Merge git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw
[linux-2.6.git] / fs / ocfs2 / alloc.c
1 /* -*- mode: c; c-basic-offset: 8; -*-
2  * vim: noexpandtab sw=8 ts=8 sts=0:
3  *
4  * alloc.c
5  *
6  * Extent allocs and frees
7  *
8  * Copyright (C) 2002, 2004 Oracle.  All rights reserved.
9  *
10  * This program is free software; you can redistribute it and/or
11  * modify it under the terms of the GNU General Public
12  * License as published by the Free Software Foundation; either
13  * version 2 of the License, or (at your option) any later version.
14  *
15  * This program is distributed in the hope that it will be useful,
16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
18  * General Public License for more details.
19  *
20  * You should have received a copy of the GNU General Public
21  * License along with this program; if not, write to the
22  * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
23  * Boston, MA 021110-1307, USA.
24  */
25
26 #include <linux/fs.h>
27 #include <linux/types.h>
28 #include <linux/slab.h>
29 #include <linux/highmem.h>
30 #include <linux/swap.h>
31
32 #define MLOG_MASK_PREFIX ML_DISK_ALLOC
33 #include <cluster/masklog.h>
34
35 #include "ocfs2.h"
36
37 #include "alloc.h"
38 #include "aops.h"
39 #include "dlmglue.h"
40 #include "extent_map.h"
41 #include "inode.h"
42 #include "journal.h"
43 #include "localalloc.h"
44 #include "suballoc.h"
45 #include "sysfile.h"
46 #include "file.h"
47 #include "super.h"
48 #include "uptodate.h"
49
50 #include "buffer_head_io.h"
51
52 static void ocfs2_free_truncate_context(struct ocfs2_truncate_context *tc);
53 static int ocfs2_cache_extent_block_free(struct ocfs2_cached_dealloc_ctxt *ctxt,
54                                          struct ocfs2_extent_block *eb);
55
56 /*
57  * Structures which describe a path through a btree, and functions to
58  * manipulate them.
59  *
60  * The idea here is to be as generic as possible with the tree
61  * manipulation code.
62  */
63 struct ocfs2_path_item {
64         struct buffer_head              *bh;
65         struct ocfs2_extent_list        *el;
66 };
67
68 #define OCFS2_MAX_PATH_DEPTH    5
69
70 struct ocfs2_path {
71         int                     p_tree_depth;
72         struct ocfs2_path_item  p_node[OCFS2_MAX_PATH_DEPTH];
73 };
74
75 #define path_root_bh(_path) ((_path)->p_node[0].bh)
76 #define path_root_el(_path) ((_path)->p_node[0].el)
77 #define path_leaf_bh(_path) ((_path)->p_node[(_path)->p_tree_depth].bh)
78 #define path_leaf_el(_path) ((_path)->p_node[(_path)->p_tree_depth].el)
79 #define path_num_items(_path) ((_path)->p_tree_depth + 1)
80
81 /*
82  * Reset the actual path elements so that we can re-use the structure
83  * to build another path. Generally, this involves freeing the buffer
84  * heads.
85  */
86 static void ocfs2_reinit_path(struct ocfs2_path *path, int keep_root)
87 {
88         int i, start = 0, depth = 0;
89         struct ocfs2_path_item *node;
90
91         if (keep_root)
92                 start = 1;
93
94         for(i = start; i < path_num_items(path); i++) {
95                 node = &path->p_node[i];
96
97                 brelse(node->bh);
98                 node->bh = NULL;
99                 node->el = NULL;
100         }
101
102         /*
103          * Tree depth may change during truncate, or insert. If we're
104          * keeping the root extent list, then make sure that our path
105          * structure reflects the proper depth.
106          */
107         if (keep_root)
108                 depth = le16_to_cpu(path_root_el(path)->l_tree_depth);
109
110         path->p_tree_depth = depth;
111 }
112
113 static void ocfs2_free_path(struct ocfs2_path *path)
114 {
115         if (path) {
116                 ocfs2_reinit_path(path, 0);
117                 kfree(path);
118         }
119 }
120
121 /*
122  * All the elements of src into dest. After this call, src could be freed
123  * without affecting dest.
124  *
125  * Both paths should have the same root. Any non-root elements of dest
126  * will be freed.
127  */
128 static void ocfs2_cp_path(struct ocfs2_path *dest, struct ocfs2_path *src)
129 {
130         int i;
131
132         BUG_ON(path_root_bh(dest) != path_root_bh(src));
133         BUG_ON(path_root_el(dest) != path_root_el(src));
134
135         ocfs2_reinit_path(dest, 1);
136
137         for(i = 1; i < OCFS2_MAX_PATH_DEPTH; i++) {
138                 dest->p_node[i].bh = src->p_node[i].bh;
139                 dest->p_node[i].el = src->p_node[i].el;
140
141                 if (dest->p_node[i].bh)
142                         get_bh(dest->p_node[i].bh);
143         }
144 }
145
146 /*
147  * Make the *dest path the same as src and re-initialize src path to
148  * have a root only.
149  */
150 static void ocfs2_mv_path(struct ocfs2_path *dest, struct ocfs2_path *src)
151 {
152         int i;
153
154         BUG_ON(path_root_bh(dest) != path_root_bh(src));
155
156         for(i = 1; i < OCFS2_MAX_PATH_DEPTH; i++) {
157                 brelse(dest->p_node[i].bh);
158
159                 dest->p_node[i].bh = src->p_node[i].bh;
160                 dest->p_node[i].el = src->p_node[i].el;
161
162                 src->p_node[i].bh = NULL;
163                 src->p_node[i].el = NULL;
164         }
165 }
166
167 /*
168  * Insert an extent block at given index.
169  *
170  * This will not take an additional reference on eb_bh.
171  */
172 static inline void ocfs2_path_insert_eb(struct ocfs2_path *path, int index,
173                                         struct buffer_head *eb_bh)
174 {
175         struct ocfs2_extent_block *eb = (struct ocfs2_extent_block *)eb_bh->b_data;
176
177         /*
178          * Right now, no root bh is an extent block, so this helps
179          * catch code errors with dinode trees. The assertion can be
180          * safely removed if we ever need to insert extent block
181          * structures at the root.
182          */
183         BUG_ON(index == 0);
184
185         path->p_node[index].bh = eb_bh;
186         path->p_node[index].el = &eb->h_list;
187 }
188
189 static struct ocfs2_path *ocfs2_new_path(struct buffer_head *root_bh,
190                                          struct ocfs2_extent_list *root_el)
191 {
192         struct ocfs2_path *path;
193
194         BUG_ON(le16_to_cpu(root_el->l_tree_depth) >= OCFS2_MAX_PATH_DEPTH);
195
196         path = kzalloc(sizeof(*path), GFP_NOFS);
197         if (path) {
198                 path->p_tree_depth = le16_to_cpu(root_el->l_tree_depth);
199                 get_bh(root_bh);
200                 path_root_bh(path) = root_bh;
201                 path_root_el(path) = root_el;
202         }
203
204         return path;
205 }
206
207 /*
208  * Allocate and initialize a new path based on a disk inode tree.
209  */
210 static struct ocfs2_path *ocfs2_new_inode_path(struct buffer_head *di_bh)
211 {
212         struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
213         struct ocfs2_extent_list *el = &di->id2.i_list;
214
215         return ocfs2_new_path(di_bh, el);
216 }
217
218 /*
219  * Convenience function to journal all components in a path.
220  */
221 static int ocfs2_journal_access_path(struct inode *inode, handle_t *handle,
222                                      struct ocfs2_path *path)
223 {
224         int i, ret = 0;
225
226         if (!path)
227                 goto out;
228
229         for(i = 0; i < path_num_items(path); i++) {
230                 ret = ocfs2_journal_access(handle, inode, path->p_node[i].bh,
231                                            OCFS2_JOURNAL_ACCESS_WRITE);
232                 if (ret < 0) {
233                         mlog_errno(ret);
234                         goto out;
235                 }
236         }
237
238 out:
239         return ret;
240 }
241
242 /*
243  * Return the index of the extent record which contains cluster #v_cluster.
244  * -1 is returned if it was not found.
245  *
246  * Should work fine on interior and exterior nodes.
247  */
248 int ocfs2_search_extent_list(struct ocfs2_extent_list *el, u32 v_cluster)
249 {
250         int ret = -1;
251         int i;
252         struct ocfs2_extent_rec *rec;
253         u32 rec_end, rec_start, clusters;
254
255         for(i = 0; i < le16_to_cpu(el->l_next_free_rec); i++) {
256                 rec = &el->l_recs[i];
257
258                 rec_start = le32_to_cpu(rec->e_cpos);
259                 clusters = ocfs2_rec_clusters(el, rec);
260
261                 rec_end = rec_start + clusters;
262
263                 if (v_cluster >= rec_start && v_cluster < rec_end) {
264                         ret = i;
265                         break;
266                 }
267         }
268
269         return ret;
270 }
271
272 enum ocfs2_contig_type {
273         CONTIG_NONE = 0,
274         CONTIG_LEFT,
275         CONTIG_RIGHT,
276         CONTIG_LEFTRIGHT,
277 };
278
279
280 /*
281  * NOTE: ocfs2_block_extent_contig(), ocfs2_extents_adjacent() and
282  * ocfs2_extent_contig only work properly against leaf nodes!
283  */
284 static int ocfs2_block_extent_contig(struct super_block *sb,
285                                      struct ocfs2_extent_rec *ext,
286                                      u64 blkno)
287 {
288         u64 blk_end = le64_to_cpu(ext->e_blkno);
289
290         blk_end += ocfs2_clusters_to_blocks(sb,
291                                     le16_to_cpu(ext->e_leaf_clusters));
292
293         return blkno == blk_end;
294 }
295
296 static int ocfs2_extents_adjacent(struct ocfs2_extent_rec *left,
297                                   struct ocfs2_extent_rec *right)
298 {
299         u32 left_range;
300
301         left_range = le32_to_cpu(left->e_cpos) +
302                 le16_to_cpu(left->e_leaf_clusters);
303
304         return (left_range == le32_to_cpu(right->e_cpos));
305 }
306
307 static enum ocfs2_contig_type
308         ocfs2_extent_contig(struct inode *inode,
309                             struct ocfs2_extent_rec *ext,
310                             struct ocfs2_extent_rec *insert_rec)
311 {
312         u64 blkno = le64_to_cpu(insert_rec->e_blkno);
313
314         /*
315          * Refuse to coalesce extent records with different flag
316          * fields - we don't want to mix unwritten extents with user
317          * data.
318          */
319         if (ext->e_flags != insert_rec->e_flags)
320                 return CONTIG_NONE;
321
322         if (ocfs2_extents_adjacent(ext, insert_rec) &&
323             ocfs2_block_extent_contig(inode->i_sb, ext, blkno))
324                         return CONTIG_RIGHT;
325
326         blkno = le64_to_cpu(ext->e_blkno);
327         if (ocfs2_extents_adjacent(insert_rec, ext) &&
328             ocfs2_block_extent_contig(inode->i_sb, insert_rec, blkno))
329                 return CONTIG_LEFT;
330
331         return CONTIG_NONE;
332 }
333
334 /*
335  * NOTE: We can have pretty much any combination of contiguousness and
336  * appending.
337  *
338  * The usefulness of APPEND_TAIL is more in that it lets us know that
339  * we'll have to update the path to that leaf.
340  */
341 enum ocfs2_append_type {
342         APPEND_NONE = 0,
343         APPEND_TAIL,
344 };
345
346 enum ocfs2_split_type {
347         SPLIT_NONE = 0,
348         SPLIT_LEFT,
349         SPLIT_RIGHT,
350 };
351
352 struct ocfs2_insert_type {
353         enum ocfs2_split_type   ins_split;
354         enum ocfs2_append_type  ins_appending;
355         enum ocfs2_contig_type  ins_contig;
356         int                     ins_contig_index;
357         int                     ins_tree_depth;
358 };
359
360 struct ocfs2_merge_ctxt {
361         enum ocfs2_contig_type  c_contig_type;
362         int                     c_has_empty_extent;
363         int                     c_split_covers_rec;
364 };
365
366 /*
367  * How many free extents have we got before we need more meta data?
368  */
369 int ocfs2_num_free_extents(struct ocfs2_super *osb,
370                            struct inode *inode,
371                            struct ocfs2_dinode *fe)
372 {
373         int retval;
374         struct ocfs2_extent_list *el;
375         struct ocfs2_extent_block *eb;
376         struct buffer_head *eb_bh = NULL;
377
378         mlog_entry_void();
379
380         if (!OCFS2_IS_VALID_DINODE(fe)) {
381                 OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
382                 retval = -EIO;
383                 goto bail;
384         }
385
386         if (fe->i_last_eb_blk) {
387                 retval = ocfs2_read_block(osb, le64_to_cpu(fe->i_last_eb_blk),
388                                           &eb_bh, OCFS2_BH_CACHED, inode);
389                 if (retval < 0) {
390                         mlog_errno(retval);
391                         goto bail;
392                 }
393                 eb = (struct ocfs2_extent_block *) eb_bh->b_data;
394                 el = &eb->h_list;
395         } else
396                 el = &fe->id2.i_list;
397
398         BUG_ON(el->l_tree_depth != 0);
399
400         retval = le16_to_cpu(el->l_count) - le16_to_cpu(el->l_next_free_rec);
401 bail:
402         if (eb_bh)
403                 brelse(eb_bh);
404
405         mlog_exit(retval);
406         return retval;
407 }
408
409 /* expects array to already be allocated
410  *
411  * sets h_signature, h_blkno, h_suballoc_bit, h_suballoc_slot, and
412  * l_count for you
413  */
414 static int ocfs2_create_new_meta_bhs(struct ocfs2_super *osb,
415                                      handle_t *handle,
416                                      struct inode *inode,
417                                      int wanted,
418                                      struct ocfs2_alloc_context *meta_ac,
419                                      struct buffer_head *bhs[])
420 {
421         int count, status, i;
422         u16 suballoc_bit_start;
423         u32 num_got;
424         u64 first_blkno;
425         struct ocfs2_extent_block *eb;
426
427         mlog_entry_void();
428
429         count = 0;
430         while (count < wanted) {
431                 status = ocfs2_claim_metadata(osb,
432                                               handle,
433                                               meta_ac,
434                                               wanted - count,
435                                               &suballoc_bit_start,
436                                               &num_got,
437                                               &first_blkno);
438                 if (status < 0) {
439                         mlog_errno(status);
440                         goto bail;
441                 }
442
443                 for(i = count;  i < (num_got + count); i++) {
444                         bhs[i] = sb_getblk(osb->sb, first_blkno);
445                         if (bhs[i] == NULL) {
446                                 status = -EIO;
447                                 mlog_errno(status);
448                                 goto bail;
449                         }
450                         ocfs2_set_new_buffer_uptodate(inode, bhs[i]);
451
452                         status = ocfs2_journal_access(handle, inode, bhs[i],
453                                                       OCFS2_JOURNAL_ACCESS_CREATE);
454                         if (status < 0) {
455                                 mlog_errno(status);
456                                 goto bail;
457                         }
458
459                         memset(bhs[i]->b_data, 0, osb->sb->s_blocksize);
460                         eb = (struct ocfs2_extent_block *) bhs[i]->b_data;
461                         /* Ok, setup the minimal stuff here. */
462                         strcpy(eb->h_signature, OCFS2_EXTENT_BLOCK_SIGNATURE);
463                         eb->h_blkno = cpu_to_le64(first_blkno);
464                         eb->h_fs_generation = cpu_to_le32(osb->fs_generation);
465                         eb->h_suballoc_slot = cpu_to_le16(osb->slot_num);
466                         eb->h_suballoc_bit = cpu_to_le16(suballoc_bit_start);
467                         eb->h_list.l_count =
468                                 cpu_to_le16(ocfs2_extent_recs_per_eb(osb->sb));
469
470                         suballoc_bit_start++;
471                         first_blkno++;
472
473                         /* We'll also be dirtied by the caller, so
474                          * this isn't absolutely necessary. */
475                         status = ocfs2_journal_dirty(handle, bhs[i]);
476                         if (status < 0) {
477                                 mlog_errno(status);
478                                 goto bail;
479                         }
480                 }
481
482                 count += num_got;
483         }
484
485         status = 0;
486 bail:
487         if (status < 0) {
488                 for(i = 0; i < wanted; i++) {
489                         if (bhs[i])
490                                 brelse(bhs[i]);
491                         bhs[i] = NULL;
492                 }
493         }
494         mlog_exit(status);
495         return status;
496 }
497
498 /*
499  * Helper function for ocfs2_add_branch() and ocfs2_shift_tree_depth().
500  *
501  * Returns the sum of the rightmost extent rec logical offset and
502  * cluster count.
503  *
504  * ocfs2_add_branch() uses this to determine what logical cluster
505  * value should be populated into the leftmost new branch records.
506  *
507  * ocfs2_shift_tree_depth() uses this to determine the # clusters
508  * value for the new topmost tree record.
509  */
510 static inline u32 ocfs2_sum_rightmost_rec(struct ocfs2_extent_list  *el)
511 {
512         int i;
513
514         i = le16_to_cpu(el->l_next_free_rec) - 1;
515
516         return le32_to_cpu(el->l_recs[i].e_cpos) +
517                 ocfs2_rec_clusters(el, &el->l_recs[i]);
518 }
519
520 /*
521  * Add an entire tree branch to our inode. eb_bh is the extent block
522  * to start at, if we don't want to start the branch at the dinode
523  * structure.
524  *
525  * last_eb_bh is required as we have to update it's next_leaf pointer
526  * for the new last extent block.
527  *
528  * the new branch will be 'empty' in the sense that every block will
529  * contain a single record with cluster count == 0.
530  */
531 static int ocfs2_add_branch(struct ocfs2_super *osb,
532                             handle_t *handle,
533                             struct inode *inode,
534                             struct buffer_head *fe_bh,
535                             struct buffer_head *eb_bh,
536                             struct buffer_head **last_eb_bh,
537                             struct ocfs2_alloc_context *meta_ac)
538 {
539         int status, new_blocks, i;
540         u64 next_blkno, new_last_eb_blk;
541         struct buffer_head *bh;
542         struct buffer_head **new_eb_bhs = NULL;
543         struct ocfs2_dinode *fe;
544         struct ocfs2_extent_block *eb;
545         struct ocfs2_extent_list  *eb_el;
546         struct ocfs2_extent_list  *el;
547         u32 new_cpos;
548
549         mlog_entry_void();
550
551         BUG_ON(!last_eb_bh || !*last_eb_bh);
552
553         fe = (struct ocfs2_dinode *) fe_bh->b_data;
554
555         if (eb_bh) {
556                 eb = (struct ocfs2_extent_block *) eb_bh->b_data;
557                 el = &eb->h_list;
558         } else
559                 el = &fe->id2.i_list;
560
561         /* we never add a branch to a leaf. */
562         BUG_ON(!el->l_tree_depth);
563
564         new_blocks = le16_to_cpu(el->l_tree_depth);
565
566         /* allocate the number of new eb blocks we need */
567         new_eb_bhs = kcalloc(new_blocks, sizeof(struct buffer_head *),
568                              GFP_KERNEL);
569         if (!new_eb_bhs) {
570                 status = -ENOMEM;
571                 mlog_errno(status);
572                 goto bail;
573         }
574
575         status = ocfs2_create_new_meta_bhs(osb, handle, inode, new_blocks,
576                                            meta_ac, new_eb_bhs);
577         if (status < 0) {
578                 mlog_errno(status);
579                 goto bail;
580         }
581
582         eb = (struct ocfs2_extent_block *)(*last_eb_bh)->b_data;
583         new_cpos = ocfs2_sum_rightmost_rec(&eb->h_list);
584
585         /* Note: new_eb_bhs[new_blocks - 1] is the guy which will be
586          * linked with the rest of the tree.
587          * conversly, new_eb_bhs[0] is the new bottommost leaf.
588          *
589          * when we leave the loop, new_last_eb_blk will point to the
590          * newest leaf, and next_blkno will point to the topmost extent
591          * block. */
592         next_blkno = new_last_eb_blk = 0;
593         for(i = 0; i < new_blocks; i++) {
594                 bh = new_eb_bhs[i];
595                 eb = (struct ocfs2_extent_block *) bh->b_data;
596                 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
597                         OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
598                         status = -EIO;
599                         goto bail;
600                 }
601                 eb_el = &eb->h_list;
602
603                 status = ocfs2_journal_access(handle, inode, bh,
604                                               OCFS2_JOURNAL_ACCESS_CREATE);
605                 if (status < 0) {
606                         mlog_errno(status);
607                         goto bail;
608                 }
609
610                 eb->h_next_leaf_blk = 0;
611                 eb_el->l_tree_depth = cpu_to_le16(i);
612                 eb_el->l_next_free_rec = cpu_to_le16(1);
613                 /*
614                  * This actually counts as an empty extent as
615                  * c_clusters == 0
616                  */
617                 eb_el->l_recs[0].e_cpos = cpu_to_le32(new_cpos);
618                 eb_el->l_recs[0].e_blkno = cpu_to_le64(next_blkno);
619                 /*
620                  * eb_el isn't always an interior node, but even leaf
621                  * nodes want a zero'd flags and reserved field so
622                  * this gets the whole 32 bits regardless of use.
623                  */
624                 eb_el->l_recs[0].e_int_clusters = cpu_to_le32(0);
625                 if (!eb_el->l_tree_depth)
626                         new_last_eb_blk = le64_to_cpu(eb->h_blkno);
627
628                 status = ocfs2_journal_dirty(handle, bh);
629                 if (status < 0) {
630                         mlog_errno(status);
631                         goto bail;
632                 }
633
634                 next_blkno = le64_to_cpu(eb->h_blkno);
635         }
636
637         /* This is a bit hairy. We want to update up to three blocks
638          * here without leaving any of them in an inconsistent state
639          * in case of error. We don't have to worry about
640          * journal_dirty erroring as it won't unless we've aborted the
641          * handle (in which case we would never be here) so reserving
642          * the write with journal_access is all we need to do. */
643         status = ocfs2_journal_access(handle, inode, *last_eb_bh,
644                                       OCFS2_JOURNAL_ACCESS_WRITE);
645         if (status < 0) {
646                 mlog_errno(status);
647                 goto bail;
648         }
649         status = ocfs2_journal_access(handle, inode, fe_bh,
650                                       OCFS2_JOURNAL_ACCESS_WRITE);
651         if (status < 0) {
652                 mlog_errno(status);
653                 goto bail;
654         }
655         if (eb_bh) {
656                 status = ocfs2_journal_access(handle, inode, eb_bh,
657                                               OCFS2_JOURNAL_ACCESS_WRITE);
658                 if (status < 0) {
659                         mlog_errno(status);
660                         goto bail;
661                 }
662         }
663
664         /* Link the new branch into the rest of the tree (el will
665          * either be on the fe, or the extent block passed in. */
666         i = le16_to_cpu(el->l_next_free_rec);
667         el->l_recs[i].e_blkno = cpu_to_le64(next_blkno);
668         el->l_recs[i].e_cpos = cpu_to_le32(new_cpos);
669         el->l_recs[i].e_int_clusters = 0;
670         le16_add_cpu(&el->l_next_free_rec, 1);
671
672         /* fe needs a new last extent block pointer, as does the
673          * next_leaf on the previously last-extent-block. */
674         fe->i_last_eb_blk = cpu_to_le64(new_last_eb_blk);
675
676         eb = (struct ocfs2_extent_block *) (*last_eb_bh)->b_data;
677         eb->h_next_leaf_blk = cpu_to_le64(new_last_eb_blk);
678
679         status = ocfs2_journal_dirty(handle, *last_eb_bh);
680         if (status < 0)
681                 mlog_errno(status);
682         status = ocfs2_journal_dirty(handle, fe_bh);
683         if (status < 0)
684                 mlog_errno(status);
685         if (eb_bh) {
686                 status = ocfs2_journal_dirty(handle, eb_bh);
687                 if (status < 0)
688                         mlog_errno(status);
689         }
690
691         /*
692          * Some callers want to track the rightmost leaf so pass it
693          * back here.
694          */
695         brelse(*last_eb_bh);
696         get_bh(new_eb_bhs[0]);
697         *last_eb_bh = new_eb_bhs[0];
698
699         status = 0;
700 bail:
701         if (new_eb_bhs) {
702                 for (i = 0; i < new_blocks; i++)
703                         if (new_eb_bhs[i])
704                                 brelse(new_eb_bhs[i]);
705                 kfree(new_eb_bhs);
706         }
707
708         mlog_exit(status);
709         return status;
710 }
711
712 /*
713  * adds another level to the allocation tree.
714  * returns back the new extent block so you can add a branch to it
715  * after this call.
716  */
717 static int ocfs2_shift_tree_depth(struct ocfs2_super *osb,
718                                   handle_t *handle,
719                                   struct inode *inode,
720                                   struct buffer_head *fe_bh,
721                                   struct ocfs2_alloc_context *meta_ac,
722                                   struct buffer_head **ret_new_eb_bh)
723 {
724         int status, i;
725         u32 new_clusters;
726         struct buffer_head *new_eb_bh = NULL;
727         struct ocfs2_dinode *fe;
728         struct ocfs2_extent_block *eb;
729         struct ocfs2_extent_list  *fe_el;
730         struct ocfs2_extent_list  *eb_el;
731
732         mlog_entry_void();
733
734         status = ocfs2_create_new_meta_bhs(osb, handle, inode, 1, meta_ac,
735                                            &new_eb_bh);
736         if (status < 0) {
737                 mlog_errno(status);
738                 goto bail;
739         }
740
741         eb = (struct ocfs2_extent_block *) new_eb_bh->b_data;
742         if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
743                 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
744                 status = -EIO;
745                 goto bail;
746         }
747
748         eb_el = &eb->h_list;
749         fe = (struct ocfs2_dinode *) fe_bh->b_data;
750         fe_el = &fe->id2.i_list;
751
752         status = ocfs2_journal_access(handle, inode, new_eb_bh,
753                                       OCFS2_JOURNAL_ACCESS_CREATE);
754         if (status < 0) {
755                 mlog_errno(status);
756                 goto bail;
757         }
758
759         /* copy the fe data into the new extent block */
760         eb_el->l_tree_depth = fe_el->l_tree_depth;
761         eb_el->l_next_free_rec = fe_el->l_next_free_rec;
762         for(i = 0; i < le16_to_cpu(fe_el->l_next_free_rec); i++)
763                 eb_el->l_recs[i] = fe_el->l_recs[i];
764
765         status = ocfs2_journal_dirty(handle, new_eb_bh);
766         if (status < 0) {
767                 mlog_errno(status);
768                 goto bail;
769         }
770
771         status = ocfs2_journal_access(handle, inode, fe_bh,
772                                       OCFS2_JOURNAL_ACCESS_WRITE);
773         if (status < 0) {
774                 mlog_errno(status);
775                 goto bail;
776         }
777
778         new_clusters = ocfs2_sum_rightmost_rec(eb_el);
779
780         /* update fe now */
781         le16_add_cpu(&fe_el->l_tree_depth, 1);
782         fe_el->l_recs[0].e_cpos = 0;
783         fe_el->l_recs[0].e_blkno = eb->h_blkno;
784         fe_el->l_recs[0].e_int_clusters = cpu_to_le32(new_clusters);
785         for(i = 1; i < le16_to_cpu(fe_el->l_next_free_rec); i++)
786                 memset(&fe_el->l_recs[i], 0, sizeof(struct ocfs2_extent_rec));
787         fe_el->l_next_free_rec = cpu_to_le16(1);
788
789         /* If this is our 1st tree depth shift, then last_eb_blk
790          * becomes the allocated extent block */
791         if (fe_el->l_tree_depth == cpu_to_le16(1))
792                 fe->i_last_eb_blk = eb->h_blkno;
793
794         status = ocfs2_journal_dirty(handle, fe_bh);
795         if (status < 0) {
796                 mlog_errno(status);
797                 goto bail;
798         }
799
800         *ret_new_eb_bh = new_eb_bh;
801         new_eb_bh = NULL;
802         status = 0;
803 bail:
804         if (new_eb_bh)
805                 brelse(new_eb_bh);
806
807         mlog_exit(status);
808         return status;
809 }
810
811 /*
812  * Should only be called when there is no space left in any of the
813  * leaf nodes. What we want to do is find the lowest tree depth
814  * non-leaf extent block with room for new records. There are three
815  * valid results of this search:
816  *
817  * 1) a lowest extent block is found, then we pass it back in
818  *    *lowest_eb_bh and return '0'
819  *
820  * 2) the search fails to find anything, but the dinode has room. We
821  *    pass NULL back in *lowest_eb_bh, but still return '0'
822  *
823  * 3) the search fails to find anything AND the dinode is full, in
824  *    which case we return > 0
825  *
826  * return status < 0 indicates an error.
827  */
828 static int ocfs2_find_branch_target(struct ocfs2_super *osb,
829                                     struct inode *inode,
830                                     struct buffer_head *fe_bh,
831                                     struct buffer_head **target_bh)
832 {
833         int status = 0, i;
834         u64 blkno;
835         struct ocfs2_dinode *fe;
836         struct ocfs2_extent_block *eb;
837         struct ocfs2_extent_list  *el;
838         struct buffer_head *bh = NULL;
839         struct buffer_head *lowest_bh = NULL;
840
841         mlog_entry_void();
842
843         *target_bh = NULL;
844
845         fe = (struct ocfs2_dinode *) fe_bh->b_data;
846         el = &fe->id2.i_list;
847
848         while(le16_to_cpu(el->l_tree_depth) > 1) {
849                 if (le16_to_cpu(el->l_next_free_rec) == 0) {
850                         ocfs2_error(inode->i_sb, "Dinode %llu has empty "
851                                     "extent list (next_free_rec == 0)",
852                                     (unsigned long long)OCFS2_I(inode)->ip_blkno);
853                         status = -EIO;
854                         goto bail;
855                 }
856                 i = le16_to_cpu(el->l_next_free_rec) - 1;
857                 blkno = le64_to_cpu(el->l_recs[i].e_blkno);
858                 if (!blkno) {
859                         ocfs2_error(inode->i_sb, "Dinode %llu has extent "
860                                     "list where extent # %d has no physical "
861                                     "block start",
862                                     (unsigned long long)OCFS2_I(inode)->ip_blkno, i);
863                         status = -EIO;
864                         goto bail;
865                 }
866
867                 if (bh) {
868                         brelse(bh);
869                         bh = NULL;
870                 }
871
872                 status = ocfs2_read_block(osb, blkno, &bh, OCFS2_BH_CACHED,
873                                           inode);
874                 if (status < 0) {
875                         mlog_errno(status);
876                         goto bail;
877                 }
878
879                 eb = (struct ocfs2_extent_block *) bh->b_data;
880                 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
881                         OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
882                         status = -EIO;
883                         goto bail;
884                 }
885                 el = &eb->h_list;
886
887                 if (le16_to_cpu(el->l_next_free_rec) <
888                     le16_to_cpu(el->l_count)) {
889                         if (lowest_bh)
890                                 brelse(lowest_bh);
891                         lowest_bh = bh;
892                         get_bh(lowest_bh);
893                 }
894         }
895
896         /* If we didn't find one and the fe doesn't have any room,
897          * then return '1' */
898         if (!lowest_bh
899             && (fe->id2.i_list.l_next_free_rec == fe->id2.i_list.l_count))
900                 status = 1;
901
902         *target_bh = lowest_bh;
903 bail:
904         if (bh)
905                 brelse(bh);
906
907         mlog_exit(status);
908         return status;
909 }
910
911 /*
912  * Grow a b-tree so that it has more records.
913  *
914  * We might shift the tree depth in which case existing paths should
915  * be considered invalid.
916  *
917  * Tree depth after the grow is returned via *final_depth.
918  *
919  * *last_eb_bh will be updated by ocfs2_add_branch().
920  */
921 static int ocfs2_grow_tree(struct inode *inode, handle_t *handle,
922                            struct buffer_head *di_bh, int *final_depth,
923                            struct buffer_head **last_eb_bh,
924                            struct ocfs2_alloc_context *meta_ac)
925 {
926         int ret, shift;
927         struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
928         int depth = le16_to_cpu(di->id2.i_list.l_tree_depth);
929         struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
930         struct buffer_head *bh = NULL;
931
932         BUG_ON(meta_ac == NULL);
933
934         shift = ocfs2_find_branch_target(osb, inode, di_bh, &bh);
935         if (shift < 0) {
936                 ret = shift;
937                 mlog_errno(ret);
938                 goto out;
939         }
940
941         /* We traveled all the way to the bottom of the allocation tree
942          * and didn't find room for any more extents - we need to add
943          * another tree level */
944         if (shift) {
945                 BUG_ON(bh);
946                 mlog(0, "need to shift tree depth (current = %d)\n", depth);
947
948                 /* ocfs2_shift_tree_depth will return us a buffer with
949                  * the new extent block (so we can pass that to
950                  * ocfs2_add_branch). */
951                 ret = ocfs2_shift_tree_depth(osb, handle, inode, di_bh,
952                                              meta_ac, &bh);
953                 if (ret < 0) {
954                         mlog_errno(ret);
955                         goto out;
956                 }
957                 depth++;
958                 if (depth == 1) {
959                         /*
960                          * Special case: we have room now if we shifted from
961                          * tree_depth 0, so no more work needs to be done.
962                          *
963                          * We won't be calling add_branch, so pass
964                          * back *last_eb_bh as the new leaf. At depth
965                          * zero, it should always be null so there's
966                          * no reason to brelse.
967                          */
968                         BUG_ON(*last_eb_bh);
969                         get_bh(bh);
970                         *last_eb_bh = bh;
971                         goto out;
972                 }
973         }
974
975         /* call ocfs2_add_branch to add the final part of the tree with
976          * the new data. */
977         mlog(0, "add branch. bh = %p\n", bh);
978         ret = ocfs2_add_branch(osb, handle, inode, di_bh, bh, last_eb_bh,
979                                meta_ac);
980         if (ret < 0) {
981                 mlog_errno(ret);
982                 goto out;
983         }
984
985 out:
986         if (final_depth)
987                 *final_depth = depth;
988         brelse(bh);
989         return ret;
990 }
991
992 /*
993  * This is only valid for leaf nodes, which are the only ones that can
994  * have empty extents anyway.
995  */
996 static inline int ocfs2_is_empty_extent(struct ocfs2_extent_rec *rec)
997 {
998         return !rec->e_leaf_clusters;
999 }
1000
1001 /*
1002  * This function will discard the rightmost extent record.
1003  */
1004 static void ocfs2_shift_records_right(struct ocfs2_extent_list *el)
1005 {
1006         int next_free = le16_to_cpu(el->l_next_free_rec);
1007         int count = le16_to_cpu(el->l_count);
1008         unsigned int num_bytes;
1009
1010         BUG_ON(!next_free);
1011         /* This will cause us to go off the end of our extent list. */
1012         BUG_ON(next_free >= count);
1013
1014         num_bytes = sizeof(struct ocfs2_extent_rec) * next_free;
1015
1016         memmove(&el->l_recs[1], &el->l_recs[0], num_bytes);
1017 }
1018
1019 static void ocfs2_rotate_leaf(struct ocfs2_extent_list *el,
1020                               struct ocfs2_extent_rec *insert_rec)
1021 {
1022         int i, insert_index, next_free, has_empty, num_bytes;
1023         u32 insert_cpos = le32_to_cpu(insert_rec->e_cpos);
1024         struct ocfs2_extent_rec *rec;
1025
1026         next_free = le16_to_cpu(el->l_next_free_rec);
1027         has_empty = ocfs2_is_empty_extent(&el->l_recs[0]);
1028
1029         BUG_ON(!next_free);
1030
1031         /* The tree code before us didn't allow enough room in the leaf. */
1032         if (el->l_next_free_rec == el->l_count && !has_empty)
1033                 BUG();
1034
1035         /*
1036          * The easiest way to approach this is to just remove the
1037          * empty extent and temporarily decrement next_free.
1038          */
1039         if (has_empty) {
1040                 /*
1041                  * If next_free was 1 (only an empty extent), this
1042                  * loop won't execute, which is fine. We still want
1043                  * the decrement above to happen.
1044                  */
1045                 for(i = 0; i < (next_free - 1); i++)
1046                         el->l_recs[i] = el->l_recs[i+1];
1047
1048                 next_free--;
1049         }
1050
1051         /*
1052          * Figure out what the new record index should be.
1053          */
1054         for(i = 0; i < next_free; i++) {
1055                 rec = &el->l_recs[i];
1056
1057                 if (insert_cpos < le32_to_cpu(rec->e_cpos))
1058                         break;
1059         }
1060         insert_index = i;
1061
1062         mlog(0, "ins %u: index %d, has_empty %d, next_free %d, count %d\n",
1063              insert_cpos, insert_index, has_empty, next_free, le16_to_cpu(el->l_count));
1064
1065         BUG_ON(insert_index < 0);
1066         BUG_ON(insert_index >= le16_to_cpu(el->l_count));
1067         BUG_ON(insert_index > next_free);
1068
1069         /*
1070          * No need to memmove if we're just adding to the tail.
1071          */
1072         if (insert_index != next_free) {
1073                 BUG_ON(next_free >= le16_to_cpu(el->l_count));
1074
1075                 num_bytes = next_free - insert_index;
1076                 num_bytes *= sizeof(struct ocfs2_extent_rec);
1077                 memmove(&el->l_recs[insert_index + 1],
1078                         &el->l_recs[insert_index],
1079                         num_bytes);
1080         }
1081
1082         /*
1083          * Either we had an empty extent, and need to re-increment or
1084          * there was no empty extent on a non full rightmost leaf node,
1085          * in which case we still need to increment.
1086          */
1087         next_free++;
1088         el->l_next_free_rec = cpu_to_le16(next_free);
1089         /*
1090          * Make sure none of the math above just messed up our tree.
1091          */
1092         BUG_ON(le16_to_cpu(el->l_next_free_rec) > le16_to_cpu(el->l_count));
1093
1094         el->l_recs[insert_index] = *insert_rec;
1095
1096 }
1097
1098 static void ocfs2_remove_empty_extent(struct ocfs2_extent_list *el)
1099 {
1100         int size, num_recs = le16_to_cpu(el->l_next_free_rec);
1101
1102         BUG_ON(num_recs == 0);
1103
1104         if (ocfs2_is_empty_extent(&el->l_recs[0])) {
1105                 num_recs--;
1106                 size = num_recs * sizeof(struct ocfs2_extent_rec);
1107                 memmove(&el->l_recs[0], &el->l_recs[1], size);
1108                 memset(&el->l_recs[num_recs], 0,
1109                        sizeof(struct ocfs2_extent_rec));
1110                 el->l_next_free_rec = cpu_to_le16(num_recs);
1111         }
1112 }
1113
1114 /*
1115  * Create an empty extent record .
1116  *
1117  * l_next_free_rec may be updated.
1118  *
1119  * If an empty extent already exists do nothing.
1120  */
1121 static void ocfs2_create_empty_extent(struct ocfs2_extent_list *el)
1122 {
1123         int next_free = le16_to_cpu(el->l_next_free_rec);
1124
1125         BUG_ON(le16_to_cpu(el->l_tree_depth) != 0);
1126
1127         if (next_free == 0)
1128                 goto set_and_inc;
1129
1130         if (ocfs2_is_empty_extent(&el->l_recs[0]))
1131                 return;
1132
1133         mlog_bug_on_msg(el->l_count == el->l_next_free_rec,
1134                         "Asked to create an empty extent in a full list:\n"
1135                         "count = %u, tree depth = %u",
1136                         le16_to_cpu(el->l_count),
1137                         le16_to_cpu(el->l_tree_depth));
1138
1139         ocfs2_shift_records_right(el);
1140
1141 set_and_inc:
1142         le16_add_cpu(&el->l_next_free_rec, 1);
1143         memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec));
1144 }
1145
1146 /*
1147  * For a rotation which involves two leaf nodes, the "root node" is
1148  * the lowest level tree node which contains a path to both leafs. This
1149  * resulting set of information can be used to form a complete "subtree"
1150  *
1151  * This function is passed two full paths from the dinode down to a
1152  * pair of adjacent leaves. It's task is to figure out which path
1153  * index contains the subtree root - this can be the root index itself
1154  * in a worst-case rotation.
1155  *
1156  * The array index of the subtree root is passed back.
1157  */
1158 static int ocfs2_find_subtree_root(struct inode *inode,
1159                                    struct ocfs2_path *left,
1160                                    struct ocfs2_path *right)
1161 {
1162         int i = 0;
1163
1164         /*
1165          * Check that the caller passed in two paths from the same tree.
1166          */
1167         BUG_ON(path_root_bh(left) != path_root_bh(right));
1168
1169         do {
1170                 i++;
1171
1172                 /*
1173                  * The caller didn't pass two adjacent paths.
1174                  */
1175                 mlog_bug_on_msg(i > left->p_tree_depth,
1176                                 "Inode %lu, left depth %u, right depth %u\n"
1177                                 "left leaf blk %llu, right leaf blk %llu\n",
1178                                 inode->i_ino, left->p_tree_depth,
1179                                 right->p_tree_depth,
1180                                 (unsigned long long)path_leaf_bh(left)->b_blocknr,
1181                                 (unsigned long long)path_leaf_bh(right)->b_blocknr);
1182         } while (left->p_node[i].bh->b_blocknr ==
1183                  right->p_node[i].bh->b_blocknr);
1184
1185         return i - 1;
1186 }
1187
1188 typedef void (path_insert_t)(void *, struct buffer_head *);
1189
1190 /*
1191  * Traverse a btree path in search of cpos, starting at root_el.
1192  *
1193  * This code can be called with a cpos larger than the tree, in which
1194  * case it will return the rightmost path.
1195  */
1196 static int __ocfs2_find_path(struct inode *inode,
1197                              struct ocfs2_extent_list *root_el, u32 cpos,
1198                              path_insert_t *func, void *data)
1199 {
1200         int i, ret = 0;
1201         u32 range;
1202         u64 blkno;
1203         struct buffer_head *bh = NULL;
1204         struct ocfs2_extent_block *eb;
1205         struct ocfs2_extent_list *el;
1206         struct ocfs2_extent_rec *rec;
1207         struct ocfs2_inode_info *oi = OCFS2_I(inode);
1208
1209         el = root_el;
1210         while (el->l_tree_depth) {
1211                 if (le16_to_cpu(el->l_next_free_rec) == 0) {
1212                         ocfs2_error(inode->i_sb,
1213                                     "Inode %llu has empty extent list at "
1214                                     "depth %u\n",
1215                                     (unsigned long long)oi->ip_blkno,
1216                                     le16_to_cpu(el->l_tree_depth));
1217                         ret = -EROFS;
1218                         goto out;
1219
1220                 }
1221
1222                 for(i = 0; i < le16_to_cpu(el->l_next_free_rec) - 1; i++) {
1223                         rec = &el->l_recs[i];
1224
1225                         /*
1226                          * In the case that cpos is off the allocation
1227                          * tree, this should just wind up returning the
1228                          * rightmost record.
1229                          */
1230                         range = le32_to_cpu(rec->e_cpos) +
1231                                 ocfs2_rec_clusters(el, rec);
1232                         if (cpos >= le32_to_cpu(rec->e_cpos) && cpos < range)
1233                             break;
1234                 }
1235
1236                 blkno = le64_to_cpu(el->l_recs[i].e_blkno);
1237                 if (blkno == 0) {
1238                         ocfs2_error(inode->i_sb,
1239                                     "Inode %llu has bad blkno in extent list "
1240                                     "at depth %u (index %d)\n",
1241                                     (unsigned long long)oi->ip_blkno,
1242                                     le16_to_cpu(el->l_tree_depth), i);
1243                         ret = -EROFS;
1244                         goto out;
1245                 }
1246
1247                 brelse(bh);
1248                 bh = NULL;
1249                 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb), blkno,
1250                                        &bh, OCFS2_BH_CACHED, inode);
1251                 if (ret) {
1252                         mlog_errno(ret);
1253                         goto out;
1254                 }
1255
1256                 eb = (struct ocfs2_extent_block *) bh->b_data;
1257                 el = &eb->h_list;
1258                 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
1259                         OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
1260                         ret = -EIO;
1261                         goto out;
1262                 }
1263
1264                 if (le16_to_cpu(el->l_next_free_rec) >
1265                     le16_to_cpu(el->l_count)) {
1266                         ocfs2_error(inode->i_sb,
1267                                     "Inode %llu has bad count in extent list "
1268                                     "at block %llu (next free=%u, count=%u)\n",
1269                                     (unsigned long long)oi->ip_blkno,
1270                                     (unsigned long long)bh->b_blocknr,
1271                                     le16_to_cpu(el->l_next_free_rec),
1272                                     le16_to_cpu(el->l_count));
1273                         ret = -EROFS;
1274                         goto out;
1275                 }
1276
1277                 if (func)
1278                         func(data, bh);
1279         }
1280
1281 out:
1282         /*
1283          * Catch any trailing bh that the loop didn't handle.
1284          */
1285         brelse(bh);
1286
1287         return ret;
1288 }
1289
1290 /*
1291  * Given an initialized path (that is, it has a valid root extent
1292  * list), this function will traverse the btree in search of the path
1293  * which would contain cpos.
1294  *
1295  * The path traveled is recorded in the path structure.
1296  *
1297  * Note that this will not do any comparisons on leaf node extent
1298  * records, so it will work fine in the case that we just added a tree
1299  * branch.
1300  */
1301 struct find_path_data {
1302         int index;
1303         struct ocfs2_path *path;
1304 };
1305 static void find_path_ins(void *data, struct buffer_head *bh)
1306 {
1307         struct find_path_data *fp = data;
1308
1309         get_bh(bh);
1310         ocfs2_path_insert_eb(fp->path, fp->index, bh);
1311         fp->index++;
1312 }
1313 static int ocfs2_find_path(struct inode *inode, struct ocfs2_path *path,
1314                            u32 cpos)
1315 {
1316         struct find_path_data data;
1317
1318         data.index = 1;
1319         data.path = path;
1320         return __ocfs2_find_path(inode, path_root_el(path), cpos,
1321                                  find_path_ins, &data);
1322 }
1323
1324 static void find_leaf_ins(void *data, struct buffer_head *bh)
1325 {
1326         struct ocfs2_extent_block *eb =(struct ocfs2_extent_block *)bh->b_data;
1327         struct ocfs2_extent_list *el = &eb->h_list;
1328         struct buffer_head **ret = data;
1329
1330         /* We want to retain only the leaf block. */
1331         if (le16_to_cpu(el->l_tree_depth) == 0) {
1332                 get_bh(bh);
1333                 *ret = bh;
1334         }
1335 }
1336 /*
1337  * Find the leaf block in the tree which would contain cpos. No
1338  * checking of the actual leaf is done.
1339  *
1340  * Some paths want to call this instead of allocating a path structure
1341  * and calling ocfs2_find_path().
1342  *
1343  * This function doesn't handle non btree extent lists.
1344  */
1345 int ocfs2_find_leaf(struct inode *inode, struct ocfs2_extent_list *root_el,
1346                     u32 cpos, struct buffer_head **leaf_bh)
1347 {
1348         int ret;
1349         struct buffer_head *bh = NULL;
1350
1351         ret = __ocfs2_find_path(inode, root_el, cpos, find_leaf_ins, &bh);
1352         if (ret) {
1353                 mlog_errno(ret);
1354                 goto out;
1355         }
1356
1357         *leaf_bh = bh;
1358 out:
1359         return ret;
1360 }
1361
1362 /*
1363  * Adjust the adjacent records (left_rec, right_rec) involved in a rotation.
1364  *
1365  * Basically, we've moved stuff around at the bottom of the tree and
1366  * we need to fix up the extent records above the changes to reflect
1367  * the new changes.
1368  *
1369  * left_rec: the record on the left.
1370  * left_child_el: is the child list pointed to by left_rec
1371  * right_rec: the record to the right of left_rec
1372  * right_child_el: is the child list pointed to by right_rec
1373  *
1374  * By definition, this only works on interior nodes.
1375  */
1376 static void ocfs2_adjust_adjacent_records(struct ocfs2_extent_rec *left_rec,
1377                                   struct ocfs2_extent_list *left_child_el,
1378                                   struct ocfs2_extent_rec *right_rec,
1379                                   struct ocfs2_extent_list *right_child_el)
1380 {
1381         u32 left_clusters, right_end;
1382
1383         /*
1384          * Interior nodes never have holes. Their cpos is the cpos of
1385          * the leftmost record in their child list. Their cluster
1386          * count covers the full theoretical range of their child list
1387          * - the range between their cpos and the cpos of the record
1388          * immediately to their right.
1389          */
1390         left_clusters = le32_to_cpu(right_child_el->l_recs[0].e_cpos);
1391         if (ocfs2_is_empty_extent(&right_child_el->l_recs[0])) {
1392                 BUG_ON(le16_to_cpu(right_child_el->l_next_free_rec) <= 1);
1393                 left_clusters = le32_to_cpu(right_child_el->l_recs[1].e_cpos);
1394         }
1395         left_clusters -= le32_to_cpu(left_rec->e_cpos);
1396         left_rec->e_int_clusters = cpu_to_le32(left_clusters);
1397
1398         /*
1399          * Calculate the rightmost cluster count boundary before
1400          * moving cpos - we will need to adjust clusters after
1401          * updating e_cpos to keep the same highest cluster count.
1402          */
1403         right_end = le32_to_cpu(right_rec->e_cpos);
1404         right_end += le32_to_cpu(right_rec->e_int_clusters);
1405
1406         right_rec->e_cpos = left_rec->e_cpos;
1407         le32_add_cpu(&right_rec->e_cpos, left_clusters);
1408
1409         right_end -= le32_to_cpu(right_rec->e_cpos);
1410         right_rec->e_int_clusters = cpu_to_le32(right_end);
1411 }
1412
1413 /*
1414  * Adjust the adjacent root node records involved in a
1415  * rotation. left_el_blkno is passed in as a key so that we can easily
1416  * find it's index in the root list.
1417  */
1418 static void ocfs2_adjust_root_records(struct ocfs2_extent_list *root_el,
1419                                       struct ocfs2_extent_list *left_el,
1420                                       struct ocfs2_extent_list *right_el,
1421                                       u64 left_el_blkno)
1422 {
1423         int i;
1424
1425         BUG_ON(le16_to_cpu(root_el->l_tree_depth) <=
1426                le16_to_cpu(left_el->l_tree_depth));
1427
1428         for(i = 0; i < le16_to_cpu(root_el->l_next_free_rec) - 1; i++) {
1429                 if (le64_to_cpu(root_el->l_recs[i].e_blkno) == left_el_blkno)
1430                         break;
1431         }
1432
1433         /*
1434          * The path walking code should have never returned a root and
1435          * two paths which are not adjacent.
1436          */
1437         BUG_ON(i >= (le16_to_cpu(root_el->l_next_free_rec) - 1));
1438
1439         ocfs2_adjust_adjacent_records(&root_el->l_recs[i], left_el,
1440                                       &root_el->l_recs[i + 1], right_el);
1441 }
1442
1443 /*
1444  * We've changed a leaf block (in right_path) and need to reflect that
1445  * change back up the subtree.
1446  *
1447  * This happens in multiple places:
1448  *   - When we've moved an extent record from the left path leaf to the right
1449  *     path leaf to make room for an empty extent in the left path leaf.
1450  *   - When our insert into the right path leaf is at the leftmost edge
1451  *     and requires an update of the path immediately to it's left. This
1452  *     can occur at the end of some types of rotation and appending inserts.
1453  */
1454 static void ocfs2_complete_edge_insert(struct inode *inode, handle_t *handle,
1455                                        struct ocfs2_path *left_path,
1456                                        struct ocfs2_path *right_path,
1457                                        int subtree_index)
1458 {
1459         int ret, i, idx;
1460         struct ocfs2_extent_list *el, *left_el, *right_el;
1461         struct ocfs2_extent_rec *left_rec, *right_rec;
1462         struct buffer_head *root_bh = left_path->p_node[subtree_index].bh;
1463
1464         /*
1465          * Update the counts and position values within all the
1466          * interior nodes to reflect the leaf rotation we just did.
1467          *
1468          * The root node is handled below the loop.
1469          *
1470          * We begin the loop with right_el and left_el pointing to the
1471          * leaf lists and work our way up.
1472          *
1473          * NOTE: within this loop, left_el and right_el always refer
1474          * to the *child* lists.
1475          */
1476         left_el = path_leaf_el(left_path);
1477         right_el = path_leaf_el(right_path);
1478         for(i = left_path->p_tree_depth - 1; i > subtree_index; i--) {
1479                 mlog(0, "Adjust records at index %u\n", i);
1480
1481                 /*
1482                  * One nice property of knowing that all of these
1483                  * nodes are below the root is that we only deal with
1484                  * the leftmost right node record and the rightmost
1485                  * left node record.
1486                  */
1487                 el = left_path->p_node[i].el;
1488                 idx = le16_to_cpu(left_el->l_next_free_rec) - 1;
1489                 left_rec = &el->l_recs[idx];
1490
1491                 el = right_path->p_node[i].el;
1492                 right_rec = &el->l_recs[0];
1493
1494                 ocfs2_adjust_adjacent_records(left_rec, left_el, right_rec,
1495                                               right_el);
1496
1497                 ret = ocfs2_journal_dirty(handle, left_path->p_node[i].bh);
1498                 if (ret)
1499                         mlog_errno(ret);
1500
1501                 ret = ocfs2_journal_dirty(handle, right_path->p_node[i].bh);
1502                 if (ret)
1503                         mlog_errno(ret);
1504
1505                 /*
1506                  * Setup our list pointers now so that the current
1507                  * parents become children in the next iteration.
1508                  */
1509                 left_el = left_path->p_node[i].el;
1510                 right_el = right_path->p_node[i].el;
1511         }
1512
1513         /*
1514          * At the root node, adjust the two adjacent records which
1515          * begin our path to the leaves.
1516          */
1517
1518         el = left_path->p_node[subtree_index].el;
1519         left_el = left_path->p_node[subtree_index + 1].el;
1520         right_el = right_path->p_node[subtree_index + 1].el;
1521
1522         ocfs2_adjust_root_records(el, left_el, right_el,
1523                                   left_path->p_node[subtree_index + 1].bh->b_blocknr);
1524
1525         root_bh = left_path->p_node[subtree_index].bh;
1526
1527         ret = ocfs2_journal_dirty(handle, root_bh);
1528         if (ret)
1529                 mlog_errno(ret);
1530 }
1531
1532 static int ocfs2_rotate_subtree_right(struct inode *inode,
1533                                       handle_t *handle,
1534                                       struct ocfs2_path *left_path,
1535                                       struct ocfs2_path *right_path,
1536                                       int subtree_index)
1537 {
1538         int ret, i;
1539         struct buffer_head *right_leaf_bh;
1540         struct buffer_head *left_leaf_bh = NULL;
1541         struct buffer_head *root_bh;
1542         struct ocfs2_extent_list *right_el, *left_el;
1543         struct ocfs2_extent_rec move_rec;
1544
1545         left_leaf_bh = path_leaf_bh(left_path);
1546         left_el = path_leaf_el(left_path);
1547
1548         if (left_el->l_next_free_rec != left_el->l_count) {
1549                 ocfs2_error(inode->i_sb,
1550                             "Inode %llu has non-full interior leaf node %llu"
1551                             "(next free = %u)",
1552                             (unsigned long long)OCFS2_I(inode)->ip_blkno,
1553                             (unsigned long long)left_leaf_bh->b_blocknr,
1554                             le16_to_cpu(left_el->l_next_free_rec));
1555                 return -EROFS;
1556         }
1557
1558         /*
1559          * This extent block may already have an empty record, so we
1560          * return early if so.
1561          */
1562         if (ocfs2_is_empty_extent(&left_el->l_recs[0]))
1563                 return 0;
1564
1565         root_bh = left_path->p_node[subtree_index].bh;
1566         BUG_ON(root_bh != right_path->p_node[subtree_index].bh);
1567
1568         ret = ocfs2_journal_access(handle, inode, root_bh,
1569                                    OCFS2_JOURNAL_ACCESS_WRITE);
1570         if (ret) {
1571                 mlog_errno(ret);
1572                 goto out;
1573         }
1574
1575         for(i = subtree_index + 1; i < path_num_items(right_path); i++) {
1576                 ret = ocfs2_journal_access(handle, inode,
1577                                            right_path->p_node[i].bh,
1578                                            OCFS2_JOURNAL_ACCESS_WRITE);
1579                 if (ret) {
1580                         mlog_errno(ret);
1581                         goto out;
1582                 }
1583
1584                 ret = ocfs2_journal_access(handle, inode,
1585                                            left_path->p_node[i].bh,
1586                                            OCFS2_JOURNAL_ACCESS_WRITE);
1587                 if (ret) {
1588                         mlog_errno(ret);
1589                         goto out;
1590                 }
1591         }
1592
1593         right_leaf_bh = path_leaf_bh(right_path);
1594         right_el = path_leaf_el(right_path);
1595
1596         /* This is a code error, not a disk corruption. */
1597         mlog_bug_on_msg(!right_el->l_next_free_rec, "Inode %llu: Rotate fails "
1598                         "because rightmost leaf block %llu is empty\n",
1599                         (unsigned long long)OCFS2_I(inode)->ip_blkno,
1600                         (unsigned long long)right_leaf_bh->b_blocknr);
1601
1602         ocfs2_create_empty_extent(right_el);
1603
1604         ret = ocfs2_journal_dirty(handle, right_leaf_bh);
1605         if (ret) {
1606                 mlog_errno(ret);
1607                 goto out;
1608         }
1609
1610         /* Do the copy now. */
1611         i = le16_to_cpu(left_el->l_next_free_rec) - 1;
1612         move_rec = left_el->l_recs[i];
1613         right_el->l_recs[0] = move_rec;
1614
1615         /*
1616          * Clear out the record we just copied and shift everything
1617          * over, leaving an empty extent in the left leaf.
1618          *
1619          * We temporarily subtract from next_free_rec so that the
1620          * shift will lose the tail record (which is now defunct).
1621          */
1622         le16_add_cpu(&left_el->l_next_free_rec, -1);
1623         ocfs2_shift_records_right(left_el);
1624         memset(&left_el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec));
1625         le16_add_cpu(&left_el->l_next_free_rec, 1);
1626
1627         ret = ocfs2_journal_dirty(handle, left_leaf_bh);
1628         if (ret) {
1629                 mlog_errno(ret);
1630                 goto out;
1631         }
1632
1633         ocfs2_complete_edge_insert(inode, handle, left_path, right_path,
1634                                 subtree_index);
1635
1636 out:
1637         return ret;
1638 }
1639
1640 /*
1641  * Given a full path, determine what cpos value would return us a path
1642  * containing the leaf immediately to the left of the current one.
1643  *
1644  * Will return zero if the path passed in is already the leftmost path.
1645  */
1646 static int ocfs2_find_cpos_for_left_leaf(struct super_block *sb,
1647                                          struct ocfs2_path *path, u32 *cpos)
1648 {
1649         int i, j, ret = 0;
1650         u64 blkno;
1651         struct ocfs2_extent_list *el;
1652
1653         BUG_ON(path->p_tree_depth == 0);
1654
1655         *cpos = 0;
1656
1657         blkno = path_leaf_bh(path)->b_blocknr;
1658
1659         /* Start at the tree node just above the leaf and work our way up. */
1660         i = path->p_tree_depth - 1;
1661         while (i >= 0) {
1662                 el = path->p_node[i].el;
1663
1664                 /*
1665                  * Find the extent record just before the one in our
1666                  * path.
1667                  */
1668                 for(j = 0; j < le16_to_cpu(el->l_next_free_rec); j++) {
1669                         if (le64_to_cpu(el->l_recs[j].e_blkno) == blkno) {
1670                                 if (j == 0) {
1671                                         if (i == 0) {
1672                                                 /*
1673                                                  * We've determined that the
1674                                                  * path specified is already
1675                                                  * the leftmost one - return a
1676                                                  * cpos of zero.
1677                                                  */
1678                                                 goto out;
1679                                         }
1680                                         /*
1681                                          * The leftmost record points to our
1682                                          * leaf - we need to travel up the
1683                                          * tree one level.
1684                                          */
1685                                         goto next_node;
1686                                 }
1687
1688                                 *cpos = le32_to_cpu(el->l_recs[j - 1].e_cpos);
1689                                 *cpos = *cpos + ocfs2_rec_clusters(el,
1690                                                            &el->l_recs[j - 1]);
1691                                 *cpos = *cpos - 1;
1692                                 goto out;
1693                         }
1694                 }
1695
1696                 /*
1697                  * If we got here, we never found a valid node where
1698                  * the tree indicated one should be.
1699                  */
1700                 ocfs2_error(sb,
1701                             "Invalid extent tree at extent block %llu\n",
1702                             (unsigned long long)blkno);
1703                 ret = -EROFS;
1704                 goto out;
1705
1706 next_node:
1707                 blkno = path->p_node[i].bh->b_blocknr;
1708                 i--;
1709         }
1710
1711 out:
1712         return ret;
1713 }
1714
1715 /*
1716  * Extend the transaction by enough credits to complete the rotation,
1717  * and still leave at least the original number of credits allocated
1718  * to this transaction.
1719  */
1720 static int ocfs2_extend_rotate_transaction(handle_t *handle, int subtree_depth,
1721                                            int op_credits,
1722                                            struct ocfs2_path *path)
1723 {
1724         int credits = (path->p_tree_depth - subtree_depth) * 2 + 1 + op_credits;
1725
1726         if (handle->h_buffer_credits < credits)
1727                 return ocfs2_extend_trans(handle, credits);
1728
1729         return 0;
1730 }
1731
1732 /*
1733  * Trap the case where we're inserting into the theoretical range past
1734  * the _actual_ left leaf range. Otherwise, we'll rotate a record
1735  * whose cpos is less than ours into the right leaf.
1736  *
1737  * It's only necessary to look at the rightmost record of the left
1738  * leaf because the logic that calls us should ensure that the
1739  * theoretical ranges in the path components above the leaves are
1740  * correct.
1741  */
1742 static int ocfs2_rotate_requires_path_adjustment(struct ocfs2_path *left_path,
1743                                                  u32 insert_cpos)
1744 {
1745         struct ocfs2_extent_list *left_el;
1746         struct ocfs2_extent_rec *rec;
1747         int next_free;
1748
1749         left_el = path_leaf_el(left_path);
1750         next_free = le16_to_cpu(left_el->l_next_free_rec);
1751         rec = &left_el->l_recs[next_free - 1];
1752
1753         if (insert_cpos > le32_to_cpu(rec->e_cpos))
1754                 return 1;
1755         return 0;
1756 }
1757
1758 static int ocfs2_leftmost_rec_contains(struct ocfs2_extent_list *el, u32 cpos)
1759 {
1760         int next_free = le16_to_cpu(el->l_next_free_rec);
1761         unsigned int range;
1762         struct ocfs2_extent_rec *rec;
1763
1764         if (next_free == 0)
1765                 return 0;
1766
1767         rec = &el->l_recs[0];
1768         if (ocfs2_is_empty_extent(rec)) {
1769                 /* Empty list. */
1770                 if (next_free == 1)
1771                         return 0;
1772                 rec = &el->l_recs[1];
1773         }
1774
1775         range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec);
1776         if (cpos >= le32_to_cpu(rec->e_cpos) && cpos < range)
1777                 return 1;
1778         return 0;
1779 }
1780
1781 /*
1782  * Rotate all the records in a btree right one record, starting at insert_cpos.
1783  *
1784  * The path to the rightmost leaf should be passed in.
1785  *
1786  * The array is assumed to be large enough to hold an entire path (tree depth).
1787  *
1788  * Upon succesful return from this function:
1789  *
1790  * - The 'right_path' array will contain a path to the leaf block
1791  *   whose range contains e_cpos.
1792  * - That leaf block will have a single empty extent in list index 0.
1793  * - In the case that the rotation requires a post-insert update,
1794  *   *ret_left_path will contain a valid path which can be passed to
1795  *   ocfs2_insert_path().
1796  */
1797 static int ocfs2_rotate_tree_right(struct inode *inode,
1798                                    handle_t *handle,
1799                                    enum ocfs2_split_type split,
1800                                    u32 insert_cpos,
1801                                    struct ocfs2_path *right_path,
1802                                    struct ocfs2_path **ret_left_path)
1803 {
1804         int ret, start, orig_credits = handle->h_buffer_credits;
1805         u32 cpos;
1806         struct ocfs2_path *left_path = NULL;
1807
1808         *ret_left_path = NULL;
1809
1810         left_path = ocfs2_new_path(path_root_bh(right_path),
1811                                    path_root_el(right_path));
1812         if (!left_path) {
1813                 ret = -ENOMEM;
1814                 mlog_errno(ret);
1815                 goto out;
1816         }
1817
1818         ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path, &cpos);
1819         if (ret) {
1820                 mlog_errno(ret);
1821                 goto out;
1822         }
1823
1824         mlog(0, "Insert: %u, first left path cpos: %u\n", insert_cpos, cpos);
1825
1826         /*
1827          * What we want to do here is:
1828          *
1829          * 1) Start with the rightmost path.
1830          *
1831          * 2) Determine a path to the leaf block directly to the left
1832          *    of that leaf.
1833          *
1834          * 3) Determine the 'subtree root' - the lowest level tree node
1835          *    which contains a path to both leaves.
1836          *
1837          * 4) Rotate the subtree.
1838          *
1839          * 5) Find the next subtree by considering the left path to be
1840          *    the new right path.
1841          *
1842          * The check at the top of this while loop also accepts
1843          * insert_cpos == cpos because cpos is only a _theoretical_
1844          * value to get us the left path - insert_cpos might very well
1845          * be filling that hole.
1846          *
1847          * Stop at a cpos of '0' because we either started at the
1848          * leftmost branch (i.e., a tree with one branch and a
1849          * rotation inside of it), or we've gone as far as we can in
1850          * rotating subtrees.
1851          */
1852         while (cpos && insert_cpos <= cpos) {
1853                 mlog(0, "Rotating a tree: ins. cpos: %u, left path cpos: %u\n",
1854                      insert_cpos, cpos);
1855
1856                 ret = ocfs2_find_path(inode, left_path, cpos);
1857                 if (ret) {
1858                         mlog_errno(ret);
1859                         goto out;
1860                 }
1861
1862                 mlog_bug_on_msg(path_leaf_bh(left_path) ==
1863                                 path_leaf_bh(right_path),
1864                                 "Inode %lu: error during insert of %u "
1865                                 "(left path cpos %u) results in two identical "
1866                                 "paths ending at %llu\n",
1867                                 inode->i_ino, insert_cpos, cpos,
1868                                 (unsigned long long)
1869                                 path_leaf_bh(left_path)->b_blocknr);
1870
1871                 if (split == SPLIT_NONE &&
1872                     ocfs2_rotate_requires_path_adjustment(left_path,
1873                                                           insert_cpos)) {
1874
1875                         /*
1876                          * We've rotated the tree as much as we
1877                          * should. The rest is up to
1878                          * ocfs2_insert_path() to complete, after the
1879                          * record insertion. We indicate this
1880                          * situation by returning the left path.
1881                          *
1882                          * The reason we don't adjust the records here
1883                          * before the record insert is that an error
1884                          * later might break the rule where a parent
1885                          * record e_cpos will reflect the actual
1886                          * e_cpos of the 1st nonempty record of the
1887                          * child list.
1888                          */
1889                         *ret_left_path = left_path;
1890                         goto out_ret_path;
1891                 }
1892
1893                 start = ocfs2_find_subtree_root(inode, left_path, right_path);
1894
1895                 mlog(0, "Subtree root at index %d (blk %llu, depth %d)\n",
1896                      start,
1897                      (unsigned long long) right_path->p_node[start].bh->b_blocknr,
1898                      right_path->p_tree_depth);
1899
1900                 ret = ocfs2_extend_rotate_transaction(handle, start,
1901                                                       orig_credits, right_path);
1902                 if (ret) {
1903                         mlog_errno(ret);
1904                         goto out;
1905                 }
1906
1907                 ret = ocfs2_rotate_subtree_right(inode, handle, left_path,
1908                                                  right_path, start);
1909                 if (ret) {
1910                         mlog_errno(ret);
1911                         goto out;
1912                 }
1913
1914                 if (split != SPLIT_NONE &&
1915                     ocfs2_leftmost_rec_contains(path_leaf_el(right_path),
1916                                                 insert_cpos)) {
1917                         /*
1918                          * A rotate moves the rightmost left leaf
1919                          * record over to the leftmost right leaf
1920                          * slot. If we're doing an extent split
1921                          * instead of a real insert, then we have to
1922                          * check that the extent to be split wasn't
1923                          * just moved over. If it was, then we can
1924                          * exit here, passing left_path back -
1925                          * ocfs2_split_extent() is smart enough to
1926                          * search both leaves.
1927                          */
1928                         *ret_left_path = left_path;
1929                         goto out_ret_path;
1930                 }
1931
1932                 /*
1933                  * There is no need to re-read the next right path
1934                  * as we know that it'll be our current left
1935                  * path. Optimize by copying values instead.
1936                  */
1937                 ocfs2_mv_path(right_path, left_path);
1938
1939                 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path,
1940                                                     &cpos);
1941                 if (ret) {
1942                         mlog_errno(ret);
1943                         goto out;
1944                 }
1945         }
1946
1947 out:
1948         ocfs2_free_path(left_path);
1949
1950 out_ret_path:
1951         return ret;
1952 }
1953
1954 static void ocfs2_update_edge_lengths(struct inode *inode, handle_t *handle,
1955                                       struct ocfs2_path *path)
1956 {
1957         int i, idx;
1958         struct ocfs2_extent_rec *rec;
1959         struct ocfs2_extent_list *el;
1960         struct ocfs2_extent_block *eb;
1961         u32 range;
1962
1963         /* Path should always be rightmost. */
1964         eb = (struct ocfs2_extent_block *)path_leaf_bh(path)->b_data;
1965         BUG_ON(eb->h_next_leaf_blk != 0ULL);
1966
1967         el = &eb->h_list;
1968         BUG_ON(le16_to_cpu(el->l_next_free_rec) == 0);
1969         idx = le16_to_cpu(el->l_next_free_rec) - 1;
1970         rec = &el->l_recs[idx];
1971         range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec);
1972
1973         for (i = 0; i < path->p_tree_depth; i++) {
1974                 el = path->p_node[i].el;
1975                 idx = le16_to_cpu(el->l_next_free_rec) - 1;
1976                 rec = &el->l_recs[idx];
1977
1978                 rec->e_int_clusters = cpu_to_le32(range);
1979                 le32_add_cpu(&rec->e_int_clusters, -le32_to_cpu(rec->e_cpos));
1980
1981                 ocfs2_journal_dirty(handle, path->p_node[i].bh);
1982         }
1983 }
1984
1985 static void ocfs2_unlink_path(struct inode *inode, handle_t *handle,
1986                               struct ocfs2_cached_dealloc_ctxt *dealloc,
1987                               struct ocfs2_path *path, int unlink_start)
1988 {
1989         int ret, i;
1990         struct ocfs2_extent_block *eb;
1991         struct ocfs2_extent_list *el;
1992         struct buffer_head *bh;
1993
1994         for(i = unlink_start; i < path_num_items(path); i++) {
1995                 bh = path->p_node[i].bh;
1996
1997                 eb = (struct ocfs2_extent_block *)bh->b_data;
1998                 /*
1999                  * Not all nodes might have had their final count
2000                  * decremented by the caller - handle this here.
2001                  */
2002                 el = &eb->h_list;
2003                 if (le16_to_cpu(el->l_next_free_rec) > 1) {
2004                         mlog(ML_ERROR,
2005                              "Inode %llu, attempted to remove extent block "
2006                              "%llu with %u records\n",
2007                              (unsigned long long)OCFS2_I(inode)->ip_blkno,
2008                              (unsigned long long)le64_to_cpu(eb->h_blkno),
2009                              le16_to_cpu(el->l_next_free_rec));
2010
2011                         ocfs2_journal_dirty(handle, bh);
2012                         ocfs2_remove_from_cache(inode, bh);
2013                         continue;
2014                 }
2015
2016                 el->l_next_free_rec = 0;
2017                 memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec));
2018
2019                 ocfs2_journal_dirty(handle, bh);
2020
2021                 ret = ocfs2_cache_extent_block_free(dealloc, eb);
2022                 if (ret)
2023                         mlog_errno(ret);
2024
2025                 ocfs2_remove_from_cache(inode, bh);
2026         }
2027 }
2028
2029 static void ocfs2_unlink_subtree(struct inode *inode, handle_t *handle,
2030                                  struct ocfs2_path *left_path,
2031                                  struct ocfs2_path *right_path,
2032                                  int subtree_index,
2033                                  struct ocfs2_cached_dealloc_ctxt *dealloc)
2034 {
2035         int i;
2036         struct buffer_head *root_bh = left_path->p_node[subtree_index].bh;
2037         struct ocfs2_extent_list *root_el = left_path->p_node[subtree_index].el;
2038         struct ocfs2_extent_list *el;
2039         struct ocfs2_extent_block *eb;
2040
2041         el = path_leaf_el(left_path);
2042
2043         eb = (struct ocfs2_extent_block *)right_path->p_node[subtree_index + 1].bh->b_data;
2044
2045         for(i = 1; i < le16_to_cpu(root_el->l_next_free_rec); i++)
2046                 if (root_el->l_recs[i].e_blkno == eb->h_blkno)
2047                         break;
2048
2049         BUG_ON(i >= le16_to_cpu(root_el->l_next_free_rec));
2050
2051         memset(&root_el->l_recs[i], 0, sizeof(struct ocfs2_extent_rec));
2052         le16_add_cpu(&root_el->l_next_free_rec, -1);
2053
2054         eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data;
2055         eb->h_next_leaf_blk = 0;
2056
2057         ocfs2_journal_dirty(handle, root_bh);
2058         ocfs2_journal_dirty(handle, path_leaf_bh(left_path));
2059
2060         ocfs2_unlink_path(inode, handle, dealloc, right_path,
2061                           subtree_index + 1);
2062 }
2063
2064 static int ocfs2_rotate_subtree_left(struct inode *inode, handle_t *handle,
2065                                      struct ocfs2_path *left_path,
2066                                      struct ocfs2_path *right_path,
2067                                      int subtree_index,
2068                                      struct ocfs2_cached_dealloc_ctxt *dealloc,
2069                                      int *deleted)
2070 {
2071         int ret, i, del_right_subtree = 0, right_has_empty = 0;
2072         struct buffer_head *root_bh, *di_bh = path_root_bh(right_path);
2073         struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
2074         struct ocfs2_extent_list *right_leaf_el, *left_leaf_el;
2075         struct ocfs2_extent_block *eb;
2076
2077         *deleted = 0;
2078
2079         right_leaf_el = path_leaf_el(right_path);
2080         left_leaf_el = path_leaf_el(left_path);
2081         root_bh = left_path->p_node[subtree_index].bh;
2082         BUG_ON(root_bh != right_path->p_node[subtree_index].bh);
2083
2084         if (!ocfs2_is_empty_extent(&left_leaf_el->l_recs[0]))
2085                 return 0;
2086
2087         eb = (struct ocfs2_extent_block *)path_leaf_bh(right_path)->b_data;
2088         if (ocfs2_is_empty_extent(&right_leaf_el->l_recs[0])) {
2089                 /*
2090                  * It's legal for us to proceed if the right leaf is
2091                  * the rightmost one and it has an empty extent. There
2092                  * are two cases to handle - whether the leaf will be
2093                  * empty after removal or not. If the leaf isn't empty
2094                  * then just remove the empty extent up front. The
2095                  * next block will handle empty leaves by flagging
2096                  * them for unlink.
2097                  *
2098                  * Non rightmost leaves will throw -EAGAIN and the
2099                  * caller can manually move the subtree and retry.
2100                  */
2101
2102                 if (eb->h_next_leaf_blk != 0ULL)
2103                         return -EAGAIN;
2104
2105                 if (le16_to_cpu(right_leaf_el->l_next_free_rec) > 1) {
2106                         ret = ocfs2_journal_access(handle, inode,
2107                                                    path_leaf_bh(right_path),
2108                                                    OCFS2_JOURNAL_ACCESS_WRITE);
2109                         if (ret) {
2110                                 mlog_errno(ret);
2111                                 goto out;
2112                         }
2113
2114                         ocfs2_remove_empty_extent(right_leaf_el);
2115                 } else
2116                         right_has_empty = 1;
2117         }
2118
2119         if (eb->h_next_leaf_blk == 0ULL &&
2120             le16_to_cpu(right_leaf_el->l_next_free_rec) == 1) {
2121                 /*
2122                  * We have to update i_last_eb_blk during the meta
2123                  * data delete.
2124                  */
2125                 ret = ocfs2_journal_access(handle, inode, di_bh,
2126                                            OCFS2_JOURNAL_ACCESS_WRITE);
2127                 if (ret) {
2128                         mlog_errno(ret);
2129                         goto out;
2130                 }
2131
2132                 del_right_subtree = 1;
2133         }
2134
2135         /*
2136          * Getting here with an empty extent in the right path implies
2137          * that it's the rightmost path and will be deleted.
2138          */
2139         BUG_ON(right_has_empty && !del_right_subtree);
2140
2141         ret = ocfs2_journal_access(handle, inode, root_bh,
2142                                    OCFS2_JOURNAL_ACCESS_WRITE);
2143         if (ret) {
2144                 mlog_errno(ret);
2145                 goto out;
2146         }
2147
2148         for(i = subtree_index + 1; i < path_num_items(right_path); i++) {
2149                 ret = ocfs2_journal_access(handle, inode,
2150                                            right_path->p_node[i].bh,
2151                                            OCFS2_JOURNAL_ACCESS_WRITE);
2152                 if (ret) {
2153                         mlog_errno(ret);
2154                         goto out;
2155                 }
2156
2157                 ret = ocfs2_journal_access(handle, inode,
2158                                            left_path->p_node[i].bh,
2159                                            OCFS2_JOURNAL_ACCESS_WRITE);
2160                 if (ret) {
2161                         mlog_errno(ret);
2162                         goto out;
2163                 }
2164         }
2165
2166         if (!right_has_empty) {
2167                 /*
2168                  * Only do this if we're moving a real
2169                  * record. Otherwise, the action is delayed until
2170                  * after removal of the right path in which case we
2171                  * can do a simple shift to remove the empty extent.
2172                  */
2173                 ocfs2_rotate_leaf(left_leaf_el, &right_leaf_el->l_recs[0]);
2174                 memset(&right_leaf_el->l_recs[0], 0,
2175                        sizeof(struct ocfs2_extent_rec));
2176         }
2177         if (eb->h_next_leaf_blk == 0ULL) {
2178                 /*
2179                  * Move recs over to get rid of empty extent, decrease
2180                  * next_free. This is allowed to remove the last
2181                  * extent in our leaf (setting l_next_free_rec to
2182                  * zero) - the delete code below won't care.
2183                  */
2184                 ocfs2_remove_empty_extent(right_leaf_el);
2185         }
2186
2187         ret = ocfs2_journal_dirty(handle, path_leaf_bh(left_path));
2188         if (ret)
2189                 mlog_errno(ret);
2190         ret = ocfs2_journal_dirty(handle, path_leaf_bh(right_path));
2191         if (ret)
2192                 mlog_errno(ret);
2193
2194         if (del_right_subtree) {
2195                 ocfs2_unlink_subtree(inode, handle, left_path, right_path,
2196                                      subtree_index, dealloc);
2197                 ocfs2_update_edge_lengths(inode, handle, left_path);
2198
2199                 eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data;
2200                 di->i_last_eb_blk = eb->h_blkno;
2201
2202                 /*
2203                  * Removal of the extent in the left leaf was skipped
2204                  * above so we could delete the right path
2205                  * 1st.
2206                  */
2207                 if (right_has_empty)
2208                         ocfs2_remove_empty_extent(left_leaf_el);
2209
2210                 ret = ocfs2_journal_dirty(handle, di_bh);
2211                 if (ret)
2212                         mlog_errno(ret);
2213
2214                 *deleted = 1;
2215         } else
2216                 ocfs2_complete_edge_insert(inode, handle, left_path, right_path,
2217                                            subtree_index);
2218
2219 out:
2220         return ret;
2221 }
2222
2223 /*
2224  * Given a full path, determine what cpos value would return us a path
2225  * containing the leaf immediately to the right of the current one.
2226  *
2227  * Will return zero if the path passed in is already the rightmost path.
2228  *
2229  * This looks similar, but is subtly different to
2230  * ocfs2_find_cpos_for_left_leaf().
2231  */
2232 static int ocfs2_find_cpos_for_right_leaf(struct super_block *sb,
2233                                           struct ocfs2_path *path, u32 *cpos)
2234 {
2235         int i, j, ret = 0;
2236         u64 blkno;
2237         struct ocfs2_extent_list *el;
2238
2239         *cpos = 0;
2240
2241         if (path->p_tree_depth == 0)
2242                 return 0;
2243
2244         blkno = path_leaf_bh(path)->b_blocknr;
2245
2246         /* Start at the tree node just above the leaf and work our way up. */
2247         i = path->p_tree_depth - 1;
2248         while (i >= 0) {
2249                 int next_free;
2250
2251                 el = path->p_node[i].el;
2252
2253                 /*
2254                  * Find the extent record just after the one in our
2255                  * path.
2256                  */
2257                 next_free = le16_to_cpu(el->l_next_free_rec);
2258                 for(j = 0; j < le16_to_cpu(el->l_next_free_rec); j++) {
2259                         if (le64_to_cpu(el->l_recs[j].e_blkno) == blkno) {
2260                                 if (j == (next_free - 1)) {
2261                                         if (i == 0) {
2262                                                 /*
2263                                                  * We've determined that the
2264                                                  * path specified is already
2265                                                  * the rightmost one - return a
2266                                                  * cpos of zero.
2267                                                  */
2268                                                 goto out;
2269                                         }
2270                                         /*
2271                                          * The rightmost record points to our
2272                                          * leaf - we need to travel up the
2273                                          * tree one level.
2274                                          */
2275                                         goto next_node;
2276                                 }
2277
2278                                 *cpos = le32_to_cpu(el->l_recs[j + 1].e_cpos);
2279                                 goto out;
2280                         }
2281                 }
2282
2283                 /*
2284                  * If we got here, we never found a valid node where
2285                  * the tree indicated one should be.
2286                  */
2287                 ocfs2_error(sb,
2288                             "Invalid extent tree at extent block %llu\n",
2289                             (unsigned long long)blkno);
2290                 ret = -EROFS;
2291                 goto out;
2292
2293 next_node:
2294                 blkno = path->p_node[i].bh->b_blocknr;
2295                 i--;
2296         }
2297
2298 out:
2299         return ret;
2300 }
2301
2302 static int ocfs2_rotate_rightmost_leaf_left(struct inode *inode,
2303                                             handle_t *handle,
2304                                             struct buffer_head *bh,
2305                                             struct ocfs2_extent_list *el)
2306 {
2307         int ret;
2308
2309         if (!ocfs2_is_empty_extent(&el->l_recs[0]))
2310                 return 0;
2311
2312         ret = ocfs2_journal_access(handle, inode, bh,
2313                                    OCFS2_JOURNAL_ACCESS_WRITE);
2314         if (ret) {
2315                 mlog_errno(ret);
2316                 goto out;
2317         }
2318
2319         ocfs2_remove_empty_extent(el);
2320
2321         ret = ocfs2_journal_dirty(handle, bh);
2322         if (ret)
2323                 mlog_errno(ret);
2324
2325 out:
2326         return ret;
2327 }
2328
2329 static int __ocfs2_rotate_tree_left(struct inode *inode,
2330                                     handle_t *handle, int orig_credits,
2331                                     struct ocfs2_path *path,
2332                                     struct ocfs2_cached_dealloc_ctxt *dealloc,
2333                                     struct ocfs2_path **empty_extent_path)
2334 {
2335         int ret, subtree_root, deleted;
2336         u32 right_cpos;
2337         struct ocfs2_path *left_path = NULL;
2338         struct ocfs2_path *right_path = NULL;
2339
2340         BUG_ON(!ocfs2_is_empty_extent(&(path_leaf_el(path)->l_recs[0])));
2341
2342         *empty_extent_path = NULL;
2343
2344         ret = ocfs2_find_cpos_for_right_leaf(inode->i_sb, path,
2345                                              &right_cpos);
2346         if (ret) {
2347                 mlog_errno(ret);
2348                 goto out;
2349         }
2350
2351         left_path = ocfs2_new_path(path_root_bh(path),
2352                                    path_root_el(path));
2353         if (!left_path) {
2354                 ret = -ENOMEM;
2355                 mlog_errno(ret);
2356                 goto out;
2357         }
2358
2359         ocfs2_cp_path(left_path, path);
2360
2361         right_path = ocfs2_new_path(path_root_bh(path),
2362                                     path_root_el(path));
2363         if (!right_path) {
2364                 ret = -ENOMEM;
2365                 mlog_errno(ret);
2366                 goto out;
2367         }
2368
2369         while (right_cpos) {
2370                 ret = ocfs2_find_path(inode, right_path, right_cpos);
2371                 if (ret) {
2372                         mlog_errno(ret);
2373                         goto out;
2374                 }
2375
2376                 subtree_root = ocfs2_find_subtree_root(inode, left_path,
2377                                                        right_path);
2378
2379                 mlog(0, "Subtree root at index %d (blk %llu, depth %d)\n",
2380                      subtree_root,
2381                      (unsigned long long)
2382                      right_path->p_node[subtree_root].bh->b_blocknr,
2383                      right_path->p_tree_depth);
2384
2385                 ret = ocfs2_extend_rotate_transaction(handle, subtree_root,
2386                                                       orig_credits, left_path);
2387                 if (ret) {
2388                         mlog_errno(ret);
2389                         goto out;
2390                 }
2391
2392                 /*
2393                  * Caller might still want to make changes to the
2394                  * tree root, so re-add it to the journal here.
2395                  */
2396                 ret = ocfs2_journal_access(handle, inode,
2397                                            path_root_bh(left_path),
2398                                            OCFS2_JOURNAL_ACCESS_WRITE);
2399                 if (ret) {
2400                         mlog_errno(ret);
2401                         goto out;
2402                 }
2403
2404                 ret = ocfs2_rotate_subtree_left(inode, handle, left_path,
2405                                                 right_path, subtree_root,
2406                                                 dealloc, &deleted);
2407                 if (ret == -EAGAIN) {
2408                         /*
2409                          * The rotation has to temporarily stop due to
2410                          * the right subtree having an empty
2411                          * extent. Pass it back to the caller for a
2412                          * fixup.
2413                          */
2414                         *empty_extent_path = right_path;
2415                         right_path = NULL;
2416                         goto out;
2417                 }
2418                 if (ret) {
2419                         mlog_errno(ret);
2420                         goto out;
2421                 }
2422
2423                 /*
2424                  * The subtree rotate might have removed records on
2425                  * the rightmost edge. If so, then rotation is
2426                  * complete.
2427                  */
2428                 if (deleted)
2429                         break;
2430
2431                 ocfs2_mv_path(left_path, right_path);
2432
2433                 ret = ocfs2_find_cpos_for_right_leaf(inode->i_sb, left_path,
2434                                                      &right_cpos);
2435                 if (ret) {
2436                         mlog_errno(ret);
2437                         goto out;
2438                 }
2439         }
2440
2441 out:
2442         ocfs2_free_path(right_path);
2443         ocfs2_free_path(left_path);
2444
2445         return ret;
2446 }
2447
2448 static int ocfs2_remove_rightmost_path(struct inode *inode, handle_t *handle,
2449                                        struct ocfs2_path *path,
2450                                        struct ocfs2_cached_dealloc_ctxt *dealloc)
2451 {
2452         int ret, subtree_index;
2453         u32 cpos;
2454         struct ocfs2_path *left_path = NULL;
2455         struct ocfs2_dinode *di;
2456         struct ocfs2_extent_block *eb;
2457         struct ocfs2_extent_list *el;
2458
2459         /*
2460          * XXX: This code assumes that the root is an inode, which is
2461          * true for now but may change as tree code gets generic.
2462          */
2463         di = (struct ocfs2_dinode *)path_root_bh(path)->b_data;
2464         if (!OCFS2_IS_VALID_DINODE(di)) {
2465                 ret = -EIO;
2466                 ocfs2_error(inode->i_sb,
2467                             "Inode %llu has invalid path root",
2468                             (unsigned long long)OCFS2_I(inode)->ip_blkno);
2469                 goto out;
2470         }
2471
2472         /*
2473          * There's two ways we handle this depending on
2474          * whether path is the only existing one.
2475          */
2476         ret = ocfs2_extend_rotate_transaction(handle, 0,
2477                                               handle->h_buffer_credits,
2478                                               path);
2479         if (ret) {
2480                 mlog_errno(ret);
2481                 goto out;
2482         }
2483
2484         ret = ocfs2_journal_access_path(inode, handle, path);
2485         if (ret) {
2486                 mlog_errno(ret);
2487                 goto out;
2488         }
2489
2490         ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, path, &cpos);
2491         if (ret) {
2492                 mlog_errno(ret);
2493                 goto out;
2494         }
2495
2496         if (cpos) {
2497                 /*
2498                  * We have a path to the left of this one - it needs
2499                  * an update too.
2500                  */
2501                 left_path = ocfs2_new_path(path_root_bh(path),
2502                                            path_root_el(path));
2503                 if (!left_path) {
2504                         ret = -ENOMEM;
2505                         mlog_errno(ret);
2506                         goto out;
2507                 }
2508
2509                 ret = ocfs2_find_path(inode, left_path, cpos);
2510                 if (ret) {
2511                         mlog_errno(ret);
2512                         goto out;
2513                 }
2514
2515                 ret = ocfs2_journal_access_path(inode, handle, left_path);
2516                 if (ret) {
2517                         mlog_errno(ret);
2518                         goto out;
2519                 }
2520
2521                 subtree_index = ocfs2_find_subtree_root(inode, left_path, path);
2522
2523                 ocfs2_unlink_subtree(inode, handle, left_path, path,
2524                                      subtree_index, dealloc);
2525                 ocfs2_update_edge_lengths(inode, handle, left_path);
2526
2527                 eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data;
2528                 di->i_last_eb_blk = eb->h_blkno;
2529         } else {
2530                 /*
2531                  * 'path' is also the leftmost path which
2532                  * means it must be the only one. This gets
2533                  * handled differently because we want to
2534                  * revert the inode back to having extents
2535                  * in-line.
2536                  */
2537                 ocfs2_unlink_path(inode, handle, dealloc, path, 1);
2538
2539                 el = &di->id2.i_list;
2540                 el->l_tree_depth = 0;
2541                 el->l_next_free_rec = 0;
2542                 memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec));
2543
2544                 di->i_last_eb_blk = 0;
2545         }
2546
2547         ocfs2_journal_dirty(handle, path_root_bh(path));
2548
2549 out:
2550         ocfs2_free_path(left_path);
2551         return ret;
2552 }
2553
2554 /*
2555  * Left rotation of btree records.
2556  *
2557  * In many ways, this is (unsurprisingly) the opposite of right
2558  * rotation. We start at some non-rightmost path containing an empty
2559  * extent in the leaf block. The code works its way to the rightmost
2560  * path by rotating records to the left in every subtree.
2561  *
2562  * This is used by any code which reduces the number of extent records
2563  * in a leaf. After removal, an empty record should be placed in the
2564  * leftmost list position.
2565  *
2566  * This won't handle a length update of the rightmost path records if
2567  * the rightmost tree leaf record is removed so the caller is
2568  * responsible for detecting and correcting that.
2569  */
2570 static int ocfs2_rotate_tree_left(struct inode *inode, handle_t *handle,
2571                                   struct ocfs2_path *path,
2572                                   struct ocfs2_cached_dealloc_ctxt *dealloc)
2573 {
2574         int ret, orig_credits = handle->h_buffer_credits;
2575         struct ocfs2_path *tmp_path = NULL, *restart_path = NULL;
2576         struct ocfs2_extent_block *eb;
2577         struct ocfs2_extent_list *el;
2578
2579         el = path_leaf_el(path);
2580         if (!ocfs2_is_empty_extent(&el->l_recs[0]))
2581                 return 0;
2582
2583         if (path->p_tree_depth == 0) {
2584 rightmost_no_delete:
2585                 /*
2586                  * In-inode extents. This is trivially handled, so do
2587                  * it up front.
2588                  */
2589                 ret = ocfs2_rotate_rightmost_leaf_left(inode, handle,
2590                                                        path_leaf_bh(path),
2591                                                        path_leaf_el(path));
2592                 if (ret)
2593                         mlog_errno(ret);
2594                 goto out;
2595         }
2596
2597         /*
2598          * Handle rightmost branch now. There's several cases:
2599          *  1) simple rotation leaving records in there. That's trivial.
2600          *  2) rotation requiring a branch delete - there's no more
2601          *     records left. Two cases of this:
2602          *     a) There are branches to the left.
2603          *     b) This is also the leftmost (the only) branch.
2604          *
2605          *  1) is handled via ocfs2_rotate_rightmost_leaf_left()
2606          *  2a) we need the left branch so that we can update it with the unlink
2607          *  2b) we need to bring the inode back to inline extents.
2608          */
2609
2610         eb = (struct ocfs2_extent_block *)path_leaf_bh(path)->b_data;
2611         el = &eb->h_list;
2612         if (eb->h_next_leaf_blk == 0) {
2613                 /*
2614                  * This gets a bit tricky if we're going to delete the
2615                  * rightmost path. Get the other cases out of the way
2616                  * 1st.
2617                  */
2618                 if (le16_to_cpu(el->l_next_free_rec) > 1)
2619                         goto rightmost_no_delete;
2620
2621                 if (le16_to_cpu(el->l_next_free_rec) == 0) {
2622                         ret = -EIO;
2623                         ocfs2_error(inode->i_sb,
2624                                     "Inode %llu has empty extent block at %llu",
2625                                     (unsigned long long)OCFS2_I(inode)->ip_blkno,
2626                                     (unsigned long long)le64_to_cpu(eb->h_blkno));
2627                         goto out;
2628                 }
2629
2630                 /*
2631                  * XXX: The caller can not trust "path" any more after
2632                  * this as it will have been deleted. What do we do?
2633                  *
2634                  * In theory the rotate-for-merge code will never get
2635                  * here because it'll always ask for a rotate in a
2636                  * nonempty list.
2637                  */
2638
2639                 ret = ocfs2_remove_rightmost_path(inode, handle, path,
2640                                                   dealloc);
2641                 if (ret)
2642                         mlog_errno(ret);
2643                 goto out;
2644         }
2645
2646         /*
2647          * Now we can loop, remembering the path we get from -EAGAIN
2648          * and restarting from there.
2649          */
2650 try_rotate:
2651         ret = __ocfs2_rotate_tree_left(inode, handle, orig_credits, path,
2652                                        dealloc, &restart_path);
2653         if (ret && ret != -EAGAIN) {
2654                 mlog_errno(ret);
2655                 goto out;
2656         }
2657
2658         while (ret == -EAGAIN) {
2659                 tmp_path = restart_path;
2660                 restart_path = NULL;
2661
2662                 ret = __ocfs2_rotate_tree_left(inode, handle, orig_credits,
2663                                                tmp_path, dealloc,
2664                                                &restart_path);
2665                 if (ret && ret != -EAGAIN) {
2666                         mlog_errno(ret);
2667                         goto out;
2668                 }
2669
2670                 ocfs2_free_path(tmp_path);
2671                 tmp_path = NULL;
2672
2673                 if (ret == 0)
2674                         goto try_rotate;
2675         }
2676
2677 out:
2678         ocfs2_free_path(tmp_path);
2679         ocfs2_free_path(restart_path);
2680         return ret;
2681 }
2682
2683 static void ocfs2_cleanup_merge(struct ocfs2_extent_list *el,
2684                                 int index)
2685 {
2686         struct ocfs2_extent_rec *rec = &el->l_recs[index];
2687         unsigned int size;
2688
2689         if (rec->e_leaf_clusters == 0) {
2690                 /*
2691                  * We consumed all of the merged-from record. An empty
2692                  * extent cannot exist anywhere but the 1st array
2693                  * position, so move things over if the merged-from
2694                  * record doesn't occupy that position.
2695                  *
2696                  * This creates a new empty extent so the caller
2697                  * should be smart enough to have removed any existing
2698                  * ones.
2699                  */
2700                 if (index > 0) {
2701                         BUG_ON(ocfs2_is_empty_extent(&el->l_recs[0]));
2702                         size = index * sizeof(struct ocfs2_extent_rec);
2703                         memmove(&el->l_recs[1], &el->l_recs[0], size);
2704                 }
2705
2706                 /*
2707                  * Always memset - the caller doesn't check whether it
2708                  * created an empty extent, so there could be junk in
2709                  * the other fields.
2710                  */
2711                 memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec));
2712         }
2713 }
2714
2715 /*
2716  * Remove split_rec clusters from the record at index and merge them
2717  * onto the beginning of the record at index + 1.
2718  */
2719 static int ocfs2_merge_rec_right(struct inode *inode, struct buffer_head *bh,
2720                                 handle_t *handle,
2721                                 struct ocfs2_extent_rec *split_rec,
2722                                 struct ocfs2_extent_list *el, int index)
2723 {
2724         int ret;
2725         unsigned int split_clusters = le16_to_cpu(split_rec->e_leaf_clusters);
2726         struct ocfs2_extent_rec *left_rec;
2727         struct ocfs2_extent_rec *right_rec;
2728
2729         BUG_ON(index >= le16_to_cpu(el->l_next_free_rec));
2730
2731         left_rec = &el->l_recs[index];
2732         right_rec = &el->l_recs[index + 1];
2733
2734         ret = ocfs2_journal_access(handle, inode, bh,
2735                                    OCFS2_JOURNAL_ACCESS_WRITE);
2736         if (ret) {
2737                 mlog_errno(ret);
2738                 goto out;
2739         }
2740
2741         le16_add_cpu(&left_rec->e_leaf_clusters, -split_clusters);
2742
2743         le32_add_cpu(&right_rec->e_cpos, -split_clusters);
2744         le64_add_cpu(&right_rec->e_blkno,
2745                      -ocfs2_clusters_to_blocks(inode->i_sb, split_clusters));
2746         le16_add_cpu(&right_rec->e_leaf_clusters, split_clusters);
2747
2748         ocfs2_cleanup_merge(el, index);
2749
2750         ret = ocfs2_journal_dirty(handle, bh);
2751         if (ret)
2752                 mlog_errno(ret);
2753
2754 out:
2755         return ret;
2756 }
2757
2758 /*
2759  * Remove split_rec clusters from the record at index and merge them
2760  * onto the tail of the record at index - 1.
2761  */
2762 static int ocfs2_merge_rec_left(struct inode *inode, struct buffer_head *bh,
2763                                 handle_t *handle,
2764                                 struct ocfs2_extent_rec *split_rec,
2765                                 struct ocfs2_extent_list *el, int index)
2766 {
2767         int ret, has_empty_extent = 0;
2768         unsigned int split_clusters = le16_to_cpu(split_rec->e_leaf_clusters);
2769         struct ocfs2_extent_rec *left_rec;
2770         struct ocfs2_extent_rec *right_rec;
2771
2772         BUG_ON(index <= 0);
2773
2774         left_rec = &el->l_recs[index - 1];
2775         right_rec = &el->l_recs[index];
2776         if (ocfs2_is_empty_extent(&el->l_recs[0]))
2777                 has_empty_extent = 1;
2778
2779         ret = ocfs2_journal_access(handle, inode, bh,
2780                                    OCFS2_JOURNAL_ACCESS_WRITE);
2781         if (ret) {
2782                 mlog_errno(ret);
2783                 goto out;
2784         }
2785
2786         if (has_empty_extent && index == 1) {
2787                 /*
2788                  * The easy case - we can just plop the record right in.
2789                  */
2790                 *left_rec = *split_rec;
2791
2792                 has_empty_extent = 0;
2793         } else {
2794                 le16_add_cpu(&left_rec->e_leaf_clusters, split_clusters);
2795         }
2796
2797         le32_add_cpu(&right_rec->e_cpos, split_clusters);
2798         le64_add_cpu(&right_rec->e_blkno,
2799                      ocfs2_clusters_to_blocks(inode->i_sb, split_clusters));
2800         le16_add_cpu(&right_rec->e_leaf_clusters, -split_clusters);
2801
2802         ocfs2_cleanup_merge(el, index);
2803
2804         ret = ocfs2_journal_dirty(handle, bh);
2805         if (ret)
2806                 mlog_errno(ret);
2807
2808 out:
2809         return ret;
2810 }
2811
2812 static int ocfs2_try_to_merge_extent(struct inode *inode,
2813                                      handle_t *handle,
2814                                      struct ocfs2_path *left_path,
2815                                      int split_index,
2816                                      struct ocfs2_extent_rec *split_rec,
2817                                      struct ocfs2_cached_dealloc_ctxt *dealloc,
2818                                      struct ocfs2_merge_ctxt *ctxt)
2819
2820 {
2821         int ret = 0;
2822         struct ocfs2_extent_list *el = path_leaf_el(left_path);
2823         struct ocfs2_extent_rec *rec = &el->l_recs[split_index];
2824
2825         BUG_ON(ctxt->c_contig_type == CONTIG_NONE);
2826
2827         if (ctxt->c_split_covers_rec && ctxt->c_has_empty_extent) {
2828                 /*
2829                  * The merge code will need to create an empty
2830                  * extent to take the place of the newly
2831                  * emptied slot. Remove any pre-existing empty
2832                  * extents - having more than one in a leaf is
2833                  * illegal.
2834                  */
2835                 ret = ocfs2_rotate_tree_left(inode, handle, left_path,
2836                                              dealloc);
2837                 if (ret) {
2838                         mlog_errno(ret);
2839                         goto out;
2840                 }
2841                 split_index--;
2842                 rec = &el->l_recs[split_index];
2843         }
2844
2845         if (ctxt->c_contig_type == CONTIG_LEFTRIGHT) {
2846                 /*
2847                  * Left-right contig implies this.
2848                  */
2849                 BUG_ON(!ctxt->c_split_covers_rec);
2850                 BUG_ON(split_index == 0);
2851
2852                 /*
2853                  * Since the leftright insert always covers the entire
2854                  * extent, this call will delete the insert record
2855                  * entirely, resulting in an empty extent record added to
2856                  * the extent block.
2857                  *
2858                  * Since the adding of an empty extent shifts
2859                  * everything back to the right, there's no need to
2860                  * update split_index here.
2861                  */
2862                 ret = ocfs2_merge_rec_left(inode, path_leaf_bh(left_path),
2863                                            handle, split_rec, el, split_index);
2864                 if (ret) {
2865                         mlog_errno(ret);
2866                         goto out;
2867                 }
2868
2869                 /*
2870                  * We can only get this from logic error above.
2871                  */
2872                 BUG_ON(!ocfs2_is_empty_extent(&el->l_recs[0]));
2873
2874                 /*
2875                  * The left merge left us with an empty extent, remove
2876                  * it.
2877                  */
2878                 ret = ocfs2_rotate_tree_left(inode, handle, left_path, dealloc);
2879                 if (ret) {
2880                         mlog_errno(ret);
2881                         goto out;
2882                 }
2883                 split_index--;
2884                 rec = &el->l_recs[split_index];
2885
2886                 /*
2887                  * Note that we don't pass split_rec here on purpose -
2888                  * we've merged it into the left side.
2889                  */
2890                 ret = ocfs2_merge_rec_right(inode, path_leaf_bh(left_path),
2891                                             handle, rec, el, split_index);
2892                 if (ret) {
2893                         mlog_errno(ret);
2894                         goto out;
2895                 }
2896
2897                 BUG_ON(!ocfs2_is_empty_extent(&el->l_recs[0]));
2898
2899                 ret = ocfs2_rotate_tree_left(inode, handle, left_path,
2900                                              dealloc);
2901                 /*
2902                  * Error from this last rotate is not critical, so
2903                  * print but don't bubble it up.
2904                  */
2905                 if (ret)
2906                         mlog_errno(ret);
2907                 ret = 0;
2908         } else {
2909                 /*
2910                  * Merge a record to the left or right.
2911                  *
2912                  * 'contig_type' is relative to the existing record,
2913                  * so for example, if we're "right contig", it's to
2914                  * the record on the left (hence the left merge).
2915                  */
2916                 if (ctxt->c_contig_type == CONTIG_RIGHT) {
2917                         ret = ocfs2_merge_rec_left(inode,
2918                                                    path_leaf_bh(left_path),
2919                                                    handle, split_rec, el,
2920                                                    split_index);
2921                         if (ret) {
2922                                 mlog_errno(ret);
2923                                 goto out;
2924                         }
2925                 } else {
2926                         ret = ocfs2_merge_rec_right(inode,
2927                                                     path_leaf_bh(left_path),
2928                                                     handle, split_rec, el,
2929                                                     split_index);
2930                         if (ret) {
2931                                 mlog_errno(ret);
2932                                 goto out;
2933                         }
2934                 }
2935
2936                 if (ctxt->c_split_covers_rec) {
2937                         /*
2938                          * The merge may have left an empty extent in
2939                          * our leaf. Try to rotate it away.
2940                          */
2941                         ret = ocfs2_rotate_tree_left(inode, handle, left_path,
2942                                                      dealloc);
2943                         if (ret)
2944                                 mlog_errno(ret);
2945                         ret = 0;
2946                 }
2947         }
2948
2949 out:
2950         return ret;
2951 }
2952
2953 static void ocfs2_subtract_from_rec(struct super_block *sb,
2954                                     enum ocfs2_split_type split,
2955                                     struct ocfs2_extent_rec *rec,
2956                                     struct ocfs2_extent_rec *split_rec)
2957 {
2958         u64 len_blocks;
2959
2960         len_blocks = ocfs2_clusters_to_blocks(sb,
2961                                 le16_to_cpu(split_rec->e_leaf_clusters));
2962
2963         if (split == SPLIT_LEFT) {
2964                 /*
2965                  * Region is on the left edge of the existing
2966                  * record.
2967                  */
2968                 le32_add_cpu(&rec->e_cpos,
2969                              le16_to_cpu(split_rec->e_leaf_clusters));
2970                 le64_add_cpu(&rec->e_blkno, len_blocks);
2971                 le16_add_cpu(&rec->e_leaf_clusters,
2972                              -le16_to_cpu(split_rec->e_leaf_clusters));
2973         } else {
2974                 /*
2975                  * Region is on the right edge of the existing
2976                  * record.
2977                  */
2978                 le16_add_cpu(&rec->e_leaf_clusters,
2979                              -le16_to_cpu(split_rec->e_leaf_clusters));
2980         }
2981 }
2982
2983 /*
2984  * Do the final bits of extent record insertion at the target leaf
2985  * list. If this leaf is part of an allocation tree, it is assumed
2986  * that the tree above has been prepared.
2987  */
2988 static void ocfs2_insert_at_leaf(struct ocfs2_extent_rec *insert_rec,
2989                                  struct ocfs2_extent_list *el,
2990                                  struct ocfs2_insert_type *insert,
2991                                  struct inode *inode)
2992 {
2993         int i = insert->ins_contig_index;
2994         unsigned int range;
2995         struct ocfs2_extent_rec *rec;
2996
2997         BUG_ON(le16_to_cpu(el->l_tree_depth) != 0);
2998
2999         if (insert->ins_split != SPLIT_NONE) {
3000                 i = ocfs2_search_extent_list(el, le32_to_cpu(insert_rec->e_cpos));
3001                 BUG_ON(i == -1);
3002                 rec = &el->l_recs[i];
3003                 ocfs2_subtract_from_rec(inode->i_sb, insert->ins_split, rec,
3004                                         insert_rec);
3005                 goto rotate;
3006         }
3007
3008         /*
3009          * Contiguous insert - either left or right.
3010          */
3011         if (insert->ins_contig != CONTIG_NONE) {
3012                 rec = &el->l_recs[i];
3013                 if (insert->ins_contig == CONTIG_LEFT) {
3014                         rec->e_blkno = insert_rec->e_blkno;
3015                         rec->e_cpos = insert_rec->e_cpos;
3016                 }
3017                 le16_add_cpu(&rec->e_leaf_clusters,
3018                              le16_to_cpu(insert_rec->e_leaf_clusters));
3019                 return;
3020         }
3021
3022         /*
3023          * Handle insert into an empty leaf.
3024          */
3025         if (le16_to_cpu(el->l_next_free_rec) == 0 ||
3026             ((le16_to_cpu(el->l_next_free_rec) == 1) &&
3027              ocfs2_is_empty_extent(&el->l_recs[0]))) {
3028                 el->l_recs[0] = *insert_rec;
3029                 el->l_next_free_rec = cpu_to_le16(1);
3030                 return;
3031         }
3032
3033         /*
3034          * Appending insert.
3035          */
3036         if (insert->ins_appending == APPEND_TAIL) {
3037                 i = le16_to_cpu(el->l_next_free_rec) - 1;
3038                 rec = &el->l_recs[i];
3039                 range = le32_to_cpu(rec->e_cpos)
3040                         + le16_to_cpu(rec->e_leaf_clusters);
3041                 BUG_ON(le32_to_cpu(insert_rec->e_cpos) < range);
3042
3043                 mlog_bug_on_msg(le16_to_cpu(el->l_next_free_rec) >=
3044                                 le16_to_cpu(el->l_count),
3045                                 "inode %lu, depth %u, count %u, next free %u, "
3046                                 "rec.cpos %u, rec.clusters %u, "
3047                                 "insert.cpos %u, insert.clusters %u\n",
3048                                 inode->i_ino,
3049                                 le16_to_cpu(el->l_tree_depth),
3050                                 le16_to_cpu(el->l_count),
3051                                 le16_to_cpu(el->l_next_free_rec),
3052                                 le32_to_cpu(el->l_recs[i].e_cpos),
3053                                 le16_to_cpu(el->l_recs[i].e_leaf_clusters),
3054                                 le32_to_cpu(insert_rec->e_cpos),
3055                                 le16_to_cpu(insert_rec->e_leaf_clusters));
3056                 i++;
3057                 el->l_recs[i] = *insert_rec;
3058                 le16_add_cpu(&el->l_next_free_rec, 1);
3059                 return;
3060         }
3061
3062 rotate:
3063         /*
3064          * Ok, we have to rotate.
3065          *
3066          * At this point, it is safe to assume that inserting into an
3067          * empty leaf and appending to a leaf have both been handled
3068          * above.
3069          *
3070          * This leaf needs to have space, either by the empty 1st
3071          * extent record, or by virtue of an l_next_rec < l_count.
3072          */
3073         ocfs2_rotate_leaf(el, insert_rec);
3074 }
3075
3076 static inline void ocfs2_update_dinode_clusters(struct inode *inode,
3077                                                 struct ocfs2_dinode *di,
3078                                                 u32 clusters)
3079 {
3080         le32_add_cpu(&di->i_clusters, clusters);
3081         spin_lock(&OCFS2_I(inode)->ip_lock);
3082         OCFS2_I(inode)->ip_clusters = le32_to_cpu(di->i_clusters);
3083         spin_unlock(&OCFS2_I(inode)->ip_lock);
3084 }
3085
3086 static void ocfs2_adjust_rightmost_records(struct inode *inode,
3087                                            handle_t *handle,
3088                                            struct ocfs2_path *path,
3089                                            struct ocfs2_extent_rec *insert_rec)
3090 {
3091         int ret, i, next_free;
3092         struct buffer_head *bh;
3093         struct ocfs2_extent_list *el;
3094         struct ocfs2_extent_rec *rec;
3095
3096         /*
3097          * Update everything except the leaf block.
3098          */
3099         for (i = 0; i < path->p_tree_depth; i++) {
3100                 bh = path->p_node[i].bh;
3101                 el = path->p_node[i].el;
3102
3103                 next_free = le16_to_cpu(el->l_next_free_rec);
3104                 if (next_free == 0) {
3105                         ocfs2_error(inode->i_sb,
3106                                     "Dinode %llu has a bad extent list",
3107                                     (unsigned long long)OCFS2_I(inode)->ip_blkno);
3108                         ret = -EIO;
3109                         return;
3110                 }
3111
3112                 rec = &el->l_recs[next_free - 1];
3113
3114                 rec->e_int_clusters = insert_rec->e_cpos;
3115                 le32_add_cpu(&rec->e_int_clusters,
3116                              le16_to_cpu(insert_rec->e_leaf_clusters));
3117                 le32_add_cpu(&rec->e_int_clusters,
3118                              -le32_to_cpu(rec->e_cpos));
3119
3120                 ret = ocfs2_journal_dirty(handle, bh);
3121                 if (ret)
3122                         mlog_errno(ret);
3123
3124         }
3125 }
3126
3127 static int ocfs2_append_rec_to_path(struct inode *inode, handle_t *handle,
3128                                     struct ocfs2_extent_rec *insert_rec,
3129                                     struct ocfs2_path *right_path,
3130                                     struct ocfs2_path **ret_left_path)
3131 {
3132         int ret, next_free;
3133         struct ocfs2_extent_list *el;
3134         struct ocfs2_path *left_path = NULL;
3135
3136         *ret_left_path = NULL;
3137
3138         /*
3139          * This shouldn't happen for non-trees. The extent rec cluster
3140          * count manipulation below only works for interior nodes.
3141          */
3142         BUG_ON(right_path->p_tree_depth == 0);
3143
3144         /*
3145          * If our appending insert is at the leftmost edge of a leaf,
3146          * then we might need to update the rightmost records of the
3147          * neighboring path.
3148          */
3149         el = path_leaf_el(right_path);
3150         next_free = le16_to_cpu(el->l_next_free_rec);
3151         if (next_free == 0 ||
3152             (next_free == 1 && ocfs2_is_empty_extent(&el->l_recs[0]))) {
3153                 u32 left_cpos;
3154
3155                 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path,
3156                                                     &left_cpos);
3157                 if (ret) {
3158                         mlog_errno(ret);
3159                         goto out;
3160                 }
3161
3162                 mlog(0, "Append may need a left path update. cpos: %u, "
3163                      "left_cpos: %u\n", le32_to_cpu(insert_rec->e_cpos),
3164                      left_cpos);
3165
3166                 /*
3167                  * No need to worry if the append is already in the
3168                  * leftmost leaf.
3169                  */
3170                 if (left_cpos) {
3171                         left_path = ocfs2_new_path(path_root_bh(right_path),
3172                                                    path_root_el(right_path));
3173                         if (!left_path) {
3174                                 ret = -ENOMEM;
3175                                 mlog_errno(ret);
3176                                 goto out;
3177                         }
3178
3179                         ret = ocfs2_find_path(inode, left_path, left_cpos);
3180                         if (ret) {
3181                                 mlog_errno(ret);
3182                                 goto out;
3183                         }
3184
3185                         /*
3186                          * ocfs2_insert_path() will pass the left_path to the
3187                          * journal for us.
3188                          */
3189                 }
3190         }
3191
3192         ret = ocfs2_journal_access_path(inode, handle, right_path);
3193         if (ret) {
3194                 mlog_errno(ret);
3195                 goto out;
3196         }
3197
3198         ocfs2_adjust_rightmost_records(inode, handle, right_path, insert_rec);
3199
3200         *ret_left_path = left_path;
3201         ret = 0;
3202 out:
3203         if (ret != 0)
3204                 ocfs2_free_path(left_path);
3205
3206         return ret;
3207 }
3208
3209 static void ocfs2_split_record(struct inode *inode,
3210                                struct ocfs2_path *left_path,
3211                                struct ocfs2_path *right_path,
3212                                struct ocfs2_extent_rec *split_rec,
3213                                enum ocfs2_split_type split)
3214 {
3215         int index;
3216         u32 cpos = le32_to_cpu(split_rec->e_cpos);
3217         struct ocfs2_extent_list *left_el = NULL, *right_el, *insert_el, *el;
3218         struct ocfs2_extent_rec *rec, *tmprec;
3219
3220         right_el = path_leaf_el(right_path);;
3221         if (left_path)
3222                 left_el = path_leaf_el(left_path);
3223
3224         el = right_el;
3225         insert_el = right_el;
3226         index = ocfs2_search_extent_list(el, cpos);
3227         if (index != -1) {
3228                 if (index == 0 && left_path) {
3229                         BUG_ON(ocfs2_is_empty_extent(&el->l_recs[0]));
3230
3231                         /*
3232                          * This typically means that the record
3233                          * started in the left path but moved to the
3234                          * right as a result of rotation. We either
3235                          * move the existing record to the left, or we
3236                          * do the later insert there.
3237                          *
3238                          * In this case, the left path should always
3239                          * exist as the rotate code will have passed
3240                          * it back for a post-insert update.
3241                          */
3242
3243                         if (split == SPLIT_LEFT) {
3244                                 /*
3245                                  * It's a left split. Since we know
3246                                  * that the rotate code gave us an
3247                                  * empty extent in the left path, we
3248                                  * can just do the insert there.
3249                                  */
3250                                 insert_el = left_el;
3251                         } else {
3252                                 /*
3253                                  * Right split - we have to move the
3254                                  * existing record over to the left
3255                                  * leaf. The insert will be into the
3256                                  * newly created empty extent in the
3257                                  * right leaf.
3258                                  */
3259                                 tmprec = &right_el->l_recs[index];
3260                                 ocfs2_rotate_leaf(left_el, tmprec);
3261                                 el = left_el;
3262
3263                                 memset(tmprec, 0, sizeof(*tmprec));
3264                                 index = ocfs2_search_extent_list(left_el, cpos);
3265                                 BUG_ON(index == -1);
3266                         }
3267                 }
3268         } else {
3269                 BUG_ON(!left_path);
3270                 BUG_ON(!ocfs2_is_empty_extent(&left_el->l_recs[0]));
3271                 /*
3272                  * Left path is easy - we can just allow the insert to
3273                  * happen.
3274                  */
3275                 el = left_el;
3276                 insert_el = left_el;
3277                 index = ocfs2_search_extent_list(el, cpos);
3278                 BUG_ON(index == -1);
3279         }
3280
3281         rec = &el->l_recs[index];
3282         ocfs2_subtract_from_rec(inode->i_sb, split, rec, split_rec);
3283         ocfs2_rotate_leaf(insert_el, split_rec);
3284 }
3285
3286 /*
3287  * This function only does inserts on an allocation b-tree. For dinode
3288  * lists, ocfs2_insert_at_leaf() is called directly.
3289  *
3290  * right_path is the path we want to do the actual insert
3291  * in. left_path should only be passed in if we need to update that
3292  * portion of the tree after an edge insert.
3293  */
3294 static int ocfs2_insert_path(struct inode *inode,
3295                              handle_t *handle,
3296                              struct ocfs2_path *left_path,
3297                              struct ocfs2_path *right_path,
3298                              struct ocfs2_extent_rec *insert_rec,
3299                              struct ocfs2_insert_type *insert)
3300 {
3301         int ret, subtree_index;
3302         struct buffer_head *leaf_bh = path_leaf_bh(right_path);
3303
3304         if (left_path) {
3305                 int credits = handle->h_buffer_credits;
3306
3307                 /*
3308                  * There's a chance that left_path got passed back to
3309                  * us without being accounted for in the
3310                  * journal. Extend our transaction here to be sure we
3311                  * can change those blocks.
3312                  */
3313                 credits += left_path->p_tree_depth;
3314
3315                 ret = ocfs2_extend_trans(handle, credits);
3316                 if (ret < 0) {
3317                         mlog_errno(ret);
3318                         goto out;
3319                 }
3320
3321                 ret = ocfs2_journal_access_path(inode, handle, left_path);
3322                 if (ret < 0) {
3323                         mlog_errno(ret);
3324                         goto out;
3325                 }
3326         }
3327
3328         /*
3329          * Pass both paths to the journal. The majority of inserts
3330          * will be touching all components anyway.
3331          */
3332         ret = ocfs2_journal_access_path(inode, handle, right_path);
3333         if (ret < 0) {
3334                 mlog_errno(ret);
3335                 goto out;
3336         }
3337
3338         if (insert->ins_split != SPLIT_NONE) {
3339                 /*
3340                  * We could call ocfs2_insert_at_leaf() for some types
3341                  * of splits, but it's easier to just let one separate
3342                  * function sort it all out.
3343                  */
3344                 ocfs2_split_record(inode, left_path, right_path,
3345                                    insert_rec, insert->ins_split);
3346
3347                 /*
3348                  * Split might have modified either leaf and we don't
3349                  * have a guarantee that the later edge insert will
3350                  * dirty this for us.
3351                  */
3352                 if (left_path)
3353                         ret = ocfs2_journal_dirty(handle,
3354                                                   path_leaf_bh(left_path));
3355                         if (ret)
3356                                 mlog_errno(ret);
3357         } else
3358                 ocfs2_insert_at_leaf(insert_rec, path_leaf_el(right_path),
3359                                      insert, inode);
3360
3361         ret = ocfs2_journal_dirty(handle, leaf_bh);
3362         if (ret)
3363                 mlog_errno(ret);
3364
3365         if (left_path) {
3366                 /*
3367                  * The rotate code has indicated that we need to fix
3368                  * up portions of the tree after the insert.
3369                  *
3370                  * XXX: Should we extend the transaction here?
3371                  */
3372                 subtree_index = ocfs2_find_subtree_root(inode, left_path,
3373                                                         right_path);
3374                 ocfs2_complete_edge_insert(inode, handle, left_path,
3375                                            right_path, subtree_index);
3376         }
3377
3378         ret = 0;
3379 out:
3380         return ret;
3381 }
3382
3383 static int ocfs2_do_insert_extent(struct inode *inode,
3384                                   handle_t *handle,
3385                                   struct buffer_head *di_bh,
3386                                   struct ocfs2_extent_rec *insert_rec,
3387                                   struct ocfs2_insert_type *type)
3388 {
3389         int ret, rotate = 0;
3390         u32 cpos;
3391         struct ocfs2_path *right_path = NULL;
3392         struct ocfs2_path *left_path = NULL;
3393         struct ocfs2_dinode *di;
3394         struct ocfs2_extent_list *el;
3395
3396         di = (struct ocfs2_dinode *) di_bh->b_data;
3397         el = &di->id2.i_list;
3398
3399         ret = ocfs2_journal_access(handle, inode, di_bh,
3400                                    OCFS2_JOURNAL_ACCESS_WRITE);
3401         if (ret) {
3402                 mlog_errno(ret);
3403                 goto out;
3404         }
3405
3406         if (le16_to_cpu(el->l_tree_depth) == 0) {
3407                 ocfs2_insert_at_leaf(insert_rec, el, type, inode);
3408                 goto out_update_clusters;
3409         }
3410
3411         right_path = ocfs2_new_inode_path(di_bh);
3412         if (!right_path) {
3413                 ret = -ENOMEM;
3414                 mlog_errno(ret);
3415                 goto out;
3416         }
3417
3418         /*
3419          * Determine the path to start with. Rotations need the
3420          * rightmost path, everything else can go directly to the
3421          * target leaf.
3422          */
3423         cpos = le32_to_cpu(insert_rec->e_cpos);
3424         if (type->ins_appending == APPEND_NONE &&
3425             type->ins_contig == CONTIG_NONE) {
3426                 rotate = 1;
3427                 cpos = UINT_MAX;
3428         }
3429
3430         ret = ocfs2_find_path(inode, right_path, cpos);
3431         if (ret) {
3432                 mlog_errno(ret);
3433                 goto out;
3434         }
3435
3436         /*
3437          * Rotations and appends need special treatment - they modify
3438          * parts of the tree's above them.
3439          *
3440          * Both might pass back a path immediate to the left of the
3441          * one being inserted to. This will be cause
3442          * ocfs2_insert_path() to modify the rightmost records of
3443          * left_path to account for an edge insert.
3444          *
3445          * XXX: When modifying this code, keep in mind that an insert
3446          * can wind up skipping both of these two special cases...
3447          */
3448         if (rotate) {
3449                 ret = ocfs2_rotate_tree_right(inode, handle, type->ins_split,
3450                                               le32_to_cpu(insert_rec->e_cpos),
3451                                               right_path, &left_path);
3452                 if (ret) {
3453                         mlog_errno(ret);
3454                         goto out;
3455                 }
3456
3457                 /*
3458                  * ocfs2_rotate_tree_right() might have extended the
3459                  * transaction without re-journaling our tree root.
3460                  */
3461                 ret = ocfs2_journal_access(handle, inode, di_bh,
3462                                            OCFS2_JOURNAL_ACCESS_WRITE);
3463                 if (ret) {
3464                         mlog_errno(ret);
3465                         goto out;
3466                 }
3467         } else if (type->ins_appending == APPEND_TAIL
3468                    && type->ins_contig != CONTIG_LEFT) {
3469                 ret = ocfs2_append_rec_to_path(inode, handle, insert_rec,
3470                                                right_path, &left_path);
3471                 if (ret) {
3472                         mlog_errno(ret);
3473                         goto out;
3474                 }
3475         }
3476
3477         ret = ocfs2_insert_path(inode, handle, left_path, right_path,
3478                                 insert_rec, type);
3479         if (ret) {
3480                 mlog_errno(ret);
3481                 goto out;
3482         }
3483
3484 out_update_clusters:
3485         if (type->ins_split == SPLIT_NONE)
3486                 ocfs2_update_dinode_clusters(inode, di,
3487                                              le16_to_cpu(insert_rec->e_leaf_clusters));
3488
3489         ret = ocfs2_journal_dirty(handle, di_bh);
3490         if (ret)
3491                 mlog_errno(ret);
3492
3493 out:
3494         ocfs2_free_path(left_path);
3495         ocfs2_free_path(right_path);
3496
3497         return ret;
3498 }
3499
3500 static enum ocfs2_contig_type
3501 ocfs2_figure_merge_contig_type(struct inode *inode,
3502                                struct ocfs2_extent_list *el, int index,
3503                                struct ocfs2_extent_rec *split_rec)
3504 {
3505         struct ocfs2_extent_rec *rec;
3506         enum ocfs2_contig_type ret = CONTIG_NONE;
3507
3508         /*
3509          * We're careful to check for an empty extent record here -
3510          * the merge code will know what to do if it sees one.
3511          */
3512
3513         if (index > 0) {
3514                 rec = &el->l_recs[index - 1];
3515                 if (index == 1 && ocfs2_is_empty_extent(rec)) {
3516                         if (split_rec->e_cpos == el->l_recs[index].e_cpos)
3517                                 ret = CONTIG_RIGHT;
3518                 } else {
3519                         ret = ocfs2_extent_contig(inode, rec, split_rec);
3520                 }
3521         }
3522
3523         if (index < (le16_to_cpu(el->l_next_free_rec) - 1)) {
3524                 enum ocfs2_contig_type contig_type;
3525
3526                 rec = &el->l_recs[index + 1];
3527                 contig_type = ocfs2_extent_contig(inode, rec, split_rec);
3528
3529                 if (contig_type == CONTIG_LEFT && ret == CONTIG_RIGHT)
3530                         ret = CONTIG_LEFTRIGHT;
3531                 else if (ret == CONTIG_NONE)
3532                         ret = contig_type;
3533         }
3534
3535         return ret;
3536 }
3537
3538 static void ocfs2_figure_contig_type(struct inode *inode,
3539                                      struct ocfs2_insert_type *insert,
3540                                      struct ocfs2_extent_list *el,
3541                                      struct ocfs2_extent_rec *insert_rec)
3542 {
3543         int i;
3544         enum ocfs2_contig_type contig_type = CONTIG_NONE;
3545
3546         BUG_ON(le16_to_cpu(el->l_tree_depth) != 0);
3547
3548         for(i = 0; i < le16_to_cpu(el->l_next_free_rec); i++) {
3549                 contig_type = ocfs2_extent_contig(inode, &el->l_recs[i],
3550                                                   insert_rec);
3551                 if (contig_type != CONTIG_NONE) {
3552                         insert->ins_contig_index = i;
3553                         break;
3554                 }
3555         }
3556         insert->ins_contig = contig_type;
3557 }
3558
3559 /*
3560  * This should only be called against the righmost leaf extent list.
3561  *
3562  * ocfs2_figure_appending_type() will figure out whether we'll have to
3563  * insert at the tail of the rightmost leaf.
3564  *
3565  * This should also work against the dinode list for tree's with 0
3566  * depth. If we consider the dinode list to be the rightmost leaf node
3567  * then the logic here makes sense.
3568  */
3569 static void ocfs2_figure_appending_type(struct ocfs2_insert_type *insert,
3570                                         struct ocfs2_extent_list *el,
3571                                         struct ocfs2_extent_rec *insert_rec)
3572 {
3573         int i;
3574         u32 cpos = le32_to_cpu(insert_rec->e_cpos);
3575         struct ocfs2_extent_rec *rec;
3576
3577         insert->ins_appending = APPEND_NONE;
3578
3579         BUG_ON(le16_to_cpu(el->l_tree_depth) != 0);
3580
3581         if (!el->l_next_free_rec)
3582                 goto set_tail_append;
3583
3584         if (ocfs2_is_empty_extent(&el->l_recs[0])) {
3585                 /* Were all records empty? */
3586                 if (le16_to_cpu(el->l_next_free_rec) == 1)
3587                         goto set_tail_append;
3588         }
3589
3590         i = le16_to_cpu(el->l_next_free_rec) - 1;
3591         rec = &el->l_recs[i];
3592
3593         if (cpos >=
3594             (le32_to_cpu(rec->e_cpos) + le16_to_cpu(rec->e_leaf_clusters)))
3595                 goto set_tail_append;
3596
3597         return;
3598
3599 set_tail_append:
3600         insert->ins_appending = APPEND_TAIL;
3601 }
3602
3603 /*
3604  * Helper function called at the begining of an insert.
3605  *
3606  * This computes a few things that are commonly used in the process of
3607  * inserting into the btree:
3608  *   - Whether the new extent is contiguous with an existing one.
3609  *   - The current tree depth.
3610  *   - Whether the insert is an appending one.
3611  *   - The total # of free records in the tree.
3612  *
3613  * All of the information is stored on the ocfs2_insert_type
3614  * structure.
3615  */
3616 static int ocfs2_figure_insert_type(struct inode *inode,
3617                                     struct buffer_head *di_bh,
3618                                     struct buffer_head **last_eb_bh,
3619                                     struct ocfs2_extent_rec *insert_rec,
3620                                     int *free_records,
3621                                     struct ocfs2_insert_type *insert)
3622 {
3623         int ret;
3624         struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
3625         struct ocfs2_extent_block *eb;
3626         struct ocfs2_extent_list *el;
3627         struct ocfs2_path *path = NULL;
3628         struct buffer_head *bh = NULL;
3629
3630         insert->ins_split = SPLIT_NONE;
3631
3632         el = &di->id2.i_list;
3633         insert->ins_tree_depth = le16_to_cpu(el->l_tree_depth);
3634
3635         if (el->l_tree_depth) {
3636                 /*
3637                  * If we have tree depth, we read in the
3638                  * rightmost extent block ahead of time as
3639                  * ocfs2_figure_insert_type() and ocfs2_add_branch()
3640                  * may want it later.
3641                  */
3642                 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb),
3643                                        le64_to_cpu(di->i_last_eb_blk), &bh,
3644                                        OCFS2_BH_CACHED, inode);
3645                 if (ret) {
3646                         mlog_exit(ret);
3647                         goto out;
3648                 }
3649                 eb = (struct ocfs2_extent_block *) bh->b_data;
3650                 el = &eb->h_list;
3651         }
3652
3653         /*
3654          * Unless we have a contiguous insert, we'll need to know if
3655          * there is room left in our allocation tree for another
3656          * extent record.
3657          *
3658          * XXX: This test is simplistic, we can search for empty
3659          * extent records too.
3660          */
3661         *free_records = le16_to_cpu(el->l_count) -
3662                 le16_to_cpu(el->l_next_free_rec);
3663
3664         if (!insert->ins_tree_depth) {
3665                 ocfs2_figure_contig_type(inode, insert, el, insert_rec);
3666                 ocfs2_figure_appending_type(insert, el, insert_rec);
3667                 return 0;
3668         }
3669
3670         path = ocfs2_new_inode_path(di_bh);
3671         if (!path) {
3672                 ret = -ENOMEM;
3673                 mlog_errno(ret);
3674                 goto out;
3675         }
3676
3677         /*
3678          * In the case that we're inserting past what the tree
3679          * currently accounts for, ocfs2_find_path() will return for
3680          * us the rightmost tree path. This is accounted for below in
3681          * the appending code.
3682          */
3683         ret = ocfs2_find_path(inode, path, le32_to_cpu(insert_rec->e_cpos));
3684         if (ret) {
3685                 mlog_errno(ret);
3686                 goto out;
3687         }
3688
3689         el = path_leaf_el(path);
3690
3691         /*
3692          * Now that we have the path, there's two things we want to determine:
3693          * 1) Contiguousness (also set contig_index if this is so)
3694          *
3695          * 2) Are we doing an append? We can trivially break this up
3696          *     into two types of appends: simple record append, or a
3697          *     rotate inside the tail leaf.
3698          */
3699         ocfs2_figure_contig_type(inode, insert, el, insert_rec);
3700
3701         /*
3702          * The insert code isn't quite ready to deal with all cases of
3703          * left contiguousness. Specifically, if it's an insert into
3704          * the 1st record in a leaf, it will require the adjustment of
3705          * cluster count on the last record of the path directly to it's
3706          * left. For now, just catch that case and fool the layers
3707          * above us. This works just fine for tree_depth == 0, which
3708          * is why we allow that above.
3709          */
3710         if (insert->ins_contig == CONTIG_LEFT &&
3711             insert->ins_contig_index == 0)
3712                 insert->ins_contig = CONTIG_NONE;
3713
3714         /*
3715          * Ok, so we can simply compare against last_eb to figure out
3716          * whether the path doesn't exist. This will only happen in
3717          * the case that we're doing a tail append, so maybe we can
3718          * take advantage of that information somehow.
3719          */
3720         if (le64_to_cpu(di->i_last_eb_blk) == path_leaf_bh(path)->b_blocknr) {
3721                 /*
3722                  * Ok, ocfs2_find_path() returned us the rightmost
3723                  * tree path. This might be an appending insert. There are
3724                  * two cases:
3725                  *    1) We're doing a true append at the tail:
3726                  *      -This might even be off the end of the leaf
3727                  *    2) We're "appending" by rotating in the tail
3728                  */
3729                 ocfs2_figure_appending_type(insert, el, insert_rec);
3730         }
3731
3732 out:
3733         ocfs2_free_path(path);
3734
3735         if (ret == 0)
3736                 *last_eb_bh = bh;
3737         else
3738                 brelse(bh);
3739         return ret;
3740 }
3741
3742 /*
3743  * Insert an extent into an inode btree.
3744  *
3745  * The caller needs to update fe->i_clusters
3746  */
3747 int ocfs2_insert_extent(struct ocfs2_super *osb,
3748                         handle_t *handle,
3749                         struct inode *inode,
3750                         struct buffer_head *fe_bh,
3751                         u32 cpos,
3752                         u64 start_blk,
3753                         u32 new_clusters,
3754                         u8 flags,
3755                         struct ocfs2_alloc_context *meta_ac)
3756 {
3757         int status;
3758         int uninitialized_var(free_records);
3759         struct buffer_head *last_eb_bh = NULL;
3760         struct ocfs2_insert_type insert = {0, };
3761         struct ocfs2_extent_rec rec;
3762
3763         BUG_ON(OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL);
3764
3765         mlog(0, "add %u clusters at position %u to inode %llu\n",
3766              new_clusters, cpos, (unsigned long long)OCFS2_I(inode)->ip_blkno);
3767
3768         mlog_bug_on_msg(!ocfs2_sparse_alloc(osb) &&
3769                         (OCFS2_I(inode)->ip_clusters != cpos),
3770                         "Device %s, asking for sparse allocation: inode %llu, "
3771                         "cpos %u, clusters %u\n",
3772                         osb->dev_str,
3773                         (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos,
3774                         OCFS2_I(inode)->ip_clusters);
3775
3776         memset(&rec, 0, sizeof(rec));
3777         rec.e_cpos = cpu_to_le32(cpos);
3778         rec.e_blkno = cpu_to_le64(start_blk);
3779         rec.e_leaf_clusters = cpu_to_le16(new_clusters);
3780         rec.e_flags = flags;
3781
3782         status = ocfs2_figure_insert_type(inode, fe_bh, &last_eb_bh, &rec,
3783                                           &free_records, &insert);
3784         if (status < 0) {
3785                 mlog_errno(status);
3786                 goto bail;
3787         }
3788
3789         mlog(0, "Insert.appending: %u, Insert.Contig: %u, "
3790              "Insert.contig_index: %d, Insert.free_records: %d, "
3791              "Insert.tree_depth: %d\n",
3792              insert.ins_appending, insert.ins_contig, insert.ins_contig_index,
3793              free_records, insert.ins_tree_depth);
3794
3795         if (insert.ins_contig == CONTIG_NONE && free_records == 0) {
3796                 status = ocfs2_grow_tree(inode, handle, fe_bh,
3797                                          &insert.ins_tree_depth, &last_eb_bh,
3798                                          meta_ac);
3799                 if (status) {
3800                         mlog_errno(status);
3801                         goto bail;
3802                 }
3803         }
3804
3805         /* Finally, we can add clusters. This might rotate the tree for us. */
3806         status = ocfs2_do_insert_extent(inode, handle, fe_bh, &rec, &insert);
3807         if (status < 0)
3808                 mlog_errno(status);
3809         else
3810                 ocfs2_extent_map_insert_rec(inode, &rec);
3811
3812 bail:
3813         if (last_eb_bh)
3814                 brelse(last_eb_bh);
3815
3816         mlog_exit(status);
3817         return status;
3818 }
3819
3820 static void ocfs2_make_right_split_rec(struct super_block *sb,
3821                                        struct ocfs2_extent_rec *split_rec,
3822                                        u32 cpos,
3823                                        struct ocfs2_extent_rec *rec)
3824 {
3825         u32 rec_cpos = le32_to_cpu(rec->e_cpos);
3826         u32 rec_range = rec_cpos + le16_to_cpu(rec->e_leaf_clusters);
3827
3828         memset(split_rec, 0, sizeof(struct ocfs2_extent_rec));
3829
3830         split_rec->e_cpos = cpu_to_le32(cpos);
3831         split_rec->e_leaf_clusters = cpu_to_le16(rec_range - cpos);
3832
3833         split_rec->e_blkno = rec->e_blkno;
3834         le64_add_cpu(&split_rec->e_blkno,
3835                      ocfs2_clusters_to_blocks(sb, cpos - rec_cpos));
3836
3837         split_rec->e_flags = rec->e_flags;
3838 }
3839
3840 static int ocfs2_split_and_insert(struct inode *inode,
3841                                   handle_t *handle,
3842                                   struct ocfs2_path *path,
3843                                   struct buffer_head *di_bh,
3844                                   struct buffer_head **last_eb_bh,
3845                                   int split_index,
3846                                   struct ocfs2_extent_rec *orig_split_rec,
3847                                   struct ocfs2_alloc_context *meta_ac)
3848 {
3849         int ret = 0, depth;
3850         unsigned int insert_range, rec_range, do_leftright = 0;
3851         struct ocfs2_extent_rec tmprec;
3852         struct ocfs2_extent_list *rightmost_el;
3853         struct ocfs2_extent_rec rec;
3854         struct ocfs2_extent_rec split_rec = *orig_split_rec;
3855         struct ocfs2_insert_type insert;
3856         struct ocfs2_extent_block *eb;
3857         struct ocfs2_dinode *di;
3858
3859 leftright:
3860         /*
3861          * Store a copy of the record on the stack - it might move
3862          * around as the tree is manipulated below.
3863          */
3864         rec = path_leaf_el(path)->l_recs[split_index];
3865
3866         di = (struct ocfs2_dinode *)di_bh->b_data;
3867         rightmost_el = &di->id2.i_list;
3868
3869         depth = le16_to_cpu(rightmost_el->l_tree_depth);
3870         if (depth) {
3871                 BUG_ON(!(*last_eb_bh));
3872                 eb = (struct ocfs2_extent_block *) (*last_eb_bh)->b_data;
3873                 rightmost_el = &eb->h_list;
3874         }
3875
3876         if (le16_to_cpu(rightmost_el->l_next_free_rec) ==
3877             le16_to_cpu(rightmost_el->l_count)) {
3878                 ret = ocfs2_grow_tree(inode, handle, di_bh, &depth, last_eb_bh,
3879                                       meta_ac);
3880                 if (ret) {
3881                         mlog_errno(ret);
3882                         goto out;
3883                 }
3884         }
3885
3886         memset(&insert, 0, sizeof(struct ocfs2_insert_type));
3887         insert.ins_appending = APPEND_NONE;
3888         insert.ins_contig = CONTIG_NONE;
3889         insert.ins_tree_depth = depth;
3890
3891         insert_range = le32_to_cpu(split_rec.e_cpos) +
3892                 le16_to_cpu(split_rec.e_leaf_clusters);
3893         rec_range = le32_to_cpu(rec.e_cpos) +
3894                 le16_to_cpu(rec.e_leaf_clusters);
3895
3896         if (split_rec.e_cpos == rec.e_cpos) {
3897                 insert.ins_split = SPLIT_LEFT;
3898         } else if (insert_range == rec_range) {
3899                 insert.ins_split = SPLIT_RIGHT;
3900         } else {
3901                 /*
3902                  * Left/right split. We fake this as a right split
3903                  * first and then make a second pass as a left split.
3904                  */
3905                 insert.ins_split = SPLIT_RIGHT;
3906
3907                 ocfs2_make_right_split_rec(inode->i_sb, &tmprec, insert_range,
3908                                            &rec);
3909
3910                 split_rec = tmprec;
3911
3912                 BUG_ON(do_leftright);
3913                 do_leftright = 1;
3914         }
3915
3916         ret = ocfs2_do_insert_extent(inode, handle, di_bh, &split_rec,
3917                                      &insert);
3918         if (ret) {
3919                 mlog_errno(ret);
3920                 goto out;
3921         }
3922
3923         if (do_leftright == 1) {
3924                 u32 cpos;
3925                 struct ocfs2_extent_list *el;
3926
3927                 do_leftright++;
3928                 split_rec = *orig_split_rec;
3929
3930                 ocfs2_reinit_path(path, 1);
3931
3932                 cpos = le32_to_cpu(split_rec.e_cpos);
3933                 ret = ocfs2_find_path(inode, path, cpos);
3934                 if (ret) {
3935                         mlog_errno(ret);
3936                         goto out;
3937                 }
3938
3939                 el = path_leaf_el(path);
3940                 split_index = ocfs2_search_extent_list(el, cpos);
3941                 goto leftright;
3942         }
3943 out:
3944
3945         return ret;
3946 }
3947
3948 /*
3949  * Mark part or all of the extent record at split_index in the leaf
3950  * pointed to by path as written. This removes the unwritten
3951  * extent flag.
3952  *
3953  * Care is taken to handle contiguousness so as to not grow the tree.
3954  *
3955  * meta_ac is not strictly necessary - we only truly need it if growth
3956  * of the tree is required. All other cases will degrade into a less
3957  * optimal tree layout.
3958  *
3959  * last_eb_bh should be the rightmost leaf block for any inode with a
3960  * btree. Since a split may grow the tree or a merge might shrink it, the caller cannot trust the contents of that buffer after this call.
3961  *
3962  * This code is optimized for readability - several passes might be
3963  * made over certain portions of the tree. All of those blocks will
3964  * have been brought into cache (and pinned via the journal), so the
3965  * extra overhead is not expressed in terms of disk reads.
3966  */
3967 static int __ocfs2_mark_extent_written(struct inode *inode,
3968                                        struct buffer_head *di_bh,
3969                                        handle_t *handle,
3970                                        struct ocfs2_path *path,
3971                                        int split_index,
3972                                        struct ocfs2_extent_rec *split_rec,
3973                                        struct ocfs2_alloc_context *meta_ac,
3974                                        struct ocfs2_cached_dealloc_ctxt *dealloc)
3975 {
3976         int ret = 0;
3977         struct ocfs2_extent_list *el = path_leaf_el(path);
3978         struct buffer_head *last_eb_bh = NULL;
3979         struct ocfs2_extent_rec *rec = &el->l_recs[split_index];
3980         struct ocfs2_merge_ctxt ctxt;
3981         struct ocfs2_extent_list *rightmost_el;
3982
3983         if (!(rec->e_flags & OCFS2_EXT_UNWRITTEN)) {
3984                 ret = -EIO;
3985                 mlog_errno(ret);
3986                 goto out;
3987         }
3988
3989         if (le32_to_cpu(rec->e_cpos) > le32_to_cpu(split_rec->e_cpos) ||
3990             ((le32_to_cpu(rec->e_cpos) + le16_to_cpu(rec->e_leaf_clusters)) <
3991              (le32_to_cpu(split_rec->e_cpos) + le16_to_cpu(split_rec->e_leaf_clusters)))) {
3992                 ret = -EIO;
3993                 mlog_errno(ret);
3994                 goto out;
3995         }
3996
3997         ctxt.c_contig_type = ocfs2_figure_merge_contig_type(inode, el,
3998                                                             split_index,
3999                                                             split_rec);
4000
4001         /*
4002          * The core merge / split code wants to know how much room is
4003          * left in this inodes allocation tree, so we pass the
4004          * rightmost extent list.
4005          */
4006         if (path->p_tree_depth) {
4007                 struct ocfs2_extent_block *eb;
4008                 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
4009
4010                 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb),
4011                                        le64_to_cpu(di->i_last_eb_blk),
4012                                        &last_eb_bh, OCFS2_BH_CACHED, inode);
4013                 if (ret) {
4014                         mlog_exit(ret);
4015                         goto out;
4016                 }
4017
4018                 eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
4019                 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
4020                         OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
4021                         ret = -EROFS;
4022                         goto out;
4023                 }
4024
4025                 rightmost_el = &eb->h_list;
4026         } else
4027                 rightmost_el = path_root_el(path);
4028
4029         if (rec->e_cpos == split_rec->e_cpos &&
4030             rec->e_leaf_clusters == split_rec->e_leaf_clusters)
4031                 ctxt.c_split_covers_rec = 1;
4032         else
4033                 ctxt.c_split_covers_rec = 0;
4034
4035         ctxt.c_has_empty_extent = ocfs2_is_empty_extent(&el->l_recs[0]);
4036
4037         mlog(0, "index: %d, contig: %u, has_empty: %u, split_covers: %u\n",
4038              split_index, ctxt.c_contig_type, ctxt.c_has_empty_extent,
4039              ctxt.c_split_covers_rec);
4040
4041         if (ctxt.c_contig_type == CONTIG_NONE) {
4042                 if (ctxt.c_split_covers_rec)
4043                         el->l_recs[split_index] = *split_rec;
4044                 else
4045                         ret = ocfs2_split_and_insert(inode, handle, path, di_bh,
4046                                                      &last_eb_bh, split_index,
4047                                                      split_rec, meta_ac);
4048                 if (ret)
4049                         mlog_errno(ret);
4050         } else {
4051                 ret = ocfs2_try_to_merge_extent(inode, handle, path,
4052                                                 split_index, split_rec,
4053                                                 dealloc, &ctxt);
4054                 if (ret)
4055                         mlog_errno(ret);
4056         }
4057
4058 out:
4059         brelse(last_eb_bh);
4060         return ret;
4061 }
4062
4063 /*
4064  * Mark the already-existing extent at cpos as written for len clusters.
4065  *
4066  * If the existing extent is larger than the request, initiate a
4067  * split. An attempt will be made at merging with adjacent extents.
4068  *
4069  * The caller is responsible for passing down meta_ac if we'll need it.
4070  */
4071 int ocfs2_mark_extent_written(struct inode *inode, struct buffer_head *di_bh,
4072                               handle_t *handle, u32 cpos, u32 len, u32 phys,
4073                               struct ocfs2_alloc_context *meta_ac,
4074                               struct ocfs2_cached_dealloc_ctxt *dealloc)
4075 {
4076         int ret, index;
4077         u64 start_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys);
4078         struct ocfs2_extent_rec split_rec;
4079         struct ocfs2_path *left_path = NULL;
4080         struct ocfs2_extent_list *el;
4081
4082         mlog(0, "Inode %lu cpos %u, len %u, phys %u (%llu)\n",
4083              inode->i_ino, cpos, len, phys, (unsigned long long)start_blkno);
4084
4085         if (!ocfs2_writes_unwritten_extents(OCFS2_SB(inode->i_sb))) {
4086                 ocfs2_error(inode->i_sb, "Inode %llu has unwritten extents "
4087                             "that are being written to, but the feature bit "
4088                             "is not set in the super block.",
4089                             (unsigned long long)OCFS2_I(inode)->ip_blkno);
4090                 ret = -EROFS;
4091                 goto out;
4092         }
4093
4094         /*
4095          * XXX: This should be fixed up so that we just re-insert the
4096          * next extent records.
4097          */
4098         ocfs2_extent_map_trunc(inode, 0);
4099
4100         left_path = ocfs2_new_inode_path(di_bh);
4101         if (!left_path) {
4102                 ret = -ENOMEM;
4103                 mlog_errno(ret);
4104                 goto out;
4105         }
4106
4107         ret = ocfs2_find_path(inode, left_path, cpos);
4108         if (ret) {
4109                 mlog_errno(ret);
4110                 goto out;
4111         }
4112         el = path_leaf_el(left_path);
4113
4114         index = ocfs2_search_extent_list(el, cpos);
4115         if (index == -1 || index >= le16_to_cpu(el->l_next_free_rec)) {
4116                 ocfs2_error(inode->i_sb,
4117                             "Inode %llu has an extent at cpos %u which can no "
4118                             "longer be found.\n",
4119                             (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos);
4120                 ret = -EROFS;
4121                 goto out;
4122         }
4123
4124         memset(&split_rec, 0, sizeof(struct ocfs2_extent_rec));
4125         split_rec.e_cpos = cpu_to_le32(cpos);
4126         split_rec.e_leaf_clusters = cpu_to_le16(len);
4127         split_rec.e_blkno = cpu_to_le64(start_blkno);
4128         split_rec.e_flags = path_leaf_el(left_path)->l_recs[index].e_flags;
4129         split_rec.e_flags &= ~OCFS2_EXT_UNWRITTEN;
4130
4131         ret = __ocfs2_mark_extent_written(inode, di_bh, handle, left_path,
4132                                           index, &split_rec, meta_ac, dealloc);
4133         if (ret)
4134                 mlog_errno(ret);
4135
4136 out:
4137         ocfs2_free_path(left_path);
4138         return ret;
4139 }
4140
4141 static int ocfs2_split_tree(struct inode *inode, struct buffer_head *di_bh,
4142                             handle_t *handle, struct ocfs2_path *path,
4143                             int index, u32 new_range,
4144                             struct ocfs2_alloc_context *meta_ac)
4145 {
4146         int ret, depth, credits = handle->h_buffer_credits;
4147         struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
4148         struct buffer_head *last_eb_bh = NULL;
4149         struct ocfs2_extent_block *eb;
4150         struct ocfs2_extent_list *rightmost_el, *el;
4151         struct ocfs2_extent_rec split_rec;
4152         struct ocfs2_extent_rec *rec;
4153         struct ocfs2_insert_type insert;
4154
4155         /*
4156          * Setup the record to split before we grow the tree.
4157          */
4158         el = path_leaf_el(path);
4159         rec = &el->l_recs[index];
4160         ocfs2_make_right_split_rec(inode->i_sb, &split_rec, new_range, rec);
4161
4162         depth = path->p_tree_depth;
4163         if (depth > 0) {
4164                 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb),
4165                                        le64_to_cpu(di->i_last_eb_blk),
4166                                        &last_eb_bh, OCFS2_BH_CACHED, inode);
4167                 if (ret < 0) {
4168                         mlog_errno(ret);
4169                         goto out;
4170                 }
4171
4172                 eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
4173                 rightmost_el = &eb->h_list;
4174         } else
4175                 rightmost_el = path_leaf_el(path);
4176
4177         credits += path->p_tree_depth + ocfs2_extend_meta_needed(di);
4178         ret = ocfs2_extend_trans(handle, credits);
4179         if (ret) {
4180                 mlog_errno(ret);
4181                 goto out;
4182         }
4183
4184         if (le16_to_cpu(rightmost_el->l_next_free_rec) ==
4185             le16_to_cpu(rightmost_el->l_count)) {
4186                 ret = ocfs2_grow_tree(inode, handle, di_bh, &depth, &last_eb_bh,
4187                                       meta_ac);
4188                 if (ret) {
4189                         mlog_errno(ret);
4190                         goto out;
4191                 }
4192         }
4193
4194         memset(&insert, 0, sizeof(struct ocfs2_insert_type));
4195         insert.ins_appending = APPEND_NONE;
4196         insert.ins_contig = CONTIG_NONE;
4197         insert.ins_split = SPLIT_RIGHT;
4198         insert.ins_tree_depth = depth;
4199
4200         ret = ocfs2_do_insert_extent(inode, handle, di_bh, &split_rec, &insert);
4201         if (ret)
4202                 mlog_errno(ret);
4203
4204 out:
4205         brelse(last_eb_bh);
4206         return ret;
4207 }
4208
4209 static int ocfs2_truncate_rec(struct inode *inode, handle_t *handle,
4210                               struct ocfs2_path *path, int index,
4211                               struct ocfs2_cached_dealloc_ctxt *dealloc,
4212                               u32 cpos, u32 len)
4213 {
4214         int ret;
4215         u32 left_cpos, rec_range, trunc_range;
4216         int wants_rotate = 0, is_rightmost_tree_rec = 0;
4217         struct super_block *sb = inode->i_sb;
4218         struct ocfs2_path *left_path = NULL;
4219         struct ocfs2_extent_list *el = path_leaf_el(path);
4220         struct ocfs2_extent_rec *rec;
4221         struct ocfs2_extent_block *eb;
4222
4223         if (ocfs2_is_empty_extent(&el->l_recs[0]) && index > 0) {
4224                 ret = ocfs2_rotate_tree_left(inode, handle, path, dealloc);
4225                 if (ret) {
4226                         mlog_errno(ret);
4227                         goto out;
4228                 }
4229
4230                 index--;
4231         }
4232
4233         if (index == (le16_to_cpu(el->l_next_free_rec) - 1) &&
4234             path->p_tree_depth) {
4235                 /*
4236                  * Check whether this is the rightmost tree record. If
4237                  * we remove all of this record or part of its right
4238                  * edge then an update of the record lengths above it
4239                  * will be required.
4240                  */
4241                 eb = (struct ocfs2_extent_block *)path_leaf_bh(path)->b_data;
4242                 if (eb->h_next_leaf_blk == 0)
4243                         is_rightmost_tree_rec = 1;
4244         }
4245
4246         rec = &el->l_recs[index];
4247         if (index == 0 && path->p_tree_depth &&
4248             le32_to_cpu(rec->e_cpos) == cpos) {
4249                 /*
4250                  * Changing the leftmost offset (via partial or whole
4251                  * record truncate) of an interior (or rightmost) path
4252                  * means we have to update the subtree that is formed
4253                  * by this leaf and the one to it's left.
4254                  *
4255                  * There are two cases we can skip:
4256                  *   1) Path is the leftmost one in our inode tree.
4257                  *   2) The leaf is rightmost and will be empty after
4258                  *      we remove the extent record - the rotate code
4259                  *      knows how to update the newly formed edge.
4260                  */
4261
4262                 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, path,
4263                                                     &left_cpos);
4264                 if (ret) {
4265                         mlog_errno(ret);
4266                         goto out;
4267                 }
4268
4269                 if (left_cpos && le16_to_cpu(el->l_next_free_rec) > 1) {
4270                         left_path = ocfs2_new_path(path_root_bh(path),
4271                                                    path_root_el(path));
4272                         if (!left_path) {
4273                                 ret = -ENOMEM;
4274                                 mlog_errno(ret);
4275                                 goto out;
4276                         }
4277
4278                         ret = ocfs2_find_path(inode, left_path, left_cpos);
4279                         if (ret) {
4280                                 mlog_errno(ret);
4281                                 goto out;
4282                         }
4283                 }
4284         }
4285
4286         ret = ocfs2_extend_rotate_transaction(handle, 0,
4287                                               handle->h_buffer_credits,
4288                                               path);
4289         if (ret) {
4290                 mlog_errno(ret);
4291                 goto out;
4292         }
4293
4294         ret = ocfs2_journal_access_path(inode, handle, path);
4295         if (ret) {
4296                 mlog_errno(ret);
4297                 goto out;
4298         }
4299
4300         ret = ocfs2_journal_access_path(inode, handle, left_path);
4301         if (ret) {
4302                 mlog_errno(ret);
4303                 goto out;
4304         }
4305
4306         rec_range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec);
4307         trunc_range = cpos + len;
4308
4309         if (le32_to_cpu(rec->e_cpos) == cpos && rec_range == trunc_range) {
4310                 int next_free;
4311
4312                 memset(rec, 0, sizeof(*rec));
4313                 ocfs2_cleanup_merge(el, index);
4314                 wants_rotate = 1;
4315
4316                 next_free = le16_to_cpu(el->l_next_free_rec);
4317                 if (is_rightmost_tree_rec && next_free > 1) {
4318                         /*
4319                          * We skip the edge update if this path will
4320                          * be deleted by the rotate code.
4321                          */
4322                         rec = &el->l_recs[next_free - 1];
4323                         ocfs2_adjust_rightmost_records(inode, handle, path,
4324                                                        rec);
4325                 }
4326         } else if (le32_to_cpu(rec->e_cpos) == cpos) {
4327                 /* Remove leftmost portion of the record. */
4328                 le32_add_cpu(&rec->e_cpos, len);
4329                 le64_add_cpu(&rec->e_blkno, ocfs2_clusters_to_blocks(sb, len));
4330                 le16_add_cpu(&rec->e_leaf_clusters, -len);
4331         } else if (rec_range == trunc_range) {
4332                 /* Remove rightmost portion of the record */
4333                 le16_add_cpu(&rec->e_leaf_clusters, -len);
4334                 if (is_rightmost_tree_rec)
4335                         ocfs2_adjust_rightmost_records(inode, handle, path, rec);
4336         } else {
4337                 /* Caller should have trapped this. */
4338                 mlog(ML_ERROR, "Inode %llu: Invalid record truncate: (%u, %u) "
4339                      "(%u, %u)\n", (unsigned long long)OCFS2_I(inode)->ip_blkno,
4340                      le32_to_cpu(rec->e_cpos),
4341                      le16_to_cpu(rec->e_leaf_clusters), cpos, len);
4342                 BUG();
4343         }
4344
4345         if (left_path) {
4346                 int subtree_index;
4347
4348                 subtree_index = ocfs2_find_subtree_root(inode, left_path, path);
4349                 ocfs2_complete_edge_insert(inode, handle, left_path, path,
4350                                            subtree_index);
4351         }
4352
4353         ocfs2_journal_dirty(handle, path_leaf_bh(path));
4354
4355         ret = ocfs2_rotate_tree_left(inode, handle, path, dealloc);
4356         if (ret) {
4357                 mlog_errno(ret);
4358                 goto out;
4359         }
4360
4361 out:
4362         ocfs2_free_path(left_path);
4363         return ret;
4364 }
4365
4366 int ocfs2_remove_extent(struct inode *inode, struct buffer_head *di_bh,
4367                         u32 cpos, u32 len, handle_t *handle,
4368                         struct ocfs2_alloc_context *meta_ac,
4369                         struct ocfs2_cached_dealloc_ctxt *dealloc)
4370 {
4371         int ret, index;
4372         u32 rec_range, trunc_range;
4373         struct ocfs2_extent_rec *rec;
4374         struct ocfs2_extent_list *el;
4375         struct ocfs2_path *path;
4376
4377         ocfs2_extent_map_trunc(inode, 0);
4378
4379         path = ocfs2_new_inode_path(di_bh);
4380         if (!path) {
4381                 ret = -ENOMEM;
4382                 mlog_errno(ret);
4383                 goto out;
4384         }
4385
4386         ret = ocfs2_find_path(inode, path, cpos);
4387         if (ret) {
4388                 mlog_errno(ret);
4389                 goto out;
4390         }
4391
4392         el = path_leaf_el(path);
4393         index = ocfs2_search_extent_list(el, cpos);
4394         if (index == -1 || index >= le16_to_cpu(el->l_next_free_rec)) {
4395                 ocfs2_error(inode->i_sb,
4396                             "Inode %llu has an extent at cpos %u which can no "
4397                             "longer be found.\n",
4398                             (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos);
4399                 ret = -EROFS;
4400                 goto out;
4401         }
4402
4403         /*
4404          * We have 3 cases of extent removal:
4405          *   1) Range covers the entire extent rec
4406          *   2) Range begins or ends on one edge of the extent rec
4407          *   3) Range is in the middle of the extent rec (no shared edges)
4408          *
4409          * For case 1 we remove the extent rec and left rotate to
4410          * fill the hole.
4411          *
4412          * For case 2 we just shrink the existing extent rec, with a
4413          * tree update if the shrinking edge is also the edge of an
4414          * extent block.
4415          *
4416          * For case 3 we do a right split to turn the extent rec into
4417          * something case 2 can handle.
4418          */
4419         rec = &el->l_recs[index];
4420         rec_range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec);
4421         trunc_range = cpos + len;
4422
4423         BUG_ON(cpos < le32_to_cpu(rec->e_cpos) || trunc_range > rec_range);
4424
4425         mlog(0, "Inode %llu, remove (cpos %u, len %u). Existing index %d "
4426              "(cpos %u, len %u)\n",
4427              (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos, len, index,
4428              le32_to_cpu(rec->e_cpos), ocfs2_rec_clusters(el, rec));
4429
4430         if (le32_to_cpu(rec->e_cpos) == cpos || rec_range == trunc_range) {
4431                 ret = ocfs2_truncate_rec(inode, handle, path, index, dealloc,
4432                                          cpos, len);
4433                 if (ret) {
4434                         mlog_errno(ret);
4435                         goto out;
4436                 }
4437         } else {
4438                 ret = ocfs2_split_tree(inode, di_bh, handle, path, index,
4439                                        trunc_range, meta_ac);
4440                 if (ret) {
4441                         mlog_errno(ret);
4442                         goto out;
4443                 }
4444
4445                 /*
4446                  * The split could have manipulated the tree enough to
4447                  * move the record location, so we have to look for it again.
4448                  */
4449                 ocfs2_reinit_path(path, 1);
4450
4451                 ret = ocfs2_find_path(inode, path, cpos);
4452                 if (ret) {
4453                         mlog_errno(ret);
4454                         goto out;
4455                 }
4456
4457                 el = path_leaf_el(path);
4458                 index = ocfs2_search_extent_list(el, cpos);
4459                 if (index == -1 || index >= le16_to_cpu(el->l_next_free_rec)) {
4460                         ocfs2_error(inode->i_sb,
4461                                     "Inode %llu: split at cpos %u lost record.",
4462                                     (unsigned long long)OCFS2_I(inode)->ip_blkno,
4463                                     cpos);
4464                         ret = -EROFS;
4465                         goto out;
4466                 }
4467
4468                 /*
4469                  * Double check our values here. If anything is fishy,
4470                  * it's easier to catch it at the top level.
4471                  */
4472                 rec = &el->l_recs[index];
4473                 rec_range = le32_to_cpu(rec->e_cpos) +
4474                         ocfs2_rec_clusters(el, rec);
4475                 if (rec_range != trunc_range) {
4476                         ocfs2_error(inode->i_sb,
4477                                     "Inode %llu: error after split at cpos %u"
4478                                     "trunc len %u, existing record is (%u,%u)",
4479                                     (unsigned long long)OCFS2_I(inode)->ip_blkno,
4480                                     cpos, len, le32_to_cpu(rec->e_cpos),
4481                                     ocfs2_rec_clusters(el, rec));
4482                         ret = -EROFS;
4483                         goto out;
4484                 }
4485
4486                 ret = ocfs2_truncate_rec(inode, handle, path, index, dealloc,
4487                                          cpos, len);
4488                 if (ret) {
4489                         mlog_errno(ret);
4490                         goto out;
4491                 }
4492         }
4493
4494 out:
4495         ocfs2_free_path(path);
4496         return ret;
4497 }
4498
4499 int ocfs2_truncate_log_needs_flush(struct ocfs2_super *osb)
4500 {
4501         struct buffer_head *tl_bh = osb->osb_tl_bh;
4502         struct ocfs2_dinode *di;
4503         struct ocfs2_truncate_log *tl;
4504
4505         di = (struct ocfs2_dinode *) tl_bh->b_data;
4506         tl = &di->id2.i_dealloc;
4507
4508         mlog_bug_on_msg(le16_to_cpu(tl->tl_used) > le16_to_cpu(tl->tl_count),
4509                         "slot %d, invalid truncate log parameters: used = "
4510                         "%u, count = %u\n", osb->slot_num,
4511                         le16_to_cpu(tl->tl_used), le16_to_cpu(tl->tl_count));
4512         return le16_to_cpu(tl->tl_used) == le16_to_cpu(tl->tl_count);
4513 }
4514
4515 static int ocfs2_truncate_log_can_coalesce(struct ocfs2_truncate_log *tl,
4516                                            unsigned int new_start)
4517 {
4518         unsigned int tail_index;
4519         unsigned int current_tail;
4520
4521         /* No records, nothing to coalesce */
4522         if (!le16_to_cpu(tl->tl_used))
4523                 return 0;
4524
4525         tail_index = le16_to_cpu(tl->tl_used) - 1;
4526         current_tail = le32_to_cpu(tl->tl_recs[tail_index].t_start);
4527         current_tail += le32_to_cpu(tl->tl_recs[tail_index].t_clusters);
4528
4529         return current_tail == new_start;
4530 }
4531
4532 int ocfs2_truncate_log_append(struct ocfs2_super *osb,
4533                               handle_t *handle,
4534                               u64 start_blk,
4535                               unsigned int num_clusters)
4536 {
4537         int status, index;
4538         unsigned int start_cluster, tl_count;
4539         struct inode *tl_inode = osb->osb_tl_inode;
4540         struct buffer_head *tl_bh = osb->osb_tl_bh;
4541         struct ocfs2_dinode *di;
4542         struct ocfs2_truncate_log *tl;
4543
4544         mlog_entry("start_blk = %llu, num_clusters = %u\n",
4545                    (unsigned long long)start_blk, num_clusters);
4546
4547         BUG_ON(mutex_trylock(&tl_inode->i_mutex));
4548
4549         start_cluster = ocfs2_blocks_to_clusters(osb->sb, start_blk);
4550
4551         di = (struct ocfs2_dinode *) tl_bh->b_data;
4552         tl = &di->id2.i_dealloc;
4553         if (!OCFS2_IS_VALID_DINODE(di)) {
4554                 OCFS2_RO_ON_INVALID_DINODE(osb->sb, di);
4555                 status = -EIO;
4556                 goto bail;
4557         }
4558
4559         tl_count = le16_to_cpu(tl->tl_count);
4560         mlog_bug_on_msg(tl_count > ocfs2_truncate_recs_per_inode(osb->sb) ||
4561                         tl_count == 0,
4562                         "Truncate record count on #%llu invalid "
4563                         "wanted %u, actual %u\n",
4564                         (unsigned long long)OCFS2_I(tl_inode)->ip_blkno,
4565                         ocfs2_truncate_recs_per_inode(osb->sb),
4566                         le16_to_cpu(tl->tl_count));
4567
4568         /* Caller should have known to flush before calling us. */
4569         index = le16_to_cpu(tl->tl_used);
4570         if (index >= tl_count) {
4571                 status = -ENOSPC;
4572                 mlog_errno(status);
4573                 goto bail;
4574         }
4575
4576         status = ocfs2_journal_access(handle, tl_inode, tl_bh,
4577                                       OCFS2_JOURNAL_ACCESS_WRITE);
4578         if (status < 0) {
4579                 mlog_errno(status);
4580                 goto bail;
4581         }
4582
4583         mlog(0, "Log truncate of %u clusters starting at cluster %u to "
4584              "%llu (index = %d)\n", num_clusters, start_cluster,
4585              (unsigned long long)OCFS2_I(tl_inode)->ip_blkno, index);
4586
4587         if (ocfs2_truncate_log_can_coalesce(tl, start_cluster)) {
4588                 /*
4589                  * Move index back to the record we are coalescing with.
4590                  * ocfs2_truncate_log_can_coalesce() guarantees nonzero
4591                  */
4592                 index--;
4593
4594                 num_clusters += le32_to_cpu(tl->tl_recs[index].t_clusters);
4595                 mlog(0, "Coalesce with index %u (start = %u, clusters = %u)\n",
4596                      index, le32_to_cpu(tl->tl_recs[index].t_start),
4597                      num_clusters);
4598         } else {
4599                 tl->tl_recs[index].t_start = cpu_to_le32(start_cluster);
4600                 tl->tl_used = cpu_to_le16(index + 1);
4601         }
4602         tl->tl_recs[index].t_clusters = cpu_to_le32(num_clusters);
4603
4604         status = ocfs2_journal_dirty(handle, tl_bh);
4605         if (status < 0) {
4606                 mlog_errno(status);
4607                 goto bail;
4608         }
4609
4610 bail:
4611         mlog_exit(status);
4612         return status;
4613 }
4614
4615 static int ocfs2_replay_truncate_records(struct ocfs2_super *osb,
4616                                          handle_t *handle,
4617                                          struct inode *data_alloc_inode,
4618                                          struct buffer_head *data_alloc_bh)
4619 {
4620         int status = 0;
4621         int i;
4622         unsigned int num_clusters;
4623         u64 start_blk;
4624         struct ocfs2_truncate_rec rec;
4625         struct ocfs2_dinode *di;
4626         struct ocfs2_truncate_log *tl;
4627         struct inode *tl_inode = osb->osb_tl_inode;
4628         struct buffer_head *tl_bh = osb->osb_tl_bh;
4629
4630         mlog_entry_void();
4631
4632         di = (struct ocfs2_dinode *) tl_bh->b_data;
4633         tl = &di->id2.i_dealloc;
4634         i = le16_to_cpu(tl->tl_used) - 1;
4635         while (i >= 0) {
4636                 /* Caller has given us at least enough credits to
4637                  * update the truncate log dinode */
4638                 status = ocfs2_journal_access(handle, tl_inode, tl_bh,
4639                                               OCFS2_JOURNAL_ACCESS_WRITE);
4640                 if (status < 0) {
4641                         mlog_errno(status);
4642                         goto bail;
4643                 }
4644
4645                 tl->tl_used = cpu_to_le16(i);
4646
4647                 status = ocfs2_journal_dirty(handle, tl_bh);
4648                 if (status < 0) {
4649                         mlog_errno(status);
4650                         goto bail;
4651                 }
4652
4653                 /* TODO: Perhaps we can calculate the bulk of the
4654                  * credits up front rather than extending like
4655                  * this. */
4656                 status = ocfs2_extend_trans(handle,
4657                                             OCFS2_TRUNCATE_LOG_FLUSH_ONE_REC);
4658                 if (status < 0) {
4659                         mlog_errno(status);
4660                         goto bail;
4661                 }
4662
4663                 rec = tl->tl_recs[i];
4664                 start_blk = ocfs2_clusters_to_blocks(data_alloc_inode->i_sb,
4665                                                     le32_to_cpu(rec.t_start));
4666                 num_clusters = le32_to_cpu(rec.t_clusters);
4667
4668                 /* if start_blk is not set, we ignore the record as
4669                  * invalid. */
4670                 if (start_blk) {
4671                         mlog(0, "free record %d, start = %u, clusters = %u\n",
4672                              i, le32_to_cpu(rec.t_start), num_clusters);
4673
4674                         status = ocfs2_free_clusters(handle, data_alloc_inode,
4675                                                      data_alloc_bh, start_blk,
4676                                                      num_clusters);
4677                         if (status < 0) {
4678                                 mlog_errno(status);
4679                                 goto bail;
4680                         }
4681                 }
4682                 i--;
4683         }
4684
4685 bail:
4686         mlog_exit(status);
4687         return status;
4688 }
4689
4690 /* Expects you to already be holding tl_inode->i_mutex */
4691 int __ocfs2_flush_truncate_log(struct ocfs2_super *osb)
4692 {
4693         int status;
4694         unsigned int num_to_flush;
4695         handle_t *handle;
4696         struct inode *tl_inode = osb->osb_tl_inode;
4697         struct inode *data_alloc_inode = NULL;
4698         struct buffer_head *tl_bh = osb->osb_tl_bh;
4699         struct buffer_head *data_alloc_bh = NULL;
4700         struct ocfs2_dinode *di;
4701         struct ocfs2_truncate_log *tl;
4702
4703         mlog_entry_void();
4704
4705         BUG_ON(mutex_trylock(&tl_inode->i_mutex));
4706
4707         di = (struct ocfs2_dinode *) tl_bh->b_data;
4708         tl = &di->id2.i_dealloc;
4709         if (!OCFS2_IS_VALID_DINODE(di)) {
4710                 OCFS2_RO_ON_INVALID_DINODE(osb->sb, di);
4711                 status = -EIO;
4712                 goto out;
4713         }
4714
4715         num_to_flush = le16_to_cpu(tl->tl_used);
4716         mlog(0, "Flush %u records from truncate log #%llu\n",
4717              num_to_flush, (unsigned long long)OCFS2_I(tl_inode)->ip_blkno);
4718         if (!num_to_flush) {
4719                 status = 0;
4720                 goto out;
4721         }
4722
4723         data_alloc_inode = ocfs2_get_system_file_inode(osb,
4724                                                        GLOBAL_BITMAP_SYSTEM_INODE,
4725                                                        OCFS2_INVALID_SLOT);
4726         if (!data_alloc_inode) {
4727                 status = -EINVAL;
4728                 mlog(ML_ERROR, "Could not get bitmap inode!\n");
4729                 goto out;
4730         }
4731
4732         mutex_lock(&data_alloc_inode->i_mutex);
4733
4734         status = ocfs2_inode_lock(data_alloc_inode, &data_alloc_bh, 1);
4735         if (status < 0) {
4736                 mlog_errno(status);
4737                 goto out_mutex;
4738         }
4739
4740         handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_UPDATE);
4741         if (IS_ERR(handle)) {
4742                 status = PTR_ERR(handle);
4743                 mlog_errno(status);
4744                 goto out_unlock;
4745         }
4746
4747         status = ocfs2_replay_truncate_records(osb, handle, data_alloc_inode,
4748                                                data_alloc_bh);
4749         if (status < 0)
4750                 mlog_errno(status);
4751
4752         ocfs2_commit_trans(osb, handle);
4753
4754 out_unlock:
4755         brelse(data_alloc_bh);
4756         ocfs2_inode_unlock(data_alloc_inode, 1);
4757
4758 out_mutex:
4759         mutex_unlock(&data_alloc_inode->i_mutex);
4760         iput(data_alloc_inode);
4761
4762 out:
4763         mlog_exit(status);
4764         return status;
4765 }
4766
4767 int ocfs2_flush_truncate_log(struct ocfs2_super *osb)
4768 {
4769         int status;
4770         struct inode *tl_inode = osb->osb_tl_inode;
4771
4772         mutex_lock(&tl_inode->i_mutex);
4773         status = __ocfs2_flush_truncate_log(osb);
4774         mutex_unlock(&tl_inode->i_mutex);
4775
4776         return status;
4777 }
4778
4779 static void ocfs2_truncate_log_worker(struct work_struct *work)
4780 {
4781         int status;
4782         struct ocfs2_super *osb =
4783                 container_of(work, struct ocfs2_super,
4784                              osb_truncate_log_wq.work);
4785
4786         mlog_entry_void();
4787
4788         status = ocfs2_flush_truncate_log(osb);
4789         if (status < 0)
4790                 mlog_errno(status);
4791
4792         mlog_exit(status);
4793 }
4794
4795 #define OCFS2_TRUNCATE_LOG_FLUSH_INTERVAL (2 * HZ)
4796 void ocfs2_schedule_truncate_log_flush(struct ocfs2_super *osb,
4797                                        int cancel)
4798 {
4799         if (osb->osb_tl_inode) {
4800                 /* We want to push off log flushes while truncates are
4801                  * still running. */
4802                 if (cancel)
4803                         cancel_delayed_work(&osb->osb_truncate_log_wq);
4804
4805                 queue_delayed_work(ocfs2_wq, &osb->osb_truncate_log_wq,
4806                                    OCFS2_TRUNCATE_LOG_FLUSH_INTERVAL);
4807         }
4808 }
4809
4810 static int ocfs2_get_truncate_log_info(struct ocfs2_super *osb,
4811                                        int slot_num,
4812                                        struct inode **tl_inode,
4813                                        struct buffer_head **tl_bh)
4814 {
4815         int status;
4816         struct inode *inode = NULL;
4817         struct buffer_head *bh = NULL;
4818
4819         inode = ocfs2_get_system_file_inode(osb,
4820                                            TRUNCATE_LOG_SYSTEM_INODE,
4821                                            slot_num);
4822         if (!inode) {
4823                 status = -EINVAL;
4824                 mlog(ML_ERROR, "Could not get load truncate log inode!\n");
4825                 goto bail;
4826         }
4827
4828         status = ocfs2_read_block(osb, OCFS2_I(inode)->ip_blkno, &bh,
4829                                   OCFS2_BH_CACHED, inode);
4830         if (status < 0) {
4831                 iput(inode);
4832                 mlog_errno(status);
4833                 goto bail;
4834         }
4835
4836         *tl_inode = inode;
4837         *tl_bh    = bh;
4838 bail:
4839         mlog_exit(status);
4840         return status;
4841 }
4842
4843 /* called during the 1st stage of node recovery. we stamp a clean
4844  * truncate log and pass back a copy for processing later. if the
4845  * truncate log does not require processing, a *tl_copy is set to
4846  * NULL. */
4847 int ocfs2_begin_truncate_log_recovery(struct ocfs2_super *osb,
4848                                       int slot_num,
4849                                       struct ocfs2_dinode **tl_copy)
4850 {
4851         int status;
4852         struct inode *tl_inode = NULL;
4853         struct buffer_head *tl_bh = NULL;
4854         struct ocfs2_dinode *di;
4855         struct ocfs2_truncate_log *tl;
4856
4857         *tl_copy = NULL;
4858
4859         mlog(0, "recover truncate log from slot %d\n", slot_num);
4860
4861         status = ocfs2_get_truncate_log_info(osb, slot_num, &tl_inode, &tl_bh);
4862         if (status < 0) {
4863                 mlog_errno(status);
4864                 goto bail;
4865         }
4866
4867         di = (struct ocfs2_dinode *) tl_bh->b_data;
4868         tl = &di->id2.i_dealloc;
4869         if (!OCFS2_IS_VALID_DINODE(di)) {
4870                 OCFS2_RO_ON_INVALID_DINODE(tl_inode->i_sb, di);
4871                 status = -EIO;
4872                 goto bail;
4873         }
4874
4875         if (le16_to_cpu(tl->tl_used)) {
4876                 mlog(0, "We'll have %u logs to recover\n",
4877                      le16_to_cpu(tl->tl_used));
4878
4879                 *tl_copy = kmalloc(tl_bh->b_size, GFP_KERNEL);
4880                 if (!(*tl_copy)) {
4881                         status = -ENOMEM;
4882                         mlog_errno(status);
4883                         goto bail;
4884                 }
4885
4886                 /* Assuming the write-out below goes well, this copy
4887                  * will be passed back to recovery for processing. */
4888                 memcpy(*tl_copy, tl_bh->b_data, tl_bh->b_size);
4889
4890                 /* All we need to do to clear the truncate log is set
4891                  * tl_used. */
4892                 tl->tl_used = 0;
4893
4894                 status = ocfs2_write_block(osb, tl_bh, tl_inode);
4895                 if (status < 0) {
4896                         mlog_errno(status);
4897                         goto bail;
4898                 }
4899         }
4900
4901 bail:
4902         if (tl_inode)
4903                 iput(tl_inode);
4904         if (tl_bh)
4905                 brelse(tl_bh);
4906
4907         if (status < 0 && (*tl_copy)) {
4908                 kfree(*tl_copy);
4909                 *tl_copy = NULL;
4910         }
4911
4912         mlog_exit(status);
4913         return status;
4914 }
4915
4916 int ocfs2_complete_truncate_log_recovery(struct ocfs2_super *osb,
4917                                          struct ocfs2_dinode *tl_copy)
4918 {
4919         int status = 0;
4920         int i;
4921         unsigned int clusters, num_recs, start_cluster;
4922         u64 start_blk;
4923         handle_t *handle;
4924         struct inode *tl_inode = osb->osb_tl_inode;
4925         struct ocfs2_truncate_log *tl;
4926
4927         mlog_entry_void();
4928
4929         if (OCFS2_I(tl_inode)->ip_blkno == le64_to_cpu(tl_copy->i_blkno)) {
4930                 mlog(ML_ERROR, "Asked to recover my own truncate log!\n");
4931                 return -EINVAL;
4932         }
4933
4934         tl = &tl_copy->id2.i_dealloc;
4935         num_recs = le16_to_cpu(tl->tl_used);
4936         mlog(0, "cleanup %u records from %llu\n", num_recs,
4937              (unsigned long long)le64_to_cpu(tl_copy->i_blkno));
4938
4939         mutex_lock(&tl_inode->i_mutex);
4940         for(i = 0; i < num_recs; i++) {
4941                 if (ocfs2_truncate_log_needs_flush(osb)) {
4942                         status = __ocfs2_flush_truncate_log(osb);
4943                         if (status < 0) {
4944                                 mlog_errno(status);
4945                                 goto bail_up;
4946                         }
4947                 }
4948
4949                 handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_UPDATE);
4950                 if (IS_ERR(handle)) {
4951                         status = PTR_ERR(handle);
4952                         mlog_errno(status);
4953                         goto bail_up;
4954                 }
4955
4956                 clusters = le32_to_cpu(tl->tl_recs[i].t_clusters);
4957                 start_cluster = le32_to_cpu(tl->tl_recs[i].t_start);
4958                 start_blk = ocfs2_clusters_to_blocks(osb->sb, start_cluster);
4959
4960                 status = ocfs2_truncate_log_append(osb, handle,
4961                                                    start_blk, clusters);
4962                 ocfs2_commit_trans(osb, handle);
4963                 if (status < 0) {
4964                         mlog_errno(status);
4965                         goto bail_up;
4966                 }
4967         }
4968
4969 bail_up:
4970         mutex_unlock(&tl_inode->i_mutex);
4971
4972         mlog_exit(status);
4973         return status;
4974 }
4975
4976 void ocfs2_truncate_log_shutdown(struct ocfs2_super *osb)
4977 {
4978         int status;
4979         struct inode *tl_inode = osb->osb_tl_inode;
4980
4981         mlog_entry_void();
4982
4983         if (tl_inode) {
4984                 cancel_delayed_work(&osb->osb_truncate_log_wq);
4985                 flush_workqueue(ocfs2_wq);
4986
4987                 status = ocfs2_flush_truncate_log(osb);
4988                 if (status < 0)
4989                         mlog_errno(status);
4990
4991                 brelse(osb->osb_tl_bh);
4992                 iput(osb->osb_tl_inode);
4993         }
4994
4995         mlog_exit_void();
4996 }
4997
4998 int ocfs2_truncate_log_init(struct ocfs2_super *osb)
4999 {
5000         int status;
5001         struct inode *tl_inode = NULL;
5002         struct buffer_head *tl_bh = NULL;
5003
5004         mlog_entry_void();
5005
5006         status = ocfs2_get_truncate_log_info(osb,
5007                                              osb->slot_num,
5008                                              &tl_inode,
5009                                              &tl_bh);
5010         if (status < 0)
5011                 mlog_errno(status);
5012
5013         /* ocfs2_truncate_log_shutdown keys on the existence of
5014          * osb->osb_tl_inode so we don't set any of the osb variables
5015          * until we're sure all is well. */
5016         INIT_DELAYED_WORK(&osb->osb_truncate_log_wq,
5017                           ocfs2_truncate_log_worker);
5018         osb->osb_tl_bh    = tl_bh;
5019         osb->osb_tl_inode = tl_inode;
5020
5021         mlog_exit(status);
5022         return status;
5023 }
5024
5025 /*
5026  * Delayed de-allocation of suballocator blocks.
5027  *
5028  * Some sets of block de-allocations might involve multiple suballocator inodes.
5029  *
5030  * The locking for this can get extremely complicated, especially when
5031  * the suballocator inodes to delete from aren't known until deep
5032  * within an unrelated codepath.
5033  *
5034  * ocfs2_extent_block structures are a good example of this - an inode
5035  * btree could have been grown by any number of nodes each allocating
5036  * out of their own suballoc inode.
5037  *
5038  * These structures allow the delay of block de-allocation until a
5039  * later time, when locking of multiple cluster inodes won't cause
5040  * deadlock.
5041  */
5042
5043 /*
5044  * Describes a single block free from a suballocator
5045  */
5046 struct ocfs2_cached_block_free {
5047         struct ocfs2_cached_block_free          *free_next;
5048         u64                                     free_blk;
5049         unsigned int                            free_bit;
5050 };
5051
5052 struct ocfs2_per_slot_free_list {
5053         struct ocfs2_per_slot_free_list         *f_next_suballocator;
5054         int                                     f_inode_type;
5055         int                                     f_slot;
5056         struct ocfs2_cached_block_free          *f_first;
5057 };
5058
5059 static int ocfs2_free_cached_items(struct ocfs2_super *osb,
5060                                    int sysfile_type,
5061                                    int slot,
5062                                    struct ocfs2_cached_block_free *head)
5063 {
5064         int ret;
5065         u64 bg_blkno;
5066         handle_t *handle;
5067         struct inode *inode;
5068         struct buffer_head *di_bh = NULL;
5069         struct ocfs2_cached_block_free *tmp;
5070
5071         inode = ocfs2_get_system_file_inode(osb, sysfile_type, slot);
5072         if (!inode) {
5073                 ret = -EINVAL;
5074                 mlog_errno(ret);
5075                 goto out;
5076         }
5077
5078         mutex_lock(&inode->i_mutex);
5079
5080         ret = ocfs2_inode_lock(inode, &di_bh, 1);
5081         if (ret) {
5082                 mlog_errno(ret);
5083                 goto out_mutex;
5084         }
5085
5086         handle = ocfs2_start_trans(osb, OCFS2_SUBALLOC_FREE);
5087         if (IS_ERR(handle)) {
5088                 ret = PTR_ERR(handle);
5089                 mlog_errno(ret);
5090                 goto out_unlock;
5091         }
5092
5093         while (head) {
5094                 bg_blkno = ocfs2_which_suballoc_group(head->free_blk,
5095                                                       head->free_bit);
5096                 mlog(0, "Free bit: (bit %u, blkno %llu)\n",
5097                      head->free_bit, (unsigned long long)head->free_blk);
5098
5099                 ret = ocfs2_free_suballoc_bits(handle, inode, di_bh,
5100                                                head->free_bit, bg_blkno, 1);
5101                 if (ret) {
5102                         mlog_errno(ret);
5103                         goto out_journal;
5104                 }
5105
5106                 ret = ocfs2_extend_trans(handle, OCFS2_SUBALLOC_FREE);
5107                 if (ret) {
5108                         mlog_errno(ret);
5109                         goto out_journal;
5110                 }
5111
5112                 tmp = head;
5113                 head = head->free_next;
5114                 kfree(tmp);
5115         }
5116
5117 out_journal:
5118         ocfs2_commit_trans(osb, handle);
5119
5120 out_unlock:
5121         ocfs2_inode_unlock(inode, 1);
5122         brelse(di_bh);
5123 out_mutex:
5124         mutex_unlock(&inode->i_mutex);
5125         iput(inode);
5126 out:
5127         while(head) {
5128                 /* Premature exit may have left some dangling items. */
5129                 tmp = head;
5130                 head = head->free_next;
5131                 kfree(tmp);
5132         }
5133
5134         return ret;
5135 }
5136
5137 int ocfs2_run_deallocs(struct ocfs2_super *osb,
5138                        struct ocfs2_cached_dealloc_ctxt *ctxt)
5139 {
5140         int ret = 0, ret2;
5141         struct ocfs2_per_slot_free_list *fl;
5142
5143         if (!ctxt)
5144                 return 0;
5145
5146         while (ctxt->c_first_suballocator) {
5147                 fl = ctxt->c_first_suballocator;
5148
5149                 if (fl->f_first) {
5150                         mlog(0, "Free items: (type %u, slot %d)\n",
5151                              fl->f_inode_type, fl->f_slot);
5152                         ret2 = ocfs2_free_cached_items(osb, fl->f_inode_type,
5153                                                        fl->f_slot, fl->f_first);
5154                         if (ret2)
5155                                 mlog_errno(ret2);
5156                         if (!ret)
5157                                 ret = ret2;
5158                 }
5159
5160                 ctxt->c_first_suballocator = fl->f_next_suballocator;
5161                 kfree(fl);
5162         }
5163
5164         return ret;
5165 }
5166
5167 static struct ocfs2_per_slot_free_list *
5168 ocfs2_find_per_slot_free_list(int type,
5169                               int slot,
5170                               struct ocfs2_cached_dealloc_ctxt *ctxt)
5171 {
5172         struct ocfs2_per_slot_free_list *fl = ctxt->c_first_suballocator;
5173
5174         while (fl) {
5175                 if (fl->f_inode_type == type && fl->f_slot == slot)
5176                         return fl;
5177
5178                 fl = fl->f_next_suballocator;
5179         }
5180
5181         fl = kmalloc(sizeof(*fl), GFP_NOFS);
5182         if (fl) {
5183                 fl->f_inode_type = type;
5184                 fl->f_slot = slot;
5185                 fl->f_first = NULL;
5186                 fl->f_next_suballocator = ctxt->c_first_suballocator;
5187
5188                 ctxt->c_first_suballocator = fl;
5189         }
5190         return fl;
5191 }
5192
5193 static int ocfs2_cache_block_dealloc(struct ocfs2_cached_dealloc_ctxt *ctxt,
5194                                      int type, int slot, u64 blkno,
5195                                      unsigned int bit)
5196 {
5197         int ret;
5198         struct ocfs2_per_slot_free_list *fl;
5199         struct ocfs2_cached_block_free *item;
5200
5201         fl = ocfs2_find_per_slot_free_list(type, slot, ctxt);
5202         if (fl == NULL) {
5203                 ret = -ENOMEM;
5204                 mlog_errno(ret);
5205                 goto out;
5206         }
5207
5208         item = kmalloc(sizeof(*item), GFP_NOFS);
5209         if (item == NULL) {
5210                 ret = -ENOMEM;
5211                 mlog_errno(ret);
5212                 goto out;
5213         }
5214
5215         mlog(0, "Insert: (type %d, slot %u, bit %u, blk %llu)\n",
5216              type, slot, bit, (unsigned long long)blkno);
5217
5218         item->free_blk = blkno;
5219         item->free_bit = bit;
5220         item->free_next = fl->f_first;
5221
5222         fl->f_first = item;
5223
5224         ret = 0;
5225 out:
5226         return ret;
5227 }
5228
5229 static int ocfs2_cache_extent_block_free(struct ocfs2_cached_dealloc_ctxt *ctxt,
5230                                          struct ocfs2_extent_block *eb)
5231 {
5232         return ocfs2_cache_block_dealloc(ctxt, EXTENT_ALLOC_SYSTEM_INODE,
5233                                          le16_to_cpu(eb->h_suballoc_slot),
5234                                          le64_to_cpu(eb->h_blkno),
5235                                          le16_to_cpu(eb->h_suballoc_bit));
5236 }
5237
5238 /* This function will figure out whether the currently last extent
5239  * block will be deleted, and if it will, what the new last extent
5240  * block will be so we can update his h_next_leaf_blk field, as well
5241  * as the dinodes i_last_eb_blk */
5242 static int ocfs2_find_new_last_ext_blk(struct inode *inode,
5243                                        unsigned int clusters_to_del,
5244                                        struct ocfs2_path *path,
5245                                        struct buffer_head **new_last_eb)
5246 {
5247         int next_free, ret = 0;
5248         u32 cpos;
5249         struct ocfs2_extent_rec *rec;
5250         struct ocfs2_extent_block *eb;
5251         struct ocfs2_extent_list *el;
5252         struct buffer_head *bh = NULL;
5253
5254         *new_last_eb = NULL;
5255
5256         /* we have no tree, so of course, no last_eb. */
5257         if (!path->p_tree_depth)
5258                 goto out;
5259
5260         /* trunc to zero special case - this makes tree_depth = 0
5261          * regardless of what it is.  */
5262         if (OCFS2_I(inode)->ip_clusters == clusters_to_del)
5263                 goto out;
5264
5265         el = path_leaf_el(path);
5266         BUG_ON(!el->l_next_free_rec);
5267
5268         /*
5269          * Make sure that this extent list will actually be empty
5270          * after we clear away the data. We can shortcut out if
5271          * there's more than one non-empty extent in the
5272          * list. Otherwise, a check of the remaining extent is
5273          * necessary.
5274          */
5275         next_free = le16_to_cpu(el->l_next_free_rec);
5276         rec = NULL;
5277         if (ocfs2_is_empty_extent(&el->l_recs[0])) {
5278                 if (next_free > 2)
5279                         goto out;
5280
5281                 /* We may have a valid extent in index 1, check it. */
5282                 if (next_free == 2)
5283                         rec = &el->l_recs[1];
5284
5285                 /*
5286                  * Fall through - no more nonempty extents, so we want
5287                  * to delete this leaf.
5288                  */
5289         } else {
5290                 if (next_free > 1)
5291                         goto out;
5292
5293                 rec = &el->l_recs[0];
5294         }
5295
5296         if (rec) {
5297                 /*
5298                  * Check it we'll only be trimming off the end of this
5299                  * cluster.
5300                  */
5301                 if (le16_to_cpu(rec->e_leaf_clusters) > clusters_to_del)
5302                         goto out;
5303         }
5304
5305         ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, path, &cpos);
5306         if (ret) {
5307                 mlog_errno(ret);
5308                 goto out;
5309         }
5310
5311         ret = ocfs2_find_leaf(inode, path_root_el(path), cpos, &bh);
5312         if (ret) {
5313                 mlog_errno(ret);
5314                 goto out;
5315         }
5316
5317         eb = (struct ocfs2_extent_block *) bh->b_data;
5318         el = &eb->h_list;
5319         if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
5320                 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
5321                 ret = -EROFS;
5322                 goto out;
5323         }
5324
5325         *new_last_eb = bh;
5326         get_bh(*new_last_eb);
5327         mlog(0, "returning block %llu, (cpos: %u)\n",
5328              (unsigned long long)le64_to_cpu(eb->h_blkno), cpos);
5329 out:
5330         brelse(bh);
5331
5332         return ret;
5333 }
5334
5335 /*
5336  * Trim some clusters off the rightmost edge of a tree. Only called
5337  * during truncate.
5338  *
5339  * The caller needs to:
5340  *   - start journaling of each path component.
5341  *   - compute and fully set up any new last ext block
5342  */
5343 static int ocfs2_trim_tree(struct inode *inode, struct ocfs2_path *path,
5344                            handle_t *handle, struct ocfs2_truncate_context *tc,
5345                            u32 clusters_to_del, u64 *delete_start)
5346 {
5347         int ret, i, index = path->p_tree_depth;
5348         u32 new_edge = 0;
5349         u64 deleted_eb = 0;
5350         struct buffer_head *bh;
5351         struct ocfs2_extent_list *el;
5352         struct ocfs2_extent_rec *rec;
5353
5354         *delete_start = 0;
5355
5356         while (index >= 0) {
5357                 bh = path->p_node[index].bh;
5358                 el = path->p_node[index].el;
5359
5360                 mlog(0, "traveling tree (index = %d, block = %llu)\n",
5361                      index,  (unsigned long long)bh->b_blocknr);
5362
5363                 BUG_ON(le16_to_cpu(el->l_next_free_rec) == 0);
5364
5365                 if (index !=
5366                     (path->p_tree_depth - le16_to_cpu(el->l_tree_depth))) {
5367                         ocfs2_error(inode->i_sb,
5368                                     "Inode %lu has invalid ext. block %llu",
5369                                     inode->i_ino,
5370                                     (unsigned long long)bh->b_blocknr);
5371                         ret = -EROFS;
5372                         goto out;
5373                 }
5374
5375 find_tail_record:
5376                 i = le16_to_cpu(el->l_next_free_rec) - 1;
5377                 rec = &el->l_recs[i];
5378
5379                 mlog(0, "Extent list before: record %d: (%u, %u, %llu), "
5380                      "next = %u\n", i, le32_to_cpu(rec->e_cpos),
5381                      ocfs2_rec_clusters(el, rec),
5382                      (unsigned long long)le64_to_cpu(rec->e_blkno),
5383                      le16_to_cpu(el->l_next_free_rec));
5384
5385                 BUG_ON(ocfs2_rec_clusters(el, rec) < clusters_to_del);
5386
5387                 if (le16_to_cpu(el->l_tree_depth) == 0) {
5388                         /*
5389                          * If the leaf block contains a single empty
5390                          * extent and no records, we can just remove
5391                          * the block.
5392                          */
5393                         if (i == 0 && ocfs2_is_empty_extent(rec)) {
5394                                 memset(rec, 0,
5395                                        sizeof(struct ocfs2_extent_rec));
5396                                 el->l_next_free_rec = cpu_to_le16(0);
5397
5398                                 goto delete;
5399                         }
5400
5401                         /*
5402                          * Remove any empty extents by shifting things
5403                          * left. That should make life much easier on
5404                          * the code below. This condition is rare
5405                          * enough that we shouldn't see a performance
5406                          * hit.
5407                          */
5408                         if (ocfs2_is_empty_extent(&el->l_recs[0])) {
5409                                 le16_add_cpu(&el->l_next_free_rec, -1);
5410
5411                                 for(i = 0;
5412                                     i < le16_to_cpu(el->l_next_free_rec); i++)
5413                                         el->l_recs[i] = el->l_recs[i + 1];
5414
5415                                 memset(&el->l_recs[i], 0,
5416                                        sizeof(struct ocfs2_extent_rec));
5417
5418                                 /*
5419                                  * We've modified our extent list. The
5420                                  * simplest way to handle this change
5421                                  * is to being the search from the
5422                                  * start again.
5423                                  */
5424                                 goto find_tail_record;
5425                         }
5426
5427                         le16_add_cpu(&rec->e_leaf_clusters, -clusters_to_del);
5428
5429                         /*
5430                          * We'll use "new_edge" on our way back up the
5431                          * tree to know what our rightmost cpos is.
5432                          */
5433                         new_edge = le16_to_cpu(rec->e_leaf_clusters);
5434                         new_edge += le32_to_cpu(rec->e_cpos);
5435
5436                         /*
5437                          * The caller will use this to delete data blocks.
5438                          */
5439                         *delete_start = le64_to_cpu(rec->e_blkno)
5440                                 + ocfs2_clusters_to_blocks(inode->i_sb,
5441                                         le16_to_cpu(rec->e_leaf_clusters));
5442
5443                         /*
5444                          * If it's now empty, remove this record.
5445                          */
5446                         if (le16_to_cpu(rec->e_leaf_clusters) == 0) {
5447                                 memset(rec, 0,
5448                                        sizeof(struct ocfs2_extent_rec));
5449                                 le16_add_cpu(&el->l_next_free_rec, -1);
5450                         }
5451                 } else {
5452                         if (le64_to_cpu(rec->e_blkno) == deleted_eb) {
5453                                 memset(rec, 0,
5454                                        sizeof(struct ocfs2_extent_rec));
5455                                 le16_add_cpu(&el->l_next_free_rec, -1);
5456
5457                                 goto delete;
5458                         }
5459
5460                         /* Can this actually happen? */
5461                         if (le16_to_cpu(el->l_next_free_rec) == 0)
5462                                 goto delete;
5463
5464                         /*
5465                          * We never actually deleted any clusters
5466                          * because our leaf was empty. There's no
5467                          * reason to adjust the rightmost edge then.
5468                          */
5469                         if (new_edge == 0)
5470                                 goto delete;
5471
5472                         rec->e_int_clusters = cpu_to_le32(new_edge);
5473                         le32_add_cpu(&rec->e_int_clusters,
5474                                      -le32_to_cpu(rec->e_cpos));
5475
5476                          /*
5477                           * A deleted child record should have been
5478                           * caught above.
5479                           */
5480                          BUG_ON(le32_to_cpu(rec->e_int_clusters) == 0);
5481                 }
5482
5483 delete:
5484                 ret = ocfs2_journal_dirty(handle, bh);
5485                 if (ret) {
5486                         mlog_errno(ret);
5487                         goto out;
5488                 }
5489
5490                 mlog(0, "extent list container %llu, after: record %d: "
5491                      "(%u, %u, %llu), next = %u.\n",
5492                      (unsigned long long)bh->b_blocknr, i,
5493                      le32_to_cpu(rec->e_cpos), ocfs2_rec_clusters(el, rec),
5494                      (unsigned long long)le64_to_cpu(rec->e_blkno),
5495                      le16_to_cpu(el->l_next_free_rec));
5496
5497                 /*
5498                  * We must be careful to only attempt delete of an
5499                  * extent block (and not the root inode block).
5500                  */
5501                 if (index > 0 && le16_to_cpu(el->l_next_free_rec) == 0) {
5502                         struct ocfs2_extent_block *eb =
5503                                 (struct ocfs2_extent_block *)bh->b_data;
5504
5505                         /*
5506                          * Save this for use when processing the
5507                          * parent block.
5508                          */
5509                         deleted_eb = le64_to_cpu(eb->h_blkno);
5510
5511                         mlog(0, "deleting this extent block.\n");
5512
5513                         ocfs2_remove_from_cache(inode, bh);
5514
5515                         BUG_ON(ocfs2_rec_clusters(el, &el->l_recs[0]));
5516                         BUG_ON(le32_to_cpu(el->l_recs[0].e_cpos));
5517                         BUG_ON(le64_to_cpu(el->l_recs[0].e_blkno));
5518
5519                         ret = ocfs2_cache_extent_block_free(&tc->tc_dealloc, eb);
5520                         /* An error here is not fatal. */
5521                         if (ret < 0)
5522                                 mlog_errno(ret);
5523                 } else {
5524                         deleted_eb = 0;
5525                 }
5526
5527                 index--;
5528         }
5529
5530         ret = 0;
5531 out:
5532         return ret;
5533 }
5534
5535 static int ocfs2_do_truncate(struct ocfs2_super *osb,
5536                              unsigned int clusters_to_del,
5537                              struct inode *inode,
5538                              struct buffer_head *fe_bh,
5539                              handle_t *handle,
5540                              struct ocfs2_truncate_context *tc,
5541                              struct ocfs2_path *path)
5542 {
5543         int status;
5544         struct ocfs2_dinode *fe;
5545         struct ocfs2_extent_block *last_eb = NULL;
5546         struct ocfs2_extent_list *el;
5547         struct buffer_head *last_eb_bh = NULL;
5548         u64 delete_blk = 0;
5549
5550         fe = (struct ocfs2_dinode *) fe_bh->b_data;
5551
5552         status = ocfs2_find_new_last_ext_blk(inode, clusters_to_del,
5553                                              path, &last_eb_bh);
5554         if (status < 0) {
5555                 mlog_errno(status);
5556                 goto bail;
5557         }
5558
5559         /*
5560          * Each component will be touched, so we might as well journal
5561          * here to avoid having to handle errors later.
5562          */
5563         status = ocfs2_journal_access_path(inode, handle, path);
5564         if (status < 0) {
5565                 mlog_errno(status);
5566                 goto bail;
5567         }
5568
5569         if (last_eb_bh) {
5570                 status = ocfs2_journal_access(handle, inode, last_eb_bh,
5571                                               OCFS2_JOURNAL_ACCESS_WRITE);
5572                 if (status < 0) {
5573                         mlog_errno(status);
5574                         goto bail;
5575                 }
5576
5577                 last_eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
5578         }
5579
5580         el = &(fe->id2.i_list);
5581
5582         /*
5583          * Lower levels depend on this never happening, but it's best
5584          * to check it up here before changing the tree.
5585          */
5586         if (el->l_tree_depth && el->l_recs[0].e_int_clusters == 0) {
5587                 ocfs2_error(inode->i_sb,
5588                             "Inode %lu has an empty extent record, depth %u\n",
5589                             inode->i_ino, le16_to_cpu(el->l_tree_depth));
5590                 status = -EROFS;
5591                 goto bail;
5592         }
5593
5594         spin_lock(&OCFS2_I(inode)->ip_lock);
5595         OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters) -
5596                                       clusters_to_del;
5597         spin_unlock(&OCFS2_I(inode)->ip_lock);
5598         le32_add_cpu(&fe->i_clusters, -clusters_to_del);
5599         inode->i_blocks = ocfs2_inode_sector_count(inode);
5600
5601         status = ocfs2_trim_tree(inode, path, handle, tc,
5602                                  clusters_to_del, &delete_blk);
5603         if (status) {
5604                 mlog_errno(status);
5605                 goto bail;
5606         }
5607
5608         if (le32_to_cpu(fe->i_clusters) == 0) {
5609                 /* trunc to zero is a special case. */
5610                 el->l_tree_depth = 0;
5611                 fe->i_last_eb_blk = 0;
5612         } else if (last_eb)
5613                 fe->i_last_eb_blk = last_eb->h_blkno;
5614
5615         status = ocfs2_journal_dirty(handle, fe_bh);
5616         if (status < 0) {
5617                 mlog_errno(status);
5618                 goto bail;
5619         }
5620
5621         if (last_eb) {
5622                 /* If there will be a new last extent block, then by
5623                  * definition, there cannot be any leaves to the right of
5624                  * him. */
5625                 last_eb->h_next_leaf_blk = 0;
5626                 status = ocfs2_journal_dirty(handle, last_eb_bh);
5627                 if (status < 0) {
5628                         mlog_errno(status);
5629                         goto bail;
5630                 }
5631         }
5632
5633         if (delete_blk) {
5634                 status = ocfs2_truncate_log_append(osb, handle, delete_blk,
5635                                                    clusters_to_del);
5636                 if (status < 0) {
5637                         mlog_errno(status);
5638                         goto bail;
5639                 }
5640         }
5641         status = 0;
5642 bail:
5643
5644         mlog_exit(status);
5645         return status;
5646 }
5647
5648 static int ocfs2_writeback_zero_func(handle_t *handle, struct buffer_head *bh)
5649 {
5650         set_buffer_uptodate(bh);
5651         mark_buffer_dirty(bh);
5652         return 0;
5653 }
5654
5655 static int ocfs2_ordered_zero_func(handle_t *handle, struct buffer_head *bh)
5656 {
5657         set_buffer_uptodate(bh);
5658         mark_buffer_dirty(bh);
5659         return ocfs2_journal_dirty_data(handle, bh);
5660 }
5661
5662 static void ocfs2_map_and_dirty_page(struct inode *inode, handle_t *handle,
5663                                      unsigned int from, unsigned int to,
5664                                      struct page *page, int zero, u64 *phys)
5665 {
5666         int ret, partial = 0;
5667
5668         ret = ocfs2_map_page_blocks(page, phys, inode, from, to, 0);
5669         if (ret)
5670                 mlog_errno(ret);
5671
5672         if (zero)
5673                 zero_user_segment(page, from, to);
5674
5675         /*
5676          * Need to set the buffers we zero'd into uptodate
5677          * here if they aren't - ocfs2_map_page_blocks()
5678          * might've skipped some
5679          */
5680         if (ocfs2_should_order_data(inode)) {
5681                 ret = walk_page_buffers(handle,
5682                                         page_buffers(page),