1 .. SPDX-License-Identifier: GPL-2.0 1 .. SPDX-License-Identifier: GPL-2.0
2 2
3 Directory Entries 3 Directory Entries
4 ----------------- 4 -----------------
5 5
6 In an ext4 filesystem, a directory is more or 6 In an ext4 filesystem, a directory is more or less a flat file that maps
7 an arbitrary byte string (usually ASCII) to an 7 an arbitrary byte string (usually ASCII) to an inode number on the
8 filesystem. There can be many directory entrie 8 filesystem. There can be many directory entries across the filesystem
9 that reference the same inode number--these ar 9 that reference the same inode number--these are known as hard links, and
10 that is why hard links cannot reference files 10 that is why hard links cannot reference files on other filesystems. As
11 such, directory entries are found by reading t 11 such, directory entries are found by reading the data block(s)
12 associated with a directory file for the parti 12 associated with a directory file for the particular directory entry that
13 is desired. 13 is desired.
14 14
15 Linear (Classic) Directories 15 Linear (Classic) Directories
16 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 16 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
17 17
18 By default, each directory lists its entries i 18 By default, each directory lists its entries in an “almost-linear”
19 array. I write “almost” because it's not a 19 array. I write “almost” because it's not a linear array in the memory
20 sense because directory entries are not split 20 sense because directory entries are not split across filesystem blocks.
21 Therefore, it is more accurate to say that a d 21 Therefore, it is more accurate to say that a directory is a series of
22 data blocks and that each block contains a lin 22 data blocks and that each block contains a linear array of directory
23 entries. The end of each per-block array is si 23 entries. The end of each per-block array is signified by reaching the
24 end of the block; the last entry in the block 24 end of the block; the last entry in the block has a record length that
25 takes it all the way to the end of the block. 25 takes it all the way to the end of the block. The end of the entire
26 directory is of course signified by reaching t 26 directory is of course signified by reaching the end of the file. Unused
27 directory entries are signified by inode = 0. 27 directory entries are signified by inode = 0. By default the filesystem
28 uses ``struct ext4_dir_entry_2`` for directory 28 uses ``struct ext4_dir_entry_2`` for directory entries unless the
29 “filetype” feature flag is not set, in whi 29 “filetype” feature flag is not set, in which case it uses
30 ``struct ext4_dir_entry``. 30 ``struct ext4_dir_entry``.
31 31
32 The original directory entry format is ``struc 32 The original directory entry format is ``struct ext4_dir_entry``, which
33 is at most 263 bytes long, though on disk you' 33 is at most 263 bytes long, though on disk you'll need to reference
34 ``dirent.rec_len`` to know for sure. 34 ``dirent.rec_len`` to know for sure.
35 35
36 .. list-table:: 36 .. list-table::
37 :widths: 8 8 24 40 37 :widths: 8 8 24 40
38 :header-rows: 1 38 :header-rows: 1
39 39
40 * - Offset 40 * - Offset
41 - Size 41 - Size
42 - Name 42 - Name
43 - Description 43 - Description
44 * - 0x0 44 * - 0x0
45 - __le32 !! 45 - \_\_le32
46 - inode 46 - inode
47 - Number of the inode that this directory 47 - Number of the inode that this directory entry points to.
48 * - 0x4 48 * - 0x4
49 - __le16 !! 49 - \_\_le16
50 - rec_len !! 50 - rec\_len
51 - Length of this directory entry. Must be 51 - Length of this directory entry. Must be a multiple of 4.
52 * - 0x6 52 * - 0x6
53 - __le16 !! 53 - \_\_le16
54 - name_len !! 54 - name\_len
55 - Length of the file name. 55 - Length of the file name.
56 * - 0x8 56 * - 0x8
57 - char 57 - char
58 - name[EXT4_NAME_LEN] !! 58 - name[EXT4\_NAME\_LEN]
59 - File name. 59 - File name.
60 60
61 Since file names cannot be longer than 255 byt 61 Since file names cannot be longer than 255 bytes, the new directory
62 entry format shortens the name_len field and u !! 62 entry format shortens the rec\_len field and uses the space for a file
63 type flag, probably to avoid having to load ev 63 type flag, probably to avoid having to load every inode during directory
64 tree traversal. This format is ``ext4_dir_entr 64 tree traversal. This format is ``ext4_dir_entry_2``, which is at most
65 263 bytes long, though on disk you'll need to 65 263 bytes long, though on disk you'll need to reference
66 ``dirent.rec_len`` to know for sure. 66 ``dirent.rec_len`` to know for sure.
67 67
68 .. list-table:: 68 .. list-table::
69 :widths: 8 8 24 40 69 :widths: 8 8 24 40
70 :header-rows: 1 70 :header-rows: 1
71 71
72 * - Offset 72 * - Offset
73 - Size 73 - Size
74 - Name 74 - Name
75 - Description 75 - Description
76 * - 0x0 76 * - 0x0
77 - __le32 !! 77 - \_\_le32
78 - inode 78 - inode
79 - Number of the inode that this directory 79 - Number of the inode that this directory entry points to.
80 * - 0x4 80 * - 0x4
81 - __le16 !! 81 - \_\_le16
82 - rec_len !! 82 - rec\_len
83 - Length of this directory entry. 83 - Length of this directory entry.
84 * - 0x6 84 * - 0x6
85 - __u8 !! 85 - \_\_u8
86 - name_len !! 86 - name\_len
87 - Length of the file name. 87 - Length of the file name.
88 * - 0x7 88 * - 0x7
89 - __u8 !! 89 - \_\_u8
90 - file_type !! 90 - file\_type
91 - File type code, see ftype_ table below. 91 - File type code, see ftype_ table below.
92 * - 0x8 92 * - 0x8
93 - char 93 - char
94 - name[EXT4_NAME_LEN] !! 94 - name[EXT4\_NAME\_LEN]
95 - File name. 95 - File name.
96 96
97 .. _ftype: 97 .. _ftype:
98 98
99 The directory file type is one of the followin 99 The directory file type is one of the following values:
100 100
101 .. list-table:: 101 .. list-table::
102 :widths: 16 64 102 :widths: 16 64
103 :header-rows: 1 103 :header-rows: 1
104 104
105 * - Value 105 * - Value
106 - Description 106 - Description
107 * - 0x0 107 * - 0x0
108 - Unknown. 108 - Unknown.
109 * - 0x1 109 * - 0x1
110 - Regular file. 110 - Regular file.
111 * - 0x2 111 * - 0x2
112 - Directory. 112 - Directory.
113 * - 0x3 113 * - 0x3
114 - Character device file. 114 - Character device file.
115 * - 0x4 115 * - 0x4
116 - Block device file. 116 - Block device file.
117 * - 0x5 117 * - 0x5
118 - FIFO. 118 - FIFO.
119 * - 0x6 119 * - 0x6
120 - Socket. 120 - Socket.
121 * - 0x7 121 * - 0x7
122 - Symbolic link. 122 - Symbolic link.
123 123
124 To support directories that are both encrypted <<
125 must also include hash information in the dire <<
126 ``ext4_extended_dir_entry_2`` to ``ext4_dir_en <<
127 for dot and dotdot, which are kept the same. T <<
128 after ``name`` and is included in the size lis <<
129 entry uses this extension, it may be up to 271 <<
130 <<
131 .. list-table:: <<
132 :widths: 8 8 24 40 <<
133 :header-rows: 1 <<
134 <<
135 * - Offset <<
136 - Size <<
137 - Name <<
138 - Description <<
139 * - 0x0 <<
140 - __le32 <<
141 - hash <<
142 - The hash of the directory name <<
143 * - 0x4 <<
144 - __le32 <<
145 - minor_hash <<
146 - The minor hash of the directory name <<
147 <<
148 <<
149 In order to add checksums to these classic dir 124 In order to add checksums to these classic directory blocks, a phony
150 ``struct ext4_dir_entry`` is placed at the end 125 ``struct ext4_dir_entry`` is placed at the end of each leaf block to
151 hold the checksum. The directory entry is 12 b 126 hold the checksum. The directory entry is 12 bytes long. The inode
152 number and name_len fields are set to zero to !! 127 number and name\_len fields are set to zero to fool old software into
153 ignoring an apparently empty directory entry, 128 ignoring an apparently empty directory entry, and the checksum is stored
154 in the place where the name normally goes. The 129 in the place where the name normally goes. The structure is
155 ``struct ext4_dir_entry_tail``: 130 ``struct ext4_dir_entry_tail``:
156 131
157 .. list-table:: 132 .. list-table::
158 :widths: 8 8 24 40 133 :widths: 8 8 24 40
159 :header-rows: 1 134 :header-rows: 1
160 135
161 * - Offset 136 * - Offset
162 - Size 137 - Size
163 - Name 138 - Name
164 - Description 139 - Description
165 * - 0x0 140 * - 0x0
166 - __le32 !! 141 - \_\_le32
167 - det_reserved_zero1 !! 142 - det\_reserved\_zero1
168 - Inode number, which must be zero. 143 - Inode number, which must be zero.
169 * - 0x4 144 * - 0x4
170 - __le16 !! 145 - \_\_le16
171 - det_rec_len !! 146 - det\_rec\_len
172 - Length of this directory entry, which m 147 - Length of this directory entry, which must be 12.
173 * - 0x6 148 * - 0x6
174 - __u8 !! 149 - \_\_u8
175 - det_reserved_zero2 !! 150 - det\_reserved\_zero2
176 - Length of the file name, which must be 151 - Length of the file name, which must be zero.
177 * - 0x7 152 * - 0x7
178 - __u8 !! 153 - \_\_u8
179 - det_reserved_ft !! 154 - det\_reserved\_ft
180 - File type, which must be 0xDE. 155 - File type, which must be 0xDE.
181 * - 0x8 156 * - 0x8
182 - __le32 !! 157 - \_\_le32
183 - det_checksum !! 158 - det\_checksum
184 - Directory leaf block checksum. 159 - Directory leaf block checksum.
185 160
186 The leaf directory block checksum is calculate 161 The leaf directory block checksum is calculated against the FS UUID, the
187 directory's inode number, the directory's inod 162 directory's inode number, the directory's inode generation number, and
188 the entire directory entry block up to (but no 163 the entire directory entry block up to (but not including) the fake
189 directory entry. 164 directory entry.
190 165
191 Hash Tree Directories 166 Hash Tree Directories
192 ~~~~~~~~~~~~~~~~~~~~~ 167 ~~~~~~~~~~~~~~~~~~~~~
193 168
194 A linear array of directory entries isn't grea 169 A linear array of directory entries isn't great for performance, so a
195 new feature was added to ext3 to provide a fas 170 new feature was added to ext3 to provide a faster (but peculiar)
196 balanced tree keyed off a hash of the director 171 balanced tree keyed off a hash of the directory entry name. If the
197 EXT4_INDEX_FL (0x1000) flag is set in the inod !! 172 EXT4\_INDEX\_FL (0x1000) flag is set in the inode, this directory uses a
198 hashed btree (htree) to organize and find dire 173 hashed btree (htree) to organize and find directory entries. For
199 backwards read-only compatibility with ext2, t 174 backwards read-only compatibility with ext2, this tree is actually
200 hidden inside the directory file, masquerading 175 hidden inside the directory file, masquerading as “empty” directory data
201 blocks! It was stated previously that the end 176 blocks! It was stated previously that the end of the linear directory
202 entry table was signified with an entry pointi 177 entry table was signified with an entry pointing to inode 0; this is
203 (ab)used to fool the old linear-scan algorithm 178 (ab)used to fool the old linear-scan algorithm into thinking that the
204 rest of the directory block is empty so that i 179 rest of the directory block is empty so that it moves on.
205 180
206 The root of the tree always lives in the first 181 The root of the tree always lives in the first data block of the
207 directory. By ext2 custom, the '.' and '..' en 182 directory. By ext2 custom, the '.' and '..' entries must appear at the
208 beginning of this first block, so they are put 183 beginning of this first block, so they are put here as two
209 ``struct ext4_dir_entry_2`` s and not stored i !! 184 ``struct ext4_dir_entry_2``\ s and not stored in the tree. The rest of
210 the root node contains metadata about the tree 185 the root node contains metadata about the tree and finally a hash->block
211 map to find nodes that are lower in the htree. 186 map to find nodes that are lower in the htree. If
212 ``dx_root.info.indirect_levels`` is non-zero t 187 ``dx_root.info.indirect_levels`` is non-zero then the htree has two
213 levels; the data block pointed to by the root 188 levels; the data block pointed to by the root node's map is an interior
214 node, which is indexed by a minor hash. Interi 189 node, which is indexed by a minor hash. Interior nodes in this tree
215 contains a zeroed out ``struct ext4_dir_entry_ 190 contains a zeroed out ``struct ext4_dir_entry_2`` followed by a
216 minor_hash->block map to find leafe nodes. Lea !! 191 minor\_hash->block map to find leafe nodes. Leaf nodes contain a linear
217 array of all ``struct ext4_dir_entry_2``; all 192 array of all ``struct ext4_dir_entry_2``; all of these entries
218 (presumably) hash to the same value. If there 193 (presumably) hash to the same value. If there is an overflow, the
219 entries simply overflow into the next leaf nod 194 entries simply overflow into the next leaf node, and the
220 least-significant bit of the hash (in the inte 195 least-significant bit of the hash (in the interior node map) that gets
221 us to this next leaf node is set. 196 us to this next leaf node is set.
222 197
223 To traverse the directory as a htree, the code 198 To traverse the directory as a htree, the code calculates the hash of
224 the desired file name and uses it to find the 199 the desired file name and uses it to find the corresponding block
225 number. If the tree is flat, the block is a li 200 number. If the tree is flat, the block is a linear array of directory
226 entries that can be searched; otherwise, the m 201 entries that can be searched; otherwise, the minor hash of the file name
227 is computed and used against this second block 202 is computed and used against this second block to find the corresponding
228 third block number. That third block number wi 203 third block number. That third block number will be a linear array of
229 directory entries. 204 directory entries.
230 205
231 To traverse the directory as a linear array (s 206 To traverse the directory as a linear array (such as the old code does),
232 the code simply reads every data block in the 207 the code simply reads every data block in the directory. The blocks used
233 for the htree will appear to have no entries ( 208 for the htree will appear to have no entries (aside from '.' and '..')
234 and so only the leaf nodes will appear to have 209 and so only the leaf nodes will appear to have any interesting content.
235 210
236 The root of the htree is in ``struct dx_root`` 211 The root of the htree is in ``struct dx_root``, which is the full length
237 of a data block: 212 of a data block:
238 213
239 .. list-table:: 214 .. list-table::
240 :widths: 8 8 24 40 215 :widths: 8 8 24 40
241 :header-rows: 1 216 :header-rows: 1
242 217
243 * - Offset 218 * - Offset
244 - Type 219 - Type
245 - Name 220 - Name
246 - Description 221 - Description
247 * - 0x0 222 * - 0x0
248 - __le32 !! 223 - \_\_le32
249 - dot.inode 224 - dot.inode
250 - inode number of this directory. 225 - inode number of this directory.
251 * - 0x4 226 * - 0x4
252 - __le16 !! 227 - \_\_le16
253 - dot.rec_len !! 228 - dot.rec\_len
254 - Length of this record, 12. 229 - Length of this record, 12.
255 * - 0x6 230 * - 0x6
256 - u8 231 - u8
257 - dot.name_len !! 232 - dot.name\_len
258 - Length of the name, 1. 233 - Length of the name, 1.
259 * - 0x7 234 * - 0x7
260 - u8 235 - u8
261 - dot.file_type !! 236 - dot.file\_type
262 - File type of this entry, 0x2 (directory 237 - File type of this entry, 0x2 (directory) (if the feature flag is set).
263 * - 0x8 238 * - 0x8
264 - char 239 - char
265 - dot.name[4] 240 - dot.name[4]
266 - “.\0\0\0” !! 241 - “.\\0\\0\\0”
267 * - 0xC 242 * - 0xC
268 - __le32 !! 243 - \_\_le32
269 - dotdot.inode 244 - dotdot.inode
270 - inode number of parent directory. 245 - inode number of parent directory.
271 * - 0x10 246 * - 0x10
272 - __le16 !! 247 - \_\_le16
273 - dotdot.rec_len !! 248 - dotdot.rec\_len
274 - block_size - 12. The record length is l !! 249 - block\_size - 12. The record length is long enough to cover all htree
275 data. 250 data.
276 * - 0x12 251 * - 0x12
277 - u8 252 - u8
278 - dotdot.name_len !! 253 - dotdot.name\_len
279 - Length of the name, 2. 254 - Length of the name, 2.
280 * - 0x13 255 * - 0x13
281 - u8 256 - u8
282 - dotdot.file_type !! 257 - dotdot.file\_type
283 - File type of this entry, 0x2 (directory 258 - File type of this entry, 0x2 (directory) (if the feature flag is set).
284 * - 0x14 259 * - 0x14
285 - char 260 - char
286 - dotdot_name[4] !! 261 - dotdot\_name[4]
287 - “..\0\0” !! 262 - “..\\0\\0”
288 * - 0x18 263 * - 0x18
289 - __le32 !! 264 - \_\_le32
290 - struct dx_root_info.reserved_zero !! 265 - struct dx\_root\_info.reserved\_zero
291 - Zero. 266 - Zero.
292 * - 0x1C 267 * - 0x1C
293 - u8 268 - u8
294 - struct dx_root_info.hash_version !! 269 - struct dx\_root\_info.hash\_version
295 - Hash type, see dirhash_ table below. 270 - Hash type, see dirhash_ table below.
296 * - 0x1D 271 * - 0x1D
297 - u8 272 - u8
298 - struct dx_root_info.info_length !! 273 - struct dx\_root\_info.info\_length
299 - Length of the tree information, 0x8. 274 - Length of the tree information, 0x8.
300 * - 0x1E 275 * - 0x1E
301 - u8 276 - u8
302 - struct dx_root_info.indirect_levels !! 277 - struct dx\_root\_info.indirect\_levels
303 - Depth of the htree. Cannot be larger th !! 278 - Depth of the htree. Cannot be larger than 3 if the INCOMPAT\_LARGEDIR
304 feature is set; cannot be larger than 2 279 feature is set; cannot be larger than 2 otherwise.
305 * - 0x1F 280 * - 0x1F
306 - u8 281 - u8
307 - struct dx_root_info.unused_flags !! 282 - struct dx\_root\_info.unused\_flags
308 - 283 -
309 * - 0x20 284 * - 0x20
310 - __le16 !! 285 - \_\_le16
311 - limit 286 - limit
312 - Maximum number of dx_entries that can f !! 287 - Maximum number of dx\_entries that can follow this header, plus 1 for
313 the header itself. 288 the header itself.
314 * - 0x22 289 * - 0x22
315 - __le16 !! 290 - \_\_le16
316 - count 291 - count
317 - Actual number of dx_entries that follow !! 292 - Actual number of dx\_entries that follow this header, plus 1 for the
318 header itself. 293 header itself.
319 * - 0x24 294 * - 0x24
320 - __le32 !! 295 - \_\_le32
321 - block 296 - block
322 - The block number (within the directory 297 - The block number (within the directory file) that goes with hash=0.
323 * - 0x28 298 * - 0x28
324 - struct dx_entry !! 299 - struct dx\_entry
325 - entries[0] 300 - entries[0]
326 - As many 8-byte ``struct dx_entry`` as f 301 - As many 8-byte ``struct dx_entry`` as fits in the rest of the data block.
327 302
328 .. _dirhash: 303 .. _dirhash:
329 304
330 The directory hash is one of the following val 305 The directory hash is one of the following values:
331 306
332 .. list-table:: 307 .. list-table::
333 :widths: 16 64 308 :widths: 16 64
334 :header-rows: 1 309 :header-rows: 1
335 310
336 * - Value 311 * - Value
337 - Description 312 - Description
338 * - 0x0 313 * - 0x0
339 - Legacy. 314 - Legacy.
340 * - 0x1 315 * - 0x1
341 - Half MD4. 316 - Half MD4.
342 * - 0x2 317 * - 0x2
343 - Tea. 318 - Tea.
344 * - 0x3 319 * - 0x3
345 - Legacy, unsigned. 320 - Legacy, unsigned.
346 * - 0x4 321 * - 0x4
347 - Half MD4, unsigned. 322 - Half MD4, unsigned.
348 * - 0x5 323 * - 0x5
349 - Tea, unsigned. 324 - Tea, unsigned.
350 * - 0x6 <<
351 - Siphash. <<
352 325
353 Interior nodes of an htree are recorded as ``s 326 Interior nodes of an htree are recorded as ``struct dx_node``, which is
354 also the full length of a data block: 327 also the full length of a data block:
355 328
356 .. list-table:: 329 .. list-table::
357 :widths: 8 8 24 40 330 :widths: 8 8 24 40
358 :header-rows: 1 331 :header-rows: 1
359 332
360 * - Offset 333 * - Offset
361 - Type 334 - Type
362 - Name 335 - Name
363 - Description 336 - Description
364 * - 0x0 337 * - 0x0
365 - __le32 !! 338 - \_\_le32
366 - fake.inode 339 - fake.inode
367 - Zero, to make it look like this entry i 340 - Zero, to make it look like this entry is not in use.
368 * - 0x4 341 * - 0x4
369 - __le16 !! 342 - \_\_le16
370 - fake.rec_len !! 343 - fake.rec\_len
371 - The size of the block, in order to hide !! 344 - The size of the block, in order to hide all of the dx\_node data.
372 * - 0x6 345 * - 0x6
373 - u8 346 - u8
374 - name_len !! 347 - name\_len
375 - Zero. There is no name for this “unus 348 - Zero. There is no name for this “unused” directory entry.
376 * - 0x7 349 * - 0x7
377 - u8 350 - u8
378 - file_type !! 351 - file\_type
379 - Zero. There is no file type for this 352 - Zero. There is no file type for this “unused” directory entry.
380 * - 0x8 353 * - 0x8
381 - __le16 !! 354 - \_\_le16
382 - limit 355 - limit
383 - Maximum number of dx_entries that can f !! 356 - Maximum number of dx\_entries that can follow this header, plus 1 for
384 the header itself. 357 the header itself.
385 * - 0xA 358 * - 0xA
386 - __le16 !! 359 - \_\_le16
387 - count 360 - count
388 - Actual number of dx_entries that follow !! 361 - Actual number of dx\_entries that follow this header, plus 1 for the
389 header itself. 362 header itself.
390 * - 0xE 363 * - 0xE
391 - __le32 !! 364 - \_\_le32
392 - block 365 - block
393 - The block number (within the directory 366 - The block number (within the directory file) that goes with the lowest
394 hash value of this block. This value is 367 hash value of this block. This value is stored in the parent block.
395 * - 0x12 368 * - 0x12
396 - struct dx_entry !! 369 - struct dx\_entry
397 - entries[0] 370 - entries[0]
398 - As many 8-byte ``struct dx_entry`` as f 371 - As many 8-byte ``struct dx_entry`` as fits in the rest of the data block.
399 372
400 The hash maps that exist in both ``struct dx_r 373 The hash maps that exist in both ``struct dx_root`` and
401 ``struct dx_node`` are recorded as ``struct dx 374 ``struct dx_node`` are recorded as ``struct dx_entry``, which is 8 bytes
402 long: 375 long:
403 376
404 .. list-table:: 377 .. list-table::
405 :widths: 8 8 24 40 378 :widths: 8 8 24 40
406 :header-rows: 1 379 :header-rows: 1
407 380
408 * - Offset 381 * - Offset
409 - Type 382 - Type
410 - Name 383 - Name
411 - Description 384 - Description
412 * - 0x0 385 * - 0x0
413 - __le32 !! 386 - \_\_le32
414 - hash 387 - hash
415 - Hash code. 388 - Hash code.
416 * - 0x4 389 * - 0x4
417 - __le32 !! 390 - \_\_le32
418 - block 391 - block
419 - Block number (within the directory file 392 - Block number (within the directory file, not filesystem blocks) of the
420 next node in the htree. 393 next node in the htree.
421 394
422 (If you think this is all quite clever and pec 395 (If you think this is all quite clever and peculiar, so does the
423 author.) 396 author.)
424 397
425 If metadata checksums are enabled, the last 8 398 If metadata checksums are enabled, the last 8 bytes of the directory
426 block (precisely the length of one dx_entry) a !! 399 block (precisely the length of one dx\_entry) are used to store a
427 ``struct dx_tail``, which contains the checksu 400 ``struct dx_tail``, which contains the checksum. The ``limit`` and
428 ``count`` entries in the dx_root/dx_node struc !! 401 ``count`` entries in the dx\_root/dx\_node structures are adjusted as
429 necessary to fit the dx_tail into the block. I !! 402 necessary to fit the dx\_tail into the block. If there is no space for
430 the dx_tail, the user is notified to run e2fsc !! 403 the dx\_tail, the user is notified to run e2fsck -D to rebuild the
431 directory index (which will ensure that there' 404 directory index (which will ensure that there's space for the checksum.
432 The dx_tail structure is 8 bytes long and look !! 405 The dx\_tail structure is 8 bytes long and looks like this:
433 406
434 .. list-table:: 407 .. list-table::
435 :widths: 8 8 24 40 408 :widths: 8 8 24 40
436 :header-rows: 1 409 :header-rows: 1
437 410
438 * - Offset 411 * - Offset
439 - Type 412 - Type
440 - Name 413 - Name
441 - Description 414 - Description
442 * - 0x0 415 * - 0x0
443 - u32 416 - u32
444 - dt_reserved !! 417 - dt\_reserved
445 - Zero. 418 - Zero.
446 * - 0x4 419 * - 0x4
447 - __le32 !! 420 - \_\_le32
448 - dt_checksum !! 421 - dt\_checksum
449 - Checksum of the htree directory block. 422 - Checksum of the htree directory block.
450 423
451 The checksum is calculated against the FS UUID 424 The checksum is calculated against the FS UUID, the htree index header
452 (dx_root or dx_node), all of the htree indices !! 425 (dx\_root or dx\_node), all of the htree indices (dx\_entry) that are in
453 use, and the tail block (dx_tail). !! 426 use, and the tail block (dx\_tail).
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.