fs/xfs/xfs_da_btree.c

   1 /*
   2  * Copyright (c) 2000-2005 Silicon Graphics, Inc.
   3  * All Rights Reserved.
   4  *
   5  * This program is free software; you can redistribute it and/or
   6  * modify it under the terms of the GNU General Public License as
   7  * published by the Free Software Foundation.
   8  *
   9  * This program is distributed in the hope that it would be useful,
  10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12  * GNU General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU General Public License
  15  * along with this program; if not, write the Free Software Foundation,
  16  * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
  17  */
  18 #include "xfs.h"
  19 #include "xfs_fs.h"
  20 #include "xfs_types.h"
  21 #include "xfs_bit.h"
  22 #include "xfs_log.h"
  23 #include "xfs_inum.h"
  24 #include "xfs_trans.h"
  25 #include "xfs_sb.h"
  26 #include "xfs_ag.h"
  27 #include "xfs_mount.h"
  28 #include "xfs_da_btree.h"
  29 #include "xfs_bmap_btree.h"
  30 #include "xfs_dir2.h"
  31 #include "xfs_dir2_format.h"
  32 #include "xfs_dir2_priv.h"
  33 #include "xfs_dinode.h"
  34 #include "xfs_inode.h"
  35 #include "xfs_inode_item.h"
  36 #include "xfs_alloc.h"
  37 #include "xfs_bmap.h"
  38 #include "xfs_attr.h"
  39 #include "xfs_attr_leaf.h"
  40 #include "xfs_error.h"
  41 #include "xfs_trace.h"
  42
  43 /*
  44  * xfs_da_btree.c
  45  *
  46  * Routines to implement directories as Btrees of hashed names.
  47  */
  48
  49 /*========================================================================
  50  * Function prototypes for the kernel.
  51  *========================================================================*/
  52
  53 /*
  54  * Routines used for growing the Btree.
  55  */
  56 STATIC int xfs_da_root_split(xfs_da_state_t *state,
  57                                             xfs_da_state_blk_t *existing_root,
  58                                             xfs_da_state_blk_t *new_child);
  59 STATIC int xfs_da_node_split(xfs_da_state_t *state,
  60                                             xfs_da_state_blk_t *existing_blk,
  61                                             xfs_da_state_blk_t *split_blk,
  62                                             xfs_da_state_blk_t *blk_to_add,
  63                                             int treelevel,
  64                                             int *result);
  65 STATIC void xfs_da_node_rebalance(xfs_da_state_t *state,
  66                                          xfs_da_state_blk_t *node_blk_1,
  67                                          xfs_da_state_blk_t *node_blk_2);
  68 STATIC void xfs_da_node_add(xfs_da_state_t *state,
  69                                    xfs_da_state_blk_t *old_node_blk,
  70                                    xfs_da_state_blk_t *new_node_blk);
  71
  72 /*
  73  * Routines used for shrinking the Btree.
  74  */
  75 STATIC int xfs_da_root_join(xfs_da_state_t *state,
  76                                            xfs_da_state_blk_t *root_blk);
  77 STATIC int xfs_da_node_toosmall(xfs_da_state_t *state, int *retval);
  78 STATIC void xfs_da_node_remove(xfs_da_state_t *state,
  79                                               xfs_da_state_blk_t *drop_blk);
  80 STATIC void xfs_da_node_unbalance(xfs_da_state_t *state,
  81                                          xfs_da_state_blk_t *src_node_blk,
  82                                          xfs_da_state_blk_t *dst_node_blk);
  83
  84 /*
  85  * Utility routines.
  86  */
  87 STATIC uint     xfs_da_node_lasthash(xfs_dabuf_t *bp, int *count);
  88 STATIC int      xfs_da_node_order(xfs_dabuf_t *node1_bp, xfs_dabuf_t *node2_bp);
  89 STATIC xfs_dabuf_t *xfs_da_buf_make(int nbuf, xfs_buf_t **bps);
  90 STATIC int      xfs_da_blk_unlink(xfs_da_state_t *state,
  91                                   xfs_da_state_blk_t *drop_blk,
  92                                   xfs_da_state_blk_t *save_blk);
  93 STATIC void     xfs_da_state_kill_altpath(xfs_da_state_t *state);
  94
  95 /*========================================================================
  96  * Routines used for growing the Btree.
  97  *========================================================================*/
  98
  99 /*
 100  * Create the initial contents of an intermediate node.
 101  */
 102 int
 103 xfs_da_node_create(xfs_da_args_t *args, xfs_dablk_t blkno, int level,
 104                                  xfs_dabuf_t **bpp, int whichfork)
 105 {
 106         xfs_da_intnode_t *node;
 107         xfs_dabuf_t *bp;
 108         int error;
 109         xfs_trans_t *tp;
 110
 111         tp = args->trans;
 112         error = xfs_da_get_buf(tp, args->dp, blkno, -1, &bp, whichfork);
 113         if (error)
 114                 return(error);
 115         ASSERT(bp != NULL);
 116         node = bp->data;
 117         node->hdr.info.forw = 0;
 118         node->hdr.info.back = 0;
 119         node->hdr.info.magic = cpu_to_be16(XFS_DA_NODE_MAGIC);
 120         node->hdr.info.pad = 0;
 121         node->hdr.count = 0;
 122         node->hdr.level = cpu_to_be16(level);
 123
 124         xfs_da_log_buf(tp, bp,
 125                 XFS_DA_LOGRANGE(node, &node->hdr, sizeof(node->hdr)));
 126
 127         *bpp = bp;
 128         return(0);
 129 }
 130
 131 /*
 132  * Split a leaf node, rebalance, then possibly split
 133  * intermediate nodes, rebalance, etc.
 134  */
 135 int                                                     /* error */
 136 xfs_da_split(xfs_da_state_t *state)
 137 {
 138         xfs_da_state_blk_t *oldblk, *newblk, *addblk;
 139         xfs_da_intnode_t *node;
 140         xfs_dabuf_t *bp;
 141         int max, action, error, i;
 142
 143         /*
 144          * Walk back up the tree splitting/inserting/adjusting as necessary.
 145          * If we need to insert and there isn't room, split the node, then
 146          * decide which fragment to insert the new block from below into.
 147          * Note that we may split the root this way, but we need more fixup.
 148          */
 149         max = state->path.active - 1;
 150         ASSERT((max >= 0) && (max < XFS_DA_NODE_MAXDEPTH));
 151         ASSERT(state->path.blk[max].magic == XFS_ATTR_LEAF_MAGIC ||
 152                state->path.blk[max].magic == XFS_DIR2_LEAFN_MAGIC);
 153
 154         addblk = &state->path.blk[max];         /* initial dummy value */
 155         for (i = max; (i >= 0) && addblk; state->path.active--, i--) {
 156                 oldblk = &state->path.blk[i];
 157                 newblk = &state->altpath.blk[i];
 158
 159                 /*
 160                  * If a leaf node then
 161                  *     Allocate a new leaf node, then rebalance across them.
 162                  * else if an intermediate node then
 163                  *     We split on the last layer, must we split the node?
 164                  */
 165                 switch (oldblk->magic) {
 166                 case XFS_ATTR_LEAF_MAGIC:
 167                         error = xfs_attr_leaf_split(state, oldblk, newblk);
 168                         if ((error != 0) && (error != ENOSPC)) {
 169                                 return(error);  /* GROT: attr is inconsistent */
 170                         }
 171                         if (!error) {
 172                                 addblk = newblk;
 173                                 break;
 174                         }
 175                         /*
 176                          * Entry wouldn't fit, split the leaf again.
 177                          */
 178                         state->extravalid = 1;
 179                         if (state->inleaf) {
 180                                 state->extraafter = 0;  /* before newblk */
 181                                 error = xfs_attr_leaf_split(state, oldblk,
 182                                                             &state->extrablk);
 183                         } else {
 184                                 state->extraafter = 1;  /* after newblk */
 185                                 error = xfs_attr_leaf_split(state, newblk,
 186                                                             &state->extrablk);
 187                         }
 188                         if (error)
 189                                 return(error);  /* GROT: attr inconsistent */
 190                         addblk = newblk;
 191                         break;
 192                 case XFS_DIR2_LEAFN_MAGIC:
 193                         error = xfs_dir2_leafn_split(state, oldblk, newblk);
 194                         if (error)
 195                                 return error;
 196                         addblk = newblk;
 197                         break;
 198                 case XFS_DA_NODE_MAGIC:
 199                         error = xfs_da_node_split(state, oldblk, newblk, addblk,
 200                                                          max - i, &action);
 201                         xfs_da_buf_done(addblk->bp);
 202                         addblk->bp = NULL;
 203                         if (error)
 204                                 return(error);  /* GROT: dir is inconsistent */
 205                         /*
 206                          * Record the newly split block for the next time thru?
 207                          */
 208                         if (action)
 209                                 addblk = newblk;
 210                         else
 211                                 addblk = NULL;
 212                         break;
 213                 }
 214
 215                 /*
 216                  * Update the btree to show the new hashval for this child.
 217                  */
 218                 xfs_da_fixhashpath(state, &state->path);
 219                 /*
 220                  * If we won't need this block again, it's getting dropped
 221                  * from the active path by the loop control, so we need
 222                  * to mark it done now.
 223                  */
 224                 if (i > 0 || !addblk)
 225                         xfs_da_buf_done(oldblk->bp);
 226         }
 227         if (!addblk)
 228                 return(0);
 229
 230         /*
 231          * Split the root node.
 232          */
 233         ASSERT(state->path.active == 0);
 234         oldblk = &state->path.blk[0];
 235         error = xfs_da_root_split(state, oldblk, addblk);
 236         if (error) {
 237                 xfs_da_buf_done(oldblk->bp);
 238                 xfs_da_buf_done(addblk->bp);
 239                 addblk->bp = NULL;
 240                 return(error);  /* GROT: dir is inconsistent */
 241         }
 242
 243         /*
 244          * Update pointers to the node which used to be block 0 and
 245          * just got bumped because of the addition of a new root node.
 246          * There might be three blocks involved if a double split occurred,
 247          * and the original block 0 could be at any position in the list.
 248          */
 249
 250         node = oldblk->bp->data;
 251         if (node->hdr.info.forw) {
 252                 if (be32_to_cpu(node->hdr.info.forw) == addblk->blkno) {
 253                         bp = addblk->bp;
 254                 } else {
 255                         ASSERT(state->extravalid);
 256                         bp = state->extrablk.bp;
 257                 }
 258                 node = bp->data;
 259                 node->hdr.info.back = cpu_to_be32(oldblk->blkno);
 260                 xfs_da_log_buf(state->args->trans, bp,
 261                     XFS_DA_LOGRANGE(node, &node->hdr.info,
 262                     sizeof(node->hdr.info)));
 263         }
 264         node = oldblk->bp->data;
 265         if (node->hdr.info.back) {
 266                 if (be32_to_cpu(node->hdr.info.back) == addblk->blkno) {
 267                         bp = addblk->bp;
 268                 } else {
 269                         ASSERT(state->extravalid);
 270                         bp = state->extrablk.bp;
 271                 }
 272                 node = bp->data;
 273                 node->hdr.info.forw = cpu_to_be32(oldblk->blkno);
 274                 xfs_da_log_buf(state->args->trans, bp,
 275                     XFS_DA_LOGRANGE(node, &node->hdr.info,
 276                     sizeof(node->hdr.info)));
 277         }
 278         xfs_da_buf_done(oldblk->bp);
 279         xfs_da_buf_done(addblk->bp);
 280         addblk->bp = NULL;
 281         return(0);
 282 }
 283
 284 /*
 285  * Split the root.  We have to create a new root and point to the two
 286  * parts (the split old root) that we just created.  Copy block zero to
 287  * the EOF, extending the inode in process.
 288  */
 289 STATIC int                                              /* error */
 290 xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
 291                                  xfs_da_state_blk_t *blk2)
 292 {
 293         xfs_da_intnode_t *node, *oldroot;
 294         xfs_da_args_t *args;
 295         xfs_dablk_t blkno;
 296         xfs_dabuf_t *bp;
 297         int error, size;
 298         xfs_inode_t *dp;
 299         xfs_trans_t *tp;
 300         xfs_mount_t *mp;
 301         xfs_dir2_leaf_t *leaf;
 302
 303         /*
 304          * Copy the existing (incorrect) block from the root node position
 305          * to a free space somewhere.
 306          */
 307         args = state->args;
 308         ASSERT(args != NULL);
 309         error = xfs_da_grow_inode(args, &blkno);
 310         if (error)
 311                 return(error);
 312         dp = args->dp;
 313         tp = args->trans;
 314         mp = state->mp;
 315         error = xfs_da_get_buf(tp, dp, blkno, -1, &bp, args->whichfork);
 316         if (error)
 317                 return(error);
 318         ASSERT(bp != NULL);
 319         node = bp->data;
 320         oldroot = blk1->bp->data;
 321         if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC)) {
 322                 size = (int)((char *)&oldroot->btree[be16_to_cpu(oldroot->hdr.count)] -
 323                              (char *)oldroot);
 324         } else {
 325                 ASSERT(oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 326                 leaf = (xfs_dir2_leaf_t *)oldroot;
 327                 size = (int)((char *)&leaf->ents[be16_to_cpu(leaf->hdr.count)] -
 328                              (char *)leaf);
 329         }
 330         memcpy(node, oldroot, size);
 331         xfs_da_log_buf(tp, bp, 0, size - 1);
 332         xfs_da_buf_done(blk1->bp);
 333         blk1->bp = bp;
 334         blk1->blkno = blkno;
 335
 336         /*
 337          * Set up the new root node.
 338          */
 339         error = xfs_da_node_create(args,
 340                 (args->whichfork == XFS_DATA_FORK) ? mp->m_dirleafblk : 0,
 341                 be16_to_cpu(node->hdr.level) + 1, &bp, args->whichfork);
 342         if (error)
 343                 return(error);
 344         node = bp->data;
 345         node->btree[0].hashval = cpu_to_be32(blk1->hashval);
 346         node->btree[0].before = cpu_to_be32(blk1->blkno);
 347         node->btree[1].hashval = cpu_to_be32(blk2->hashval);
 348         node->btree[1].before = cpu_to_be32(blk2->blkno);
 349         node->hdr.count = cpu_to_be16(2);
 350
 351 #ifdef DEBUG
 352         if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)) {
 353                 ASSERT(blk1->blkno >= mp->m_dirleafblk &&
 354                        blk1->blkno < mp->m_dirfreeblk);
 355                 ASSERT(blk2->blkno >= mp->m_dirleafblk &&
 356                        blk2->blkno < mp->m_dirfreeblk);
 357         }
 358 #endif
 359
 360         /* Header is already logged by xfs_da_node_create */
 361         xfs_da_log_buf(tp, bp,
 362                 XFS_DA_LOGRANGE(node, node->btree,
 363                         sizeof(xfs_da_node_entry_t) * 2));
 364         xfs_da_buf_done(bp);
 365
 366         return(0);
 367 }
 368
 369 /*
 370  * Split the node, rebalance, then add the new entry.
 371  */
 372 STATIC int                                              /* error */
 373 xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
 374                                  xfs_da_state_blk_t *newblk,
 375                                  xfs_da_state_blk_t *addblk,
 376                                  int treelevel, int *result)
 377 {
 378         xfs_da_intnode_t *node;
 379         xfs_dablk_t blkno;
 380         int newcount, error;
 381         int useextra;
 382
 383         node = oldblk->bp->data;
 384         ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 385
 386         /*
 387          * With V2 dirs the extra block is data or freespace.
 388          */
 389         useextra = state->extravalid && state->args->whichfork == XFS_ATTR_FORK;
 390         newcount = 1 + useextra;
 391         /*
 392          * Do we have to split the node?
 393          */
 394         if ((be16_to_cpu(node->hdr.count) + newcount) > state->node_ents) {
 395                 /*
 396                  * Allocate a new node, add to the doubly linked chain of
 397                  * nodes, then move some of our excess entries into it.
 398                  */
 399                 error = xfs_da_grow_inode(state->args, &blkno);
 400                 if (error)
 401                         return(error);  /* GROT: dir is inconsistent */
 402
 403                 error = xfs_da_node_create(state->args, blkno, treelevel,
 404                                            &newblk->bp, state->args->whichfork);
 405                 if (error)
 406                         return(error);  /* GROT: dir is inconsistent */
 407                 newblk->blkno = blkno;
 408                 newblk->magic = XFS_DA_NODE_MAGIC;
 409                 xfs_da_node_rebalance(state, oldblk, newblk);
 410                 error = xfs_da_blk_link(state, oldblk, newblk);
 411                 if (error)
 412                         return(error);
 413                 *result = 1;
 414         } else {
 415                 *result = 0;
 416         }
 417
 418         /*
 419          * Insert the new entry(s) into the correct block
 420          * (updating last hashval in the process).
 421          *
 422          * xfs_da_node_add() inserts BEFORE the given index,
 423          * and as a result of using node_lookup_int() we always
 424          * point to a valid entry (not after one), but a split
 425          * operation always results in a new block whose hashvals
 426          * FOLLOW the current block.
 427          *
 428          * If we had double-split op below us, then add the extra block too.
 429          */
 430         node = oldblk->bp->data;
 431         if (oldblk->index <= be16_to_cpu(node->hdr.count)) {
 432                 oldblk->index++;
 433                 xfs_da_node_add(state, oldblk, addblk);
 434                 if (useextra) {
 435                         if (state->extraafter)
 436                                 oldblk->index++;
 437                         xfs_da_node_add(state, oldblk, &state->extrablk);
 438                         state->extravalid = 0;
 439                 }
 440         } else {
 441                 newblk->index++;
 442                 xfs_da_node_add(state, newblk, addblk);
 443                 if (useextra) {
 444                         if (state->extraafter)
 445                                 newblk->index++;
 446                         xfs_da_node_add(state, newblk, &state->extrablk);
 447                         state->extravalid = 0;
 448                 }
 449         }
 450
 451         return(0);
 452 }
 453
 454 /*
 455  * Balance the btree elements between two intermediate nodes,
 456  * usually one full and one empty.
 457  *
 458  * NOTE: if blk2 is empty, then it will get the upper half of blk1.
 459  */
 460 STATIC void
 461 xfs_da_node_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
 462                                      xfs_da_state_blk_t *blk2)
 463 {
 464         xfs_da_intnode_t *node1, *node2, *tmpnode;
 465         xfs_da_node_entry_t *btree_s, *btree_d;
 466         int count, tmp;
 467         xfs_trans_t *tp;
 468
 469         node1 = blk1->bp->data;
 470         node2 = blk2->bp->data;
 471         /*
 472          * Figure out how many entries need to move, and in which direction.
 473          * Swap the nodes around if that makes it simpler.
 474          */
 475         if ((be16_to_cpu(node1->hdr.count) > 0) && (be16_to_cpu(node2->hdr.count) > 0) &&
 476             ((be32_to_cpu(node2->btree[0].hashval) < be32_to_cpu(node1->btree[0].hashval)) ||
 477              (be32_to_cpu(node2->btree[be16_to_cpu(node2->hdr.count)-1].hashval) <
 478               be32_to_cpu(node1->btree[be16_to_cpu(node1->hdr.count)-1].hashval)))) {
 479                 tmpnode = node1;
 480                 node1 = node2;
 481                 node2 = tmpnode;
 482         }
 483         ASSERT(node1->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 484         ASSERT(node2->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 485         count = (be16_to_cpu(node1->hdr.count) - be16_to_cpu(node2->hdr.count)) / 2;
 486         if (count == 0)
 487                 return;
 488         tp = state->args->trans;
 489         /*
 490          * Two cases: high-to-low and low-to-high.
 491          */
 492         if (count > 0) {
 493                 /*
 494                  * Move elements in node2 up to make a hole.
 495                  */
 496                 if ((tmp = be16_to_cpu(node2->hdr.count)) > 0) {
 497                         tmp *= (uint)sizeof(xfs_da_node_entry_t);
 498                         btree_s = &node2->btree[0];
 499                         btree_d = &node2->btree[count];
 500                         memmove(btree_d, btree_s, tmp);
 501                 }
 502
 503                 /*
 504                  * Move the req'd B-tree elements from high in node1 to
 505                  * low in node2.
 506                  */
 507                 be16_add_cpu(&node2->hdr.count, count);
 508                 tmp = count * (uint)sizeof(xfs_da_node_entry_t);
 509                 btree_s = &node1->btree[be16_to_cpu(node1->hdr.count) - count];
 510                 btree_d = &node2->btree[0];
 511                 memcpy(btree_d, btree_s, tmp);
 512                 be16_add_cpu(&node1->hdr.count, -count);
 513         } else {
 514                 /*
 515                  * Move the req'd B-tree elements from low in node2 to
 516                  * high in node1.
 517                  */
 518                 count = -count;
 519                 tmp = count * (uint)sizeof(xfs_da_node_entry_t);
 520                 btree_s = &node2->btree[0];
 521                 btree_d = &node1->btree[be16_to_cpu(node1->hdr.count)];
 522                 memcpy(btree_d, btree_s, tmp);
 523                 be16_add_cpu(&node1->hdr.count, count);
 524                 xfs_da_log_buf(tp, blk1->bp,
 525                         XFS_DA_LOGRANGE(node1, btree_d, tmp));
 526
 527                 /*
 528                  * Move elements in node2 down to fill the hole.
 529                  */
 530                 tmp  = be16_to_cpu(node2->hdr.count) - count;
 531                 tmp *= (uint)sizeof(xfs_da_node_entry_t);
 532                 btree_s = &node2->btree[count];
 533                 btree_d = &node2->btree[0];
 534                 memmove(btree_d, btree_s, tmp);
 535                 be16_add_cpu(&node2->hdr.count, -count);
 536         }
 537
 538         /*
 539          * Log header of node 1 and all current bits of node 2.
 540          */
 541         xfs_da_log_buf(tp, blk1->bp,
 542                 XFS_DA_LOGRANGE(node1, &node1->hdr, sizeof(node1->hdr)));
 543         xfs_da_log_buf(tp, blk2->bp,
 544                 XFS_DA_LOGRANGE(node2, &node2->hdr,
 545                         sizeof(node2->hdr) +
 546                         sizeof(node2->btree[0]) * be16_to_cpu(node2->hdr.count)));
 547
 548         /*
 549          * Record the last hashval from each block for upward propagation.
 550          * (note: don't use the swapped node pointers)
 551          */
 552         node1 = blk1->bp->data;
 553         node2 = blk2->bp->data;
 554         blk1->hashval = be32_to_cpu(node1->btree[be16_to_cpu(node1->hdr.count)-1].hashval);
 555         blk2->hashval = be32_to_cpu(node2->btree[be16_to_cpu(node2->hdr.count)-1].hashval);
 556
 557         /*
 558          * Adjust the expected index for insertion.
 559          */
 560         if (blk1->index >= be16_to_cpu(node1->hdr.count)) {
 561                 blk2->index = blk1->index - be16_to_cpu(node1->hdr.count);
 562                 blk1->index = be16_to_cpu(node1->hdr.count) + 1;        /* make it invalid */
 563         }
 564 }
 565
 566 /*
 567  * Add a new entry to an intermediate node.
 568  */
 569 STATIC void
 570 xfs_da_node_add(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
 571                                xfs_da_state_blk_t *newblk)
 572 {
 573         xfs_da_intnode_t *node;
 574         xfs_da_node_entry_t *btree;
 575         int tmp;
 576
 577         node = oldblk->bp->data;
 578         ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 579         ASSERT((oldblk->index >= 0) && (oldblk->index <= be16_to_cpu(node->hdr.count)));
 580         ASSERT(newblk->blkno != 0);
 581         if (state->args->whichfork == XFS_DATA_FORK)
 582                 ASSERT(newblk->blkno >= state->mp->m_dirleafblk &&
 583                        newblk->blkno < state->mp->m_dirfreeblk);
 584
 585         /*
 586          * We may need to make some room before we insert the new node.
 587          */
 588         tmp = 0;
 589         btree = &node->btree[ oldblk->index ];
 590         if (oldblk->index < be16_to_cpu(node->hdr.count)) {
 591                 tmp = (be16_to_cpu(node->hdr.count) - oldblk->index) * (uint)sizeof(*btree);
 592                 memmove(btree + 1, btree, tmp);
 593         }
 594         btree->hashval = cpu_to_be32(newblk->hashval);
 595         btree->before = cpu_to_be32(newblk->blkno);
 596         xfs_da_log_buf(state->args->trans, oldblk->bp,
 597                 XFS_DA_LOGRANGE(node, btree, tmp + sizeof(*btree)));
 598         be16_add_cpu(&node->hdr.count, 1);
 599         xfs_da_log_buf(state->args->trans, oldblk->bp,
 600                 XFS_DA_LOGRANGE(node, &node->hdr, sizeof(node->hdr)));
 601
 602         /*
 603          * Copy the last hash value from the oldblk to propagate upwards.
 604          */
 605         oldblk->hashval = be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1 ].hashval);
 606 }
 607
 608 /*========================================================================
 609  * Routines used for shrinking the Btree.
 610  *========================================================================*/
 611
 612 /*
 613  * Deallocate an empty leaf node, remove it from its parent,
 614  * possibly deallocating that block, etc...
 615  */
 616 int
 617 xfs_da_join(xfs_da_state_t *state)
 618 {
 619         xfs_da_state_blk_t *drop_blk, *save_blk;
 620         int action, error;
 621
 622         action = 0;
 623         drop_blk = &state->path.blk[ state->path.active-1 ];
 624         save_blk = &state->altpath.blk[ state->path.active-1 ];
 625         ASSERT(state->path.blk[0].magic == XFS_DA_NODE_MAGIC);
 626         ASSERT(drop_blk->magic == XFS_ATTR_LEAF_MAGIC ||
 627                drop_blk->magic == XFS_DIR2_LEAFN_MAGIC);
 628
 629         /*
 630          * Walk back up the tree joining/deallocating as necessary.
 631          * When we stop dropping blocks, break out.
 632          */
 633         for (  ; state->path.active >= 2; drop_blk--, save_blk--,
 634                  state->path.active--) {
 635                 /*
 636                  * See if we can combine the block with a neighbor.
 637                  *   (action == 0) => no options, just leave
 638                  *   (action == 1) => coalesce, then unlink
 639                  *   (action == 2) => block empty, unlink it
 640                  */
 641                 switch (drop_blk->magic) {
 642                 case XFS_ATTR_LEAF_MAGIC:
 643                         error = xfs_attr_leaf_toosmall(state, &action);
 644                         if (error)
 645                                 return(error);
 646                         if (action == 0)
 647                                 return(0);
 648                         xfs_attr_leaf_unbalance(state, drop_blk, save_blk);
 649                         break;
 650                 case XFS_DIR2_LEAFN_MAGIC:
 651                         error = xfs_dir2_leafn_toosmall(state, &action);
 652                         if (error)
 653                                 return error;
 654                         if (action == 0)
 655                                 return 0;
 656                         xfs_dir2_leafn_unbalance(state, drop_blk, save_blk);
 657                         break;
 658                 case XFS_DA_NODE_MAGIC:
 659                         /*
 660                          * Remove the offending node, fixup hashvals,
 661                          * check for a toosmall neighbor.
 662                          */
 663                         xfs_da_node_remove(state, drop_blk);
 664                         xfs_da_fixhashpath(state, &state->path);
 665                         error = xfs_da_node_toosmall(state, &action);
 666                         if (error)
 667                                 return(error);
 668                         if (action == 0)
 669                                 return 0;
 670                         xfs_da_node_unbalance(state, drop_blk, save_blk);
 671                         break;
 672                 }
 673                 xfs_da_fixhashpath(state, &state->altpath);
 674                 error = xfs_da_blk_unlink(state, drop_blk, save_blk);
 675                 xfs_da_state_kill_altpath(state);
 676                 if (error)
 677                         return(error);
 678                 error = xfs_da_shrink_inode(state->args, drop_blk->blkno,
 679                                                          drop_blk->bp);
 680                 drop_blk->bp = NULL;
 681                 if (error)
 682                         return(error);
 683         }
 684         /*
 685          * We joined all the way to the top.  If it turns out that
 686          * we only have one entry in the root, make the child block
 687          * the new root.
 688          */
 689         xfs_da_node_remove(state, drop_blk);
 690         xfs_da_fixhashpath(state, &state->path);
 691         error = xfs_da_root_join(state, &state->path.blk[0]);
 692         return(error);
 693 }
 694
 695 /*
 696  * We have only one entry in the root.  Copy the only remaining child of
 697  * the old root to block 0 as the new root node.
 698  */
 699 STATIC int
 700 xfs_da_root_join(xfs_da_state_t *state, xfs_da_state_blk_t *root_blk)
 701 {
 702         xfs_da_intnode_t *oldroot;
 703         /* REFERENCED */
 704         xfs_da_blkinfo_t *blkinfo;
 705         xfs_da_args_t *args;
 706         xfs_dablk_t child;
 707         xfs_dabuf_t *bp;
 708         int error;
 709
 710         args = state->args;
 711         ASSERT(args != NULL);
 712         ASSERT(root_blk->magic == XFS_DA_NODE_MAGIC);
 713         oldroot = root_blk->bp->data;
 714         ASSERT(oldroot->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 715         ASSERT(!oldroot->hdr.info.forw);
 716         ASSERT(!oldroot->hdr.info.back);
 717
 718         /*
 719          * If the root has more than one child, then don't do anything.
 720          */
 721         if (be16_to_cpu(oldroot->hdr.count) > 1)
 722                 return(0);
 723
 724         /*
 725          * Read in the (only) child block, then copy those bytes into
 726          * the root block's buffer and free the original child block.
 727          */
 728         child = be32_to_cpu(oldroot->btree[0].before);
 729         ASSERT(child != 0);
 730         error = xfs_da_read_buf(args->trans, args->dp, child, -1, &bp,
 731                                              args->whichfork);
 732         if (error)
 733                 return(error);
 734         ASSERT(bp != NULL);
 735         blkinfo = bp->data;
 736         if (be16_to_cpu(oldroot->hdr.level) == 1) {
 737                 ASSERT(blkinfo->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
 738                        blkinfo->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 739         } else {
 740                 ASSERT(blkinfo->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 741         }
 742         ASSERT(!blkinfo->forw);
 743         ASSERT(!blkinfo->back);
 744         memcpy(root_blk->bp->data, bp->data, state->blocksize);
 745         xfs_da_log_buf(args->trans, root_blk->bp, 0, state->blocksize - 1);
 746         error = xfs_da_shrink_inode(args, child, bp);
 747         return(error);
 748 }
 749
 750 /*
 751  * Check a node block and its neighbors to see if the block should be
 752  * collapsed into one or the other neighbor.  Always keep the block
 753  * with the smaller block number.
 754  * If the current block is over 50% full, don't try to join it, return 0.
 755  * If the block is empty, fill in the state structure and return 2.
 756  * If it can be collapsed, fill in the state structure and return 1.
 757  * If nothing can be done, return 0.
 758  */
 759 STATIC int
 760 xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
 761 {
 762         xfs_da_intnode_t *node;
 763         xfs_da_state_blk_t *blk;
 764         xfs_da_blkinfo_t *info;
 765         int count, forward, error, retval, i;
 766         xfs_dablk_t blkno;
 767         xfs_dabuf_t *bp;
 768
 769         /*
 770          * Check for the degenerate case of the block being over 50% full.
 771          * If so, it's not worth even looking to see if we might be able
 772          * to coalesce with a sibling.
 773          */
 774         blk = &state->path.blk[ state->path.active-1 ];
 775         info = blk->bp->data;
 776         ASSERT(info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 777         node = (xfs_da_intnode_t *)info;
 778         count = be16_to_cpu(node->hdr.count);
 779         if (count > (state->node_ents >> 1)) {
 780                 *action = 0;    /* blk over 50%, don't try to join */
 781                 return(0);      /* blk over 50%, don't try to join */
 782         }
 783
 784         /*
 785          * Check for the degenerate case of the block being empty.
 786          * If the block is empty, we'll simply delete it, no need to
 787          * coalesce it with a sibling block.  We choose (arbitrarily)
 788          * to merge with the forward block unless it is NULL.
 789          */
 790         if (count == 0) {
 791                 /*
 792                  * Make altpath point to the block we want to keep and
 793                  * path point to the block we want to drop (this one).
 794                  */
 795                 forward = (info->forw != 0);
 796                 memcpy(&state->altpath, &state->path, sizeof(state->path));
 797                 error = xfs_da_path_shift(state, &state->altpath, forward,
 798                                                  0, &retval);
 799                 if (error)
 800                         return(error);
 801                 if (retval) {
 802                         *action = 0;
 803                 } else {
 804                         *action = 2;
 805                 }
 806                 return(0);
 807         }
 808
 809         /*
 810          * Examine each sibling block to see if we can coalesce with
 811          * at least 25% free space to spare.  We need to figure out
 812          * whether to merge with the forward or the backward block.
 813          * We prefer coalescing with the lower numbered sibling so as
 814          * to shrink a directory over time.
 815          */
 816         /* start with smaller blk num */
 817         forward = (be32_to_cpu(info->forw) < be32_to_cpu(info->back));
 818         for (i = 0; i < 2; forward = !forward, i++) {
 819                 if (forward)
 820                         blkno = be32_to_cpu(info->forw);
 821                 else
 822                         blkno = be32_to_cpu(info->back);
 823                 if (blkno == 0)
 824                         continue;
 825                 error = xfs_da_read_buf(state->args->trans, state->args->dp,
 826                                         blkno, -1, &bp, state->args->whichfork);
 827                 if (error)
 828                         return(error);
 829                 ASSERT(bp != NULL);
 830
 831                 node = (xfs_da_intnode_t *)info;
 832                 count  = state->node_ents;
 833                 count -= state->node_ents >> 2;
 834                 count -= be16_to_cpu(node->hdr.count);
 835                 node = bp->data;
 836                 ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 837                 count -= be16_to_cpu(node->hdr.count);
 838                 xfs_da_brelse(state->args->trans, bp);
 839                 if (count >= 0)
 840                         break;  /* fits with at least 25% to spare */
 841         }
 842         if (i >= 2) {
 843                 *action = 0;
 844                 return(0);
 845         }
 846
 847         /*
 848          * Make altpath point to the block we want to keep (the lower
 849          * numbered block) and path point to the block we want to drop.
 850          */
 851         memcpy(&state->altpath, &state->path, sizeof(state->path));
 852         if (blkno < blk->blkno) {
 853                 error = xfs_da_path_shift(state, &state->altpath, forward,
 854                                                  0, &retval);
 855                 if (error) {
 856                         return(error);
 857                 }
 858                 if (retval) {
 859                         *action = 0;
 860                         return(0);
 861                 }
 862         } else {
 863                 error = xfs_da_path_shift(state, &state->path, forward,
 864                                                  0, &retval);
 865                 if (error) {
 866                         return(error);
 867                 }
 868                 if (retval) {
 869                         *action = 0;
 870                         return(0);
 871                 }
 872         }
 873         *action = 1;
 874         return(0);
 875 }
 876
 877 /*
 878  * Walk back up the tree adjusting hash values as necessary,
 879  * when we stop making changes, return.
 880  */
 881 void
 882 xfs_da_fixhashpath(xfs_da_state_t *state, xfs_da_state_path_t *path)
 883 {
 884         xfs_da_state_blk_t *blk;
 885         xfs_da_intnode_t *node;
 886         xfs_da_node_entry_t *btree;
 887         xfs_dahash_t lasthash=0;
 888         int level, count;
 889
 890         level = path->active-1;
 891         blk = &path->blk[ level ];
 892         switch (blk->magic) {
 893         case XFS_ATTR_LEAF_MAGIC:
 894                 lasthash = xfs_attr_leaf_lasthash(blk->bp, &count);
 895                 if (count == 0)
 896                         return;
 897                 break;
 898         case XFS_DIR2_LEAFN_MAGIC:
 899                 lasthash = xfs_dir2_leafn_lasthash(blk->bp, &count);
 900                 if (count == 0)
 901                         return;
 902                 break;
 903         case XFS_DA_NODE_MAGIC:
 904                 lasthash = xfs_da_node_lasthash(blk->bp, &count);
 905                 if (count == 0)
 906                         return;
 907                 break;
 908         }
 909         for (blk--, level--; level >= 0; blk--, level--) {
 910                 node = blk->bp->data;
 911                 ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 912                 btree = &node->btree[ blk->index ];
 913                 if (be32_to_cpu(btree->hashval) == lasthash)
 914                         break;
 915                 blk->hashval = lasthash;
 916                 btree->hashval = cpu_to_be32(lasthash);
 917                 xfs_da_log_buf(state->args->trans, blk->bp,
 918                                   XFS_DA_LOGRANGE(node, btree, sizeof(*btree)));
 919
 920                 lasthash = be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1].hashval);
 921         }
 922 }
 923
 924 /*
 925  * Remove an entry from an intermediate node.
 926  */
 927 STATIC void
 928 xfs_da_node_remove(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk)
 929 {
 930         xfs_da_intnode_t *node;
 931         xfs_da_node_entry_t *btree;
 932         int tmp;
 933
 934         node = drop_blk->bp->data;
 935         ASSERT(drop_blk->index < be16_to_cpu(node->hdr.count));
 936         ASSERT(drop_blk->index >= 0);
 937
 938         /*
 939          * Copy over the offending entry, or just zero it out.
 940          */
 941         btree = &node->btree[drop_blk->index];
 942         if (drop_blk->index < (be16_to_cpu(node->hdr.count)-1)) {
 943                 tmp  = be16_to_cpu(node->hdr.count) - drop_blk->index - 1;
 944                 tmp *= (uint)sizeof(xfs_da_node_entry_t);
 945                 memmove(btree, btree + 1, tmp);
 946                 xfs_da_log_buf(state->args->trans, drop_blk->bp,
 947                     XFS_DA_LOGRANGE(node, btree, tmp));
 948                 btree = &node->btree[be16_to_cpu(node->hdr.count)-1];
 949         }
 950         memset((char *)btree, 0, sizeof(xfs_da_node_entry_t));
 951         xfs_da_log_buf(state->args->trans, drop_blk->bp,
 952             XFS_DA_LOGRANGE(node, btree, sizeof(*btree)));
 953         be16_add_cpu(&node->hdr.count, -1);
 954         xfs_da_log_buf(state->args->trans, drop_blk->bp,
 955             XFS_DA_LOGRANGE(node, &node->hdr, sizeof(node->hdr)));
 956
 957         /*
 958          * Copy the last hash value from the block to propagate upwards.
 959          */
 960         btree--;
 961         drop_blk->hashval = be32_to_cpu(btree->hashval);
 962 }
 963
 964 /*
 965  * Unbalance the btree elements between two intermediate nodes,
 966  * move all Btree elements from one node into another.
 967  */
 968 STATIC void
 969 xfs_da_node_unbalance(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
 970                                      xfs_da_state_blk_t *save_blk)
 971 {
 972         xfs_da_intnode_t *drop_node, *save_node;
 973         xfs_da_node_entry_t *btree;
 974         int tmp;
 975         xfs_trans_t *tp;
 976
 977         drop_node = drop_blk->bp->data;
 978         save_node = save_blk->bp->data;
 979         ASSERT(drop_node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 980         ASSERT(save_node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 981         tp = state->args->trans;
 982
 983         /*
 984          * If the dying block has lower hashvals, then move all the
 985          * elements in the remaining block up to make a hole.
 986          */
 987         if ((be32_to_cpu(drop_node->btree[0].hashval) < be32_to_cpu(save_node->btree[ 0 ].hashval)) ||
 988             (be32_to_cpu(drop_node->btree[be16_to_cpu(drop_node->hdr.count)-1].hashval) <
 989              be32_to_cpu(save_node->btree[be16_to_cpu(save_node->hdr.count)-1].hashval)))
 990         {
 991                 btree = &save_node->btree[be16_to_cpu(drop_node->hdr.count)];
 992                 tmp = be16_to_cpu(save_node->hdr.count) * (uint)sizeof(xfs_da_node_entry_t);
 993                 memmove(btree, &save_node->btree[0], tmp);
 994                 btree = &save_node->btree[0];
 995                 xfs_da_log_buf(tp, save_blk->bp,
 996                         XFS_DA_LOGRANGE(save_node, btree,
 997                                 (be16_to_cpu(save_node->hdr.count) + be16_to_cpu(drop_node->hdr.count)) *
 998                                 sizeof(xfs_da_node_entry_t)));
 999         } else {
1000                 btree = &save_node->btree[be16_to_cpu(save_node->hdr.count)];
1001                 xfs_da_log_buf(tp, save_blk->bp,
1002                         XFS_DA_LOGRANGE(save_node, btree,
1003                                 be16_to_cpu(drop_node->hdr.count) *
1004                                 sizeof(xfs_da_node_entry_t)));
1005         }
1006
1007         /*
1008          * Move all the B-tree elements from drop_blk to save_blk.
1009          */
1010         tmp = be16_to_cpu(drop_node->hdr.count) * (uint)sizeof(xfs_da_node_entry_t);
1011         memcpy(btree, &drop_node->btree[0], tmp);
1012         be16_add_cpu(&save_node->hdr.count, be16_to_cpu(drop_node->hdr.count));
1013
1014         xfs_da_log_buf(tp, save_blk->bp,
1015                 XFS_DA_LOGRANGE(save_node, &save_node->hdr,
1016                         sizeof(save_node->hdr)));
1017
1018         /*
1019          * Save the last hashval in the remaining block for upward propagation.
1020          */
1021         save_blk->hashval = be32_to_cpu(save_node->btree[be16_to_cpu(save_node->hdr.count)-1].hashval);
1022 }
1023
1024 /*========================================================================
1025  * Routines used for finding things in the Btree.
1026  *========================================================================*/
1027
1028 /*
1029  * Walk down the Btree looking for a particular filename, filling
1030  * in the state structure as we go.
1031  *
1032  * We will set the state structure to point to each of the elements
1033  * in each of the nodes where either the hashval is or should be.
1034  *
1035  * We support duplicate hashval's so for each entry in the current
1036  * node that could contain the desired hashval, descend.  This is a
1037  * pruned depth-first tree search.
1038  */
1039 int                                                     /* error */
1040 xfs_da_node_lookup_int(xfs_da_state_t *state, int *result)
1041 {
1042         xfs_da_state_blk_t *blk;
1043         xfs_da_blkinfo_t *curr;
1044         xfs_da_intnode_t *node;
1045         xfs_da_node_entry_t *btree;
1046         xfs_dablk_t blkno;
1047         int probe, span, max, error, retval;
1048         xfs_dahash_t hashval, btreehashval;
1049         xfs_da_args_t *args;
1050
1051         args = state->args;
1052
1053         /*
1054          * Descend thru the B-tree searching each level for the right
1055          * node to use, until the right hashval is found.
1056          */
1057         blkno = (args->whichfork == XFS_DATA_FORK)? state->mp->m_dirleafblk : 0;
1058         for (blk = &state->path.blk[0], state->path.active = 1;
1059                          state->path.active <= XFS_DA_NODE_MAXDEPTH;
1060                          blk++, state->path.active++) {
1061                 /*
1062                  * Read the next node down in the tree.
1063                  */
1064                 blk->blkno = blkno;
1065                 error = xfs_da_read_buf(args->trans, args->dp, blkno,
1066                                         -1, &blk->bp, args->whichfork);
1067                 if (error) {
1068                         blk->blkno = 0;
1069                         state->path.active--;
1070                         return(error);
1071                 }
1072                 curr = blk->bp->data;
1073                 blk->magic = be16_to_cpu(curr->magic);
1074                 ASSERT(blk->magic == XFS_DA_NODE_MAGIC ||
1075                        blk->magic == XFS_DIR2_LEAFN_MAGIC ||
1076                        blk->magic == XFS_ATTR_LEAF_MAGIC);
1077
1078                 /*
1079                  * Search an intermediate node for a match.
1080                  */
1081                 if (blk->magic == XFS_DA_NODE_MAGIC) {
1082                         node = blk->bp->data;
1083                         max = be16_to_cpu(node->hdr.count);
1084                         blk->hashval = be32_to_cpu(node->btree[max-1].hashval);
1085
1086                         /*
1087                          * Binary search.  (note: small blocks will skip loop)
1088                          */
1089                         probe = span = max / 2;
1090                         hashval = args->hashval;
1091                         for (btree = &node->btree[probe]; span > 4;
1092                                    btree = &node->btree[probe]) {
1093                                 span /= 2;
1094                                 btreehashval = be32_to_cpu(btree->hashval);
1095                                 if (btreehashval < hashval)
1096                                         probe += span;
1097                                 else if (btreehashval > hashval)
1098                                         probe -= span;
1099                                 else
1100                                         break;
1101                         }
1102                         ASSERT((probe >= 0) && (probe < max));
1103                         ASSERT((span <= 4) || (be32_to_cpu(btree->hashval) == hashval));
1104
1105                         /*
1106                          * Since we may have duplicate hashval's, find the first
1107                          * matching hashval in the node.
1108                          */
1109                         while ((probe > 0) && (be32_to_cpu(btree->hashval) >= hashval)) {
1110                                 btree--;
1111                                 probe--;
1112                         }
1113                         while ((probe < max) && (be32_to_cpu(btree->hashval) < hashval)) {
1114                                 btree++;
1115                                 probe++;
1116                         }
1117
1118                         /*
1119                          * Pick the right block to descend on.
1120                          */
1121                         if (probe == max) {
1122                                 blk->index = max-1;
1123                                 blkno = be32_to_cpu(node->btree[max-1].before);
1124                         } else {
1125                                 blk->index = probe;
1126                                 blkno = be32_to_cpu(btree->before);
1127                         }
1128                 } else if (blk->magic == XFS_ATTR_LEAF_MAGIC) {
1129                         blk->hashval = xfs_attr_leaf_lasthash(blk->bp, NULL);
1130                         break;
1131                 } else if (blk->magic == XFS_DIR2_LEAFN_MAGIC) {
1132                         blk->hashval = xfs_dir2_leafn_lasthash(blk->bp, NULL);
1133                         break;
1134                 }
1135         }
1136
1137         /*
1138          * A leaf block that ends in the hashval that we are interested in
1139          * (final hashval == search hashval) means that the next block may
1140          * contain more entries with the same hashval, shift upward to the
1141          * next leaf and keep searching.
1142          */
1143         for (;;) {
1144                 if (blk->magic == XFS_DIR2_LEAFN_MAGIC) {
1145                         retval = xfs_dir2_leafn_lookup_int(blk->bp, args,
1146                                                         &blk->index, state);
1147                 } else if (blk->magic == XFS_ATTR_LEAF_MAGIC) {
1148                         retval = xfs_attr_leaf_lookup_int(blk->bp, args);
1149                         blk->index = args->index;
1150                         args->blkno = blk->blkno;
1151                 } else {
1152                         ASSERT(0);
1153                         return XFS_ERROR(EFSCORRUPTED);
1154                 }
1155                 if (((retval == ENOENT) || (retval == ENOATTR)) &&
1156                     (blk->hashval == args->hashval)) {
1157                         error = xfs_da_path_shift(state, &state->path, 1, 1,
1158                                                          &retval);
1159                         if (error)
1160                                 return(error);
1161                         if (retval == 0) {
1162                                 continue;
1163                         } else if (blk->magic == XFS_ATTR_LEAF_MAGIC) {
1164                                 /* path_shift() gives ENOENT */
1165                                 retval = XFS_ERROR(ENOATTR);
1166                         }
1167                 }
1168                 break;
1169         }
1170         *result = retval;
1171         return(0);
1172 }
1173
1174 /*========================================================================
1175  * Utility routines.
1176  *========================================================================*/
1177
1178 /*
1179  * Link a new block into a doubly linked list of blocks (of whatever type).
1180  */
1181 int                                                     /* error */
1182 xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
1183                                xfs_da_state_blk_t *new_blk)
1184 {
1185         xfs_da_blkinfo_t *old_info, *new_info, *tmp_info;
1186         xfs_da_args_t *args;
1187         int before=0, error;
1188         xfs_dabuf_t *bp;
1189
1190         /*
1191          * Set up environment.
1192          */
1193         args = state->args;
1194         ASSERT(args != NULL);
1195         old_info = old_blk->bp->data;
1196         new_info = new_blk->bp->data;
1197         ASSERT(old_blk->magic == XFS_DA_NODE_MAGIC ||
1198                old_blk->magic == XFS_DIR2_LEAFN_MAGIC ||
1199                old_blk->magic == XFS_ATTR_LEAF_MAGIC);
1200         ASSERT(old_blk->magic == be16_to_cpu(old_info->magic));
1201         ASSERT(new_blk->magic == be16_to_cpu(new_info->magic));
1202         ASSERT(old_blk->magic == new_blk->magic);
1203
1204         switch (old_blk->magic) {
1205         case XFS_ATTR_LEAF_MAGIC:
1206                 before = xfs_attr_leaf_order(old_blk->bp, new_blk->bp);
1207                 break;
1208         case XFS_DIR2_LEAFN_MAGIC:
1209                 before = xfs_dir2_leafn_order(old_blk->bp, new_blk->bp);
1210                 break;
1211         case XFS_DA_NODE_MAGIC:
1212                 before = xfs_da_node_order(old_blk->bp, new_blk->bp);
1213                 break;
1214         }
1215
1216         /*
1217          * Link blocks in appropriate order.
1218          */
1219         if (before) {
1220                 /*
1221                  * Link new block in before existing block.
1222                  */
1223                 new_info->forw = cpu_to_be32(old_blk->blkno);
1224                 new_info->back = old_info->back;
1225                 if (old_info->back) {
1226                         error = xfs_da_read_buf(args->trans, args->dp,
1227                                                 be32_to_cpu(old_info->back),
1228                                                 -1, &bp, args->whichfork);
1229                         if (error)
1230                                 return(error);
1231                         ASSERT(bp != NULL);
1232                         tmp_info = bp->data;
1233                         ASSERT(be16_to_cpu(tmp_info->magic) == be16_to_cpu(old_info->magic));
1234                         ASSERT(be32_to_cpu(tmp_info->forw) == old_blk->blkno);
1235                         tmp_info->forw = cpu_to_be32(new_blk->blkno);
1236                         xfs_da_log_buf(args->trans, bp, 0, sizeof(*tmp_info)-1);
1237                         xfs_da_buf_done(bp);
1238                 }
1239                 old_info->back = cpu_to_be32(new_blk->blkno);
1240         } else {
1241                 /*
1242                  * Link new block in after existing block.
1243                  */
1244                 new_info->forw = old_info->forw;
1245                 new_info->back = cpu_to_be32(old_blk->blkno);
1246                 if (old_info->forw) {
1247                         error = xfs_da_read_buf(args->trans, args->dp,
1248                                                 be32_to_cpu(old_info->forw),
1249                                                 -1, &bp, args->whichfork);
1250                         if (error)
1251                                 return(error);
1252                         ASSERT(bp != NULL);
1253                         tmp_info = bp->data;
1254                         ASSERT(tmp_info->magic == old_info->magic);
1255                         ASSERT(be32_to_cpu(tmp_info->back) == old_blk->blkno);
1256                         tmp_info->back = cpu_to_be32(new_blk->blkno);
1257                         xfs_da_log_buf(args->trans, bp, 0, sizeof(*tmp_info)-1);
1258                         xfs_da_buf_done(bp);
1259                 }
1260                 old_info->forw = cpu_to_be32(new_blk->blkno);
1261         }
1262
1263         xfs_da_log_buf(args->trans, old_blk->bp, 0, sizeof(*tmp_info) - 1);
1264         xfs_da_log_buf(args->trans, new_blk->bp, 0, sizeof(*tmp_info) - 1);
1265         return(0);
1266 }
1267
1268 /*
1269  * Compare two intermediate nodes for "order".
1270  */
1271 STATIC int
1272 xfs_da_node_order(xfs_dabuf_t *node1_bp, xfs_dabuf_t *node2_bp)
1273 {
1274         xfs_da_intnode_t *node1, *node2;
1275
1276         node1 = node1_bp->data;
1277         node2 = node2_bp->data;
1278         ASSERT(node1->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC) &&
1279                node2->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
1280         if ((be16_to_cpu(node1->hdr.count) > 0) && (be16_to_cpu(node2->hdr.count) > 0) &&
1281             ((be32_to_cpu(node2->btree[0].hashval) <
1282               be32_to_cpu(node1->btree[0].hashval)) ||
1283              (be32_to_cpu(node2->btree[be16_to_cpu(node2->hdr.count)-1].hashval) <
1284               be32_to_cpu(node1->btree[be16_to_cpu(node1->hdr.count)-1].hashval)))) {
1285                 return(1);
1286         }
1287         return(0);
1288 }
1289
1290 /*
1291  * Pick up the last hashvalue from an intermediate node.
1292  */
1293 STATIC uint
1294 xfs_da_node_lasthash(xfs_dabuf_t *bp, int *count)
1295 {
1296         xfs_da_intnode_t *node;
1297
1298         node = bp->data;
1299         ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
1300         if (count)
1301                 *count = be16_to_cpu(node->hdr.count);
1302         if (!node->hdr.count)
1303                 return(0);
1304         return be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1].hashval);
1305 }
1306
1307 /*
1308  * Unlink a block from a doubly linked list of blocks.
1309  */
1310 STATIC int                                              /* error */
1311 xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
1312                                  xfs_da_state_blk_t *save_blk)
1313 {
1314         xfs_da_blkinfo_t *drop_info, *save_info, *tmp_info;
1315         xfs_da_args_t *args;
1316         xfs_dabuf_t *bp;
1317         int error;
1318
1319         /*
1320          * Set up environment.
1321          */
1322         args = state->args;
1323         ASSERT(args != NULL);
1324         save_info = save_blk->bp->data;
1325         drop_info = drop_blk->bp->data;
1326         ASSERT(save_blk->magic == XFS_DA_NODE_MAGIC ||
1327                save_blk->magic == XFS_DIR2_LEAFN_MAGIC ||
1328                save_blk->magic == XFS_ATTR_LEAF_MAGIC);
1329         ASSERT(save_blk->magic == be16_to_cpu(save_info->magic));
1330         ASSERT(drop_blk->magic == be16_to_cpu(drop_info->magic));
1331         ASSERT(save_blk->magic == drop_blk->magic);
1332         ASSERT((be32_to_cpu(save_info->forw) == drop_blk->blkno) ||
1333                (be32_to_cpu(save_info->back) == drop_blk->blkno));
1334         ASSERT((be32_to_cpu(drop_info->forw) == save_blk->blkno) ||
1335                (be32_to_cpu(drop_info->back) == save_blk->blkno));
1336
1337         /*
1338          * Unlink the leaf block from the doubly linked chain of leaves.
1339          */
1340         if (be32_to_cpu(save_info->back) == drop_blk->blkno) {
1341                 save_info->back = drop_info->back;
1342                 if (drop_info->back) {
1343                         error = xfs_da_read_buf(args->trans, args->dp,
1344                                                 be32_to_cpu(drop_info->back),
1345                                                 -1, &bp, args->whichfork);
1346                         if (error)
1347                                 return(error);
1348                         ASSERT(bp != NULL);
1349                         tmp_info = bp->data;
1350                         ASSERT(tmp_info->magic == save_info->magic);
1351                         ASSERT(be32_to_cpu(tmp_info->forw) == drop_blk->blkno);
1352                         tmp_info->forw = cpu_to_be32(save_blk->blkno);
1353                         xfs_da_log_buf(args->trans, bp, 0,
1354                                                     sizeof(*tmp_info) - 1);
1355                         xfs_da_buf_done(bp);
1356                 }
1357         } else {
1358                 save_info->forw = drop_info->forw;
1359                 if (drop_info->forw) {
1360                         error = xfs_da_read_buf(args->trans, args->dp,
1361                                                 be32_to_cpu(drop_info->forw),
1362                                                 -1, &bp, args->whichfork);
1363                         if (error)
1364                                 return(error);
1365                         ASSERT(bp != NULL);
1366                         tmp_info = bp->data;
1367                         ASSERT(tmp_info->magic == save_info->magic);
1368                         ASSERT(be32_to_cpu(tmp_info->back) == drop_blk->blkno);
1369                         tmp_info->back = cpu_to_be32(save_blk->blkno);
1370                         xfs_da_log_buf(args->trans, bp, 0,
1371                                                     sizeof(*tmp_info) - 1);
1372                         xfs_da_buf_done(bp);
1373                 }
1374         }
1375
1376         xfs_da_log_buf(args->trans, save_blk->bp, 0, sizeof(*save_info) - 1);
1377         return(0);
1378 }
1379
1380 /*
1381  * Move a path "forward" or "!forward" one block at the current level.
1382  *
1383  * This routine will adjust a "path" to point to the next block
1384  * "forward" (higher hashvalues) or "!forward" (lower hashvals) in the
1385  * Btree, including updating pointers to the intermediate nodes between
1386  * the new bottom and the root.
1387  */
1388 int                                                     /* error */
1389 xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
1390                                  int forward, int release, int *result)
1391 {
1392         xfs_da_state_blk_t *blk;
1393         xfs_da_blkinfo_t *info;
1394         xfs_da_intnode_t *node;
1395         xfs_da_args_t *args;
1396         xfs_dablk_t blkno=0;
1397         int level, error;
1398
1399         /*
1400          * Roll up the Btree looking for the first block where our
1401          * current index is not at the edge of the block.  Note that
1402          * we skip the bottom layer because we want the sibling block.
1403          */
1404         args = state->args;
1405         ASSERT(args != NULL);
1406         ASSERT(path != NULL);
1407         ASSERT((path->active > 0) && (path->active < XFS_DA_NODE_MAXDEPTH));
1408         level = (path->active-1) - 1;   /* skip bottom layer in path */
1409         for (blk = &path->blk[level]; level >= 0; blk--, level--) {
1410                 ASSERT(blk->bp != NULL);
1411                 node = blk->bp->data;
1412                 ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
1413                 if (forward && (blk->index < be16_to_cpu(node->hdr.count)-1)) {
1414                         blk->index++;
1415                         blkno = be32_to_cpu(node->btree[blk->index].before);
1416                         break;
1417                 } else if (!forward && (blk->index > 0)) {
1418                         blk->index--;
1419                         blkno = be32_to_cpu(node->btree[blk->index].before);
1420                         break;
1421                 }
1422         }
1423         if (level < 0) {
1424                 *result = XFS_ERROR(ENOENT);    /* we're out of our tree */
1425                 ASSERT(args->op_flags & XFS_DA_OP_OKNOENT);
1426                 return(0);
1427         }
1428
1429         /*
1430          * Roll down the edge of the subtree until we reach the
1431          * same depth we were at originally.
1432          */
1433         for (blk++, level++; level < path->active; blk++, level++) {
1434                 /*
1435                  * Release the old block.
1436                  * (if it's dirty, trans won't actually let go)
1437                  */
1438                 if (release)
1439                         xfs_da_brelse(args->trans, blk->bp);
1440
1441                 /*
1442                  * Read the next child block.
1443                  */
1444                 blk->blkno = blkno;
1445                 error = xfs_da_read_buf(args->trans, args->dp, blkno, -1,
1446                                                      &blk->bp, args->whichfork);
1447                 if (error)
1448                         return(error);
1449                 ASSERT(blk->bp != NULL);
1450                 info = blk->bp->data;
1451                 ASSERT(info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC) ||
1452                        info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
1453                        info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
1454                 blk->magic = be16_to_cpu(info->magic);
1455                 if (blk->magic == XFS_DA_NODE_MAGIC) {
1456                         node = (xfs_da_intnode_t *)info;
1457                         blk->hashval = be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1].hashval);
1458                         if (forward)
1459                                 blk->index = 0;
1460                         else
1461                                 blk->index = be16_to_cpu(node->hdr.count)-1;
1462                         blkno = be32_to_cpu(node->btree[blk->index].before);
1463                 } else {
1464                         ASSERT(level == path->active-1);
1465                         blk->index = 0;
1466                         switch(blk->magic) {
1467                         case XFS_ATTR_LEAF_MAGIC:
1468                                 blk->hashval = xfs_attr_leaf_lasthash(blk->bp,
1469                                                                       NULL);
1470                                 break;
1471                         case XFS_DIR2_LEAFN_MAGIC:
1472                                 blk->hashval = xfs_dir2_leafn_lasthash(blk->bp,
1473                                                                        NULL);
1474                                 break;
1475                         default:
1476                                 ASSERT(blk->magic == XFS_ATTR_LEAF_MAGIC ||
1477                                        blk->magic == XFS_DIR2_LEAFN_MAGIC);
1478                                 break;
1479                         }
1480                 }
1481         }
1482         *result = 0;
1483         return(0);
1484 }
1485
1486
1487 /*========================================================================
1488  * Utility routines.
1489  *========================================================================*/
1490
1491 /*
1492  * Implement a simple hash on a character string.
1493  * Rotate the hash value by 7 bits, then XOR each character in.
1494  * This is implemented with some source-level loop unrolling.
1495  */
1496 xfs_dahash_t
1497 xfs_da_hashname(const __uint8_t *name, int namelen)
1498 {
1499         xfs_dahash_t hash;
1500
1501         /*
1502          * Do four characters at a time as long as we can.
1503          */
1504         for (hash = 0; namelen >= 4; namelen -= 4, name += 4)
1505                 hash = (name[0] << 21) ^ (name[1] << 14) ^ (name[2] << 7) ^
1506                        (name[3] << 0) ^ rol32(hash, 7 * 4);
1507
1508         /*
1509          * Now do the rest of the characters.
1510          */
1511         switch (namelen) {
1512         case 3:
1513                 return (name[0] << 14) ^ (name[1] << 7) ^ (name[2] << 0) ^
1514                        rol32(hash, 7 * 3);
1515         case 2:
1516                 return (name[0] << 7) ^ (name[1] << 0) ^ rol32(hash, 7 * 2);
1517         case 1:
1518                 return (name[0] << 0) ^ rol32(hash, 7 * 1);
1519         default: /* case 0: */
1520                 return hash;
1521         }
1522 }
1523
1524 enum xfs_dacmp
1525 xfs_da_compname(
1526         struct xfs_da_args *args,
1527         const unsigned char *name,
1528         int             len)
1529 {
1530         return (args->namelen == len && memcmp(args->name, name, len) == 0) ?
1531                                         XFS_CMP_EXACT : XFS_CMP_DIFFERENT;
1532 }
1533
1534 static xfs_dahash_t
1535 xfs_default_hashname(
1536         struct xfs_name *name)
1537 {
1538         return xfs_da_hashname(name->name, name->len);
1539 }
1540
1541 const struct xfs_nameops xfs_default_nameops = {
1542         .hashname       = xfs_default_hashname,
1543         .compname       = xfs_da_compname
1544 };
1545
1546 int
1547 xfs_da_grow_inode_int(
1548         struct xfs_da_args      *args,
1549         xfs_fileoff_t           *bno,
1550         int                     count)
1551 {
1552         struct xfs_trans        *tp = args->trans;
1553         struct xfs_inode        *dp = args->dp;
1554         int                     w = args->whichfork;
1555         xfs_drfsbno_t           nblks = dp->i_d.di_nblocks;
1556         struct xfs_bmbt_irec    map, *mapp;
1557         int                     nmap, error, got, i, mapi;
1558
1559         /*
1560          * Find a spot in the file space to put the new block.
1561          */
1562         error = xfs_bmap_first_unused(tp, dp, count, bno, w);
1563         if (error)
1564                 return error;
1565
1566         /*
1567          * Try mapping it in one filesystem block.
1568          */
1569         nmap = 1;
1570         ASSERT(args->firstblock != NULL);
1571         error = xfs_bmapi(tp, dp, *bno, count,
1572                         xfs_bmapi_aflag(w)|XFS_BMAPI_WRITE|XFS_BMAPI_METADATA|
1573                         XFS_BMAPI_CONTIG,
1574                         args->firstblock, args->total, &map, &nmap,
1575                         args->flist);
1576         if (error)
1577                 return error;
1578
1579         ASSERT(nmap <= 1);
1580         if (nmap == 1) {
1581                 mapp = &map;
1582                 mapi = 1;
1583         } else if (nmap == 0 && count > 1) {
1584                 xfs_fileoff_t           b;
1585                 int                     c;
1586
1587                 /*
1588                  * If we didn't get it and the block might work if fragmented,
1589                  * try without the CONTIG flag.  Loop until we get it all.
1590                  */
1591                 mapp = kmem_alloc(sizeof(*mapp) * count, KM_SLEEP);
1592                 for (b = *bno, mapi = 0; b < *bno + count; ) {
1593                         nmap = MIN(XFS_BMAP_MAX_NMAP, count);
1594                         c = (int)(*bno + count - b);
1595                         error = xfs_bmapi(tp, dp, b, c,
1596                                         xfs_bmapi_aflag(w)|XFS_BMAPI_WRITE|
1597                                         XFS_BMAPI_METADATA,
1598                                         args->firstblock, args->total,
1599                                         &mapp[mapi], &nmap, args->flist);
1600                         if (error)
1601                                 goto out_free_map;
1602                         if (nmap < 1)
1603                                 break;
1604                         mapi += nmap;
1605                         b = mapp[mapi - 1].br_startoff +
1606                             mapp[mapi - 1].br_blockcount;
1607                 }
1608         } else {
1609                 mapi = 0;
1610                 mapp = NULL;
1611         }
1612
1613         /*
1614          * Count the blocks we got, make sure it matches the total.
1615          */
1616         for (i = 0, got = 0; i < mapi; i++)
1617                 got += mapp[i].br_blockcount;
1618         if (got != count || mapp[0].br_startoff != *bno ||
1619             mapp[mapi - 1].br_startoff + mapp[mapi - 1].br_blockcount !=
1620             *bno + count) {
1621                 error = XFS_ERROR(ENOSPC);
1622                 goto out_free_map;
1623         }
1624
1625         /* account for newly allocated blocks in reserved blocks total */
1626         args->total -= dp->i_d.di_nblocks - nblks;
1627
1628 out_free_map:
1629         if (mapp != &map)
1630                 kmem_free(mapp);
1631         return error;
1632 }
1633
1634 /*
1635  * Add a block to the btree ahead of the file.
1636  * Return the new block number to the caller.
1637  */
1638 int
1639 xfs_da_grow_inode(
1640         struct xfs_da_args      *args,
1641         xfs_dablk_t             *new_blkno)
1642 {
1643         xfs_fileoff_t           bno;
1644         int                     count;
1645         int                     error;
1646
1647         if (args->whichfork == XFS_DATA_FORK) {
1648                 bno = args->dp->i_mount->m_dirleafblk;
1649                 count = args->dp->i_mount->m_dirblkfsbs;
1650         } else {
1651                 bno = 0;
1652                 count = 1;
1653         }
1654
1655         error = xfs_da_grow_inode_int(args, &bno, count);
1656         if (!error)
1657                 *new_blkno = (xfs_dablk_t)bno;
1658         return error;
1659 }
1660
1661 /*
1662  * Ick.  We need to always be able to remove a btree block, even
1663  * if there's no space reservation because the filesystem is full.
1664  * This is called if xfs_bunmapi on a btree block fails due to ENOSPC.
1665  * It swaps the target block with the last block in the file.  The
1666  * last block in the file can always be removed since it can't cause
1667  * a bmap btree split to do that.
1668  */
1669 STATIC int
1670 xfs_da_swap_lastblock(xfs_da_args_t *args, xfs_dablk_t *dead_blknop,
1671                       xfs_dabuf_t **dead_bufp)
1672 {
1673         xfs_dablk_t dead_blkno, last_blkno, sib_blkno, par_blkno;
1674         xfs_dabuf_t *dead_buf, *last_buf, *sib_buf, *par_buf;
1675         xfs_fileoff_t lastoff;
1676         xfs_inode_t *ip;
1677         xfs_trans_t *tp;
1678         xfs_mount_t *mp;
1679         int error, w, entno, level, dead_level;
1680         xfs_da_blkinfo_t *dead_info, *sib_info;
1681         xfs_da_intnode_t *par_node, *dead_node;
1682         xfs_dir2_leaf_t *dead_leaf2;
1683         xfs_dahash_t dead_hash;
1684
1685         dead_buf = *dead_bufp;
1686         dead_blkno = *dead_blknop;
1687         tp = args->trans;
1688         ip = args->dp;
1689         w = args->whichfork;
1690         ASSERT(w == XFS_DATA_FORK);
1691         mp = ip->i_mount;
1692         lastoff = mp->m_dirfreeblk;
1693         error = xfs_bmap_last_before(tp, ip, &lastoff, w);
1694         if (error)
1695                 return error;
1696         if (unlikely(lastoff == 0)) {
1697                 XFS_ERROR_REPORT("xfs_da_swap_lastblock(1)", XFS_ERRLEVEL_LOW,
1698                                  mp);
1699                 return XFS_ERROR(EFSCORRUPTED);
1700         }
1701         /*
1702          * Read the last block in the btree space.
1703          */
1704         last_blkno = (xfs_dablk_t)lastoff - mp->m_dirblkfsbs;
1705         if ((error = xfs_da_read_buf(tp, ip, last_blkno, -1, &last_buf, w)))
1706                 return error;
1707         /*
1708          * Copy the last block into the dead buffer and log it.
1709          */
1710         memcpy(dead_buf->data, last_buf->data, mp->m_dirblksize);
1711         xfs_da_log_buf(tp, dead_buf, 0, mp->m_dirblksize - 1);
1712         dead_info = dead_buf->data;
1713         /*
1714          * Get values from the moved block.
1715          */
1716         if (dead_info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)) {
1717                 dead_leaf2 = (xfs_dir2_leaf_t *)dead_info;
1718                 dead_level = 0;
1719                 dead_hash = be32_to_cpu(dead_leaf2->ents[be16_to_cpu(dead_leaf2->hdr.count) - 1].hashval);
1720         } else {
1721                 ASSERT(dead_info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
1722                 dead_node = (xfs_da_intnode_t *)dead_info;
1723                 dead_level = be16_to_cpu(dead_node->hdr.level);
1724                 dead_hash = be32_to_cpu(dead_node->btree[be16_to_cpu(dead_node->hdr.count) - 1].hashval);
1725         }
1726         sib_buf = par_buf = NULL;
1727         /*
1728          * If the moved block has a left sibling, fix up the pointers.
1729          */
1730         if ((sib_blkno = be32_to_cpu(dead_info->back))) {
1731                 if ((error = xfs_da_read_buf(tp, ip, sib_blkno, -1, &sib_buf, w)))
1732                         goto done;
1733                 sib_info = sib_buf->data;
1734                 if (unlikely(
1735                     be32_to_cpu(sib_info->forw) != last_blkno ||
1736                     sib_info->magic != dead_info->magic)) {
1737                         XFS_ERROR_REPORT("xfs_da_swap_lastblock(2)",
1738                                          XFS_ERRLEVEL_LOW, mp);
1739                         error = XFS_ERROR(EFSCORRUPTED);
1740                         goto done;
1741                 }
1742                 sib_info->forw = cpu_to_be32(dead_blkno);
1743                 xfs_da_log_buf(tp, sib_buf,
1744                         XFS_DA_LOGRANGE(sib_info, &sib_info->forw,
1745                                         sizeof(sib_info->forw)));
1746                 xfs_da_buf_done(sib_buf);
1747                 sib_buf = NULL;
1748         }
1749         /*
1750          * If the moved block has a right sibling, fix up the pointers.
1751          */
1752         if ((sib_blkno = be32_to_cpu(dead_info->forw))) {
1753                 if ((error = xfs_da_read_buf(tp, ip, sib_blkno, -1, &sib_buf, w)))
1754                         goto done;
1755                 sib_info = sib_buf->data;
1756                 if (unlikely(
1757                        be32_to_cpu(sib_info->back) != last_blkno ||
1758                        sib_info->magic != dead_info->magic)) {
1759                         XFS_ERROR_REPORT("xfs_da_swap_lastblock(3)",
1760                                          XFS_ERRLEVEL_LOW, mp);
1761                         error = XFS_ERROR(EFSCORRUPTED);
1762                         goto done;
1763                 }
1764                 sib_info->back = cpu_to_be32(dead_blkno);
1765                 xfs_da_log_buf(tp, sib_buf,
1766                         XFS_DA_LOGRANGE(sib_info, &sib_info->back,
1767                                         sizeof(sib_info->back)));
1768                 xfs_da_buf_done(sib_buf);
1769                 sib_buf = NULL;
1770         }
1771         par_blkno = mp->m_dirleafblk;
1772         level = -1;
1773         /*
1774          * Walk down the tree looking for the parent of the moved block.
1775          */
1776         for (;;) {
1777                 if ((error = xfs_da_read_buf(tp, ip, par_blkno, -1, &par_buf, w)))
1778                         goto done;
1779                 par_node = par_buf->data;
1780                 if (unlikely(par_node->hdr.info.magic !=
1781                     cpu_to_be16(XFS_DA_NODE_MAGIC) ||
1782                     (level >= 0 && level != be16_to_cpu(par_node->hdr.level) + 1))) {
1783                         XFS_ERROR_REPORT("xfs_da_swap_lastblock(4)",
1784                                          XFS_ERRLEVEL_LOW, mp);
1785                         error = XFS_ERROR(EFSCORRUPTED);
1786                         goto done;
1787                 }
1788                 level = be16_to_cpu(par_node->hdr.level);
1789                 for (entno = 0;
1790                      entno < be16_to_cpu(par_node->hdr.count) &&
1791                      be32_to_cpu(par_node->btree[entno].hashval) < dead_hash;
1792                      entno++)
1793                         continue;
1794                 if (unlikely(entno == be16_to_cpu(par_node->hdr.count))) {
1795                         XFS_ERROR_REPORT("xfs_da_swap_lastblock(5)",
1796                                          XFS_ERRLEVEL_LOW, mp);
1797                         error = XFS_ERROR(EFSCORRUPTED);
1798                         goto done;
1799                 }
1800                 par_blkno = be32_to_cpu(par_node->btree[entno].before);
1801                 if (level == dead_level + 1)
1802                         break;
1803                 xfs_da_brelse(tp, par_buf);
1804                 par_buf = NULL;
1805         }
1806         /*
1807          * We're in the right parent block.
1808          * Look for the right entry.
1809          */
1810         for (;;) {
1811                 for (;
1812                      entno < be16_to_cpu(par_node->hdr.count) &&
1813                      be32_to_cpu(par_node->btree[entno].before) != last_blkno;
1814                      entno++)
1815                         continue;
1816                 if (entno < be16_to_cpu(par_node->hdr.count))
1817                         break;
1818                 par_blkno = be32_to_cpu(par_node->hdr.info.forw);
1819                 xfs_da_brelse(tp, par_buf);
1820                 par_buf = NULL;
1821                 if (unlikely(par_blkno == 0)) {
1822                         XFS_ERROR_REPORT("xfs_da_swap_lastblock(6)",
1823                                          XFS_ERRLEVEL_LOW, mp);
1824                         error = XFS_ERROR(EFSCORRUPTED);
1825                         goto done;
1826                 }
1827                 if ((error = xfs_da_read_buf(tp, ip, par_blkno, -1, &par_buf, w)))
1828                         goto done;
1829                 par_node = par_buf->data;
1830                 if (unlikely(
1831                     be16_to_cpu(par_node->hdr.level) != level ||
1832                     par_node->hdr.info.magic != cpu_to_be16(XFS_DA_NODE_MAGIC))) {
1833                         XFS_ERROR_REPORT("xfs_da_swap_lastblock(7)",
1834                                          XFS_ERRLEVEL_LOW, mp);
1835                         error = XFS_ERROR(EFSCORRUPTED);
1836                         goto done;
1837                 }
1838                 entno = 0;
1839         }
1840         /*
1841          * Update the parent entry pointing to the moved block.
1842          */
1843         par_node->btree[entno].before = cpu_to_be32(dead_blkno);
1844         xfs_da_log_buf(tp, par_buf,
1845                 XFS_DA_LOGRANGE(par_node, &par_node->btree[entno].before,
1846                                 sizeof(par_node->btree[entno].before)));
1847         xfs_da_buf_done(par_buf);
1848         xfs_da_buf_done(dead_buf);
1849         *dead_blknop = last_blkno;
1850         *dead_bufp = last_buf;
1851         return 0;
1852 done:
1853         if (par_buf)
1854                 xfs_da_brelse(tp, par_buf);
1855         if (sib_buf)
1856                 xfs_da_brelse(tp, sib_buf);
1857         xfs_da_brelse(tp, last_buf);
1858         return error;
1859 }
1860
1861 /*
1862  * Remove a btree block from a directory or attribute.
1863  */
1864 int
1865 xfs_da_shrink_inode(xfs_da_args_t *args, xfs_dablk_t dead_blkno,
1866                     xfs_dabuf_t *dead_buf)
1867 {
1868         xfs_inode_t *dp;
1869         int done, error, w, count;
1870         xfs_trans_t *tp;
1871         xfs_mount_t *mp;
1872
1873         dp = args->dp;
1874         w = args->whichfork;
1875         tp = args->trans;
1876         mp = dp->i_mount;
1877         if (w == XFS_DATA_FORK)
1878                 count = mp->m_dirblkfsbs;
1879         else
1880                 count = 1;
1881         for (;;) {
1882                 /*
1883                  * Remove extents.  If we get ENOSPC for a dir we have to move
1884                  * the last block to the place we want to kill.
1885                  */
1886                 if ((error = xfs_bunmapi(tp, dp, dead_blkno, count,
1887                                 xfs_bmapi_aflag(w)|XFS_BMAPI_METADATA,
1888                                 0, args->firstblock, args->flist,
1889                                 &done)) == ENOSPC) {
1890                         if (w != XFS_DATA_FORK)
1891                                 break;
1892                         if ((error = xfs_da_swap_lastblock(args, &dead_blkno,
1893                                         &dead_buf)))
1894                                 break;
1895                 } else {
1896                         break;
1897                 }
1898         }
1899         xfs_da_binval(tp, dead_buf);
1900         return error;
1901 }
1902
1903 /*
1904  * See if the mapping(s) for this btree block are valid, i.e.
1905  * don't contain holes, are logically contiguous, and cover the whole range.
1906  */
1907 STATIC int
1908 xfs_da_map_covers_blocks(
1909         int             nmap,
1910         xfs_bmbt_irec_t *mapp,
1911         xfs_dablk_t     bno,
1912         int             count)
1913 {
1914         int             i;
1915         xfs_fileoff_t   off;
1916
1917         for (i = 0, off = bno; i < nmap; i++) {
1918                 if (mapp[i].br_startblock == HOLESTARTBLOCK ||
1919                     mapp[i].br_startblock == DELAYSTARTBLOCK) {
1920                         return 0;
1921                 }
1922                 if (off != mapp[i].br_startoff) {
1923                         return 0;
1924                 }
1925                 off += mapp[i].br_blockcount;
1926         }
1927         return off == bno + count;
1928 }
1929
1930 /*
1931  * Make a dabuf.
1932  * Used for get_buf, read_buf, read_bufr, and reada_buf.
1933  */
1934 STATIC int
1935 xfs_da_do_buf(
1936         xfs_trans_t     *trans,
1937         xfs_inode_t     *dp,
1938         xfs_dablk_t     bno,
1939         xfs_daddr_t     *mappedbnop,
1940         xfs_dabuf_t     **bpp,
1941         int             whichfork,
1942         int             caller)
1943 {
1944         xfs_buf_t       *bp = NULL;
1945         xfs_buf_t       **bplist;
1946         int             error=0;
1947         int             i;
1948         xfs_bmbt_irec_t map;
1949         xfs_bmbt_irec_t *mapp;
1950         xfs_daddr_t     mappedbno;
1951         xfs_mount_t     *mp;
1952         int             nbplist=0;
1953         int             nfsb;
1954         int             nmap;
1955         xfs_dabuf_t     *rbp;
1956
1957         mp = dp->i_mount;
1958         nfsb = (whichfork == XFS_DATA_FORK) ? mp->m_dirblkfsbs : 1;
1959         mappedbno = *mappedbnop;
1960         /*
1961          * Caller doesn't have a mapping.  -2 means don't complain
1962          * if we land in a hole.
1963          */
1964         if (mappedbno == -1 || mappedbno == -2) {
1965                 /*
1966                  * Optimize the one-block case.
1967                  */
1968                 if (nfsb == 1) {
1969                         xfs_fsblock_t   fsb;
1970
1971                         if ((error =
1972                             xfs_bmapi_single(trans, dp, whichfork, &fsb,
1973                                     (xfs_fileoff_t)bno))) {
1974                                 return error;
1975                         }
1976                         mapp = &map;
1977                         if (fsb == NULLFSBLOCK) {
1978                                 nmap = 0;
1979                         } else {
1980                                 map.br_startblock = fsb;
1981                                 map.br_startoff = (xfs_fileoff_t)bno;
1982                                 map.br_blockcount = 1;
1983                                 nmap = 1;
1984                         }
1985                 } else {
1986                         mapp = kmem_alloc(sizeof(*mapp) * nfsb, KM_SLEEP);
1987                         nmap = nfsb;
1988                         if ((error = xfs_bmapi(trans, dp, (xfs_fileoff_t)bno,
1989                                         nfsb,
1990                                         XFS_BMAPI_METADATA |
1991                                                 xfs_bmapi_aflag(whichfork),
1992                                         NULL, 0, mapp, &nmap, NULL)))
1993                                 goto exit0;
1994                 }
1995         } else {
1996                 map.br_startblock = XFS_DADDR_TO_FSB(mp, mappedbno);
1997                 map.br_startoff = (xfs_fileoff_t)bno;
1998                 map.br_blockcount = nfsb;
1999                 mapp = &map;
2000                 nmap = 1;
2001         }
2002         if (!xfs_da_map_covers_blocks(nmap, mapp, bno, nfsb)) {
2003                 error = mappedbno == -2 ? 0 : XFS_ERROR(EFSCORRUPTED);
2004                 if (unlikely(error == EFSCORRUPTED)) {
2005                         if (xfs_error_level >= XFS_ERRLEVEL_LOW) {
2006                                 xfs_alert(mp, "%s: bno %lld dir: inode %lld",
2007                                         __func__, (long long)bno,
2008                                         (long long)dp->i_ino);
2009                                 for (i = 0; i < nmap; i++) {
2010                                         xfs_alert(mp,
2011 "[%02d] br_startoff %lld br_startblock %lld br_blockcount %lld br_state %d",
2012                                                 i,
2013                                                 (long long)mapp[i].br_startoff,
2014                                                 (long long)mapp[i].br_startblock,
2015                                                 (long long)mapp[i].br_blockcount,
2016                                                 mapp[i].br_state);
2017                                 }
2018                         }
2019                         XFS_ERROR_REPORT("xfs_da_do_buf(1)",
2020                                          XFS_ERRLEVEL_LOW, mp);
2021                 }
2022                 goto exit0;
2023         }
2024         if (caller != 3 && nmap > 1) {
2025                 bplist = kmem_alloc(sizeof(*bplist) * nmap, KM_SLEEP);
2026                 nbplist = 0;
2027         } else
2028                 bplist = NULL;
2029         /*
2030          * Turn the mapping(s) into buffer(s).
2031          */
2032         for (i = 0; i < nmap; i++) {
2033                 int     nmapped;
2034
2035                 mappedbno = XFS_FSB_TO_DADDR(mp, mapp[i].br_startblock);
2036                 if (i == 0)
2037                         *mappedbnop = mappedbno;
2038                 nmapped = (int)XFS_FSB_TO_BB(mp, mapp[i].br_blockcount);
2039                 switch (caller) {
2040                 case 0:
2041                         bp = xfs_trans_get_buf(trans, mp->m_ddev_targp,
2042                                 mappedbno, nmapped, 0);
2043                         error = bp ? XFS_BUF_GETERROR(bp) : XFS_ERROR(EIO);
2044                         break;
2045                 case 1:
2046                 case 2:
2047                         bp = NULL;
2048                         error = xfs_trans_read_buf(mp, trans, mp->m_ddev_targp,
2049                                 mappedbno, nmapped, 0, &bp);
2050                         break;
2051                 case 3:
2052                         xfs_buf_readahead(mp->m_ddev_targp, mappedbno, nmapped);
2053                         error = 0;
2054                         bp = NULL;
2055                         break;
2056                 }
2057                 if (error) {
2058                         if (bp)
2059                                 xfs_trans_brelse(trans, bp);
2060                         goto exit1;
2061                 }
2062                 if (!bp)
2063                         continue;
2064                 if (caller == 1) {
2065                         if (whichfork == XFS_ATTR_FORK) {
2066                                 XFS_BUF_SET_VTYPE_REF(bp, B_FS_ATTR_BTREE,
2067                                                 XFS_ATTR_BTREE_REF);
2068                         } else {
2069                                 XFS_BUF_SET_VTYPE_REF(bp, B_FS_DIR_BTREE,
2070                                                 XFS_DIR_BTREE_REF);
2071                         }
2072                 }
2073                 if (bplist) {
2074                         bplist[nbplist++] = bp;
2075                 }
2076         }
2077         /*
2078          * Build a dabuf structure.
2079          */
2080         if (bplist) {
2081                 rbp = xfs_da_buf_make(nbplist, bplist);
2082         } else if (bp)
2083                 rbp = xfs_da_buf_make(1, &bp);
2084         else
2085                 rbp = NULL;
2086         /*
2087          * For read_buf, check the magic number.
2088          */
2089         if (caller == 1) {
2090                 xfs_dir2_data_hdr_t     *hdr = rbp->data;
2091                 xfs_dir2_free_t         *free = rbp->data;
2092                 xfs_da_blkinfo_t        *info = rbp->data;
2093                 uint                    magic, magic1;
2094
2095                 magic = be16_to_cpu(info->magic);
2096                 magic1 = be32_to_cpu(hdr->magic);
2097                 if (unlikely(
2098                     XFS_TEST_ERROR((magic != XFS_DA_NODE_MAGIC) &&
2099                                    (magic != XFS_ATTR_LEAF_MAGIC) &&
2100                                    (magic != XFS_DIR2_LEAF1_MAGIC) &&
2101                                    (magic != XFS_DIR2_LEAFN_MAGIC) &&
2102                                    (magic1 != XFS_DIR2_BLOCK_MAGIC) &&
2103                                    (magic1 != XFS_DIR2_DATA_MAGIC) &&
2104                                    (free->hdr.magic != cpu_to_be32(XFS_DIR2_FREE_MAGIC)),
2105                                 mp, XFS_ERRTAG_DA_READ_BUF,
2106                                 XFS_RANDOM_DA_READ_BUF))) {
2107                         trace_xfs_da_btree_corrupt(rbp->bps[0], _RET_IP_);
2108                         XFS_CORRUPTION_ERROR("xfs_da_do_buf(2)",
2109                                              XFS_ERRLEVEL_LOW, mp, info);
2110                         error = XFS_ERROR(EFSCORRUPTED);
2111                         xfs_da_brelse(trans, rbp);
2112                         nbplist = 0;
2113                         goto exit1;
2114                 }
2115         }
2116         if (bplist) {
2117                 kmem_free(bplist);
2118         }
2119         if (mapp != &map) {
2120                 kmem_free(mapp);
2121         }
2122         if (bpp)
2123                 *bpp = rbp;
2124         return 0;
2125 exit1:
2126         if (bplist) {
2127                 for (i = 0; i < nbplist; i++)
2128                         xfs_trans_brelse(trans, bplist[i]);
2129                 kmem_free(bplist);
2130         }
2131 exit0:
2132         if (mapp != &map)
2133                 kmem_free(mapp);
2134         if (bpp)
2135                 *bpp = NULL;
2136         return error;
2137 }
2138
2139 /*
2140  * Get a buffer for the dir/attr block.
2141  */
2142 int
2143 xfs_da_get_buf(
2144         xfs_trans_t     *trans,
2145         xfs_inode_t     *dp,
2146         xfs_dablk_t     bno,
2147         xfs_daddr_t             mappedbno,
2148         xfs_dabuf_t     **bpp,
2149         int             whichfork)
2150 {
2151         return xfs_da_do_buf(trans, dp, bno, &mappedbno, bpp, whichfork, 0);
2152 }
2153
2154 /*
2155  * Get a buffer for the dir/attr block, fill in the contents.
2156  */
2157 int
2158 xfs_da_read_buf(
2159         xfs_trans_t     *trans,
2160         xfs_inode_t     *dp,
2161         xfs_dablk_t     bno,
2162         xfs_daddr_t             mappedbno,
2163         xfs_dabuf_t     **bpp,
2164         int             whichfork)
2165 {
2166         return xfs_da_do_buf(trans, dp, bno, &mappedbno, bpp, whichfork, 1);
2167 }
2168
2169 /*
2170  * Readahead the dir/attr block.
2171  */
2172 xfs_daddr_t
2173 xfs_da_reada_buf(
2174         xfs_trans_t     *trans,
2175         xfs_inode_t     *dp,
2176         xfs_dablk_t     bno,
2177         int             whichfork)
2178 {
2179         xfs_daddr_t             rval;
2180
2181         rval = -1;
2182         if (xfs_da_do_buf(trans, dp, bno, &rval, NULL, whichfork, 3))
2183                 return -1;
2184         else
2185                 return rval;
2186 }
2187
2188 kmem_zone_t *xfs_da_state_zone; /* anchor for state struct zone */
2189 kmem_zone_t *xfs_dabuf_zone;            /* dabuf zone */
2190
2191 /*
2192  * Allocate a dir-state structure.
2193  * We don't put them on the stack since they're large.
2194  */
2195 xfs_da_state_t *
2196 xfs_da_state_alloc(void)
2197 {
2198         return kmem_zone_zalloc(xfs_da_state_zone, KM_NOFS);
2199 }
2200
2201 /*
2202  * Kill the altpath contents of a da-state structure.
2203  */
2204 STATIC void
2205 xfs_da_state_kill_altpath(xfs_da_state_t *state)
2206 {
2207         int     i;
2208
2209         for (i = 0; i < state->altpath.active; i++) {
2210                 if (state->altpath.blk[i].bp) {
2211                         if (state->altpath.blk[i].bp != state->path.blk[i].bp)
2212                                 xfs_da_buf_done(state->altpath.blk[i].bp);
2213                         state->altpath.blk[i].bp = NULL;
2214                 }
2215         }
2216         state->altpath.active = 0;
2217 }
2218
2219 /*
2220  * Free a da-state structure.
2221  */
2222 void
2223 xfs_da_state_free(xfs_da_state_t *state)
2224 {
2225         int     i;
2226
2227         xfs_da_state_kill_altpath(state);
2228         for (i = 0; i < state->path.active; i++) {
2229                 if (state->path.blk[i].bp)
2230                         xfs_da_buf_done(state->path.blk[i].bp);
2231         }
2232         if (state->extravalid && state->extrablk.bp)
2233                 xfs_da_buf_done(state->extrablk.bp);
2234 #ifdef DEBUG
2235         memset((char *)state, 0, sizeof(*state));
2236 #endif /* DEBUG */
2237         kmem_zone_free(xfs_da_state_zone, state);
2238 }
2239
2240 /*
2241  * Create a dabuf.
2242  */
2243 /* ARGSUSED */
2244 STATIC xfs_dabuf_t *
2245 xfs_da_buf_make(int nbuf, xfs_buf_t **bps)
2246 {
2247         xfs_buf_t       *bp;
2248         xfs_dabuf_t     *dabuf;
2249         int             i;
2250         int             off;
2251
2252         if (nbuf == 1)
2253                 dabuf = kmem_zone_alloc(xfs_dabuf_zone, KM_NOFS);
2254         else
2255                 dabuf = kmem_alloc(XFS_DA_BUF_SIZE(nbuf), KM_NOFS);
2256         dabuf->dirty = 0;
2257         if (nbuf == 1) {
2258                 dabuf->nbuf = 1;
2259                 bp = bps[0];
2260                 dabuf->bbcount = (short)BTOBB(XFS_BUF_COUNT(bp));
2261                 dabuf->data = XFS_BUF_PTR(bp);
2262                 dabuf->bps[0] = bp;
2263         } else {
2264                 dabuf->nbuf = nbuf;
2265                 for (i = 0, dabuf->bbcount = 0; i < nbuf; i++) {
2266                         dabuf->bps[i] = bp = bps[i];
2267                         dabuf->bbcount += BTOBB(XFS_BUF_COUNT(bp));
2268                 }
2269                 dabuf->data = kmem_alloc(BBTOB(dabuf->bbcount), KM_SLEEP);
2270                 for (i = off = 0; i < nbuf; i++, off += XFS_BUF_COUNT(bp)) {
2271                         bp = bps[i];
2272                         memcpy((char *)dabuf->data + off, XFS_BUF_PTR(bp),
2273                                 XFS_BUF_COUNT(bp));
2274                 }
2275         }
2276         return dabuf;
2277 }
2278
2279 /*
2280  * Un-dirty a dabuf.
2281  */
2282 STATIC void
2283 xfs_da_buf_clean(xfs_dabuf_t *dabuf)
2284 {
2285         xfs_buf_t       *bp;
2286         int             i;
2287         int             off;
2288
2289         if (dabuf->dirty) {
2290                 ASSERT(dabuf->nbuf > 1);
2291                 dabuf->dirty = 0;
2292                 for (i = off = 0; i < dabuf->nbuf;
2293                                 i++, off += XFS_BUF_COUNT(bp)) {
2294                         bp = dabuf->bps[i];
2295                         memcpy(XFS_BUF_PTR(bp), (char *)dabuf->data + off,
2296                                 XFS_BUF_COUNT(bp));
2297                 }
2298         }
2299 }
2300
2301 /*
2302  * Release a dabuf.
2303  */
2304 void
2305 xfs_da_buf_done(xfs_dabuf_t *dabuf)
2306 {
2307         ASSERT(dabuf);
2308         ASSERT(dabuf->nbuf && dabuf->data && dabuf->bbcount && dabuf->bps[0]);
2309         if (dabuf->dirty)
2310                 xfs_da_buf_clean(dabuf);
2311         if (dabuf->nbuf > 1) {
2312                 kmem_free(dabuf->data);
2313                 kmem_free(dabuf);
2314         } else {
2315                 kmem_zone_free(xfs_dabuf_zone, dabuf);
2316         }
2317 }
2318
2319 /*
2320  * Log transaction from a dabuf.
2321  */
2322 void
2323 xfs_da_log_buf(xfs_trans_t *tp, xfs_dabuf_t *dabuf, uint first, uint last)
2324 {
2325         xfs_buf_t       *bp;
2326         uint            f;
2327         int             i;
2328         uint            l;
2329         int             off;
2330
2331         ASSERT(dabuf->nbuf && dabuf->data && dabuf->bbcount && dabuf->bps[0]);
2332         if (dabuf->nbuf == 1) {
2333                 ASSERT(dabuf->data == (void *)XFS_BUF_PTR(dabuf->bps[0]));
2334                 xfs_trans_log_buf(tp, dabuf->bps[0], first, last);
2335                 return;
2336         }
2337         dabuf->dirty = 1;
2338         ASSERT(first <= last);
2339         for (i = off = 0; i < dabuf->nbuf; i++, off += XFS_BUF_COUNT(bp)) {
2340                 bp = dabuf->bps[i];
2341                 f = off;
2342                 l = f + XFS_BUF_COUNT(bp) - 1;
2343                 if (f < first)
2344                         f = first;
2345                 if (l > last)
2346                         l = last;
2347                 if (f <= l)
2348                         xfs_trans_log_buf(tp, bp, f - off, l - off);
2349                 /*
2350                  * B_DONE is set by xfs_trans_log buf.
2351                  * If we don't set it on a new buffer (get not read)
2352                  * then if we don't put anything in the buffer it won't
2353                  * be set, and at commit it it released into the cache,
2354                  * and then a read will fail.
2355                  */
2356                 else if (!(XFS_BUF_ISDONE(bp)))
2357                   XFS_BUF_DONE(bp);
2358         }
2359         ASSERT(last < off);
2360 }
2361
2362 /*
2363  * Release dabuf from a transaction.
2364  * Have to free up the dabuf before the buffers are released,
2365  * since the synchronization on the dabuf is really the lock on the buffer.
2366  */
2367 void
2368 xfs_da_brelse(xfs_trans_t *tp, xfs_dabuf_t *dabuf)
2369 {
2370         xfs_buf_t       *bp;
2371         xfs_buf_t       **bplist;
2372         int             i;
2373         int             nbuf;
2374
2375         ASSERT(dabuf->nbuf && dabuf->data && dabuf->bbcount && dabuf->bps[0]);
2376         if ((nbuf = dabuf->nbuf) == 1) {
2377                 bplist = &bp;
2378                 bp = dabuf->bps[0];
2379         } else {
2380                 bplist = kmem_alloc(nbuf * sizeof(*bplist), KM_SLEEP);
2381                 memcpy(bplist, dabuf->bps, nbuf * sizeof(*bplist));
2382         }
2383         xfs_da_buf_done(dabuf);
2384         for (i = 0; i < nbuf; i++)
2385                 xfs_trans_brelse(tp, bplist[i]);
2386         if (bplist != &bp)
2387                 kmem_free(bplist);
2388 }
2389
2390 /*
2391  * Invalidate dabuf from a transaction.
2392  */
2393 void
2394 xfs_da_binval(xfs_trans_t *tp, xfs_dabuf_t *dabuf)
2395 {
2396         xfs_buf_t       *bp;
2397         xfs_buf_t       **bplist;
2398         int             i;
2399         int             nbuf;
2400
2401         ASSERT(dabuf->nbuf && dabuf->data && dabuf->bbcount && dabuf->bps[0]);
2402         if ((nbuf = dabuf->nbuf) == 1) {
2403                 bplist = &bp;
2404                 bp = dabuf->bps[0];
2405         } else {
2406                 bplist = kmem_alloc(nbuf * sizeof(*bplist), KM_SLEEP);
2407                 memcpy(bplist, dabuf->bps, nbuf * sizeof(*bplist));
2408         }
2409         xfs_da_buf_done(dabuf);
2410         for (i = 0; i < nbuf; i++)
2411                 xfs_trans_binval(tp, bplist[i]);
2412         if (bplist != &bp)
2413                 kmem_free(bplist);
2414 }
2415
2416 /*
2417  * Get the first daddr from a dabuf.
2418  */
2419 xfs_daddr_t
2420 xfs_da_blkno(xfs_dabuf_t *dabuf)
2421 {
2422         ASSERT(dabuf->nbuf);
2423         ASSERT(dabuf->data);
2424         return XFS_BUF_ADDR(dabuf->bps[0]);
2425 }