2 * SPDX-License-Identifier: BSD-3-Clause and BSD-2-Clause-FreeBSD
4 * Copyright (c) 2002 Networks Associates Technology, Inc.
7 * This software was developed for the FreeBSD Project by Marshall
8 * Kirk McKusick and Network Associates Laboratories, the Security
9 * Research Division of Network Associates, Inc. under DARPA/SPAWAR
10 * contract N66001-01-C-8035 ("CBOSS"), as part of the DARPA CHATS
13 * Redistribution and use in source and binary forms, with or without
14 * modification, are permitted provided that the following conditions
16 * 1. Redistributions of source code must retain the above copyright
17 * notice, this list of conditions and the following disclaimer.
18 * 2. Redistributions in binary form must reproduce the above copyright
19 * notice, this list of conditions and the following disclaimer in the
20 * documentation and/or other materials provided with the distribution.
22 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 * Copyright (c) 1980, 1986, 1993
35 * The Regents of the University of California. All rights reserved.
37 * Redistribution and use in source and binary forms, with or without
38 * modification, are permitted provided that the following conditions
40 * 1. Redistributions of source code must retain the above copyright
41 * notice, this list of conditions and the following disclaimer.
42 * 2. Redistributions in binary form must reproduce the above copyright
43 * notice, this list of conditions and the following disclaimer in the
44 * documentation and/or other materials provided with the distribution.
45 * 3. Neither the name of the University nor the names of its contributors
46 * may be used to endorse or promote products derived from this software
47 * without specific prior written permission.
49 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
50 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
51 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
52 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
53 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
54 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
55 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
56 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
57 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
58 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
61 * @(#)fsck.h 8.4 (Berkeley) 5/9/95
72 #include <sys/queue.h>
74 #define MAXDUP 10 /* limit on dup blks (per inode) */
75 #define MAXBAD 10 /* limit on bad blks (per inode) */
76 #define MINBUFS 100 /* minimum number of buffers required */
77 #define INOBUFSIZE 64*1024 /* size of buffer to read inodes in pass1 */
78 #define ZEROBUFSIZE (dev_bsize * 128) /* size of zero buffer used by -Z */
81 struct ufs1_dinode dp1;
82 struct ufs2_dinode dp2;
84 #define DIP(dp, field) \
85 ((sblock.fs_magic == FS_UFS1_MAGIC) ? \
86 (dp)->dp1.field : (dp)->dp2.field)
88 #define DIP_SET(dp, field, val) do { \
89 if (sblock.fs_magic == FS_UFS1_MAGIC) \
90 (dp)->dp1.field = (val); \
92 (dp)->dp2.field = (val); \
96 * Each inode on the file system is described by the following structure.
97 * The linkcnt is initially set to the value in the inode. Each time it
98 * is found during the descent in passes 2, 3, and 4 the count is
99 * decremented. Any inodes whose count is non-zero after pass 4 needs to
100 * have its link count adjusted by the value remaining in ino_linkcnt.
103 u_char ino_state; /* state of inode, see below */
104 u_char ino_type:4; /* type of inode */
105 u_char ino_idtype:4; /* idesc id_type, SNAP or ADDR */
106 u_short ino_linkcnt; /* number of links not found */
111 #define USTATE 0x1 /* inode not allocated */
112 #define FSTATE 0x2 /* inode is file */
113 #define FZLINK 0x3 /* inode is file with a link count of zero */
114 #define DSTATE 0x4 /* inode is directory */
115 #define DZLINK 0x5 /* inode is directory with a zero link count */
116 #define DFOUND 0x6 /* directory found during descent */
117 /* 0x7 UNUSED - see S_IS_DVALID() definition */
118 #define DCLEAR 0x8 /* directory is to be cleared */
119 #define FCLEAR 0x9 /* file is to be cleared */
120 /* DUNFOUND === (state == DSTATE || state == DZLINK) */
121 #define S_IS_DUNFOUND(state) (((state) & ~0x1) == DSTATE)
122 /* DVALID === (state == DSTATE || state == DZLINK || state == DFOUND) */
123 #define S_IS_DVALID(state) (((state) & ~0x3) == DSTATE)
124 #define INO_IS_DUNFOUND(ino) S_IS_DUNFOUND(inoinfo(ino)->ino_state)
125 #define INO_IS_DVALID(ino) S_IS_DVALID(inoinfo(ino)->ino_state)
127 * Inode state information is contained on per cylinder group lists
128 * which are described by the following structure.
130 extern struct inostatlist {
131 long il_numalloced; /* number of inodes allocated in this cg */
132 struct inostat *il_stat;/* inostat info for this cylinder group */
136 * Structure to reference a dinode.
139 struct bufarea *i_bp; /* buffer containing the dinode */
140 union dinode *i_dp; /* pointer to dinode in buffer */
141 ino_t i_number; /* inode number */
145 * Size of hash tables
147 #define HASHSIZE 2048
148 #define HASH(x) ((x * 2654435761) & (HASHSIZE - 1))
151 * buffer cache structure.
154 TAILQ_ENTRY(bufarea) b_list; /* LRU buffer queue */
155 LIST_ENTRY(bufarea) b_hash; /* hash list */
156 ufs2_daddr_t b_bno; /* disk block number */
157 int b_size; /* size of I/O */
158 int b_errs; /* I/O error */
159 int b_flags; /* B_ flags below */
160 int b_type; /* BT_ type below */
161 int b_refcnt; /* ref count of users */
162 int b_index; /* for BT_LEVEL, ptr index */
163 /* for BT_INODES, first inum */
165 char *b_buf; /* buffer space */
166 ufs1_daddr_t *b_indir1; /* UFS1 indirect block */
167 ufs2_daddr_t *b_indir2; /* UFS2 indirect block */
168 struct fs *b_fs; /* super block */
169 struct cg *b_cg; /* cylinder group */
170 struct ufs1_dinode *b_dinode1; /* UFS1 inode block */
171 struct ufs2_dinode *b_dinode2; /* UFS2 inode block */
175 #define IBLK(bp, i) \
176 ((sblock.fs_magic == FS_UFS1_MAGIC) ? \
177 (bp)->b_un.b_indir1[i] : (bp)->b_un.b_indir2[i])
179 #define IBLK_SET(bp, i, val) do { \
180 if (sblock.fs_magic == FS_UFS1_MAGIC) \
181 (bp)->b_un.b_indir1[i] = (val); \
183 (bp)->b_un.b_indir2[i] = (val); \
189 #define B_DIRTY 0x00000001 /* Buffer is dirty */
191 * Type of data in buffer
193 #define BT_UNKNOWN 0 /* Buffer type is unknown */
194 #define BT_SUPERBLK 1 /* Buffer holds a superblock */
195 #define BT_CYLGRP 2 /* Buffer holds a cylinder group map */
196 #define BT_LEVEL1 3 /* Buffer holds single level indirect */
197 #define BT_LEVEL2 4 /* Buffer holds double level indirect */
198 #define BT_LEVEL3 5 /* Buffer holds triple level indirect */
199 #define BT_EXTATTR 6 /* Buffer holds external attribute data */
200 #define BT_INODES 7 /* Buffer holds inodes */
201 #define BT_DIRDATA 8 /* Buffer holds directory data */
202 #define BT_DATA 9 /* Buffer holds user data */
203 #define BT_NUMBUFTYPES 10
208 "Single Level Indirect", \
209 "Double Level Indirect", \
210 "Triple Level Indirect", \
211 "External Attribute", \
213 "Directory Contents", \
215 extern char *buftype[];
216 #define BT_BUFTYPE(type) \
217 type < BT_NUMBUFTYPES ? buftype[type] : buftype[BT_UNKNOWN]
218 extern long readcnt[BT_NUMBUFTYPES];
219 extern long totalreadcnt[BT_NUMBUFTYPES];
220 extern struct timespec readtime[BT_NUMBUFTYPES];
221 extern struct timespec totalreadtime[BT_NUMBUFTYPES];
222 extern struct timespec startprog;
224 extern struct bufarea *icachebp; /* inode cache buffer */
225 extern struct bufarea sblk; /* file system superblock */
226 extern struct bufarea *pdirbp; /* current directory contents */
228 #define dirty(bp) do { \
230 pfatal("SETTING DIRTY FLAG IN READ_ONLY MODE\n"); \
232 (bp)->b_flags |= B_DIRTY; \
234 #define initbarea(bp, type) do { \
235 (bp)->b_bno = (ufs2_daddr_t)-4; \
239 (bp)->b_type = type; \
240 (bp)->b_refcnt = 0; \
244 #define sbdirty() dirty(&sblk)
245 #define sblock (*sblk.b_un.b_fs)
247 enum fixstate {DONTKNOW, NOFIX, FIX, IGNORE};
248 extern ino_t cursnapshot;
251 enum fixstate id_fix; /* policy on fixing errors */
252 int (*id_func)(struct inodesc *);
253 /* function to be applied to blocks of inode */
254 struct bufarea *id_bp; /* ckinode: buffer with indirect pointers */
255 union dinode *id_dp; /* ckinode: dinode being traversed */
256 ino_t id_number; /* inode number described */
257 ino_t id_parent; /* for DATA nodes, their parent */
258 ufs_lbn_t id_lbn; /* logical block number of current block */
259 ufs2_daddr_t id_blkno; /* current block number being examined */
260 int id_level; /* level of indirection of this block */
261 int id_numfrags; /* number of frags contained in block */
262 ufs_lbn_t id_lballoc; /* pass1: last LBN that is allocated */
263 off_t id_filesize; /* for DATA nodes, the size of the directory */
264 ufs2_daddr_t id_entryno;/* for DATA nodes, current entry number */
265 int id_loc; /* for DATA nodes, current location in dir */
266 struct direct *id_dirp; /* for DATA nodes, ptr to current entry */
267 char *id_name; /* for DATA nodes, name to find or enter */
268 char id_type; /* type of descriptor, DATA, ADDR, or SNAP */
271 #define DATA 1 /* a directory */
272 #define SNAP 2 /* a snapshot */
273 #define ADDR 3 /* anything but a directory or a snapshot */
276 * Linked list of duplicate blocks.
278 * The list is composed of two parts. The first part of the
279 * list (from duplist through the node pointed to by muldup)
280 * contains a single copy of each duplicate block that has been
281 * found. The second part of the list (from muldup to the end)
282 * contains duplicate blocks that have been found more than once.
283 * To check if a block has been found as a duplicate it is only
284 * necessary to search from duplist through muldup. To find the
285 * total number of times that a block has been found as a duplicate
286 * the entire list must be searched for occurrences of the block
287 * in question. The following diagram shows a sample list where
288 * w (found twice), x (found once), y (found three times), and z
289 * (found once) are duplicate block numbers:
291 * w -> y -> x -> z -> y -> w -> y
300 extern struct dups *duplist; /* head of dup list */
301 extern struct dups *muldup; /* end of unique duplicate dup block numbers */
304 * Inode cache data structures.
307 SLIST_ENTRY(inoinfo) i_hash; /* hash list */
308 ino_t i_number; /* inode number of this entry */
309 ino_t i_parent; /* inode number of parent */
310 ino_t i_dotdot; /* inode number of `..' */
311 size_t i_isize; /* size of inode */
312 u_int i_flags; /* flags, see below */
313 u_int i_numblks; /* size of block array in bytes */
314 ufs2_daddr_t i_blks[1]; /* actually longer */
316 extern SLIST_HEAD(inohash, inoinfo) *inphash;
317 extern struct inoinfo **inpsort;
319 * flags for struct inoinfo
321 #define INFO_NEW 0x0000001 /* replaced broken directory */
323 extern long dirhash, inplast;
324 extern unsigned long numdirs, listmax;
325 extern long countdirs; /* number of directories we actually found */
327 #define MIBSIZE 3 /* size of fsck sysctl MIBs */
328 extern int adjblkcnt[MIBSIZE]; /* MIB cmd to adjust inode block count */
329 extern int adjrefcnt[MIBSIZE]; /* MIB cmd to adjust inode reference count */
330 extern int adjndir[MIBSIZE]; /* MIB cmd to adjust number of directories */
331 extern int adjnbfree[MIBSIZE]; /* MIB cmd to adjust number of free blocks */
332 extern int adjnifree[MIBSIZE]; /* MIB cmd to adjust number of free inodes */
333 extern int adjnffree[MIBSIZE]; /* MIB cmd to adjust number of free frags */
334 extern int adjnumclusters[MIBSIZE]; /* MIB cmd adjust number of free clusters */
335 extern int freefiles[MIBSIZE]; /* MIB cmd to free a set of files */
336 extern int freedirs[MIBSIZE]; /* MIB cmd to free a set of directories */
337 extern int freeblks[MIBSIZE]; /* MIB cmd to free a set of data blocks */
338 extern int setsize[MIBSIZE]; /* MIB cmd to set inode size */
339 extern struct fsck_cmd cmd; /* sysctl file system update commands */
341 extern int bkgrdcheck; /* determine if background check is possible */
342 extern int bkgrdsumadj; /* whether the kernel has the ability to adjust
343 the superblock summary fields */
344 extern off_t bflag; /* location of alternate super block */
345 extern int bkgrdflag; /* use a snapshot to run on an active system */
346 extern char *blockmap; /* ptr to primary blk allocation map */
347 extern char *cdevname; /* name of device being checked */
348 extern char ckclean; /* only do work if not cleanly unmounted */
349 extern int ckhashadd; /* check hashes to be added */
350 extern char *copybuf; /* buffer to copy snapshot blocks */
351 extern int cvtlevel; /* convert to newer file system format */
352 extern long dev_bsize; /* computed value of DEV_BSIZE */
353 extern u_int real_dev_bsize; /* actual disk sector size, not overridden */
354 extern int debug; /* output debugging info */
355 extern int Eflag; /* delete empty data blocks */
356 extern int fsmodified; /* 1 => write done to file system */
357 extern int fsreadfd; /* file descriptor for reading file system */
358 extern int fswritefd; /* file descriptor for writing file system */
359 extern char havesb; /* superblock has been read */
360 extern int inoopt; /* trim out unused inodes */
361 extern ino_t lfdir; /* lost & found directory inode number */
362 extern int lfmode; /* lost & found directory creation mode */
363 extern const char *lfname; /* lost & found directory name */
364 extern ufs2_daddr_t maxfsblock; /* number of blocks in the file system */
365 extern ino_t maxino; /* number of inodes in file system */
366 extern ufs2_daddr_t n_blks; /* number of blocks in use */
367 extern ino_t n_files; /* number of files in use */
368 extern char nflag; /* assume a no response */
369 extern char preen; /* just fix normal inconsistencies */
370 extern char rerun; /* rerun fsck. Only used in non-preen mode */
371 extern char resolved; /* cleared if unresolved changes => not clean */
372 extern int returntosingle; /* 1 => return to single user mode on exit */
373 extern long secsize; /* actual disk sector size */
374 extern char skipclean; /* skip clean file systems if preening */
375 extern int snapcnt; /* number of active snapshots */
376 extern struct inode snaplist[FSMAXSNAP + 1]; /* list of active snapshots */
377 extern char snapname[BUFSIZ]; /* when doing snapshots, the name of the file */
378 extern int sujrecovery; /* 1 => doing check using the journal */
379 extern int surrender; /* Give up if reads fail */
380 extern char usedsoftdep; /* just fix soft dependency inconsistencies */
381 extern int wantrestart; /* Restart fsck on early termination */
382 extern char yflag; /* assume a yes response */
383 extern int zflag; /* zero unused directory space */
384 extern int Zflag; /* zero empty data blocks */
386 extern volatile sig_atomic_t got_siginfo; /* received a SIGINFO */
387 extern volatile sig_atomic_t got_sigalarm; /* received a SIGALRM */
389 #define clearinode(dp) \
390 if (sblock.fs_magic == FS_UFS1_MAGIC) { \
391 (dp)->dp1 = zino.dp1; \
393 (dp)->dp2 = zino.dp2; \
395 extern union dinode zino;
397 #define setbmap(blkno) setbit(blockmap, blkno)
398 #define testbmap(blkno) isset(blockmap, blkno)
399 #define clrbmap(blkno) clrbit(blockmap, blkno)
407 #define EEXIT 8 /* Standard error exit. */
408 #define ERERUN 16 /* fsck needs to be re-run. */
411 int flushentry(void);
413 * Wrapper for malloc() that flushes the cylinder group cache to try
421 while ((retval = malloc(size)) == NULL)
422 if (flushentry() == 0)
428 * Wrapper for calloc() that flushes the cylinder group cache to try
432 Calloc(size_t cnt, size_t size)
436 while ((retval = calloc(cnt, size)) == NULL)
437 if (flushentry() == 0)
445 void adjust(struct inodesc *, int lcnt);
446 void alarmhandler(int sig);
447 ufs2_daddr_t allocblk(long cg, long frags, ufs2_daddr_t (*checkblkavail)
448 (ufs2_daddr_t blkno, long frags));
449 ino_t allocdir(ino_t parent, ino_t request, int mode);
450 ino_t allocino(ino_t request, int type);
451 void binval(struct bufarea *);
452 void blkerror(ino_t ino, const char *type, ufs2_daddr_t blk);
453 char *blockcheck(char *name);
454 int blread(int fd, char *buf, ufs2_daddr_t blk, long size);
456 void blwrite(int fd, char *buf, ufs2_daddr_t blk, ssize_t size);
457 void blerase(int fd, ufs2_daddr_t blk, long size);
458 void blzero(int fd, ufs2_daddr_t blk, long size);
459 void brelse(struct bufarea *);
460 struct inoinfo *cacheino(union dinode *dp, ino_t inumber);
463 void cgdirty(struct bufarea *);
464 struct bufarea *cglookup(int cg);
465 int changeino(ino_t dir, const char *name, ino_t newnum);
466 void check_blkcnt(struct inode *ip);
467 int check_cgmagic(int cg, struct bufarea *cgbp, int requestrebuild);
468 int chkrange(ufs2_daddr_t blk, int cnt);
469 void ckfini(int markclean);
470 int ckinode(union dinode *dp, struct inodesc *);
471 void clri(struct inodesc *, const char *type, int flag);
472 int clearentry(struct inodesc *);
473 void copyonwrite(struct fs *, struct bufarea *,
474 ufs2_daddr_t (*checkblkavail)(ufs2_daddr_t, long));
475 void direrror(ino_t ino, const char *errmesg);
476 int dirscan(struct inodesc *);
477 int dofix(struct inodesc *, const char *msg);
478 int eascan(struct inodesc *, struct ufs2_dinode *dp);
479 void fileerror(ino_t cwd, ino_t ino, const char *errmesg);
480 void finalIOstats(void);
481 int findino(struct inodesc *);
482 int findname(struct inodesc *);
483 void flush(int fd, struct bufarea *bp);
484 int freeblock(struct inodesc *);
485 void freedirino(ino_t ino, ino_t parent);
486 void freeino(ino_t ino);
487 void freeinodebuf(void);
489 void fsutilinit(void);
490 int ftypeok(union dinode *dp);
491 void getblk(struct bufarea *bp, ufs2_daddr_t blk, long size);
492 struct bufarea *getdatablk(ufs2_daddr_t blkno, long size, int type);
493 struct inoinfo *getinoinfo(ino_t inumber);
494 union dinode *getnextinode(ino_t inumber, int rebuildcg);
495 void getpathname(char *namebuf, ino_t curdir, ino_t ino);
496 void ginode(ino_t, struct inode *);
497 void gjournal_check(const char *filesys);
498 void infohandler(int sig);
499 void irelse(struct inode *);
500 ufs2_daddr_t ino_blkatoff(union dinode *, ino_t, ufs_lbn_t, int *,
502 void inocleanup(void);
503 void inodirty(struct inode *);
504 struct inostat *inoinfo(ino_t inum);
505 void IOstats(char *what);
506 int linkup(ino_t orphan, ino_t parentdir, char *name);
507 int makeentry(ino_t parent, ino_t ino, const char *name);
508 int openfilesys(char *dev);
509 void panic(const char *fmt, ...) __printflike(1, 2);
512 int pass1check(struct inodesc *);
517 void pfatal(const char *fmt, ...) __printflike(1, 2);
518 void propagate(void);
519 void prtbuf(struct bufarea *, const char *, ...) __printflike(2, 3);
520 void prtinode(struct inode *);
521 void pwarn(const char *fmt, ...) __printflike(1, 2);
523 int removecachedino(ino_t);
524 int reply(const char *question);
525 void rwerror(const char *mesg, ufs2_daddr_t blk);
526 void sblock_init(void);
527 void setinodebuf(int, ino_t);
528 int setup(char *dev);
529 int snapblkfree(struct fs *, ufs2_daddr_t, long, ino_t,
530 ufs2_daddr_t (*)(ufs2_daddr_t, long));
531 void snapremove(ino_t);
532 void snapflush(ufs2_daddr_t (*checkblkavail)(ufs2_daddr_t, long));
533 ufs2_daddr_t std_checkblkavail(ufs2_daddr_t blkno, long frags);
534 ufs2_daddr_t suj_checkblkavail(ufs2_daddr_t, long);
535 int suj_check(const char *filesys);
536 void update_maps(struct cg *, struct cg*, int);
538 #endif /* !_FSCK_H_ */