ffs_subr.c 22.8 KB
Newer Older
1
/*-
2
3
 * SPDX-License-Identifier: BSD-3-Clause
 *
Rodney W. Grimes's avatar
Rodney W. Grimes committed
4
5
6
7
8
9
10
11
12
13
14
 * Copyright (c) 1982, 1986, 1989, 1993
 *	The Regents of the University of California.  All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
Ed Maste's avatar
Ed Maste committed
15
 * 3. Neither the name of the University nor the names of its contributors
Rodney W. Grimes's avatar
Rodney W. Grimes committed
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
31
 *	@(#)ffs_subr.c	8.5 (Berkeley) 3/21/95
Rodney W. Grimes's avatar
Rodney W. Grimes committed
32
33
 */

David E. O'Brien's avatar
David E. O'Brien committed
34
35
36
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

Rodney W. Grimes's avatar
Rodney W. Grimes committed
37
#include <sys/param.h>
38
#include <sys/limits.h>
Rodney W. Grimes's avatar
Rodney W. Grimes committed
39

40
#ifndef _KERNEL
41
42
43
44
45
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <time.h>
#include <sys/errno.h>
46
#include <ufs/ufs/dinode.h>
47
#include <ufs/ffs/fs.h>
48

49
uint32_t calculate_crc32c(uint32_t, const void *, size_t);
50
uint32_t ffs_calc_sbhash(struct fs *);
51
52
53
struct malloc_type;
#define UFS_MALLOC(size, type, flags) malloc(size)
#define UFS_FREE(ptr, type) free(ptr)
54
#define maxphys MAXPHYS
55
56

#else /* _KERNEL */
Rodney W. Grimes's avatar
Rodney W. Grimes committed
57
#include <sys/systm.h>
58
#include <sys/gsb_crc32.h>
59
#include <sys/lock.h>
60
61
#include <sys/malloc.h>
#include <sys/mount.h>
Rodney W. Grimes's avatar
Rodney W. Grimes committed
62
#include <sys/vnode.h>
63
#include <sys/bio.h>
Rodney W. Grimes's avatar
Rodney W. Grimes committed
64
#include <sys/buf.h>
65
66
#include <sys/ucred.h>

Rodney W. Grimes's avatar
Rodney W. Grimes committed
67
68
#include <ufs/ufs/quota.h>
#include <ufs/ufs/inode.h>
69
70
71
#include <ufs/ufs/extattr.h>
#include <ufs/ufs/ufsmount.h>
#include <ufs/ufs/ufs_extern.h>
72
#include <ufs/ffs/ffs_extern.h>
73
#include <ufs/ffs/fs.h>
Rodney W. Grimes's avatar
Rodney W. Grimes committed
74

75
76
77
#define UFS_MALLOC(size, type, flags) malloc(size, type, flags)
#define UFS_FREE(ptr, type) free(ptr, type)

78
#endif /* _KERNEL */
Rodney W. Grimes's avatar
Rodney W. Grimes committed
79

80
81
82
83
84
85
/*
 * Verify an inode check-hash.
 */
int
ffs_verify_dinode_ckhash(struct fs *fs, struct ufs2_dinode *dip)
{
86
	uint32_t ckhash, save_ckhash;
87
88
89
90
91
92
93
94
95
96
97
98

	/*
	 * Return success if unallocated or we are not doing inode check-hash.
	 */
	if (dip->di_mode == 0 || (fs->fs_metackhash & CK_INODE) == 0)
		return (0);
	/*
	 * Exclude di_ckhash from the crc32 calculation, e.g., always use
	 * a check-hash value of zero when calculating the check-hash.
	 */
	save_ckhash = dip->di_ckhash;
	dip->di_ckhash = 0;
99
	ckhash = calculate_crc32c(~0L, (void *)dip, sizeof(*dip));
100
	dip->di_ckhash = save_ckhash;
101
102
103
	if (save_ckhash == ckhash)
		return (0);
	return (EINVAL);
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
}

/*
 * Update an inode check-hash.
 */
void
ffs_update_dinode_ckhash(struct fs *fs, struct ufs2_dinode *dip)
{

	if (dip->di_mode == 0 || (fs->fs_metackhash & CK_INODE) == 0)
		return;
	/*
	 * Exclude old di_ckhash from the crc32 calculation, e.g., always use
	 * a check-hash value of zero when calculating the new check-hash.
	 */
	dip->di_ckhash = 0;
	dip->di_ckhash = calculate_crc32c(~0L, (void *)dip, sizeof(*dip));
}

123
124
125
126
127
/*
 * These are the low-level functions that actually read and write
 * the superblock and its associated data.
 */
static off_t sblock_try[] = SBLOCKSEARCH;
128
static int readsuper(void *, struct fs **, off_t, int, int,
129
	int (*)(void *, off_t, void **, int));
130
static int validate_sblock(struct fs *, int);
131
132
133
134
135
136
137
138
139

/*
 * Read a superblock from the devfd device.
 *
 * If an alternate superblock is specified, it is read. Otherwise the
 * set of locations given in the SBLOCKSEARCH list is searched for a
 * superblock. Memory is allocated for the superblock by the readfunc and
 * is returned. If filltype is non-NULL, additional memory is allocated
 * of type filltype and filled in with the superblock summary information.
140
 * All memory is freed when any error is returned.
141
142
143
144
145
146
 *
 * If a superblock is found, zero is returned. Otherwise one of the
 * following error values is returned:
 *     EIO: non-existent or truncated superblock.
 *     EIO: error reading summary information.
 *     ENOENT: no usable known superblock found.
147
 *     ENOMEM: failed to allocate space for the superblock.
148
149
150
151
 *     EINVAL: The previous newfs operation on this volume did not complete.
 *         The administrator must complete newfs before using this volume.
 */
int
152
ffs_sbget(void *devfd, struct fs **fsp, off_t altsblock,
153
154
155
156
    struct malloc_type *filltype,
    int (*readfunc)(void *devfd, off_t loc, void **bufp, int size))
{
	struct fs *fs;
157
	struct fs_summary_info *fs_si;
158
159
	int i, error;
	uint64_t size, blks;
160
161
162
163
	uint8_t *space;
	int32_t *lp;
	char *buf;

164
	fs = NULL;
165
	*fsp = NULL;
166
	if (altsblock >= 0) {
167
		if ((error = readsuper(devfd, &fs, altsblock, 1, 0,
168
169
170
		     readfunc)) != 0) {
			if (fs != NULL)
				UFS_FREE(fs, filltype);
171
			return (error);
172
		}
173
174
	} else {
		for (i = 0; sblock_try[i] != -1; i++) {
175
			if ((error = readsuper(devfd, &fs, sblock_try[i], 0,
176
			     altsblock, readfunc)) == 0)
177
				break;
178
179
180
181
			if (fs != NULL) {
				UFS_FREE(fs, filltype);
				fs = NULL;
			}
182
			if (error == ENOENT)
183
				continue;
184
			return (error);
185
186
187
188
189
190
191
192
193
194
195
196
		}
		if (sblock_try[i] == -1)
			return (ENOENT);
	}
	/*
	 * Read in the superblock summary information.
	 */
	size = fs->fs_cssize;
	blks = howmany(size, fs->fs_fsize);
	if (fs->fs_contigsumsize > 0)
		size += fs->fs_ncg * sizeof(int32_t);
	size += fs->fs_ncg * sizeof(u_int8_t);
197
	if ((fs_si = UFS_MALLOC(sizeof(*fs_si), filltype, M_NOWAIT)) == NULL) {
198
		UFS_FREE(fs, filltype);
199
		return (ENOMEM);
200
201
202
	}
	bzero(fs_si, sizeof(*fs_si));
	fs->fs_si = fs_si;
203
	if ((space = UFS_MALLOC(size, filltype, M_NOWAIT)) == NULL) {
204
		UFS_FREE(fs->fs_si, filltype);
205
		UFS_FREE(fs, filltype);
206
		return (ENOMEM);
207
	}
208
209
210
211
212
	fs->fs_csp = (struct csum *)space;
	for (i = 0; i < blks; i += fs->fs_frag) {
		size = fs->fs_bsize;
		if (i + fs->fs_frag > blks)
			size = (blks - i) * fs->fs_fsize;
213
		buf = NULL;
214
		error = (*readfunc)(devfd,
215
		    dbtob(fsbtodb(fs, fs->fs_csaddr + i)), (void **)&buf, size);
216
		if (error) {
217
218
			if (buf != NULL)
				UFS_FREE(buf, filltype);
219
			UFS_FREE(fs->fs_csp, filltype);
220
			UFS_FREE(fs->fs_si, filltype);
221
			UFS_FREE(fs, filltype);
222
			return (error);
223
224
225
226
227
228
229
230
231
232
233
234
235
236
		}
		memcpy(space, buf, size);
		UFS_FREE(buf, filltype);
		space += size;
	}
	if (fs->fs_contigsumsize > 0) {
		fs->fs_maxcluster = lp = (int32_t *)space;
		for (i = 0; i < fs->fs_ncg; i++)
			*lp++ = fs->fs_contigsumsize;
		space = (uint8_t *)lp;
	}
	size = fs->fs_ncg * sizeof(u_int8_t);
	fs->fs_contigdirs = (u_int8_t *)space;
	bzero(fs->fs_contigdirs, size);
237
	*fsp = fs;
238
239
240
241
242
243
244
245
	return (0);
}

/*
 * Try to read a superblock from the location specified by sblockloc.
 * Return zero on success or an errno on failure.
 */
static int
246
readsuper(void *devfd, struct fs **fsp, off_t sblockloc, int isaltsblk,
247
    int chkhash, int (*readfunc)(void *devfd, off_t loc, void **bufp, int size))
248
249
{
	struct fs *fs;
250
251
	int error, res;
	uint32_t ckhash;
252
253
254
255
256
257
258

	error = (*readfunc)(devfd, sblockloc, (void **)fsp, SBLOCKSIZE);
	if (error != 0)
		return (error);
	fs = *fsp;
	if (fs->fs_magic == FS_BAD_MAGIC)
		return (EINVAL);
259
260
	if ((error = validate_sblock(fs, isaltsblk)) != 0)
		return (error);
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
	/*
	 * If the filesystem has been run on a kernel without
	 * metadata check hashes, disable them.
	 */
	if ((fs->fs_flags & FS_METACKHASH) == 0)
		fs->fs_metackhash = 0;
	/*
	 * Clear any check-hashes that are not maintained
	 * by this kernel. Also clear any unsupported flags.
	 */
	fs->fs_metackhash &= CK_SUPPORTED;
	fs->fs_flags &= FS_SUPPORTED;
	if (fs->fs_ckhash != (ckhash = ffs_calc_sbhash(fs))) {
		if (chkhash == STDSB_NOMSG)
			return (EINTEGRITY);
		if (chkhash == STDSB_NOHASHFAIL_NOMSG)
			return (0);
278
#ifdef _KERNEL
279
280
281
282
		res = uprintf("Superblock check-hash failed: recorded "
		    "check-hash 0x%x != computed check-hash 0x%x%s\n",
		    fs->fs_ckhash, ckhash,
		    chkhash == STDSB_NOHASHFAIL ? " (Ignored)" : "");
283
#else
284
		res = 0;
285
#endif
286
287
288
289
290
291
292
293
294
295
296
297
298
		/*
		 * Print check-hash failure if no controlling terminal
		 * in kernel or always if in user-mode (libufs).
		 */
		if (res == 0)
			printf("Superblock check-hash failed: recorded "
			    "check-hash 0x%x != computed check-hash "
			    "0x%x%s\n", fs->fs_ckhash, ckhash,
			    chkhash == STDSB_NOHASHFAIL ?
			    " (Ignored)" : "");
		if (chkhash == STDSB)
			return (EINTEGRITY);
		/* chkhash == STDSB_NOHASHFAIL */
299
300
		return (0);
	}
301
302
303
304
305
	/* Have to set for old filesystems that predate this field */
	fs->fs_sblockactualloc = sblockloc;
	/* Not yet any summary information */
	fs->fs_si = NULL;
	return (0);
306
307
}

308
309
310
311
/*
 * Verify the filesystem values.
 */
#define ILOG2(num) (fls(num) - 1)
Kirk McKusick's avatar
Kirk McKusick committed
312
#undef CHK
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
#define CHK(lhs, op, rhs, fmt)						\
	if (lhs op rhs) {						\
		printf("UFS%d superblock failed: %s (" #fmt ") %s %s ("	\
		    #fmt ")\n", fs->fs_magic == FS_UFS1_MAGIC ? 1 : 2,	\
		    #lhs, (intmax_t)lhs, #op, #rhs, (intmax_t)rhs);	\
		return (ENOENT);					\
	}
#define CHK2(lhs1, op1, rhs1, lhs2, op2, rhs2, fmt)			\
	if (lhs1 op1 rhs1 && lhs2 op2 rhs2) {				\
		printf("UFS%d superblock failed: %s (" #fmt ") %s %s ("	\
		    #fmt ") && %s (" #fmt ") %s %s (" #fmt ")\n",	\
		    fs->fs_magic == FS_UFS1_MAGIC ? 1 : 2, #lhs1, 	\
		    (intmax_t)lhs1, #op1, #rhs1, (intmax_t)rhs1, #lhs2,	\
		    (intmax_t)lhs2, #op2, #rhs2, (intmax_t)rhs2);	\
		return (ENOENT);					\
	}
329
330
331
332
333
334
335
336
337

static int
validate_sblock(struct fs *fs, int isaltsblk)
{
	int i, sectorsize;
	u_int64_t maxfilesize, minfpg, sizepb;

	sectorsize = dbtob(1);
	if (fs->fs_magic == FS_UFS2_MAGIC) {
338
339
340
341
342
343
344
345
346
347
348
		if (!isaltsblk) {
			CHK(fs->fs_sblockloc, !=, SBLOCK_UFS2, %#jx);
			CHK2(fs->fs_sblockactualloc, !=, SBLOCK_UFS2,
			    fs->fs_sblockactualloc, !=, 0, %jd);
		}
		CHK(fs->fs_maxsymlinklen, !=, ((UFS_NDADDR + UFS_NIADDR) *
			sizeof(ufs2_daddr_t)), %jd);
		CHK(fs->fs_nindir, !=, fs->fs_bsize / sizeof(ufs2_daddr_t),
		    %jd);
		CHK(fs->fs_inopb, !=, fs->fs_bsize / sizeof(struct ufs2_dinode),
		    %jd);
349
	} else if (fs->fs_magic == FS_UFS1_MAGIC) {
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
		if (!isaltsblk) {
			CHK(fs->fs_sblockloc, >, SBLOCK_UFS1, %#jx);
			CHK2(fs->fs_sblockactualloc, !=, SBLOCK_UFS1,
			    fs->fs_sblockactualloc, !=, 0, %jd);
		}
		CHK(fs->fs_nindir, !=, fs->fs_bsize / sizeof(ufs1_daddr_t),
		    %jd);
		CHK(fs->fs_inopb, !=, fs->fs_bsize / sizeof(struct ufs1_dinode),
		    %jd);
		CHK(fs->fs_maxsymlinklen, !=, ((UFS_NDADDR + UFS_NIADDR) *
			sizeof(ufs1_daddr_t)), %jd);
		CHK(fs->fs_old_inodefmt, !=, FS_44INODEFMT, %jd);
		CHK(fs->fs_old_cgoffset, !=, 0, %jd);
		CHK(fs->fs_old_cgmask, !=, 0xffffffff, %#jx);
		CHK(fs->fs_old_rotdelay, !=, 0, %jd);
		CHK(fs->fs_old_rps, !=, 60, %jd);
		CHK(fs->fs_old_nspf, !=, fs->fs_fsize / sectorsize, %jd);
		CHK(fs->fs_old_cpg, !=, 1, %jd);
		CHK(fs->fs_old_interleave, !=, 1, %jd);
		CHK(fs->fs_old_trackskew, !=, 0, %jd);
		CHK(fs->fs_old_cpc, !=, 0, %jd);
		CHK(fs->fs_old_postblformat, !=, 1, %jd);
		CHK(fs->fs_old_nrpos, !=, 1, %jd);
		CHK(fs->fs_old_spc, !=, fs->fs_fpg * fs->fs_old_nspf, %jd);
		CHK(fs->fs_old_nsect, !=, fs->fs_old_spc, %jd);
		CHK(fs->fs_old_npsect, !=, fs->fs_old_spc, %jd);
		CHK(fs->fs_old_ncyl, !=, fs->fs_ncg, %jd);
377
	} else {
378
		/* Bad magic number, so assume not a superblock */
379
380
		return (ENOENT);
	}
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
	CHK(fs->fs_bsize, <, MINBSIZE, %jd);
	CHK(fs->fs_bsize, >, MAXBSIZE, %jd);
	CHK(fs->fs_bsize, <, roundup(sizeof(struct fs), DEV_BSIZE), %jd);
	CHK(fs->fs_sbsize, >, SBLOCKSIZE, %jd);
	CHK(fs->fs_sbsize, <, fs->fs_fsize, %jd);
	CHK(powerof2(fs->fs_bsize), ==, 0, %jd);
	CHK(fs->fs_fsize, <, sectorsize, %jd);
	CHK(fs->fs_fsize, >, fs->fs_bsize, %jd);
	CHK(fs->fs_fsize * MAXFRAG, <, fs->fs_bsize, %jd);
	CHK(powerof2(fs->fs_fsize), ==, 0, %jd);
	CHK(fs->fs_maxbsize, <, fs->fs_bsize, %jd);
	CHK(powerof2(fs->fs_maxbsize), ==, 0, %jd);
	CHK(fs->fs_maxbsize, >, FS_MAXCONTIG * fs->fs_bsize, %jd);
	CHK(fs->fs_bmask, !=, ~(fs->fs_bsize - 1), %#jx);
	CHK(fs->fs_fmask, !=, ~(fs->fs_fsize - 1), %#jx);
	CHK(fs->fs_qbmask, !=, ~fs->fs_bmask, %#jx);
	CHK(fs->fs_qfmask, !=, ~fs->fs_fmask, %#jx);
	CHK(fs->fs_bshift, !=, ILOG2(fs->fs_bsize), %jd);
	CHK(fs->fs_fshift, !=, ILOG2(fs->fs_fsize), %jd);
	CHK(fs->fs_frag, !=, numfrags(fs, fs->fs_bsize), %jd);
	CHK(fs->fs_fragshift, !=, ILOG2(fs->fs_frag), %jd);
	CHK(fs->fs_frag, >, MAXFRAG, %jd);
	CHK(fs->fs_fsbtodb, !=, ILOG2(fs->fs_fsize / sectorsize), %jd);
	CHK(fs->fs_sblkno, !=, roundup(
	    howmany(fs->fs_sblockloc + SBLOCKSIZE, fs->fs_fsize),
	    fs->fs_frag), %jd);
	CHK(fs->fs_cblkno, !=, fs->fs_sblkno +
	    roundup(howmany(SBLOCKSIZE, fs->fs_fsize), fs->fs_frag), %jd);
	CHK(fs->fs_iblkno, !=, fs->fs_cblkno + fs->fs_frag, %jd);
	CHK(fs->fs_dblkno, !=, fs->fs_iblkno + fs->fs_ipg / INOPF(fs), %jd);
	CHK(fs->fs_cgsize, >, fs->fs_bsize, %jd);
	CHK(fs->fs_csaddr, !=, cgdmin(fs, 0), %jd);
	CHK(fs->fs_cssize, !=,
		fragroundup(fs, fs->fs_ncg * sizeof(struct csum)), %jd);
	CHK(fs->fs_dsize, !=, fs->fs_size - fs->fs_sblkno -
416
		fs->fs_ncg * (fs->fs_dblkno - fs->fs_sblkno) -
417
418
419
420
		howmany(fs->fs_cssize, fs->fs_fsize), %jd);
	CHK(fs->fs_metaspace, <, 0, %jd);
	CHK(fs->fs_metaspace, >, fs->fs_fpg / 2, %jd);
	CHK(fs->fs_minfree, >, 99, %jd%%);
421
422
423
424
425
	maxfilesize = fs->fs_bsize * UFS_NDADDR - 1;
	for (sizepb = fs->fs_bsize, i = 0; i < UFS_NIADDR; i++) {
		sizepb *= NINDIR(fs);
		maxfilesize += sizepb;
	}
426
	CHK(fs->fs_maxfilesize, !=, maxfilesize, %jd);
427
428
429
430
431
432
433
434
435
436
437
438
439
	/*
	 * These values have a tight interaction with each other that
	 * makes it hard to tightly bound them. So we can only check
	 * that they are within a broader possible range.
	 *
	 * Calculate minfpg, the minimum number of fragments that can be
	 * in a cylinder group. The value 12289 is calculated in newfs(8)
	 * when creating the smallest block size UFS version 1 filesystem
	 * (4096 block size) with no fragments (4096 fragment size). That
	 * number may be depressed even further for very small filesystems
	 * since newfs(8) strives to have at least four cylinder groups.
	 */
	minfpg = MIN(12289, fs->fs_size / 4);
440
441
442
443
444
445
446
447
448
449
	CHK(fs->fs_ncg, <, 1, %jd);
	CHK(fs->fs_ncg, >, (fs->fs_size / minfpg) + 1, %jd);
	CHK(fs->fs_fpg, <, minfpg, %jd);
	CHK(fs->fs_fpg, >, fs->fs_size, %jd);
	CHK(fs->fs_ipg * fs->fs_ncg, >, (((int64_t)(1)) << 32) - INOPB(fs),
	    %jd);
	CHK(fs->fs_ipg, >, fs->fs_fpg, %jd);
	CHK(fs->fs_size, <, 8 * fs->fs_frag, %jd);
	CHK(fs->fs_size, <=, (fs->fs_ncg - 1) * fs->fs_fpg, %jd);
	CHK(fs->fs_size, >, fs->fs_ncg * fs->fs_fpg, %jd);
450
	/*
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
	 * With file system clustering it is possible to allocate
	 * many contiguous blocks. The kernel variable maxphys defines
	 * the maximum transfer size permitted by the controller and/or
	 * buffering. The fs_maxcontig parameter controls the maximum
	 * number of blocks that the filesystem will read or write
	 * in a single transfer. It is calculated when the filesystem
	 * is created as maxphys / fs_bsize. The loader uses a maxphys
	 * of 128K even when running on a system that supports larger
	 * values. If the filesystem was built on a system that supports
	 * a larger maxphys (1M is typical) it will have configured
	 * fs_maxcontig for that larger system. So we bound the upper
	 * allowable limit for fs_maxconfig to be able to at least 
	 * work with a 1M maxphys on the smallest block size filesystem:
	 * 1M / 4096 == 256. There is no harm in allowing the mounting of
	 * filesystems that make larger than maxphys I/O requests because
	 * those (mostly 32-bit machines) can (very slowly) handle I/O
	 * requests that exceed maxphys.
468
	 */
469
470
471
472
473
474
	CHK(fs->fs_maxcontig, <, 1, %jd);
	CHK(fs->fs_maxcontig, >, MAX(256, maxphys / fs->fs_bsize), %jd);
	CHK(fs->fs_maxcontig, <, 0, %jd);
	CHK2(fs->fs_maxcontig, ==, 0, fs->fs_contigsumsize, !=, 0, %jd);
	CHK2(fs->fs_maxcontig, >, 1, fs->fs_contigsumsize, !=,
	    MIN(fs->fs_maxcontig, FS_MAXCONTIG), %jd);
475
476
477
	return (0);
}

478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
/*
 * Write a superblock to the devfd device from the memory pointed to by fs.
 * Write out the superblock summary information if it is present.
 *
 * If the write is successful, zero is returned. Otherwise one of the
 * following error values is returned:
 *     EIO: failed to write superblock.
 *     EIO: failed to write superblock summary information.
 */
int
ffs_sbput(void *devfd, struct fs *fs, off_t loc,
    int (*writefunc)(void *devfd, off_t loc, void *buf, int size))
{
	int i, error, blks, size;
	uint8_t *space;

	/*
	 * If there is summary information, write it first, so if there
	 * is an error, the superblock will not be marked as clean.
	 */
498
	if (fs->fs_si != NULL && fs->fs_csp != NULL) {
499
500
501
502
503
504
505
506
507
508
509
510
511
512
		blks = howmany(fs->fs_cssize, fs->fs_fsize);
		space = (uint8_t *)fs->fs_csp;
		for (i = 0; i < blks; i += fs->fs_frag) {
			size = fs->fs_bsize;
			if (i + fs->fs_frag > blks)
				size = (blks - i) * fs->fs_fsize;
			if ((error = (*writefunc)(devfd,
			     dbtob(fsbtodb(fs, fs->fs_csaddr + i)),
			     space, size)) != 0)
				return (error);
			space += size;
		}
	}
	fs->fs_fmod = 0;
513
514
515
516
517
518
519
520
521
522
523
524
525
526
#ifndef _KERNEL
	{
		struct fs_summary_info *fs_si;

		fs->fs_time = time(NULL);
		/* Clear the pointers for the duration of writing. */
		fs_si = fs->fs_si;
		fs->fs_si = NULL;
		fs->fs_ckhash = ffs_calc_sbhash(fs);
		error = (*writefunc)(devfd, loc, fs, fs->fs_sbsize);
		fs->fs_si = fs_si;
	}
#else /* _KERNEL */
	fs->fs_time = time_second;
527
	fs->fs_ckhash = ffs_calc_sbhash(fs);
528
529
530
	error = (*writefunc)(devfd, loc, fs, fs->fs_sbsize);
#endif /* _KERNEL */
	return (error);
531
532
}

533
534
535
/*
 * Calculate the check-hash for a superblock.
 */
536
537
uint32_t
ffs_calc_sbhash(struct fs *fs)
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
{
	uint32_t ckhash, save_ckhash;

	/*
	 * A filesystem that was using a superblock ckhash may be moved
	 * to an older kernel that does not support ckhashes. The
	 * older kernel will clear the FS_METACKHASH flag indicating
	 * that it does not update hashes. When the disk is moved back
	 * to a kernel capable of ckhashes it disables them on mount:
	 *
	 *	if ((fs->fs_flags & FS_METACKHASH) == 0)
	 *		fs->fs_metackhash = 0;
	 *
	 * This leaves (fs->fs_metackhash & CK_SUPERBLOCK) == 0) with an
	 * old stale value in the fs->fs_ckhash field. Thus the need to
	 * just accept what is there.
	 */
	if ((fs->fs_metackhash & CK_SUPERBLOCK) == 0)
		return (fs->fs_ckhash);

	save_ckhash = fs->fs_ckhash;
	fs->fs_ckhash = 0;
	/*
	 * If newly read from disk, the caller is responsible for
	 * verifying that fs->fs_sbsize <= SBLOCKSIZE.
	 */
	ckhash = calculate_crc32c(~0L, (void *)fs, fs->fs_sbsize);
	fs->fs_ckhash = save_ckhash;
	return (ckhash);
}

Rodney W. Grimes's avatar
Rodney W. Grimes committed
569
/*
Rodney W. Grimes's avatar
Rodney W. Grimes committed
570
 * Update the frsum fields to reflect addition or deletion
Rodney W. Grimes's avatar
Rodney W. Grimes committed
571
572
573
 * of some frags.
 */
void
Ed Maste's avatar
Ed Maste committed
574
ffs_fragacct(struct fs *fs, int fragmap, int32_t fraglist[], int cnt)
Rodney W. Grimes's avatar
Rodney W. Grimes committed
575
576
{
	int inblk;
Poul-Henning Kamp's avatar
Poul-Henning Kamp committed
577
578
	int field, subfield;
	int siz, pos;
Rodney W. Grimes's avatar
Rodney W. Grimes committed
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605

	inblk = (int)(fragtbl[fs->fs_frag][fragmap]) << 1;
	fragmap <<= 1;
	for (siz = 1; siz < fs->fs_frag; siz++) {
		if ((inblk & (1 << (siz + (fs->fs_frag % NBBY)))) == 0)
			continue;
		field = around[siz];
		subfield = inside[siz];
		for (pos = siz; pos <= fs->fs_frag; pos++) {
			if ((fragmap & field) == subfield) {
				fraglist[siz] += cnt;
				pos += siz;
				field <<= siz;
				subfield <<= siz;
			}
			field <<= 1;
			subfield <<= 1;
		}
	}
}

/*
 * block operations
 *
 * check if a block is available
 */
int
Ed Maste's avatar
Ed Maste committed
606
ffs_isblock(struct fs *fs, unsigned char *cp, ufs1_daddr_t h)
Rodney W. Grimes's avatar
Rodney W. Grimes committed
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
{
	unsigned char mask;

	switch ((int)fs->fs_frag) {
	case 8:
		return (cp[h] == 0xff);
	case 4:
		mask = 0x0f << ((h & 0x1) << 2);
		return ((cp[h >> 1] & mask) == mask);
	case 2:
		mask = 0x03 << ((h & 0x3) << 1);
		return ((cp[h >> 2] & mask) == mask);
	case 1:
		mask = 0x01 << (h & 0x7);
		return ((cp[h >> 3] & mask) == mask);
	default:
623
#ifdef _KERNEL
Rodney W. Grimes's avatar
Rodney W. Grimes committed
624
		panic("ffs_isblock");
625
626
627
628
629
630
631
632
633
634
#endif
		break;
	}
	return (0);
}

/*
 * check if a block is free
 */
int
Ed Maste's avatar
Ed Maste committed
635
ffs_isfreeblock(struct fs *fs, u_char *cp, ufs1_daddr_t h)
636
{
637

638
639
640
641
642
643
644
645
646
647
648
649
650
651
	switch ((int)fs->fs_frag) {
	case 8:
		return (cp[h] == 0);
	case 4:
		return ((cp[h >> 1] & (0x0f << ((h & 0x1) << 2))) == 0);
	case 2:
		return ((cp[h >> 2] & (0x03 << ((h & 0x3) << 1))) == 0);
	case 1:
		return ((cp[h >> 3] & (0x01 << (h & 0x7))) == 0);
	default:
#ifdef _KERNEL
		panic("ffs_isfreeblock");
#endif
		break;
Rodney W. Grimes's avatar
Rodney W. Grimes committed
652
	}
653
	return (0);
Rodney W. Grimes's avatar
Rodney W. Grimes committed
654
655
656
657
658
659
}

/*
 * take a block out of the map
 */
void
Ed Maste's avatar
Ed Maste committed
660
ffs_clrblock(struct fs *fs, u_char *cp, ufs1_daddr_t h)
Rodney W. Grimes's avatar
Rodney W. Grimes committed
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
{

	switch ((int)fs->fs_frag) {
	case 8:
		cp[h] = 0;
		return;
	case 4:
		cp[h >> 1] &= ~(0x0f << ((h & 0x1) << 2));
		return;
	case 2:
		cp[h >> 2] &= ~(0x03 << ((h & 0x3) << 1));
		return;
	case 1:
		cp[h >> 3] &= ~(0x01 << (h & 0x7));
		return;
	default:
677
#ifdef _KERNEL
Rodney W. Grimes's avatar
Rodney W. Grimes committed
678
		panic("ffs_clrblock");
679
680
#endif
		break;
Rodney W. Grimes's avatar
Rodney W. Grimes committed
681
682
683
684
685
686
687
	}
}

/*
 * put a block into the map
 */
void
Ed Maste's avatar
Ed Maste committed
688
ffs_setblock(struct fs *fs, unsigned char *cp, ufs1_daddr_t h)
Rodney W. Grimes's avatar
Rodney W. Grimes committed
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
{

	switch ((int)fs->fs_frag) {
	case 8:
		cp[h] = 0xff;
		return;
	case 4:
		cp[h >> 1] |= (0x0f << ((h & 0x1) << 2));
		return;
	case 2:
		cp[h >> 2] |= (0x03 << ((h & 0x3) << 1));
		return;
	case 1:
		cp[h >> 3] |= (0x01 << (h & 0x7));
		return;
	default:
705
#ifdef _KERNEL
Rodney W. Grimes's avatar
Rodney W. Grimes committed
706
		panic("ffs_setblock");
707
708
709
710
711
712
713
714
715
716
717
#endif
		break;
	}
}

/*
 * Update the cluster map because of an allocation or free.
 *
 * Cnt == 1 means free; cnt == -1 means allocating.
 */
void
Ed Maste's avatar
Ed Maste committed
718
ffs_clusteracct(struct fs *fs, struct cg *cgp, ufs1_daddr_t blkno, int cnt)
719
720
721
722
{
	int32_t *sump;
	int32_t *lp;
	u_char *freemapp, *mapp;
723
724
	int i, start, end, forw, back, map;
	u_int bit;
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745

	if (fs->fs_contigsumsize <= 0)
		return;
	freemapp = cg_clustersfree(cgp);
	sump = cg_clustersum(cgp);
	/*
	 * Allocate or clear the actual block.
	 */
	if (cnt > 0)
		setbit(freemapp, blkno);
	else
		clrbit(freemapp, blkno);
	/*
	 * Find the size of the cluster going forward.
	 */
	start = blkno + 1;
	end = start + fs->fs_contigsumsize;
	if (end >= cgp->cg_nclusterblks)
		end = cgp->cg_nclusterblks;
	mapp = &freemapp[start / NBBY];
	map = *mapp++;
746
	bit = 1U << (start % NBBY);
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
	for (i = start; i < end; i++) {
		if ((map & bit) == 0)
			break;
		if ((i & (NBBY - 1)) != (NBBY - 1)) {
			bit <<= 1;
		} else {
			map = *mapp++;
			bit = 1;
		}
	}
	forw = i - start;
	/*
	 * Find the size of the cluster going backward.
	 */
	start = blkno - 1;
	end = start - fs->fs_contigsumsize;
	if (end < 0)
		end = -1;
	mapp = &freemapp[start / NBBY];
	map = *mapp--;
767
	bit = 1U << (start % NBBY);
768
769
770
771
772
773
774
	for (i = start; i > end; i--) {
		if ((map & bit) == 0)
			break;
		if ((i & (NBBY - 1)) != 0) {
			bit >>= 1;
		} else {
			map = *mapp--;
775
			bit = 1U << (NBBY - 1);
776
		}
Rodney W. Grimes's avatar
Rodney W. Grimes committed
777
	}
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
	back = start - i;
	/*
	 * Account for old cluster and the possibly new forward and
	 * back clusters.
	 */
	i = back + forw + 1;
	if (i > fs->fs_contigsumsize)
		i = fs->fs_contigsumsize;
	sump[i] += cnt;
	if (back > 0)
		sump[back] -= cnt;
	if (forw > 0)
		sump[forw] -= cnt;
	/*
	 * Update cluster summary information.
	 */
	lp = &sump[fs->fs_contigsumsize];
	for (i = fs->fs_contigsumsize; i > 0; i--)
		if (*lp-- > 0)
			break;
	fs->fs_maxcluster[cgp->cg_cgx] = i;
Rodney W. Grimes's avatar
Rodney W. Grimes committed
799
}