2 * Squashfs - a compressed read only filesystem for Linux
4 * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008
5 * Phillip Lougher <phillip@lougher.demon.co.uk>
7 * This program is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation; either version 2,
10 * or (at your option) any later version.
12 * This program is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 * GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with this program; if not, write to the Free Software
19 * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
25 * This file implements code to do filename lookup in directories.
27 * Like inodes, directories are packed into compressed metadata blocks, stored
28 * in a directory table. Directories are accessed using the start address of
29 * the metablock containing the directory and the offset into the
30 * decompressed block (<block, offset>).
32 * Directories are organised in a slightly complex way, and are not simply
33 * a list of file names. The organisation takes advantage of the
34 * fact that (in most cases) the inodes of the files will be in the same
35 * compressed metadata block, and therefore, can share the start block.
36 * Directories are therefore organised in a two level list, a directory
37 * header containing the shared start block value, and a sequence of directory
38 * entries, each of which share the shared start block. A new directory header
39 * is written once/if the inode start block changes. The directory
40 * header/directory entry list is repeated as many times as necessary.
42 * Directories are sorted, and can contain a directory index to speed up
43 * file lookup. Directory indexes store one entry per metablock, each entry
44 * storing the index/filename mapping to the first directory header
45 * in each metadata block. Directories are sorted in alphabetical order,
46 * and at lookup the index is scanned linearly looking for the first filename
47 * alphabetically larger than the filename being looked up. At this point the
48 * location of the metadata block the filename is in has been found.
49 * The general idea of the index is ensure only one metadata block needs to be
50 * decompressed to do a lookup irrespective of the length of the directory.
51 * This scheme has the advantage that it doesn't require extra memory overhead
52 * and doesn't require much extra storage on disk.
56 #include <linux/vfs.h>
57 #include <linux/slab.h>
58 #include <linux/string.h>
59 #include <linux/dcache.h>
60 #include <linux/xattr.h>
62 #include "squashfs_fs.h"
63 #include "squashfs_fs_sb.h"
64 #include "squashfs_fs_i.h"
69 * Lookup name in the directory index, returning the location of the metadata
70 * block containing it, and the directory index this represents.
72 * If we get an error reading the index then return the part of the index
73 * (if any) we have managed to read - the index isn't essential, just
76 static int get_dir_index_using_name(struct super_block
*sb
,
77 u64
*next_block
, int *next_offset
, u64 index_start
,
78 int index_offset
, int i_count
, const char *name
,
81 struct squashfs_sb_info
*msblk
= sb
->s_fs_info
;
82 int i
, size
, length
= 0, err
;
83 struct squashfs_dir_index
*index
;
86 TRACE("Entered get_dir_index_using_name, i_count %d\n", i_count
);
88 index
= kmalloc(sizeof(*index
) + SQUASHFS_NAME_LEN
* 2 + 2, GFP_KERNEL
);
90 ERROR("Failed to allocate squashfs_dir_index\n");
94 str
= &index
->name
[SQUASHFS_NAME_LEN
+ 1];
95 strncpy(str
, name
, len
);
98 for (i
= 0; i
< i_count
; i
++) {
99 err
= squashfs_read_metadata(sb
, index
, &index_start
,
100 &index_offset
, sizeof(*index
));
105 size
= le32_to_cpu(index
->size
) + 1;
107 err
= squashfs_read_metadata(sb
, index
->name
, &index_start
,
108 &index_offset
, size
);
112 index
->name
[size
] = '\0';
114 if (strcmp(index
->name
, str
) > 0)
117 length
= le32_to_cpu(index
->index
);
118 *next_block
= le32_to_cpu(index
->start_block
) +
119 msblk
->directory_table
;
122 *next_offset
= (length
+ *next_offset
) % SQUASHFS_METADATA_SIZE
;
127 * Return index (f_pos) of the looked up metadata block. Translate
128 * from internal f_pos to external f_pos which is offset by 3 because
129 * we invent "." and ".." entries which are not actually stored in the
136 static struct dentry
*squashfs_lookup(struct inode
*dir
, struct dentry
*dentry
,
137 struct nameidata
*nd
)
139 const unsigned char *name
= dentry
->d_name
.name
;
140 int len
= dentry
->d_name
.len
;
141 struct inode
*inode
= NULL
;
142 struct squashfs_sb_info
*msblk
= dir
->i_sb
->s_fs_info
;
143 struct squashfs_dir_header dirh
;
144 struct squashfs_dir_entry
*dire
;
145 u64 block
= squashfs_i(dir
)->start
+ msblk
->directory_table
;
146 int offset
= squashfs_i(dir
)->offset
;
147 int err
, length
= 0, dir_count
, size
;
149 TRACE("Entered squashfs_lookup [%llx:%x]\n", block
, offset
);
151 dire
= kmalloc(sizeof(*dire
) + SQUASHFS_NAME_LEN
+ 1, GFP_KERNEL
);
153 ERROR("Failed to allocate squashfs_dir_entry\n");
154 return ERR_PTR(-ENOMEM
);
157 if (len
> SQUASHFS_NAME_LEN
) {
162 length
= get_dir_index_using_name(dir
->i_sb
, &block
, &offset
,
163 squashfs_i(dir
)->dir_idx_start
,
164 squashfs_i(dir
)->dir_idx_offset
,
165 squashfs_i(dir
)->dir_idx_cnt
, name
, len
);
167 while (length
< i_size_read(dir
)) {
169 * Read directory header.
171 err
= squashfs_read_metadata(dir
->i_sb
, &dirh
, &block
,
172 &offset
, sizeof(dirh
));
176 length
+= sizeof(dirh
);
178 dir_count
= le32_to_cpu(dirh
.count
) + 1;
179 while (dir_count
--) {
181 * Read directory entry.
183 err
= squashfs_read_metadata(dir
->i_sb
, dire
, &block
,
184 &offset
, sizeof(*dire
));
188 size
= le16_to_cpu(dire
->size
) + 1;
190 err
= squashfs_read_metadata(dir
->i_sb
, dire
->name
,
191 &block
, &offset
, size
);
195 length
+= sizeof(*dire
) + size
;
197 if (name
[0] < dire
->name
[0])
200 if (len
== size
&& !strncmp(name
, dire
->name
, len
)) {
201 unsigned int blk
, off
, ino_num
;
203 blk
= le32_to_cpu(dirh
.start_block
);
204 off
= le16_to_cpu(dire
->offset
);
205 ino_num
= le32_to_cpu(dirh
.inode_number
) +
206 (short) le16_to_cpu(dire
->inode_number
);
207 ino
= SQUASHFS_MKINODE(blk
, off
);
209 TRACE("calling squashfs_iget for directory "
210 "entry %s, inode %x:%x, %d\n", name
,
213 inode
= squashfs_iget(dir
->i_sb
, ino
, ino_num
);
215 err
= PTR_ERR(inode
);
227 return d_splice_alias(inode
, dentry
);
228 d_add(dentry
, inode
);
232 ERROR("Unable to read directory block [%llx:%x]\n",
233 squashfs_i(dir
)->start
+ msblk
->directory_table
,
234 squashfs_i(dir
)->offset
);
241 const struct inode_operations squashfs_dir_inode_ops
= {
242 .lookup
= squashfs_lookup
,
243 .getxattr
= generic_getxattr
,
244 .listxattr
= squashfs_listxattr