 1427     struct super_block * p_s_sb = p_s_inode->i_sb;
 1428     int n_block_size = p_s_sb->s_blocksize;
 1429     int cut_bytes;
 1430 
 1431     if (n_new_file_size != p_s_inode->i_size)
 1432         BUG ();
 1433 
 1434     /* the page being sent in could be NULL if there was an i/o error
 1435     ** reading in the last block.  The user will hit problems trying to
 1436     ** read the file, but for now we just skip the indirect2direct
 1437     */
 1438     if (atomic_read(&p_s_inode->i_count) > 1 || 
 1439         !tail_has_to_be_packed (p_s_inode) || 
 1440         !page || (p_s_inode->u.reiserfs_i.i_flags & i_nopack_mask)) {
 1441         // leave tail in an unformatted node    
 1442         *p_c_mode = M_SKIP_BALANCING;
 1443         cut_bytes = n_block_size - (n_new_file_size & (n_block_size - 1));
 1444         pathrelse(p_s_path);
 1445         return cut_bytes;
 1446     }
 1447     /* Permorm the conversion to a direct_item. */
 1448     /*return indirect_to_direct (p_s_inode, p_s_path, p_s_item_key, n_new_file_size, p_c_mode);*/
 1449     return indirect2direct (th, p_s_inode, page, p_s_path, p_s_item_key, n_new_file_size, p_c_mode);
 1450 }
 1451 
 1452 
 1453 /* we did indirect_to_direct conversion. And we have inserted direct
 1454    item successesfully, but there were no disk space to cut unfm
 1455    pointer being converted. Therefore we have to delete inserted
 1456    direct item(s) */
 1457 static void indirect_to_direct_roll_back (struct reiserfs_transaction_handle *th, struct inode * inode, struct path * path)
 1458 {
 1459     struct cpu_key tail_key;
 1460     int tail_len;
 1461     int removed;
 1462 
 1463     make_cpu_key (&tail_key, inode, inode->i_size + 1, TYPE_DIRECT, 4);// !!!!
 1464     tail_key.key_length = 4;
 1465 
 1466     tail_len = (cpu_key_k_offset (&tail_key) & (inode->i_sb->s_blocksize - 1)) - 1;
 1467     while (tail_len) {
 1468         /* look for the last byte of the tail */
 1469         if (search_for_position_by_key (inode->i_sb, &tail_key, path) == POSITION_NOT_FOUND)
 1470             reiserfs_panic (inode->i_sb, "vs-5615: indirect_to_direct_roll_back: found invalid item");
 1471         RFALSE( path->pos_in_item != ih_item_len(PATH_PITEM_HEAD (path)) - 1,
 1472                 "vs-5616: appended bytes found");
 1473         PATH_LAST_POSITION (path) --;
 1474         
 1475         removed = reiserfs_delete_item (th, path, &tail_key, inode, 0/*unbh not needed*/);
 1476         RFALSE( removed <= 0 || removed > tail_len,
 1477                 "vs-5617: there was tail %d bytes, removed item length %d bytes",
 1478                 tail_len, removed);
 1479         tail_len -= removed;
 1480         set_cpu_key_k_offset (&tail_key, cpu_key_k_offset (&tail_key) - removed);
 1481     }
 1482     reiserfs_warning (inode->i_sb, "indirect_to_direct_roll_back: indirect_to_direct conversion has been rolled back due to lack of disk space\n");
 1483     //mark_file_without_tail (inode);
 1484     mark_inode_dirty (inode);
 1485 }
 1486 
 1487 
 1488 /* (Truncate or cut entry) or delete object item. Returns < 0 on failure */
 1489 int reiserfs_cut_from_item (struct reiserfs_transaction_handle *th, 
 1490                             struct path * p_s_path,
 1491                             struct cpu_key * p_s_item_key,
 1492                             struct inode * p_s_inode,
 1493                             struct page *page, 
 1494                             loff_t n_new_file_size)
 1495 {
 1496     struct super_block * p_s_sb = p_s_inode->i_sb;
 1497     /* Every function which is going to call do_balance must first
 1498        create a tree_balance structure.  Then it must fill up this
 1499        structure by using the init_tb_struct and fix_nodes functions.
 1500        After that we can make tree balancing. */
 1501     struct tree_balance s_cut_balance;
 1502     int n_cut_size = 0,        /* Amount to be cut. */
 1503         n_ret_value = CARRY_ON,
 1504         n_removed = 0,     /* Number of the removed unformatted nodes. */
 1505         n_is_inode_locked = 0;
 1506     char                c_mode;            /* Mode of the balance. */
 1507     int retval2 = -1;
 1508     
 1509     
 1510     init_tb_struct(th, &s_cut_balance, p_s_inode->i_sb, p_s_path, n_cut_size);
 1511 
 1512 
 1513     /* Repeat this loop until we either cut the item without needing
 1514        to balance, or we fix_nodes without schedule occuring */
 1515     while ( 1 ) {
 1516         /* Determine the balance mode, position of the first byte to
 1517            be cut, and size to be cut.  In case of the indirect item
 1518            free unformatted nodes which are pointed to by the cut
 1519            pointers. */
 1520       
 1521         c_mode = prepare_for_delete_or_cut(th, p_s_inode, p_s_path, p_s_item_key, &n_removed, 
 1522                                            &n_cut_size, n_new_file_size);
 1523         if ( c_mode == M_CONVERT )  {
 1524             /* convert last unformatted node to direct item or leave
 1525                tail in the unformatted node */
 1526             RFALSE( n_ret_value != CARRY_ON, "PAP-5570: can not convert twice");
 1527 
 1528             n_ret_value = maybe_indirect_to_direct (th, p_s_inode, page, p_s_path, p_s_item_key,
 1529                                                     n_new_file_size, &c_mode);
 1530             if ( c_mode == M_SKIP_BALANCING )
 1531                 /* tail has been left in the unformatted node */
 1532                 return n_ret_value;
 1533 
 1534             n_is_inode_locked = 1;
 1535           
 1536             /* removing of last unformatted node will change value we
 1537                have to return to truncate. Save it */
 1538             retval2 = n_ret_value;
 1539             /*retval2 = p_s_sb->s_blocksize - (n_new_file_size & (p_s_sb->s_blocksize - 1));*/
 1540           
 1541             /* So, we have performed the first part of the conversion:
 1542                inserting the new direct item.  Now we are removing the
 1543                last unformatted node pointer. Set key to search for
 1544                it. */
 1545             set_cpu_key_k_type (p_s_item_key, TYPE_INDIRECT);
 1546             p_s_item_key->key_length = 4;
 1547             n_new_file_size -= (n_new_file_size & (p_s_sb->s_blocksize - 1));
 1548             set_cpu_key_k_offset (p_s_item_key, n_new_file_size + 1);
 1549             if ( search_for_position_by_key(p_s_sb, p_s_item_key, p_s_path) == POSITION_NOT_FOUND ){
 1550                 print_block (PATH_PLAST_BUFFER (p_s_path), 3, PATH_LAST_POSITION (p_s_path) - 1, PATH_LAST_POSITION (p_s_path) + 1);
 1551                 reiserfs_panic(p_s_sb, "PAP-5580: reiserfs_cut_from_item: item to convert does not exist (%K)", p_s_item_key);
 1552             }
 1553             continue;
 1554         }
 1555         if (n_cut_size == 0) {
 1556             pathrelse (p_s_path);
 1557             return 0;
 1558         }
 1559 
 1560         s_cut_balance.insert_size[0] = n_cut_size;
 1561         
 1562         n_ret_value = fix_nodes(c_mode, &s_cut_balance, NULL, 0);
 1563         if ( n_ret_value != REPEAT_SEARCH )
 1564             break;
 1565         
 1566         PROC_INFO_INC( p_s_sb, cut_from_item_restarted );
 1567 
 1568         n_ret_value = search_for_position_by_key(p_s_sb, p_s_item_key, p_s_path);
 1569         if (n_ret_value == POSITION_FOUND)
 1570             continue;
 1571 
 1572         reiserfs_warning (p_s_sb, "PAP-5610: reiserfs_cut_from_item: item %K not found\n", p_s_item_key);
 1573         unfix_nodes (&s_cut_balance);
 1574         return (n_ret_value == IO_ERROR) ? -EIO : -ENOENT;
 1575     } /* while */
 1576   
 1577     // check fix_nodes results (IO_ERROR or NO_DISK_SPACE)
 1578     if ( n_ret_value != CARRY_ON ) {
 1579         if ( n_is_inode_locked ) {
 1580             // FIXME: this seems to be not needed: we are always able
 1581             // to cut item
 1582             indirect_to_direct_roll_back (th, p_s_inode, p_s_path);
 1583         }
 1584         if (n_ret_value == NO_DISK_SPACE)
 1585             reiserfs_warning (p_s_sb, "NO_DISK_SPACE\n");
 1586         unfix_nodes (&s_cut_balance);
 1587         return -EIO;
 1588     }
 1589 
 1590     /* go ahead and perform balancing */
 1591     
 1592     RFALSE( c_mode == M_PASTE || c_mode == M_INSERT, "illegal mode");
 1593 
 1594     /* Calculate number of bytes that need to be cut from the item. */
 1595     if (retval2 == -1)
 1596         n_ret_value = calc_deleted_bytes_number(&s_cut_balance, c_mode);
 1597     else
 1598         n_ret_value = retval2;
 1599     
 1600     if ( c_mode == M_DELETE ) {
 1601         struct item_head * p_le_ih = PATH_PITEM_HEAD (s_cut_balance.tb_path);
 1602         
 1603         if ( is_direct_le_ih (p_le_ih) && (le_ih_k_offset (p_le_ih) & (p_s_sb->s_blocksize - 1)) == 1 ) {
 1604             /* we delete first part of tail which was stored in direct
 1605                item(s) */
 1606             // FIXME: this is to keep 3.5 happy
 1607             p_s_inode->u.reiserfs_i.i_first_direct_byte = U32_MAX;
 1608             p_s_inode->i_blocks -= p_s_sb->s_blocksize / 512;
 1609         }
 1610     }
 1611 
 1612 #ifdef CONFIG_REISERFS_CHECK
 1613     if (n_is_inode_locked) {
 1614         struct item_head * le_ih = PATH_PITEM_HEAD (s_cut_balance.tb_path);
 1615         /* we are going to complete indirect2direct conversion. Make
 1616            sure, that we exactly remove last unformatted node pointer
 1617            of the item */
 1618         if (!is_indirect_le_ih (le_ih))
 1619             reiserfs_panic (p_s_sb, "vs-5652: reiserfs_cut_from_item: "
 1620                             "item must be indirect %h", le_ih);
 1621 
 1622         if (c_mode == M_DELETE && ih_item_len(le_ih) != UNFM_P_SIZE)
 1623             reiserfs_panic (p_s_sb, "vs-5653: reiserfs_cut_from_item: "
 1624                             "completing indirect2direct conversion indirect item %h "
 1625                             "being deleted must be of 4 byte long", le_ih);
 1626 
 1627         if (c_mode == M_CUT && s_cut_balance.insert_size[0] != -UNFM_P_SIZE) {
 1628             reiserfs_panic (p_s_sb, "vs-5654: reiserfs_cut_from_item: "
 1629                             "can not complete indirect2direct conversion of %h (CUT, insert_size==%d)",
 1630                             le_ih, s_cut_balance.insert_size[0]);
 1631         }
 1632         /* it would be useful to make sure, that right neighboring
 1633            item is direct item of this file */
 1634     }
 1635 #endif
 1636     
 1637     do_balance(&s_cut_balance, NULL, NULL, c_mode);
 1638     if ( n_is_inode_locked ) {
 1639         /* we've done an indirect->direct conversion.  when the data block
 1640         ** was freed, it was removed from the list of blocks that must
 1641         ** be flushed before the transaction commits, so we don't need to
 1642         ** deal with it here.
 1643         */
 1644         p_s_inode->u.reiserfs_i.i_flags &= ~i_pack_on_close_mask;
 1645     }
 1646     return n_ret_value;
 1647 }
 1648 
 1649 
 1650 static void truncate_directory (struct reiserfs_transaction_handle *th, struct inode * inode)
 1651 {
 1652     if (inode->i_nlink)
 1653         reiserfs_warning (th->t_super, "vs-5655: truncate_directory: link count != 0\n");
 1654 
 1655     set_le_key_k_offset (KEY_FORMAT_3_5, INODE_PKEY (inode), DOT_OFFSET);
 1656     set_le_key_k_type (KEY_FORMAT_3_5, INODE_PKEY (inode), TYPE_DIRENTRY);
 1657     reiserfs_delete_solid_item (th, INODE_PKEY (inode));
 1658 
 1659     set_le_key_k_offset (KEY_FORMAT_3_5, INODE_PKEY (inode), SD_OFFSET);
 1660     set_le_key_k_type (KEY_FORMAT_3_5, INODE_PKEY (inode), TYPE_STAT_DATA);    
 1661 }
 1662 
 1663 
 1664 
 1665 
 1666 /* Truncate file to the new size. Note, this must be called with a transaction
 1667    already started */
 1668 void reiserfs_do_truncate (struct reiserfs_transaction_handle *th, 
 1669                            struct  inode * p_s_inode, /* ->i_size contains new
 1670                                                          size */
 1671                            struct page *page, /* up to date for last block */
 1672                            int update_timestamps  /* when it is called by
 1673                                                      file_release to convert
 1674                                                      the tail - no timestamps
 1675                                                      should be updated */
 1676     ) {
 1677     INITIALIZE_PATH (s_search_path);       /* Path to the current object item. */
 1678     struct item_head    * p_le_ih;         /* Pointer to an item header. */
 1679     struct cpu_key      s_item_key;     /* Key to search for a previous file item. */
 1680     loff_t         n_file_size,    /* Old file size. */
 1681         n_new_file_size;/* New file size. */
 1682     int                   n_deleted;      /* Number of deleted or truncated bytes. */
 1683     int retval;
 1684 
 1685     if ( ! (S_ISREG(p_s_inode->i_mode) || S_ISDIR(p_s_inode->i_mode) || S_ISLNK(p_s_inode->i_mode)) )
 1686         return;
 1687 
 1688     if (S_ISDIR(p_s_inode->i_mode)) {
 1689         // deletion of directory - no need to update timestamps
 1690         truncate_directory (th, p_s_inode);
 1691         return;
 1692     }
 1693 
 1694     /* Get new file size. */
 1695     n_new_file_size = p_s_inode->i_size;
 1696 
 1697     // FIXME: note, that key type is unimportant here
 1698     make_cpu_key (&s_item_key, p_s_inode, max_reiserfs_offset (p_s_inode), TYPE_DIRECT, 3);
 1699 
 1700     retval = search_for_position_by_key(p_s_inode->i_sb, &s_item_key, &s_search_path);
 1701     if (retval == IO_ERROR) {
 1702         reiserfs_warning (p_s_inode->i_sb, "vs-5657: reiserfs_do_truncate: "
 1703                           "i/o failure occurred trying to truncate %K\n", &s_item_key);
 1704         return;
 1705     }
 1706     if (retval == POSITION_FOUND || retval == FILE_NOT_FOUND) {
 1707         pathrelse (&s_search_path);
 1708         reiserfs_warning (p_s_inode->i_sb, "PAP-5660: reiserfs_do_truncate: "
 1709                           "wrong result %d of search for %K\n", retval, &s_item_key);
 1710         return;
 1711     }
 1712 
 1713     s_search_path.pos_in_item --;
 1714 
 1715     /* Get real file size (total length of all file items) */
 1716     p_le_ih = PATH_PITEM_HEAD(&s_search_path);
 1717     if ( is_statdata_le_ih (p_le_ih) )
 1718         n_file_size = 0;
 1719     else {
 1720         loff_t offset = le_ih_k_offset (p_le_ih);
 1721         int bytes = op_bytes_number (p_le_ih,p_s_inode->i_sb->s_blocksize);
 1722 
 1723         /* this may mismatch with real file size: if last direct item
 1724            had no padding zeros and last unformatted node had no free
 1725            space, this file would have this file size */
 1726         n_file_size = offset + bytes - 1;
 1727     }
 1728 
 1729     if ( n_file_size == 0 || n_file_size < n_new_file_size ) {
 1730         goto update_and_out ;
 1731     }
 1732 
 1733     /* Update key to search for the last file item. */
 1734     set_cpu_key_k_offset (&s_item_key, n_file_size);
 1735 
 1736     do  {
 1737         /* Cut or delete file item. */
 1738         n_deleted = reiserfs_cut_from_item(th, &s_search_path, &s_item_key, p_s_inode,  page, n_new_file_size);
 1739         if (n_deleted < 0) {
 1740             reiserfs_warning (th->t_super, "vs-5665: reiserfs_truncate_file: cut_from_item failed\n");
 1741             reiserfs_check_path(&s_search_path) ;
 1742             return;
 1743         }
 1744 
 1745         RFALSE( n_deleted > n_file_size,
 1746                 "PAP-5670: reiserfs_truncate_file returns too big number: deleted %d, file_size %lu, item_key %K",
 1747                 n_deleted, n_file_size, &s_item_key);
 1748 
 1749         /* Change key to search the last file item. */
 1750         n_file_size -= n_deleted;
 1751 
 1752         set_cpu_key_k_offset (&s_item_key, n_file_size);
 1753 
 1754         /* While there are bytes to truncate and previous file item is presented in the tree. */
 1755 
 1756         /*
 1757         ** This loop could take a really long time, and could log 
 1758         ** many more blocks than a transaction can hold.  So, we do a polite
 1759         ** journal end here, and if the transaction needs ending, we make
 1760         ** sure the file is consistent before ending the current trans
 1761         ** and starting a new one
 1762         */
 1763         if (journal_transaction_should_end(th, th->t_blocks_allocated)) {
 1764           int orig_len_alloc = th->t_blocks_allocated ;
 1765           decrement_counters_in_path(&s_search_path) ;
 1766 
 1767           if (update_timestamps) {
 1768               p_s_inode->i_mtime = p_s_inode->i_ctime = CURRENT_TIME;
 1769           } 
 1770           reiserfs_update_sd(th, p_s_inode) ;
 1771 
 1772           journal_end(th, p_s_inode->i_sb, orig_len_alloc) ;
 1773           journal_begin(th, p_s_inode->i_sb, orig_len_alloc) ;
 1774           reiserfs_update_inode_transaction(p_s_inode) ;
 1775         }
 1776     } while ( n_file_size > ROUND_UP (n_new_file_size) &&
 1777               search_for_position_by_key(p_s_inode->i_sb, &s_item_key, &s_search_path) == POSITION_FOUND )  ;
 1778 
 1779     RFALSE( n_file_size > ROUND_UP (n_new_file_size),
 1780             "PAP-5680: truncate did not finish: new_file_size %Ld, current %Ld, oid %d\n",
 1781             n_new_file_size, n_file_size, s_item_key.on_disk_key.k_objectid);
 1782 
 1783 update_and_out:
 1784     if (update_timestamps) {
 1785         // this is truncate, not file closing
 1786         p_s_inode->i_mtime = p_s_inode->i_ctime = CURRENT_TIME;
 1787     }
 1788     reiserfs_update_sd (th, p_s_inode);
 1789 
 1790     pathrelse(&s_search_path) ;
 1791 }
 1792 
 1793 
 1794 #ifdef CONFIG_REISERFS_CHECK
 1795 // this makes sure, that we __append__, not overwrite or add holes
 1796 static void check_research_for_paste (struct path * path, 
 1797                                       const struct cpu_key * p_s_key)
 1798 {
 1799     struct item_head * found_ih = get_ih (path);
 1800     
 1801     if (is_direct_le_ih (found_ih)) {
 1802         if (le_ih_k_offset (found_ih) + op_bytes_number (found_ih, get_last_bh (path)->b_size) !=
 1803             cpu_key_k_offset (p_s_key) ||
 1804             op_bytes_number (found_ih, get_last_bh (path)->b_size) != pos_in_item (path))
 1805             reiserfs_panic (0, "PAP-5720: check_research_for_paste: "
 1806                             "found direct item %h or position (%d) does not match to key %K",
 1807                             found_ih, pos_in_item (path), p_s_key);
 1808     }
 1809     if (is_indirect_le_ih (found_ih)) {
 1810         if (le_ih_k_offset (found_ih) + op_bytes_number (found_ih, get_last_bh (path)->b_size) != cpu_key_k_offset (p_s_key) || 
 1811             I_UNFM_NUM (found_ih) != pos_in_item (path) ||
 1812             get_ih_free_space (found_ih) != 0)
 1813             reiserfs_panic (0, "PAP-5730: check_research_for_paste: "
 1814                             "found indirect item (%h) or position (%d) does not match to key (%K)",
 1815                             found_ih, pos_in_item (path), p_s_key);
 1816     }
 1817 }
 1818 #endif /* config reiserfs check */
 1819 
 1820 
 1821 /* Paste bytes to the existing item. Returns bytes number pasted into the item. */
 1822 int reiserfs_paste_into_item (struct reiserfs_transaction_handle *th, 
 1823                               struct path         * p_s_search_path,    /* Path to the pasted item.          */
 1824                               const struct cpu_key      * p_s_key,              /* Key to search for the needed item.*/
 1825                               const char          * p_c_body,           /* Pointer to the bytes to paste.    */
 1826                               int                   n_pasted_size)      /* Size of pasted bytes.             */
 1827 {
 1828     struct tree_balance s_paste_balance;
 1829     int                 retval;
 1830 
 1831     init_tb_struct(th, &s_paste_balance, th->t_super, p_s_search_path, n_pasted_size);
 1832 #ifdef DISPLACE_NEW_PACKING_LOCALITIES
 1833     s_paste_balance.key = p_s_key->on_disk_key;
 1834 #endif
 1835     
 1836     while ( (retval = fix_nodes(M_PASTE, &s_paste_balance, NULL, p_c_body)) == REPEAT_SEARCH ) {
 1837         /* file system changed while we were in the fix_nodes */
 1838         PROC_INFO_INC( th -> t_super, paste_into_item_restarted );
 1839         retval = search_for_position_by_key (th->t_super, p_s_key, p_s_search_path);
 1840         if (retval == IO_ERROR) {
 1841             retval = -EIO ;
 1842             goto error_out ;
 1843         }
 1844         if (retval == POSITION_FOUND) {
 1845             reiserfs_warning (th->t_super, "PAP-5710: reiserfs_paste_into_item: entry or pasted byte (%K) exists\n", p_s_key);
 1846             retval = -EEXIST ;
 1847             goto error_out ;
 1848         }
 1849         
 1850 #ifdef CONFIG_REISERFS_CHECK
 1851         check_research_for_paste (p_s_search_path, p_s_key);
 1852 #endif
 1853     }
 1854 
 1855     /* Perform balancing after all resources are collected by fix_nodes, and
 1856        accessing them will not risk triggering schedule. */
 1857     if ( retval == CARRY_ON ) {
 1858         do_balance(&s_paste_balance, NULL/*ih*/, p_c_body, M_PASTE);
 1859         return 0;
 1860     }
 1861     retval = (retval == NO_DISK_SPACE) ? -ENOSPC : -EIO;
 1862 error_out:
 1863     /* this also releases the path */
 1864     unfix_nodes(&s_paste_balance);
 1865     return retval ;
 1866 }
 1867 
 1868 
 1869 /* Insert new item into the buffer at the path. */
 1870 int reiserfs_insert_item(struct reiserfs_transaction_handle *th, 
 1871                          struct path         *  p_s_path,         /* Path to the inserteded item.         */
 1872                          const struct cpu_key      * key,
 1873                          struct item_head    *  p_s_ih,           /* Pointer to the item header to insert.*/
 1874                          const char          *  p_c_body)         /* Pointer to the bytes to insert.      */
 1875 {
 1876     struct tree_balance s_ins_balance;
 1877     int                 retval;
 1878 
 1879     init_tb_struct(th, &s_ins_balance, th->t_super, p_s_path, IH_SIZE + ih_item_len(p_s_ih));
 1880 #ifdef DISPLACE_NEW_PACKING_LOCALITIES
 1881     s_ins_balance.key = key->on_disk_key;
 1882 #endif
 1883 
 1884     /*
 1885     if (p_c_body == 0)
 1886       n_zeros_num = ih_item_len(p_s_ih);
 1887     */
 1888     //    le_key2cpu_key (&key, &(p_s_ih->ih_key));
 1889 
 1890     while ( (retval = fix_nodes(M_INSERT, &s_ins_balance, p_s_ih, p_c_body)) == REPEAT_SEARCH) {
 1891         /* file system changed while we were in the fix_nodes */
 1892         PROC_INFO_INC( th -> t_super, insert_item_restarted );
 1893         retval = search_item (th->t_super, key, p_s_path);
 1894         if (retval == IO_ERROR) {
 1895             retval = -EIO;
 1896             goto error_out ;
 1897         }
 1898         if (retval == ITEM_FOUND) {
 1899             reiserfs_warning (th->t_super, "PAP-5760: reiserfs_insert_item: "
 1900                               "key %K already exists in the tree\n", key);
 1901             retval = -EEXIST ;
 1902             goto error_out; 
 1903         }
 1904     }
 1905 
 1906     /* make balancing after all resources will be collected at a time */ 
 1907     if ( retval == CARRY_ON ) {
 1908         do_balance (&s_ins_balance, p_s_ih, p_c_body, M_INSERT);
 1909         return 0;
 1910     }
 1911 
 1912     retval = (retval == NO_DISK_SPACE) ? -ENOSPC : -EIO;
 1913 error_out:
 1914     /* also releases the path */
 1915     unfix_nodes(&s_ins_balance);
 1916     return retval; 
 1917 }
 1918 
 1919 
 1920 
 1921 

Cache object: 5cc6f31cc3d2017bae63523035ba5c4b


[ source navigation ] [ diff markup ] [ identifier search ] [ freetext search ] [ file search ] [ list types ] [ track identifier ]


This page is part of the FreeBSD/Linux Linux Kernel Cross-Reference, and was automatically generated using a modified version of the LXR engine.