blob: f815cc81e7a287bd7198dc874d7954938c188464 [file] [log] [blame]
Alex Tomasa86c6182006-10-11 01:21:03 -07001/*
2 * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com
3 * Written by Alex Tomas <alex@clusterfs.com>
4 *
5 * Architecture independence:
6 * Copyright (c) 2005, Bull S.A.
7 * Written by Pierre Peiffer <pierre.peiffer@bull.net>
8 *
9 * This program is free software; you can redistribute it and/or modify
10 * it under the terms of the GNU General Public License version 2 as
11 * published by the Free Software Foundation.
12 *
13 * This program is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
17 *
18 * You should have received a copy of the GNU General Public Licens
19 * along with this program; if not, write to the Free Software
20 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-
21 */
22
23/*
24 * Extents support for EXT4
25 *
26 * TODO:
27 * - ext4*_error() should be used in some situations
28 * - analyze all BUG()/BUG_ON(), use -EIO where appropriate
29 * - smart tree reduction
30 */
31
32#include <linux/module.h>
33#include <linux/fs.h>
34#include <linux/time.h>
Mingming Caocd02ff02007-10-16 18:38:25 -040035#include <linux/jbd2.h>
Alex Tomasa86c6182006-10-11 01:21:03 -070036#include <linux/highuid.h>
37#include <linux/pagemap.h>
38#include <linux/quotaops.h>
39#include <linux/string.h>
40#include <linux/slab.h>
Amit Aroraa2df2a62007-07-17 21:42:41 -040041#include <linux/falloc.h>
Alex Tomasa86c6182006-10-11 01:21:03 -070042#include <asm/uaccess.h>
Eric Sandeen6873fa02008-10-07 00:46:36 -040043#include <linux/fiemap.h>
Christoph Hellwig3dcf5452008-04-29 18:13:32 -040044#include "ext4_jbd2.h"
45#include "ext4_extents.h"
Alex Tomasa86c6182006-10-11 01:21:03 -070046
Jiaying Zhang0562e0b2011-03-21 21:38:05 -040047#include <trace/events/ext4.h>
48
Allison Hendersond583fb82011-05-25 07:41:43 -040049static int ext4_split_extent(handle_t *handle,
50 struct inode *inode,
51 struct ext4_ext_path *path,
52 struct ext4_map_blocks *map,
53 int split_flag,
54 int flags);
55
Jan Kara487caee2009-08-17 22:17:20 -040056static int ext4_ext_truncate_extend_restart(handle_t *handle,
57 struct inode *inode,
58 int needed)
Alex Tomasa86c6182006-10-11 01:21:03 -070059{
60 int err;
61
Frank Mayhar03901312009-01-07 00:06:22 -050062 if (!ext4_handle_valid(handle))
63 return 0;
Alex Tomasa86c6182006-10-11 01:21:03 -070064 if (handle->h_buffer_credits > needed)
Shen Feng9102e4f2008-07-11 19:27:31 -040065 return 0;
66 err = ext4_journal_extend(handle, needed);
Theodore Ts'o0123c932008-08-01 20:57:54 -040067 if (err <= 0)
Shen Feng9102e4f2008-07-11 19:27:31 -040068 return err;
Jan Kara487caee2009-08-17 22:17:20 -040069 err = ext4_truncate_restart_trans(handle, inode, needed);
Dmitry Monakhov0617b832010-05-17 01:00:00 -040070 if (err == 0)
71 err = -EAGAIN;
Jan Kara487caee2009-08-17 22:17:20 -040072
73 return err;
Alex Tomasa86c6182006-10-11 01:21:03 -070074}
75
76/*
77 * could return:
78 * - EROFS
79 * - ENOMEM
80 */
81static int ext4_ext_get_access(handle_t *handle, struct inode *inode,
82 struct ext4_ext_path *path)
83{
84 if (path->p_bh) {
85 /* path points to block */
86 return ext4_journal_get_write_access(handle, path->p_bh);
87 }
88 /* path points to leaf/index in inode body */
89 /* we use in-core data, no need to protect them */
90 return 0;
91}
92
93/*
94 * could return:
95 * - EROFS
96 * - ENOMEM
97 * - EIO
98 */
99static int ext4_ext_dirty(handle_t *handle, struct inode *inode,
100 struct ext4_ext_path *path)
101{
102 int err;
103 if (path->p_bh) {
104 /* path points to block */
Frank Mayhar03901312009-01-07 00:06:22 -0500105 err = ext4_handle_dirty_metadata(handle, inode, path->p_bh);
Alex Tomasa86c6182006-10-11 01:21:03 -0700106 } else {
107 /* path points to leaf/index in inode body */
108 err = ext4_mark_inode_dirty(handle, inode);
109 }
110 return err;
111}
112
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700113static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
Alex Tomasa86c6182006-10-11 01:21:03 -0700114 struct ext4_ext_path *path,
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -0500115 ext4_lblk_t block)
Alex Tomasa86c6182006-10-11 01:21:03 -0700116{
117 struct ext4_inode_info *ei = EXT4_I(inode);
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700118 ext4_fsblk_t bg_start;
Valerie Clement74d34872008-02-15 13:43:07 -0500119 ext4_fsblk_t last_block;
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700120 ext4_grpblk_t colour;
Theodore Ts'oa4912122009-03-12 12:18:34 -0400121 ext4_group_t block_group;
122 int flex_size = ext4_flex_bg_size(EXT4_SB(inode->i_sb));
Alex Tomasa86c6182006-10-11 01:21:03 -0700123 int depth;
124
125 if (path) {
126 struct ext4_extent *ex;
127 depth = path->p_depth;
128
Kazuya Mioad4fb9c2011-01-10 12:12:28 -0500129 /*
130 * Try to predict block placement assuming that we are
131 * filling in a file which will eventually be
132 * non-sparse --- i.e., in the case of libbfd writing
133 * an ELF object sections out-of-order but in a way
134 * the eventually results in a contiguous object or
135 * executable file, or some database extending a table
136 * space file. However, this is actually somewhat
137 * non-ideal if we are writing a sparse file such as
138 * qemu or KVM writing a raw image file that is going
139 * to stay fairly sparse, since it will end up
140 * fragmenting the file system's free space. Maybe we
141 * should have some hueristics or some way to allow
142 * userspace to pass a hint to file system,
Tao Mab8d65682011-01-21 23:21:31 +0800143 * especially if the latter case turns out to be
Kazuya Mioad4fb9c2011-01-10 12:12:28 -0500144 * common.
145 */
Avantika Mathur7e028972006-12-06 20:41:33 -0800146 ex = path[depth].p_ext;
Kazuya Mioad4fb9c2011-01-10 12:12:28 -0500147 if (ex) {
148 ext4_fsblk_t ext_pblk = ext4_ext_pblock(ex);
149 ext4_lblk_t ext_block = le32_to_cpu(ex->ee_block);
150
151 if (block > ext_block)
152 return ext_pblk + (block - ext_block);
153 else
154 return ext_pblk - (ext_block - block);
155 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700156
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700157 /* it looks like index is empty;
158 * try to find starting block from index itself */
Alex Tomasa86c6182006-10-11 01:21:03 -0700159 if (path[depth].p_bh)
160 return path[depth].p_bh->b_blocknr;
161 }
162
163 /* OK. use inode's group */
Theodore Ts'oa4912122009-03-12 12:18:34 -0400164 block_group = ei->i_block_group;
165 if (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) {
166 /*
167 * If there are at least EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME
Theodore Ts'o60e66792010-05-17 07:00:00 -0400168 * block groups per flexgroup, reserve the first block
169 * group for directories and special files. Regular
Theodore Ts'oa4912122009-03-12 12:18:34 -0400170 * files will start at the second block group. This
Theodore Ts'o60e66792010-05-17 07:00:00 -0400171 * tends to speed up directory access and improves
Theodore Ts'oa4912122009-03-12 12:18:34 -0400172 * fsck times.
173 */
174 block_group &= ~(flex_size-1);
175 if (S_ISREG(inode->i_mode))
176 block_group++;
177 }
Akinobu Mita5661bd62010-03-03 23:53:39 -0500178 bg_start = ext4_group_first_block_no(inode->i_sb, block_group);
Valerie Clement74d34872008-02-15 13:43:07 -0500179 last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1;
180
Theodore Ts'oa4912122009-03-12 12:18:34 -0400181 /*
182 * If we are doing delayed allocation, we don't need take
183 * colour into account.
184 */
185 if (test_opt(inode->i_sb, DELALLOC))
186 return bg_start;
187
Valerie Clement74d34872008-02-15 13:43:07 -0500188 if (bg_start + EXT4_BLOCKS_PER_GROUP(inode->i_sb) <= last_block)
189 colour = (current->pid % 16) *
Alex Tomasa86c6182006-10-11 01:21:03 -0700190 (EXT4_BLOCKS_PER_GROUP(inode->i_sb) / 16);
Valerie Clement74d34872008-02-15 13:43:07 -0500191 else
192 colour = (current->pid % 16) * ((last_block - bg_start) / 16);
Alex Tomasa86c6182006-10-11 01:21:03 -0700193 return bg_start + colour + block;
194}
195
Aneesh Kumar K.V654b4902008-07-11 19:27:31 -0400196/*
197 * Allocation for a meta data block
198 */
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700199static ext4_fsblk_t
Aneesh Kumar K.V654b4902008-07-11 19:27:31 -0400200ext4_ext_new_meta_block(handle_t *handle, struct inode *inode,
Alex Tomasa86c6182006-10-11 01:21:03 -0700201 struct ext4_ext_path *path,
Allison Henderson55f020d2011-05-25 07:41:26 -0400202 struct ext4_extent *ex, int *err, unsigned int flags)
Alex Tomasa86c6182006-10-11 01:21:03 -0700203{
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700204 ext4_fsblk_t goal, newblock;
Alex Tomasa86c6182006-10-11 01:21:03 -0700205
206 goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block));
Allison Henderson55f020d2011-05-25 07:41:26 -0400207 newblock = ext4_new_meta_blocks(handle, inode, goal, flags,
208 NULL, err);
Alex Tomasa86c6182006-10-11 01:21:03 -0700209 return newblock;
210}
211
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400212static inline int ext4_ext_space_block(struct inode *inode, int check)
Alex Tomasa86c6182006-10-11 01:21:03 -0700213{
214 int size;
215
216 size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
217 / sizeof(struct ext4_extent);
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400218 if (!check) {
Robert P. J. Daybbf2f9f2007-02-17 19:20:16 +0100219#ifdef AGGRESSIVE_TEST
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400220 if (size > 6)
221 size = 6;
Alex Tomasa86c6182006-10-11 01:21:03 -0700222#endif
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400223 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700224 return size;
225}
226
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400227static inline int ext4_ext_space_block_idx(struct inode *inode, int check)
Alex Tomasa86c6182006-10-11 01:21:03 -0700228{
229 int size;
230
231 size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
232 / sizeof(struct ext4_extent_idx);
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400233 if (!check) {
Robert P. J. Daybbf2f9f2007-02-17 19:20:16 +0100234#ifdef AGGRESSIVE_TEST
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400235 if (size > 5)
236 size = 5;
Alex Tomasa86c6182006-10-11 01:21:03 -0700237#endif
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400238 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700239 return size;
240}
241
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400242static inline int ext4_ext_space_root(struct inode *inode, int check)
Alex Tomasa86c6182006-10-11 01:21:03 -0700243{
244 int size;
245
246 size = sizeof(EXT4_I(inode)->i_data);
247 size -= sizeof(struct ext4_extent_header);
248 size /= sizeof(struct ext4_extent);
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400249 if (!check) {
Robert P. J. Daybbf2f9f2007-02-17 19:20:16 +0100250#ifdef AGGRESSIVE_TEST
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400251 if (size > 3)
252 size = 3;
Alex Tomasa86c6182006-10-11 01:21:03 -0700253#endif
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400254 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700255 return size;
256}
257
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400258static inline int ext4_ext_space_root_idx(struct inode *inode, int check)
Alex Tomasa86c6182006-10-11 01:21:03 -0700259{
260 int size;
261
262 size = sizeof(EXT4_I(inode)->i_data);
263 size -= sizeof(struct ext4_extent_header);
264 size /= sizeof(struct ext4_extent_idx);
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400265 if (!check) {
Robert P. J. Daybbf2f9f2007-02-17 19:20:16 +0100266#ifdef AGGRESSIVE_TEST
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400267 if (size > 4)
268 size = 4;
Alex Tomasa86c6182006-10-11 01:21:03 -0700269#endif
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400270 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700271 return size;
272}
273
Mingming Caod2a17632008-07-14 17:52:37 -0400274/*
275 * Calculate the number of metadata blocks needed
276 * to allocate @blocks
277 * Worse case is one block per extent
278 */
Theodore Ts'o01f49d02011-01-10 12:13:03 -0500279int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock)
Mingming Caod2a17632008-07-14 17:52:37 -0400280{
Theodore Ts'o9d0be502010-01-01 02:41:30 -0500281 struct ext4_inode_info *ei = EXT4_I(inode);
282 int idxs, num = 0;
Mingming Caod2a17632008-07-14 17:52:37 -0400283
Theodore Ts'o9d0be502010-01-01 02:41:30 -0500284 idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
285 / sizeof(struct ext4_extent_idx));
Mingming Caod2a17632008-07-14 17:52:37 -0400286
287 /*
Theodore Ts'o9d0be502010-01-01 02:41:30 -0500288 * If the new delayed allocation block is contiguous with the
289 * previous da block, it can share index blocks with the
290 * previous block, so we only need to allocate a new index
291 * block every idxs leaf blocks. At ldxs**2 blocks, we need
292 * an additional index block, and at ldxs**3 blocks, yet
293 * another index blocks.
Mingming Caod2a17632008-07-14 17:52:37 -0400294 */
Theodore Ts'o9d0be502010-01-01 02:41:30 -0500295 if (ei->i_da_metadata_calc_len &&
296 ei->i_da_metadata_calc_last_lblock+1 == lblock) {
297 if ((ei->i_da_metadata_calc_len % idxs) == 0)
298 num++;
299 if ((ei->i_da_metadata_calc_len % (idxs*idxs)) == 0)
300 num++;
301 if ((ei->i_da_metadata_calc_len % (idxs*idxs*idxs)) == 0) {
302 num++;
303 ei->i_da_metadata_calc_len = 0;
304 } else
305 ei->i_da_metadata_calc_len++;
306 ei->i_da_metadata_calc_last_lblock++;
307 return num;
308 }
Mingming Caod2a17632008-07-14 17:52:37 -0400309
Theodore Ts'o9d0be502010-01-01 02:41:30 -0500310 /*
311 * In the worst case we need a new set of index blocks at
312 * every level of the inode's extent tree.
313 */
314 ei->i_da_metadata_calc_len = 1;
315 ei->i_da_metadata_calc_last_lblock = lblock;
316 return ext_depth(inode) + 1;
Mingming Caod2a17632008-07-14 17:52:37 -0400317}
318
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400319static int
320ext4_ext_max_entries(struct inode *inode, int depth)
321{
322 int max;
323
324 if (depth == ext_depth(inode)) {
325 if (depth == 0)
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400326 max = ext4_ext_space_root(inode, 1);
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400327 else
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400328 max = ext4_ext_space_root_idx(inode, 1);
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400329 } else {
330 if (depth == 0)
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400331 max = ext4_ext_space_block(inode, 1);
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400332 else
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400333 max = ext4_ext_space_block_idx(inode, 1);
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400334 }
335
336 return max;
337}
338
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -0400339static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext)
340{
Theodore Ts'obf89d162010-10-27 21:30:14 -0400341 ext4_fsblk_t block = ext4_ext_pblock(ext);
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -0400342 int len = ext4_ext_get_actual_len(ext);
Theodore Ts'oe84a26c2009-04-22 20:52:25 -0400343
Theodore Ts'o6fd058f2009-05-17 15:38:01 -0400344 return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len);
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -0400345}
346
347static int ext4_valid_extent_idx(struct inode *inode,
348 struct ext4_extent_idx *ext_idx)
349{
Theodore Ts'obf89d162010-10-27 21:30:14 -0400350 ext4_fsblk_t block = ext4_idx_pblock(ext_idx);
Theodore Ts'oe84a26c2009-04-22 20:52:25 -0400351
Theodore Ts'o6fd058f2009-05-17 15:38:01 -0400352 return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, 1);
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -0400353}
354
355static int ext4_valid_extent_entries(struct inode *inode,
356 struct ext4_extent_header *eh,
357 int depth)
358{
359 struct ext4_extent *ext;
360 struct ext4_extent_idx *ext_idx;
361 unsigned short entries;
362 if (eh->eh_entries == 0)
363 return 1;
364
365 entries = le16_to_cpu(eh->eh_entries);
366
367 if (depth == 0) {
368 /* leaf entries */
369 ext = EXT_FIRST_EXTENT(eh);
370 while (entries) {
371 if (!ext4_valid_extent(inode, ext))
372 return 0;
373 ext++;
374 entries--;
375 }
376 } else {
377 ext_idx = EXT_FIRST_INDEX(eh);
378 while (entries) {
379 if (!ext4_valid_extent_idx(inode, ext_idx))
380 return 0;
381 ext_idx++;
382 entries--;
383 }
384 }
385 return 1;
386}
387
Theodore Ts'oc398eda2010-07-27 11:56:40 -0400388static int __ext4_ext_check(const char *function, unsigned int line,
389 struct inode *inode, struct ext4_extent_header *eh,
390 int depth)
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400391{
392 const char *error_msg;
393 int max = 0;
394
395 if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) {
396 error_msg = "invalid magic";
397 goto corrupted;
398 }
399 if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) {
400 error_msg = "unexpected eh_depth";
401 goto corrupted;
402 }
403 if (unlikely(eh->eh_max == 0)) {
404 error_msg = "invalid eh_max";
405 goto corrupted;
406 }
407 max = ext4_ext_max_entries(inode, depth);
408 if (unlikely(le16_to_cpu(eh->eh_max) > max)) {
409 error_msg = "too large eh_max";
410 goto corrupted;
411 }
412 if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) {
413 error_msg = "invalid eh_entries";
414 goto corrupted;
415 }
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -0400416 if (!ext4_valid_extent_entries(inode, eh, depth)) {
417 error_msg = "invalid extent entries";
418 goto corrupted;
419 }
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400420 return 0;
421
422corrupted:
Theodore Ts'oc398eda2010-07-27 11:56:40 -0400423 ext4_error_inode(inode, function, line, 0,
Theodore Ts'o24676da2010-05-16 21:00:00 -0400424 "bad header/extent: %s - magic %x, "
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400425 "entries %u, max %u(%u), depth %u(%u)",
Theodore Ts'o24676da2010-05-16 21:00:00 -0400426 error_msg, le16_to_cpu(eh->eh_magic),
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400427 le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
428 max, le16_to_cpu(eh->eh_depth), depth);
429
430 return -EIO;
431}
432
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -0400433#define ext4_ext_check(inode, eh, depth) \
Theodore Ts'oc398eda2010-07-27 11:56:40 -0400434 __ext4_ext_check(__func__, __LINE__, inode, eh, depth)
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400435
Aneesh Kumar K.V7a262f72009-03-27 16:39:58 -0400436int ext4_ext_check_inode(struct inode *inode)
437{
438 return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode));
439}
440
Alex Tomasa86c6182006-10-11 01:21:03 -0700441#ifdef EXT_DEBUG
442static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path)
443{
444 int k, l = path->p_depth;
445
446 ext_debug("path:");
447 for (k = 0; k <= l; k++, path++) {
448 if (path->p_idx) {
Mingming Cao2ae02102006-10-11 01:21:11 -0700449 ext_debug(" %d->%llu", le32_to_cpu(path->p_idx->ei_block),
Theodore Ts'obf89d162010-10-27 21:30:14 -0400450 ext4_idx_pblock(path->p_idx));
Alex Tomasa86c6182006-10-11 01:21:03 -0700451 } else if (path->p_ext) {
Mingming553f9002009-09-18 13:34:55 -0400452 ext_debug(" %d:[%d]%d:%llu ",
Alex Tomasa86c6182006-10-11 01:21:03 -0700453 le32_to_cpu(path->p_ext->ee_block),
Mingming553f9002009-09-18 13:34:55 -0400454 ext4_ext_is_uninitialized(path->p_ext),
Amit Aroraa2df2a62007-07-17 21:42:41 -0400455 ext4_ext_get_actual_len(path->p_ext),
Theodore Ts'obf89d162010-10-27 21:30:14 -0400456 ext4_ext_pblock(path->p_ext));
Alex Tomasa86c6182006-10-11 01:21:03 -0700457 } else
458 ext_debug(" []");
459 }
460 ext_debug("\n");
461}
462
463static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path)
464{
465 int depth = ext_depth(inode);
466 struct ext4_extent_header *eh;
467 struct ext4_extent *ex;
468 int i;
469
470 if (!path)
471 return;
472
473 eh = path[depth].p_hdr;
474 ex = EXT_FIRST_EXTENT(eh);
475
Mingming553f9002009-09-18 13:34:55 -0400476 ext_debug("Displaying leaf extents for inode %lu\n", inode->i_ino);
477
Alex Tomasa86c6182006-10-11 01:21:03 -0700478 for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
Mingming553f9002009-09-18 13:34:55 -0400479 ext_debug("%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block),
480 ext4_ext_is_uninitialized(ex),
Theodore Ts'obf89d162010-10-27 21:30:14 -0400481 ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex));
Alex Tomasa86c6182006-10-11 01:21:03 -0700482 }
483 ext_debug("\n");
484}
Yongqiang Yang1b16da72011-05-25 17:41:48 -0400485
486static void ext4_ext_show_move(struct inode *inode, struct ext4_ext_path *path,
487 ext4_fsblk_t newblock, int level)
488{
489 int depth = ext_depth(inode);
490 struct ext4_extent *ex;
491
492 if (depth != level) {
493 struct ext4_extent_idx *idx;
494 idx = path[level].p_idx;
495 while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) {
496 ext_debug("%d: move %d:%llu in new index %llu\n", level,
497 le32_to_cpu(idx->ei_block),
498 ext4_idx_pblock(idx),
499 newblock);
500 idx++;
501 }
502
503 return;
504 }
505
506 ex = path[depth].p_ext;
507 while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) {
508 ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n",
509 le32_to_cpu(ex->ee_block),
510 ext4_ext_pblock(ex),
511 ext4_ext_is_uninitialized(ex),
512 ext4_ext_get_actual_len(ex),
513 newblock);
514 ex++;
515 }
516}
517
Alex Tomasa86c6182006-10-11 01:21:03 -0700518#else
Theodore Ts'oaf5bc922008-09-08 22:25:24 -0400519#define ext4_ext_show_path(inode, path)
520#define ext4_ext_show_leaf(inode, path)
Yongqiang Yang1b16da72011-05-25 17:41:48 -0400521#define ext4_ext_show_move(inode, path, newblock, level)
Alex Tomasa86c6182006-10-11 01:21:03 -0700522#endif
523
Aneesh Kumar K.Vb35905c2008-02-25 16:54:37 -0500524void ext4_ext_drop_refs(struct ext4_ext_path *path)
Alex Tomasa86c6182006-10-11 01:21:03 -0700525{
526 int depth = path->p_depth;
527 int i;
528
529 for (i = 0; i <= depth; i++, path++)
530 if (path->p_bh) {
531 brelse(path->p_bh);
532 path->p_bh = NULL;
533 }
534}
535
536/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700537 * ext4_ext_binsearch_idx:
538 * binary search for the closest index of the given block
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400539 * the header must be checked before calling this
Alex Tomasa86c6182006-10-11 01:21:03 -0700540 */
541static void
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -0500542ext4_ext_binsearch_idx(struct inode *inode,
543 struct ext4_ext_path *path, ext4_lblk_t block)
Alex Tomasa86c6182006-10-11 01:21:03 -0700544{
545 struct ext4_extent_header *eh = path->p_hdr;
546 struct ext4_extent_idx *r, *l, *m;
547
Alex Tomasa86c6182006-10-11 01:21:03 -0700548
Eric Sandeenbba90742008-01-28 23:58:27 -0500549 ext_debug("binsearch for %u(idx): ", block);
Alex Tomasa86c6182006-10-11 01:21:03 -0700550
551 l = EXT_FIRST_INDEX(eh) + 1;
Dmitry Monakhove9f410b2007-07-18 09:09:15 -0400552 r = EXT_LAST_INDEX(eh);
Alex Tomasa86c6182006-10-11 01:21:03 -0700553 while (l <= r) {
554 m = l + (r - l) / 2;
555 if (block < le32_to_cpu(m->ei_block))
556 r = m - 1;
557 else
558 l = m + 1;
Dmitry Monakhov26d535e2007-07-18 08:33:37 -0400559 ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ei_block),
560 m, le32_to_cpu(m->ei_block),
561 r, le32_to_cpu(r->ei_block));
Alex Tomasa86c6182006-10-11 01:21:03 -0700562 }
563
564 path->p_idx = l - 1;
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700565 ext_debug(" -> %d->%lld ", le32_to_cpu(path->p_idx->ei_block),
Theodore Ts'obf89d162010-10-27 21:30:14 -0400566 ext4_idx_pblock(path->p_idx));
Alex Tomasa86c6182006-10-11 01:21:03 -0700567
568#ifdef CHECK_BINSEARCH
569 {
570 struct ext4_extent_idx *chix, *ix;
571 int k;
572
573 chix = ix = EXT_FIRST_INDEX(eh);
574 for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) {
575 if (k != 0 &&
576 le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) {
Theodore Ts'o4776004f2008-09-08 23:00:52 -0400577 printk(KERN_DEBUG "k=%d, ix=0x%p, "
578 "first=0x%p\n", k,
579 ix, EXT_FIRST_INDEX(eh));
580 printk(KERN_DEBUG "%u <= %u\n",
Alex Tomasa86c6182006-10-11 01:21:03 -0700581 le32_to_cpu(ix->ei_block),
582 le32_to_cpu(ix[-1].ei_block));
583 }
584 BUG_ON(k && le32_to_cpu(ix->ei_block)
Dave Kleikamp8c55e202007-05-24 13:04:54 -0400585 <= le32_to_cpu(ix[-1].ei_block));
Alex Tomasa86c6182006-10-11 01:21:03 -0700586 if (block < le32_to_cpu(ix->ei_block))
587 break;
588 chix = ix;
589 }
590 BUG_ON(chix != path->p_idx);
591 }
592#endif
593
594}
595
596/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700597 * ext4_ext_binsearch:
598 * binary search for closest extent of the given block
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400599 * the header must be checked before calling this
Alex Tomasa86c6182006-10-11 01:21:03 -0700600 */
601static void
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -0500602ext4_ext_binsearch(struct inode *inode,
603 struct ext4_ext_path *path, ext4_lblk_t block)
Alex Tomasa86c6182006-10-11 01:21:03 -0700604{
605 struct ext4_extent_header *eh = path->p_hdr;
606 struct ext4_extent *r, *l, *m;
607
Alex Tomasa86c6182006-10-11 01:21:03 -0700608 if (eh->eh_entries == 0) {
609 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700610 * this leaf is empty:
611 * we get such a leaf in split/add case
Alex Tomasa86c6182006-10-11 01:21:03 -0700612 */
613 return;
614 }
615
Eric Sandeenbba90742008-01-28 23:58:27 -0500616 ext_debug("binsearch for %u: ", block);
Alex Tomasa86c6182006-10-11 01:21:03 -0700617
618 l = EXT_FIRST_EXTENT(eh) + 1;
Dmitry Monakhove9f410b2007-07-18 09:09:15 -0400619 r = EXT_LAST_EXTENT(eh);
Alex Tomasa86c6182006-10-11 01:21:03 -0700620
621 while (l <= r) {
622 m = l + (r - l) / 2;
623 if (block < le32_to_cpu(m->ee_block))
624 r = m - 1;
625 else
626 l = m + 1;
Dmitry Monakhov26d535e2007-07-18 08:33:37 -0400627 ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ee_block),
628 m, le32_to_cpu(m->ee_block),
629 r, le32_to_cpu(r->ee_block));
Alex Tomasa86c6182006-10-11 01:21:03 -0700630 }
631
632 path->p_ext = l - 1;
Mingming553f9002009-09-18 13:34:55 -0400633 ext_debug(" -> %d:%llu:[%d]%d ",
Dave Kleikamp8c55e202007-05-24 13:04:54 -0400634 le32_to_cpu(path->p_ext->ee_block),
Theodore Ts'obf89d162010-10-27 21:30:14 -0400635 ext4_ext_pblock(path->p_ext),
Mingming553f9002009-09-18 13:34:55 -0400636 ext4_ext_is_uninitialized(path->p_ext),
Amit Aroraa2df2a62007-07-17 21:42:41 -0400637 ext4_ext_get_actual_len(path->p_ext));
Alex Tomasa86c6182006-10-11 01:21:03 -0700638
639#ifdef CHECK_BINSEARCH
640 {
641 struct ext4_extent *chex, *ex;
642 int k;
643
644 chex = ex = EXT_FIRST_EXTENT(eh);
645 for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) {
646 BUG_ON(k && le32_to_cpu(ex->ee_block)
Dave Kleikamp8c55e202007-05-24 13:04:54 -0400647 <= le32_to_cpu(ex[-1].ee_block));
Alex Tomasa86c6182006-10-11 01:21:03 -0700648 if (block < le32_to_cpu(ex->ee_block))
649 break;
650 chex = ex;
651 }
652 BUG_ON(chex != path->p_ext);
653 }
654#endif
655
656}
657
658int ext4_ext_tree_init(handle_t *handle, struct inode *inode)
659{
660 struct ext4_extent_header *eh;
661
662 eh = ext_inode_hdr(inode);
663 eh->eh_depth = 0;
664 eh->eh_entries = 0;
665 eh->eh_magic = EXT4_EXT_MAGIC;
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400666 eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0));
Alex Tomasa86c6182006-10-11 01:21:03 -0700667 ext4_mark_inode_dirty(handle, inode);
668 ext4_ext_invalidate_cache(inode);
669 return 0;
670}
671
672struct ext4_ext_path *
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -0500673ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
674 struct ext4_ext_path *path)
Alex Tomasa86c6182006-10-11 01:21:03 -0700675{
676 struct ext4_extent_header *eh;
677 struct buffer_head *bh;
678 short int depth, i, ppos = 0, alloc = 0;
679
680 eh = ext_inode_hdr(inode);
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400681 depth = ext_depth(inode);
Alex Tomasa86c6182006-10-11 01:21:03 -0700682
683 /* account possible depth increase */
684 if (!path) {
Avantika Mathur5d4958f2006-12-06 20:41:35 -0800685 path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 2),
Alex Tomasa86c6182006-10-11 01:21:03 -0700686 GFP_NOFS);
687 if (!path)
688 return ERR_PTR(-ENOMEM);
689 alloc = 1;
690 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700691 path[0].p_hdr = eh;
Shen Feng1973adc2008-07-11 19:27:31 -0400692 path[0].p_bh = NULL;
Alex Tomasa86c6182006-10-11 01:21:03 -0700693
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400694 i = depth;
Alex Tomasa86c6182006-10-11 01:21:03 -0700695 /* walk through the tree */
696 while (i) {
Aneesh Kumar K.V7a262f72009-03-27 16:39:58 -0400697 int need_to_validate = 0;
698
Alex Tomasa86c6182006-10-11 01:21:03 -0700699 ext_debug("depth %d: num %d, max %d\n",
700 ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
Alex Tomasc29c0ae2007-07-18 09:19:09 -0400701
Alex Tomasa86c6182006-10-11 01:21:03 -0700702 ext4_ext_binsearch_idx(inode, path + ppos, block);
Theodore Ts'obf89d162010-10-27 21:30:14 -0400703 path[ppos].p_block = ext4_idx_pblock(path[ppos].p_idx);
Alex Tomasa86c6182006-10-11 01:21:03 -0700704 path[ppos].p_depth = i;
705 path[ppos].p_ext = NULL;
706
Aneesh Kumar K.V7a262f72009-03-27 16:39:58 -0400707 bh = sb_getblk(inode->i_sb, path[ppos].p_block);
708 if (unlikely(!bh))
Alex Tomasa86c6182006-10-11 01:21:03 -0700709 goto err;
Aneesh Kumar K.V7a262f72009-03-27 16:39:58 -0400710 if (!bh_uptodate_or_lock(bh)) {
Jiaying Zhang0562e0b2011-03-21 21:38:05 -0400711 trace_ext4_ext_load_extent(inode, block,
712 path[ppos].p_block);
Aneesh Kumar K.V7a262f72009-03-27 16:39:58 -0400713 if (bh_submit_read(bh) < 0) {
714 put_bh(bh);
715 goto err;
716 }
717 /* validate the extent entries */
718 need_to_validate = 1;
719 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700720 eh = ext_block_hdr(bh);
721 ppos++;
Frank Mayhar273df552010-03-02 11:46:09 -0500722 if (unlikely(ppos > depth)) {
723 put_bh(bh);
724 EXT4_ERROR_INODE(inode,
725 "ppos %d > depth %d", ppos, depth);
726 goto err;
727 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700728 path[ppos].p_bh = bh;
729 path[ppos].p_hdr = eh;
730 i--;
731
Aneesh Kumar K.V7a262f72009-03-27 16:39:58 -0400732 if (need_to_validate && ext4_ext_check(inode, eh, i))
Alex Tomasa86c6182006-10-11 01:21:03 -0700733 goto err;
734 }
735
736 path[ppos].p_depth = i;
Alex Tomasa86c6182006-10-11 01:21:03 -0700737 path[ppos].p_ext = NULL;
738 path[ppos].p_idx = NULL;
739
Alex Tomasa86c6182006-10-11 01:21:03 -0700740 /* find extent */
741 ext4_ext_binsearch(inode, path + ppos, block);
Shen Feng1973adc2008-07-11 19:27:31 -0400742 /* if not an empty leaf */
743 if (path[ppos].p_ext)
Theodore Ts'obf89d162010-10-27 21:30:14 -0400744 path[ppos].p_block = ext4_ext_pblock(path[ppos].p_ext);
Alex Tomasa86c6182006-10-11 01:21:03 -0700745
746 ext4_ext_show_path(inode, path);
747
748 return path;
749
750err:
751 ext4_ext_drop_refs(path);
752 if (alloc)
753 kfree(path);
754 return ERR_PTR(-EIO);
755}
756
757/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700758 * ext4_ext_insert_index:
759 * insert new index [@logical;@ptr] into the block at @curp;
760 * check where to insert: before @curp or after @curp
Alex Tomasa86c6182006-10-11 01:21:03 -0700761 */
Theodore Ts'o1f109d52010-10-27 21:30:14 -0400762static int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
763 struct ext4_ext_path *curp,
764 int logical, ext4_fsblk_t ptr)
Alex Tomasa86c6182006-10-11 01:21:03 -0700765{
766 struct ext4_extent_idx *ix;
767 int len, err;
768
Avantika Mathur7e028972006-12-06 20:41:33 -0800769 err = ext4_ext_get_access(handle, inode, curp);
770 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -0700771 return err;
772
Frank Mayhar273df552010-03-02 11:46:09 -0500773 if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) {
774 EXT4_ERROR_INODE(inode,
775 "logical %d == ei_block %d!",
776 logical, le32_to_cpu(curp->p_idx->ei_block));
777 return -EIO;
778 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700779 len = EXT_MAX_INDEX(curp->p_hdr) - curp->p_idx;
780 if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
781 /* insert after */
782 if (curp->p_idx != EXT_LAST_INDEX(curp->p_hdr)) {
783 len = (len - 1) * sizeof(struct ext4_extent_idx);
784 len = len < 0 ? 0 : len;
Dmitry Monakhov26d535e2007-07-18 08:33:37 -0400785 ext_debug("insert new index %d after: %llu. "
Alex Tomasa86c6182006-10-11 01:21:03 -0700786 "move %d from 0x%p to 0x%p\n",
787 logical, ptr, len,
788 (curp->p_idx + 1), (curp->p_idx + 2));
789 memmove(curp->p_idx + 2, curp->p_idx + 1, len);
790 }
791 ix = curp->p_idx + 1;
792 } else {
793 /* insert before */
794 len = len * sizeof(struct ext4_extent_idx);
795 len = len < 0 ? 0 : len;
Dmitry Monakhov26d535e2007-07-18 08:33:37 -0400796 ext_debug("insert new index %d before: %llu. "
Alex Tomasa86c6182006-10-11 01:21:03 -0700797 "move %d from 0x%p to 0x%p\n",
798 logical, ptr, len,
799 curp->p_idx, (curp->p_idx + 1));
800 memmove(curp->p_idx + 1, curp->p_idx, len);
801 ix = curp->p_idx;
802 }
803
804 ix->ei_block = cpu_to_le32(logical);
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700805 ext4_idx_store_pblock(ix, ptr);
Marcin Slusarze8546d02008-04-17 10:38:59 -0400806 le16_add_cpu(&curp->p_hdr->eh_entries, 1);
Alex Tomasa86c6182006-10-11 01:21:03 -0700807
Frank Mayhar273df552010-03-02 11:46:09 -0500808 if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries)
809 > le16_to_cpu(curp->p_hdr->eh_max))) {
810 EXT4_ERROR_INODE(inode,
811 "logical %d == ei_block %d!",
812 logical, le32_to_cpu(curp->p_idx->ei_block));
813 return -EIO;
814 }
815 if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) {
816 EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!");
817 return -EIO;
818 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700819
820 err = ext4_ext_dirty(handle, inode, curp);
821 ext4_std_error(inode->i_sb, err);
822
823 return err;
824}
825
826/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700827 * ext4_ext_split:
828 * inserts new subtree into the path, using free index entry
829 * at depth @at:
830 * - allocates all needed blocks (new leaf and all intermediate index blocks)
831 * - makes decision where to split
832 * - moves remaining extents and index entries (right to the split point)
833 * into the newly allocated blocks
834 * - initializes subtree
Alex Tomasa86c6182006-10-11 01:21:03 -0700835 */
836static int ext4_ext_split(handle_t *handle, struct inode *inode,
Allison Henderson55f020d2011-05-25 07:41:26 -0400837 unsigned int flags,
838 struct ext4_ext_path *path,
839 struct ext4_extent *newext, int at)
Alex Tomasa86c6182006-10-11 01:21:03 -0700840{
841 struct buffer_head *bh = NULL;
842 int depth = ext_depth(inode);
843 struct ext4_extent_header *neh;
844 struct ext4_extent_idx *fidx;
Alex Tomasa86c6182006-10-11 01:21:03 -0700845 int i = at, k, m, a;
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700846 ext4_fsblk_t newblock, oldblock;
Alex Tomasa86c6182006-10-11 01:21:03 -0700847 __le32 border;
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700848 ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */
Alex Tomasa86c6182006-10-11 01:21:03 -0700849 int err = 0;
850
851 /* make decision: where to split? */
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700852 /* FIXME: now decision is simplest: at current extent */
Alex Tomasa86c6182006-10-11 01:21:03 -0700853
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700854 /* if current leaf will be split, then we should use
Alex Tomasa86c6182006-10-11 01:21:03 -0700855 * border from split point */
Frank Mayhar273df552010-03-02 11:46:09 -0500856 if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) {
857 EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!");
858 return -EIO;
859 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700860 if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
861 border = path[depth].p_ext[1].ee_block;
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700862 ext_debug("leaf will be split."
Alex Tomasa86c6182006-10-11 01:21:03 -0700863 " next leaf starts at %d\n",
Dave Kleikamp8c55e202007-05-24 13:04:54 -0400864 le32_to_cpu(border));
Alex Tomasa86c6182006-10-11 01:21:03 -0700865 } else {
866 border = newext->ee_block;
867 ext_debug("leaf will be added."
868 " next leaf starts at %d\n",
Dave Kleikamp8c55e202007-05-24 13:04:54 -0400869 le32_to_cpu(border));
Alex Tomasa86c6182006-10-11 01:21:03 -0700870 }
871
872 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700873 * If error occurs, then we break processing
874 * and mark filesystem read-only. index won't
Alex Tomasa86c6182006-10-11 01:21:03 -0700875 * be inserted and tree will be in consistent
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700876 * state. Next mount will repair buffers too.
Alex Tomasa86c6182006-10-11 01:21:03 -0700877 */
878
879 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700880 * Get array to track all allocated blocks.
881 * We need this to handle errors and free blocks
882 * upon them.
Alex Tomasa86c6182006-10-11 01:21:03 -0700883 */
Avantika Mathur5d4958f2006-12-06 20:41:35 -0800884 ablocks = kzalloc(sizeof(ext4_fsblk_t) * depth, GFP_NOFS);
Alex Tomasa86c6182006-10-11 01:21:03 -0700885 if (!ablocks)
886 return -ENOMEM;
Alex Tomasa86c6182006-10-11 01:21:03 -0700887
888 /* allocate all needed blocks */
889 ext_debug("allocate %d blocks for indexes/leaf\n", depth - at);
890 for (a = 0; a < depth - at; a++) {
Aneesh Kumar K.V654b4902008-07-11 19:27:31 -0400891 newblock = ext4_ext_new_meta_block(handle, inode, path,
Allison Henderson55f020d2011-05-25 07:41:26 -0400892 newext, &err, flags);
Alex Tomasa86c6182006-10-11 01:21:03 -0700893 if (newblock == 0)
894 goto cleanup;
895 ablocks[a] = newblock;
896 }
897
898 /* initialize new leaf */
899 newblock = ablocks[--a];
Frank Mayhar273df552010-03-02 11:46:09 -0500900 if (unlikely(newblock == 0)) {
901 EXT4_ERROR_INODE(inode, "newblock == 0!");
902 err = -EIO;
903 goto cleanup;
904 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700905 bh = sb_getblk(inode->i_sb, newblock);
906 if (!bh) {
907 err = -EIO;
908 goto cleanup;
909 }
910 lock_buffer(bh);
911
Avantika Mathur7e028972006-12-06 20:41:33 -0800912 err = ext4_journal_get_create_access(handle, bh);
913 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -0700914 goto cleanup;
915
916 neh = ext_block_hdr(bh);
917 neh->eh_entries = 0;
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400918 neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
Alex Tomasa86c6182006-10-11 01:21:03 -0700919 neh->eh_magic = EXT4_EXT_MAGIC;
920 neh->eh_depth = 0;
Alex Tomasa86c6182006-10-11 01:21:03 -0700921
Randy Dunlapd0d856e2006-10-11 01:21:07 -0700922 /* move remainder of path[depth] to the new leaf */
Frank Mayhar273df552010-03-02 11:46:09 -0500923 if (unlikely(path[depth].p_hdr->eh_entries !=
924 path[depth].p_hdr->eh_max)) {
925 EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!",
926 path[depth].p_hdr->eh_entries,
927 path[depth].p_hdr->eh_max);
928 err = -EIO;
929 goto cleanup;
930 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700931 /* start copy from next extent */
Yongqiang Yang1b16da72011-05-25 17:41:48 -0400932 m = EXT_MAX_EXTENT(path[depth].p_hdr) - path[depth].p_ext++;
933 ext4_ext_show_move(inode, path, newblock, depth);
Alex Tomasa86c6182006-10-11 01:21:03 -0700934 if (m) {
Yongqiang Yang1b16da72011-05-25 17:41:48 -0400935 struct ext4_extent *ex;
936 ex = EXT_FIRST_EXTENT(neh);
937 memmove(ex, path[depth].p_ext, sizeof(struct ext4_extent) * m);
Marcin Slusarze8546d02008-04-17 10:38:59 -0400938 le16_add_cpu(&neh->eh_entries, m);
Alex Tomasa86c6182006-10-11 01:21:03 -0700939 }
940
941 set_buffer_uptodate(bh);
942 unlock_buffer(bh);
943
Frank Mayhar03901312009-01-07 00:06:22 -0500944 err = ext4_handle_dirty_metadata(handle, inode, bh);
Avantika Mathur7e028972006-12-06 20:41:33 -0800945 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -0700946 goto cleanup;
947 brelse(bh);
948 bh = NULL;
949
950 /* correct old leaf */
951 if (m) {
Avantika Mathur7e028972006-12-06 20:41:33 -0800952 err = ext4_ext_get_access(handle, inode, path + depth);
953 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -0700954 goto cleanup;
Marcin Slusarze8546d02008-04-17 10:38:59 -0400955 le16_add_cpu(&path[depth].p_hdr->eh_entries, -m);
Avantika Mathur7e028972006-12-06 20:41:33 -0800956 err = ext4_ext_dirty(handle, inode, path + depth);
957 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -0700958 goto cleanup;
959
960 }
961
962 /* create intermediate indexes */
963 k = depth - at - 1;
Frank Mayhar273df552010-03-02 11:46:09 -0500964 if (unlikely(k < 0)) {
965 EXT4_ERROR_INODE(inode, "k %d < 0!", k);
966 err = -EIO;
967 goto cleanup;
968 }
Alex Tomasa86c6182006-10-11 01:21:03 -0700969 if (k)
970 ext_debug("create %d intermediate indices\n", k);
971 /* insert new index into current index block */
972 /* current depth stored in i var */
973 i = depth - 1;
974 while (k--) {
975 oldblock = newblock;
976 newblock = ablocks[--a];
Eric Sandeenbba90742008-01-28 23:58:27 -0500977 bh = sb_getblk(inode->i_sb, newblock);
Alex Tomasa86c6182006-10-11 01:21:03 -0700978 if (!bh) {
979 err = -EIO;
980 goto cleanup;
981 }
982 lock_buffer(bh);
983
Avantika Mathur7e028972006-12-06 20:41:33 -0800984 err = ext4_journal_get_create_access(handle, bh);
985 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -0700986 goto cleanup;
987
988 neh = ext_block_hdr(bh);
989 neh->eh_entries = cpu_to_le16(1);
990 neh->eh_magic = EXT4_EXT_MAGIC;
Theodore Ts'o55ad63b2009-08-28 10:40:33 -0400991 neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
Alex Tomasa86c6182006-10-11 01:21:03 -0700992 neh->eh_depth = cpu_to_le16(depth - i);
993 fidx = EXT_FIRST_INDEX(neh);
994 fidx->ei_block = border;
Alex Tomasf65e6fb2006-10-11 01:21:05 -0700995 ext4_idx_store_pblock(fidx, oldblock);
Alex Tomasa86c6182006-10-11 01:21:03 -0700996
Eric Sandeenbba90742008-01-28 23:58:27 -0500997 ext_debug("int.index at %d (block %llu): %u -> %llu\n",
998 i, newblock, le32_to_cpu(border), oldblock);
Alex Tomasa86c6182006-10-11 01:21:03 -0700999
Yongqiang Yang1b16da72011-05-25 17:41:48 -04001000 /* move remainder of path[i] to the new index block */
Frank Mayhar273df552010-03-02 11:46:09 -05001001 if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) !=
1002 EXT_LAST_INDEX(path[i].p_hdr))) {
1003 EXT4_ERROR_INODE(inode,
1004 "EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!",
1005 le32_to_cpu(path[i].p_ext->ee_block));
1006 err = -EIO;
1007 goto cleanup;
1008 }
Yongqiang Yang1b16da72011-05-25 17:41:48 -04001009 /* start copy indexes */
1010 m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++;
1011 ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
1012 EXT_MAX_INDEX(path[i].p_hdr));
1013 ext4_ext_show_move(inode, path, newblock, i);
Alex Tomasa86c6182006-10-11 01:21:03 -07001014 if (m) {
Yongqiang Yang1b16da72011-05-25 17:41:48 -04001015 memmove(++fidx, path[i].p_idx,
Alex Tomasa86c6182006-10-11 01:21:03 -07001016 sizeof(struct ext4_extent_idx) * m);
Marcin Slusarze8546d02008-04-17 10:38:59 -04001017 le16_add_cpu(&neh->eh_entries, m);
Alex Tomasa86c6182006-10-11 01:21:03 -07001018 }
1019 set_buffer_uptodate(bh);
1020 unlock_buffer(bh);
1021
Frank Mayhar03901312009-01-07 00:06:22 -05001022 err = ext4_handle_dirty_metadata(handle, inode, bh);
Avantika Mathur7e028972006-12-06 20:41:33 -08001023 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001024 goto cleanup;
1025 brelse(bh);
1026 bh = NULL;
1027
1028 /* correct old index */
1029 if (m) {
1030 err = ext4_ext_get_access(handle, inode, path + i);
1031 if (err)
1032 goto cleanup;
Marcin Slusarze8546d02008-04-17 10:38:59 -04001033 le16_add_cpu(&path[i].p_hdr->eh_entries, -m);
Alex Tomasa86c6182006-10-11 01:21:03 -07001034 err = ext4_ext_dirty(handle, inode, path + i);
1035 if (err)
1036 goto cleanup;
1037 }
1038
1039 i--;
1040 }
1041
1042 /* insert new index */
Alex Tomasa86c6182006-10-11 01:21:03 -07001043 err = ext4_ext_insert_index(handle, inode, path + at,
1044 le32_to_cpu(border), newblock);
1045
1046cleanup:
1047 if (bh) {
1048 if (buffer_locked(bh))
1049 unlock_buffer(bh);
1050 brelse(bh);
1051 }
1052
1053 if (err) {
1054 /* free all allocated blocks in error case */
1055 for (i = 0; i < depth; i++) {
1056 if (!ablocks[i])
1057 continue;
Peter Huewe7dc57612011-02-21 21:01:42 -05001058 ext4_free_blocks(handle, inode, NULL, ablocks[i], 1,
Theodore Ts'oe6362602009-11-23 07:17:05 -05001059 EXT4_FREE_BLOCKS_METADATA);
Alex Tomasa86c6182006-10-11 01:21:03 -07001060 }
1061 }
1062 kfree(ablocks);
1063
1064 return err;
1065}
1066
1067/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001068 * ext4_ext_grow_indepth:
1069 * implements tree growing procedure:
1070 * - allocates new block
1071 * - moves top-level data (index block or leaf) into the new block
1072 * - initializes new top-level, creating index that points to the
1073 * just created block
Alex Tomasa86c6182006-10-11 01:21:03 -07001074 */
1075static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
Allison Henderson55f020d2011-05-25 07:41:26 -04001076 unsigned int flags,
1077 struct ext4_ext_path *path,
1078 struct ext4_extent *newext)
Alex Tomasa86c6182006-10-11 01:21:03 -07001079{
1080 struct ext4_ext_path *curp = path;
1081 struct ext4_extent_header *neh;
Alex Tomasa86c6182006-10-11 01:21:03 -07001082 struct buffer_head *bh;
Alex Tomasf65e6fb2006-10-11 01:21:05 -07001083 ext4_fsblk_t newblock;
Alex Tomasa86c6182006-10-11 01:21:03 -07001084 int err = 0;
1085
Allison Henderson55f020d2011-05-25 07:41:26 -04001086 newblock = ext4_ext_new_meta_block(handle, inode, path,
1087 newext, &err, flags);
Alex Tomasa86c6182006-10-11 01:21:03 -07001088 if (newblock == 0)
1089 return err;
1090
1091 bh = sb_getblk(inode->i_sb, newblock);
1092 if (!bh) {
1093 err = -EIO;
1094 ext4_std_error(inode->i_sb, err);
1095 return err;
1096 }
1097 lock_buffer(bh);
1098
Avantika Mathur7e028972006-12-06 20:41:33 -08001099 err = ext4_journal_get_create_access(handle, bh);
1100 if (err) {
Alex Tomasa86c6182006-10-11 01:21:03 -07001101 unlock_buffer(bh);
1102 goto out;
1103 }
1104
1105 /* move top-level index/leaf into new block */
1106 memmove(bh->b_data, curp->p_hdr, sizeof(EXT4_I(inode)->i_data));
1107
1108 /* set size of new block */
1109 neh = ext_block_hdr(bh);
1110 /* old root could have indexes or leaves
1111 * so calculate e_max right way */
1112 if (ext_depth(inode))
Theodore Ts'o55ad63b2009-08-28 10:40:33 -04001113 neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
Alex Tomasa86c6182006-10-11 01:21:03 -07001114 else
Theodore Ts'o55ad63b2009-08-28 10:40:33 -04001115 neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
Alex Tomasa86c6182006-10-11 01:21:03 -07001116 neh->eh_magic = EXT4_EXT_MAGIC;
1117 set_buffer_uptodate(bh);
1118 unlock_buffer(bh);
1119
Frank Mayhar03901312009-01-07 00:06:22 -05001120 err = ext4_handle_dirty_metadata(handle, inode, bh);
Avantika Mathur7e028972006-12-06 20:41:33 -08001121 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001122 goto out;
1123
1124 /* create index in new top-level index: num,max,pointer */
Avantika Mathur7e028972006-12-06 20:41:33 -08001125 err = ext4_ext_get_access(handle, inode, curp);
1126 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001127 goto out;
1128
1129 curp->p_hdr->eh_magic = EXT4_EXT_MAGIC;
Theodore Ts'o55ad63b2009-08-28 10:40:33 -04001130 curp->p_hdr->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode, 0));
Alex Tomasa86c6182006-10-11 01:21:03 -07001131 curp->p_hdr->eh_entries = cpu_to_le16(1);
1132 curp->p_idx = EXT_FIRST_INDEX(curp->p_hdr);
Dmitry Monakhove9f410b2007-07-18 09:09:15 -04001133
1134 if (path[0].p_hdr->eh_depth)
1135 curp->p_idx->ei_block =
1136 EXT_FIRST_INDEX(path[0].p_hdr)->ei_block;
1137 else
1138 curp->p_idx->ei_block =
1139 EXT_FIRST_EXTENT(path[0].p_hdr)->ee_block;
Alex Tomasf65e6fb2006-10-11 01:21:05 -07001140 ext4_idx_store_pblock(curp->p_idx, newblock);
Alex Tomasa86c6182006-10-11 01:21:03 -07001141
1142 neh = ext_inode_hdr(inode);
Mingming Cao2ae02102006-10-11 01:21:11 -07001143 ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
Alex Tomasa86c6182006-10-11 01:21:03 -07001144 le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
Andi Kleen5a0790c2010-06-14 13:28:03 -04001145 le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
Theodore Ts'obf89d162010-10-27 21:30:14 -04001146 ext4_idx_pblock(EXT_FIRST_INDEX(neh)));
Alex Tomasa86c6182006-10-11 01:21:03 -07001147
1148 neh->eh_depth = cpu_to_le16(path->p_depth + 1);
1149 err = ext4_ext_dirty(handle, inode, curp);
1150out:
1151 brelse(bh);
1152
1153 return err;
1154}
1155
1156/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001157 * ext4_ext_create_new_leaf:
1158 * finds empty index and adds new leaf.
1159 * if no free index is found, then it requests in-depth growing.
Alex Tomasa86c6182006-10-11 01:21:03 -07001160 */
1161static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
Allison Henderson55f020d2011-05-25 07:41:26 -04001162 unsigned int flags,
1163 struct ext4_ext_path *path,
1164 struct ext4_extent *newext)
Alex Tomasa86c6182006-10-11 01:21:03 -07001165{
1166 struct ext4_ext_path *curp;
1167 int depth, i, err = 0;
1168
1169repeat:
1170 i = depth = ext_depth(inode);
1171
1172 /* walk up to the tree and look for free index entry */
1173 curp = path + depth;
1174 while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) {
1175 i--;
1176 curp--;
1177 }
1178
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001179 /* we use already allocated block for index block,
1180 * so subsequent data blocks should be contiguous */
Alex Tomasa86c6182006-10-11 01:21:03 -07001181 if (EXT_HAS_FREE_INDEX(curp)) {
1182 /* if we found index with free entry, then use that
1183 * entry: create all needed subtree and add new leaf */
Allison Henderson55f020d2011-05-25 07:41:26 -04001184 err = ext4_ext_split(handle, inode, flags, path, newext, i);
Shen Feng787e0982008-07-11 19:27:31 -04001185 if (err)
1186 goto out;
Alex Tomasa86c6182006-10-11 01:21:03 -07001187
1188 /* refill path */
1189 ext4_ext_drop_refs(path);
1190 path = ext4_ext_find_extent(inode,
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001191 (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1192 path);
Alex Tomasa86c6182006-10-11 01:21:03 -07001193 if (IS_ERR(path))
1194 err = PTR_ERR(path);
1195 } else {
1196 /* tree is full, time to grow in depth */
Allison Henderson55f020d2011-05-25 07:41:26 -04001197 err = ext4_ext_grow_indepth(handle, inode, flags,
1198 path, newext);
Alex Tomasa86c6182006-10-11 01:21:03 -07001199 if (err)
1200 goto out;
1201
1202 /* refill path */
1203 ext4_ext_drop_refs(path);
1204 path = ext4_ext_find_extent(inode,
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001205 (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1206 path);
Alex Tomasa86c6182006-10-11 01:21:03 -07001207 if (IS_ERR(path)) {
1208 err = PTR_ERR(path);
1209 goto out;
1210 }
1211
1212 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001213 * only first (depth 0 -> 1) produces free space;
1214 * in all other cases we have to split the grown tree
Alex Tomasa86c6182006-10-11 01:21:03 -07001215 */
1216 depth = ext_depth(inode);
1217 if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) {
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001218 /* now we need to split */
Alex Tomasa86c6182006-10-11 01:21:03 -07001219 goto repeat;
1220 }
1221 }
1222
1223out:
1224 return err;
1225}
1226
1227/*
Alex Tomas1988b512008-01-28 23:58:27 -05001228 * search the closest allocated block to the left for *logical
1229 * and returns it at @logical + it's physical address at @phys
1230 * if *logical is the smallest allocated block, the function
1231 * returns 0 at @phys
1232 * return value contains 0 (success) or error code
1233 */
Theodore Ts'o1f109d52010-10-27 21:30:14 -04001234static int ext4_ext_search_left(struct inode *inode,
1235 struct ext4_ext_path *path,
1236 ext4_lblk_t *logical, ext4_fsblk_t *phys)
Alex Tomas1988b512008-01-28 23:58:27 -05001237{
1238 struct ext4_extent_idx *ix;
1239 struct ext4_extent *ex;
Aneesh Kumar K.Vb939e372008-01-28 23:58:27 -05001240 int depth, ee_len;
Alex Tomas1988b512008-01-28 23:58:27 -05001241
Frank Mayhar273df552010-03-02 11:46:09 -05001242 if (unlikely(path == NULL)) {
1243 EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
1244 return -EIO;
1245 }
Alex Tomas1988b512008-01-28 23:58:27 -05001246 depth = path->p_depth;
1247 *phys = 0;
1248
1249 if (depth == 0 && path->p_ext == NULL)
1250 return 0;
1251
1252 /* usually extent in the path covers blocks smaller
1253 * then *logical, but it can be that extent is the
1254 * first one in the file */
1255
1256 ex = path[depth].p_ext;
Aneesh Kumar K.Vb939e372008-01-28 23:58:27 -05001257 ee_len = ext4_ext_get_actual_len(ex);
Alex Tomas1988b512008-01-28 23:58:27 -05001258 if (*logical < le32_to_cpu(ex->ee_block)) {
Frank Mayhar273df552010-03-02 11:46:09 -05001259 if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
1260 EXT4_ERROR_INODE(inode,
1261 "EXT_FIRST_EXTENT != ex *logical %d ee_block %d!",
1262 *logical, le32_to_cpu(ex->ee_block));
1263 return -EIO;
1264 }
Alex Tomas1988b512008-01-28 23:58:27 -05001265 while (--depth >= 0) {
1266 ix = path[depth].p_idx;
Frank Mayhar273df552010-03-02 11:46:09 -05001267 if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
1268 EXT4_ERROR_INODE(inode,
1269 "ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!",
1270 ix != NULL ? ix->ei_block : 0,
1271 EXT_FIRST_INDEX(path[depth].p_hdr) != NULL ?
1272 EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block : 0,
1273 depth);
1274 return -EIO;
1275 }
Alex Tomas1988b512008-01-28 23:58:27 -05001276 }
1277 return 0;
1278 }
1279
Frank Mayhar273df552010-03-02 11:46:09 -05001280 if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
1281 EXT4_ERROR_INODE(inode,
1282 "logical %d < ee_block %d + ee_len %d!",
1283 *logical, le32_to_cpu(ex->ee_block), ee_len);
1284 return -EIO;
1285 }
Alex Tomas1988b512008-01-28 23:58:27 -05001286
Aneesh Kumar K.Vb939e372008-01-28 23:58:27 -05001287 *logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
Theodore Ts'obf89d162010-10-27 21:30:14 -04001288 *phys = ext4_ext_pblock(ex) + ee_len - 1;
Alex Tomas1988b512008-01-28 23:58:27 -05001289 return 0;
1290}
1291
1292/*
1293 * search the closest allocated block to the right for *logical
1294 * and returns it at @logical + it's physical address at @phys
1295 * if *logical is the smallest allocated block, the function
1296 * returns 0 at @phys
1297 * return value contains 0 (success) or error code
1298 */
Theodore Ts'o1f109d52010-10-27 21:30:14 -04001299static int ext4_ext_search_right(struct inode *inode,
1300 struct ext4_ext_path *path,
1301 ext4_lblk_t *logical, ext4_fsblk_t *phys)
Alex Tomas1988b512008-01-28 23:58:27 -05001302{
1303 struct buffer_head *bh = NULL;
1304 struct ext4_extent_header *eh;
1305 struct ext4_extent_idx *ix;
1306 struct ext4_extent *ex;
1307 ext4_fsblk_t block;
Eric Sandeen395a87b2009-03-10 18:18:47 -04001308 int depth; /* Note, NOT eh_depth; depth from top of tree */
1309 int ee_len;
Alex Tomas1988b512008-01-28 23:58:27 -05001310
Frank Mayhar273df552010-03-02 11:46:09 -05001311 if (unlikely(path == NULL)) {
1312 EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
1313 return -EIO;
1314 }
Alex Tomas1988b512008-01-28 23:58:27 -05001315 depth = path->p_depth;
1316 *phys = 0;
1317
1318 if (depth == 0 && path->p_ext == NULL)
1319 return 0;
1320
1321 /* usually extent in the path covers blocks smaller
1322 * then *logical, but it can be that extent is the
1323 * first one in the file */
1324
1325 ex = path[depth].p_ext;
Aneesh Kumar K.Vb939e372008-01-28 23:58:27 -05001326 ee_len = ext4_ext_get_actual_len(ex);
Alex Tomas1988b512008-01-28 23:58:27 -05001327 if (*logical < le32_to_cpu(ex->ee_block)) {
Frank Mayhar273df552010-03-02 11:46:09 -05001328 if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
1329 EXT4_ERROR_INODE(inode,
1330 "first_extent(path[%d].p_hdr) != ex",
1331 depth);
1332 return -EIO;
1333 }
Alex Tomas1988b512008-01-28 23:58:27 -05001334 while (--depth >= 0) {
1335 ix = path[depth].p_idx;
Frank Mayhar273df552010-03-02 11:46:09 -05001336 if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
1337 EXT4_ERROR_INODE(inode,
1338 "ix != EXT_FIRST_INDEX *logical %d!",
1339 *logical);
1340 return -EIO;
1341 }
Alex Tomas1988b512008-01-28 23:58:27 -05001342 }
1343 *logical = le32_to_cpu(ex->ee_block);
Theodore Ts'obf89d162010-10-27 21:30:14 -04001344 *phys = ext4_ext_pblock(ex);
Alex Tomas1988b512008-01-28 23:58:27 -05001345 return 0;
1346 }
1347
Frank Mayhar273df552010-03-02 11:46:09 -05001348 if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
1349 EXT4_ERROR_INODE(inode,
1350 "logical %d < ee_block %d + ee_len %d!",
1351 *logical, le32_to_cpu(ex->ee_block), ee_len);
1352 return -EIO;
1353 }
Alex Tomas1988b512008-01-28 23:58:27 -05001354
1355 if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
1356 /* next allocated block in this leaf */
1357 ex++;
1358 *logical = le32_to_cpu(ex->ee_block);
Theodore Ts'obf89d162010-10-27 21:30:14 -04001359 *phys = ext4_ext_pblock(ex);
Alex Tomas1988b512008-01-28 23:58:27 -05001360 return 0;
1361 }
1362
1363 /* go up and search for index to the right */
1364 while (--depth >= 0) {
1365 ix = path[depth].p_idx;
1366 if (ix != EXT_LAST_INDEX(path[depth].p_hdr))
Wu Fengguang25f1ee32008-11-25 17:24:23 -05001367 goto got_index;
Alex Tomas1988b512008-01-28 23:58:27 -05001368 }
1369
Wu Fengguang25f1ee32008-11-25 17:24:23 -05001370 /* we've gone up to the root and found no index to the right */
1371 return 0;
Alex Tomas1988b512008-01-28 23:58:27 -05001372
Wu Fengguang25f1ee32008-11-25 17:24:23 -05001373got_index:
Alex Tomas1988b512008-01-28 23:58:27 -05001374 /* we've found index to the right, let's
1375 * follow it and find the closest allocated
1376 * block to the right */
1377 ix++;
Theodore Ts'obf89d162010-10-27 21:30:14 -04001378 block = ext4_idx_pblock(ix);
Alex Tomas1988b512008-01-28 23:58:27 -05001379 while (++depth < path->p_depth) {
1380 bh = sb_bread(inode->i_sb, block);
1381 if (bh == NULL)
1382 return -EIO;
1383 eh = ext_block_hdr(bh);
Eric Sandeen395a87b2009-03-10 18:18:47 -04001384 /* subtract from p_depth to get proper eh_depth */
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -04001385 if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
Alex Tomas1988b512008-01-28 23:58:27 -05001386 put_bh(bh);
1387 return -EIO;
1388 }
1389 ix = EXT_FIRST_INDEX(eh);
Theodore Ts'obf89d162010-10-27 21:30:14 -04001390 block = ext4_idx_pblock(ix);
Alex Tomas1988b512008-01-28 23:58:27 -05001391 put_bh(bh);
1392 }
1393
1394 bh = sb_bread(inode->i_sb, block);
1395 if (bh == NULL)
1396 return -EIO;
1397 eh = ext_block_hdr(bh);
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -04001398 if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
Alex Tomas1988b512008-01-28 23:58:27 -05001399 put_bh(bh);
1400 return -EIO;
1401 }
1402 ex = EXT_FIRST_EXTENT(eh);
1403 *logical = le32_to_cpu(ex->ee_block);
Theodore Ts'obf89d162010-10-27 21:30:14 -04001404 *phys = ext4_ext_pblock(ex);
Alex Tomas1988b512008-01-28 23:58:27 -05001405 put_bh(bh);
1406 return 0;
Alex Tomas1988b512008-01-28 23:58:27 -05001407}
1408
1409/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001410 * ext4_ext_next_allocated_block:
Lukas Czernerf17722f2011-06-06 00:05:17 -04001411 * returns allocated block in subsequent extent or EXT_MAX_BLOCKS.
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001412 * NOTE: it considers block number from index entry as
1413 * allocated block. Thus, index entries have to be consistent
1414 * with leaves.
Alex Tomasa86c6182006-10-11 01:21:03 -07001415 */
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001416static ext4_lblk_t
Alex Tomasa86c6182006-10-11 01:21:03 -07001417ext4_ext_next_allocated_block(struct ext4_ext_path *path)
1418{
1419 int depth;
1420
1421 BUG_ON(path == NULL);
1422 depth = path->p_depth;
1423
1424 if (depth == 0 && path->p_ext == NULL)
Lukas Czernerf17722f2011-06-06 00:05:17 -04001425 return EXT_MAX_BLOCKS;
Alex Tomasa86c6182006-10-11 01:21:03 -07001426
1427 while (depth >= 0) {
1428 if (depth == path->p_depth) {
1429 /* leaf */
1430 if (path[depth].p_ext !=
1431 EXT_LAST_EXTENT(path[depth].p_hdr))
1432 return le32_to_cpu(path[depth].p_ext[1].ee_block);
1433 } else {
1434 /* index */
1435 if (path[depth].p_idx !=
1436 EXT_LAST_INDEX(path[depth].p_hdr))
1437 return le32_to_cpu(path[depth].p_idx[1].ei_block);
1438 }
1439 depth--;
1440 }
1441
Lukas Czernerf17722f2011-06-06 00:05:17 -04001442 return EXT_MAX_BLOCKS;
Alex Tomasa86c6182006-10-11 01:21:03 -07001443}
1444
1445/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001446 * ext4_ext_next_leaf_block:
Lukas Czernerf17722f2011-06-06 00:05:17 -04001447 * returns first allocated block from next leaf or EXT_MAX_BLOCKS
Alex Tomasa86c6182006-10-11 01:21:03 -07001448 */
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001449static ext4_lblk_t ext4_ext_next_leaf_block(struct inode *inode,
Andrew Morton63f57932006-10-11 01:21:24 -07001450 struct ext4_ext_path *path)
Alex Tomasa86c6182006-10-11 01:21:03 -07001451{
1452 int depth;
1453
1454 BUG_ON(path == NULL);
1455 depth = path->p_depth;
1456
1457 /* zero-tree has no leaf blocks at all */
1458 if (depth == 0)
Lukas Czernerf17722f2011-06-06 00:05:17 -04001459 return EXT_MAX_BLOCKS;
Alex Tomasa86c6182006-10-11 01:21:03 -07001460
1461 /* go to index block */
1462 depth--;
1463
1464 while (depth >= 0) {
1465 if (path[depth].p_idx !=
1466 EXT_LAST_INDEX(path[depth].p_hdr))
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001467 return (ext4_lblk_t)
1468 le32_to_cpu(path[depth].p_idx[1].ei_block);
Alex Tomasa86c6182006-10-11 01:21:03 -07001469 depth--;
1470 }
1471
Lukas Czernerf17722f2011-06-06 00:05:17 -04001472 return EXT_MAX_BLOCKS;
Alex Tomasa86c6182006-10-11 01:21:03 -07001473}
1474
1475/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001476 * ext4_ext_correct_indexes:
1477 * if leaf gets modified and modified extent is first in the leaf,
1478 * then we have to correct all indexes above.
Alex Tomasa86c6182006-10-11 01:21:03 -07001479 * TODO: do we need to correct tree in all cases?
1480 */
Aneesh Kumar K.V1d03ec92008-01-28 23:58:27 -05001481static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
Alex Tomasa86c6182006-10-11 01:21:03 -07001482 struct ext4_ext_path *path)
1483{
1484 struct ext4_extent_header *eh;
1485 int depth = ext_depth(inode);
1486 struct ext4_extent *ex;
1487 __le32 border;
1488 int k, err = 0;
1489
1490 eh = path[depth].p_hdr;
1491 ex = path[depth].p_ext;
Frank Mayhar273df552010-03-02 11:46:09 -05001492
1493 if (unlikely(ex == NULL || eh == NULL)) {
1494 EXT4_ERROR_INODE(inode,
1495 "ex %p == NULL or eh %p == NULL", ex, eh);
1496 return -EIO;
1497 }
Alex Tomasa86c6182006-10-11 01:21:03 -07001498
1499 if (depth == 0) {
1500 /* there is no tree at all */
1501 return 0;
1502 }
1503
1504 if (ex != EXT_FIRST_EXTENT(eh)) {
1505 /* we correct tree if first leaf got modified only */
1506 return 0;
1507 }
1508
1509 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001510 * TODO: we need correction if border is smaller than current one
Alex Tomasa86c6182006-10-11 01:21:03 -07001511 */
1512 k = depth - 1;
1513 border = path[depth].p_ext->ee_block;
Avantika Mathur7e028972006-12-06 20:41:33 -08001514 err = ext4_ext_get_access(handle, inode, path + k);
1515 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001516 return err;
1517 path[k].p_idx->ei_block = border;
Avantika Mathur7e028972006-12-06 20:41:33 -08001518 err = ext4_ext_dirty(handle, inode, path + k);
1519 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001520 return err;
1521
1522 while (k--) {
1523 /* change all left-side indexes */
1524 if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr))
1525 break;
Avantika Mathur7e028972006-12-06 20:41:33 -08001526 err = ext4_ext_get_access(handle, inode, path + k);
1527 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001528 break;
1529 path[k].p_idx->ei_block = border;
Avantika Mathur7e028972006-12-06 20:41:33 -08001530 err = ext4_ext_dirty(handle, inode, path + k);
1531 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001532 break;
1533 }
1534
1535 return err;
1536}
1537
Akira Fujita748de672009-06-17 19:24:03 -04001538int
Alex Tomasa86c6182006-10-11 01:21:03 -07001539ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1,
1540 struct ext4_extent *ex2)
1541{
Amit Arora749269f2007-07-18 09:02:56 -04001542 unsigned short ext1_ee_len, ext2_ee_len, max_len;
Amit Aroraa2df2a62007-07-17 21:42:41 -04001543
1544 /*
1545 * Make sure that either both extents are uninitialized, or
1546 * both are _not_.
1547 */
1548 if (ext4_ext_is_uninitialized(ex1) ^ ext4_ext_is_uninitialized(ex2))
1549 return 0;
1550
Amit Arora749269f2007-07-18 09:02:56 -04001551 if (ext4_ext_is_uninitialized(ex1))
1552 max_len = EXT_UNINIT_MAX_LEN;
1553 else
1554 max_len = EXT_INIT_MAX_LEN;
1555
Amit Aroraa2df2a62007-07-17 21:42:41 -04001556 ext1_ee_len = ext4_ext_get_actual_len(ex1);
1557 ext2_ee_len = ext4_ext_get_actual_len(ex2);
1558
1559 if (le32_to_cpu(ex1->ee_block) + ext1_ee_len !=
Andrew Morton63f57932006-10-11 01:21:24 -07001560 le32_to_cpu(ex2->ee_block))
Alex Tomasa86c6182006-10-11 01:21:03 -07001561 return 0;
1562
Suparna Bhattacharya471d4012006-10-11 01:21:06 -07001563 /*
1564 * To allow future support for preallocated extents to be added
1565 * as an RO_COMPAT feature, refuse to merge to extents if
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001566 * this can result in the top bit of ee_len being set.
Suparna Bhattacharya471d4012006-10-11 01:21:06 -07001567 */
Amit Arora749269f2007-07-18 09:02:56 -04001568 if (ext1_ee_len + ext2_ee_len > max_len)
Suparna Bhattacharya471d4012006-10-11 01:21:06 -07001569 return 0;
Robert P. J. Daybbf2f9f2007-02-17 19:20:16 +01001570#ifdef AGGRESSIVE_TEST
Aneesh Kumar K.Vb939e372008-01-28 23:58:27 -05001571 if (ext1_ee_len >= 4)
Alex Tomasa86c6182006-10-11 01:21:03 -07001572 return 0;
1573#endif
1574
Theodore Ts'obf89d162010-10-27 21:30:14 -04001575 if (ext4_ext_pblock(ex1) + ext1_ee_len == ext4_ext_pblock(ex2))
Alex Tomasa86c6182006-10-11 01:21:03 -07001576 return 1;
1577 return 0;
1578}
1579
1580/*
Amit Arora56055d32007-07-17 21:42:38 -04001581 * This function tries to merge the "ex" extent to the next extent in the tree.
1582 * It always tries to merge towards right. If you want to merge towards
1583 * left, pass "ex - 1" as argument instead of "ex".
1584 * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns
1585 * 1 if they got merged.
1586 */
Yongqiang Yang197217a2011-05-03 11:45:29 -04001587static int ext4_ext_try_to_merge_right(struct inode *inode,
Theodore Ts'o1f109d52010-10-27 21:30:14 -04001588 struct ext4_ext_path *path,
1589 struct ext4_extent *ex)
Amit Arora56055d32007-07-17 21:42:38 -04001590{
1591 struct ext4_extent_header *eh;
1592 unsigned int depth, len;
1593 int merge_done = 0;
1594 int uninitialized = 0;
1595
1596 depth = ext_depth(inode);
1597 BUG_ON(path[depth].p_hdr == NULL);
1598 eh = path[depth].p_hdr;
1599
1600 while (ex < EXT_LAST_EXTENT(eh)) {
1601 if (!ext4_can_extents_be_merged(inode, ex, ex + 1))
1602 break;
1603 /* merge with next extent! */
1604 if (ext4_ext_is_uninitialized(ex))
1605 uninitialized = 1;
1606 ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1607 + ext4_ext_get_actual_len(ex + 1));
1608 if (uninitialized)
1609 ext4_ext_mark_uninitialized(ex);
1610
1611 if (ex + 1 < EXT_LAST_EXTENT(eh)) {
1612 len = (EXT_LAST_EXTENT(eh) - ex - 1)
1613 * sizeof(struct ext4_extent);
1614 memmove(ex + 1, ex + 2, len);
1615 }
Marcin Slusarze8546d02008-04-17 10:38:59 -04001616 le16_add_cpu(&eh->eh_entries, -1);
Amit Arora56055d32007-07-17 21:42:38 -04001617 merge_done = 1;
1618 WARN_ON(eh->eh_entries == 0);
1619 if (!eh->eh_entries)
Theodore Ts'o24676da2010-05-16 21:00:00 -04001620 EXT4_ERROR_INODE(inode, "eh->eh_entries = 0!");
Amit Arora56055d32007-07-17 21:42:38 -04001621 }
1622
1623 return merge_done;
1624}
1625
1626/*
Yongqiang Yang197217a2011-05-03 11:45:29 -04001627 * This function tries to merge the @ex extent to neighbours in the tree.
1628 * return 1 if merge left else 0.
1629 */
1630static int ext4_ext_try_to_merge(struct inode *inode,
1631 struct ext4_ext_path *path,
1632 struct ext4_extent *ex) {
1633 struct ext4_extent_header *eh;
1634 unsigned int depth;
1635 int merge_done = 0;
1636 int ret = 0;
1637
1638 depth = ext_depth(inode);
1639 BUG_ON(path[depth].p_hdr == NULL);
1640 eh = path[depth].p_hdr;
1641
1642 if (ex > EXT_FIRST_EXTENT(eh))
1643 merge_done = ext4_ext_try_to_merge_right(inode, path, ex - 1);
1644
1645 if (!merge_done)
1646 ret = ext4_ext_try_to_merge_right(inode, path, ex);
1647
1648 return ret;
1649}
1650
1651/*
Amit Arora25d14f92007-05-24 13:04:13 -04001652 * check if a portion of the "newext" extent overlaps with an
1653 * existing extent.
1654 *
1655 * If there is an overlap discovered, it updates the length of the newext
1656 * such that there will be no overlap, and then returns 1.
1657 * If there is no overlap found, it returns 0.
1658 */
Theodore Ts'o1f109d52010-10-27 21:30:14 -04001659static unsigned int ext4_ext_check_overlap(struct inode *inode,
1660 struct ext4_extent *newext,
1661 struct ext4_ext_path *path)
Amit Arora25d14f92007-05-24 13:04:13 -04001662{
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001663 ext4_lblk_t b1, b2;
Amit Arora25d14f92007-05-24 13:04:13 -04001664 unsigned int depth, len1;
1665 unsigned int ret = 0;
1666
1667 b1 = le32_to_cpu(newext->ee_block);
Amit Aroraa2df2a62007-07-17 21:42:41 -04001668 len1 = ext4_ext_get_actual_len(newext);
Amit Arora25d14f92007-05-24 13:04:13 -04001669 depth = ext_depth(inode);
1670 if (!path[depth].p_ext)
1671 goto out;
1672 b2 = le32_to_cpu(path[depth].p_ext->ee_block);
1673
1674 /*
1675 * get the next allocated block if the extent in the path
Theodore Ts'o2b2d6d02008-07-26 16:15:44 -04001676 * is before the requested block(s)
Amit Arora25d14f92007-05-24 13:04:13 -04001677 */
1678 if (b2 < b1) {
1679 b2 = ext4_ext_next_allocated_block(path);
Lukas Czernerf17722f2011-06-06 00:05:17 -04001680 if (b2 == EXT_MAX_BLOCKS)
Amit Arora25d14f92007-05-24 13:04:13 -04001681 goto out;
1682 }
1683
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001684 /* check for wrap through zero on extent logical start block*/
Amit Arora25d14f92007-05-24 13:04:13 -04001685 if (b1 + len1 < b1) {
Lukas Czernerf17722f2011-06-06 00:05:17 -04001686 len1 = EXT_MAX_BLOCKS - b1;
Amit Arora25d14f92007-05-24 13:04:13 -04001687 newext->ee_len = cpu_to_le16(len1);
1688 ret = 1;
1689 }
1690
1691 /* check for overlap */
1692 if (b1 + len1 > b2) {
1693 newext->ee_len = cpu_to_le16(b2 - b1);
1694 ret = 1;
1695 }
1696out:
1697 return ret;
1698}
1699
1700/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001701 * ext4_ext_insert_extent:
1702 * tries to merge requsted extent into the existing extent or
1703 * inserts requested extent as new one into the tree,
1704 * creating new leaf in the no-space case.
Alex Tomasa86c6182006-10-11 01:21:03 -07001705 */
1706int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
1707 struct ext4_ext_path *path,
Mingming Cao00314622009-09-28 15:49:08 -04001708 struct ext4_extent *newext, int flag)
Alex Tomasa86c6182006-10-11 01:21:03 -07001709{
Theodore Ts'oaf5bc922008-09-08 22:25:24 -04001710 struct ext4_extent_header *eh;
Alex Tomasa86c6182006-10-11 01:21:03 -07001711 struct ext4_extent *ex, *fex;
1712 struct ext4_extent *nearex; /* nearest extent */
1713 struct ext4_ext_path *npath = NULL;
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001714 int depth, len, err;
1715 ext4_lblk_t next;
Amit Aroraa2df2a62007-07-17 21:42:41 -04001716 unsigned uninitialized = 0;
Allison Henderson55f020d2011-05-25 07:41:26 -04001717 int flags = 0;
Alex Tomasa86c6182006-10-11 01:21:03 -07001718
Frank Mayhar273df552010-03-02 11:46:09 -05001719 if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
1720 EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
1721 return -EIO;
1722 }
Alex Tomasa86c6182006-10-11 01:21:03 -07001723 depth = ext_depth(inode);
1724 ex = path[depth].p_ext;
Frank Mayhar273df552010-03-02 11:46:09 -05001725 if (unlikely(path[depth].p_hdr == NULL)) {
1726 EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
1727 return -EIO;
1728 }
Alex Tomasa86c6182006-10-11 01:21:03 -07001729
1730 /* try to insert block into found extent and return */
Jiaying Zhang744692d2010-03-04 16:14:02 -05001731 if (ex && !(flag & EXT4_GET_BLOCKS_PRE_IO)
Mingming Cao00314622009-09-28 15:49:08 -04001732 && ext4_can_extents_be_merged(inode, ex, newext)) {
Mingming553f9002009-09-18 13:34:55 -04001733 ext_debug("append [%d]%d block to %d:[%d]%d (from %llu)\n",
Theodore Ts'obf89d162010-10-27 21:30:14 -04001734 ext4_ext_is_uninitialized(newext),
1735 ext4_ext_get_actual_len(newext),
1736 le32_to_cpu(ex->ee_block),
1737 ext4_ext_is_uninitialized(ex),
1738 ext4_ext_get_actual_len(ex),
1739 ext4_ext_pblock(ex));
Avantika Mathur7e028972006-12-06 20:41:33 -08001740 err = ext4_ext_get_access(handle, inode, path + depth);
1741 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001742 return err;
Amit Aroraa2df2a62007-07-17 21:42:41 -04001743
1744 /*
1745 * ext4_can_extents_be_merged should have checked that either
1746 * both extents are uninitialized, or both aren't. Thus we
1747 * need to check only one of them here.
1748 */
1749 if (ext4_ext_is_uninitialized(ex))
1750 uninitialized = 1;
1751 ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1752 + ext4_ext_get_actual_len(newext));
1753 if (uninitialized)
1754 ext4_ext_mark_uninitialized(ex);
Alex Tomasa86c6182006-10-11 01:21:03 -07001755 eh = path[depth].p_hdr;
1756 nearex = ex;
1757 goto merge;
1758 }
1759
1760repeat:
1761 depth = ext_depth(inode);
1762 eh = path[depth].p_hdr;
1763 if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max))
1764 goto has_space;
1765
1766 /* probably next leaf has space for us? */
1767 fex = EXT_LAST_EXTENT(eh);
1768 next = ext4_ext_next_leaf_block(inode, path);
1769 if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block)
Lukas Czernerf17722f2011-06-06 00:05:17 -04001770 && next != EXT_MAX_BLOCKS) {
Alex Tomasa86c6182006-10-11 01:21:03 -07001771 ext_debug("next leaf block - %d\n", next);
1772 BUG_ON(npath != NULL);
1773 npath = ext4_ext_find_extent(inode, next, NULL);
1774 if (IS_ERR(npath))
1775 return PTR_ERR(npath);
1776 BUG_ON(npath->p_depth != path->p_depth);
1777 eh = npath[depth].p_hdr;
1778 if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) {
Lucas De Marchi25985ed2011-03-30 22:57:33 -03001779 ext_debug("next leaf isn't full(%d)\n",
Alex Tomasa86c6182006-10-11 01:21:03 -07001780 le16_to_cpu(eh->eh_entries));
1781 path = npath;
1782 goto repeat;
1783 }
1784 ext_debug("next leaf has no free space(%d,%d)\n",
1785 le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
1786 }
1787
1788 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07001789 * There is no free space in the found leaf.
1790 * We're gonna add a new leaf in the tree.
Alex Tomasa86c6182006-10-11 01:21:03 -07001791 */
Allison Henderson55f020d2011-05-25 07:41:26 -04001792 if (flag & EXT4_GET_BLOCKS_PUNCH_OUT_EXT)
1793 flags = EXT4_MB_USE_ROOT_BLOCKS;
1794 err = ext4_ext_create_new_leaf(handle, inode, flags, path, newext);
Alex Tomasa86c6182006-10-11 01:21:03 -07001795 if (err)
1796 goto cleanup;
1797 depth = ext_depth(inode);
1798 eh = path[depth].p_hdr;
1799
1800has_space:
1801 nearex = path[depth].p_ext;
1802
Avantika Mathur7e028972006-12-06 20:41:33 -08001803 err = ext4_ext_get_access(handle, inode, path + depth);
1804 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07001805 goto cleanup;
1806
1807 if (!nearex) {
1808 /* there is no extent in this leaf, create first one */
Mingming553f9002009-09-18 13:34:55 -04001809 ext_debug("first extent in the leaf: %d:%llu:[%d]%d\n",
Dave Kleikamp8c55e202007-05-24 13:04:54 -04001810 le32_to_cpu(newext->ee_block),
Theodore Ts'obf89d162010-10-27 21:30:14 -04001811 ext4_ext_pblock(newext),
Mingming553f9002009-09-18 13:34:55 -04001812 ext4_ext_is_uninitialized(newext),
Amit Aroraa2df2a62007-07-17 21:42:41 -04001813 ext4_ext_get_actual_len(newext));
Alex Tomasa86c6182006-10-11 01:21:03 -07001814 path[depth].p_ext = EXT_FIRST_EXTENT(eh);
1815 } else if (le32_to_cpu(newext->ee_block)
Dave Kleikamp8c55e202007-05-24 13:04:54 -04001816 > le32_to_cpu(nearex->ee_block)) {
Alex Tomasa86c6182006-10-11 01:21:03 -07001817/* BUG_ON(newext->ee_block == nearex->ee_block); */
1818 if (nearex != EXT_LAST_EXTENT(eh)) {
1819 len = EXT_MAX_EXTENT(eh) - nearex;
1820 len = (len - 1) * sizeof(struct ext4_extent);
1821 len = len < 0 ? 0 : len;
Mingming553f9002009-09-18 13:34:55 -04001822 ext_debug("insert %d:%llu:[%d]%d after: nearest 0x%p, "
Alex Tomasa86c6182006-10-11 01:21:03 -07001823 "move %d from 0x%p to 0x%p\n",
Dave Kleikamp8c55e202007-05-24 13:04:54 -04001824 le32_to_cpu(newext->ee_block),
Theodore Ts'obf89d162010-10-27 21:30:14 -04001825 ext4_ext_pblock(newext),
Mingming553f9002009-09-18 13:34:55 -04001826 ext4_ext_is_uninitialized(newext),
Amit Aroraa2df2a62007-07-17 21:42:41 -04001827 ext4_ext_get_actual_len(newext),
Alex Tomasa86c6182006-10-11 01:21:03 -07001828 nearex, len, nearex + 1, nearex + 2);
1829 memmove(nearex + 2, nearex + 1, len);
1830 }
1831 path[depth].p_ext = nearex + 1;
1832 } else {
1833 BUG_ON(newext->ee_block == nearex->ee_block);
1834 len = (EXT_MAX_EXTENT(eh) - nearex) * sizeof(struct ext4_extent);
1835 len = len < 0 ? 0 : len;
Mingming553f9002009-09-18 13:34:55 -04001836 ext_debug("insert %d:%llu:[%d]%d before: nearest 0x%p, "
Alex Tomasa86c6182006-10-11 01:21:03 -07001837 "move %d from 0x%p to 0x%p\n",
1838 le32_to_cpu(newext->ee_block),
Theodore Ts'obf89d162010-10-27 21:30:14 -04001839 ext4_ext_pblock(newext),
Mingming553f9002009-09-18 13:34:55 -04001840 ext4_ext_is_uninitialized(newext),
Amit Aroraa2df2a62007-07-17 21:42:41 -04001841 ext4_ext_get_actual_len(newext),
Alex Tomasa86c6182006-10-11 01:21:03 -07001842 nearex, len, nearex + 1, nearex + 2);
1843 memmove(nearex + 1, nearex, len);
1844 path[depth].p_ext = nearex;
1845 }
1846
Marcin Slusarze8546d02008-04-17 10:38:59 -04001847 le16_add_cpu(&eh->eh_entries, 1);
Alex Tomasa86c6182006-10-11 01:21:03 -07001848 nearex = path[depth].p_ext;
1849 nearex->ee_block = newext->ee_block;
Theodore Ts'obf89d162010-10-27 21:30:14 -04001850 ext4_ext_store_pblock(nearex, ext4_ext_pblock(newext));
Alex Tomasa86c6182006-10-11 01:21:03 -07001851 nearex->ee_len = newext->ee_len;
Alex Tomasa86c6182006-10-11 01:21:03 -07001852
1853merge:
1854 /* try to merge extents to the right */
Jiaying Zhang744692d2010-03-04 16:14:02 -05001855 if (!(flag & EXT4_GET_BLOCKS_PRE_IO))
Mingming Cao00314622009-09-28 15:49:08 -04001856 ext4_ext_try_to_merge(inode, path, nearex);
Alex Tomasa86c6182006-10-11 01:21:03 -07001857
1858 /* try to merge extents to the left */
1859
1860 /* time to correct all indexes above */
1861 err = ext4_ext_correct_indexes(handle, inode, path);
1862 if (err)
1863 goto cleanup;
1864
1865 err = ext4_ext_dirty(handle, inode, path + depth);
1866
1867cleanup:
1868 if (npath) {
1869 ext4_ext_drop_refs(npath);
1870 kfree(npath);
1871 }
Alex Tomasa86c6182006-10-11 01:21:03 -07001872 ext4_ext_invalidate_cache(inode);
1873 return err;
1874}
1875
Theodore Ts'o1f109d52010-10-27 21:30:14 -04001876static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
1877 ext4_lblk_t num, ext_prepare_callback func,
1878 void *cbdata)
Eric Sandeen6873fa02008-10-07 00:46:36 -04001879{
1880 struct ext4_ext_path *path = NULL;
1881 struct ext4_ext_cache cbex;
1882 struct ext4_extent *ex;
1883 ext4_lblk_t next, start = 0, end = 0;
1884 ext4_lblk_t last = block + num;
1885 int depth, exists, err = 0;
1886
1887 BUG_ON(func == NULL);
1888 BUG_ON(inode == NULL);
1889
Lukas Czernerf17722f2011-06-06 00:05:17 -04001890 while (block < last && block != EXT_MAX_BLOCKS) {
Eric Sandeen6873fa02008-10-07 00:46:36 -04001891 num = last - block;
1892 /* find extent for this block */
Theodore Ts'ofab3a542009-12-09 21:30:02 -05001893 down_read(&EXT4_I(inode)->i_data_sem);
Eric Sandeen6873fa02008-10-07 00:46:36 -04001894 path = ext4_ext_find_extent(inode, block, path);
Theodore Ts'ofab3a542009-12-09 21:30:02 -05001895 up_read(&EXT4_I(inode)->i_data_sem);
Eric Sandeen6873fa02008-10-07 00:46:36 -04001896 if (IS_ERR(path)) {
1897 err = PTR_ERR(path);
1898 path = NULL;
1899 break;
1900 }
1901
1902 depth = ext_depth(inode);
Frank Mayhar273df552010-03-02 11:46:09 -05001903 if (unlikely(path[depth].p_hdr == NULL)) {
1904 EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
1905 err = -EIO;
1906 break;
1907 }
Eric Sandeen6873fa02008-10-07 00:46:36 -04001908 ex = path[depth].p_ext;
1909 next = ext4_ext_next_allocated_block(path);
1910
1911 exists = 0;
1912 if (!ex) {
1913 /* there is no extent yet, so try to allocate
1914 * all requested space */
1915 start = block;
1916 end = block + num;
1917 } else if (le32_to_cpu(ex->ee_block) > block) {
1918 /* need to allocate space before found extent */
1919 start = block;
1920 end = le32_to_cpu(ex->ee_block);
1921 if (block + num < end)
1922 end = block + num;
1923 } else if (block >= le32_to_cpu(ex->ee_block)
1924 + ext4_ext_get_actual_len(ex)) {
1925 /* need to allocate space after found extent */
1926 start = block;
1927 end = block + num;
1928 if (end >= next)
1929 end = next;
1930 } else if (block >= le32_to_cpu(ex->ee_block)) {
1931 /*
1932 * some part of requested space is covered
1933 * by found extent
1934 */
1935 start = block;
1936 end = le32_to_cpu(ex->ee_block)
1937 + ext4_ext_get_actual_len(ex);
1938 if (block + num < end)
1939 end = block + num;
1940 exists = 1;
1941 } else {
1942 BUG();
1943 }
1944 BUG_ON(end <= start);
1945
1946 if (!exists) {
1947 cbex.ec_block = start;
1948 cbex.ec_len = end - start;
1949 cbex.ec_start = 0;
Eric Sandeen6873fa02008-10-07 00:46:36 -04001950 } else {
1951 cbex.ec_block = le32_to_cpu(ex->ee_block);
1952 cbex.ec_len = ext4_ext_get_actual_len(ex);
Theodore Ts'obf89d162010-10-27 21:30:14 -04001953 cbex.ec_start = ext4_ext_pblock(ex);
Eric Sandeen6873fa02008-10-07 00:46:36 -04001954 }
1955
Frank Mayhar273df552010-03-02 11:46:09 -05001956 if (unlikely(cbex.ec_len == 0)) {
1957 EXT4_ERROR_INODE(inode, "cbex.ec_len == 0");
1958 err = -EIO;
1959 break;
1960 }
Lukas Czernerc03f8aa2011-06-06 00:06:52 -04001961 err = func(inode, next, &cbex, ex, cbdata);
Eric Sandeen6873fa02008-10-07 00:46:36 -04001962 ext4_ext_drop_refs(path);
1963
1964 if (err < 0)
1965 break;
1966
1967 if (err == EXT_REPEAT)
1968 continue;
1969 else if (err == EXT_BREAK) {
1970 err = 0;
1971 break;
1972 }
1973
1974 if (ext_depth(inode) != depth) {
1975 /* depth was changed. we have to realloc path */
1976 kfree(path);
1977 path = NULL;
1978 }
1979
1980 block = cbex.ec_block + cbex.ec_len;
1981 }
1982
1983 if (path) {
1984 ext4_ext_drop_refs(path);
1985 kfree(path);
1986 }
1987
1988 return err;
1989}
1990
Avantika Mathur09b88252006-12-06 20:41:36 -08001991static void
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05001992ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block,
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05001993 __u32 len, ext4_fsblk_t start)
Alex Tomasa86c6182006-10-11 01:21:03 -07001994{
1995 struct ext4_ext_cache *cex;
1996 BUG_ON(len == 0);
Theodore Ts'o2ec0ae32009-05-15 09:07:28 -04001997 spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
Alex Tomasa86c6182006-10-11 01:21:03 -07001998 cex = &EXT4_I(inode)->i_cached_extent;
Alex Tomasa86c6182006-10-11 01:21:03 -07001999 cex->ec_block = block;
2000 cex->ec_len = len;
2001 cex->ec_start = start;
Theodore Ts'o2ec0ae32009-05-15 09:07:28 -04002002 spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
Alex Tomasa86c6182006-10-11 01:21:03 -07002003}
2004
2005/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002006 * ext4_ext_put_gap_in_cache:
2007 * calculate boundaries of the gap that the requested block fits into
Alex Tomasa86c6182006-10-11 01:21:03 -07002008 * and cache this gap
2009 */
Avantika Mathur09b88252006-12-06 20:41:36 -08002010static void
Alex Tomasa86c6182006-10-11 01:21:03 -07002011ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002012 ext4_lblk_t block)
Alex Tomasa86c6182006-10-11 01:21:03 -07002013{
2014 int depth = ext_depth(inode);
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002015 unsigned long len;
2016 ext4_lblk_t lblock;
Alex Tomasa86c6182006-10-11 01:21:03 -07002017 struct ext4_extent *ex;
2018
2019 ex = path[depth].p_ext;
2020 if (ex == NULL) {
2021 /* there is no extent yet, so gap is [0;-] */
2022 lblock = 0;
Lukas Czernerf17722f2011-06-06 00:05:17 -04002023 len = EXT_MAX_BLOCKS;
Alex Tomasa86c6182006-10-11 01:21:03 -07002024 ext_debug("cache gap(whole file):");
2025 } else if (block < le32_to_cpu(ex->ee_block)) {
2026 lblock = block;
2027 len = le32_to_cpu(ex->ee_block) - block;
Eric Sandeenbba90742008-01-28 23:58:27 -05002028 ext_debug("cache gap(before): %u [%u:%u]",
2029 block,
2030 le32_to_cpu(ex->ee_block),
2031 ext4_ext_get_actual_len(ex));
Alex Tomasa86c6182006-10-11 01:21:03 -07002032 } else if (block >= le32_to_cpu(ex->ee_block)
Amit Aroraa2df2a62007-07-17 21:42:41 -04002033 + ext4_ext_get_actual_len(ex)) {
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002034 ext4_lblk_t next;
Dave Kleikamp8c55e202007-05-24 13:04:54 -04002035 lblock = le32_to_cpu(ex->ee_block)
Amit Aroraa2df2a62007-07-17 21:42:41 -04002036 + ext4_ext_get_actual_len(ex);
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002037
2038 next = ext4_ext_next_allocated_block(path);
Eric Sandeenbba90742008-01-28 23:58:27 -05002039 ext_debug("cache gap(after): [%u:%u] %u",
2040 le32_to_cpu(ex->ee_block),
2041 ext4_ext_get_actual_len(ex),
2042 block);
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002043 BUG_ON(next == lblock);
2044 len = next - lblock;
Alex Tomasa86c6182006-10-11 01:21:03 -07002045 } else {
2046 lblock = len = 0;
2047 BUG();
2048 }
2049
Eric Sandeenbba90742008-01-28 23:58:27 -05002050 ext_debug(" -> %u:%lu\n", lblock, len);
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05002051 ext4_ext_put_in_cache(inode, lblock, len, 0);
Alex Tomasa86c6182006-10-11 01:21:03 -07002052}
2053
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05002054/*
Allison Hendersona4bb6b62011-05-25 07:41:50 -04002055 * ext4_ext_in_cache()
2056 * Checks to see if the given block is in the cache.
2057 * If it is, the cached extent is stored in the given
2058 * cache extent pointer. If the cached extent is a hole,
2059 * this routine should be used instead of
2060 * ext4_ext_in_cache if the calling function needs to
2061 * know the size of the hole.
2062 *
2063 * @inode: The files inode
2064 * @block: The block to look for in the cache
2065 * @ex: Pointer where the cached extent will be stored
2066 * if it contains block
2067 *
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05002068 * Return 0 if cache is invalid; 1 if the cache is valid
2069 */
Allison Hendersona4bb6b62011-05-25 07:41:50 -04002070static int ext4_ext_check_cache(struct inode *inode, ext4_lblk_t block,
2071 struct ext4_ext_cache *ex){
Alex Tomasa86c6182006-10-11 01:21:03 -07002072 struct ext4_ext_cache *cex;
Vivek Haldar77f41352011-05-22 21:24:16 -04002073 struct ext4_sb_info *sbi;
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05002074 int ret = 0;
Alex Tomasa86c6182006-10-11 01:21:03 -07002075
Theodore Ts'o60e66792010-05-17 07:00:00 -04002076 /*
Theodore Ts'o2ec0ae32009-05-15 09:07:28 -04002077 * We borrow i_block_reservation_lock to protect i_cached_extent
2078 */
2079 spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
Alex Tomasa86c6182006-10-11 01:21:03 -07002080 cex = &EXT4_I(inode)->i_cached_extent;
Vivek Haldar77f41352011-05-22 21:24:16 -04002081 sbi = EXT4_SB(inode->i_sb);
Alex Tomasa86c6182006-10-11 01:21:03 -07002082
2083 /* has cache valid data? */
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05002084 if (cex->ec_len == 0)
Theodore Ts'o2ec0ae32009-05-15 09:07:28 -04002085 goto errout;
Alex Tomasa86c6182006-10-11 01:21:03 -07002086
Akinobu Mita731eb1a2010-03-03 23:55:01 -05002087 if (in_range(block, cex->ec_block, cex->ec_len)) {
Allison Hendersona4bb6b62011-05-25 07:41:50 -04002088 memcpy(ex, cex, sizeof(struct ext4_ext_cache));
Eric Sandeenbba90742008-01-28 23:58:27 -05002089 ext_debug("%u cached by %u:%u:%llu\n",
2090 block,
2091 cex->ec_block, cex->ec_len, cex->ec_start);
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05002092 ret = 1;
Alex Tomasa86c6182006-10-11 01:21:03 -07002093 }
Theodore Ts'o2ec0ae32009-05-15 09:07:28 -04002094errout:
Vivek Haldar77f41352011-05-22 21:24:16 -04002095 if (!ret)
2096 sbi->extent_cache_misses++;
2097 else
2098 sbi->extent_cache_hits++;
Theodore Ts'o2ec0ae32009-05-15 09:07:28 -04002099 spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
2100 return ret;
Alex Tomasa86c6182006-10-11 01:21:03 -07002101}
2102
2103/*
Allison Hendersona4bb6b62011-05-25 07:41:50 -04002104 * ext4_ext_in_cache()
2105 * Checks to see if the given block is in the cache.
2106 * If it is, the cached extent is stored in the given
2107 * extent pointer.
2108 *
2109 * @inode: The files inode
2110 * @block: The block to look for in the cache
2111 * @ex: Pointer where the cached extent will be stored
2112 * if it contains block
2113 *
2114 * Return 0 if cache is invalid; 1 if the cache is valid
2115 */
2116static int
2117ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
2118 struct ext4_extent *ex)
2119{
2120 struct ext4_ext_cache cex;
2121 int ret = 0;
2122
2123 if (ext4_ext_check_cache(inode, block, &cex)) {
2124 ex->ee_block = cpu_to_le32(cex.ec_block);
2125 ext4_ext_store_pblock(ex, cex.ec_start);
2126 ex->ee_len = cpu_to_le16(cex.ec_len);
2127 ret = 1;
2128 }
2129
2130 return ret;
2131}
2132
2133
2134/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002135 * ext4_ext_rm_idx:
2136 * removes index from the index block.
2137 * It's used in truncate case only, thus all requests are for
2138 * last index in the block only.
Alex Tomasa86c6182006-10-11 01:21:03 -07002139 */
Aneesh Kumar K.V1d03ec92008-01-28 23:58:27 -05002140static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
Alex Tomasa86c6182006-10-11 01:21:03 -07002141 struct ext4_ext_path *path)
2142{
Alex Tomasa86c6182006-10-11 01:21:03 -07002143 int err;
Alex Tomasf65e6fb2006-10-11 01:21:05 -07002144 ext4_fsblk_t leaf;
Alex Tomasa86c6182006-10-11 01:21:03 -07002145
2146 /* free index block */
2147 path--;
Theodore Ts'obf89d162010-10-27 21:30:14 -04002148 leaf = ext4_idx_pblock(path->p_idx);
Frank Mayhar273df552010-03-02 11:46:09 -05002149 if (unlikely(path->p_hdr->eh_entries == 0)) {
2150 EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0");
2151 return -EIO;
2152 }
Avantika Mathur7e028972006-12-06 20:41:33 -08002153 err = ext4_ext_get_access(handle, inode, path);
2154 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07002155 return err;
Marcin Slusarze8546d02008-04-17 10:38:59 -04002156 le16_add_cpu(&path->p_hdr->eh_entries, -1);
Avantika Mathur7e028972006-12-06 20:41:33 -08002157 err = ext4_ext_dirty(handle, inode, path);
2158 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07002159 return err;
Mingming Cao2ae02102006-10-11 01:21:11 -07002160 ext_debug("index is empty, remove it, free block %llu\n", leaf);
Peter Huewe7dc57612011-02-21 21:01:42 -05002161 ext4_free_blocks(handle, inode, NULL, leaf, 1,
Theodore Ts'oe6362602009-11-23 07:17:05 -05002162 EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
Alex Tomasa86c6182006-10-11 01:21:03 -07002163 return err;
2164}
2165
2166/*
Mingming Caoee12b632008-08-19 22:16:05 -04002167 * ext4_ext_calc_credits_for_single_extent:
2168 * This routine returns max. credits that needed to insert an extent
2169 * to the extent tree.
2170 * When pass the actual path, the caller should calculate credits
2171 * under i_data_sem.
Alex Tomasa86c6182006-10-11 01:21:03 -07002172 */
Mingming Cao525f4ed2008-08-19 22:15:58 -04002173int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks,
Alex Tomasa86c6182006-10-11 01:21:03 -07002174 struct ext4_ext_path *path)
2175{
Alex Tomasa86c6182006-10-11 01:21:03 -07002176 if (path) {
Mingming Caoee12b632008-08-19 22:16:05 -04002177 int depth = ext_depth(inode);
Mingming Caof3bd1f32008-08-19 22:16:03 -04002178 int ret = 0;
Mingming Caoee12b632008-08-19 22:16:05 -04002179
Alex Tomasa86c6182006-10-11 01:21:03 -07002180 /* probably there is space in leaf? */
Alex Tomasa86c6182006-10-11 01:21:03 -07002181 if (le16_to_cpu(path[depth].p_hdr->eh_entries)
Mingming Caoee12b632008-08-19 22:16:05 -04002182 < le16_to_cpu(path[depth].p_hdr->eh_max)) {
2183
2184 /*
2185 * There are some space in the leaf tree, no
2186 * need to account for leaf block credit
2187 *
2188 * bitmaps and block group descriptor blocks
2189 * and other metadat blocks still need to be
2190 * accounted.
2191 */
Mingming Cao525f4ed2008-08-19 22:15:58 -04002192 /* 1 bitmap, 1 block group descriptor */
Mingming Caoee12b632008-08-19 22:16:05 -04002193 ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb);
Aneesh Kumar K.V5887e982009-07-05 23:12:04 -04002194 return ret;
Mingming Caoee12b632008-08-19 22:16:05 -04002195 }
Alex Tomasa86c6182006-10-11 01:21:03 -07002196 }
2197
Mingming Cao525f4ed2008-08-19 22:15:58 -04002198 return ext4_chunk_trans_blocks(inode, nrblocks);
Mingming Caoee12b632008-08-19 22:16:05 -04002199}
Alex Tomasa86c6182006-10-11 01:21:03 -07002200
Mingming Caoee12b632008-08-19 22:16:05 -04002201/*
2202 * How many index/leaf blocks need to change/allocate to modify nrblocks?
2203 *
2204 * if nrblocks are fit in a single extent (chunk flag is 1), then
2205 * in the worse case, each tree level index/leaf need to be changed
2206 * if the tree split due to insert a new extent, then the old tree
2207 * index/leaf need to be updated too
2208 *
2209 * If the nrblocks are discontiguous, they could cause
2210 * the whole tree split more than once, but this is really rare.
2211 */
Mingming Cao525f4ed2008-08-19 22:15:58 -04002212int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
Mingming Caoee12b632008-08-19 22:16:05 -04002213{
2214 int index;
2215 int depth = ext_depth(inode);
Alex Tomasa86c6182006-10-11 01:21:03 -07002216
Mingming Caoee12b632008-08-19 22:16:05 -04002217 if (chunk)
2218 index = depth * 2;
2219 else
2220 index = depth * 3;
Alex Tomasa86c6182006-10-11 01:21:03 -07002221
Mingming Caoee12b632008-08-19 22:16:05 -04002222 return index;
Alex Tomasa86c6182006-10-11 01:21:03 -07002223}
2224
2225static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
2226 struct ext4_extent *ex,
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002227 ext4_lblk_t from, ext4_lblk_t to)
Alex Tomasa86c6182006-10-11 01:21:03 -07002228{
Amit Aroraa2df2a62007-07-17 21:42:41 -04002229 unsigned short ee_len = ext4_ext_get_actual_len(ex);
Theodore Ts'oe6362602009-11-23 07:17:05 -05002230 int flags = EXT4_FREE_BLOCKS_FORGET;
Alex Tomasa86c6182006-10-11 01:21:03 -07002231
Alex Tomasc9de5602008-01-29 00:19:52 -05002232 if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
Theodore Ts'oe6362602009-11-23 07:17:05 -05002233 flags |= EXT4_FREE_BLOCKS_METADATA;
Alex Tomasa86c6182006-10-11 01:21:03 -07002234#ifdef EXTENTS_STATS
2235 {
2236 struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
Alex Tomasa86c6182006-10-11 01:21:03 -07002237 spin_lock(&sbi->s_ext_stats_lock);
2238 sbi->s_ext_blocks += ee_len;
2239 sbi->s_ext_extents++;
2240 if (ee_len < sbi->s_ext_min)
2241 sbi->s_ext_min = ee_len;
2242 if (ee_len > sbi->s_ext_max)
2243 sbi->s_ext_max = ee_len;
2244 if (ext_depth(inode) > sbi->s_depth_max)
2245 sbi->s_depth_max = ext_depth(inode);
2246 spin_unlock(&sbi->s_ext_stats_lock);
2247 }
2248#endif
2249 if (from >= le32_to_cpu(ex->ee_block)
Amit Aroraa2df2a62007-07-17 21:42:41 -04002250 && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
Alex Tomasa86c6182006-10-11 01:21:03 -07002251 /* tail removal */
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002252 ext4_lblk_t num;
Alex Tomasf65e6fb2006-10-11 01:21:05 -07002253 ext4_fsblk_t start;
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002254
Amit Aroraa2df2a62007-07-17 21:42:41 -04002255 num = le32_to_cpu(ex->ee_block) + ee_len - from;
Theodore Ts'obf89d162010-10-27 21:30:14 -04002256 start = ext4_ext_pblock(ex) + ee_len - num;
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002257 ext_debug("free last %u blocks starting %llu\n", num, start);
Peter Huewe7dc57612011-02-21 21:01:42 -05002258 ext4_free_blocks(handle, inode, NULL, start, num, flags);
Alex Tomasa86c6182006-10-11 01:21:03 -07002259 } else if (from == le32_to_cpu(ex->ee_block)
Amit Aroraa2df2a62007-07-17 21:42:41 -04002260 && to <= le32_to_cpu(ex->ee_block) + ee_len - 1) {
Allison Hendersond583fb82011-05-25 07:41:43 -04002261 /* head removal */
2262 ext4_lblk_t num;
2263 ext4_fsblk_t start;
2264
2265 num = to - from;
2266 start = ext4_ext_pblock(ex);
2267
2268 ext_debug("free first %u blocks starting %llu\n", num, start);
2269 ext4_free_blocks(handle, inode, 0, start, num, flags);
2270
Alex Tomasa86c6182006-10-11 01:21:03 -07002271 } else {
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002272 printk(KERN_INFO "strange request: removal(2) "
2273 "%u-%u from %u:%u\n",
2274 from, to, le32_to_cpu(ex->ee_block), ee_len);
Alex Tomasa86c6182006-10-11 01:21:03 -07002275 }
2276 return 0;
2277}
2278
Allison Hendersond583fb82011-05-25 07:41:43 -04002279
2280/*
2281 * ext4_ext_rm_leaf() Removes the extents associated with the
2282 * blocks appearing between "start" and "end", and splits the extents
2283 * if "start" and "end" appear in the same extent
2284 *
2285 * @handle: The journal handle
2286 * @inode: The files inode
2287 * @path: The path to the leaf
2288 * @start: The first block to remove
2289 * @end: The last block to remove
2290 */
Alex Tomasa86c6182006-10-11 01:21:03 -07002291static int
2292ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
Allison Hendersond583fb82011-05-25 07:41:43 -04002293 struct ext4_ext_path *path, ext4_lblk_t start,
2294 ext4_lblk_t end)
Alex Tomasa86c6182006-10-11 01:21:03 -07002295{
2296 int err = 0, correct_index = 0;
2297 int depth = ext_depth(inode), credits;
2298 struct ext4_extent_header *eh;
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002299 ext4_lblk_t a, b, block;
2300 unsigned num;
2301 ext4_lblk_t ex_ee_block;
Alex Tomasa86c6182006-10-11 01:21:03 -07002302 unsigned short ex_ee_len;
Amit Aroraa2df2a62007-07-17 21:42:41 -04002303 unsigned uninitialized = 0;
Alex Tomasa86c6182006-10-11 01:21:03 -07002304 struct ext4_extent *ex;
Allison Hendersond583fb82011-05-25 07:41:43 -04002305 struct ext4_map_blocks map;
Alex Tomasa86c6182006-10-11 01:21:03 -07002306
Alex Tomasc29c0ae2007-07-18 09:19:09 -04002307 /* the header must be checked already in ext4_ext_remove_space() */
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002308 ext_debug("truncate since %u in leaf\n", start);
Alex Tomasa86c6182006-10-11 01:21:03 -07002309 if (!path[depth].p_hdr)
2310 path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
2311 eh = path[depth].p_hdr;
Frank Mayhar273df552010-03-02 11:46:09 -05002312 if (unlikely(path[depth].p_hdr == NULL)) {
2313 EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
2314 return -EIO;
2315 }
Alex Tomasa86c6182006-10-11 01:21:03 -07002316 /* find where to start removing */
2317 ex = EXT_LAST_EXTENT(eh);
2318
2319 ex_ee_block = le32_to_cpu(ex->ee_block);
Amit Aroraa2df2a62007-07-17 21:42:41 -04002320 ex_ee_len = ext4_ext_get_actual_len(ex);
Alex Tomasa86c6182006-10-11 01:21:03 -07002321
2322 while (ex >= EXT_FIRST_EXTENT(eh) &&
2323 ex_ee_block + ex_ee_len > start) {
Aneesh Kumar K.Va41f2072009-06-10 14:22:55 -04002324
2325 if (ext4_ext_is_uninitialized(ex))
2326 uninitialized = 1;
2327 else
2328 uninitialized = 0;
2329
Mingming553f9002009-09-18 13:34:55 -04002330 ext_debug("remove ext %u:[%d]%d\n", ex_ee_block,
2331 uninitialized, ex_ee_len);
Alex Tomasa86c6182006-10-11 01:21:03 -07002332 path[depth].p_ext = ex;
2333
2334 a = ex_ee_block > start ? ex_ee_block : start;
Allison Hendersond583fb82011-05-25 07:41:43 -04002335 b = ex_ee_block+ex_ee_len - 1 < end ?
2336 ex_ee_block+ex_ee_len - 1 : end;
Alex Tomasa86c6182006-10-11 01:21:03 -07002337
2338 ext_debug(" border %u:%u\n", a, b);
2339
Allison Hendersond583fb82011-05-25 07:41:43 -04002340 /* If this extent is beyond the end of the hole, skip it */
2341 if (end <= ex_ee_block) {
2342 ex--;
2343 ex_ee_block = le32_to_cpu(ex->ee_block);
2344 ex_ee_len = ext4_ext_get_actual_len(ex);
2345 continue;
2346 } else if (a != ex_ee_block &&
2347 b != ex_ee_block + ex_ee_len - 1) {
2348 /*
2349 * If this is a truncate, then this condition should
2350 * never happen because at least one of the end points
2351 * needs to be on the edge of the extent.
2352 */
Lukas Czernerf17722f2011-06-06 00:05:17 -04002353 if (end == EXT_MAX_BLOCKS - 1) {
Allison Hendersond583fb82011-05-25 07:41:43 -04002354 ext_debug(" bad truncate %u:%u\n",
2355 start, end);
2356 block = 0;
2357 num = 0;
2358 err = -EIO;
2359 goto out;
2360 }
2361 /*
2362 * else this is a hole punch, so the extent needs to
2363 * be split since neither edge of the hole is on the
2364 * extent edge
2365 */
2366 else{
2367 map.m_pblk = ext4_ext_pblock(ex);
2368 map.m_lblk = ex_ee_block;
2369 map.m_len = b - ex_ee_block;
2370
2371 err = ext4_split_extent(handle,
2372 inode, path, &map, 0,
2373 EXT4_GET_BLOCKS_PUNCH_OUT_EXT |
2374 EXT4_GET_BLOCKS_PRE_IO);
2375
2376 if (err < 0)
2377 goto out;
2378
2379 ex_ee_len = ext4_ext_get_actual_len(ex);
2380
2381 b = ex_ee_block+ex_ee_len - 1 < end ?
2382 ex_ee_block+ex_ee_len - 1 : end;
2383
2384 /* Then remove tail of this extent */
2385 block = ex_ee_block;
2386 num = a - block;
2387 }
Alex Tomasa86c6182006-10-11 01:21:03 -07002388 } else if (a != ex_ee_block) {
2389 /* remove tail of the extent */
2390 block = ex_ee_block;
2391 num = a - block;
2392 } else if (b != ex_ee_block + ex_ee_len - 1) {
2393 /* remove head of the extent */
Allison Hendersond583fb82011-05-25 07:41:43 -04002394 block = b;
2395 num = ex_ee_block + ex_ee_len - b;
2396
2397 /*
2398 * If this is a truncate, this condition
2399 * should never happen
2400 */
Lukas Czernerf17722f2011-06-06 00:05:17 -04002401 if (end == EXT_MAX_BLOCKS - 1) {
Allison Hendersond583fb82011-05-25 07:41:43 -04002402 ext_debug(" bad truncate %u:%u\n",
2403 start, end);
2404 err = -EIO;
2405 goto out;
2406 }
Alex Tomasa86c6182006-10-11 01:21:03 -07002407 } else {
2408 /* remove whole extent: excellent! */
2409 block = ex_ee_block;
2410 num = 0;
Allison Hendersond583fb82011-05-25 07:41:43 -04002411 if (a != ex_ee_block) {
2412 ext_debug(" bad truncate %u:%u\n",
2413 start, end);
2414 err = -EIO;
2415 goto out;
2416 }
2417
2418 if (b != ex_ee_block + ex_ee_len - 1) {
2419 ext_debug(" bad truncate %u:%u\n",
2420 start, end);
2421 err = -EIO;
2422 goto out;
2423 }
Alex Tomasa86c6182006-10-11 01:21:03 -07002424 }
2425
Theodore Ts'o34071da2008-08-01 21:59:19 -04002426 /*
2427 * 3 for leaf, sb, and inode plus 2 (bmap and group
2428 * descriptor) for each block group; assume two block
2429 * groups plus ex_ee_len/blocks_per_block_group for
2430 * the worst case
2431 */
2432 credits = 7 + 2*(ex_ee_len/EXT4_BLOCKS_PER_GROUP(inode->i_sb));
Alex Tomasa86c6182006-10-11 01:21:03 -07002433 if (ex == EXT_FIRST_EXTENT(eh)) {
2434 correct_index = 1;
2435 credits += (ext_depth(inode)) + 1;
2436 }
Dmitry Monakhov5aca07e2009-12-08 22:42:15 -05002437 credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);
Alex Tomasa86c6182006-10-11 01:21:03 -07002438
Jan Kara487caee2009-08-17 22:17:20 -04002439 err = ext4_ext_truncate_extend_restart(handle, inode, credits);
Shen Feng9102e4f2008-07-11 19:27:31 -04002440 if (err)
Alex Tomasa86c6182006-10-11 01:21:03 -07002441 goto out;
Alex Tomasa86c6182006-10-11 01:21:03 -07002442
2443 err = ext4_ext_get_access(handle, inode, path + depth);
2444 if (err)
2445 goto out;
2446
2447 err = ext4_remove_blocks(handle, inode, ex, a, b);
2448 if (err)
2449 goto out;
2450
2451 if (num == 0) {
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002452 /* this extent is removed; mark slot entirely unused */
Alex Tomasf65e6fb2006-10-11 01:21:05 -07002453 ext4_ext_store_pblock(ex, 0);
Allison Hendersond583fb82011-05-25 07:41:43 -04002454 } else if (block != ex_ee_block) {
2455 /*
2456 * If this was a head removal, then we need to update
2457 * the physical block since it is now at a different
2458 * location
2459 */
2460 ext4_ext_store_pblock(ex, ext4_ext_pblock(ex) + (b-a));
Alex Tomasa86c6182006-10-11 01:21:03 -07002461 }
2462
2463 ex->ee_block = cpu_to_le32(block);
2464 ex->ee_len = cpu_to_le16(num);
Amit Arora749269f2007-07-18 09:02:56 -04002465 /*
2466 * Do not mark uninitialized if all the blocks in the
2467 * extent have been removed.
2468 */
2469 if (uninitialized && num)
Amit Aroraa2df2a62007-07-17 21:42:41 -04002470 ext4_ext_mark_uninitialized(ex);
Alex Tomasa86c6182006-10-11 01:21:03 -07002471
2472 err = ext4_ext_dirty(handle, inode, path + depth);
2473 if (err)
2474 goto out;
2475
Allison Hendersond583fb82011-05-25 07:41:43 -04002476 /*
2477 * If the extent was completely released,
2478 * we need to remove it from the leaf
2479 */
2480 if (num == 0) {
Lukas Czernerf17722f2011-06-06 00:05:17 -04002481 if (end != EXT_MAX_BLOCKS - 1) {
Allison Hendersond583fb82011-05-25 07:41:43 -04002482 /*
2483 * For hole punching, we need to scoot all the
2484 * extents up when an extent is removed so that
2485 * we dont have blank extents in the middle
2486 */
2487 memmove(ex, ex+1, (EXT_LAST_EXTENT(eh) - ex) *
2488 sizeof(struct ext4_extent));
2489
2490 /* Now get rid of the one at the end */
2491 memset(EXT_LAST_EXTENT(eh), 0,
2492 sizeof(struct ext4_extent));
2493 }
2494 le16_add_cpu(&eh->eh_entries, -1);
2495 }
2496
Mingming Cao2ae02102006-10-11 01:21:11 -07002497 ext_debug("new extent: %u:%u:%llu\n", block, num,
Theodore Ts'obf89d162010-10-27 21:30:14 -04002498 ext4_ext_pblock(ex));
Alex Tomasa86c6182006-10-11 01:21:03 -07002499 ex--;
2500 ex_ee_block = le32_to_cpu(ex->ee_block);
Amit Aroraa2df2a62007-07-17 21:42:41 -04002501 ex_ee_len = ext4_ext_get_actual_len(ex);
Alex Tomasa86c6182006-10-11 01:21:03 -07002502 }
2503
2504 if (correct_index && eh->eh_entries)
2505 err = ext4_ext_correct_indexes(handle, inode, path);
2506
2507 /* if this leaf is free, then we should
2508 * remove it from index block above */
2509 if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL)
2510 err = ext4_ext_rm_idx(handle, inode, path + depth);
2511
2512out:
2513 return err;
2514}
2515
2516/*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002517 * ext4_ext_more_to_rm:
2518 * returns 1 if current index has to be freed (even partial)
Alex Tomasa86c6182006-10-11 01:21:03 -07002519 */
Avantika Mathur09b88252006-12-06 20:41:36 -08002520static int
Alex Tomasa86c6182006-10-11 01:21:03 -07002521ext4_ext_more_to_rm(struct ext4_ext_path *path)
2522{
2523 BUG_ON(path->p_idx == NULL);
2524
2525 if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr))
2526 return 0;
2527
2528 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002529 * if truncate on deeper level happened, it wasn't partial,
Alex Tomasa86c6182006-10-11 01:21:03 -07002530 * so we have to consider current index for truncation
2531 */
2532 if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block)
2533 return 0;
2534 return 1;
2535}
2536
Allison Hendersond583fb82011-05-25 07:41:43 -04002537static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
2538 ext4_lblk_t end)
Alex Tomasa86c6182006-10-11 01:21:03 -07002539{
2540 struct super_block *sb = inode->i_sb;
2541 int depth = ext_depth(inode);
2542 struct ext4_ext_path *path;
2543 handle_t *handle;
Dmitry Monakhov0617b832010-05-17 01:00:00 -04002544 int i, err;
Alex Tomasa86c6182006-10-11 01:21:03 -07002545
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002546 ext_debug("truncate since %u\n", start);
Alex Tomasa86c6182006-10-11 01:21:03 -07002547
2548 /* probably first extent we're gonna free will be last in block */
2549 handle = ext4_journal_start(inode, depth + 1);
2550 if (IS_ERR(handle))
2551 return PTR_ERR(handle);
2552
Dmitry Monakhov0617b832010-05-17 01:00:00 -04002553again:
Alex Tomasa86c6182006-10-11 01:21:03 -07002554 ext4_ext_invalidate_cache(inode);
2555
2556 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002557 * We start scanning from right side, freeing all the blocks
2558 * after i_size and walking into the tree depth-wise.
Alex Tomasa86c6182006-10-11 01:21:03 -07002559 */
Dmitry Monakhov0617b832010-05-17 01:00:00 -04002560 depth = ext_depth(inode);
Josef Bacik216553c2008-04-29 22:02:02 -04002561 path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_NOFS);
Alex Tomasa86c6182006-10-11 01:21:03 -07002562 if (path == NULL) {
2563 ext4_journal_stop(handle);
2564 return -ENOMEM;
2565 }
Dmitry Monakhov0617b832010-05-17 01:00:00 -04002566 path[0].p_depth = depth;
Alex Tomasa86c6182006-10-11 01:21:03 -07002567 path[0].p_hdr = ext_inode_hdr(inode);
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -04002568 if (ext4_ext_check(inode, path[0].p_hdr, depth)) {
Alex Tomasa86c6182006-10-11 01:21:03 -07002569 err = -EIO;
2570 goto out;
2571 }
Dmitry Monakhov0617b832010-05-17 01:00:00 -04002572 i = err = 0;
Alex Tomasa86c6182006-10-11 01:21:03 -07002573
2574 while (i >= 0 && err == 0) {
2575 if (i == depth) {
2576 /* this is leaf block */
Allison Hendersond583fb82011-05-25 07:41:43 -04002577 err = ext4_ext_rm_leaf(handle, inode, path,
2578 start, end);
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002579 /* root level has p_bh == NULL, brelse() eats this */
Alex Tomasa86c6182006-10-11 01:21:03 -07002580 brelse(path[i].p_bh);
2581 path[i].p_bh = NULL;
2582 i--;
2583 continue;
2584 }
2585
2586 /* this is index block */
2587 if (!path[i].p_hdr) {
2588 ext_debug("initialize header\n");
2589 path[i].p_hdr = ext_block_hdr(path[i].p_bh);
Alex Tomasa86c6182006-10-11 01:21:03 -07002590 }
2591
Alex Tomasa86c6182006-10-11 01:21:03 -07002592 if (!path[i].p_idx) {
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002593 /* this level hasn't been touched yet */
Alex Tomasa86c6182006-10-11 01:21:03 -07002594 path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
2595 path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
2596 ext_debug("init index ptr: hdr 0x%p, num %d\n",
2597 path[i].p_hdr,
2598 le16_to_cpu(path[i].p_hdr->eh_entries));
2599 } else {
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002600 /* we were already here, see at next index */
Alex Tomasa86c6182006-10-11 01:21:03 -07002601 path[i].p_idx--;
2602 }
2603
2604 ext_debug("level %d - index, first 0x%p, cur 0x%p\n",
2605 i, EXT_FIRST_INDEX(path[i].p_hdr),
2606 path[i].p_idx);
2607 if (ext4_ext_more_to_rm(path + i)) {
Alex Tomasc29c0ae2007-07-18 09:19:09 -04002608 struct buffer_head *bh;
Alex Tomasa86c6182006-10-11 01:21:03 -07002609 /* go to the next level */
Mingming Cao2ae02102006-10-11 01:21:11 -07002610 ext_debug("move to level %d (block %llu)\n",
Theodore Ts'obf89d162010-10-27 21:30:14 -04002611 i + 1, ext4_idx_pblock(path[i].p_idx));
Alex Tomasa86c6182006-10-11 01:21:03 -07002612 memset(path + i + 1, 0, sizeof(*path));
Theodore Ts'obf89d162010-10-27 21:30:14 -04002613 bh = sb_bread(sb, ext4_idx_pblock(path[i].p_idx));
Alex Tomasc29c0ae2007-07-18 09:19:09 -04002614 if (!bh) {
Alex Tomasa86c6182006-10-11 01:21:03 -07002615 /* should we reset i_size? */
2616 err = -EIO;
2617 break;
2618 }
Alex Tomasc29c0ae2007-07-18 09:19:09 -04002619 if (WARN_ON(i + 1 > depth)) {
2620 err = -EIO;
2621 break;
2622 }
Aneesh Kumar K.V56b19862009-03-12 09:51:20 -04002623 if (ext4_ext_check(inode, ext_block_hdr(bh),
Alex Tomasc29c0ae2007-07-18 09:19:09 -04002624 depth - i - 1)) {
2625 err = -EIO;
2626 break;
2627 }
2628 path[i + 1].p_bh = bh;
Alex Tomasa86c6182006-10-11 01:21:03 -07002629
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002630 /* save actual number of indexes since this
2631 * number is changed at the next iteration */
Alex Tomasa86c6182006-10-11 01:21:03 -07002632 path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries);
2633 i++;
2634 } else {
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002635 /* we finished processing this index, go up */
Alex Tomasa86c6182006-10-11 01:21:03 -07002636 if (path[i].p_hdr->eh_entries == 0 && i > 0) {
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002637 /* index is empty, remove it;
Alex Tomasa86c6182006-10-11 01:21:03 -07002638 * handle must be already prepared by the
2639 * truncatei_leaf() */
2640 err = ext4_ext_rm_idx(handle, inode, path + i);
2641 }
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002642 /* root level has p_bh == NULL, brelse() eats this */
Alex Tomasa86c6182006-10-11 01:21:03 -07002643 brelse(path[i].p_bh);
2644 path[i].p_bh = NULL;
2645 i--;
2646 ext_debug("return to level %d\n", i);
2647 }
2648 }
2649
2650 /* TODO: flexible tree reduction should be here */
2651 if (path->p_hdr->eh_entries == 0) {
2652 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07002653 * truncate to zero freed all the tree,
2654 * so we need to correct eh_depth
Alex Tomasa86c6182006-10-11 01:21:03 -07002655 */
2656 err = ext4_ext_get_access(handle, inode, path);
2657 if (err == 0) {
2658 ext_inode_hdr(inode)->eh_depth = 0;
2659 ext_inode_hdr(inode)->eh_max =
Theodore Ts'o55ad63b2009-08-28 10:40:33 -04002660 cpu_to_le16(ext4_ext_space_root(inode, 0));
Alex Tomasa86c6182006-10-11 01:21:03 -07002661 err = ext4_ext_dirty(handle, inode, path);
2662 }
2663 }
2664out:
Alex Tomasa86c6182006-10-11 01:21:03 -07002665 ext4_ext_drop_refs(path);
2666 kfree(path);
Dmitry Monakhov0617b832010-05-17 01:00:00 -04002667 if (err == -EAGAIN)
2668 goto again;
Alex Tomasa86c6182006-10-11 01:21:03 -07002669 ext4_journal_stop(handle);
2670
2671 return err;
2672}
2673
2674/*
2675 * called at mount time
2676 */
2677void ext4_ext_init(struct super_block *sb)
2678{
2679 /*
2680 * possible initialization would be here
2681 */
2682
Theodore Ts'o83982b62009-01-06 14:53:16 -05002683 if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) {
Theodore Ts'o90576c02009-09-29 15:51:30 -04002684#if defined(AGGRESSIVE_TEST) || defined(CHECK_BINSEARCH) || defined(EXTENTS_STATS)
Theodore Ts'o4776004f2008-09-08 23:00:52 -04002685 printk(KERN_INFO "EXT4-fs: file extents enabled");
Robert P. J. Daybbf2f9f2007-02-17 19:20:16 +01002686#ifdef AGGRESSIVE_TEST
2687 printk(", aggressive tests");
Alex Tomasa86c6182006-10-11 01:21:03 -07002688#endif
2689#ifdef CHECK_BINSEARCH
2690 printk(", check binsearch");
2691#endif
2692#ifdef EXTENTS_STATS
2693 printk(", stats");
2694#endif
2695 printk("\n");
Theodore Ts'o90576c02009-09-29 15:51:30 -04002696#endif
Alex Tomasa86c6182006-10-11 01:21:03 -07002697#ifdef EXTENTS_STATS
2698 spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock);
2699 EXT4_SB(sb)->s_ext_min = 1 << 30;
2700 EXT4_SB(sb)->s_ext_max = 0;
2701#endif
2702 }
2703}
2704
2705/*
2706 * called at umount time
2707 */
2708void ext4_ext_release(struct super_block *sb)
2709{
Theodore Ts'o83982b62009-01-06 14:53:16 -05002710 if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS))
Alex Tomasa86c6182006-10-11 01:21:03 -07002711 return;
2712
2713#ifdef EXTENTS_STATS
2714 if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) {
2715 struct ext4_sb_info *sbi = EXT4_SB(sb);
2716 printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n",
2717 sbi->s_ext_blocks, sbi->s_ext_extents,
2718 sbi->s_ext_blocks / sbi->s_ext_extents);
2719 printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n",
2720 sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max);
2721 }
2722#endif
2723}
2724
Aneesh Kumar K.V093a0882008-04-29 08:11:12 -04002725/* FIXME!! we need to try to merge to left or right after zero-out */
2726static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
2727{
Lukas Czerner24075182010-10-27 21:30:06 -04002728 ext4_fsblk_t ee_pblock;
2729 unsigned int ee_len;
Jing Zhangb7203032010-05-12 00:00:00 -04002730 int ret;
Aneesh Kumar K.V093a0882008-04-29 08:11:12 -04002731
Aneesh Kumar K.V093a0882008-04-29 08:11:12 -04002732 ee_len = ext4_ext_get_actual_len(ex);
Theodore Ts'obf89d162010-10-27 21:30:14 -04002733 ee_pblock = ext4_ext_pblock(ex);
Aneesh Kumar K.V093a0882008-04-29 08:11:12 -04002734
Theodore Ts'oa107e5a2010-10-27 23:44:47 -04002735 ret = sb_issue_zeroout(inode->i_sb, ee_pblock, ee_len, GFP_NOFS);
Lukas Czerner24075182010-10-27 21:30:06 -04002736 if (ret > 0)
2737 ret = 0;
Aneesh Kumar K.V093a0882008-04-29 08:11:12 -04002738
Lukas Czerner24075182010-10-27 21:30:06 -04002739 return ret;
Aneesh Kumar K.V093a0882008-04-29 08:11:12 -04002740}
2741
Yongqiang Yang47ea3bb2011-05-03 12:23:07 -04002742/*
2743 * used by extent splitting.
2744 */
2745#define EXT4_EXT_MAY_ZEROOUT 0x1 /* safe to zeroout if split fails \
2746 due to ENOSPC */
2747#define EXT4_EXT_MARK_UNINIT1 0x2 /* mark first half uninitialized */
2748#define EXT4_EXT_MARK_UNINIT2 0x4 /* mark second half uninitialized */
2749
2750/*
2751 * ext4_split_extent_at() splits an extent at given block.
2752 *
2753 * @handle: the journal handle
2754 * @inode: the file inode
2755 * @path: the path to the extent
2756 * @split: the logical block where the extent is splitted.
2757 * @split_flags: indicates if the extent could be zeroout if split fails, and
2758 * the states(init or uninit) of new extents.
2759 * @flags: flags used to insert new extent to extent tree.
2760 *
2761 *
2762 * Splits extent [a, b] into two extents [a, @split) and [@split, b], states
2763 * of which are deterimined by split_flag.
2764 *
2765 * There are two cases:
2766 * a> the extent are splitted into two extent.
2767 * b> split is not needed, and just mark the extent.
2768 *
2769 * return 0 on success.
2770 */
2771static int ext4_split_extent_at(handle_t *handle,
2772 struct inode *inode,
2773 struct ext4_ext_path *path,
2774 ext4_lblk_t split,
2775 int split_flag,
2776 int flags)
2777{
2778 ext4_fsblk_t newblock;
2779 ext4_lblk_t ee_block;
2780 struct ext4_extent *ex, newex, orig_ex;
2781 struct ext4_extent *ex2 = NULL;
2782 unsigned int ee_len, depth;
2783 int err = 0;
2784
2785 ext_debug("ext4_split_extents_at: inode %lu, logical"
2786 "block %llu\n", inode->i_ino, (unsigned long long)split);
2787
2788 ext4_ext_show_leaf(inode, path);
2789
2790 depth = ext_depth(inode);
2791 ex = path[depth].p_ext;
2792 ee_block = le32_to_cpu(ex->ee_block);
2793 ee_len = ext4_ext_get_actual_len(ex);
2794 newblock = split - ee_block + ext4_ext_pblock(ex);
2795
2796 BUG_ON(split < ee_block || split >= (ee_block + ee_len));
2797
2798 err = ext4_ext_get_access(handle, inode, path + depth);
2799 if (err)
2800 goto out;
2801
2802 if (split == ee_block) {
2803 /*
2804 * case b: block @split is the block that the extent begins with
2805 * then we just change the state of the extent, and splitting
2806 * is not needed.
2807 */
2808 if (split_flag & EXT4_EXT_MARK_UNINIT2)
2809 ext4_ext_mark_uninitialized(ex);
2810 else
2811 ext4_ext_mark_initialized(ex);
2812
2813 if (!(flags & EXT4_GET_BLOCKS_PRE_IO))
2814 ext4_ext_try_to_merge(inode, path, ex);
2815
2816 err = ext4_ext_dirty(handle, inode, path + depth);
2817 goto out;
2818 }
2819
2820 /* case a */
2821 memcpy(&orig_ex, ex, sizeof(orig_ex));
2822 ex->ee_len = cpu_to_le16(split - ee_block);
2823 if (split_flag & EXT4_EXT_MARK_UNINIT1)
2824 ext4_ext_mark_uninitialized(ex);
2825
2826 /*
2827 * path may lead to new leaf, not to original leaf any more
2828 * after ext4_ext_insert_extent() returns,
2829 */
2830 err = ext4_ext_dirty(handle, inode, path + depth);
2831 if (err)
2832 goto fix_extent_len;
2833
2834 ex2 = &newex;
2835 ex2->ee_block = cpu_to_le32(split);
2836 ex2->ee_len = cpu_to_le16(ee_len - (split - ee_block));
2837 ext4_ext_store_pblock(ex2, newblock);
2838 if (split_flag & EXT4_EXT_MARK_UNINIT2)
2839 ext4_ext_mark_uninitialized(ex2);
2840
2841 err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
2842 if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
2843 err = ext4_ext_zeroout(inode, &orig_ex);
2844 if (err)
2845 goto fix_extent_len;
2846 /* update the extent length and mark as initialized */
2847 ex->ee_len = cpu_to_le32(ee_len);
2848 ext4_ext_try_to_merge(inode, path, ex);
2849 err = ext4_ext_dirty(handle, inode, path + depth);
2850 goto out;
2851 } else if (err)
2852 goto fix_extent_len;
2853
2854out:
2855 ext4_ext_show_leaf(inode, path);
2856 return err;
2857
2858fix_extent_len:
2859 ex->ee_len = orig_ex.ee_len;
2860 ext4_ext_dirty(handle, inode, path + depth);
2861 return err;
2862}
2863
2864/*
2865 * ext4_split_extents() splits an extent and mark extent which is covered
2866 * by @map as split_flags indicates
2867 *
2868 * It may result in splitting the extent into multiple extents (upto three)
2869 * There are three possibilities:
2870 * a> There is no split required
2871 * b> Splits in two extents: Split is happening at either end of the extent
2872 * c> Splits in three extents: Somone is splitting in middle of the extent
2873 *
2874 */
2875static int ext4_split_extent(handle_t *handle,
2876 struct inode *inode,
2877 struct ext4_ext_path *path,
2878 struct ext4_map_blocks *map,
2879 int split_flag,
2880 int flags)
2881{
2882 ext4_lblk_t ee_block;
2883 struct ext4_extent *ex;
2884 unsigned int ee_len, depth;
2885 int err = 0;
2886 int uninitialized;
2887 int split_flag1, flags1;
2888
2889 depth = ext_depth(inode);
2890 ex = path[depth].p_ext;
2891 ee_block = le32_to_cpu(ex->ee_block);
2892 ee_len = ext4_ext_get_actual_len(ex);
2893 uninitialized = ext4_ext_is_uninitialized(ex);
2894
2895 if (map->m_lblk + map->m_len < ee_block + ee_len) {
2896 split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ?
2897 EXT4_EXT_MAY_ZEROOUT : 0;
2898 flags1 = flags | EXT4_GET_BLOCKS_PRE_IO;
2899 if (uninitialized)
2900 split_flag1 |= EXT4_EXT_MARK_UNINIT1 |
2901 EXT4_EXT_MARK_UNINIT2;
2902 err = ext4_split_extent_at(handle, inode, path,
2903 map->m_lblk + map->m_len, split_flag1, flags1);
Yongqiang Yang93917412011-05-22 20:49:12 -04002904 if (err)
2905 goto out;
Yongqiang Yang47ea3bb2011-05-03 12:23:07 -04002906 }
2907
2908 ext4_ext_drop_refs(path);
2909 path = ext4_ext_find_extent(inode, map->m_lblk, path);
2910 if (IS_ERR(path))
2911 return PTR_ERR(path);
2912
2913 if (map->m_lblk >= ee_block) {
2914 split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ?
2915 EXT4_EXT_MAY_ZEROOUT : 0;
2916 if (uninitialized)
2917 split_flag1 |= EXT4_EXT_MARK_UNINIT1;
2918 if (split_flag & EXT4_EXT_MARK_UNINIT2)
2919 split_flag1 |= EXT4_EXT_MARK_UNINIT2;
2920 err = ext4_split_extent_at(handle, inode, path,
2921 map->m_lblk, split_flag1, flags);
2922 if (err)
2923 goto out;
2924 }
2925
2926 ext4_ext_show_leaf(inode, path);
2927out:
2928 return err ? err : map->m_len;
2929}
2930
Aneesh Kumar K.V3977c962008-04-17 10:38:59 -04002931#define EXT4_EXT_ZERO_LEN 7
Amit Arora56055d32007-07-17 21:42:38 -04002932/*
Theodore Ts'oe35fd662010-05-16 19:00:00 -04002933 * This function is called by ext4_ext_map_blocks() if someone tries to write
Amit Arora56055d32007-07-17 21:42:38 -04002934 * to an uninitialized extent. It may result in splitting the uninitialized
Lucas De Marchi25985ed2011-03-30 22:57:33 -03002935 * extent into multiple extents (up to three - one initialized and two
Amit Arora56055d32007-07-17 21:42:38 -04002936 * uninitialized).
2937 * There are three possibilities:
2938 * a> There is no split required: Entire extent should be initialized
2939 * b> Splits in two extents: Write is happening at either end of the extent
2940 * c> Splits in three extents: Somone is writing in middle of the extent
2941 */
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002942static int ext4_ext_convert_to_initialized(handle_t *handle,
Theodore Ts'oe35fd662010-05-16 19:00:00 -04002943 struct inode *inode,
2944 struct ext4_map_blocks *map,
2945 struct ext4_ext_path *path)
Amit Arora56055d32007-07-17 21:42:38 -04002946{
Yongqiang Yang667eff32011-05-03 12:25:07 -04002947 struct ext4_map_blocks split_map;
2948 struct ext4_extent zero_ex;
2949 struct ext4_extent *ex;
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04002950 ext4_lblk_t ee_block, eof_block;
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05002951 unsigned int allocated, ee_len, depth;
Amit Arora56055d32007-07-17 21:42:38 -04002952 int err = 0;
Yongqiang Yang667eff32011-05-03 12:25:07 -04002953 int split_flag = 0;
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04002954
2955 ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
2956 "block %llu, max_blocks %u\n", inode->i_ino,
Theodore Ts'oe35fd662010-05-16 19:00:00 -04002957 (unsigned long long)map->m_lblk, map->m_len);
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04002958
2959 eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
2960 inode->i_sb->s_blocksize_bits;
Theodore Ts'oe35fd662010-05-16 19:00:00 -04002961 if (eof_block < map->m_lblk + map->m_len)
2962 eof_block = map->m_lblk + map->m_len;
Amit Arora56055d32007-07-17 21:42:38 -04002963
2964 depth = ext_depth(inode);
Amit Arora56055d32007-07-17 21:42:38 -04002965 ex = path[depth].p_ext;
2966 ee_block = le32_to_cpu(ex->ee_block);
2967 ee_len = ext4_ext_get_actual_len(ex);
Theodore Ts'oe35fd662010-05-16 19:00:00 -04002968 allocated = ee_len - (map->m_lblk - ee_block);
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04002969
Yongqiang Yang667eff32011-05-03 12:25:07 -04002970 WARN_ON(map->m_lblk < ee_block);
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04002971 /*
2972 * It is safe to convert extent to initialized via explicit
2973 * zeroout only if extent is fully insde i_size or new_size.
2974 */
Yongqiang Yang667eff32011-05-03 12:25:07 -04002975 split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04002976
Aneesh Kumar K.V3977c962008-04-17 10:38:59 -04002977 /* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */
Yongqiang Yang667eff32011-05-03 12:25:07 -04002978 if (ee_len <= 2*EXT4_EXT_ZERO_LEN &&
2979 (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
2980 err = ext4_ext_zeroout(inode, ex);
Aneesh Kumar K.V3977c962008-04-17 10:38:59 -04002981 if (err)
Aneesh Kumar K.Vd03856b2008-08-02 18:51:32 -04002982 goto out;
Aneesh Kumar K.Vd03856b2008-08-02 18:51:32 -04002983
2984 err = ext4_ext_get_access(handle, inode, path + depth);
2985 if (err)
2986 goto out;
Yongqiang Yang667eff32011-05-03 12:25:07 -04002987 ext4_ext_mark_initialized(ex);
2988 ext4_ext_try_to_merge(inode, path, ex);
2989 err = ext4_ext_dirty(handle, inode, path + depth);
2990 goto out;
Amit Arora56055d32007-07-17 21:42:38 -04002991 }
Yongqiang Yang667eff32011-05-03 12:25:07 -04002992
Amit Arora56055d32007-07-17 21:42:38 -04002993 /*
Yongqiang Yang667eff32011-05-03 12:25:07 -04002994 * four cases:
2995 * 1. split the extent into three extents.
2996 * 2. split the extent into two extents, zeroout the first half.
2997 * 3. split the extent into two extents, zeroout the second half.
2998 * 4. split the extent into two extents with out zeroout.
Amit Arora56055d32007-07-17 21:42:38 -04002999 */
Yongqiang Yang667eff32011-05-03 12:25:07 -04003000 split_map.m_lblk = map->m_lblk;
3001 split_map.m_len = map->m_len;
3002
3003 if (allocated > map->m_len) {
3004 if (allocated <= EXT4_EXT_ZERO_LEN &&
3005 (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
3006 /* case 3 */
3007 zero_ex.ee_block =
Allison Henderson9b940f82011-05-16 10:11:09 -04003008 cpu_to_le32(map->m_lblk);
3009 zero_ex.ee_len = cpu_to_le16(allocated);
Yongqiang Yang667eff32011-05-03 12:25:07 -04003010 ext4_ext_store_pblock(&zero_ex,
3011 ext4_ext_pblock(ex) + map->m_lblk - ee_block);
3012 err = ext4_ext_zeroout(inode, &zero_ex);
Amit Arora56055d32007-07-17 21:42:38 -04003013 if (err)
3014 goto out;
Yongqiang Yang667eff32011-05-03 12:25:07 -04003015 split_map.m_lblk = map->m_lblk;
3016 split_map.m_len = allocated;
3017 } else if ((map->m_lblk - ee_block + map->m_len <
3018 EXT4_EXT_ZERO_LEN) &&
3019 (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
3020 /* case 2 */
3021 if (map->m_lblk != ee_block) {
3022 zero_ex.ee_block = ex->ee_block;
3023 zero_ex.ee_len = cpu_to_le16(map->m_lblk -
3024 ee_block);
3025 ext4_ext_store_pblock(&zero_ex,
3026 ext4_ext_pblock(ex));
3027 err = ext4_ext_zeroout(inode, &zero_ex);
3028 if (err)
3029 goto out;
3030 }
3031
Yongqiang Yang667eff32011-05-03 12:25:07 -04003032 split_map.m_lblk = ee_block;
Allison Henderson9b940f82011-05-16 10:11:09 -04003033 split_map.m_len = map->m_lblk - ee_block + map->m_len;
3034 allocated = map->m_len;
Amit Arora56055d32007-07-17 21:42:38 -04003035 }
3036 }
Yongqiang Yang667eff32011-05-03 12:25:07 -04003037
3038 allocated = ext4_split_extent(handle, inode, path,
3039 &split_map, split_flag, 0);
3040 if (allocated < 0)
3041 err = allocated;
3042
Amit Arora56055d32007-07-17 21:42:38 -04003043out:
3044 return err ? err : allocated;
3045}
3046
Aneesh Kumar K.Vc278bfe2008-01-28 23:58:27 -05003047/*
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003048 * This function is called by ext4_ext_map_blocks() from
Mingming Cao00314622009-09-28 15:49:08 -04003049 * ext4_get_blocks_dio_write() when DIO to write
3050 * to an uninitialized extent.
3051 *
Paul Bollefd018fe2011-02-15 00:05:43 +01003052 * Writing to an uninitialized extent may result in splitting the uninitialized
Uwe Kleine-Königb5950762010-11-01 15:38:34 -04003053 * extent into multiple /initialized uninitialized extents (up to three)
Mingming Cao00314622009-09-28 15:49:08 -04003054 * There are three possibilities:
3055 * a> There is no split required: Entire extent should be uninitialized
3056 * b> Splits in two extents: Write is happening at either end of the extent
3057 * c> Splits in three extents: Somone is writing in middle of the extent
3058 *
3059 * One of more index blocks maybe needed if the extent tree grow after
Uwe Kleine-Königb5950762010-11-01 15:38:34 -04003060 * the uninitialized extent split. To prevent ENOSPC occur at the IO
Mingming Cao00314622009-09-28 15:49:08 -04003061 * complete, we need to split the uninitialized extent before DIO submit
Uwe Kleine-König421f91d2010-06-11 12:17:00 +02003062 * the IO. The uninitialized extent called at this time will be split
Mingming Cao00314622009-09-28 15:49:08 -04003063 * into three uninitialized extent(at most). After IO complete, the part
3064 * being filled will be convert to initialized by the end_io callback function
3065 * via ext4_convert_unwritten_extents().
Mingmingba230c32009-11-06 04:01:23 -05003066 *
3067 * Returns the size of uninitialized extent to be written on success.
Mingming Cao00314622009-09-28 15:49:08 -04003068 */
3069static int ext4_split_unwritten_extents(handle_t *handle,
3070 struct inode *inode,
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003071 struct ext4_map_blocks *map,
Mingming Cao00314622009-09-28 15:49:08 -04003072 struct ext4_ext_path *path,
Mingming Cao00314622009-09-28 15:49:08 -04003073 int flags)
3074{
Yongqiang Yang667eff32011-05-03 12:25:07 -04003075 ext4_lblk_t eof_block;
3076 ext4_lblk_t ee_block;
3077 struct ext4_extent *ex;
3078 unsigned int ee_len;
3079 int split_flag = 0, depth;
Mingming Cao00314622009-09-28 15:49:08 -04003080
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04003081 ext_debug("ext4_split_unwritten_extents: inode %lu, logical"
3082 "block %llu, max_blocks %u\n", inode->i_ino,
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003083 (unsigned long long)map->m_lblk, map->m_len);
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04003084
3085 eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
3086 inode->i_sb->s_blocksize_bits;
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003087 if (eof_block < map->m_lblk + map->m_len)
3088 eof_block = map->m_lblk + map->m_len;
Mingming Cao00314622009-09-28 15:49:08 -04003089 /*
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04003090 * It is safe to convert extent to initialized via explicit
3091 * zeroout only if extent is fully insde i_size or new_size.
3092 */
Yongqiang Yang667eff32011-05-03 12:25:07 -04003093 depth = ext_depth(inode);
3094 ex = path[depth].p_ext;
3095 ee_block = le32_to_cpu(ex->ee_block);
3096 ee_len = ext4_ext_get_actual_len(ex);
Dmitry Monakhov21ca0872010-05-16 06:00:00 -04003097
Yongqiang Yang667eff32011-05-03 12:25:07 -04003098 split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
3099 split_flag |= EXT4_EXT_MARK_UNINIT2;
Mingming Cao00314622009-09-28 15:49:08 -04003100
Yongqiang Yang667eff32011-05-03 12:25:07 -04003101 flags |= EXT4_GET_BLOCKS_PRE_IO;
3102 return ext4_split_extent(handle, inode, path, map, split_flag, flags);
Mingming Cao00314622009-09-28 15:49:08 -04003103}
Yongqiang Yang197217a2011-05-03 11:45:29 -04003104
Jiaying Zhangc7064ef2010-03-02 13:28:44 -05003105static int ext4_convert_unwritten_extents_endio(handle_t *handle,
Mingming Cao00314622009-09-28 15:49:08 -04003106 struct inode *inode,
3107 struct ext4_ext_path *path)
3108{
3109 struct ext4_extent *ex;
3110 struct ext4_extent_header *eh;
3111 int depth;
3112 int err = 0;
Mingming Cao00314622009-09-28 15:49:08 -04003113
3114 depth = ext_depth(inode);
3115 eh = path[depth].p_hdr;
3116 ex = path[depth].p_ext;
3117
Yongqiang Yang197217a2011-05-03 11:45:29 -04003118 ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical"
3119 "block %llu, max_blocks %u\n", inode->i_ino,
3120 (unsigned long long)le32_to_cpu(ex->ee_block),
3121 ext4_ext_get_actual_len(ex));
3122
Mingming Cao00314622009-09-28 15:49:08 -04003123 err = ext4_ext_get_access(handle, inode, path + depth);
3124 if (err)
3125 goto out;
3126 /* first mark the extent as initialized */
3127 ext4_ext_mark_initialized(ex);
3128
Yongqiang Yang197217a2011-05-03 11:45:29 -04003129 /* note: ext4_ext_correct_indexes() isn't needed here because
3130 * borders are not changed
Mingming Cao00314622009-09-28 15:49:08 -04003131 */
Yongqiang Yang197217a2011-05-03 11:45:29 -04003132 ext4_ext_try_to_merge(inode, path, ex);
3133
Mingming Cao00314622009-09-28 15:49:08 -04003134 /* Mark modified extent as dirty */
3135 err = ext4_ext_dirty(handle, inode, path + depth);
3136out:
3137 ext4_ext_show_leaf(inode, path);
3138 return err;
3139}
3140
Aneesh Kumar K.V515f41c2009-12-29 23:39:06 -05003141static void unmap_underlying_metadata_blocks(struct block_device *bdev,
3142 sector_t block, int count)
3143{
3144 int i;
3145 for (i = 0; i < count; i++)
3146 unmap_underlying_metadata(bdev, block + i);
3147}
3148
Theodore Ts'o58590b02010-10-27 21:23:12 -04003149/*
3150 * Handle EOFBLOCKS_FL flag, clearing it if necessary
3151 */
3152static int check_eofblocks_fl(handle_t *handle, struct inode *inode,
Eric Sandeend002ebf2011-01-10 13:03:35 -05003153 ext4_lblk_t lblk,
Theodore Ts'o58590b02010-10-27 21:23:12 -04003154 struct ext4_ext_path *path,
3155 unsigned int len)
3156{
3157 int i, depth;
3158 struct ext4_extent_header *eh;
Sergey Senozhatsky65922cb2011-03-23 14:08:27 -04003159 struct ext4_extent *last_ex;
Theodore Ts'o58590b02010-10-27 21:23:12 -04003160
3161 if (!ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))
3162 return 0;
3163
3164 depth = ext_depth(inode);
3165 eh = path[depth].p_hdr;
Theodore Ts'o58590b02010-10-27 21:23:12 -04003166
3167 if (unlikely(!eh->eh_entries)) {
3168 EXT4_ERROR_INODE(inode, "eh->eh_entries == 0 and "
3169 "EOFBLOCKS_FL set");
3170 return -EIO;
3171 }
3172 last_ex = EXT_LAST_EXTENT(eh);
3173 /*
3174 * We should clear the EOFBLOCKS_FL flag if we are writing the
3175 * last block in the last extent in the file. We test this by
3176 * first checking to see if the caller to
3177 * ext4_ext_get_blocks() was interested in the last block (or
3178 * a block beyond the last block) in the current extent. If
3179 * this turns out to be false, we can bail out from this
3180 * function immediately.
3181 */
Eric Sandeend002ebf2011-01-10 13:03:35 -05003182 if (lblk + len < le32_to_cpu(last_ex->ee_block) +
Theodore Ts'o58590b02010-10-27 21:23:12 -04003183 ext4_ext_get_actual_len(last_ex))
3184 return 0;
3185 /*
3186 * If the caller does appear to be planning to write at or
3187 * beyond the end of the current extent, we then test to see
3188 * if the current extent is the last extent in the file, by
3189 * checking to make sure it was reached via the rightmost node
3190 * at each level of the tree.
3191 */
3192 for (i = depth-1; i >= 0; i--)
3193 if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
3194 return 0;
3195 ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
3196 return ext4_mark_inode_dirty(handle, inode);
3197}
3198
Mingming Cao00314622009-09-28 15:49:08 -04003199static int
3200ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode,
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003201 struct ext4_map_blocks *map,
Mingming Cao00314622009-09-28 15:49:08 -04003202 struct ext4_ext_path *path, int flags,
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003203 unsigned int allocated, ext4_fsblk_t newblock)
Mingming Cao00314622009-09-28 15:49:08 -04003204{
3205 int ret = 0;
3206 int err = 0;
Mingming Cao8d5d02e2009-09-28 15:48:29 -04003207 ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
Mingming Cao00314622009-09-28 15:49:08 -04003208
3209 ext_debug("ext4_ext_handle_uninitialized_extents: inode %lu, logical"
3210 "block %llu, max_blocks %u, flags %d, allocated %u",
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003211 inode->i_ino, (unsigned long long)map->m_lblk, map->m_len,
Mingming Cao00314622009-09-28 15:49:08 -04003212 flags, allocated);
3213 ext4_ext_show_leaf(inode, path);
3214
Jiaying Zhangc7064ef2010-03-02 13:28:44 -05003215 /* get_block() before submit the IO, split the extent */
Jiaying Zhang744692d2010-03-04 16:14:02 -05003216 if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003217 ret = ext4_split_unwritten_extents(handle, inode, map,
3218 path, flags);
Mingming5f524952009-11-10 10:48:04 -05003219 /*
3220 * Flag the inode(non aio case) or end_io struct (aio case)
Lucas De Marchi25985ed2011-03-30 22:57:33 -03003221 * that this IO needs to conversion to written when IO is
Mingming5f524952009-11-10 10:48:04 -05003222 * completed
3223 */
Eric Sandeene9e3bce2011-02-12 08:17:34 -05003224 if (io && !(io->flag & EXT4_IO_END_UNWRITTEN)) {
Theodore Ts'obd2d0212010-10-27 21:30:10 -04003225 io->flag = EXT4_IO_END_UNWRITTEN;
Eric Sandeene9e3bce2011-02-12 08:17:34 -05003226 atomic_inc(&EXT4_I(inode)->i_aiodio_unwritten);
3227 } else
Theodore Ts'o19f5fb72010-01-24 14:34:07 -05003228 ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN);
Jiaying Zhang744692d2010-03-04 16:14:02 -05003229 if (ext4_should_dioread_nolock(inode))
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003230 map->m_flags |= EXT4_MAP_UNINIT;
Mingming Cao00314622009-09-28 15:49:08 -04003231 goto out;
3232 }
Jiaying Zhangc7064ef2010-03-02 13:28:44 -05003233 /* IO end_io complete, convert the filled extent to written */
Jiaying Zhang744692d2010-03-04 16:14:02 -05003234 if ((flags & EXT4_GET_BLOCKS_CONVERT)) {
Jiaying Zhangc7064ef2010-03-02 13:28:44 -05003235 ret = ext4_convert_unwritten_extents_endio(handle, inode,
Mingming Cao00314622009-09-28 15:49:08 -04003236 path);
Theodore Ts'o58590b02010-10-27 21:23:12 -04003237 if (ret >= 0) {
Jan Karab436b9b2009-12-08 23:51:10 -05003238 ext4_update_inode_fsync_trans(handle, inode, 1);
Eric Sandeend002ebf2011-01-10 13:03:35 -05003239 err = check_eofblocks_fl(handle, inode, map->m_lblk,
3240 path, map->m_len);
Theodore Ts'o58590b02010-10-27 21:23:12 -04003241 } else
3242 err = ret;
Mingming Cao00314622009-09-28 15:49:08 -04003243 goto out2;
3244 }
3245 /* buffered IO case */
3246 /*
3247 * repeat fallocate creation request
3248 * we already have an unwritten extent
3249 */
3250 if (flags & EXT4_GET_BLOCKS_UNINIT_EXT)
3251 goto map_out;
3252
3253 /* buffered READ or buffered write_begin() lookup */
3254 if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3255 /*
3256 * We have blocks reserved already. We
3257 * return allocated blocks so that delalloc
3258 * won't do block reservation for us. But
3259 * the buffer head will be unmapped so that
3260 * a read from the block returns 0s.
3261 */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003262 map->m_flags |= EXT4_MAP_UNWRITTEN;
Mingming Cao00314622009-09-28 15:49:08 -04003263 goto out1;
3264 }
3265
3266 /* buffered write, writepage time, convert*/
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003267 ret = ext4_ext_convert_to_initialized(handle, inode, map, path);
Theodore Ts'o58590b02010-10-27 21:23:12 -04003268 if (ret >= 0) {
Jan Karab436b9b2009-12-08 23:51:10 -05003269 ext4_update_inode_fsync_trans(handle, inode, 1);
Eric Sandeend002ebf2011-01-10 13:03:35 -05003270 err = check_eofblocks_fl(handle, inode, map->m_lblk, path,
3271 map->m_len);
Theodore Ts'o58590b02010-10-27 21:23:12 -04003272 if (err < 0)
3273 goto out2;
3274 }
3275
Mingming Cao00314622009-09-28 15:49:08 -04003276out:
3277 if (ret <= 0) {
3278 err = ret;
3279 goto out2;
3280 } else
3281 allocated = ret;
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003282 map->m_flags |= EXT4_MAP_NEW;
Aneesh Kumar K.V515f41c2009-12-29 23:39:06 -05003283 /*
3284 * if we allocated more blocks than requested
3285 * we need to make sure we unmap the extra block
3286 * allocated. The actual needed block will get
3287 * unmapped later when we find the buffer_head marked
3288 * new.
3289 */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003290 if (allocated > map->m_len) {
Aneesh Kumar K.V515f41c2009-12-29 23:39:06 -05003291 unmap_underlying_metadata_blocks(inode->i_sb->s_bdev,
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003292 newblock + map->m_len,
3293 allocated - map->m_len);
3294 allocated = map->m_len;
Aneesh Kumar K.V515f41c2009-12-29 23:39:06 -05003295 }
Aneesh Kumar K.V5f634d02010-01-25 04:00:31 -05003296
3297 /*
3298 * If we have done fallocate with the offset that is already
3299 * delayed allocated, we would have block reservation
3300 * and quota reservation done in the delayed write path.
3301 * But fallocate would have already updated quota and block
3302 * count for this offset. So cancel these reservation
3303 */
Aneesh Kumar K.V1296cc82010-01-15 01:27:59 -05003304 if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
Aneesh Kumar K.V5f634d02010-01-25 04:00:31 -05003305 ext4_da_update_reserve_space(inode, allocated, 0);
3306
Mingming Cao00314622009-09-28 15:49:08 -04003307map_out:
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003308 map->m_flags |= EXT4_MAP_MAPPED;
Mingming Cao00314622009-09-28 15:49:08 -04003309out1:
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003310 if (allocated > map->m_len)
3311 allocated = map->m_len;
Mingming Cao00314622009-09-28 15:49:08 -04003312 ext4_ext_show_leaf(inode, path);
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003313 map->m_pblk = newblock;
3314 map->m_len = allocated;
Mingming Cao00314622009-09-28 15:49:08 -04003315out2:
3316 if (path) {
3317 ext4_ext_drop_refs(path);
3318 kfree(path);
3319 }
3320 return err ? err : allocated;
3321}
Theodore Ts'o58590b02010-10-27 21:23:12 -04003322
Mingming Cao00314622009-09-28 15:49:08 -04003323/*
Mingming Caof5ab0d12008-02-25 15:29:55 -05003324 * Block allocation/map/preallocation routine for extents based files
3325 *
3326 *
Aneesh Kumar K.Vc278bfe2008-01-28 23:58:27 -05003327 * Need to be called with
Aneesh Kumar K.V0e855ac2008-01-28 23:58:26 -05003328 * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block
3329 * (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem)
Mingming Caof5ab0d12008-02-25 15:29:55 -05003330 *
3331 * return > 0, number of of blocks already mapped/allocated
3332 * if create == 0 and these are pre-allocated blocks
3333 * buffer head is unmapped
3334 * otherwise blocks are mapped
3335 *
3336 * return = 0, if plain look up failed (blocks have not been allocated)
3337 * buffer head is unmapped
3338 *
3339 * return < 0, error case.
Aneesh Kumar K.Vc278bfe2008-01-28 23:58:27 -05003340 */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003341int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
3342 struct ext4_map_blocks *map, int flags)
Alex Tomasa86c6182006-10-11 01:21:03 -07003343{
3344 struct ext4_ext_path *path = NULL;
Theodore Ts'o58590b02010-10-27 21:23:12 -04003345 struct ext4_extent newex, *ex;
Jiaying Zhang0562e0b2011-03-21 21:38:05 -04003346 ext4_fsblk_t newblock = 0;
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05003347 int err = 0, depth, ret;
Theodore Ts'o498e5f22008-11-05 00:14:04 -05003348 unsigned int allocated = 0;
Allison Hendersone8613042011-05-25 07:41:46 -04003349 unsigned int punched_out = 0;
3350 unsigned int result = 0;
Alex Tomasc9de5602008-01-29 00:19:52 -05003351 struct ext4_allocation_request ar;
Mingming Cao8d5d02e2009-09-28 15:48:29 -04003352 ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
Allison Hendersone8613042011-05-25 07:41:46 -04003353 struct ext4_map_blocks punch_map;
Alex Tomasa86c6182006-10-11 01:21:03 -07003354
Mingming84fe3be2009-09-01 08:44:37 -04003355 ext_debug("blocks %u/%u requested for inode %lu\n",
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003356 map->m_lblk, map->m_len, inode->i_ino);
Jiaying Zhang0562e0b2011-03-21 21:38:05 -04003357 trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags);
Alex Tomasa86c6182006-10-11 01:21:03 -07003358
3359 /* check in cache */
Allison Hendersone8613042011-05-25 07:41:46 -04003360 if (ext4_ext_in_cache(inode, map->m_lblk, &newex) &&
3361 ((flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) == 0)) {
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05003362 if (!newex.ee_start_lo && !newex.ee_start_hi) {
Theodore Ts'oc2177052009-05-14 00:58:52 -04003363 if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
Amit Arora56055d32007-07-17 21:42:38 -04003364 /*
3365 * block isn't allocated yet and
3366 * user doesn't want to allocate it
3367 */
Alex Tomasa86c6182006-10-11 01:21:03 -07003368 goto out2;
3369 }
3370 /* we should allocate requested block */
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05003371 } else {
Alex Tomasa86c6182006-10-11 01:21:03 -07003372 /* block is already allocated */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003373 newblock = map->m_lblk
Dave Kleikamp8c55e202007-05-24 13:04:54 -04003374 - le32_to_cpu(newex.ee_block)
Theodore Ts'obf89d162010-10-27 21:30:14 -04003375 + ext4_ext_pblock(&newex);
Randy Dunlapd0d856e2006-10-11 01:21:07 -07003376 /* number of remaining blocks in the extent */
Aneesh Kumar K.Vb939e372008-01-28 23:58:27 -05003377 allocated = ext4_ext_get_actual_len(&newex) -
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003378 (map->m_lblk - le32_to_cpu(newex.ee_block));
Alex Tomasa86c6182006-10-11 01:21:03 -07003379 goto out;
Alex Tomasa86c6182006-10-11 01:21:03 -07003380 }
3381 }
3382
3383 /* find extent for this block */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003384 path = ext4_ext_find_extent(inode, map->m_lblk, NULL);
Alex Tomasa86c6182006-10-11 01:21:03 -07003385 if (IS_ERR(path)) {
3386 err = PTR_ERR(path);
3387 path = NULL;
3388 goto out2;
3389 }
3390
3391 depth = ext_depth(inode);
3392
3393 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07003394 * consistent leaf must not be empty;
3395 * this situation is possible, though, _during_ tree modification;
Alex Tomasa86c6182006-10-11 01:21:03 -07003396 * this is why assert can't be put in ext4_ext_find_extent()
3397 */
Frank Mayhar273df552010-03-02 11:46:09 -05003398 if (unlikely(path[depth].p_ext == NULL && depth != 0)) {
3399 EXT4_ERROR_INODE(inode, "bad extent address "
Theodore Ts'of70f3622010-05-16 23:00:00 -04003400 "lblock: %lu, depth: %d pblock %lld",
3401 (unsigned long) map->m_lblk, depth,
3402 path[depth].p_block);
Surbhi Palande034fb4c2009-12-14 09:53:52 -05003403 err = -EIO;
3404 goto out2;
3405 }
Alex Tomasa86c6182006-10-11 01:21:03 -07003406
Avantika Mathur7e028972006-12-06 20:41:33 -08003407 ex = path[depth].p_ext;
3408 if (ex) {
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05003409 ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
Theodore Ts'obf89d162010-10-27 21:30:14 -04003410 ext4_fsblk_t ee_start = ext4_ext_pblock(ex);
Amit Aroraa2df2a62007-07-17 21:42:41 -04003411 unsigned short ee_len;
Suparna Bhattacharya471d4012006-10-11 01:21:06 -07003412
3413 /*
Suparna Bhattacharya471d4012006-10-11 01:21:06 -07003414 * Uninitialized extents are treated as holes, except that
Amit Arora56055d32007-07-17 21:42:38 -04003415 * we split out initialized portions during a write.
Suparna Bhattacharya471d4012006-10-11 01:21:06 -07003416 */
Amit Aroraa2df2a62007-07-17 21:42:41 -04003417 ee_len = ext4_ext_get_actual_len(ex);
Randy Dunlapd0d856e2006-10-11 01:21:07 -07003418 /* if found extent covers block, simply return it */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003419 if (in_range(map->m_lblk, ee_block, ee_len)) {
3420 newblock = map->m_lblk - ee_block + ee_start;
Randy Dunlapd0d856e2006-10-11 01:21:07 -07003421 /* number of remaining blocks in the extent */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003422 allocated = ee_len - (map->m_lblk - ee_block);
3423 ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk,
3424 ee_block, ee_len, newblock);
Amit Arora56055d32007-07-17 21:42:38 -04003425
Allison Hendersone8613042011-05-25 07:41:46 -04003426 if ((flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) == 0) {
3427 /*
3428 * Do not put uninitialized extent
3429 * in the cache
3430 */
3431 if (!ext4_ext_is_uninitialized(ex)) {
3432 ext4_ext_put_in_cache(inode, ee_block,
3433 ee_len, ee_start);
3434 goto out;
3435 }
3436 ret = ext4_ext_handle_uninitialized_extents(
3437 handle, inode, map, path, flags,
3438 allocated, newblock);
3439 return ret;
Amit Arora56055d32007-07-17 21:42:38 -04003440 }
Allison Hendersone8613042011-05-25 07:41:46 -04003441
3442 /*
3443 * Punch out the map length, but only to the
3444 * end of the extent
3445 */
3446 punched_out = allocated < map->m_len ?
3447 allocated : map->m_len;
3448
3449 /*
3450 * Sense extents need to be converted to
3451 * uninitialized, they must fit in an
3452 * uninitialized extent
3453 */
3454 if (punched_out > EXT_UNINIT_MAX_LEN)
3455 punched_out = EXT_UNINIT_MAX_LEN;
3456
3457 punch_map.m_lblk = map->m_lblk;
3458 punch_map.m_pblk = newblock;
3459 punch_map.m_len = punched_out;
3460 punch_map.m_flags = 0;
3461
3462 /* Check to see if the extent needs to be split */
3463 if (punch_map.m_len != ee_len ||
3464 punch_map.m_lblk != ee_block) {
3465
3466 ret = ext4_split_extent(handle, inode,
3467 path, &punch_map, 0,
3468 EXT4_GET_BLOCKS_PUNCH_OUT_EXT |
3469 EXT4_GET_BLOCKS_PRE_IO);
3470
3471 if (ret < 0) {
3472 err = ret;
3473 goto out2;
3474 }
3475 /*
3476 * find extent for the block at
3477 * the start of the hole
3478 */
3479 ext4_ext_drop_refs(path);
3480 kfree(path);
3481
3482 path = ext4_ext_find_extent(inode,
3483 map->m_lblk, NULL);
3484 if (IS_ERR(path)) {
3485 err = PTR_ERR(path);
3486 path = NULL;
3487 goto out2;
3488 }
3489
3490 depth = ext_depth(inode);
3491 ex = path[depth].p_ext;
3492 ee_len = ext4_ext_get_actual_len(ex);
3493 ee_block = le32_to_cpu(ex->ee_block);
3494 ee_start = ext4_ext_pblock(ex);
3495
3496 }
3497
3498 ext4_ext_mark_uninitialized(ex);
3499
3500 err = ext4_ext_remove_space(inode, map->m_lblk,
3501 map->m_lblk + punched_out);
3502
3503 goto out2;
Alex Tomasa86c6182006-10-11 01:21:03 -07003504 }
3505 }
3506
3507 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07003508 * requested block isn't allocated yet;
Alex Tomasa86c6182006-10-11 01:21:03 -07003509 * we couldn't try to create block if create flag is zero
3510 */
Theodore Ts'oc2177052009-05-14 00:58:52 -04003511 if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
Amit Arora56055d32007-07-17 21:42:38 -04003512 /*
3513 * put just found gap into cache to speed up
3514 * subsequent requests
3515 */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003516 ext4_ext_put_gap_in_cache(inode, path, map->m_lblk);
Alex Tomasa86c6182006-10-11 01:21:03 -07003517 goto out2;
3518 }
3519 /*
Theodore Ts'oc2ea3fd2008-10-10 09:40:52 -04003520 * Okay, we need to do block allocation.
Andrew Morton63f57932006-10-11 01:21:24 -07003521 */
Alex Tomasa86c6182006-10-11 01:21:03 -07003522
Alex Tomasc9de5602008-01-29 00:19:52 -05003523 /* find neighbour allocated blocks */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003524 ar.lleft = map->m_lblk;
Alex Tomasc9de5602008-01-29 00:19:52 -05003525 err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
3526 if (err)
3527 goto out2;
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003528 ar.lright = map->m_lblk;
Alex Tomasc9de5602008-01-29 00:19:52 -05003529 err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright);
3530 if (err)
3531 goto out2;
Amit Arora25d14f92007-05-24 13:04:13 -04003532
Amit Arora749269f2007-07-18 09:02:56 -04003533 /*
3534 * See if request is beyond maximum number of blocks we can have in
3535 * a single extent. For an initialized extent this limit is
3536 * EXT_INIT_MAX_LEN and for an uninitialized extent this limit is
3537 * EXT_UNINIT_MAX_LEN.
3538 */
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003539 if (map->m_len > EXT_INIT_MAX_LEN &&
Theodore Ts'oc2177052009-05-14 00:58:52 -04003540 !(flags & EXT4_GET_BLOCKS_UNINIT_EXT))
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003541 map->m_len = EXT_INIT_MAX_LEN;
3542 else if (map->m_len > EXT_UNINIT_MAX_LEN &&
Theodore Ts'oc2177052009-05-14 00:58:52 -04003543 (flags & EXT4_GET_BLOCKS_UNINIT_EXT))
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003544 map->m_len = EXT_UNINIT_MAX_LEN;
Amit Arora749269f2007-07-18 09:02:56 -04003545
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003546 /* Check if we can really insert (m_lblk)::(m_lblk + m_len) extent */
3547 newex.ee_block = cpu_to_le32(map->m_lblk);
3548 newex.ee_len = cpu_to_le16(map->m_len);
Amit Arora25d14f92007-05-24 13:04:13 -04003549 err = ext4_ext_check_overlap(inode, &newex, path);
3550 if (err)
Aneesh Kumar K.Vb939e372008-01-28 23:58:27 -05003551 allocated = ext4_ext_get_actual_len(&newex);
Amit Arora25d14f92007-05-24 13:04:13 -04003552 else
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003553 allocated = map->m_len;
Alex Tomasc9de5602008-01-29 00:19:52 -05003554
3555 /* allocate new block */
3556 ar.inode = inode;
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003557 ar.goal = ext4_ext_find_goal(inode, path, map->m_lblk);
3558 ar.logical = map->m_lblk;
Alex Tomasc9de5602008-01-29 00:19:52 -05003559 ar.len = allocated;
3560 if (S_ISREG(inode->i_mode))
3561 ar.flags = EXT4_MB_HINT_DATA;
3562 else
3563 /* disable in-core preallocation for non-regular files */
3564 ar.flags = 0;
Vivek Haldar556b27a2011-05-25 07:41:54 -04003565 if (flags & EXT4_GET_BLOCKS_NO_NORMALIZE)
3566 ar.flags |= EXT4_MB_HINT_NOPREALLOC;
Alex Tomasc9de5602008-01-29 00:19:52 -05003567 newblock = ext4_mb_new_blocks(handle, &ar, &err);
Alex Tomasa86c6182006-10-11 01:21:03 -07003568 if (!newblock)
3569 goto out2;
Mingming84fe3be2009-09-01 08:44:37 -04003570 ext_debug("allocate new block: goal %llu, found %llu/%u\n",
Theodore Ts'o498e5f22008-11-05 00:14:04 -05003571 ar.goal, newblock, allocated);
Alex Tomasa86c6182006-10-11 01:21:03 -07003572
3573 /* try to insert new extent into found leaf and return */
Alex Tomasf65e6fb2006-10-11 01:21:05 -07003574 ext4_ext_store_pblock(&newex, newblock);
Alex Tomasc9de5602008-01-29 00:19:52 -05003575 newex.ee_len = cpu_to_le16(ar.len);
Mingming Cao8d5d02e2009-09-28 15:48:29 -04003576 /* Mark uninitialized */
3577 if (flags & EXT4_GET_BLOCKS_UNINIT_EXT){
Amit Aroraa2df2a62007-07-17 21:42:41 -04003578 ext4_ext_mark_uninitialized(&newex);
Mingming Cao8d5d02e2009-09-28 15:48:29 -04003579 /*
Jiaying Zhang744692d2010-03-04 16:14:02 -05003580 * io_end structure was created for every IO write to an
Lucas De Marchi25985ed2011-03-30 22:57:33 -03003581 * uninitialized extent. To avoid unnecessary conversion,
Jiaying Zhang744692d2010-03-04 16:14:02 -05003582 * here we flag the IO that really needs the conversion.
Mingming5f524952009-11-10 10:48:04 -05003583 * For non asycn direct IO case, flag the inode state
Lucas De Marchi25985ed2011-03-30 22:57:33 -03003584 * that we need to perform conversion when IO is done.
Mingming Cao8d5d02e2009-09-28 15:48:29 -04003585 */
Jiaying Zhang744692d2010-03-04 16:14:02 -05003586 if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
Eric Sandeene9e3bce2011-02-12 08:17:34 -05003587 if (io && !(io->flag & EXT4_IO_END_UNWRITTEN)) {
Theodore Ts'obd2d0212010-10-27 21:30:10 -04003588 io->flag = EXT4_IO_END_UNWRITTEN;
Eric Sandeene9e3bce2011-02-12 08:17:34 -05003589 atomic_inc(&EXT4_I(inode)->i_aiodio_unwritten);
3590 } else
Theodore Ts'o19f5fb72010-01-24 14:34:07 -05003591 ext4_set_inode_state(inode,
3592 EXT4_STATE_DIO_UNWRITTEN);
Mingming5f524952009-11-10 10:48:04 -05003593 }
Jiaying Zhang744692d2010-03-04 16:14:02 -05003594 if (ext4_should_dioread_nolock(inode))
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003595 map->m_flags |= EXT4_MAP_UNINIT;
Mingming Cao8d5d02e2009-09-28 15:48:29 -04003596 }
Jiaying Zhangc8d46e42010-02-24 09:52:53 -05003597
Eric Sandeend002ebf2011-01-10 13:03:35 -05003598 err = check_eofblocks_fl(handle, inode, map->m_lblk, path, ar.len);
Theodore Ts'o58590b02010-10-27 21:23:12 -04003599 if (err)
3600 goto out2;
3601
Mingming Cao00314622009-09-28 15:49:08 -04003602 err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
Alex Tomas315054f2007-05-24 13:04:25 -04003603 if (err) {
3604 /* free data blocks we just allocated */
Alex Tomasc9de5602008-01-29 00:19:52 -05003605 /* not a good idea to call discard here directly,
3606 * but otherwise we'd need to call it every free() */
Theodore Ts'oc2ea3fd2008-10-10 09:40:52 -04003607 ext4_discard_preallocations(inode);
Peter Huewe7dc57612011-02-21 21:01:42 -05003608 ext4_free_blocks(handle, inode, NULL, ext4_ext_pblock(&newex),
Theodore Ts'oe6362602009-11-23 07:17:05 -05003609 ext4_ext_get_actual_len(&newex), 0);
Alex Tomasa86c6182006-10-11 01:21:03 -07003610 goto out2;
Alex Tomas315054f2007-05-24 13:04:25 -04003611 }
Alex Tomasa86c6182006-10-11 01:21:03 -07003612
Alex Tomasa86c6182006-10-11 01:21:03 -07003613 /* previous routine could use block we allocated */
Theodore Ts'obf89d162010-10-27 21:30:14 -04003614 newblock = ext4_ext_pblock(&newex);
Aneesh Kumar K.Vb939e372008-01-28 23:58:27 -05003615 allocated = ext4_ext_get_actual_len(&newex);
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003616 if (allocated > map->m_len)
3617 allocated = map->m_len;
3618 map->m_flags |= EXT4_MAP_NEW;
Alex Tomasa86c6182006-10-11 01:21:03 -07003619
Jan Karab436b9b2009-12-08 23:51:10 -05003620 /*
Aneesh Kumar K.V5f634d02010-01-25 04:00:31 -05003621 * Update reserved blocks/metadata blocks after successful
3622 * block allocation which had been deferred till now.
3623 */
Aneesh Kumar K.V1296cc82010-01-15 01:27:59 -05003624 if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
Aneesh Kumar K.V5f634d02010-01-25 04:00:31 -05003625 ext4_da_update_reserve_space(inode, allocated, 1);
3626
3627 /*
Jan Karab436b9b2009-12-08 23:51:10 -05003628 * Cache the extent and update transaction to commit on fdatasync only
3629 * when it is _not_ an uninitialized extent.
3630 */
3631 if ((flags & EXT4_GET_BLOCKS_UNINIT_EXT) == 0) {
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05003632 ext4_ext_put_in_cache(inode, map->m_lblk, allocated, newblock);
Jan Karab436b9b2009-12-08 23:51:10 -05003633 ext4_update_inode_fsync_trans(handle, inode, 1);
3634 } else
3635 ext4_update_inode_fsync_trans(handle, inode, 0);
Alex Tomasa86c6182006-10-11 01:21:03 -07003636out:
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003637 if (allocated > map->m_len)
3638 allocated = map->m_len;
Alex Tomasa86c6182006-10-11 01:21:03 -07003639 ext4_ext_show_leaf(inode, path);
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003640 map->m_flags |= EXT4_MAP_MAPPED;
3641 map->m_pblk = newblock;
3642 map->m_len = allocated;
Alex Tomasa86c6182006-10-11 01:21:03 -07003643out2:
3644 if (path) {
3645 ext4_ext_drop_refs(path);
3646 kfree(path);
3647 }
Jiaying Zhang0562e0b2011-03-21 21:38:05 -04003648 trace_ext4_ext_map_blocks_exit(inode, map->m_lblk,
3649 newblock, map->m_len, err ? err : allocated);
Allison Hendersone8613042011-05-25 07:41:46 -04003650
3651 result = (flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) ?
3652 punched_out : allocated;
3653
3654 return err ? err : result;
Alex Tomasa86c6182006-10-11 01:21:03 -07003655}
3656
Jan Karacf108bc2008-07-11 19:27:31 -04003657void ext4_ext_truncate(struct inode *inode)
Alex Tomasa86c6182006-10-11 01:21:03 -07003658{
3659 struct address_space *mapping = inode->i_mapping;
3660 struct super_block *sb = inode->i_sb;
Aneesh Kumar K.V725d26d2008-01-28 23:58:27 -05003661 ext4_lblk_t last_block;
Alex Tomasa86c6182006-10-11 01:21:03 -07003662 handle_t *handle;
3663 int err = 0;
3664
3665 /*
Jiaying Zhang3889fd52011-01-10 12:47:05 -05003666 * finish any pending end_io work so we won't run the risk of
3667 * converting any truncated blocks to initialized later
3668 */
3669 ext4_flush_completed_IO(inode);
3670
3671 /*
Alex Tomasa86c6182006-10-11 01:21:03 -07003672 * probably first extent we're gonna free will be last in block
3673 */
Mingming Caof3bd1f32008-08-19 22:16:03 -04003674 err = ext4_writepage_trans_blocks(inode);
Alex Tomasa86c6182006-10-11 01:21:03 -07003675 handle = ext4_journal_start(inode, err);
Jan Karacf108bc2008-07-11 19:27:31 -04003676 if (IS_ERR(handle))
Alex Tomasa86c6182006-10-11 01:21:03 -07003677 return;
Alex Tomasa86c6182006-10-11 01:21:03 -07003678
Jan Karacf108bc2008-07-11 19:27:31 -04003679 if (inode->i_size & (sb->s_blocksize - 1))
3680 ext4_block_truncate_page(handle, mapping, inode->i_size);
Alex Tomasa86c6182006-10-11 01:21:03 -07003681
Jan Kara9ddfc3d2008-07-11 19:27:31 -04003682 if (ext4_orphan_add(handle, inode))
3683 goto out_stop;
3684
Aneesh Kumar K.V0e855ac2008-01-28 23:58:26 -05003685 down_write(&EXT4_I(inode)->i_data_sem);
Alex Tomasa86c6182006-10-11 01:21:03 -07003686 ext4_ext_invalidate_cache(inode);
3687
Theodore Ts'oc2ea3fd2008-10-10 09:40:52 -04003688 ext4_discard_preallocations(inode);
Alex Tomasc9de5602008-01-29 00:19:52 -05003689
Alex Tomasa86c6182006-10-11 01:21:03 -07003690 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07003691 * TODO: optimization is possible here.
3692 * Probably we need not scan at all,
3693 * because page truncation is enough.
Alex Tomasa86c6182006-10-11 01:21:03 -07003694 */
Alex Tomasa86c6182006-10-11 01:21:03 -07003695
3696 /* we have to know where to truncate from in crash case */
3697 EXT4_I(inode)->i_disksize = inode->i_size;
3698 ext4_mark_inode_dirty(handle, inode);
3699
3700 last_block = (inode->i_size + sb->s_blocksize - 1)
3701 >> EXT4_BLOCK_SIZE_BITS(sb);
Lukas Czernerf17722f2011-06-06 00:05:17 -04003702 err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1);
Alex Tomasa86c6182006-10-11 01:21:03 -07003703
3704 /* In a multi-transaction truncate, we only make the final
Amit Arora56055d32007-07-17 21:42:38 -04003705 * transaction synchronous.
3706 */
Alex Tomasa86c6182006-10-11 01:21:03 -07003707 if (IS_SYNC(inode))
Frank Mayhar03901312009-01-07 00:06:22 -05003708 ext4_handle_sync(handle);
Alex Tomasa86c6182006-10-11 01:21:03 -07003709
Jan Kara9ddfc3d2008-07-11 19:27:31 -04003710 up_write(&EXT4_I(inode)->i_data_sem);
Eric Gouriouf6d2f6b2011-05-22 21:33:00 -04003711
3712out_stop:
Alex Tomasa86c6182006-10-11 01:21:03 -07003713 /*
Randy Dunlapd0d856e2006-10-11 01:21:07 -07003714 * If this was a simple ftruncate() and the file will remain alive,
Alex Tomasa86c6182006-10-11 01:21:03 -07003715 * then we need to clear up the orphan record which we created above.
3716 * However, if this was a real unlink then we were called by
3717 * ext4_delete_inode(), and we allow that function to clean up the
3718 * orphan info for us.
3719 */
3720 if (inode->i_nlink)
3721 ext4_orphan_del(handle, inode);
3722
Solofo Ramangalahyef737722008-04-29 22:00:41 -04003723 inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
3724 ext4_mark_inode_dirty(handle, inode);
Alex Tomasa86c6182006-10-11 01:21:03 -07003725 ext4_journal_stop(handle);
3726}
3727
Aneesh Kumar K.Vfd287842008-04-29 08:11:12 -04003728static void ext4_falloc_update_inode(struct inode *inode,
3729 int mode, loff_t new_size, int update_ctime)
3730{
3731 struct timespec now;
3732
3733 if (update_ctime) {
3734 now = current_fs_time(inode->i_sb);
3735 if (!timespec_equal(&inode->i_ctime, &now))
3736 inode->i_ctime = now;
3737 }
3738 /*
3739 * Update only when preallocation was requested beyond
3740 * the file size.
3741 */
Aneesh Kumar K.Vcf17fea2008-09-13 13:06:18 -04003742 if (!(mode & FALLOC_FL_KEEP_SIZE)) {
3743 if (new_size > i_size_read(inode))
3744 i_size_write(inode, new_size);
3745 if (new_size > EXT4_I(inode)->i_disksize)
3746 ext4_update_i_disksize(inode, new_size);
Jiaying Zhangc8d46e42010-02-24 09:52:53 -05003747 } else {
3748 /*
3749 * Mark that we allocate beyond EOF so the subsequent truncate
3750 * can proceed even if the new size is the same as i_size.
3751 */
3752 if (new_size > i_size_read(inode))
Dmitry Monakhov12e9b892010-05-16 22:00:00 -04003753 ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
Aneesh Kumar K.Vfd287842008-04-29 08:11:12 -04003754 }
3755
3756}
3757
Amit Aroraa2df2a62007-07-17 21:42:41 -04003758/*
Christoph Hellwig2fe17c12011-01-14 13:07:43 +01003759 * preallocate space for a file. This implements ext4's fallocate file
Amit Aroraa2df2a62007-07-17 21:42:41 -04003760 * operation, which gets called from sys_fallocate system call.
3761 * For block-mapped files, posix_fallocate should fall back to the method
3762 * of writing zeroes to the required new blocks (the same behavior which is
3763 * expected for file systems which do not support fallocate() system call).
3764 */
Christoph Hellwig2fe17c12011-01-14 13:07:43 +01003765long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
Amit Aroraa2df2a62007-07-17 21:42:41 -04003766{
Christoph Hellwig2fe17c12011-01-14 13:07:43 +01003767 struct inode *inode = file->f_path.dentry->d_inode;
Amit Aroraa2df2a62007-07-17 21:42:41 -04003768 handle_t *handle;
Aneesh Kumar K.Vfd287842008-04-29 08:11:12 -04003769 loff_t new_size;
Theodore Ts'o498e5f22008-11-05 00:14:04 -05003770 unsigned int max_blocks;
Amit Aroraa2df2a62007-07-17 21:42:41 -04003771 int ret = 0;
3772 int ret2 = 0;
3773 int retries = 0;
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003774 struct ext4_map_blocks map;
Amit Aroraa2df2a62007-07-17 21:42:41 -04003775 unsigned int credits, blkbits = inode->i_blkbits;
3776
3777 /*
3778 * currently supporting (pre)allocate mode for extent-based
3779 * files _only_
3780 */
Dmitry Monakhov12e9b892010-05-16 22:00:00 -04003781 if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
Amit Aroraa2df2a62007-07-17 21:42:41 -04003782 return -EOPNOTSUPP;
3783
Allison Hendersona4bb6b62011-05-25 07:41:50 -04003784 /* Return error if mode is not supported */
3785 if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
3786 return -EOPNOTSUPP;
3787
3788 if (mode & FALLOC_FL_PUNCH_HOLE)
3789 return ext4_punch_hole(file, offset, len);
3790
Jiaying Zhang0562e0b2011-03-21 21:38:05 -04003791 trace_ext4_fallocate_enter(inode, offset, len, mode);
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003792 map.m_lblk = offset >> blkbits;
Aneesh Kumar K.Vfd287842008-04-29 08:11:12 -04003793 /*
3794 * We can't just convert len to max_blocks because
3795 * If blocksize = 4096 offset = 3072 and len = 2048
3796 */
Amit Aroraa2df2a62007-07-17 21:42:41 -04003797 max_blocks = (EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits)
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003798 - map.m_lblk;
Amit Aroraa2df2a62007-07-17 21:42:41 -04003799 /*
Mingming Caof3bd1f32008-08-19 22:16:03 -04003800 * credits to insert 1 extent into extent tree
Amit Aroraa2df2a62007-07-17 21:42:41 -04003801 */
Mingming Caof3bd1f32008-08-19 22:16:03 -04003802 credits = ext4_chunk_trans_blocks(inode, max_blocks);
Aneesh Kumar K.V55bd7252008-02-15 12:47:21 -05003803 mutex_lock(&inode->i_mutex);
Nikanth Karthikesan6d19c422010-05-16 14:00:00 -04003804 ret = inode_newsize_ok(inode, (len + offset));
3805 if (ret) {
3806 mutex_unlock(&inode->i_mutex);
Jiaying Zhang0562e0b2011-03-21 21:38:05 -04003807 trace_ext4_fallocate_exit(inode, offset, max_blocks, ret);
Nikanth Karthikesan6d19c422010-05-16 14:00:00 -04003808 return ret;
3809 }
Amit Aroraa2df2a62007-07-17 21:42:41 -04003810retry:
3811 while (ret >= 0 && ret < max_blocks) {
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003812 map.m_lblk = map.m_lblk + ret;
3813 map.m_len = max_blocks = max_blocks - ret;
Amit Aroraa2df2a62007-07-17 21:42:41 -04003814 handle = ext4_journal_start(inode, credits);
3815 if (IS_ERR(handle)) {
3816 ret = PTR_ERR(handle);
3817 break;
3818 }
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003819 ret = ext4_map_blocks(handle, inode, &map,
Vivek Haldar556b27a2011-05-25 07:41:54 -04003820 EXT4_GET_BLOCKS_CREATE_UNINIT_EXT |
3821 EXT4_GET_BLOCKS_NO_NORMALIZE);
Aneesh Kumar K.V221879c2008-01-28 23:58:27 -05003822 if (ret <= 0) {
Aneesh Kumar K.V2c986152008-02-25 15:41:35 -05003823#ifdef EXT4FS_DEBUG
3824 WARN_ON(ret <= 0);
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003825 printk(KERN_ERR "%s: ext4_ext_map_blocks "
Aneesh Kumar K.V2c986152008-02-25 15:41:35 -05003826 "returned error inode#%lu, block=%u, "
Thadeu Lima de Souza Cascardo9fd97842009-01-26 19:26:26 -05003827 "max_blocks=%u", __func__,
Kazuya Mioa6371b62010-10-27 21:30:15 -04003828 inode->i_ino, map.m_lblk, max_blocks);
Aneesh Kumar K.V2c986152008-02-25 15:41:35 -05003829#endif
Amit Aroraa2df2a62007-07-17 21:42:41 -04003830 ext4_mark_inode_dirty(handle, inode);
3831 ret2 = ext4_journal_stop(handle);
3832 break;
3833 }
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003834 if ((map.m_lblk + ret) >= (EXT4_BLOCK_ALIGN(offset + len,
Aneesh Kumar K.Vfd287842008-04-29 08:11:12 -04003835 blkbits) >> blkbits))
3836 new_size = offset + len;
3837 else
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003838 new_size = (map.m_lblk + ret) << blkbits;
Amit Aroraa2df2a62007-07-17 21:42:41 -04003839
Aneesh Kumar K.Vfd287842008-04-29 08:11:12 -04003840 ext4_falloc_update_inode(inode, mode, new_size,
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003841 (map.m_flags & EXT4_MAP_NEW));
Amit Aroraa2df2a62007-07-17 21:42:41 -04003842 ext4_mark_inode_dirty(handle, inode);
3843 ret2 = ext4_journal_stop(handle);
3844 if (ret2)
3845 break;
3846 }
Aneesh Kumar K.Vfd287842008-04-29 08:11:12 -04003847 if (ret == -ENOSPC &&
3848 ext4_should_retry_alloc(inode->i_sb, &retries)) {
3849 ret = 0;
Amit Aroraa2df2a62007-07-17 21:42:41 -04003850 goto retry;
Amit Aroraa2df2a62007-07-17 21:42:41 -04003851 }
Aneesh Kumar K.V55bd7252008-02-15 12:47:21 -05003852 mutex_unlock(&inode->i_mutex);
Jiaying Zhang0562e0b2011-03-21 21:38:05 -04003853 trace_ext4_fallocate_exit(inode, offset, max_blocks,
3854 ret > 0 ? ret2 : ret);
Amit Aroraa2df2a62007-07-17 21:42:41 -04003855 return ret > 0 ? ret2 : ret;
3856}
Eric Sandeen6873fa02008-10-07 00:46:36 -04003857
3858/*
Mingming Cao00314622009-09-28 15:49:08 -04003859 * This function convert a range of blocks to written extents
3860 * The caller of this function will pass the start offset and the size.
3861 * all unwritten extents within this range will be converted to
3862 * written extents.
3863 *
3864 * This function is called from the direct IO end io call back
3865 * function, to convert the fallocated extents after IO is completed.
Mingming109f5562009-11-10 10:48:08 -05003866 * Returns 0 on success.
Mingming Cao00314622009-09-28 15:49:08 -04003867 */
3868int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
Eric Sandeena1de02d2010-02-04 23:58:38 -05003869 ssize_t len)
Mingming Cao00314622009-09-28 15:49:08 -04003870{
3871 handle_t *handle;
Mingming Cao00314622009-09-28 15:49:08 -04003872 unsigned int max_blocks;
3873 int ret = 0;
3874 int ret2 = 0;
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003875 struct ext4_map_blocks map;
Mingming Cao00314622009-09-28 15:49:08 -04003876 unsigned int credits, blkbits = inode->i_blkbits;
3877
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003878 map.m_lblk = offset >> blkbits;
Mingming Cao00314622009-09-28 15:49:08 -04003879 /*
3880 * We can't just convert len to max_blocks because
3881 * If blocksize = 4096 offset = 3072 and len = 2048
3882 */
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003883 max_blocks = ((EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits) -
3884 map.m_lblk);
Mingming Cao00314622009-09-28 15:49:08 -04003885 /*
3886 * credits to insert 1 extent into extent tree
3887 */
3888 credits = ext4_chunk_trans_blocks(inode, max_blocks);
3889 while (ret >= 0 && ret < max_blocks) {
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003890 map.m_lblk += ret;
3891 map.m_len = (max_blocks -= ret);
Mingming Cao00314622009-09-28 15:49:08 -04003892 handle = ext4_journal_start(inode, credits);
3893 if (IS_ERR(handle)) {
3894 ret = PTR_ERR(handle);
3895 break;
3896 }
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003897 ret = ext4_map_blocks(handle, inode, &map,
Jiaying Zhangc7064ef2010-03-02 13:28:44 -05003898 EXT4_GET_BLOCKS_IO_CONVERT_EXT);
Mingming Cao00314622009-09-28 15:49:08 -04003899 if (ret <= 0) {
3900 WARN_ON(ret <= 0);
Theodore Ts'oe35fd662010-05-16 19:00:00 -04003901 printk(KERN_ERR "%s: ext4_ext_map_blocks "
Mingming Cao00314622009-09-28 15:49:08 -04003902 "returned error inode#%lu, block=%u, "
3903 "max_blocks=%u", __func__,
Theodore Ts'o2ed88682010-05-16 20:00:00 -04003904 inode->i_ino, map.m_lblk, map.m_len);
Mingming Cao00314622009-09-28 15:49:08 -04003905 }
3906 ext4_mark_inode_dirty(handle, inode);
3907 ret2 = ext4_journal_stop(handle);
3908 if (ret <= 0 || ret2 )
3909 break;
3910 }
3911 return ret > 0 ? ret2 : ret;
3912}
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003913
Mingming Cao00314622009-09-28 15:49:08 -04003914/*
Eric Sandeen6873fa02008-10-07 00:46:36 -04003915 * Callback function called for each extent to gather FIEMAP information.
3916 */
Lukas Czernerc03f8aa2011-06-06 00:06:52 -04003917static int ext4_ext_fiemap_cb(struct inode *inode, ext4_lblk_t next,
Eric Sandeen6873fa02008-10-07 00:46:36 -04003918 struct ext4_ext_cache *newex, struct ext4_extent *ex,
3919 void *data)
3920{
Eric Sandeen6873fa02008-10-07 00:46:36 -04003921 __u64 logical;
3922 __u64 physical;
3923 __u64 length;
3924 __u32 flags = 0;
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003925 int ret = 0;
3926 struct fiemap_extent_info *fieinfo = data;
3927 unsigned char blksize_bits;
Eric Sandeen6873fa02008-10-07 00:46:36 -04003928
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003929 blksize_bits = inode->i_sb->s_blocksize_bits;
3930 logical = (__u64)newex->ec_block << blksize_bits;
Eric Sandeen6873fa02008-10-07 00:46:36 -04003931
Theodore Ts'ob05e6ae2011-01-10 12:13:26 -05003932 if (newex->ec_start == 0) {
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003933 /*
3934 * No extent in extent-tree contains block @newex->ec_start,
3935 * then the block may stay in 1)a hole or 2)delayed-extent.
3936 *
3937 * Holes or delayed-extents are processed as follows.
3938 * 1. lookup dirty pages with specified range in pagecache.
3939 * If no page is got, then there is no delayed-extent and
3940 * return with EXT_CONTINUE.
3941 * 2. find the 1st mapped buffer,
3942 * 3. check if the mapped buffer is both in the request range
3943 * and a delayed buffer. If not, there is no delayed-extent,
3944 * then return.
3945 * 4. a delayed-extent is found, the extent will be collected.
3946 */
3947 ext4_lblk_t end = 0;
3948 pgoff_t last_offset;
3949 pgoff_t offset;
3950 pgoff_t index;
Yongqiang Yangb2213492011-05-24 11:36:58 -04003951 pgoff_t start_index = 0;
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003952 struct page **pages = NULL;
Eric Sandeen6873fa02008-10-07 00:46:36 -04003953 struct buffer_head *bh = NULL;
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003954 struct buffer_head *head = NULL;
3955 unsigned int nr_pages = PAGE_SIZE / sizeof(struct page *);
3956
3957 pages = kmalloc(PAGE_SIZE, GFP_KERNEL);
3958 if (pages == NULL)
3959 return -ENOMEM;
Eric Sandeen6873fa02008-10-07 00:46:36 -04003960
3961 offset = logical >> PAGE_SHIFT;
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003962repeat:
3963 last_offset = offset;
3964 head = NULL;
3965 ret = find_get_pages_tag(inode->i_mapping, &offset,
3966 PAGECACHE_TAG_DIRTY, nr_pages, pages);
Eric Sandeen6873fa02008-10-07 00:46:36 -04003967
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003968 if (!(flags & FIEMAP_EXTENT_DELALLOC)) {
3969 /* First time, try to find a mapped buffer. */
3970 if (ret == 0) {
3971out:
3972 for (index = 0; index < ret; index++)
3973 page_cache_release(pages[index]);
3974 /* just a hole. */
3975 kfree(pages);
3976 return EXT_CONTINUE;
3977 }
Yongqiang Yangb2213492011-05-24 11:36:58 -04003978 index = 0;
Eric Sandeen6873fa02008-10-07 00:46:36 -04003979
Yongqiang Yangb2213492011-05-24 11:36:58 -04003980next_page:
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003981 /* Try to find the 1st mapped buffer. */
Yongqiang Yangb2213492011-05-24 11:36:58 -04003982 end = ((__u64)pages[index]->index << PAGE_SHIFT) >>
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003983 blksize_bits;
Yongqiang Yangb2213492011-05-24 11:36:58 -04003984 if (!page_has_buffers(pages[index]))
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003985 goto out;
Yongqiang Yangb2213492011-05-24 11:36:58 -04003986 head = page_buffers(pages[index]);
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003987 if (!head)
3988 goto out;
Eric Sandeen6873fa02008-10-07 00:46:36 -04003989
Yongqiang Yangb2213492011-05-24 11:36:58 -04003990 index++;
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05003991 bh = head;
3992 do {
Yongqiang Yangb2213492011-05-24 11:36:58 -04003993 if (end >= newex->ec_block +
3994 newex->ec_len)
3995 /* The buffer is out of
3996 * the request range.
3997 */
3998 goto out;
3999
4000 if (buffer_mapped(bh) &&
4001 end >= newex->ec_block) {
4002 start_index = index - 1;
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004003 /* get the 1st mapped buffer. */
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004004 goto found_mapped_buffer;
4005 }
Yongqiang Yangb2213492011-05-24 11:36:58 -04004006
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004007 bh = bh->b_this_page;
4008 end++;
4009 } while (bh != head);
4010
Yongqiang Yangb2213492011-05-24 11:36:58 -04004011 /* No mapped buffer in the range found in this page,
4012 * We need to look up next page.
4013 */
4014 if (index >= ret) {
4015 /* There is no page left, but we need to limit
4016 * newex->ec_len.
4017 */
4018 newex->ec_len = end - newex->ec_block;
4019 goto out;
4020 }
4021 goto next_page;
Eric Sandeen6873fa02008-10-07 00:46:36 -04004022 } else {
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004023 /*Find contiguous delayed buffers. */
4024 if (ret > 0 && pages[0]->index == last_offset)
4025 head = page_buffers(pages[0]);
4026 bh = head;
Yongqiang Yangb2213492011-05-24 11:36:58 -04004027 index = 1;
4028 start_index = 0;
Eric Sandeen6873fa02008-10-07 00:46:36 -04004029 }
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004030
4031found_mapped_buffer:
4032 if (bh != NULL && buffer_delay(bh)) {
4033 /* 1st or contiguous delayed buffer found. */
4034 if (!(flags & FIEMAP_EXTENT_DELALLOC)) {
4035 /*
4036 * 1st delayed buffer found, record
4037 * the start of extent.
4038 */
4039 flags |= FIEMAP_EXTENT_DELALLOC;
4040 newex->ec_block = end;
4041 logical = (__u64)end << blksize_bits;
4042 }
4043 /* Find contiguous delayed buffers. */
4044 do {
4045 if (!buffer_delay(bh))
4046 goto found_delayed_extent;
4047 bh = bh->b_this_page;
4048 end++;
4049 } while (bh != head);
4050
Yongqiang Yangb2213492011-05-24 11:36:58 -04004051 for (; index < ret; index++) {
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004052 if (!page_has_buffers(pages[index])) {
4053 bh = NULL;
4054 break;
4055 }
4056 head = page_buffers(pages[index]);
4057 if (!head) {
4058 bh = NULL;
4059 break;
4060 }
Yongqiang Yangb2213492011-05-24 11:36:58 -04004061
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004062 if (pages[index]->index !=
Yongqiang Yangb2213492011-05-24 11:36:58 -04004063 pages[start_index]->index + index
4064 - start_index) {
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004065 /* Blocks are not contiguous. */
4066 bh = NULL;
4067 break;
4068 }
4069 bh = head;
4070 do {
4071 if (!buffer_delay(bh))
4072 /* Delayed-extent ends. */
4073 goto found_delayed_extent;
4074 bh = bh->b_this_page;
4075 end++;
4076 } while (bh != head);
4077 }
4078 } else if (!(flags & FIEMAP_EXTENT_DELALLOC))
4079 /* a hole found. */
4080 goto out;
4081
4082found_delayed_extent:
4083 newex->ec_len = min(end - newex->ec_block,
4084 (ext4_lblk_t)EXT_INIT_MAX_LEN);
4085 if (ret == nr_pages && bh != NULL &&
4086 newex->ec_len < EXT_INIT_MAX_LEN &&
4087 buffer_delay(bh)) {
4088 /* Have not collected an extent and continue. */
4089 for (index = 0; index < ret; index++)
4090 page_cache_release(pages[index]);
4091 goto repeat;
4092 }
4093
4094 for (index = 0; index < ret; index++)
4095 page_cache_release(pages[index]);
4096 kfree(pages);
Eric Sandeen6873fa02008-10-07 00:46:36 -04004097 }
4098
4099 physical = (__u64)newex->ec_start << blksize_bits;
4100 length = (__u64)newex->ec_len << blksize_bits;
4101
4102 if (ex && ext4_ext_is_uninitialized(ex))
4103 flags |= FIEMAP_EXTENT_UNWRITTEN;
4104
Lukas Czernerc03f8aa2011-06-06 00:06:52 -04004105 if (next == EXT_MAX_BLOCKS)
Eric Sandeen6873fa02008-10-07 00:46:36 -04004106 flags |= FIEMAP_EXTENT_LAST;
4107
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004108 ret = fiemap_fill_next_extent(fieinfo, logical, physical,
Eric Sandeen6873fa02008-10-07 00:46:36 -04004109 length, flags);
Yongqiang Yang6d9c85e2011-02-27 17:25:47 -05004110 if (ret < 0)
4111 return ret;
4112 if (ret == 1)
Eric Sandeen6873fa02008-10-07 00:46:36 -04004113 return EXT_BREAK;
Eric Sandeen6873fa02008-10-07 00:46:36 -04004114 return EXT_CONTINUE;
4115}
4116
4117/* fiemap flags we can handle specified here */
4118#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR)
4119
Aneesh Kumar K.V3a06d772008-11-22 15:04:59 -05004120static int ext4_xattr_fiemap(struct inode *inode,
4121 struct fiemap_extent_info *fieinfo)
Eric Sandeen6873fa02008-10-07 00:46:36 -04004122{
4123 __u64 physical = 0;
4124 __u64 length;
4125 __u32 flags = FIEMAP_EXTENT_LAST;
4126 int blockbits = inode->i_sb->s_blocksize_bits;
4127 int error = 0;
4128
4129 /* in-inode? */
Theodore Ts'o19f5fb72010-01-24 14:34:07 -05004130 if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
Eric Sandeen6873fa02008-10-07 00:46:36 -04004131 struct ext4_iloc iloc;
4132 int offset; /* offset of xattr in inode */
4133
4134 error = ext4_get_inode_loc(inode, &iloc);
4135 if (error)
4136 return error;
4137 physical = iloc.bh->b_blocknr << blockbits;
4138 offset = EXT4_GOOD_OLD_INODE_SIZE +
4139 EXT4_I(inode)->i_extra_isize;
4140 physical += offset;
4141 length = EXT4_SB(inode->i_sb)->s_inode_size - offset;
4142 flags |= FIEMAP_EXTENT_DATA_INLINE;
Curt Wohlgemuthfd2dd9f2010-04-03 17:44:16 -04004143 brelse(iloc.bh);
Eric Sandeen6873fa02008-10-07 00:46:36 -04004144 } else { /* external block */
4145 physical = EXT4_I(inode)->i_file_acl << blockbits;
4146 length = inode->i_sb->s_blocksize;
4147 }
4148
4149 if (physical)
4150 error = fiemap_fill_next_extent(fieinfo, 0, physical,
4151 length, flags);
4152 return (error < 0 ? error : 0);
4153}
4154
Allison Hendersona4bb6b62011-05-25 07:41:50 -04004155/*
4156 * ext4_ext_punch_hole
4157 *
4158 * Punches a hole of "length" bytes in a file starting
4159 * at byte "offset"
4160 *
4161 * @inode: The inode of the file to punch a hole in
4162 * @offset: The starting byte offset of the hole
4163 * @length: The length of the hole
4164 *
4165 * Returns the number of blocks removed or negative on err
4166 */
4167int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length)
4168{
4169 struct inode *inode = file->f_path.dentry->d_inode;
4170 struct super_block *sb = inode->i_sb;
4171 struct ext4_ext_cache cache_ex;
4172 ext4_lblk_t first_block, last_block, num_blocks, iblock, max_blocks;
4173 struct address_space *mapping = inode->i_mapping;
4174 struct ext4_map_blocks map;
4175 handle_t *handle;
4176 loff_t first_block_offset, last_block_offset, block_len;
4177 loff_t first_page, last_page, first_page_offset, last_page_offset;
4178 int ret, credits, blocks_released, err = 0;
4179
4180 first_block = (offset + sb->s_blocksize - 1) >>
4181 EXT4_BLOCK_SIZE_BITS(sb);
4182 last_block = (offset + length) >> EXT4_BLOCK_SIZE_BITS(sb);
4183
4184 first_block_offset = first_block << EXT4_BLOCK_SIZE_BITS(sb);
4185 last_block_offset = last_block << EXT4_BLOCK_SIZE_BITS(sb);
4186
4187 first_page = (offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
4188 last_page = (offset + length) >> PAGE_CACHE_SHIFT;
4189
4190 first_page_offset = first_page << PAGE_CACHE_SHIFT;
4191 last_page_offset = last_page << PAGE_CACHE_SHIFT;
4192
4193 /*
4194 * Write out all dirty pages to avoid race conditions
4195 * Then release them.
4196 */
4197 if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) {
4198 err = filemap_write_and_wait_range(mapping,
4199 first_page_offset == 0 ? 0 : first_page_offset-1,
4200 last_page_offset);
4201
4202 if (err)
4203 return err;
4204 }
4205
4206 /* Now release the pages */
4207 if (last_page_offset > first_page_offset) {
4208 truncate_inode_pages_range(mapping, first_page_offset,
4209 last_page_offset-1);
4210 }
4211
4212 /* finish any pending end_io work */
4213 ext4_flush_completed_IO(inode);
4214
4215 credits = ext4_writepage_trans_blocks(inode);
4216 handle = ext4_journal_start(inode, credits);
4217 if (IS_ERR(handle))
4218 return PTR_ERR(handle);
4219
4220 err = ext4_orphan_add(handle, inode);
4221 if (err)
4222 goto out;
4223
4224 /*
4225 * Now we need to zero out the un block aligned data.
4226 * If the file is smaller than a block, just
4227 * zero out the middle
4228 */
4229 if (first_block > last_block)
4230 ext4_block_zero_page_range(handle, mapping, offset, length);
4231 else {
4232 /* zero out the head of the hole before the first block */
4233 block_len = first_block_offset - offset;
4234 if (block_len > 0)
4235 ext4_block_zero_page_range(handle, mapping,
4236 offset, block_len);
4237
4238 /* zero out the tail of the hole after the last block */
4239 block_len = offset + length - last_block_offset;
4240 if (block_len > 0) {
4241 ext4_block_zero_page_range(handle, mapping,
4242 last_block_offset, block_len);
4243 }
4244 }
4245
4246 /* If there are no blocks to remove, return now */
4247 if (first_block >= last_block)
4248 goto out;
4249
4250 down_write(&EXT4_I(inode)->i_data_sem);
4251 ext4_ext_invalidate_cache(inode);
4252 ext4_discard_preallocations(inode);
4253
4254 /*
4255 * Loop over all the blocks and identify blocks
4256 * that need to be punched out
4257 */
4258 iblock = first_block;
4259 blocks_released = 0;
4260 while (iblock < last_block) {
4261 max_blocks = last_block - iblock;
4262 num_blocks = 1;
4263 memset(&map, 0, sizeof(map));
4264 map.m_lblk = iblock;
4265 map.m_len = max_blocks;
4266 ret = ext4_ext_map_blocks(handle, inode, &map,
4267 EXT4_GET_BLOCKS_PUNCH_OUT_EXT);
4268
4269 if (ret > 0) {
4270 blocks_released += ret;
4271 num_blocks = ret;
4272 } else if (ret == 0) {
4273 /*
4274 * If map blocks could not find the block,
4275 * then it is in a hole. If the hole was
4276 * not already cached, then map blocks should
4277 * put it in the cache. So we can get the hole
4278 * out of the cache
4279 */
4280 memset(&cache_ex, 0, sizeof(cache_ex));
4281 if ((ext4_ext_check_cache(inode, iblock, &cache_ex)) &&
4282 !cache_ex.ec_start) {
4283
4284 /* The hole is cached */
4285 num_blocks = cache_ex.ec_block +
4286 cache_ex.ec_len - iblock;
4287
4288 } else {
4289 /* The block could not be identified */
4290 err = -EIO;
4291 break;
4292 }
4293 } else {
4294 /* Map blocks error */
4295 err = ret;
4296 break;
4297 }
4298
4299 if (num_blocks == 0) {
4300 /* This condition should never happen */
4301 ext_debug("Block lookup failed");
4302 err = -EIO;
4303 break;
4304 }
4305
4306 iblock += num_blocks;
4307 }
4308
4309 if (blocks_released > 0) {
4310 ext4_ext_invalidate_cache(inode);
4311 ext4_discard_preallocations(inode);
4312 }
4313
4314 if (IS_SYNC(inode))
4315 ext4_handle_sync(handle);
4316
4317 up_write(&EXT4_I(inode)->i_data_sem);
4318
4319out:
4320 ext4_orphan_del(handle, inode);
4321 inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
4322 ext4_mark_inode_dirty(handle, inode);
4323 ext4_journal_stop(handle);
4324 return err;
4325}
Eric Sandeen6873fa02008-10-07 00:46:36 -04004326int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
4327 __u64 start, __u64 len)
4328{
4329 ext4_lblk_t start_blk;
Eric Sandeen6873fa02008-10-07 00:46:36 -04004330 int error = 0;
4331
4332 /* fallback to generic here if not in extents fmt */
Dmitry Monakhov12e9b892010-05-16 22:00:00 -04004333 if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
Eric Sandeen6873fa02008-10-07 00:46:36 -04004334 return generic_block_fiemap(inode, fieinfo, start, len,
4335 ext4_get_block);
4336
4337 if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS))
4338 return -EBADR;
4339
4340 if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
4341 error = ext4_xattr_fiemap(inode, fieinfo);
4342 } else {
Leonard Michlmayraca92ff2010-03-04 17:07:28 -05004343 ext4_lblk_t len_blks;
4344 __u64 last_blk;
4345
Eric Sandeen6873fa02008-10-07 00:46:36 -04004346 start_blk = start >> inode->i_sb->s_blocksize_bits;
Leonard Michlmayraca92ff2010-03-04 17:07:28 -05004347 last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
Lukas Czernerf17722f2011-06-06 00:05:17 -04004348 if (last_blk >= EXT_MAX_BLOCKS)
4349 last_blk = EXT_MAX_BLOCKS-1;
Leonard Michlmayraca92ff2010-03-04 17:07:28 -05004350 len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
Eric Sandeen6873fa02008-10-07 00:46:36 -04004351
4352 /*
4353 * Walk the extent tree gathering extent information.
4354 * ext4_ext_fiemap_cb will push extents back to user.
4355 */
Eric Sandeen6873fa02008-10-07 00:46:36 -04004356 error = ext4_ext_walk_space(inode, start_blk, len_blks,
4357 ext4_ext_fiemap_cb, fieinfo);
Eric Sandeen6873fa02008-10-07 00:46:36 -04004358 }
4359
4360 return error;
4361}