1 /* $OpenBSD: ext2fs.h,v 1.26 2021/03/11 13:31:35 jsg Exp $ */
2 /* $NetBSD: ext2fs.h,v 1.10 2000/01/28 16:00:23 bouyer Exp $ */
3
4 /*
5 * Copyright (c) 1997 Manuel Bouyer.
6 * Copyright (c) 1982, 1986, 1993
7 * The Regents of the University of California. All rights reserved.
8 *
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
11 * are met:
12 * 1. Redistributions of source code must retain the above copyright
13 * notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 * notice, this list of conditions and the following disclaimer in the
16 * documentation and/or other materials provided with the distribution.
17 * 3. Neither the name of the University nor the names of its contributors
18 * may be used to endorse or promote products derived from this software
19 * without specific prior written permission.
20 *
21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31 * SUCH DAMAGE.
32 *
33 * @(#)fs.h 8.10 (Berkeley) 10/27/94
34 * Modified for ext2fs by Manuel Bouyer.
35 */
36
37 #include <sys/endian.h>
38
39 /*
40 * Each disk drive contains some number of file systems.
41 * A file system consists of a number of cylinder groups.
42 * Each cylinder group has inodes and data.
43 *
44 * A file system is described by its super-block, which in turn
45 * describes the cylinder groups. The super-block is critical
46 * data and is replicated in each cylinder group to protect against
47 * catastrophic loss. This is done at `newfs' time and the critical
48 * super-block data does not change, so the copies need not be
49 * referenced further unless disaster strikes.
50 *
51 * The first boot and super blocks are given in absolute disk addresses.
52 * The byte-offset forms are preferred, as they don't imply a sector size.
53 */
54 #define BBSIZE 1024
55 #define SBSIZE 1024
56 #define BBOFF ((off_t)(0))
57 #define SBOFF ((off_t)(BBOFF + BBSIZE))
58 #define BBLOCK ((daddr_t)(0))
59 #define SBLOCK ((daddr_t)(BBLOCK + BBSIZE / DEV_BSIZE))
60
61 /*
62 * Inodes are, like in UFS, 32-bit unsigned integers and therefore ufsino_t.
63 * Disk blocks are 32-bit, if the filesystem isn't operating in 64-bit mode
64 * (the incompatible ext4 64BIT flag). More work is needed to properly use
65 * daddr_t as the disk block data type on both BE and LE architectures.
66 * XXX disk blocks are simply u_int32_t for now.
67 */
68
69 /*
70 * MINBSIZE is the smallest allowable block size.
71 * MINBSIZE must be big enough to hold a cylinder group block,
72 * thus changes to (struct cg) must keep its size within MINBSIZE.
73 * Note that super blocks are always of size SBSIZE,
74 * and that both SBSIZE and MAXBSIZE must be >= MINBSIZE.
75 * FSIZE means fragment size.
76 */
77 #define LOG_MINBSIZE 10
78 #define MINBSIZE (1 << LOG_MINBSIZE)
79 #define LOG_MINFSIZE 10
80 #define MINFSIZE (1 << LOG_MINFSIZE)
81
82 /*
83 * The path name on which the file system is mounted is maintained
84 * in fs_fsmnt. MAXMNTLEN defines the amount of space allocated in
85 * the super block for this name.
86 */
87 #define MAXMNTLEN 512
88
89 /*
90 * MINFREE gives the minimum acceptable percentage of file system
91 * blocks which may be free. If the freelist drops below this level
92 * only the superuser may continue to allocate blocks. This may
93 * be set to 0 if no reserve of free blocks is deemed necessary,
94 * however throughput drops by fifty percent if the file system
95 * is run at between 95% and 100% full; thus the minimum default
96 * value of fs_minfree is 5%. However, to get good clustering
97 * performance, 10% is a better choice. hence we use 10% as our
98 * default value. With 10% free space, fragmentation is not a
99 * problem, so we choose to optimize for time.
100 */
101 #define MINFREE 5
102
103 /*
104 * Super block for an ext2fs file system.
105 */
106 struct ext2fs {
107 u_int32_t e2fs_icount; /* Inode count */
108 u_int32_t e2fs_bcount; /* blocks count */
109 u_int32_t e2fs_rbcount; /* reserved blocks count */
110 u_int32_t e2fs_fbcount; /* free blocks count */
111 u_int32_t e2fs_ficount; /* free inodes count */
112 u_int32_t e2fs_first_dblock; /* first data block */
113 u_int32_t e2fs_log_bsize; /* block size = 1024*(2^e2fs_log_bsize) */
114 u_int32_t e2fs_log_fsize; /* fragment size log2 */
115 u_int32_t e2fs_bpg; /* blocks per group */
116 u_int32_t e2fs_fpg; /* frags per group */
117 u_int32_t e2fs_ipg; /* inodes per group */
118 u_int32_t e2fs_mtime; /* mount time */
119 u_int32_t e2fs_wtime; /* write time */
120 u_int16_t e2fs_mnt_count; /* mount count */
121 u_int16_t e2fs_max_mnt_count; /* max mount count */
122 u_int16_t e2fs_magic; /* magic number */
123 u_int16_t e2fs_state; /* file system state */
124 u_int16_t e2fs_beh; /* behavior on errors */
125 u_int16_t e2fs_minrev; /* minor revision level */
126 u_int32_t e2fs_lastfsck; /* time of last fsck */
127 u_int32_t e2fs_fsckintv; /* max time between fscks */
128 u_int32_t e2fs_creator; /* creator OS */
129 u_int32_t e2fs_rev; /* revision level */
130 u_int16_t e2fs_ruid; /* default uid for reserved blocks */
131 u_int16_t e2fs_rgid; /* default gid for reserved blocks */
132 /* EXT2_DYNAMIC_REV superblocks */
133 u_int32_t e2fs_first_ino; /* first non-reserved inode */
134 u_int16_t e2fs_inode_size; /* size of inode structure */
135 u_int16_t e2fs_block_group_nr; /* block grp number of this sblk*/
136 u_int32_t e2fs_features_compat; /* compatible feature set */
137 u_int32_t e2fs_features_incompat; /* incompatible feature set */
138 u_int32_t e2fs_features_rocompat; /* RO-compatible feature set */
139 u_int8_t e2fs_uuid[16]; /* 128-bit uuid for volume */
140 char e2fs_vname[16]; /* volume name */
141 char e2fs_fsmnt[64]; /* name mounted on */
142 u_int32_t e2fs_algo; /* For compression */
143 u_int8_t e2fs_prealloc; /* # of blocks to preallocate */
144 u_int8_t e2fs_dir_prealloc; /* # of blocks to preallocate for dir */
145 u_int16_t e2fs_reserved_ngdb; /* # of reserved gd blocks for resize */
146 /* Ext3 JBD2 journaling. */
147 u_int8_t e2fs_journal_uuid[16];
148 u_int32_t e2fs_journal_ino;
149 u_int32_t e2fs_journal_dev;
150 u_int32_t e2fs_last_orphan; /* start of list of inodes to delete */
151 u_int32_t e2fs_hash_seed[4]; /* htree hash seed */
152 u_int8_t e2fs_def_hash_version;
153 u_int8_t e2fs_journal_backup_type;
154 u_int16_t e2fs_gdesc_size;
155 u_int32_t e2fs_default_mount_opts;
156 u_int32_t e2fs_first_meta_bg;
157 u_int32_t e2fs_mkfs_time;
158 u_int32_t e2fs_journal_backup[17];
159 u_int32_t reserved2[172];
160 };
161
162
163 /* in-memory data for ext2fs */
164 struct m_ext2fs {
165 struct ext2fs e2fs;
166 u_char e2fs_fsmnt[MAXMNTLEN]; /* name mounted on */
167 int8_t e2fs_ronly; /* mounted read-only flag */
168 int8_t e2fs_fmod; /* super block modified flag */
169 int32_t e2fs_fsize; /* fragment size */
170 int32_t e2fs_bsize; /* block size */
171 int32_t e2fs_bshift; /* ``lblkno'' calc of logical blkno */
172 int32_t e2fs_bmask; /* ``blkoff'' calc of blk offsets */
173 int64_t e2fs_qbmask; /* ~fs_bmask - for use with quad size */
174 int32_t e2fs_fsbtodb; /* fsbtodb and dbtofsb shift constant */
175 int32_t e2fs_ncg; /* number of cylinder groups */
176 int32_t e2fs_ngdb; /* number of group descriptor block */
177 int32_t e2fs_ipb; /* number of inodes per block */
178 int32_t e2fs_itpg; /* number of inode table per group */
179 off_t e2fs_maxfilesize; /* depends on LARGE/HUGE flags */
180 struct ext2_gd *e2fs_gd; /* group descriptors */
181 };
182
183 static inline int
e2fs_overflow(struct m_ext2fs * fs,off_t lower,off_t value)184 e2fs_overflow(struct m_ext2fs *fs, off_t lower, off_t value)
185 {
186 return (value < lower || value > fs->e2fs_maxfilesize);
187 }
188
189 /*
190 * Filesystem identification
191 */
192 #define E2FS_MAGIC 0xef53 /* the ext2fs magic number */
193 #define E2FS_REV0 0 /* revision levels */
194 #define E2FS_REV1 1 /* revision levels */
195
196 /* compatible/incompatible features */
197 #define EXT2F_COMPAT_PREALLOC 0x0001
198 #define EXT2F_COMPAT_IMAGIC_INODES 0x0002
199 #define EXT2F_COMPAT_HAS_JOURNAL 0x0004
200 #define EXT2F_COMPAT_EXT_ATTR 0x0008
201 #define EXT2F_COMPAT_RESIZE 0x0010
202 #define EXT2F_COMPAT_DIR_INDEX 0x0020
203 #define EXT2F_COMPAT_SPARSE_SUPER2 0x0200
204
205 #define EXT2F_ROCOMPAT_SPARSE_SUPER 0x0001
206 #define EXT2F_ROCOMPAT_LARGE_FILE 0x0002
207 #define EXT2F_ROCOMPAT_BTREE_DIR 0x0004
208 #define EXT2F_ROCOMPAT_HUGE_FILE 0x0008
209 #define EXT2F_ROCOMPAT_GDT_CSUM 0x0010
210 #define EXT2F_ROCOMPAT_DIR_NLINK 0x0020
211 #define EXT2F_ROCOMPAT_EXTRA_ISIZE 0x0040
212 #define EXT2F_ROCOMPAT_QUOTA 0x0100
213 #define EXT2F_ROCOMPAT_BIGALLOC 0x0200
214 #define EXT2F_ROCOMPAT_METADATA_CKSUM 0x0400
215 #define EXT2F_ROCOMPAT_READONLY 0x1000
216 #define EXT2F_ROCOMPAT_PROJECT 0x2000
217
218 #define EXT2F_INCOMPAT_COMP 0x0001
219 #define EXT2F_INCOMPAT_FTYPE 0x0002
220 #define EXT2F_INCOMPAT_RECOVER 0x0004
221 #define EXT2F_INCOMPAT_JOURNAL_DEV 0x0008
222 #define EXT2F_INCOMPAT_META_BG 0x0010
223 #define EXT2F_INCOMPAT_EXTENTS 0x0040
224 #define EXT2F_INCOMPAT_64BIT 0x0080
225 #define EXT2F_INCOMPAT_MMP 0x0100
226 #define EXT2F_INCOMPAT_FLEX_BG 0x0200
227 #define EXT2F_INCOMPAT_EA_INODE 0x0400
228 #define EXT2F_INCOMPAT_DIRDATA 0x1000
229 #define EXT2F_INCOMPAT_CSUM_SEED 0x2000
230 #define EXT2F_INCOMPAT_LARGEDIR 0x4000
231 #define EXT2F_INCOMPAT_INLINE_DATA 0x8000
232 #define EXT2F_INCOMPAT_ENCRYPT 0x10000
233
234 struct ext2_feature {
235 uint32_t mask;
236 const char *name;
237 };
238
239 static const struct ext2_feature ro_compat[] = {
240 { EXT2F_ROCOMPAT_SPARSE_SUPER, "sparse_super" },
241 { EXT2F_ROCOMPAT_LARGE_FILE, "large_file" },
242 { EXT2F_ROCOMPAT_BTREE_DIR, "btree_dir" },
243 { EXT2F_ROCOMPAT_HUGE_FILE, "huge_file" },
244 { EXT2F_ROCOMPAT_GDT_CSUM, "uninit_bg" },
245 { EXT2F_ROCOMPAT_DIR_NLINK, "dir_nlink" },
246 { EXT2F_ROCOMPAT_EXTRA_ISIZE, "extra_isize" },
247 { EXT2F_ROCOMPAT_QUOTA, "quota" },
248 { EXT2F_ROCOMPAT_BIGALLOC, "bigalloc" },
249 { EXT2F_ROCOMPAT_METADATA_CKSUM, "metadata_csum" },
250 { EXT2F_ROCOMPAT_READONLY, "read-only" },
251 { EXT2F_ROCOMPAT_PROJECT, "project" }
252 };
253
254 static const struct ext2_feature incompat[] = {
255 { EXT2F_INCOMPAT_COMP, "compression" },
256 { EXT2F_INCOMPAT_FTYPE, "filetype" },
257 { EXT2F_INCOMPAT_RECOVER, "needs_recovery" },
258 { EXT2F_INCOMPAT_JOURNAL_DEV, "journal_dev" },
259 { EXT2F_INCOMPAT_META_BG, "meta_bg" },
260 { EXT2F_INCOMPAT_EXTENTS, "extents" },
261 { EXT2F_INCOMPAT_64BIT, "64bit" },
262 { EXT2F_INCOMPAT_MMP, "mmp" },
263 { EXT2F_INCOMPAT_FLEX_BG, "flex_bg" },
264 { EXT2F_INCOMPAT_EA_INODE, "ea_inode" },
265 { EXT2F_INCOMPAT_DIRDATA, "dirdata" },
266 { EXT2F_INCOMPAT_CSUM_SEED, "metadata_csum_seed" },
267 { EXT2F_INCOMPAT_LARGEDIR, "large_dir" },
268 { EXT2F_INCOMPAT_INLINE_DATA, "inline_data" },
269 { EXT2F_INCOMPAT_ENCRYPT, "encrypt" }
270 };
271
272 /* features supported in this implementation */
273 #define EXT2F_COMPAT_SUPP 0x0000
274 #define EXT2F_ROCOMPAT_SUPP (EXT2F_ROCOMPAT_SPARSE_SUPER | \
275 EXT2F_ROCOMPAT_LARGE_FILE)
276 #define EXT2F_INCOMPAT_SUPP (EXT2F_INCOMPAT_FTYPE)
277 #define EXT4F_RO_INCOMPAT_SUPP (EXT2F_INCOMPAT_EXTENTS | \
278 EXT2F_INCOMPAT_FLEX_BG | \
279 EXT2F_INCOMPAT_META_BG | \
280 EXT2F_INCOMPAT_RECOVER)
281
282 /*
283 * Definitions of behavior on errors
284 */
285 #define E2FS_BEH_CONTINUE 1 /* continue operation */
286 #define E2FS_BEH_READONLY 2 /* remount fs read only */
287 #define E2FS_BEH_PANIC 3 /* cause panic */
288 #define E2FS_BEH_DEFAULT E2FS_BEH_CONTINUE
289
290 /*
291 * OS identification
292 */
293 #define E2FS_OS_LINUX 0
294 #define E2FS_OS_HURD 1
295 #define E2FS_OS_MASIX 2
296
297 /*
298 * Filesystem clean flags
299 */
300 #define E2FS_ISCLEAN 0x01
301 #define E2FS_ERRORS 0x02
302
303 /* ext2 file system block group descriptor */
304
305 struct ext2_gd {
306 u_int32_t ext2bgd_b_bitmap; /* blocks bitmap block */
307 u_int32_t ext2bgd_i_bitmap; /* inodes bitmap block */
308 u_int32_t ext2bgd_i_tables; /* inodes table block */
309 u_int16_t ext2bgd_nbfree; /* number of free blocks */
310 u_int16_t ext2bgd_nifree; /* number of free inodes */
311 u_int16_t ext2bgd_ndirs; /* number of directories */
312 u_int16_t reserved;
313 u_int32_t reserved2[3];
314 };
315
316 /*
317 * If the EXT2F_ROCOMPAT_SPARSE_SUPER flag is set, the cylinder group has a
318 * copy of the super and cylinder group descriptors blocks only if it's
319 * a power of 3, 5 or 7
320 */
321
322 static __inline__ int cg_has_sb(int) __attribute__((__unused__));
323 static __inline int
cg_has_sb(int i)324 cg_has_sb(int i)
325 {
326 int a3 ,a5 , a7;
327
328 if (i == 0 || i == 1)
329 return 1;
330 for (a3 = 3, a5 = 5, a7 = 7;
331 a3 <= i || a5 <= i || a7 <= i;
332 a3 *= 3, a5 *= 5, a7 *= 7)
333 if (i == a3 || i == a5 || i == a7)
334 return 1;
335 return 0;
336 }
337
338 /*
339 * Ext2 metadata is stored in little-endian byte order.
340 * JBD2 journal used in ext3 and ext4 is big-endian!
341 */
342 #if BYTE_ORDER == LITTLE_ENDIAN
343 #define e2fs_sbload(old, new) memcpy((new), (old), SBSIZE);
344 #define e2fs_cgload(old, new, size) memcpy((new), (old), (size));
345 #define e2fs_sbsave(old, new) memcpy((new), (old), SBSIZE);
346 #define e2fs_cgsave(old, new, size) memcpy((new), (old), (size));
347 #else
348 void e2fs_sb_bswap(struct ext2fs *, struct ext2fs *);
349 void e2fs_cg_bswap(struct ext2_gd *, struct ext2_gd *, int);
350 #define e2fs_sbload(old, new) e2fs_sb_bswap((old), (new))
351 #define e2fs_cgload(old, new, size) e2fs_cg_bswap((old), (new), (size));
352 #define e2fs_sbsave(old, new) e2fs_sb_bswap((old), (new))
353 #define e2fs_cgsave(old, new, size) e2fs_cg_bswap((old), (new), (size));
354 #endif
355
356 /*
357 * Turn file system block numbers into disk block addresses.
358 * This maps file system blocks to device size blocks.
359 */
360 #define fsbtodb(fs, b) ((b) << (fs)->e2fs_fsbtodb)
361 #define dbtofsb(fs, b) ((b) >> (fs)->e2fs_fsbtodb)
362
363 /*
364 * Macros for handling inode numbers:
365 * inode number to file system block offset.
366 * inode number to cylinder group number.
367 * inode number to file system block address.
368 */
369 #define ino_to_cg(fs, x) (((x) - 1) / (fs)->e2fs.e2fs_ipg)
370 #define ino_to_fsba(fs, x) \
371 ((fs)->e2fs_gd[ino_to_cg(fs, x)].ext2bgd_i_tables + \
372 (((x)-1) % (fs)->e2fs.e2fs_ipg)/(fs)->e2fs_ipb)
373 #define ino_to_fsbo(fs, x) (((x)-1) % (fs)->e2fs_ipb)
374
375 /*
376 * Give cylinder group number for a file system block.
377 * Give cylinder group block number for a file system block.
378 */
379 #define dtog(fs, d) (((d) - (fs)->e2fs.e2fs_first_dblock) / (fs)->e2fs.e2fs_fpg)
380 #define dtogd(fs, d) \
381 (((d) - (fs)->e2fs.e2fs_first_dblock) % (fs)->e2fs.e2fs_fpg)
382
383 /*
384 * The following macros optimize certain frequently calculated
385 * quantities by using shifts and masks in place of divisions
386 * modulos and multiplications.
387 */
388 #define blkoff(fs, loc) /* calculates (loc % fs->e2fs_bsize) */ \
389 ((loc) & (fs)->e2fs_qbmask)
390 #define lblktosize(fs, blk) /* calculates (blk * fs->e2fs_bsize) */ \
391 ((blk) << (fs)->e2fs_bshift)
392 #define lblkno(fs, loc) /* calculates (loc / fs->e2fs_bsize) */ \
393 ((loc) >> (fs)->e2fs_bshift)
394 #define blkroundup(fs, size) /* calculates roundup(size, fs->e2fs_bsize) */ \
395 (((size) + (fs)->e2fs_qbmask) & (fs)->e2fs_bmask)
396 #define fragroundup(fs, size) /* calculates roundup(size, fs->e2fs_bsize) */ \
397 (((size) + (fs)->e2fs_qbmask) & (fs)->e2fs_bmask)
398 /*
399 * Determine the number of available frags given a
400 * percentage to hold in reserve.
401 */
402 #define freespace(fs) \
403 ((fs)->e2fs.e2fs_fbcount - (fs)->e2fs.e2fs_rbcount)
404
405 /*
406 * Number of indirects in a file system block.
407 */
408 #define NINDIR(fs) ((fs)->e2fs_bsize / sizeof(u_int32_t))
409