blob: 9a40603c4d4b2b400a6df91c34b76525dac8945d [file] [log] [blame]
Mark Fashehccd979b2005-12-15 14:31:24 -08001/* -*- 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
31#define MLOG_MASK_PREFIX ML_DISK_ALLOC
32#include <cluster/masklog.h>
33
34#include "ocfs2.h"
35
36#include "alloc.h"
37#include "dlmglue.h"
38#include "extent_map.h"
39#include "inode.h"
40#include "journal.h"
41#include "localalloc.h"
42#include "suballoc.h"
43#include "sysfile.h"
44#include "file.h"
45#include "super.h"
46#include "uptodate.h"
47
48#include "buffer_head_io.h"
49
Mark Fashehccd979b2005-12-15 14:31:24 -080050static void ocfs2_free_truncate_context(struct ocfs2_truncate_context *tc);
51
Mark Fashehdcd05382007-01-16 11:32:23 -080052/*
53 * Structures which describe a path through a btree, and functions to
54 * manipulate them.
55 *
56 * The idea here is to be as generic as possible with the tree
57 * manipulation code.
58 */
59struct ocfs2_path_item {
60 struct buffer_head *bh;
61 struct ocfs2_extent_list *el;
62};
63
64#define OCFS2_MAX_PATH_DEPTH 5
65
66struct ocfs2_path {
67 int p_tree_depth;
68 struct ocfs2_path_item p_node[OCFS2_MAX_PATH_DEPTH];
69};
70
71#define path_root_bh(_path) ((_path)->p_node[0].bh)
72#define path_root_el(_path) ((_path)->p_node[0].el)
73#define path_leaf_bh(_path) ((_path)->p_node[(_path)->p_tree_depth].bh)
74#define path_leaf_el(_path) ((_path)->p_node[(_path)->p_tree_depth].el)
75#define path_num_items(_path) ((_path)->p_tree_depth + 1)
76
77/*
78 * Reset the actual path elements so that we can re-use the structure
79 * to build another path. Generally, this involves freeing the buffer
80 * heads.
81 */
82static void ocfs2_reinit_path(struct ocfs2_path *path, int keep_root)
83{
84 int i, start = 0, depth = 0;
85 struct ocfs2_path_item *node;
86
87 if (keep_root)
88 start = 1;
89
90 for(i = start; i < path_num_items(path); i++) {
91 node = &path->p_node[i];
92
93 brelse(node->bh);
94 node->bh = NULL;
95 node->el = NULL;
96 }
97
98 /*
99 * Tree depth may change during truncate, or insert. If we're
100 * keeping the root extent list, then make sure that our path
101 * structure reflects the proper depth.
102 */
103 if (keep_root)
104 depth = le16_to_cpu(path_root_el(path)->l_tree_depth);
105
106 path->p_tree_depth = depth;
107}
108
109static void ocfs2_free_path(struct ocfs2_path *path)
110{
111 if (path) {
112 ocfs2_reinit_path(path, 0);
113 kfree(path);
114 }
115}
116
117/*
118 * Make the *dest path the same as src and re-initialize src path to
119 * have a root only.
120 */
121static void ocfs2_mv_path(struct ocfs2_path *dest, struct ocfs2_path *src)
122{
123 int i;
124
125 BUG_ON(path_root_bh(dest) != path_root_bh(src));
126
127 for(i = 1; i < OCFS2_MAX_PATH_DEPTH; i++) {
128 brelse(dest->p_node[i].bh);
129
130 dest->p_node[i].bh = src->p_node[i].bh;
131 dest->p_node[i].el = src->p_node[i].el;
132
133 src->p_node[i].bh = NULL;
134 src->p_node[i].el = NULL;
135 }
136}
137
138/*
139 * Insert an extent block at given index.
140 *
141 * This will not take an additional reference on eb_bh.
142 */
143static inline void ocfs2_path_insert_eb(struct ocfs2_path *path, int index,
144 struct buffer_head *eb_bh)
145{
146 struct ocfs2_extent_block *eb = (struct ocfs2_extent_block *)eb_bh->b_data;
147
148 /*
149 * Right now, no root bh is an extent block, so this helps
150 * catch code errors with dinode trees. The assertion can be
151 * safely removed if we ever need to insert extent block
152 * structures at the root.
153 */
154 BUG_ON(index == 0);
155
156 path->p_node[index].bh = eb_bh;
157 path->p_node[index].el = &eb->h_list;
158}
159
160static struct ocfs2_path *ocfs2_new_path(struct buffer_head *root_bh,
161 struct ocfs2_extent_list *root_el)
162{
163 struct ocfs2_path *path;
164
165 BUG_ON(le16_to_cpu(root_el->l_tree_depth) >= OCFS2_MAX_PATH_DEPTH);
166
167 path = kzalloc(sizeof(*path), GFP_NOFS);
168 if (path) {
169 path->p_tree_depth = le16_to_cpu(root_el->l_tree_depth);
170 get_bh(root_bh);
171 path_root_bh(path) = root_bh;
172 path_root_el(path) = root_el;
173 }
174
175 return path;
176}
177
178/*
179 * Allocate and initialize a new path based on a disk inode tree.
180 */
181static struct ocfs2_path *ocfs2_new_inode_path(struct buffer_head *di_bh)
182{
183 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
184 struct ocfs2_extent_list *el = &di->id2.i_list;
185
186 return ocfs2_new_path(di_bh, el);
187}
188
189/*
190 * Convenience function to journal all components in a path.
191 */
192static int ocfs2_journal_access_path(struct inode *inode, handle_t *handle,
193 struct ocfs2_path *path)
194{
195 int i, ret = 0;
196
197 if (!path)
198 goto out;
199
200 for(i = 0; i < path_num_items(path); i++) {
201 ret = ocfs2_journal_access(handle, inode, path->p_node[i].bh,
202 OCFS2_JOURNAL_ACCESS_WRITE);
203 if (ret < 0) {
204 mlog_errno(ret);
205 goto out;
206 }
207 }
208
209out:
210 return ret;
211}
212
213enum ocfs2_contig_type {
214 CONTIG_NONE = 0,
215 CONTIG_LEFT,
216 CONTIG_RIGHT
217};
218
219static int ocfs2_block_extent_contig(struct super_block *sb,
220 struct ocfs2_extent_rec *ext,
221 u64 blkno)
Mark Fashehccd979b2005-12-15 14:31:24 -0800222{
223 return blkno == (le64_to_cpu(ext->e_blkno) +
Mark Fashehdcd05382007-01-16 11:32:23 -0800224 ocfs2_clusters_to_blocks(sb,
Mark Fashehccd979b2005-12-15 14:31:24 -0800225 le32_to_cpu(ext->e_clusters)));
226}
227
Mark Fashehdcd05382007-01-16 11:32:23 -0800228static int ocfs2_extents_adjacent(struct ocfs2_extent_rec *left,
229 struct ocfs2_extent_rec *right)
230{
231 return (le32_to_cpu(left->e_cpos) + le32_to_cpu(left->e_clusters) ==
232 le32_to_cpu(right->e_cpos));
233}
234
235static enum ocfs2_contig_type
236 ocfs2_extent_contig(struct inode *inode,
237 struct ocfs2_extent_rec *ext,
238 struct ocfs2_extent_rec *insert_rec)
239{
240 u64 blkno = le64_to_cpu(insert_rec->e_blkno);
241
242 if (ocfs2_extents_adjacent(ext, insert_rec) &&
243 ocfs2_block_extent_contig(inode->i_sb, ext, blkno))
244 return CONTIG_RIGHT;
245
246 blkno = le64_to_cpu(ext->e_blkno);
247 if (ocfs2_extents_adjacent(insert_rec, ext) &&
248 ocfs2_block_extent_contig(inode->i_sb, insert_rec, blkno))
249 return CONTIG_LEFT;
250
251 return CONTIG_NONE;
252}
253
254/*
255 * NOTE: We can have pretty much any combination of contiguousness and
256 * appending.
257 *
258 * The usefulness of APPEND_TAIL is more in that it lets us know that
259 * we'll have to update the path to that leaf.
260 */
261enum ocfs2_append_type {
262 APPEND_NONE = 0,
263 APPEND_TAIL,
264};
265
266struct ocfs2_insert_type {
267 enum ocfs2_append_type ins_appending;
268 enum ocfs2_contig_type ins_contig;
269 int ins_contig_index;
270 int ins_free_records;
271 int ins_tree_depth;
272};
273
Mark Fashehccd979b2005-12-15 14:31:24 -0800274/*
275 * How many free extents have we got before we need more meta data?
276 */
277int ocfs2_num_free_extents(struct ocfs2_super *osb,
278 struct inode *inode,
279 struct ocfs2_dinode *fe)
280{
281 int retval;
282 struct ocfs2_extent_list *el;
283 struct ocfs2_extent_block *eb;
284 struct buffer_head *eb_bh = NULL;
285
286 mlog_entry_void();
287
288 if (!OCFS2_IS_VALID_DINODE(fe)) {
289 OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
290 retval = -EIO;
291 goto bail;
292 }
293
294 if (fe->i_last_eb_blk) {
295 retval = ocfs2_read_block(osb, le64_to_cpu(fe->i_last_eb_blk),
296 &eb_bh, OCFS2_BH_CACHED, inode);
297 if (retval < 0) {
298 mlog_errno(retval);
299 goto bail;
300 }
301 eb = (struct ocfs2_extent_block *) eb_bh->b_data;
302 el = &eb->h_list;
303 } else
304 el = &fe->id2.i_list;
305
306 BUG_ON(el->l_tree_depth != 0);
307
308 retval = le16_to_cpu(el->l_count) - le16_to_cpu(el->l_next_free_rec);
309bail:
310 if (eb_bh)
311 brelse(eb_bh);
312
313 mlog_exit(retval);
314 return retval;
315}
316
317/* expects array to already be allocated
318 *
319 * sets h_signature, h_blkno, h_suballoc_bit, h_suballoc_slot, and
320 * l_count for you
321 */
322static int ocfs2_create_new_meta_bhs(struct ocfs2_super *osb,
Mark Fasheh1fabe142006-10-09 18:11:45 -0700323 handle_t *handle,
Mark Fashehccd979b2005-12-15 14:31:24 -0800324 struct inode *inode,
325 int wanted,
326 struct ocfs2_alloc_context *meta_ac,
327 struct buffer_head *bhs[])
328{
329 int count, status, i;
330 u16 suballoc_bit_start;
331 u32 num_got;
332 u64 first_blkno;
333 struct ocfs2_extent_block *eb;
334
335 mlog_entry_void();
336
337 count = 0;
338 while (count < wanted) {
339 status = ocfs2_claim_metadata(osb,
340 handle,
341 meta_ac,
342 wanted - count,
343 &suballoc_bit_start,
344 &num_got,
345 &first_blkno);
346 if (status < 0) {
347 mlog_errno(status);
348 goto bail;
349 }
350
351 for(i = count; i < (num_got + count); i++) {
352 bhs[i] = sb_getblk(osb->sb, first_blkno);
353 if (bhs[i] == NULL) {
354 status = -EIO;
355 mlog_errno(status);
356 goto bail;
357 }
358 ocfs2_set_new_buffer_uptodate(inode, bhs[i]);
359
360 status = ocfs2_journal_access(handle, inode, bhs[i],
361 OCFS2_JOURNAL_ACCESS_CREATE);
362 if (status < 0) {
363 mlog_errno(status);
364 goto bail;
365 }
366
367 memset(bhs[i]->b_data, 0, osb->sb->s_blocksize);
368 eb = (struct ocfs2_extent_block *) bhs[i]->b_data;
369 /* Ok, setup the minimal stuff here. */
370 strcpy(eb->h_signature, OCFS2_EXTENT_BLOCK_SIGNATURE);
371 eb->h_blkno = cpu_to_le64(first_blkno);
372 eb->h_fs_generation = cpu_to_le32(osb->fs_generation);
373
374#ifndef OCFS2_USE_ALL_METADATA_SUBALLOCATORS
375 /* we always use slot zero's suballocator */
376 eb->h_suballoc_slot = 0;
377#else
378 eb->h_suballoc_slot = cpu_to_le16(osb->slot_num);
379#endif
380 eb->h_suballoc_bit = cpu_to_le16(suballoc_bit_start);
381 eb->h_list.l_count =
382 cpu_to_le16(ocfs2_extent_recs_per_eb(osb->sb));
383
384 suballoc_bit_start++;
385 first_blkno++;
386
387 /* We'll also be dirtied by the caller, so
388 * this isn't absolutely necessary. */
389 status = ocfs2_journal_dirty(handle, bhs[i]);
390 if (status < 0) {
391 mlog_errno(status);
392 goto bail;
393 }
394 }
395
396 count += num_got;
397 }
398
399 status = 0;
400bail:
401 if (status < 0) {
402 for(i = 0; i < wanted; i++) {
403 if (bhs[i])
404 brelse(bhs[i]);
405 bhs[i] = NULL;
406 }
407 }
408 mlog_exit(status);
409 return status;
410}
411
412/*
Mark Fashehdcd05382007-01-16 11:32:23 -0800413 * Helper function for ocfs2_add_branch() and ocfs2_shift_tree_depth().
414 *
415 * Returns the sum of the rightmost extent rec logical offset and
416 * cluster count.
417 *
418 * ocfs2_add_branch() uses this to determine what logical cluster
419 * value should be populated into the leftmost new branch records.
420 *
421 * ocfs2_shift_tree_depth() uses this to determine the # clusters
422 * value for the new topmost tree record.
423 */
424static inline u32 ocfs2_sum_rightmost_rec(struct ocfs2_extent_list *el)
425{
426 int i;
427
428 i = le16_to_cpu(el->l_next_free_rec) - 1;
429
430 return le32_to_cpu(el->l_recs[i].e_cpos) +
431 le32_to_cpu(el->l_recs[i].e_clusters);
432}
433
434/*
Mark Fashehccd979b2005-12-15 14:31:24 -0800435 * Add an entire tree branch to our inode. eb_bh is the extent block
436 * to start at, if we don't want to start the branch at the dinode
437 * structure.
438 *
439 * last_eb_bh is required as we have to update it's next_leaf pointer
440 * for the new last extent block.
441 *
442 * the new branch will be 'empty' in the sense that every block will
443 * contain a single record with e_clusters == 0.
444 */
445static int ocfs2_add_branch(struct ocfs2_super *osb,
Mark Fasheh1fabe142006-10-09 18:11:45 -0700446 handle_t *handle,
Mark Fashehccd979b2005-12-15 14:31:24 -0800447 struct inode *inode,
448 struct buffer_head *fe_bh,
449 struct buffer_head *eb_bh,
450 struct buffer_head *last_eb_bh,
451 struct ocfs2_alloc_context *meta_ac)
452{
453 int status, new_blocks, i;
454 u64 next_blkno, new_last_eb_blk;
455 struct buffer_head *bh;
456 struct buffer_head **new_eb_bhs = NULL;
457 struct ocfs2_dinode *fe;
458 struct ocfs2_extent_block *eb;
459 struct ocfs2_extent_list *eb_el;
460 struct ocfs2_extent_list *el;
Mark Fashehdcd05382007-01-16 11:32:23 -0800461 u32 new_cpos;
Mark Fashehccd979b2005-12-15 14:31:24 -0800462
463 mlog_entry_void();
464
465 BUG_ON(!last_eb_bh);
466
467 fe = (struct ocfs2_dinode *) fe_bh->b_data;
468
469 if (eb_bh) {
470 eb = (struct ocfs2_extent_block *) eb_bh->b_data;
471 el = &eb->h_list;
472 } else
473 el = &fe->id2.i_list;
474
475 /* we never add a branch to a leaf. */
476 BUG_ON(!el->l_tree_depth);
477
478 new_blocks = le16_to_cpu(el->l_tree_depth);
479
480 /* allocate the number of new eb blocks we need */
481 new_eb_bhs = kcalloc(new_blocks, sizeof(struct buffer_head *),
482 GFP_KERNEL);
483 if (!new_eb_bhs) {
484 status = -ENOMEM;
485 mlog_errno(status);
486 goto bail;
487 }
488
489 status = ocfs2_create_new_meta_bhs(osb, handle, inode, new_blocks,
490 meta_ac, new_eb_bhs);
491 if (status < 0) {
492 mlog_errno(status);
493 goto bail;
494 }
495
Mark Fashehdcd05382007-01-16 11:32:23 -0800496 eb = (struct ocfs2_extent_block *)last_eb_bh->b_data;
497 new_cpos = ocfs2_sum_rightmost_rec(&eb->h_list);
498
Mark Fashehccd979b2005-12-15 14:31:24 -0800499 /* Note: new_eb_bhs[new_blocks - 1] is the guy which will be
500 * linked with the rest of the tree.
501 * conversly, new_eb_bhs[0] is the new bottommost leaf.
502 *
503 * when we leave the loop, new_last_eb_blk will point to the
504 * newest leaf, and next_blkno will point to the topmost extent
505 * block. */
506 next_blkno = new_last_eb_blk = 0;
507 for(i = 0; i < new_blocks; i++) {
508 bh = new_eb_bhs[i];
509 eb = (struct ocfs2_extent_block *) bh->b_data;
510 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
511 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
512 status = -EIO;
513 goto bail;
514 }
515 eb_el = &eb->h_list;
516
517 status = ocfs2_journal_access(handle, inode, bh,
518 OCFS2_JOURNAL_ACCESS_CREATE);
519 if (status < 0) {
520 mlog_errno(status);
521 goto bail;
522 }
523
524 eb->h_next_leaf_blk = 0;
525 eb_el->l_tree_depth = cpu_to_le16(i);
526 eb_el->l_next_free_rec = cpu_to_le16(1);
Mark Fashehdcd05382007-01-16 11:32:23 -0800527 /*
528 * This actually counts as an empty extent as
529 * c_clusters == 0
530 */
531 eb_el->l_recs[0].e_cpos = cpu_to_le32(new_cpos);
Mark Fashehccd979b2005-12-15 14:31:24 -0800532 eb_el->l_recs[0].e_blkno = cpu_to_le64(next_blkno);
533 eb_el->l_recs[0].e_clusters = cpu_to_le32(0);
534 if (!eb_el->l_tree_depth)
535 new_last_eb_blk = le64_to_cpu(eb->h_blkno);
536
537 status = ocfs2_journal_dirty(handle, bh);
538 if (status < 0) {
539 mlog_errno(status);
540 goto bail;
541 }
542
543 next_blkno = le64_to_cpu(eb->h_blkno);
544 }
545
546 /* This is a bit hairy. We want to update up to three blocks
547 * here without leaving any of them in an inconsistent state
548 * in case of error. We don't have to worry about
549 * journal_dirty erroring as it won't unless we've aborted the
550 * handle (in which case we would never be here) so reserving
551 * the write with journal_access is all we need to do. */
552 status = ocfs2_journal_access(handle, inode, last_eb_bh,
553 OCFS2_JOURNAL_ACCESS_WRITE);
554 if (status < 0) {
555 mlog_errno(status);
556 goto bail;
557 }
558 status = ocfs2_journal_access(handle, inode, fe_bh,
559 OCFS2_JOURNAL_ACCESS_WRITE);
560 if (status < 0) {
561 mlog_errno(status);
562 goto bail;
563 }
564 if (eb_bh) {
565 status = ocfs2_journal_access(handle, inode, eb_bh,
566 OCFS2_JOURNAL_ACCESS_WRITE);
567 if (status < 0) {
568 mlog_errno(status);
569 goto bail;
570 }
571 }
572
573 /* Link the new branch into the rest of the tree (el will
574 * either be on the fe, or the extent block passed in. */
575 i = le16_to_cpu(el->l_next_free_rec);
576 el->l_recs[i].e_blkno = cpu_to_le64(next_blkno);
Mark Fashehdcd05382007-01-16 11:32:23 -0800577 el->l_recs[i].e_cpos = cpu_to_le32(new_cpos);
Mark Fashehccd979b2005-12-15 14:31:24 -0800578 el->l_recs[i].e_clusters = 0;
579 le16_add_cpu(&el->l_next_free_rec, 1);
580
581 /* fe needs a new last extent block pointer, as does the
582 * next_leaf on the previously last-extent-block. */
583 fe->i_last_eb_blk = cpu_to_le64(new_last_eb_blk);
584
585 eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
586 eb->h_next_leaf_blk = cpu_to_le64(new_last_eb_blk);
587
588 status = ocfs2_journal_dirty(handle, last_eb_bh);
589 if (status < 0)
590 mlog_errno(status);
591 status = ocfs2_journal_dirty(handle, fe_bh);
592 if (status < 0)
593 mlog_errno(status);
594 if (eb_bh) {
595 status = ocfs2_journal_dirty(handle, eb_bh);
596 if (status < 0)
597 mlog_errno(status);
598 }
599
600 status = 0;
601bail:
602 if (new_eb_bhs) {
603 for (i = 0; i < new_blocks; i++)
604 if (new_eb_bhs[i])
605 brelse(new_eb_bhs[i]);
606 kfree(new_eb_bhs);
607 }
608
609 mlog_exit(status);
610 return status;
611}
612
613/*
614 * adds another level to the allocation tree.
615 * returns back the new extent block so you can add a branch to it
616 * after this call.
617 */
618static int ocfs2_shift_tree_depth(struct ocfs2_super *osb,
Mark Fasheh1fabe142006-10-09 18:11:45 -0700619 handle_t *handle,
Mark Fashehccd979b2005-12-15 14:31:24 -0800620 struct inode *inode,
621 struct buffer_head *fe_bh,
622 struct ocfs2_alloc_context *meta_ac,
623 struct buffer_head **ret_new_eb_bh)
624{
625 int status, i;
Mark Fashehdcd05382007-01-16 11:32:23 -0800626 u32 new_clusters;
Mark Fashehccd979b2005-12-15 14:31:24 -0800627 struct buffer_head *new_eb_bh = NULL;
628 struct ocfs2_dinode *fe;
629 struct ocfs2_extent_block *eb;
630 struct ocfs2_extent_list *fe_el;
631 struct ocfs2_extent_list *eb_el;
632
633 mlog_entry_void();
634
635 status = ocfs2_create_new_meta_bhs(osb, handle, inode, 1, meta_ac,
636 &new_eb_bh);
637 if (status < 0) {
638 mlog_errno(status);
639 goto bail;
640 }
641
642 eb = (struct ocfs2_extent_block *) new_eb_bh->b_data;
643 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
644 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
645 status = -EIO;
646 goto bail;
647 }
648
649 eb_el = &eb->h_list;
650 fe = (struct ocfs2_dinode *) fe_bh->b_data;
651 fe_el = &fe->id2.i_list;
652
653 status = ocfs2_journal_access(handle, inode, new_eb_bh,
654 OCFS2_JOURNAL_ACCESS_CREATE);
655 if (status < 0) {
656 mlog_errno(status);
657 goto bail;
658 }
659
660 /* copy the fe data into the new extent block */
661 eb_el->l_tree_depth = fe_el->l_tree_depth;
662 eb_el->l_next_free_rec = fe_el->l_next_free_rec;
663 for(i = 0; i < le16_to_cpu(fe_el->l_next_free_rec); i++) {
664 eb_el->l_recs[i].e_cpos = fe_el->l_recs[i].e_cpos;
665 eb_el->l_recs[i].e_clusters = fe_el->l_recs[i].e_clusters;
666 eb_el->l_recs[i].e_blkno = fe_el->l_recs[i].e_blkno;
667 }
668
669 status = ocfs2_journal_dirty(handle, new_eb_bh);
670 if (status < 0) {
671 mlog_errno(status);
672 goto bail;
673 }
674
675 status = ocfs2_journal_access(handle, inode, fe_bh,
676 OCFS2_JOURNAL_ACCESS_WRITE);
677 if (status < 0) {
678 mlog_errno(status);
679 goto bail;
680 }
681
Mark Fashehdcd05382007-01-16 11:32:23 -0800682 new_clusters = ocfs2_sum_rightmost_rec(eb_el);
683
Mark Fashehccd979b2005-12-15 14:31:24 -0800684 /* update fe now */
685 le16_add_cpu(&fe_el->l_tree_depth, 1);
686 fe_el->l_recs[0].e_cpos = 0;
687 fe_el->l_recs[0].e_blkno = eb->h_blkno;
Mark Fashehdcd05382007-01-16 11:32:23 -0800688 fe_el->l_recs[0].e_clusters = cpu_to_le32(new_clusters);
Mark Fashehccd979b2005-12-15 14:31:24 -0800689 for(i = 1; i < le16_to_cpu(fe_el->l_next_free_rec); i++) {
690 fe_el->l_recs[i].e_cpos = 0;
691 fe_el->l_recs[i].e_clusters = 0;
692 fe_el->l_recs[i].e_blkno = 0;
693 }
694 fe_el->l_next_free_rec = cpu_to_le16(1);
695
696 /* If this is our 1st tree depth shift, then last_eb_blk
697 * becomes the allocated extent block */
698 if (fe_el->l_tree_depth == cpu_to_le16(1))
699 fe->i_last_eb_blk = eb->h_blkno;
700
701 status = ocfs2_journal_dirty(handle, fe_bh);
702 if (status < 0) {
703 mlog_errno(status);
704 goto bail;
705 }
706
707 *ret_new_eb_bh = new_eb_bh;
708 new_eb_bh = NULL;
709 status = 0;
710bail:
711 if (new_eb_bh)
712 brelse(new_eb_bh);
713
714 mlog_exit(status);
715 return status;
716}
717
718/*
Mark Fashehccd979b2005-12-15 14:31:24 -0800719 * Should only be called when there is no space left in any of the
720 * leaf nodes. What we want to do is find the lowest tree depth
721 * non-leaf extent block with room for new records. There are three
722 * valid results of this search:
723 *
724 * 1) a lowest extent block is found, then we pass it back in
725 * *lowest_eb_bh and return '0'
726 *
727 * 2) the search fails to find anything, but the dinode has room. We
728 * pass NULL back in *lowest_eb_bh, but still return '0'
729 *
730 * 3) the search fails to find anything AND the dinode is full, in
731 * which case we return > 0
732 *
733 * return status < 0 indicates an error.
734 */
735static int ocfs2_find_branch_target(struct ocfs2_super *osb,
736 struct inode *inode,
737 struct buffer_head *fe_bh,
738 struct buffer_head **target_bh)
739{
740 int status = 0, i;
741 u64 blkno;
742 struct ocfs2_dinode *fe;
743 struct ocfs2_extent_block *eb;
744 struct ocfs2_extent_list *el;
745 struct buffer_head *bh = NULL;
746 struct buffer_head *lowest_bh = NULL;
747
748 mlog_entry_void();
749
750 *target_bh = NULL;
751
752 fe = (struct ocfs2_dinode *) fe_bh->b_data;
753 el = &fe->id2.i_list;
754
755 while(le16_to_cpu(el->l_tree_depth) > 1) {
756 if (le16_to_cpu(el->l_next_free_rec) == 0) {
Mark Fashehb0697052006-03-03 10:24:33 -0800757 ocfs2_error(inode->i_sb, "Dinode %llu has empty "
Mark Fashehccd979b2005-12-15 14:31:24 -0800758 "extent list (next_free_rec == 0)",
Mark Fashehb0697052006-03-03 10:24:33 -0800759 (unsigned long long)OCFS2_I(inode)->ip_blkno);
Mark Fashehccd979b2005-12-15 14:31:24 -0800760 status = -EIO;
761 goto bail;
762 }
763 i = le16_to_cpu(el->l_next_free_rec) - 1;
764 blkno = le64_to_cpu(el->l_recs[i].e_blkno);
765 if (!blkno) {
Mark Fashehb0697052006-03-03 10:24:33 -0800766 ocfs2_error(inode->i_sb, "Dinode %llu has extent "
Mark Fashehccd979b2005-12-15 14:31:24 -0800767 "list where extent # %d has no physical "
768 "block start",
Mark Fashehb0697052006-03-03 10:24:33 -0800769 (unsigned long long)OCFS2_I(inode)->ip_blkno, i);
Mark Fashehccd979b2005-12-15 14:31:24 -0800770 status = -EIO;
771 goto bail;
772 }
773
774 if (bh) {
775 brelse(bh);
776 bh = NULL;
777 }
778
779 status = ocfs2_read_block(osb, blkno, &bh, OCFS2_BH_CACHED,
780 inode);
781 if (status < 0) {
782 mlog_errno(status);
783 goto bail;
784 }
785
786 eb = (struct ocfs2_extent_block *) bh->b_data;
787 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
788 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
789 status = -EIO;
790 goto bail;
791 }
792 el = &eb->h_list;
793
794 if (le16_to_cpu(el->l_next_free_rec) <
795 le16_to_cpu(el->l_count)) {
796 if (lowest_bh)
797 brelse(lowest_bh);
798 lowest_bh = bh;
799 get_bh(lowest_bh);
800 }
801 }
802
803 /* If we didn't find one and the fe doesn't have any room,
804 * then return '1' */
805 if (!lowest_bh
806 && (fe->id2.i_list.l_next_free_rec == fe->id2.i_list.l_count))
807 status = 1;
808
809 *target_bh = lowest_bh;
810bail:
811 if (bh)
812 brelse(bh);
813
814 mlog_exit(status);
815 return status;
816}
817
Mark Fashehdcd05382007-01-16 11:32:23 -0800818static inline int ocfs2_is_empty_extent(struct ocfs2_extent_rec *rec)
819{
820 return !rec->e_clusters;
821}
822
823/*
824 * This function will discard the rightmost extent record.
825 */
826static void ocfs2_shift_records_right(struct ocfs2_extent_list *el)
827{
828 int next_free = le16_to_cpu(el->l_next_free_rec);
829 int count = le16_to_cpu(el->l_count);
830 unsigned int num_bytes;
831
832 BUG_ON(!next_free);
833 /* This will cause us to go off the end of our extent list. */
834 BUG_ON(next_free >= count);
835
836 num_bytes = sizeof(struct ocfs2_extent_rec) * next_free;
837
838 memmove(&el->l_recs[1], &el->l_recs[0], num_bytes);
839}
840
841static void ocfs2_rotate_leaf(struct ocfs2_extent_list *el,
842 struct ocfs2_extent_rec *insert_rec)
843{
844 int i, insert_index, next_free, has_empty, num_bytes;
845 u32 insert_cpos = le32_to_cpu(insert_rec->e_cpos);
846 struct ocfs2_extent_rec *rec;
847
848 next_free = le16_to_cpu(el->l_next_free_rec);
849 has_empty = ocfs2_is_empty_extent(&el->l_recs[0]);
850
851 BUG_ON(!next_free);
852
853 /* The tree code before us didn't allow enough room in the leaf. */
854 if (el->l_next_free_rec == el->l_count && !has_empty)
855 BUG();
856
857 /*
858 * The easiest way to approach this is to just remove the
859 * empty extent and temporarily decrement next_free.
860 */
861 if (has_empty) {
862 /*
863 * If next_free was 1 (only an empty extent), this
864 * loop won't execute, which is fine. We still want
865 * the decrement above to happen.
866 */
867 for(i = 0; i < (next_free - 1); i++)
868 el->l_recs[i] = el->l_recs[i+1];
869
870 next_free--;
871 }
872
873 /*
874 * Figure out what the new record index should be.
875 */
876 for(i = 0; i < next_free; i++) {
877 rec = &el->l_recs[i];
878
879 if (insert_cpos < le32_to_cpu(rec->e_cpos))
880 break;
881 }
882 insert_index = i;
883
884 mlog(0, "ins %u: index %d, has_empty %d, next_free %d, count %d\n",
885 insert_cpos, insert_index, has_empty, next_free, le16_to_cpu(el->l_count));
886
887 BUG_ON(insert_index < 0);
888 BUG_ON(insert_index >= le16_to_cpu(el->l_count));
889 BUG_ON(insert_index > next_free);
890
891 /*
892 * No need to memmove if we're just adding to the tail.
893 */
894 if (insert_index != next_free) {
895 BUG_ON(next_free >= le16_to_cpu(el->l_count));
896
897 num_bytes = next_free - insert_index;
898 num_bytes *= sizeof(struct ocfs2_extent_rec);
899 memmove(&el->l_recs[insert_index + 1],
900 &el->l_recs[insert_index],
901 num_bytes);
902 }
903
904 /*
905 * Either we had an empty extent, and need to re-increment or
906 * there was no empty extent on a non full rightmost leaf node,
907 * in which case we still need to increment.
908 */
909 next_free++;
910 el->l_next_free_rec = cpu_to_le16(next_free);
911 /*
912 * Make sure none of the math above just messed up our tree.
913 */
914 BUG_ON(le16_to_cpu(el->l_next_free_rec) > le16_to_cpu(el->l_count));
915
916 el->l_recs[insert_index] = *insert_rec;
917
918}
919
920/*
921 * Create an empty extent record .
922 *
923 * l_next_free_rec may be updated.
924 *
925 * If an empty extent already exists do nothing.
926 */
927static void ocfs2_create_empty_extent(struct ocfs2_extent_list *el)
928{
929 int next_free = le16_to_cpu(el->l_next_free_rec);
930
931 if (next_free == 0)
932 goto set_and_inc;
933
934 if (ocfs2_is_empty_extent(&el->l_recs[0]))
935 return;
936
937 mlog_bug_on_msg(el->l_count == el->l_next_free_rec,
938 "Asked to create an empty extent in a full list:\n"
939 "count = %u, tree depth = %u",
940 le16_to_cpu(el->l_count),
941 le16_to_cpu(el->l_tree_depth));
942
943 ocfs2_shift_records_right(el);
944
945set_and_inc:
946 le16_add_cpu(&el->l_next_free_rec, 1);
947 memset(&el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec));
948}
949
950/*
951 * For a rotation which involves two leaf nodes, the "root node" is
952 * the lowest level tree node which contains a path to both leafs. This
953 * resulting set of information can be used to form a complete "subtree"
954 *
955 * This function is passed two full paths from the dinode down to a
956 * pair of adjacent leaves. It's task is to figure out which path
957 * index contains the subtree root - this can be the root index itself
958 * in a worst-case rotation.
959 *
960 * The array index of the subtree root is passed back.
961 */
962static int ocfs2_find_subtree_root(struct inode *inode,
963 struct ocfs2_path *left,
964 struct ocfs2_path *right)
965{
966 int i = 0;
967
968 /*
969 * Check that the caller passed in two paths from the same tree.
970 */
971 BUG_ON(path_root_bh(left) != path_root_bh(right));
972
973 do {
974 i++;
975
976 /*
977 * The caller didn't pass two adjacent paths.
978 */
979 mlog_bug_on_msg(i > left->p_tree_depth,
980 "Inode %lu, left depth %u, right depth %u\n"
981 "left leaf blk %llu, right leaf blk %llu\n",
982 inode->i_ino, left->p_tree_depth,
983 right->p_tree_depth,
984 (unsigned long long)path_leaf_bh(left)->b_blocknr,
985 (unsigned long long)path_leaf_bh(right)->b_blocknr);
986 } while (left->p_node[i].bh->b_blocknr ==
987 right->p_node[i].bh->b_blocknr);
988
989 return i - 1;
990}
991
992typedef void (path_insert_t)(void *, struct buffer_head *);
993
994/*
995 * Traverse a btree path in search of cpos, starting at root_el.
996 *
997 * This code can be called with a cpos larger than the tree, in which
998 * case it will return the rightmost path.
999 */
1000static int __ocfs2_find_path(struct inode *inode,
1001 struct ocfs2_extent_list *root_el, u32 cpos,
1002 path_insert_t *func, void *data)
1003{
1004 int i, ret = 0;
1005 u32 range;
1006 u64 blkno;
1007 struct buffer_head *bh = NULL;
1008 struct ocfs2_extent_block *eb;
1009 struct ocfs2_extent_list *el;
1010 struct ocfs2_extent_rec *rec;
1011 struct ocfs2_inode_info *oi = OCFS2_I(inode);
1012
1013 el = root_el;
1014 while (el->l_tree_depth) {
1015 if (le16_to_cpu(el->l_next_free_rec) == 0) {
1016 ocfs2_error(inode->i_sb,
1017 "Inode %llu has empty extent list at "
1018 "depth %u\n",
1019 (unsigned long long)oi->ip_blkno,
1020 le16_to_cpu(el->l_tree_depth));
1021 ret = -EROFS;
1022 goto out;
1023
1024 }
1025
1026 for(i = 0; i < le16_to_cpu(el->l_next_free_rec) - 1; i++) {
1027 rec = &el->l_recs[i];
1028
1029 /*
1030 * In the case that cpos is off the allocation
1031 * tree, this should just wind up returning the
1032 * rightmost record.
1033 */
1034 range = le32_to_cpu(rec->e_cpos) +
1035 le32_to_cpu(rec->e_clusters);
1036 if (cpos >= le32_to_cpu(rec->e_cpos) && cpos < range)
1037 break;
1038 }
1039
1040 blkno = le64_to_cpu(el->l_recs[i].e_blkno);
1041 if (blkno == 0) {
1042 ocfs2_error(inode->i_sb,
1043 "Inode %llu has bad blkno in extent list "
1044 "at depth %u (index %d)\n",
1045 (unsigned long long)oi->ip_blkno,
1046 le16_to_cpu(el->l_tree_depth), i);
1047 ret = -EROFS;
1048 goto out;
1049 }
1050
1051 brelse(bh);
1052 bh = NULL;
1053 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb), blkno,
1054 &bh, OCFS2_BH_CACHED, inode);
1055 if (ret) {
1056 mlog_errno(ret);
1057 goto out;
1058 }
1059
1060 eb = (struct ocfs2_extent_block *) bh->b_data;
1061 el = &eb->h_list;
1062 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
1063 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
1064 ret = -EIO;
1065 goto out;
1066 }
1067
1068 if (le16_to_cpu(el->l_next_free_rec) >
1069 le16_to_cpu(el->l_count)) {
1070 ocfs2_error(inode->i_sb,
1071 "Inode %llu has bad count in extent list "
1072 "at block %llu (next free=%u, count=%u)\n",
1073 (unsigned long long)oi->ip_blkno,
1074 (unsigned long long)bh->b_blocknr,
1075 le16_to_cpu(el->l_next_free_rec),
1076 le16_to_cpu(el->l_count));
1077 ret = -EROFS;
1078 goto out;
1079 }
1080
1081 if (func)
1082 func(data, bh);
1083 }
1084
1085out:
1086 /*
1087 * Catch any trailing bh that the loop didn't handle.
1088 */
1089 brelse(bh);
1090
1091 return ret;
1092}
1093
1094/*
1095 * Given an initialized path (that is, it has a valid root extent
1096 * list), this function will traverse the btree in search of the path
1097 * which would contain cpos.
1098 *
1099 * The path traveled is recorded in the path structure.
1100 *
1101 * Note that this will not do any comparisons on leaf node extent
1102 * records, so it will work fine in the case that we just added a tree
1103 * branch.
1104 */
1105struct find_path_data {
1106 int index;
1107 struct ocfs2_path *path;
1108};
1109static void find_path_ins(void *data, struct buffer_head *bh)
1110{
1111 struct find_path_data *fp = data;
1112
1113 get_bh(bh);
1114 ocfs2_path_insert_eb(fp->path, fp->index, bh);
1115 fp->index++;
1116}
1117static int ocfs2_find_path(struct inode *inode, struct ocfs2_path *path,
1118 u32 cpos)
1119{
1120 struct find_path_data data;
1121
1122 data.index = 1;
1123 data.path = path;
1124 return __ocfs2_find_path(inode, path_root_el(path), cpos,
1125 find_path_ins, &data);
1126}
1127
1128static void find_leaf_ins(void *data, struct buffer_head *bh)
1129{
1130 struct ocfs2_extent_block *eb =(struct ocfs2_extent_block *)bh->b_data;
1131 struct ocfs2_extent_list *el = &eb->h_list;
1132 struct buffer_head **ret = data;
1133
1134 /* We want to retain only the leaf block. */
1135 if (le16_to_cpu(el->l_tree_depth) == 0) {
1136 get_bh(bh);
1137 *ret = bh;
1138 }
1139}
1140/*
1141 * Find the leaf block in the tree which would contain cpos. No
1142 * checking of the actual leaf is done.
1143 *
1144 * Some paths want to call this instead of allocating a path structure
1145 * and calling ocfs2_find_path().
1146 *
1147 * This function doesn't handle non btree extent lists.
1148 */
Mark Fasheh363041a2007-01-17 12:31:35 -08001149int ocfs2_find_leaf(struct inode *inode, struct ocfs2_extent_list *root_el,
1150 u32 cpos, struct buffer_head **leaf_bh)
Mark Fashehdcd05382007-01-16 11:32:23 -08001151{
1152 int ret;
1153 struct buffer_head *bh = NULL;
1154
1155 ret = __ocfs2_find_path(inode, root_el, cpos, find_leaf_ins, &bh);
1156 if (ret) {
1157 mlog_errno(ret);
1158 goto out;
1159 }
1160
1161 *leaf_bh = bh;
1162out:
1163 return ret;
1164}
1165
1166/*
1167 * Adjust the adjacent records (left_rec, right_rec) involved in a rotation.
1168 *
1169 * Basically, we've moved stuff around at the bottom of the tree and
1170 * we need to fix up the extent records above the changes to reflect
1171 * the new changes.
1172 *
1173 * left_rec: the record on the left.
1174 * left_child_el: is the child list pointed to by left_rec
1175 * right_rec: the record to the right of left_rec
1176 * right_child_el: is the child list pointed to by right_rec
1177 *
1178 * By definition, this only works on interior nodes.
1179 */
1180static void ocfs2_adjust_adjacent_records(struct ocfs2_extent_rec *left_rec,
1181 struct ocfs2_extent_list *left_child_el,
1182 struct ocfs2_extent_rec *right_rec,
1183 struct ocfs2_extent_list *right_child_el)
1184{
1185 u32 left_clusters, right_end;
1186
1187 /*
1188 * Interior nodes never have holes. Their cpos is the cpos of
1189 * the leftmost record in their child list. Their cluster
1190 * count covers the full theoretical range of their child list
1191 * - the range between their cpos and the cpos of the record
1192 * immediately to their right.
1193 */
1194 left_clusters = le32_to_cpu(right_child_el->l_recs[0].e_cpos);
1195 left_clusters -= le32_to_cpu(left_rec->e_cpos);
1196 left_rec->e_clusters = cpu_to_le32(left_clusters);
1197
1198 /*
1199 * Calculate the rightmost cluster count boundary before
1200 * moving cpos - we will need to adjust e_clusters after
1201 * updating e_cpos to keep the same highest cluster count.
1202 */
1203 right_end = le32_to_cpu(right_rec->e_cpos);
1204 right_end += le32_to_cpu(right_rec->e_clusters);
1205
1206 right_rec->e_cpos = left_rec->e_cpos;
1207 le32_add_cpu(&right_rec->e_cpos, left_clusters);
1208
1209 right_end -= le32_to_cpu(right_rec->e_cpos);
1210 right_rec->e_clusters = cpu_to_le32(right_end);
1211}
1212
1213/*
1214 * Adjust the adjacent root node records involved in a
1215 * rotation. left_el_blkno is passed in as a key so that we can easily
1216 * find it's index in the root list.
1217 */
1218static void ocfs2_adjust_root_records(struct ocfs2_extent_list *root_el,
1219 struct ocfs2_extent_list *left_el,
1220 struct ocfs2_extent_list *right_el,
1221 u64 left_el_blkno)
1222{
1223 int i;
1224
1225 BUG_ON(le16_to_cpu(root_el->l_tree_depth) <=
1226 le16_to_cpu(left_el->l_tree_depth));
1227
1228 for(i = 0; i < le16_to_cpu(root_el->l_next_free_rec) - 1; i++) {
1229 if (le64_to_cpu(root_el->l_recs[i].e_blkno) == left_el_blkno)
1230 break;
1231 }
1232
1233 /*
1234 * The path walking code should have never returned a root and
1235 * two paths which are not adjacent.
1236 */
1237 BUG_ON(i >= (le16_to_cpu(root_el->l_next_free_rec) - 1));
1238
1239 ocfs2_adjust_adjacent_records(&root_el->l_recs[i], left_el,
1240 &root_el->l_recs[i + 1], right_el);
1241}
1242
1243/*
1244 * We've changed a leaf block (in right_path) and need to reflect that
1245 * change back up the subtree.
1246 *
1247 * This happens in multiple places:
1248 * - When we've moved an extent record from the left path leaf to the right
1249 * path leaf to make room for an empty extent in the left path leaf.
1250 * - When our insert into the right path leaf is at the leftmost edge
1251 * and requires an update of the path immediately to it's left. This
1252 * can occur at the end of some types of rotation and appending inserts.
1253 */
1254static void ocfs2_complete_edge_insert(struct inode *inode, handle_t *handle,
1255 struct ocfs2_path *left_path,
1256 struct ocfs2_path *right_path,
1257 int subtree_index)
1258{
1259 int ret, i, idx;
1260 struct ocfs2_extent_list *el, *left_el, *right_el;
1261 struct ocfs2_extent_rec *left_rec, *right_rec;
1262 struct buffer_head *root_bh = left_path->p_node[subtree_index].bh;
1263
1264 /*
1265 * Update the counts and position values within all the
1266 * interior nodes to reflect the leaf rotation we just did.
1267 *
1268 * The root node is handled below the loop.
1269 *
1270 * We begin the loop with right_el and left_el pointing to the
1271 * leaf lists and work our way up.
1272 *
1273 * NOTE: within this loop, left_el and right_el always refer
1274 * to the *child* lists.
1275 */
1276 left_el = path_leaf_el(left_path);
1277 right_el = path_leaf_el(right_path);
1278 for(i = left_path->p_tree_depth - 1; i > subtree_index; i--) {
1279 mlog(0, "Adjust records at index %u\n", i);
1280
1281 /*
1282 * One nice property of knowing that all of these
1283 * nodes are below the root is that we only deal with
1284 * the leftmost right node record and the rightmost
1285 * left node record.
1286 */
1287 el = left_path->p_node[i].el;
1288 idx = le16_to_cpu(left_el->l_next_free_rec) - 1;
1289 left_rec = &el->l_recs[idx];
1290
1291 el = right_path->p_node[i].el;
1292 right_rec = &el->l_recs[0];
1293
1294 ocfs2_adjust_adjacent_records(left_rec, left_el, right_rec,
1295 right_el);
1296
1297 ret = ocfs2_journal_dirty(handle, left_path->p_node[i].bh);
1298 if (ret)
1299 mlog_errno(ret);
1300
1301 ret = ocfs2_journal_dirty(handle, right_path->p_node[i].bh);
1302 if (ret)
1303 mlog_errno(ret);
1304
1305 /*
1306 * Setup our list pointers now so that the current
1307 * parents become children in the next iteration.
1308 */
1309 left_el = left_path->p_node[i].el;
1310 right_el = right_path->p_node[i].el;
1311 }
1312
1313 /*
1314 * At the root node, adjust the two adjacent records which
1315 * begin our path to the leaves.
1316 */
1317
1318 el = left_path->p_node[subtree_index].el;
1319 left_el = left_path->p_node[subtree_index + 1].el;
1320 right_el = right_path->p_node[subtree_index + 1].el;
1321
1322 ocfs2_adjust_root_records(el, left_el, right_el,
1323 left_path->p_node[subtree_index + 1].bh->b_blocknr);
1324
1325 root_bh = left_path->p_node[subtree_index].bh;
1326
1327 ret = ocfs2_journal_dirty(handle, root_bh);
1328 if (ret)
1329 mlog_errno(ret);
1330}
1331
1332static int ocfs2_rotate_subtree_right(struct inode *inode,
1333 handle_t *handle,
1334 struct ocfs2_path *left_path,
1335 struct ocfs2_path *right_path,
1336 int subtree_index)
1337{
1338 int ret, i;
1339 struct buffer_head *right_leaf_bh;
1340 struct buffer_head *left_leaf_bh = NULL;
1341 struct buffer_head *root_bh;
1342 struct ocfs2_extent_list *right_el, *left_el;
1343 struct ocfs2_extent_rec move_rec;
1344
1345 left_leaf_bh = path_leaf_bh(left_path);
1346 left_el = path_leaf_el(left_path);
1347
1348 if (left_el->l_next_free_rec != left_el->l_count) {
1349 ocfs2_error(inode->i_sb,
1350 "Inode %llu has non-full interior leaf node %llu"
1351 "(next free = %u)",
1352 (unsigned long long)OCFS2_I(inode)->ip_blkno,
1353 (unsigned long long)left_leaf_bh->b_blocknr,
1354 le16_to_cpu(left_el->l_next_free_rec));
1355 return -EROFS;
1356 }
1357
1358 /*
1359 * This extent block may already have an empty record, so we
1360 * return early if so.
1361 */
1362 if (ocfs2_is_empty_extent(&left_el->l_recs[0]))
1363 return 0;
1364
1365 root_bh = left_path->p_node[subtree_index].bh;
1366 BUG_ON(root_bh != right_path->p_node[subtree_index].bh);
1367
1368 ret = ocfs2_journal_access(handle, inode, root_bh,
1369 OCFS2_JOURNAL_ACCESS_WRITE);
1370 if (ret) {
1371 mlog_errno(ret);
1372 goto out;
1373 }
1374
1375 for(i = subtree_index + 1; i < path_num_items(right_path); i++) {
1376 ret = ocfs2_journal_access(handle, inode,
1377 right_path->p_node[i].bh,
1378 OCFS2_JOURNAL_ACCESS_WRITE);
1379 if (ret) {
1380 mlog_errno(ret);
1381 goto out;
1382 }
1383
1384 ret = ocfs2_journal_access(handle, inode,
1385 left_path->p_node[i].bh,
1386 OCFS2_JOURNAL_ACCESS_WRITE);
1387 if (ret) {
1388 mlog_errno(ret);
1389 goto out;
1390 }
1391 }
1392
1393 right_leaf_bh = path_leaf_bh(right_path);
1394 right_el = path_leaf_el(right_path);
1395
1396 /* This is a code error, not a disk corruption. */
1397 mlog_bug_on_msg(!right_el->l_next_free_rec, "Inode %llu: Rotate fails "
1398 "because rightmost leaf block %llu is empty\n",
1399 (unsigned long long)OCFS2_I(inode)->ip_blkno,
1400 (unsigned long long)right_leaf_bh->b_blocknr);
1401
1402 ocfs2_create_empty_extent(right_el);
1403
1404 ret = ocfs2_journal_dirty(handle, right_leaf_bh);
1405 if (ret) {
1406 mlog_errno(ret);
1407 goto out;
1408 }
1409
1410 /* Do the copy now. */
1411 i = le16_to_cpu(left_el->l_next_free_rec) - 1;
1412 move_rec = left_el->l_recs[i];
1413 right_el->l_recs[0] = move_rec;
1414
1415 /*
1416 * Clear out the record we just copied and shift everything
1417 * over, leaving an empty extent in the left leaf.
1418 *
1419 * We temporarily subtract from next_free_rec so that the
1420 * shift will lose the tail record (which is now defunct).
1421 */
1422 le16_add_cpu(&left_el->l_next_free_rec, -1);
1423 ocfs2_shift_records_right(left_el);
1424 memset(&left_el->l_recs[0], 0, sizeof(struct ocfs2_extent_rec));
1425 le16_add_cpu(&left_el->l_next_free_rec, 1);
1426
1427 ret = ocfs2_journal_dirty(handle, left_leaf_bh);
1428 if (ret) {
1429 mlog_errno(ret);
1430 goto out;
1431 }
1432
1433 ocfs2_complete_edge_insert(inode, handle, left_path, right_path,
1434 subtree_index);
1435
1436out:
1437 return ret;
1438}
1439
1440/*
1441 * Given a full path, determine what cpos value would return us a path
1442 * containing the leaf immediately to the left of the current one.
1443 *
1444 * Will return zero if the path passed in is already the leftmost path.
1445 */
1446static int ocfs2_find_cpos_for_left_leaf(struct super_block *sb,
1447 struct ocfs2_path *path, u32 *cpos)
1448{
1449 int i, j, ret = 0;
1450 u64 blkno;
1451 struct ocfs2_extent_list *el;
1452
1453 *cpos = 0;
1454
1455 blkno = path_leaf_bh(path)->b_blocknr;
1456
1457 /* Start at the tree node just above the leaf and work our way up. */
1458 i = path->p_tree_depth - 1;
1459 while (i >= 0) {
1460 el = path->p_node[i].el;
1461
1462 /*
1463 * Find the extent record just before the one in our
1464 * path.
1465 */
1466 for(j = 0; j < le16_to_cpu(el->l_next_free_rec); j++) {
1467 if (le64_to_cpu(el->l_recs[j].e_blkno) == blkno) {
1468 if (j == 0) {
1469 if (i == 0) {
1470 /*
1471 * We've determined that the
1472 * path specified is already
1473 * the leftmost one - return a
1474 * cpos of zero.
1475 */
1476 goto out;
1477 }
1478 /*
1479 * The leftmost record points to our
1480 * leaf - we need to travel up the
1481 * tree one level.
1482 */
1483 goto next_node;
1484 }
1485
1486 *cpos = le32_to_cpu(el->l_recs[j - 1].e_cpos);
1487 *cpos = *cpos + le32_to_cpu(el->l_recs[j - 1].e_clusters) - 1;
1488 goto out;
1489 }
1490 }
1491
1492 /*
1493 * If we got here, we never found a valid node where
1494 * the tree indicated one should be.
1495 */
1496 ocfs2_error(sb,
1497 "Invalid extent tree at extent block %llu\n",
1498 (unsigned long long)blkno);
1499 ret = -EROFS;
1500 goto out;
1501
1502next_node:
1503 blkno = path->p_node[i].bh->b_blocknr;
1504 i--;
1505 }
1506
1507out:
1508 return ret;
1509}
1510
1511static int ocfs2_extend_rotate_transaction(handle_t *handle, int subtree_depth,
1512 struct ocfs2_path *path)
1513{
1514 int credits = (path->p_tree_depth - subtree_depth) * 2 + 1;
1515
1516 if (handle->h_buffer_credits < credits)
1517 return ocfs2_extend_trans(handle, credits);
1518
1519 return 0;
1520}
1521
1522/*
1523 * Trap the case where we're inserting into the theoretical range past
1524 * the _actual_ left leaf range. Otherwise, we'll rotate a record
1525 * whose cpos is less than ours into the right leaf.
1526 *
1527 * It's only necessary to look at the rightmost record of the left
1528 * leaf because the logic that calls us should ensure that the
1529 * theoretical ranges in the path components above the leaves are
1530 * correct.
1531 */
1532static int ocfs2_rotate_requires_path_adjustment(struct ocfs2_path *left_path,
1533 u32 insert_cpos)
1534{
1535 struct ocfs2_extent_list *left_el;
1536 struct ocfs2_extent_rec *rec;
1537 int next_free;
1538
1539 left_el = path_leaf_el(left_path);
1540 next_free = le16_to_cpu(left_el->l_next_free_rec);
1541 rec = &left_el->l_recs[next_free - 1];
1542
1543 if (insert_cpos > le32_to_cpu(rec->e_cpos))
1544 return 1;
1545 return 0;
1546}
1547
1548/*
1549 * Rotate all the records in a btree right one record, starting at insert_cpos.
1550 *
1551 * The path to the rightmost leaf should be passed in.
1552 *
1553 * The array is assumed to be large enough to hold an entire path (tree depth).
1554 *
1555 * Upon succesful return from this function:
1556 *
1557 * - The 'right_path' array will contain a path to the leaf block
1558 * whose range contains e_cpos.
1559 * - That leaf block will have a single empty extent in list index 0.
1560 * - In the case that the rotation requires a post-insert update,
1561 * *ret_left_path will contain a valid path which can be passed to
1562 * ocfs2_insert_path().
1563 */
1564static int ocfs2_rotate_tree_right(struct inode *inode,
1565 handle_t *handle,
1566 u32 insert_cpos,
1567 struct ocfs2_path *right_path,
1568 struct ocfs2_path **ret_left_path)
1569{
1570 int ret, start;
1571 u32 cpos;
1572 struct ocfs2_path *left_path = NULL;
1573
1574 *ret_left_path = NULL;
1575
1576 left_path = ocfs2_new_path(path_root_bh(right_path),
1577 path_root_el(right_path));
1578 if (!left_path) {
1579 ret = -ENOMEM;
1580 mlog_errno(ret);
1581 goto out;
1582 }
1583
1584 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path, &cpos);
1585 if (ret) {
1586 mlog_errno(ret);
1587 goto out;
1588 }
1589
1590 mlog(0, "Insert: %u, first left path cpos: %u\n", insert_cpos, cpos);
1591
1592 /*
1593 * What we want to do here is:
1594 *
1595 * 1) Start with the rightmost path.
1596 *
1597 * 2) Determine a path to the leaf block directly to the left
1598 * of that leaf.
1599 *
1600 * 3) Determine the 'subtree root' - the lowest level tree node
1601 * which contains a path to both leaves.
1602 *
1603 * 4) Rotate the subtree.
1604 *
1605 * 5) Find the next subtree by considering the left path to be
1606 * the new right path.
1607 *
1608 * The check at the top of this while loop also accepts
1609 * insert_cpos == cpos because cpos is only a _theoretical_
1610 * value to get us the left path - insert_cpos might very well
1611 * be filling that hole.
1612 *
1613 * Stop at a cpos of '0' because we either started at the
1614 * leftmost branch (i.e., a tree with one branch and a
1615 * rotation inside of it), or we've gone as far as we can in
1616 * rotating subtrees.
1617 */
1618 while (cpos && insert_cpos <= cpos) {
1619 mlog(0, "Rotating a tree: ins. cpos: %u, left path cpos: %u\n",
1620 insert_cpos, cpos);
1621
1622 ret = ocfs2_find_path(inode, left_path, cpos);
1623 if (ret) {
1624 mlog_errno(ret);
1625 goto out;
1626 }
1627
1628 mlog_bug_on_msg(path_leaf_bh(left_path) ==
1629 path_leaf_bh(right_path),
1630 "Inode %lu: error during insert of %u "
1631 "(left path cpos %u) results in two identical "
1632 "paths ending at %llu\n",
1633 inode->i_ino, insert_cpos, cpos,
1634 (unsigned long long)
1635 path_leaf_bh(left_path)->b_blocknr);
1636
1637 if (ocfs2_rotate_requires_path_adjustment(left_path,
1638 insert_cpos)) {
1639 mlog(0, "Path adjustment required\n");
1640
1641 /*
1642 * We've rotated the tree as much as we
1643 * should. The rest is up to
1644 * ocfs2_insert_path() to complete, after the
1645 * record insertion. We indicate this
1646 * situation by returning the left path.
1647 *
1648 * The reason we don't adjust the records here
1649 * before the record insert is that an error
1650 * later might break the rule where a parent
1651 * record e_cpos will reflect the actual
1652 * e_cpos of the 1st nonempty record of the
1653 * child list.
1654 */
1655 *ret_left_path = left_path;
1656 goto out_ret_path;
1657 }
1658
1659 start = ocfs2_find_subtree_root(inode, left_path, right_path);
1660
1661 mlog(0, "Subtree root at index %d (blk %llu, depth %d)\n",
1662 start,
1663 (unsigned long long) right_path->p_node[start].bh->b_blocknr,
1664 right_path->p_tree_depth);
1665
1666 ret = ocfs2_extend_rotate_transaction(handle, start,
1667 right_path);
1668 if (ret) {
1669 mlog_errno(ret);
1670 goto out;
1671 }
1672
1673 ret = ocfs2_rotate_subtree_right(inode, handle, left_path,
1674 right_path, start);
1675 if (ret) {
1676 mlog_errno(ret);
1677 goto out;
1678 }
1679
1680 /*
1681 * There is no need to re-read the next right path
1682 * as we know that it'll be our current left
1683 * path. Optimize by copying values instead.
1684 */
1685 ocfs2_mv_path(right_path, left_path);
1686
1687 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path,
1688 &cpos);
1689 if (ret) {
1690 mlog_errno(ret);
1691 goto out;
1692 }
1693 }
1694
1695out:
1696 ocfs2_free_path(left_path);
1697
1698out_ret_path:
1699 return ret;
1700}
1701
1702/*
1703 * Do the final bits of extent record insertion at the target leaf
1704 * list. If this leaf is part of an allocation tree, it is assumed
1705 * that the tree above has been prepared.
1706 */
1707static void ocfs2_insert_at_leaf(struct ocfs2_extent_rec *insert_rec,
1708 struct ocfs2_extent_list *el,
1709 struct ocfs2_insert_type *insert,
1710 struct inode *inode)
1711{
1712 int i = insert->ins_contig_index;
1713 unsigned int range;
1714 struct ocfs2_extent_rec *rec;
1715
1716 BUG_ON(el->l_tree_depth);
1717
1718 /*
1719 * Contiguous insert - either left or right.
1720 */
1721 if (insert->ins_contig != CONTIG_NONE) {
1722 rec = &el->l_recs[i];
1723 if (insert->ins_contig == CONTIG_LEFT) {
1724 rec->e_blkno = insert_rec->e_blkno;
1725 rec->e_cpos = insert_rec->e_cpos;
1726 }
1727 le32_add_cpu(&rec->e_clusters,
1728 le32_to_cpu(insert_rec->e_clusters));
1729 return;
1730 }
1731
1732 /*
1733 * Handle insert into an empty leaf.
1734 */
1735 if (le16_to_cpu(el->l_next_free_rec) == 0 ||
1736 ((le16_to_cpu(el->l_next_free_rec) == 1) &&
1737 ocfs2_is_empty_extent(&el->l_recs[0]))) {
1738 el->l_recs[0] = *insert_rec;
1739 el->l_next_free_rec = cpu_to_le16(1);
1740 return;
1741 }
1742
1743 /*
1744 * Appending insert.
1745 */
1746 if (insert->ins_appending == APPEND_TAIL) {
1747 i = le16_to_cpu(el->l_next_free_rec) - 1;
1748 rec = &el->l_recs[i];
1749 range = le32_to_cpu(rec->e_cpos) + le32_to_cpu(rec->e_clusters);
1750 BUG_ON(le32_to_cpu(insert_rec->e_cpos) < range);
1751
1752 mlog_bug_on_msg(le16_to_cpu(el->l_next_free_rec) >=
1753 le16_to_cpu(el->l_count),
1754 "inode %lu, depth %u, count %u, next free %u, "
1755 "rec.cpos %u, rec.clusters %u, "
1756 "insert.cpos %u, insert.clusters %u\n",
1757 inode->i_ino,
1758 le16_to_cpu(el->l_tree_depth),
1759 le16_to_cpu(el->l_count),
1760 le16_to_cpu(el->l_next_free_rec),
1761 le32_to_cpu(el->l_recs[i].e_cpos),
1762 le32_to_cpu(el->l_recs[i].e_clusters),
1763 le32_to_cpu(insert_rec->e_cpos),
1764 le32_to_cpu(insert_rec->e_clusters));
1765 i++;
1766 el->l_recs[i] = *insert_rec;
1767 le16_add_cpu(&el->l_next_free_rec, 1);
1768 return;
1769 }
1770
1771 /*
1772 * Ok, we have to rotate.
1773 *
1774 * At this point, it is safe to assume that inserting into an
1775 * empty leaf and appending to a leaf have both been handled
1776 * above.
1777 *
1778 * This leaf needs to have space, either by the empty 1st
1779 * extent record, or by virtue of an l_next_rec < l_count.
1780 */
1781 ocfs2_rotate_leaf(el, insert_rec);
1782}
1783
1784static inline void ocfs2_update_dinode_clusters(struct inode *inode,
1785 struct ocfs2_dinode *di,
1786 u32 clusters)
1787{
1788 le32_add_cpu(&di->i_clusters, clusters);
1789 spin_lock(&OCFS2_I(inode)->ip_lock);
1790 OCFS2_I(inode)->ip_clusters = le32_to_cpu(di->i_clusters);
1791 spin_unlock(&OCFS2_I(inode)->ip_lock);
1792}
1793
1794static int ocfs2_append_rec_to_path(struct inode *inode, handle_t *handle,
1795 struct ocfs2_extent_rec *insert_rec,
1796 struct ocfs2_path *right_path,
1797 struct ocfs2_path **ret_left_path)
1798{
1799 int ret, i, next_free;
1800 struct buffer_head *bh;
1801 struct ocfs2_extent_list *el;
1802 struct ocfs2_path *left_path = NULL;
1803
1804 *ret_left_path = NULL;
1805
1806 /*
1807 * If our appending insert is at the leftmost edge of a leaf,
1808 * then we might need to update the rightmost records of the
1809 * neighboring path.
1810 */
1811 el = path_leaf_el(right_path);
1812 next_free = le16_to_cpu(el->l_next_free_rec);
1813 if (next_free == 0 ||
1814 (next_free == 1 && ocfs2_is_empty_extent(&el->l_recs[0]))) {
1815 u32 left_cpos;
1816
1817 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, right_path,
1818 &left_cpos);
1819 if (ret) {
1820 mlog_errno(ret);
1821 goto out;
1822 }
1823
1824 mlog(0, "Append may need a left path update. cpos: %u, "
1825 "left_cpos: %u\n", le32_to_cpu(insert_rec->e_cpos),
1826 left_cpos);
1827
1828 /*
1829 * No need to worry if the append is already in the
1830 * leftmost leaf.
1831 */
1832 if (left_cpos) {
1833 left_path = ocfs2_new_path(path_root_bh(right_path),
1834 path_root_el(right_path));
1835 if (!left_path) {
1836 ret = -ENOMEM;
1837 mlog_errno(ret);
1838 goto out;
1839 }
1840
1841 ret = ocfs2_find_path(inode, left_path, left_cpos);
1842 if (ret) {
1843 mlog_errno(ret);
1844 goto out;
1845 }
1846
1847 /*
1848 * ocfs2_insert_path() will pass the left_path to the
1849 * journal for us.
1850 */
1851 }
1852 }
1853
1854 ret = ocfs2_journal_access_path(inode, handle, right_path);
1855 if (ret) {
1856 mlog_errno(ret);
1857 goto out;
1858 }
1859
1860 el = path_root_el(right_path);
1861 bh = path_root_bh(right_path);
1862 i = 0;
1863 while (1) {
1864 next_free = le16_to_cpu(el->l_next_free_rec);
1865 if (next_free == 0) {
1866 ocfs2_error(inode->i_sb,
1867 "Dinode %llu has a bad extent list",
1868 (unsigned long long)OCFS2_I(inode)->ip_blkno);
1869 ret = -EIO;
1870 goto out;
1871 }
1872
1873 el->l_recs[next_free - 1].e_clusters = insert_rec->e_cpos;
1874 le32_add_cpu(&el->l_recs[next_free - 1].e_clusters,
1875 le32_to_cpu(insert_rec->e_clusters));
1876 le32_add_cpu(&el->l_recs[next_free - 1].e_clusters,
1877 -le32_to_cpu(el->l_recs[next_free - 1].e_cpos));
1878
1879 ret = ocfs2_journal_dirty(handle, bh);
1880 if (ret)
1881 mlog_errno(ret);
1882
1883 if (++i >= right_path->p_tree_depth)
1884 break;
1885
1886 bh = right_path->p_node[i].bh;
1887 el = right_path->p_node[i].el;
1888 }
1889
1890 *ret_left_path = left_path;
1891 ret = 0;
1892out:
1893 if (ret != 0)
1894 ocfs2_free_path(left_path);
1895
1896 return ret;
1897}
1898
1899/*
1900 * This function only does inserts on an allocation b-tree. For dinode
1901 * lists, ocfs2_insert_at_leaf() is called directly.
1902 *
1903 * right_path is the path we want to do the actual insert
1904 * in. left_path should only be passed in if we need to update that
1905 * portion of the tree after an edge insert.
1906 */
1907static int ocfs2_insert_path(struct inode *inode,
1908 handle_t *handle,
1909 struct ocfs2_path *left_path,
1910 struct ocfs2_path *right_path,
1911 struct ocfs2_extent_rec *insert_rec,
1912 struct ocfs2_insert_type *insert)
1913{
1914 int ret, subtree_index;
1915 struct buffer_head *leaf_bh = path_leaf_bh(right_path);
1916 struct ocfs2_extent_list *el;
1917
1918 /*
1919 * Pass both paths to the journal. The majority of inserts
1920 * will be touching all components anyway.
1921 */
1922 ret = ocfs2_journal_access_path(inode, handle, right_path);
1923 if (ret < 0) {
1924 mlog_errno(ret);
1925 goto out;
1926 }
1927
1928 if (left_path) {
1929 int credits = handle->h_buffer_credits;
1930
1931 /*
1932 * There's a chance that left_path got passed back to
1933 * us without being accounted for in the
1934 * journal. Extend our transaction here to be sure we
1935 * can change those blocks.
1936 */
1937 credits += left_path->p_tree_depth;
1938
1939 ret = ocfs2_extend_trans(handle, credits);
1940 if (ret < 0) {
1941 mlog_errno(ret);
1942 goto out;
1943 }
1944
1945 ret = ocfs2_journal_access_path(inode, handle, left_path);
1946 if (ret < 0) {
1947 mlog_errno(ret);
1948 goto out;
1949 }
1950 }
1951
1952 el = path_leaf_el(right_path);
1953
1954 ocfs2_insert_at_leaf(insert_rec, el, insert, inode);
1955 ret = ocfs2_journal_dirty(handle, leaf_bh);
1956 if (ret)
1957 mlog_errno(ret);
1958
1959 if (left_path) {
1960 /*
1961 * The rotate code has indicated that we need to fix
1962 * up portions of the tree after the insert.
1963 *
1964 * XXX: Should we extend the transaction here?
1965 */
1966 subtree_index = ocfs2_find_subtree_root(inode, left_path,
1967 right_path);
1968 ocfs2_complete_edge_insert(inode, handle, left_path,
1969 right_path, subtree_index);
1970 }
1971
1972 ret = 0;
1973out:
1974 return ret;
1975}
1976
1977static int ocfs2_do_insert_extent(struct inode *inode,
1978 handle_t *handle,
1979 struct buffer_head *di_bh,
1980 struct ocfs2_extent_rec *insert_rec,
1981 struct ocfs2_insert_type *type)
1982{
1983 int ret, rotate = 0;
1984 u32 cpos;
1985 struct ocfs2_path *right_path = NULL;
1986 struct ocfs2_path *left_path = NULL;
1987 struct ocfs2_dinode *di;
1988 struct ocfs2_extent_list *el;
1989
1990 di = (struct ocfs2_dinode *) di_bh->b_data;
1991 el = &di->id2.i_list;
1992
1993 ret = ocfs2_journal_access(handle, inode, di_bh,
1994 OCFS2_JOURNAL_ACCESS_WRITE);
1995 if (ret) {
1996 mlog_errno(ret);
1997 goto out;
1998 }
1999
2000 if (le16_to_cpu(el->l_tree_depth) == 0) {
2001 ocfs2_insert_at_leaf(insert_rec, el, type, inode);
2002 goto out_update_clusters;
2003 }
2004
2005 right_path = ocfs2_new_inode_path(di_bh);
2006 if (!right_path) {
2007 ret = -ENOMEM;
2008 mlog_errno(ret);
2009 goto out;
2010 }
2011
2012 /*
2013 * Determine the path to start with. Rotations need the
2014 * rightmost path, everything else can go directly to the
2015 * target leaf.
2016 */
2017 cpos = le32_to_cpu(insert_rec->e_cpos);
2018 if (type->ins_appending == APPEND_NONE &&
2019 type->ins_contig == CONTIG_NONE) {
2020 rotate = 1;
2021 cpos = UINT_MAX;
2022 }
2023
2024 ret = ocfs2_find_path(inode, right_path, cpos);
2025 if (ret) {
2026 mlog_errno(ret);
2027 goto out;
2028 }
2029
2030 /*
2031 * Rotations and appends need special treatment - they modify
2032 * parts of the tree's above them.
2033 *
2034 * Both might pass back a path immediate to the left of the
2035 * one being inserted to. This will be cause
2036 * ocfs2_insert_path() to modify the rightmost records of
2037 * left_path to account for an edge insert.
2038 *
2039 * XXX: When modifying this code, keep in mind that an insert
2040 * can wind up skipping both of these two special cases...
2041 */
2042 if (rotate) {
2043 ret = ocfs2_rotate_tree_right(inode, handle,
2044 le32_to_cpu(insert_rec->e_cpos),
2045 right_path, &left_path);
2046 if (ret) {
2047 mlog_errno(ret);
2048 goto out;
2049 }
2050 } else if (type->ins_appending == APPEND_TAIL
2051 && type->ins_contig != CONTIG_LEFT) {
2052 ret = ocfs2_append_rec_to_path(inode, handle, insert_rec,
2053 right_path, &left_path);
2054 if (ret) {
2055 mlog_errno(ret);
2056 goto out;
2057 }
2058 }
2059
2060 ret = ocfs2_insert_path(inode, handle, left_path, right_path,
2061 insert_rec, type);
2062 if (ret) {
2063 mlog_errno(ret);
2064 goto out;
2065 }
2066
2067out_update_clusters:
2068 ocfs2_update_dinode_clusters(inode, di,
2069 le32_to_cpu(insert_rec->e_clusters));
2070
2071 ret = ocfs2_journal_dirty(handle, di_bh);
2072 if (ret)
2073 mlog_errno(ret);
2074
2075out:
2076 ocfs2_free_path(left_path);
2077 ocfs2_free_path(right_path);
2078
2079 return ret;
2080}
2081
2082static void ocfs2_figure_contig_type(struct inode *inode,
2083 struct ocfs2_insert_type *insert,
2084 struct ocfs2_extent_list *el,
2085 struct ocfs2_extent_rec *insert_rec)
2086{
2087 int i;
2088 enum ocfs2_contig_type contig_type = CONTIG_NONE;
2089
2090 for(i = 0; i < le16_to_cpu(el->l_next_free_rec); i++) {
2091 contig_type = ocfs2_extent_contig(inode, &el->l_recs[i],
2092 insert_rec);
2093 if (contig_type != CONTIG_NONE) {
2094 insert->ins_contig_index = i;
2095 break;
2096 }
2097 }
2098 insert->ins_contig = contig_type;
2099}
2100
2101/*
2102 * This should only be called against the righmost leaf extent list.
2103 *
2104 * ocfs2_figure_appending_type() will figure out whether we'll have to
2105 * insert at the tail of the rightmost leaf.
2106 *
2107 * This should also work against the dinode list for tree's with 0
2108 * depth. If we consider the dinode list to be the rightmost leaf node
2109 * then the logic here makes sense.
2110 */
2111static void ocfs2_figure_appending_type(struct ocfs2_insert_type *insert,
2112 struct ocfs2_extent_list *el,
2113 struct ocfs2_extent_rec *insert_rec)
2114{
2115 int i;
2116 u32 cpos = le32_to_cpu(insert_rec->e_cpos);
2117 struct ocfs2_extent_rec *rec;
2118
2119 insert->ins_appending = APPEND_NONE;
2120
2121 BUG_ON(el->l_tree_depth);
2122
2123 if (!el->l_next_free_rec)
2124 goto set_tail_append;
2125
2126 if (ocfs2_is_empty_extent(&el->l_recs[0])) {
2127 /* Were all records empty? */
2128 if (le16_to_cpu(el->l_next_free_rec) == 1)
2129 goto set_tail_append;
2130 }
2131
2132 i = le16_to_cpu(el->l_next_free_rec) - 1;
2133 rec = &el->l_recs[i];
2134
2135 if (cpos >= (le32_to_cpu(rec->e_cpos) + le32_to_cpu(rec->e_clusters)))
2136 goto set_tail_append;
2137
2138 return;
2139
2140set_tail_append:
2141 insert->ins_appending = APPEND_TAIL;
2142}
2143
2144/*
2145 * Helper function called at the begining of an insert.
2146 *
2147 * This computes a few things that are commonly used in the process of
2148 * inserting into the btree:
2149 * - Whether the new extent is contiguous with an existing one.
2150 * - The current tree depth.
2151 * - Whether the insert is an appending one.
2152 * - The total # of free records in the tree.
2153 *
2154 * All of the information is stored on the ocfs2_insert_type
2155 * structure.
2156 */
2157static int ocfs2_figure_insert_type(struct inode *inode,
2158 struct buffer_head *di_bh,
2159 struct buffer_head **last_eb_bh,
2160 struct ocfs2_extent_rec *insert_rec,
2161 struct ocfs2_insert_type *insert)
2162{
2163 int ret;
2164 struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
2165 struct ocfs2_extent_block *eb;
2166 struct ocfs2_extent_list *el;
2167 struct ocfs2_path *path = NULL;
2168 struct buffer_head *bh = NULL;
2169
2170 el = &di->id2.i_list;
2171 insert->ins_tree_depth = le16_to_cpu(el->l_tree_depth);
2172
2173 if (el->l_tree_depth) {
2174 /*
2175 * If we have tree depth, we read in the
2176 * rightmost extent block ahead of time as
2177 * ocfs2_figure_insert_type() and ocfs2_add_branch()
2178 * may want it later.
2179 */
2180 ret = ocfs2_read_block(OCFS2_SB(inode->i_sb),
2181 le64_to_cpu(di->i_last_eb_blk), &bh,
2182 OCFS2_BH_CACHED, inode);
2183 if (ret) {
2184 mlog_exit(ret);
2185 goto out;
2186 }
2187 eb = (struct ocfs2_extent_block *) bh->b_data;
2188 el = &eb->h_list;
2189 }
2190
2191 /*
2192 * Unless we have a contiguous insert, we'll need to know if
2193 * there is room left in our allocation tree for another
2194 * extent record.
2195 *
2196 * XXX: This test is simplistic, we can search for empty
2197 * extent records too.
2198 */
2199 insert->ins_free_records = le16_to_cpu(el->l_count) -
2200 le16_to_cpu(el->l_next_free_rec);
2201
2202 if (!insert->ins_tree_depth) {
2203 ocfs2_figure_contig_type(inode, insert, el, insert_rec);
2204 ocfs2_figure_appending_type(insert, el, insert_rec);
2205 return 0;
2206 }
2207
2208 path = ocfs2_new_inode_path(di_bh);
2209 if (!path) {
2210 ret = -ENOMEM;
2211 mlog_errno(ret);
2212 goto out;
2213 }
2214
2215 /*
2216 * In the case that we're inserting past what the tree
2217 * currently accounts for, ocfs2_find_path() will return for
2218 * us the rightmost tree path. This is accounted for below in
2219 * the appending code.
2220 */
2221 ret = ocfs2_find_path(inode, path, le32_to_cpu(insert_rec->e_cpos));
2222 if (ret) {
2223 mlog_errno(ret);
2224 goto out;
2225 }
2226
2227 el = path_leaf_el(path);
2228
2229 /*
2230 * Now that we have the path, there's two things we want to determine:
2231 * 1) Contiguousness (also set contig_index if this is so)
2232 *
2233 * 2) Are we doing an append? We can trivially break this up
2234 * into two types of appends: simple record append, or a
2235 * rotate inside the tail leaf.
2236 */
2237 ocfs2_figure_contig_type(inode, insert, el, insert_rec);
2238
2239 /*
2240 * The insert code isn't quite ready to deal with all cases of
2241 * left contiguousness. Specifically, if it's an insert into
2242 * the 1st record in a leaf, it will require the adjustment of
2243 * e_clusters on the last record of the path directly to it's
2244 * left. For now, just catch that case and fool the layers
2245 * above us. This works just fine for tree_depth == 0, which
2246 * is why we allow that above.
2247 */
2248 if (insert->ins_contig == CONTIG_LEFT &&
2249 insert->ins_contig_index == 0)
2250 insert->ins_contig = CONTIG_NONE;
2251
2252 /*
2253 * Ok, so we can simply compare against last_eb to figure out
2254 * whether the path doesn't exist. This will only happen in
2255 * the case that we're doing a tail append, so maybe we can
2256 * take advantage of that information somehow.
2257 */
2258 if (le64_to_cpu(di->i_last_eb_blk) == path_leaf_bh(path)->b_blocknr) {
2259 /*
2260 * Ok, ocfs2_find_path() returned us the rightmost
2261 * tree path. This might be an appending insert. There are
2262 * two cases:
2263 * 1) We're doing a true append at the tail:
2264 * -This might even be off the end of the leaf
2265 * 2) We're "appending" by rotating in the tail
2266 */
2267 ocfs2_figure_appending_type(insert, el, insert_rec);
2268 }
2269
2270out:
2271 ocfs2_free_path(path);
2272
2273 if (ret == 0)
2274 *last_eb_bh = bh;
2275 else
2276 brelse(bh);
2277 return ret;
2278}
2279
2280/*
2281 * Insert an extent into an inode btree.
2282 *
2283 * The caller needs to update fe->i_clusters
2284 */
Mark Fashehccd979b2005-12-15 14:31:24 -08002285int ocfs2_insert_extent(struct ocfs2_super *osb,
Mark Fasheh1fabe142006-10-09 18:11:45 -07002286 handle_t *handle,
Mark Fashehccd979b2005-12-15 14:31:24 -08002287 struct inode *inode,
2288 struct buffer_head *fe_bh,
Mark Fashehdcd05382007-01-16 11:32:23 -08002289 u32 cpos,
Mark Fashehccd979b2005-12-15 14:31:24 -08002290 u64 start_blk,
2291 u32 new_clusters,
2292 struct ocfs2_alloc_context *meta_ac)
2293{
Mark Fashehdcd05382007-01-16 11:32:23 -08002294 int status, shift;
Mark Fashehccd979b2005-12-15 14:31:24 -08002295 struct buffer_head *last_eb_bh = NULL;
2296 struct buffer_head *bh = NULL;
Mark Fashehdcd05382007-01-16 11:32:23 -08002297 struct ocfs2_insert_type insert = {0, };
2298 struct ocfs2_extent_rec rec;
Mark Fashehccd979b2005-12-15 14:31:24 -08002299
Mark Fashehdcd05382007-01-16 11:32:23 -08002300 mlog(0, "add %u clusters at position %u to inode %llu\n",
2301 new_clusters, cpos, (unsigned long long)OCFS2_I(inode)->ip_blkno);
Mark Fashehccd979b2005-12-15 14:31:24 -08002302
Mark Fashehdcd05382007-01-16 11:32:23 -08002303 mlog_bug_on_msg(!ocfs2_sparse_alloc(osb) &&
2304 (OCFS2_I(inode)->ip_clusters != cpos),
2305 "Device %s, asking for sparse allocation: inode %llu, "
2306 "cpos %u, clusters %u\n",
2307 osb->dev_str,
2308 (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos,
2309 OCFS2_I(inode)->ip_clusters);
Mark Fashehccd979b2005-12-15 14:31:24 -08002310
Mark Fashehdcd05382007-01-16 11:32:23 -08002311 rec.e_cpos = cpu_to_le32(cpos);
2312 rec.e_blkno = cpu_to_le64(start_blk);
2313 rec.e_clusters = cpu_to_le32(new_clusters);
Mark Fashehccd979b2005-12-15 14:31:24 -08002314
Mark Fashehdcd05382007-01-16 11:32:23 -08002315 status = ocfs2_figure_insert_type(inode, fe_bh, &last_eb_bh, &rec,
2316 &insert);
2317 if (status < 0) {
2318 mlog_errno(status);
2319 goto bail;
Mark Fashehccd979b2005-12-15 14:31:24 -08002320 }
2321
Mark Fashehdcd05382007-01-16 11:32:23 -08002322 mlog(0, "Insert.appending: %u, Insert.Contig: %u, "
2323 "Insert.contig_index: %d, Insert.free_records: %d, "
2324 "Insert.tree_depth: %d\n",
2325 insert.ins_appending, insert.ins_contig, insert.ins_contig_index,
2326 insert.ins_free_records, insert.ins_tree_depth);
Mark Fashehccd979b2005-12-15 14:31:24 -08002327
Mark Fashehdcd05382007-01-16 11:32:23 -08002328 /*
2329 * Avoid growing the tree unless we're out of records and the
2330 * insert type requres one.
2331 */
2332 if (insert.ins_contig != CONTIG_NONE || insert.ins_free_records)
2333 goto out_add;
Mark Fashehccd979b2005-12-15 14:31:24 -08002334
2335 shift = ocfs2_find_branch_target(osb, inode, fe_bh, &bh);
2336 if (shift < 0) {
2337 status = shift;
2338 mlog_errno(status);
2339 goto bail;
2340 }
2341
2342 /* We traveled all the way to the bottom of the allocation tree
2343 * and didn't find room for any more extents - we need to add
2344 * another tree level */
2345 if (shift) {
Mark Fashehccd979b2005-12-15 14:31:24 -08002346 BUG_ON(bh);
Mark Fashehdcd05382007-01-16 11:32:23 -08002347 mlog(0, "need to shift tree depth "
2348 "(current = %d)\n", insert.ins_tree_depth);
Mark Fashehccd979b2005-12-15 14:31:24 -08002349
2350 /* ocfs2_shift_tree_depth will return us a buffer with
2351 * the new extent block (so we can pass that to
2352 * ocfs2_add_branch). */
2353 status = ocfs2_shift_tree_depth(osb, handle, inode, fe_bh,
2354 meta_ac, &bh);
2355 if (status < 0) {
2356 mlog_errno(status);
2357 goto bail;
2358 }
Mark Fashehdcd05382007-01-16 11:32:23 -08002359 insert.ins_tree_depth++;
Mark Fashehccd979b2005-12-15 14:31:24 -08002360 /* Special case: we have room now if we shifted from
2361 * tree_depth 0 */
Mark Fashehdcd05382007-01-16 11:32:23 -08002362 if (insert.ins_tree_depth == 1)
Mark Fashehccd979b2005-12-15 14:31:24 -08002363 goto out_add;
2364 }
2365
2366 /* call ocfs2_add_branch to add the final part of the tree with
2367 * the new data. */
Mark Fashehdcd05382007-01-16 11:32:23 -08002368 mlog(0, "add branch. bh = %p\n", bh);
Mark Fashehccd979b2005-12-15 14:31:24 -08002369 status = ocfs2_add_branch(osb, handle, inode, fe_bh, bh, last_eb_bh,
2370 meta_ac);
2371 if (status < 0) {
2372 mlog_errno(status);
2373 goto bail;
2374 }
2375
2376out_add:
Mark Fashehdcd05382007-01-16 11:32:23 -08002377 /* Finally, we can add clusters. This might rotate the tree for us. */
2378 status = ocfs2_do_insert_extent(inode, handle, fe_bh, &rec, &insert);
Mark Fashehccd979b2005-12-15 14:31:24 -08002379 if (status < 0)
2380 mlog_errno(status);
2381
2382bail:
2383 if (bh)
2384 brelse(bh);
2385
2386 if (last_eb_bh)
2387 brelse(last_eb_bh);
2388
2389 mlog_exit(status);
2390 return status;
2391}
2392
2393static inline int ocfs2_truncate_log_needs_flush(struct ocfs2_super *osb)
2394{
2395 struct buffer_head *tl_bh = osb->osb_tl_bh;
2396 struct ocfs2_dinode *di;
2397 struct ocfs2_truncate_log *tl;
2398
2399 di = (struct ocfs2_dinode *) tl_bh->b_data;
2400 tl = &di->id2.i_dealloc;
2401
2402 mlog_bug_on_msg(le16_to_cpu(tl->tl_used) > le16_to_cpu(tl->tl_count),
2403 "slot %d, invalid truncate log parameters: used = "
2404 "%u, count = %u\n", osb->slot_num,
2405 le16_to_cpu(tl->tl_used), le16_to_cpu(tl->tl_count));
2406 return le16_to_cpu(tl->tl_used) == le16_to_cpu(tl->tl_count);
2407}
2408
2409static int ocfs2_truncate_log_can_coalesce(struct ocfs2_truncate_log *tl,
2410 unsigned int new_start)
2411{
2412 unsigned int tail_index;
2413 unsigned int current_tail;
2414
2415 /* No records, nothing to coalesce */
2416 if (!le16_to_cpu(tl->tl_used))
2417 return 0;
2418
2419 tail_index = le16_to_cpu(tl->tl_used) - 1;
2420 current_tail = le32_to_cpu(tl->tl_recs[tail_index].t_start);
2421 current_tail += le32_to_cpu(tl->tl_recs[tail_index].t_clusters);
2422
2423 return current_tail == new_start;
2424}
2425
2426static int ocfs2_truncate_log_append(struct ocfs2_super *osb,
Mark Fasheh1fabe142006-10-09 18:11:45 -07002427 handle_t *handle,
Mark Fashehccd979b2005-12-15 14:31:24 -08002428 u64 start_blk,
2429 unsigned int num_clusters)
2430{
2431 int status, index;
2432 unsigned int start_cluster, tl_count;
2433 struct inode *tl_inode = osb->osb_tl_inode;
2434 struct buffer_head *tl_bh = osb->osb_tl_bh;
2435 struct ocfs2_dinode *di;
2436 struct ocfs2_truncate_log *tl;
2437
Mark Fashehb0697052006-03-03 10:24:33 -08002438 mlog_entry("start_blk = %llu, num_clusters = %u\n",
2439 (unsigned long long)start_blk, num_clusters);
Mark Fashehccd979b2005-12-15 14:31:24 -08002440
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08002441 BUG_ON(mutex_trylock(&tl_inode->i_mutex));
Mark Fashehccd979b2005-12-15 14:31:24 -08002442
2443 start_cluster = ocfs2_blocks_to_clusters(osb->sb, start_blk);
2444
2445 di = (struct ocfs2_dinode *) tl_bh->b_data;
2446 tl = &di->id2.i_dealloc;
2447 if (!OCFS2_IS_VALID_DINODE(di)) {
2448 OCFS2_RO_ON_INVALID_DINODE(osb->sb, di);
2449 status = -EIO;
2450 goto bail;
2451 }
2452
2453 tl_count = le16_to_cpu(tl->tl_count);
2454 mlog_bug_on_msg(tl_count > ocfs2_truncate_recs_per_inode(osb->sb) ||
2455 tl_count == 0,
Mark Fashehb0697052006-03-03 10:24:33 -08002456 "Truncate record count on #%llu invalid "
2457 "wanted %u, actual %u\n",
2458 (unsigned long long)OCFS2_I(tl_inode)->ip_blkno,
Mark Fashehccd979b2005-12-15 14:31:24 -08002459 ocfs2_truncate_recs_per_inode(osb->sb),
2460 le16_to_cpu(tl->tl_count));
2461
2462 /* Caller should have known to flush before calling us. */
2463 index = le16_to_cpu(tl->tl_used);
2464 if (index >= tl_count) {
2465 status = -ENOSPC;
2466 mlog_errno(status);
2467 goto bail;
2468 }
2469
2470 status = ocfs2_journal_access(handle, tl_inode, tl_bh,
2471 OCFS2_JOURNAL_ACCESS_WRITE);
2472 if (status < 0) {
2473 mlog_errno(status);
2474 goto bail;
2475 }
2476
2477 mlog(0, "Log truncate of %u clusters starting at cluster %u to "
Mark Fashehb0697052006-03-03 10:24:33 -08002478 "%llu (index = %d)\n", num_clusters, start_cluster,
2479 (unsigned long long)OCFS2_I(tl_inode)->ip_blkno, index);
Mark Fashehccd979b2005-12-15 14:31:24 -08002480
2481 if (ocfs2_truncate_log_can_coalesce(tl, start_cluster)) {
2482 /*
2483 * Move index back to the record we are coalescing with.
2484 * ocfs2_truncate_log_can_coalesce() guarantees nonzero
2485 */
2486 index--;
2487
2488 num_clusters += le32_to_cpu(tl->tl_recs[index].t_clusters);
2489 mlog(0, "Coalesce with index %u (start = %u, clusters = %u)\n",
2490 index, le32_to_cpu(tl->tl_recs[index].t_start),
2491 num_clusters);
2492 } else {
2493 tl->tl_recs[index].t_start = cpu_to_le32(start_cluster);
2494 tl->tl_used = cpu_to_le16(index + 1);
2495 }
2496 tl->tl_recs[index].t_clusters = cpu_to_le32(num_clusters);
2497
2498 status = ocfs2_journal_dirty(handle, tl_bh);
2499 if (status < 0) {
2500 mlog_errno(status);
2501 goto bail;
2502 }
2503
2504bail:
2505 mlog_exit(status);
2506 return status;
2507}
2508
2509static int ocfs2_replay_truncate_records(struct ocfs2_super *osb,
Mark Fasheh1fabe142006-10-09 18:11:45 -07002510 handle_t *handle,
Mark Fashehccd979b2005-12-15 14:31:24 -08002511 struct inode *data_alloc_inode,
2512 struct buffer_head *data_alloc_bh)
2513{
2514 int status = 0;
2515 int i;
2516 unsigned int num_clusters;
2517 u64 start_blk;
2518 struct ocfs2_truncate_rec rec;
2519 struct ocfs2_dinode *di;
2520 struct ocfs2_truncate_log *tl;
2521 struct inode *tl_inode = osb->osb_tl_inode;
2522 struct buffer_head *tl_bh = osb->osb_tl_bh;
2523
2524 mlog_entry_void();
2525
2526 di = (struct ocfs2_dinode *) tl_bh->b_data;
2527 tl = &di->id2.i_dealloc;
2528 i = le16_to_cpu(tl->tl_used) - 1;
2529 while (i >= 0) {
2530 /* Caller has given us at least enough credits to
2531 * update the truncate log dinode */
2532 status = ocfs2_journal_access(handle, tl_inode, tl_bh,
2533 OCFS2_JOURNAL_ACCESS_WRITE);
2534 if (status < 0) {
2535 mlog_errno(status);
2536 goto bail;
2537 }
2538
2539 tl->tl_used = cpu_to_le16(i);
2540
2541 status = ocfs2_journal_dirty(handle, tl_bh);
2542 if (status < 0) {
2543 mlog_errno(status);
2544 goto bail;
2545 }
2546
2547 /* TODO: Perhaps we can calculate the bulk of the
2548 * credits up front rather than extending like
2549 * this. */
2550 status = ocfs2_extend_trans(handle,
2551 OCFS2_TRUNCATE_LOG_FLUSH_ONE_REC);
2552 if (status < 0) {
2553 mlog_errno(status);
2554 goto bail;
2555 }
2556
2557 rec = tl->tl_recs[i];
2558 start_blk = ocfs2_clusters_to_blocks(data_alloc_inode->i_sb,
2559 le32_to_cpu(rec.t_start));
2560 num_clusters = le32_to_cpu(rec.t_clusters);
2561
2562 /* if start_blk is not set, we ignore the record as
2563 * invalid. */
2564 if (start_blk) {
2565 mlog(0, "free record %d, start = %u, clusters = %u\n",
2566 i, le32_to_cpu(rec.t_start), num_clusters);
2567
2568 status = ocfs2_free_clusters(handle, data_alloc_inode,
2569 data_alloc_bh, start_blk,
2570 num_clusters);
2571 if (status < 0) {
2572 mlog_errno(status);
2573 goto bail;
2574 }
2575 }
2576 i--;
2577 }
2578
2579bail:
2580 mlog_exit(status);
2581 return status;
2582}
2583
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08002584/* Expects you to already be holding tl_inode->i_mutex */
Mark Fashehccd979b2005-12-15 14:31:24 -08002585static int __ocfs2_flush_truncate_log(struct ocfs2_super *osb)
2586{
2587 int status;
2588 unsigned int num_to_flush;
Mark Fasheh1fabe142006-10-09 18:11:45 -07002589 handle_t *handle;
Mark Fashehccd979b2005-12-15 14:31:24 -08002590 struct inode *tl_inode = osb->osb_tl_inode;
2591 struct inode *data_alloc_inode = NULL;
2592 struct buffer_head *tl_bh = osb->osb_tl_bh;
2593 struct buffer_head *data_alloc_bh = NULL;
2594 struct ocfs2_dinode *di;
2595 struct ocfs2_truncate_log *tl;
2596
2597 mlog_entry_void();
2598
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08002599 BUG_ON(mutex_trylock(&tl_inode->i_mutex));
Mark Fashehccd979b2005-12-15 14:31:24 -08002600
2601 di = (struct ocfs2_dinode *) tl_bh->b_data;
2602 tl = &di->id2.i_dealloc;
2603 if (!OCFS2_IS_VALID_DINODE(di)) {
2604 OCFS2_RO_ON_INVALID_DINODE(osb->sb, di);
2605 status = -EIO;
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002606 goto out;
Mark Fashehccd979b2005-12-15 14:31:24 -08002607 }
2608
2609 num_to_flush = le16_to_cpu(tl->tl_used);
Mark Fashehb0697052006-03-03 10:24:33 -08002610 mlog(0, "Flush %u records from truncate log #%llu\n",
2611 num_to_flush, (unsigned long long)OCFS2_I(tl_inode)->ip_blkno);
Mark Fashehccd979b2005-12-15 14:31:24 -08002612 if (!num_to_flush) {
2613 status = 0;
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002614 goto out;
Mark Fashehccd979b2005-12-15 14:31:24 -08002615 }
2616
2617 data_alloc_inode = ocfs2_get_system_file_inode(osb,
2618 GLOBAL_BITMAP_SYSTEM_INODE,
2619 OCFS2_INVALID_SLOT);
2620 if (!data_alloc_inode) {
2621 status = -EINVAL;
2622 mlog(ML_ERROR, "Could not get bitmap inode!\n");
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002623 goto out;
Mark Fashehccd979b2005-12-15 14:31:24 -08002624 }
2625
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002626 mutex_lock(&data_alloc_inode->i_mutex);
2627
Mark Fasheh4bcec182006-10-09 16:02:40 -07002628 status = ocfs2_meta_lock(data_alloc_inode, &data_alloc_bh, 1);
Mark Fashehccd979b2005-12-15 14:31:24 -08002629 if (status < 0) {
2630 mlog_errno(status);
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002631 goto out_mutex;
Mark Fashehccd979b2005-12-15 14:31:24 -08002632 }
2633
Mark Fasheh65eff9c2006-10-09 17:26:22 -07002634 handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_UPDATE);
Mark Fashehccd979b2005-12-15 14:31:24 -08002635 if (IS_ERR(handle)) {
2636 status = PTR_ERR(handle);
Mark Fashehccd979b2005-12-15 14:31:24 -08002637 mlog_errno(status);
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002638 goto out_unlock;
Mark Fashehccd979b2005-12-15 14:31:24 -08002639 }
2640
2641 status = ocfs2_replay_truncate_records(osb, handle, data_alloc_inode,
2642 data_alloc_bh);
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002643 if (status < 0)
Mark Fashehccd979b2005-12-15 14:31:24 -08002644 mlog_errno(status);
Mark Fashehccd979b2005-12-15 14:31:24 -08002645
Mark Fasheh02dc1af2006-10-09 16:48:10 -07002646 ocfs2_commit_trans(osb, handle);
Mark Fashehccd979b2005-12-15 14:31:24 -08002647
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002648out_unlock:
2649 brelse(data_alloc_bh);
2650 ocfs2_meta_unlock(data_alloc_inode, 1);
Mark Fashehccd979b2005-12-15 14:31:24 -08002651
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002652out_mutex:
2653 mutex_unlock(&data_alloc_inode->i_mutex);
2654 iput(data_alloc_inode);
Mark Fashehccd979b2005-12-15 14:31:24 -08002655
Mark Fashehe08dc8b2006-10-05 15:58:48 -07002656out:
Mark Fashehccd979b2005-12-15 14:31:24 -08002657 mlog_exit(status);
2658 return status;
2659}
2660
2661int ocfs2_flush_truncate_log(struct ocfs2_super *osb)
2662{
2663 int status;
2664 struct inode *tl_inode = osb->osb_tl_inode;
2665
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08002666 mutex_lock(&tl_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08002667 status = __ocfs2_flush_truncate_log(osb);
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08002668 mutex_unlock(&tl_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08002669
2670 return status;
2671}
2672
David Howellsc4028952006-11-22 14:57:56 +00002673static void ocfs2_truncate_log_worker(struct work_struct *work)
Mark Fashehccd979b2005-12-15 14:31:24 -08002674{
2675 int status;
David Howellsc4028952006-11-22 14:57:56 +00002676 struct ocfs2_super *osb =
2677 container_of(work, struct ocfs2_super,
2678 osb_truncate_log_wq.work);
Mark Fashehccd979b2005-12-15 14:31:24 -08002679
2680 mlog_entry_void();
2681
2682 status = ocfs2_flush_truncate_log(osb);
2683 if (status < 0)
2684 mlog_errno(status);
2685
2686 mlog_exit(status);
2687}
2688
2689#define OCFS2_TRUNCATE_LOG_FLUSH_INTERVAL (2 * HZ)
2690void ocfs2_schedule_truncate_log_flush(struct ocfs2_super *osb,
2691 int cancel)
2692{
2693 if (osb->osb_tl_inode) {
2694 /* We want to push off log flushes while truncates are
2695 * still running. */
2696 if (cancel)
2697 cancel_delayed_work(&osb->osb_truncate_log_wq);
2698
2699 queue_delayed_work(ocfs2_wq, &osb->osb_truncate_log_wq,
2700 OCFS2_TRUNCATE_LOG_FLUSH_INTERVAL);
2701 }
2702}
2703
2704static int ocfs2_get_truncate_log_info(struct ocfs2_super *osb,
2705 int slot_num,
2706 struct inode **tl_inode,
2707 struct buffer_head **tl_bh)
2708{
2709 int status;
2710 struct inode *inode = NULL;
2711 struct buffer_head *bh = NULL;
2712
2713 inode = ocfs2_get_system_file_inode(osb,
2714 TRUNCATE_LOG_SYSTEM_INODE,
2715 slot_num);
2716 if (!inode) {
2717 status = -EINVAL;
2718 mlog(ML_ERROR, "Could not get load truncate log inode!\n");
2719 goto bail;
2720 }
2721
2722 status = ocfs2_read_block(osb, OCFS2_I(inode)->ip_blkno, &bh,
2723 OCFS2_BH_CACHED, inode);
2724 if (status < 0) {
2725 iput(inode);
2726 mlog_errno(status);
2727 goto bail;
2728 }
2729
2730 *tl_inode = inode;
2731 *tl_bh = bh;
2732bail:
2733 mlog_exit(status);
2734 return status;
2735}
2736
2737/* called during the 1st stage of node recovery. we stamp a clean
2738 * truncate log and pass back a copy for processing later. if the
2739 * truncate log does not require processing, a *tl_copy is set to
2740 * NULL. */
2741int ocfs2_begin_truncate_log_recovery(struct ocfs2_super *osb,
2742 int slot_num,
2743 struct ocfs2_dinode **tl_copy)
2744{
2745 int status;
2746 struct inode *tl_inode = NULL;
2747 struct buffer_head *tl_bh = NULL;
2748 struct ocfs2_dinode *di;
2749 struct ocfs2_truncate_log *tl;
2750
2751 *tl_copy = NULL;
2752
2753 mlog(0, "recover truncate log from slot %d\n", slot_num);
2754
2755 status = ocfs2_get_truncate_log_info(osb, slot_num, &tl_inode, &tl_bh);
2756 if (status < 0) {
2757 mlog_errno(status);
2758 goto bail;
2759 }
2760
2761 di = (struct ocfs2_dinode *) tl_bh->b_data;
2762 tl = &di->id2.i_dealloc;
2763 if (!OCFS2_IS_VALID_DINODE(di)) {
2764 OCFS2_RO_ON_INVALID_DINODE(tl_inode->i_sb, di);
2765 status = -EIO;
2766 goto bail;
2767 }
2768
2769 if (le16_to_cpu(tl->tl_used)) {
2770 mlog(0, "We'll have %u logs to recover\n",
2771 le16_to_cpu(tl->tl_used));
2772
2773 *tl_copy = kmalloc(tl_bh->b_size, GFP_KERNEL);
2774 if (!(*tl_copy)) {
2775 status = -ENOMEM;
2776 mlog_errno(status);
2777 goto bail;
2778 }
2779
2780 /* Assuming the write-out below goes well, this copy
2781 * will be passed back to recovery for processing. */
2782 memcpy(*tl_copy, tl_bh->b_data, tl_bh->b_size);
2783
2784 /* All we need to do to clear the truncate log is set
2785 * tl_used. */
2786 tl->tl_used = 0;
2787
2788 status = ocfs2_write_block(osb, tl_bh, tl_inode);
2789 if (status < 0) {
2790 mlog_errno(status);
2791 goto bail;
2792 }
2793 }
2794
2795bail:
2796 if (tl_inode)
2797 iput(tl_inode);
2798 if (tl_bh)
2799 brelse(tl_bh);
2800
2801 if (status < 0 && (*tl_copy)) {
2802 kfree(*tl_copy);
2803 *tl_copy = NULL;
2804 }
2805
2806 mlog_exit(status);
2807 return status;
2808}
2809
2810int ocfs2_complete_truncate_log_recovery(struct ocfs2_super *osb,
2811 struct ocfs2_dinode *tl_copy)
2812{
2813 int status = 0;
2814 int i;
2815 unsigned int clusters, num_recs, start_cluster;
2816 u64 start_blk;
Mark Fasheh1fabe142006-10-09 18:11:45 -07002817 handle_t *handle;
Mark Fashehccd979b2005-12-15 14:31:24 -08002818 struct inode *tl_inode = osb->osb_tl_inode;
2819 struct ocfs2_truncate_log *tl;
2820
2821 mlog_entry_void();
2822
2823 if (OCFS2_I(tl_inode)->ip_blkno == le64_to_cpu(tl_copy->i_blkno)) {
2824 mlog(ML_ERROR, "Asked to recover my own truncate log!\n");
2825 return -EINVAL;
2826 }
2827
2828 tl = &tl_copy->id2.i_dealloc;
2829 num_recs = le16_to_cpu(tl->tl_used);
Mark Fashehb0697052006-03-03 10:24:33 -08002830 mlog(0, "cleanup %u records from %llu\n", num_recs,
2831 (unsigned long long)tl_copy->i_blkno);
Mark Fashehccd979b2005-12-15 14:31:24 -08002832
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08002833 mutex_lock(&tl_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08002834 for(i = 0; i < num_recs; i++) {
2835 if (ocfs2_truncate_log_needs_flush(osb)) {
2836 status = __ocfs2_flush_truncate_log(osb);
2837 if (status < 0) {
2838 mlog_errno(status);
2839 goto bail_up;
2840 }
2841 }
2842
Mark Fasheh65eff9c2006-10-09 17:26:22 -07002843 handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_UPDATE);
Mark Fashehccd979b2005-12-15 14:31:24 -08002844 if (IS_ERR(handle)) {
2845 status = PTR_ERR(handle);
2846 mlog_errno(status);
2847 goto bail_up;
2848 }
2849
2850 clusters = le32_to_cpu(tl->tl_recs[i].t_clusters);
2851 start_cluster = le32_to_cpu(tl->tl_recs[i].t_start);
2852 start_blk = ocfs2_clusters_to_blocks(osb->sb, start_cluster);
2853
2854 status = ocfs2_truncate_log_append(osb, handle,
2855 start_blk, clusters);
Mark Fasheh02dc1af2006-10-09 16:48:10 -07002856 ocfs2_commit_trans(osb, handle);
Mark Fashehccd979b2005-12-15 14:31:24 -08002857 if (status < 0) {
2858 mlog_errno(status);
2859 goto bail_up;
2860 }
2861 }
2862
2863bail_up:
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08002864 mutex_unlock(&tl_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08002865
2866 mlog_exit(status);
2867 return status;
2868}
2869
2870void ocfs2_truncate_log_shutdown(struct ocfs2_super *osb)
2871{
2872 int status;
2873 struct inode *tl_inode = osb->osb_tl_inode;
2874
2875 mlog_entry_void();
2876
2877 if (tl_inode) {
2878 cancel_delayed_work(&osb->osb_truncate_log_wq);
2879 flush_workqueue(ocfs2_wq);
2880
2881 status = ocfs2_flush_truncate_log(osb);
2882 if (status < 0)
2883 mlog_errno(status);
2884
2885 brelse(osb->osb_tl_bh);
2886 iput(osb->osb_tl_inode);
2887 }
2888
2889 mlog_exit_void();
2890}
2891
2892int ocfs2_truncate_log_init(struct ocfs2_super *osb)
2893{
2894 int status;
2895 struct inode *tl_inode = NULL;
2896 struct buffer_head *tl_bh = NULL;
2897
2898 mlog_entry_void();
2899
2900 status = ocfs2_get_truncate_log_info(osb,
2901 osb->slot_num,
2902 &tl_inode,
2903 &tl_bh);
2904 if (status < 0)
2905 mlog_errno(status);
2906
2907 /* ocfs2_truncate_log_shutdown keys on the existence of
2908 * osb->osb_tl_inode so we don't set any of the osb variables
2909 * until we're sure all is well. */
David Howellsc4028952006-11-22 14:57:56 +00002910 INIT_DELAYED_WORK(&osb->osb_truncate_log_wq,
2911 ocfs2_truncate_log_worker);
Mark Fashehccd979b2005-12-15 14:31:24 -08002912 osb->osb_tl_bh = tl_bh;
2913 osb->osb_tl_inode = tl_inode;
2914
2915 mlog_exit(status);
2916 return status;
2917}
2918
2919/* This function will figure out whether the currently last extent
2920 * block will be deleted, and if it will, what the new last extent
2921 * block will be so we can update his h_next_leaf_blk field, as well
2922 * as the dinodes i_last_eb_blk */
Mark Fashehdcd05382007-01-16 11:32:23 -08002923static int ocfs2_find_new_last_ext_blk(struct inode *inode,
Mark Fasheh3a0782d2007-01-17 12:53:31 -08002924 unsigned int clusters_to_del,
Mark Fashehdcd05382007-01-16 11:32:23 -08002925 struct ocfs2_path *path,
Mark Fashehccd979b2005-12-15 14:31:24 -08002926 struct buffer_head **new_last_eb)
2927{
Mark Fasheh3a0782d2007-01-17 12:53:31 -08002928 int next_free, ret = 0;
Mark Fashehdcd05382007-01-16 11:32:23 -08002929 u32 cpos;
Mark Fasheh3a0782d2007-01-17 12:53:31 -08002930 struct ocfs2_extent_rec *rec;
Mark Fashehccd979b2005-12-15 14:31:24 -08002931 struct ocfs2_extent_block *eb;
2932 struct ocfs2_extent_list *el;
2933 struct buffer_head *bh = NULL;
2934
2935 *new_last_eb = NULL;
2936
Mark Fashehccd979b2005-12-15 14:31:24 -08002937 /* we have no tree, so of course, no last_eb. */
Mark Fashehdcd05382007-01-16 11:32:23 -08002938 if (!path->p_tree_depth)
2939 goto out;
Mark Fashehccd979b2005-12-15 14:31:24 -08002940
2941 /* trunc to zero special case - this makes tree_depth = 0
2942 * regardless of what it is. */
Mark Fasheh3a0782d2007-01-17 12:53:31 -08002943 if (OCFS2_I(inode)->ip_clusters == clusters_to_del)
Mark Fashehdcd05382007-01-16 11:32:23 -08002944 goto out;
Mark Fashehccd979b2005-12-15 14:31:24 -08002945
Mark Fashehdcd05382007-01-16 11:32:23 -08002946 el = path_leaf_el(path);
Mark Fashehccd979b2005-12-15 14:31:24 -08002947 BUG_ON(!el->l_next_free_rec);
2948
Mark Fasheh3a0782d2007-01-17 12:53:31 -08002949 /*
2950 * Make sure that this extent list will actually be empty
2951 * after we clear away the data. We can shortcut out if
2952 * there's more than one non-empty extent in the
2953 * list. Otherwise, a check of the remaining extent is
2954 * necessary.
2955 */
2956 next_free = le16_to_cpu(el->l_next_free_rec);
2957 rec = NULL;
Mark Fashehdcd05382007-01-16 11:32:23 -08002958 if (ocfs2_is_empty_extent(&el->l_recs[0])) {
Mark Fasheh3a0782d2007-01-17 12:53:31 -08002959 if (next_free > 2)
Mark Fashehdcd05382007-01-16 11:32:23 -08002960 goto out;
Mark Fasheh3a0782d2007-01-17 12:53:31 -08002961
2962 /* We may have a valid extent in index 1, check it. */
2963 if (next_free == 2)
2964 rec = &el->l_recs[1];
2965
2966 /*
2967 * Fall through - no more nonempty extents, so we want
2968 * to delete this leaf.
2969 */
2970 } else {
2971 if (next_free > 1)
2972 goto out;
2973
2974 rec = &el->l_recs[0];
2975 }
2976
2977 if (rec) {
2978 /*
2979 * Check it we'll only be trimming off the end of this
2980 * cluster.
2981 */
2982 if (le16_to_cpu(rec->e_clusters) > clusters_to_del)
2983 goto out;
2984 }
Mark Fashehccd979b2005-12-15 14:31:24 -08002985
Mark Fashehdcd05382007-01-16 11:32:23 -08002986 ret = ocfs2_find_cpos_for_left_leaf(inode->i_sb, path, &cpos);
2987 if (ret) {
2988 mlog_errno(ret);
2989 goto out;
2990 }
Mark Fashehccd979b2005-12-15 14:31:24 -08002991
Mark Fashehdcd05382007-01-16 11:32:23 -08002992 ret = ocfs2_find_leaf(inode, path_root_el(path), cpos, &bh);
2993 if (ret) {
2994 mlog_errno(ret);
2995 goto out;
2996 }
Mark Fashehccd979b2005-12-15 14:31:24 -08002997
Mark Fashehdcd05382007-01-16 11:32:23 -08002998 eb = (struct ocfs2_extent_block *) bh->b_data;
2999 el = &eb->h_list;
3000 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
3001 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
3002 ret = -EROFS;
3003 goto out;
3004 }
Mark Fashehccd979b2005-12-15 14:31:24 -08003005
3006 *new_last_eb = bh;
3007 get_bh(*new_last_eb);
Mark Fashehdcd05382007-01-16 11:32:23 -08003008 mlog(0, "returning block %llu, (cpos: %u)\n",
3009 (unsigned long long)le64_to_cpu(eb->h_blkno), cpos);
3010out:
3011 brelse(bh);
Mark Fashehccd979b2005-12-15 14:31:24 -08003012
Mark Fashehdcd05382007-01-16 11:32:23 -08003013 return ret;
Mark Fashehccd979b2005-12-15 14:31:24 -08003014}
3015
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003016/*
3017 * Trim some clusters off the rightmost edge of a tree. Only called
3018 * during truncate.
3019 *
3020 * The caller needs to:
3021 * - start journaling of each path component.
3022 * - compute and fully set up any new last ext block
3023 */
3024static int ocfs2_trim_tree(struct inode *inode, struct ocfs2_path *path,
3025 handle_t *handle, struct ocfs2_truncate_context *tc,
3026 u32 clusters_to_del, u64 *delete_start)
3027{
3028 int ret, i, index = path->p_tree_depth;
3029 u32 new_edge = 0;
3030 u64 deleted_eb = 0;
3031 struct buffer_head *bh;
3032 struct ocfs2_extent_list *el;
3033 struct ocfs2_extent_rec *rec;
3034
3035 *delete_start = 0;
3036
3037 while (index >= 0) {
3038 bh = path->p_node[index].bh;
3039 el = path->p_node[index].el;
3040
3041 mlog(0, "traveling tree (index = %d, block = %llu)\n",
3042 index, (unsigned long long)bh->b_blocknr);
3043
3044 BUG_ON(le16_to_cpu(el->l_next_free_rec) == 0);
3045
3046 if (index !=
3047 (path->p_tree_depth - le16_to_cpu(el->l_tree_depth))) {
3048 ocfs2_error(inode->i_sb,
3049 "Inode %lu has invalid ext. block %llu",
3050 inode->i_ino,
3051 (unsigned long long)bh->b_blocknr);
3052 ret = -EROFS;
3053 goto out;
3054 }
3055
3056find_tail_record:
3057 i = le16_to_cpu(el->l_next_free_rec) - 1;
3058 rec = &el->l_recs[i];
3059
3060 mlog(0, "Extent list before: record %d: (%u, %u, %llu), "
3061 "next = %u\n", i, le32_to_cpu(rec->e_cpos),
3062 le32_to_cpu(rec->e_clusters),
3063 (unsigned long long)le64_to_cpu(rec->e_blkno),
3064 le16_to_cpu(el->l_next_free_rec));
3065
3066 BUG_ON(le32_to_cpu(rec->e_clusters) < clusters_to_del);
3067
3068 if (le16_to_cpu(el->l_tree_depth) == 0) {
3069 /*
3070 * If the leaf block contains a single empty
3071 * extent and no records, we can just remove
3072 * the block.
3073 */
3074 if (i == 0 && ocfs2_is_empty_extent(rec)) {
3075 memset(rec, 0,
3076 sizeof(struct ocfs2_extent_rec));
3077 el->l_next_free_rec = cpu_to_le16(0);
3078
3079 goto delete;
3080 }
3081
3082 /*
3083 * Remove any empty extents by shifting things
3084 * left. That should make life much easier on
3085 * the code below. This condition is rare
3086 * enough that we shouldn't see a performance
3087 * hit.
3088 */
3089 if (ocfs2_is_empty_extent(&el->l_recs[0])) {
3090 le16_add_cpu(&el->l_next_free_rec, -1);
3091
3092 for(i = 0;
3093 i < le16_to_cpu(el->l_next_free_rec); i++)
3094 el->l_recs[i] = el->l_recs[i + 1];
3095
3096 memset(&el->l_recs[i], 0,
3097 sizeof(struct ocfs2_extent_rec));
3098
3099 /*
3100 * We've modified our extent list. The
3101 * simplest way to handle this change
3102 * is to being the search from the
3103 * start again.
3104 */
3105 goto find_tail_record;
3106 }
3107
3108 le32_add_cpu(&rec->e_clusters, -clusters_to_del);
3109
3110 /*
3111 * We'll use "new_edge" on our way back up the
3112 * tree to know what our rightmost cpos is.
3113 */
3114 new_edge = le32_to_cpu(rec->e_clusters);
3115 new_edge += le32_to_cpu(rec->e_cpos);
3116
3117 /*
3118 * The caller will use this to delete data blocks.
3119 */
3120 *delete_start = le64_to_cpu(rec->e_blkno)
3121 + ocfs2_clusters_to_blocks(inode->i_sb,
3122 le32_to_cpu(rec->e_clusters));
3123
3124 /*
3125 * If it's now empty, remove this record.
3126 */
3127 if (le32_to_cpu(rec->e_clusters) == 0) {
3128 memset(rec, 0,
3129 sizeof(struct ocfs2_extent_rec));
3130 le16_add_cpu(&el->l_next_free_rec, -1);
3131 }
3132 } else {
3133 if (le64_to_cpu(rec->e_blkno) == deleted_eb) {
3134 memset(rec, 0,
3135 sizeof(struct ocfs2_extent_rec));
3136 le16_add_cpu(&el->l_next_free_rec, -1);
3137
3138 goto delete;
3139 }
3140
3141 /* Can this actually happen? */
3142 if (le16_to_cpu(el->l_next_free_rec) == 0)
3143 goto delete;
3144
3145 /*
3146 * We never actually deleted any clusters
3147 * because our leaf was empty. There's no
3148 * reason to adjust the rightmost edge then.
3149 */
3150 if (new_edge == 0)
3151 goto delete;
3152
3153 rec->e_clusters = cpu_to_le32(new_edge);
3154 le32_add_cpu(&rec->e_clusters,
3155 -le32_to_cpu(rec->e_cpos));
3156
3157 /*
3158 * A deleted child record should have been
3159 * caught above.
3160 */
3161 BUG_ON(le32_to_cpu(rec->e_clusters) == 0);
3162 }
3163
3164delete:
3165 ret = ocfs2_journal_dirty(handle, bh);
3166 if (ret) {
3167 mlog_errno(ret);
3168 goto out;
3169 }
3170
3171 mlog(0, "extent list container %llu, after: record %d: "
3172 "(%u, %u, %llu), next = %u.\n",
3173 (unsigned long long)bh->b_blocknr, i,
3174 le32_to_cpu(rec->e_cpos), le32_to_cpu(rec->e_clusters),
3175 (unsigned long long)le64_to_cpu(rec->e_blkno),
3176 le16_to_cpu(el->l_next_free_rec));
3177
3178 /*
3179 * We must be careful to only attempt delete of an
3180 * extent block (and not the root inode block).
3181 */
3182 if (index > 0 && le16_to_cpu(el->l_next_free_rec) == 0) {
3183 struct ocfs2_extent_block *eb =
3184 (struct ocfs2_extent_block *)bh->b_data;
3185
3186 /*
3187 * Save this for use when processing the
3188 * parent block.
3189 */
3190 deleted_eb = le64_to_cpu(eb->h_blkno);
3191
3192 mlog(0, "deleting this extent block.\n");
3193
3194 ocfs2_remove_from_cache(inode, bh);
3195
3196 BUG_ON(le32_to_cpu(el->l_recs[0].e_clusters));
3197 BUG_ON(le32_to_cpu(el->l_recs[0].e_cpos));
3198 BUG_ON(le64_to_cpu(el->l_recs[0].e_blkno));
3199
3200 if (le16_to_cpu(eb->h_suballoc_slot) == 0) {
3201 /*
3202 * This code only understands how to
3203 * lock the suballocator in slot 0,
3204 * which is fine because allocation is
3205 * only ever done out of that
3206 * suballocator too. A future version
3207 * might change that however, so avoid
3208 * a free if we don't know how to
3209 * handle it. This way an fs incompat
3210 * bit will not be necessary.
3211 */
3212 ret = ocfs2_free_extent_block(handle,
3213 tc->tc_ext_alloc_inode,
3214 tc->tc_ext_alloc_bh,
3215 eb);
3216
3217 /* An error here is not fatal. */
3218 if (ret < 0)
3219 mlog_errno(ret);
3220 }
3221 } else {
3222 deleted_eb = 0;
3223 }
3224
3225 index--;
3226 }
3227
3228 ret = 0;
3229out:
3230 return ret;
3231}
3232
Mark Fashehccd979b2005-12-15 14:31:24 -08003233static int ocfs2_do_truncate(struct ocfs2_super *osb,
3234 unsigned int clusters_to_del,
3235 struct inode *inode,
3236 struct buffer_head *fe_bh,
Mark Fasheh1fabe142006-10-09 18:11:45 -07003237 handle_t *handle,
Mark Fashehdcd05382007-01-16 11:32:23 -08003238 struct ocfs2_truncate_context *tc,
3239 struct ocfs2_path *path)
Mark Fashehccd979b2005-12-15 14:31:24 -08003240{
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003241 int status;
Mark Fashehccd979b2005-12-15 14:31:24 -08003242 struct ocfs2_dinode *fe;
Mark Fashehccd979b2005-12-15 14:31:24 -08003243 struct ocfs2_extent_block *last_eb = NULL;
3244 struct ocfs2_extent_list *el;
Mark Fashehccd979b2005-12-15 14:31:24 -08003245 struct buffer_head *last_eb_bh = NULL;
Mark Fashehccd979b2005-12-15 14:31:24 -08003246 u64 delete_blk = 0;
3247
3248 fe = (struct ocfs2_dinode *) fe_bh->b_data;
3249
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003250 status = ocfs2_find_new_last_ext_blk(inode, clusters_to_del,
Mark Fashehdcd05382007-01-16 11:32:23 -08003251 path, &last_eb_bh);
Mark Fashehccd979b2005-12-15 14:31:24 -08003252 if (status < 0) {
3253 mlog_errno(status);
3254 goto bail;
3255 }
Mark Fashehccd979b2005-12-15 14:31:24 -08003256
Mark Fashehdcd05382007-01-16 11:32:23 -08003257 /*
3258 * Each component will be touched, so we might as well journal
3259 * here to avoid having to handle errors later.
3260 */
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003261 status = ocfs2_journal_access_path(inode, handle, path);
3262 if (status < 0) {
3263 mlog_errno(status);
3264 goto bail;
Mark Fashehdcd05382007-01-16 11:32:23 -08003265 }
3266
3267 if (last_eb_bh) {
3268 status = ocfs2_journal_access(handle, inode, last_eb_bh,
3269 OCFS2_JOURNAL_ACCESS_WRITE);
3270 if (status < 0) {
3271 mlog_errno(status);
3272 goto bail;
3273 }
3274
3275 last_eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
3276 }
3277
3278 el = &(fe->id2.i_list);
3279
3280 /*
3281 * Lower levels depend on this never happening, but it's best
3282 * to check it up here before changing the tree.
3283 */
3284 if (el->l_tree_depth && ocfs2_is_empty_extent(&el->l_recs[0])) {
3285 ocfs2_error(inode->i_sb,
3286 "Inode %lu has an empty extent record, depth %u\n",
3287 inode->i_ino, le16_to_cpu(el->l_tree_depth));
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003288 status = -EROFS;
Mark Fashehccd979b2005-12-15 14:31:24 -08003289 goto bail;
3290 }
Mark Fashehccd979b2005-12-15 14:31:24 -08003291
3292 spin_lock(&OCFS2_I(inode)->ip_lock);
3293 OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters) -
3294 clusters_to_del;
3295 spin_unlock(&OCFS2_I(inode)->ip_lock);
3296 le32_add_cpu(&fe->i_clusters, -clusters_to_del);
Mark Fashehccd979b2005-12-15 14:31:24 -08003297
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003298 status = ocfs2_trim_tree(inode, path, handle, tc,
3299 clusters_to_del, &delete_blk);
3300 if (status) {
3301 mlog_errno(status);
3302 goto bail;
Mark Fashehccd979b2005-12-15 14:31:24 -08003303 }
3304
Mark Fashehdcd05382007-01-16 11:32:23 -08003305 if (le32_to_cpu(fe->i_clusters) == 0) {
Mark Fashehccd979b2005-12-15 14:31:24 -08003306 /* trunc to zero is a special case. */
3307 el->l_tree_depth = 0;
3308 fe->i_last_eb_blk = 0;
3309 } else if (last_eb)
3310 fe->i_last_eb_blk = last_eb->h_blkno;
3311
3312 status = ocfs2_journal_dirty(handle, fe_bh);
3313 if (status < 0) {
3314 mlog_errno(status);
3315 goto bail;
3316 }
3317
3318 if (last_eb) {
3319 /* If there will be a new last extent block, then by
3320 * definition, there cannot be any leaves to the right of
3321 * him. */
Mark Fashehccd979b2005-12-15 14:31:24 -08003322 last_eb->h_next_leaf_blk = 0;
3323 status = ocfs2_journal_dirty(handle, last_eb_bh);
3324 if (status < 0) {
3325 mlog_errno(status);
3326 goto bail;
3327 }
3328 }
3329
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003330 if (delete_blk) {
3331 status = ocfs2_truncate_log_append(osb, handle, delete_blk,
3332 clusters_to_del);
Mark Fashehccd979b2005-12-15 14:31:24 -08003333 if (status < 0) {
3334 mlog_errno(status);
3335 goto bail;
3336 }
Mark Fashehccd979b2005-12-15 14:31:24 -08003337 }
3338 status = 0;
3339bail:
Mark Fashehdcd05382007-01-16 11:32:23 -08003340
Mark Fashehccd979b2005-12-15 14:31:24 -08003341 mlog_exit(status);
3342 return status;
3343}
3344
3345/*
3346 * It is expected, that by the time you call this function,
3347 * inode->i_size and fe->i_size have been adjusted.
3348 *
3349 * WARNING: This will kfree the truncate context
3350 */
3351int ocfs2_commit_truncate(struct ocfs2_super *osb,
3352 struct inode *inode,
3353 struct buffer_head *fe_bh,
3354 struct ocfs2_truncate_context *tc)
3355{
3356 int status, i, credits, tl_sem = 0;
Mark Fashehdcd05382007-01-16 11:32:23 -08003357 u32 clusters_to_del, new_highest_cpos, range;
Mark Fashehccd979b2005-12-15 14:31:24 -08003358 struct ocfs2_extent_list *el;
Mark Fasheh1fabe142006-10-09 18:11:45 -07003359 handle_t *handle = NULL;
Mark Fashehccd979b2005-12-15 14:31:24 -08003360 struct inode *tl_inode = osb->osb_tl_inode;
Mark Fashehdcd05382007-01-16 11:32:23 -08003361 struct ocfs2_path *path = NULL;
Mark Fashehccd979b2005-12-15 14:31:24 -08003362
3363 mlog_entry_void();
3364
3365 down_write(&OCFS2_I(inode)->ip_alloc_sem);
3366
Mark Fashehdcd05382007-01-16 11:32:23 -08003367 new_highest_cpos = ocfs2_clusters_for_bytes(osb->sb,
Mark Fashehccd979b2005-12-15 14:31:24 -08003368 i_size_read(inode));
3369
Mark Fashehdcd05382007-01-16 11:32:23 -08003370 path = ocfs2_new_inode_path(fe_bh);
3371 if (!path) {
3372 status = -ENOMEM;
3373 mlog_errno(status);
3374 goto bail;
3375 }
Mark Fashehccd979b2005-12-15 14:31:24 -08003376start:
Mark Fashehdcd05382007-01-16 11:32:23 -08003377 /*
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003378 * Check that we still have allocation to delete.
3379 */
3380 if (OCFS2_I(inode)->ip_clusters == 0) {
3381 status = 0;
3382 goto bail;
3383 }
3384
3385 /*
Mark Fashehdcd05382007-01-16 11:32:23 -08003386 * Truncate always works against the rightmost tree branch.
3387 */
3388 status = ocfs2_find_path(inode, path, UINT_MAX);
3389 if (status) {
3390 mlog_errno(status);
3391 goto bail;
Mark Fashehccd979b2005-12-15 14:31:24 -08003392 }
3393
Mark Fashehdcd05382007-01-16 11:32:23 -08003394 mlog(0, "inode->ip_clusters = %u, tree_depth = %u\n",
3395 OCFS2_I(inode)->ip_clusters, path->p_tree_depth);
3396
3397 /*
3398 * By now, el will point to the extent list on the bottom most
3399 * portion of this tree. Only the tail record is considered in
3400 * each pass.
3401 *
3402 * We handle the following cases, in order:
3403 * - empty extent: delete the remaining branch
3404 * - remove the entire record
3405 * - remove a partial record
3406 * - no record needs to be removed (truncate has completed)
3407 */
3408 el = path_leaf_el(path);
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003409 if (le16_to_cpu(el->l_next_free_rec) == 0) {
3410 ocfs2_error(inode->i_sb,
3411 "Inode %llu has empty extent block at %llu\n",
3412 (unsigned long long)OCFS2_I(inode)->ip_blkno,
3413 (unsigned long long)path_leaf_bh(path)->b_blocknr);
3414 status = -EROFS;
3415 goto bail;
3416 }
3417
Mark Fashehccd979b2005-12-15 14:31:24 -08003418 i = le16_to_cpu(el->l_next_free_rec) - 1;
Mark Fashehdcd05382007-01-16 11:32:23 -08003419 range = le32_to_cpu(el->l_recs[i].e_cpos) +
3420 le32_to_cpu(el->l_recs[i].e_clusters);
3421 if (i == 0 && ocfs2_is_empty_extent(&el->l_recs[i])) {
3422 clusters_to_del = 0;
3423 } else if (le32_to_cpu(el->l_recs[i].e_cpos) >= new_highest_cpos) {
Mark Fashehccd979b2005-12-15 14:31:24 -08003424 clusters_to_del = le32_to_cpu(el->l_recs[i].e_clusters);
Mark Fashehdcd05382007-01-16 11:32:23 -08003425 } else if (range > new_highest_cpos) {
Mark Fashehccd979b2005-12-15 14:31:24 -08003426 clusters_to_del = (le32_to_cpu(el->l_recs[i].e_clusters) +
3427 le32_to_cpu(el->l_recs[i].e_cpos)) -
Mark Fashehdcd05382007-01-16 11:32:23 -08003428 new_highest_cpos;
3429 } else {
3430 status = 0;
3431 goto bail;
3432 }
Mark Fashehccd979b2005-12-15 14:31:24 -08003433
Mark Fashehdcd05382007-01-16 11:32:23 -08003434 mlog(0, "clusters_to_del = %u in this pass, tail blk=%llu\n",
3435 clusters_to_del, (unsigned long long)path_leaf_bh(path)->b_blocknr);
3436
3437 BUG_ON(clusters_to_del == 0);
Mark Fashehccd979b2005-12-15 14:31:24 -08003438
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08003439 mutex_lock(&tl_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08003440 tl_sem = 1;
3441 /* ocfs2_truncate_log_needs_flush guarantees us at least one
3442 * record is free for use. If there isn't any, we flush to get
3443 * an empty truncate log. */
3444 if (ocfs2_truncate_log_needs_flush(osb)) {
3445 status = __ocfs2_flush_truncate_log(osb);
3446 if (status < 0) {
3447 mlog_errno(status);
3448 goto bail;
3449 }
3450 }
3451
3452 credits = ocfs2_calc_tree_trunc_credits(osb->sb, clusters_to_del,
Mark Fashehdcd05382007-01-16 11:32:23 -08003453 (struct ocfs2_dinode *)fe_bh->b_data,
3454 el);
Mark Fasheh65eff9c2006-10-09 17:26:22 -07003455 handle = ocfs2_start_trans(osb, credits);
Mark Fashehccd979b2005-12-15 14:31:24 -08003456 if (IS_ERR(handle)) {
3457 status = PTR_ERR(handle);
3458 handle = NULL;
3459 mlog_errno(status);
3460 goto bail;
3461 }
3462
Mark Fashehdcd05382007-01-16 11:32:23 -08003463 status = ocfs2_do_truncate(osb, clusters_to_del, inode, fe_bh, handle,
3464 tc, path);
Mark Fashehccd979b2005-12-15 14:31:24 -08003465 if (status < 0) {
3466 mlog_errno(status);
3467 goto bail;
3468 }
3469
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08003470 mutex_unlock(&tl_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08003471 tl_sem = 0;
3472
Mark Fasheh02dc1af2006-10-09 16:48:10 -07003473 ocfs2_commit_trans(osb, handle);
Mark Fashehccd979b2005-12-15 14:31:24 -08003474 handle = NULL;
3475
Mark Fashehdcd05382007-01-16 11:32:23 -08003476 ocfs2_reinit_path(path, 1);
3477
3478 /*
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003479 * The check above will catch the case where we've truncated
3480 * away all allocation.
Mark Fashehdcd05382007-01-16 11:32:23 -08003481 */
Mark Fasheh3a0782d2007-01-17 12:53:31 -08003482 goto start;
3483
Mark Fashehccd979b2005-12-15 14:31:24 -08003484bail:
3485 up_write(&OCFS2_I(inode)->ip_alloc_sem);
3486
3487 ocfs2_schedule_truncate_log_flush(osb, 1);
3488
3489 if (tl_sem)
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08003490 mutex_unlock(&tl_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08003491
3492 if (handle)
Mark Fasheh02dc1af2006-10-09 16:48:10 -07003493 ocfs2_commit_trans(osb, handle);
Mark Fashehccd979b2005-12-15 14:31:24 -08003494
Mark Fashehdcd05382007-01-16 11:32:23 -08003495 ocfs2_free_path(path);
Mark Fashehccd979b2005-12-15 14:31:24 -08003496
3497 /* This will drop the ext_alloc cluster lock for us */
3498 ocfs2_free_truncate_context(tc);
3499
3500 mlog_exit(status);
3501 return status;
3502}
3503
Mark Fashehccd979b2005-12-15 14:31:24 -08003504/*
3505 * Expects the inode to already be locked. This will figure out which
3506 * inodes need to be locked and will put them on the returned truncate
3507 * context.
3508 */
3509int ocfs2_prepare_truncate(struct ocfs2_super *osb,
3510 struct inode *inode,
3511 struct buffer_head *fe_bh,
3512 struct ocfs2_truncate_context **tc)
3513{
Mark Fashehdcd05382007-01-16 11:32:23 -08003514 int status, metadata_delete, i;
Mark Fashehccd979b2005-12-15 14:31:24 -08003515 unsigned int new_i_clusters;
3516 struct ocfs2_dinode *fe;
3517 struct ocfs2_extent_block *eb;
3518 struct ocfs2_extent_list *el;
3519 struct buffer_head *last_eb_bh = NULL;
3520 struct inode *ext_alloc_inode = NULL;
3521 struct buffer_head *ext_alloc_bh = NULL;
3522
3523 mlog_entry_void();
3524
3525 *tc = NULL;
3526
3527 new_i_clusters = ocfs2_clusters_for_bytes(osb->sb,
3528 i_size_read(inode));
3529 fe = (struct ocfs2_dinode *) fe_bh->b_data;
3530
3531 mlog(0, "fe->i_clusters = %u, new_i_clusters = %u, fe->i_size ="
Mark Fashehb0697052006-03-03 10:24:33 -08003532 "%llu\n", fe->i_clusters, new_i_clusters,
3533 (unsigned long long)fe->i_size);
Mark Fashehccd979b2005-12-15 14:31:24 -08003534
Robert P. J. Daycd861282006-12-13 00:34:52 -08003535 *tc = kzalloc(sizeof(struct ocfs2_truncate_context), GFP_KERNEL);
Mark Fashehccd979b2005-12-15 14:31:24 -08003536 if (!(*tc)) {
3537 status = -ENOMEM;
3538 mlog_errno(status);
3539 goto bail;
3540 }
3541
3542 metadata_delete = 0;
3543 if (fe->id2.i_list.l_tree_depth) {
3544 /* If we have a tree, then the truncate may result in
3545 * metadata deletes. Figure this out from the
3546 * rightmost leaf block.*/
3547 status = ocfs2_read_block(osb, le64_to_cpu(fe->i_last_eb_blk),
3548 &last_eb_bh, OCFS2_BH_CACHED, inode);
3549 if (status < 0) {
3550 mlog_errno(status);
3551 goto bail;
3552 }
3553 eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
3554 if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
3555 OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
3556
3557 brelse(last_eb_bh);
3558 status = -EIO;
3559 goto bail;
3560 }
3561 el = &(eb->h_list);
Mark Fashehdcd05382007-01-16 11:32:23 -08003562
3563 i = 0;
3564 if (ocfs2_is_empty_extent(&el->l_recs[0]))
3565 i = 1;
3566 /*
3567 * XXX: Should we check that next_free_rec contains
3568 * the extent?
3569 */
3570 if (le32_to_cpu(el->l_recs[i].e_cpos) >= new_i_clusters)
Mark Fashehccd979b2005-12-15 14:31:24 -08003571 metadata_delete = 1;
3572 }
3573
3574 (*tc)->tc_last_eb_bh = last_eb_bh;
3575
3576 if (metadata_delete) {
3577 mlog(0, "Will have to delete metadata for this trunc. "
3578 "locking allocator.\n");
3579 ext_alloc_inode = ocfs2_get_system_file_inode(osb, EXTENT_ALLOC_SYSTEM_INODE, 0);
3580 if (!ext_alloc_inode) {
3581 status = -ENOMEM;
3582 mlog_errno(status);
3583 goto bail;
3584 }
3585
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08003586 mutex_lock(&ext_alloc_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08003587 (*tc)->tc_ext_alloc_inode = ext_alloc_inode;
3588
Mark Fasheh4bcec182006-10-09 16:02:40 -07003589 status = ocfs2_meta_lock(ext_alloc_inode, &ext_alloc_bh, 1);
Mark Fashehccd979b2005-12-15 14:31:24 -08003590 if (status < 0) {
3591 mlog_errno(status);
3592 goto bail;
3593 }
3594 (*tc)->tc_ext_alloc_bh = ext_alloc_bh;
3595 (*tc)->tc_ext_alloc_locked = 1;
3596 }
3597
3598 status = 0;
3599bail:
3600 if (status < 0) {
3601 if (*tc)
3602 ocfs2_free_truncate_context(*tc);
3603 *tc = NULL;
3604 }
3605 mlog_exit_void();
3606 return status;
3607}
3608
3609static void ocfs2_free_truncate_context(struct ocfs2_truncate_context *tc)
3610{
3611 if (tc->tc_ext_alloc_inode) {
3612 if (tc->tc_ext_alloc_locked)
3613 ocfs2_meta_unlock(tc->tc_ext_alloc_inode, 1);
3614
Jes Sorensen1b1dcc12006-01-09 15:59:24 -08003615 mutex_unlock(&tc->tc_ext_alloc_inode->i_mutex);
Mark Fashehccd979b2005-12-15 14:31:24 -08003616 iput(tc->tc_ext_alloc_inode);
3617 }
3618
3619 if (tc->tc_ext_alloc_bh)
3620 brelse(tc->tc_ext_alloc_bh);
3621
3622 if (tc->tc_last_eb_bh)
3623 brelse(tc->tc_last_eb_bh);
3624
3625 kfree(tc);
3626}