aboutsummaryrefslogtreecommitdiff
path: root/sys/fs/ext2fs/ext2_vfsops.c
blob: 98d4f0a73b2cd10d278bad5183b768d21728449e (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
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
377
378
379
380
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
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
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
569
570
571
572
573
574
575
576
577
578
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
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
/*-
 *  modified for EXT2FS support in Lites 1.1
 *
 *  Aug 1995, Godmar Back (gback@cs.utah.edu)
 *  University of Utah, Department of Computer Science
 */
/*-
 * Copyright (c) 1989, 1991, 1993, 1994
 *	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.
 * 3. Neither the name of the University nor the names of its contributors
 *    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.
 *
 *	@(#)ffs_vfsops.c	8.8 (Berkeley) 4/18/94
 * $FreeBSD$
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/namei.h>
#include <sys/priv.h>
#include <sys/proc.h>
#include <sys/kernel.h>
#include <sys/vnode.h>
#include <sys/mount.h>
#include <sys/bio.h>
#include <sys/buf.h>
#include <sys/conf.h>
#include <sys/endian.h>
#include <sys/fcntl.h>
#include <sys/malloc.h>
#include <sys/stat.h>
#include <sys/mutex.h>

#include <geom/geom.h>
#include <geom/geom_vfs.h>

#include <fs/ext2fs/ext2_mount.h>
#include <fs/ext2fs/inode.h>

#include <fs/ext2fs/fs.h>
#include <fs/ext2fs/ext2fs.h>
#include <fs/ext2fs/ext2_dinode.h>
#include <fs/ext2fs/ext2_extern.h>

static int	ext2_flushfiles(struct mount *mp, int flags, struct thread *td);
static int	ext2_mountfs(struct vnode *, struct mount *);
static int	ext2_reload(struct mount *mp, struct thread *td);
static int	ext2_sbupdate(struct ext2mount *, int);
static int	ext2_cgupdate(struct ext2mount *, int);
static vfs_unmount_t		ext2_unmount;
static vfs_root_t		ext2_root;
static vfs_statfs_t		ext2_statfs;
static vfs_sync_t		ext2_sync;
static vfs_vget_t		ext2_vget;
static vfs_fhtovp_t		ext2_fhtovp;
static vfs_mount_t		ext2_mount;

MALLOC_DEFINE(M_EXT2NODE, "ext2_node", "EXT2 vnode private part");
static MALLOC_DEFINE(M_EXT2MNT, "ext2_mount", "EXT2 mount structure");

static struct vfsops ext2fs_vfsops = {
	.vfs_fhtovp =		ext2_fhtovp,
	.vfs_mount =		ext2_mount,
	.vfs_root =		ext2_root,	/* root inode via vget */
	.vfs_statfs =		ext2_statfs,
	.vfs_sync =		ext2_sync,
	.vfs_unmount =		ext2_unmount,
	.vfs_vget =		ext2_vget,
};

VFS_SET(ext2fs_vfsops, ext2fs, 0);

static int	ext2_check_sb_compat(struct ext2fs *es, struct cdev *dev,
		    int ronly);
static int	compute_sb_data(struct vnode * devvp,
		    struct ext2fs * es, struct m_ext2fs * fs);

static const char *ext2_opts[] = { "acls", "async", "noatime", "noclusterr", 
    "noclusterw", "noexec", "export", "force", "from", "multilabel",
    "suiddir", "nosymfollow", "sync", "union", NULL };

/*
 * VFS Operations.
 *
 * mount system call
 */
static int
ext2_mount(struct mount *mp)
{
	struct vfsoptlist *opts;
	struct vnode *devvp;
	struct thread *td;
	struct ext2mount *ump = NULL;
	struct m_ext2fs *fs;
	struct nameidata nd, *ndp = &nd;
	accmode_t accmode;
	char *path, *fspec;
	int error, flags, len;

	td = curthread;
	opts = mp->mnt_optnew;

	if (vfs_filteropt(opts, ext2_opts))
		return (EINVAL);

	vfs_getopt(opts, "fspath", (void **)&path, NULL);
	/* Double-check the length of path.. */
	if (strlen(path) >= MAXMNTLEN)
		return (ENAMETOOLONG);

	fspec = NULL;
	error = vfs_getopt(opts, "from", (void **)&fspec, &len);
	if (!error && fspec[len - 1] != '\0')
		return (EINVAL);

	/*
	 * If updating, check whether changing from read-only to
	 * read/write; if there is no device name, that's all we do.
	 */
	if (mp->mnt_flag & MNT_UPDATE) {
		ump = VFSTOEXT2(mp);
		fs = ump->um_e2fs;
		error = 0;
		if (fs->e2fs_ronly == 0 &&
		    vfs_flagopt(opts, "ro", NULL, 0)) {
			error = VFS_SYNC(mp, MNT_WAIT);
			if (error)
				return (error);
			flags = WRITECLOSE;
			if (mp->mnt_flag & MNT_FORCE)
				flags |= FORCECLOSE;
			error = ext2_flushfiles(mp, flags, td);
			if (error == 0 && fs->e2fs_wasvalid && ext2_cgupdate(ump, MNT_WAIT) == 0) {
				fs->e2fs->e2fs_state |= E2FS_ISCLEAN;
				ext2_sbupdate(ump, MNT_WAIT);
			}
			fs->e2fs_ronly = 1;
			vfs_flagopt(opts, "ro", &mp->mnt_flag, MNT_RDONLY);
			g_topology_lock();
			g_access(ump->um_cp, 0, -1, 0);
			g_topology_unlock();
		}
		if (!error && (mp->mnt_flag & MNT_RELOAD))
			error = ext2_reload(mp, td);
		if (error)
			return (error);
		devvp = ump->um_devvp;
		if (fs->e2fs_ronly && !vfs_flagopt(opts, "ro", NULL, 0)) {
			if (ext2_check_sb_compat(fs->e2fs, devvp->v_rdev, 0))
				return (EPERM);

			/*
			 * If upgrade to read-write by non-root, then verify
			 * that user has necessary permissions on the device.
			 */
			vn_lock(devvp, LK_EXCLUSIVE | LK_RETRY);
			error = VOP_ACCESS(devvp, VREAD | VWRITE,
			    td->td_ucred, td);
			if (error)
				error = priv_check(td, PRIV_VFS_MOUNT_PERM);
			if (error) {
				VOP_UNLOCK(devvp, 0);
				return (error);
			}
			VOP_UNLOCK(devvp, 0);
			g_topology_lock();
			error = g_access(ump->um_cp, 0, 1, 0);
			g_topology_unlock();
			if (error)
				return (error);

			if ((fs->e2fs->e2fs_state & E2FS_ISCLEAN) == 0 ||
			    (fs->e2fs->e2fs_state & E2FS_ERRORS)) {
				if (mp->mnt_flag & MNT_FORCE) {
					printf(
"WARNING: %s was not properly dismounted\n", fs->e2fs_fsmnt);
				} else {
					printf(
"WARNING: R/W mount of %s denied.  Filesystem is not clean - run fsck\n",
					    fs->e2fs_fsmnt);
					return (EPERM);
				}
			}
			fs->e2fs->e2fs_state &= ~E2FS_ISCLEAN;
			(void)ext2_cgupdate(ump, MNT_WAIT);
			fs->e2fs_ronly = 0;
			MNT_ILOCK(mp);
			mp->mnt_flag &= ~MNT_RDONLY;
			MNT_IUNLOCK(mp);
		}
		if (vfs_flagopt(opts, "export", NULL, 0)) {
			/* Process export requests in vfs_mount.c. */
			return (error);
		}
	}

	/*
	 * Not an update, or updating the name: look up the name
	 * and verify that it refers to a sensible disk device.
	 */
	if (fspec == NULL)
		return (EINVAL);
	NDINIT(ndp, LOOKUP, FOLLOW | LOCKLEAF, UIO_SYSSPACE, fspec, td);
	if ((error = namei(ndp)) != 0)
		return (error);
	NDFREE(ndp, NDF_ONLY_PNBUF);
	devvp = ndp->ni_vp;

	if (!vn_isdisk(devvp, &error)) {
		vput(devvp);
		return (error);
	}

	/*
	 * If mount by non-root, then verify that user has necessary
	 * permissions on the device.
	 *
	 * XXXRW: VOP_ACCESS() enough?
	 */
	accmode = VREAD;
	if ((mp->mnt_flag & MNT_RDONLY) == 0)
		accmode |= VWRITE;
	error = VOP_ACCESS(devvp, accmode, td->td_ucred, td);
	if (error)
		error = priv_check(td, PRIV_VFS_MOUNT_PERM);
	if (error) {
		vput(devvp);
		return (error);
	}

	if ((mp->mnt_flag & MNT_UPDATE) == 0) {
		error = ext2_mountfs(devvp, mp);
	} else {
		if (devvp != ump->um_devvp) {
			vput(devvp);
			return (EINVAL);	/* needs translation */
		} else
			vput(devvp);
	}
	if (error) {
		vrele(devvp);
		return (error);
	}
	ump = VFSTOEXT2(mp);
	fs = ump->um_e2fs;

	/*
	 * Note that this strncpy() is ok because of a check at the start
	 * of ext2_mount().
	 */
	strncpy(fs->e2fs_fsmnt, path, MAXMNTLEN);
	fs->e2fs_fsmnt[MAXMNTLEN - 1] = '\0';
	vfs_mountedfrom(mp, fspec);
	return (0);
}

static int
ext2_check_sb_compat(struct ext2fs *es, struct cdev *dev, int ronly)
{
	uint32_t i, mask;

	if (es->e2fs_magic != E2FS_MAGIC) {
		printf("ext2fs: %s: wrong magic number %#x (expected %#x)\n",
		    devtoname(dev), es->e2fs_magic, E2FS_MAGIC);
		return (1);
	}
	if (es->e2fs_rev > E2FS_REV0) {
		mask = es->e2fs_features_incompat & ~(EXT2F_INCOMPAT_SUPP |
		    EXT4F_RO_INCOMPAT_SUPP);
		if (mask) {
			printf("WARNING: mount of %s denied due to "
			    "unsupported optional features:\n", devtoname(dev));
			for (i = 0;
			    i < sizeof(incompat)/sizeof(struct ext2_feature);
			    i++)
				if (mask & incompat[i].mask)
					printf("%s ", incompat[i].name);
			printf("\n");
			return (1);
		}
		mask = es->e2fs_features_rocompat & ~EXT2F_ROCOMPAT_SUPP;
		if (!ronly && mask) {
			printf("WARNING: R/W mount of %s denied due to "
			    "unsupported optional features:\n", devtoname(dev));
			for (i = 0;
			    i < sizeof(ro_compat)/sizeof(struct ext2_feature);
			    i++)
				if (mask & ro_compat[i].mask)
					printf("%s ", ro_compat[i].name);
			printf("\n");
			return (1);
		}
	}
	return (0);
}

/*
 * This computes the fields of the m_ext2fs structure from the
 * data in the ext2fs structure read in.
 */
static int
compute_sb_data(struct vnode *devvp, struct ext2fs *es,
    struct m_ext2fs *fs)
{
	int db_count, error;
	int i;
	int logic_sb_block = 1;	/* XXX for now */
	struct buf *bp;
	uint32_t e2fs_descpb;

	fs->e2fs_bshift = EXT2_MIN_BLOCK_LOG_SIZE + es->e2fs_log_bsize;
	fs->e2fs_bsize = 1U << fs->e2fs_bshift;
	fs->e2fs_fsbtodb = es->e2fs_log_bsize + 1;
	fs->e2fs_qbmask = fs->e2fs_bsize - 1;
	fs->e2fs_fsize = EXT2_MIN_FRAG_SIZE << es->e2fs_log_fsize;
	if (fs->e2fs_fsize)
		fs->e2fs_fpb = fs->e2fs_bsize / fs->e2fs_fsize;
	fs->e2fs_bpg = es->e2fs_bpg;
	fs->e2fs_fpg = es->e2fs_fpg;
	fs->e2fs_ipg = es->e2fs_ipg;
	if (es->e2fs_rev == E2FS_REV0) {
		fs->e2fs_isize = E2FS_REV0_INODE_SIZE;
	} else {
		fs->e2fs_isize = es->e2fs_inode_size;

		/*
		 * Simple sanity check for superblock inode size value.
		 */
		if (EXT2_INODE_SIZE(fs) < E2FS_REV0_INODE_SIZE ||
		    EXT2_INODE_SIZE(fs) > fs->e2fs_bsize ||
		    (fs->e2fs_isize & (fs->e2fs_isize - 1)) != 0) {
			printf("ext2fs: invalid inode size %d\n",
			    fs->e2fs_isize);
			return (EIO);
		}
	}
	/* Check for extra isize in big inodes. */
	if (EXT2_HAS_RO_COMPAT_FEATURE(fs, EXT2F_ROCOMPAT_EXTRA_ISIZE) &&
	    EXT2_INODE_SIZE(fs) < sizeof(struct ext2fs_dinode)) {
		printf("ext2fs: no space for extra inode timestamps\n");
		return (EINVAL);
	}
	/* Check for group descriptor size */
	if (EXT2_HAS_INCOMPAT_FEATURE(fs, EXT2F_INCOMPAT_64BIT) &&
	    (es->e3fs_desc_size != sizeof(struct ext2_gd))) {
		printf("ext2fs: group descriptor size unsupported %d\n",
		    es->e3fs_desc_size);
		return (EINVAL);
	}

	fs->e2fs_ipb = fs->e2fs_bsize / EXT2_INODE_SIZE(fs);
	fs->e2fs_itpg = fs->e2fs_ipg / fs->e2fs_ipb;
	/* s_resuid / s_resgid ? */
	fs->e2fs_gcount = howmany(es->e2fs_bcount - es->e2fs_first_dblock,
	    EXT2_BLOCKS_PER_GROUP(fs));
	e2fs_descpb = fs->e2fs_bsize / sizeof(struct ext2_gd);
	db_count = howmany(fs->e2fs_gcount, e2fs_descpb);
	fs->e2fs_gdbcount = db_count;
	fs->e2fs_gd = malloc(db_count * fs->e2fs_bsize,
	    M_EXT2MNT, M_WAITOK);
	fs->e2fs_contigdirs = malloc(fs->e2fs_gcount *
	    sizeof(*fs->e2fs_contigdirs), M_EXT2MNT, M_WAITOK | M_ZERO);

	/*
	 * Adjust logic_sb_block.
	 * Godmar thinks: if the blocksize is greater than 1024, then
	 * the superblock is logically part of block zero.
	 */
	if (fs->e2fs_bsize > SBSIZE)
		logic_sb_block = 0;
	for (i = 0; i < db_count; i++) {
		error = bread(devvp,
		    fsbtodb(fs, logic_sb_block + i + 1),
		    fs->e2fs_bsize, NOCRED, &bp);
		if (error) {
			free(fs->e2fs_contigdirs, M_EXT2MNT);
			free(fs->e2fs_gd, M_EXT2MNT);
			brelse(bp);
			return (error);
		}
		e2fs_cgload((struct ext2_gd *)bp->b_data,
		    &fs->e2fs_gd[
		    i * fs->e2fs_bsize / sizeof(struct ext2_gd)],
		    fs->e2fs_bsize);
		brelse(bp);
		bp = NULL;
	}
	/* Verify cg csum */
	if (EXT2_HAS_RO_COMPAT_FEATURE(fs, EXT2F_ROCOMPAT_GDT_CSUM)) {
		error = ext2_gd_csum_verify(fs, devvp->v_rdev);
		if (error)
			return (error);
	}
	/* Initialization for the ext2 Orlov allocator variant. */
	fs->e2fs_total_dir = 0;
	for (i = 0; i < fs->e2fs_gcount; i++)
		fs->e2fs_total_dir += fs->e2fs_gd[i].ext2bgd_ndirs;

	if (es->e2fs_rev == E2FS_REV0 ||
	    !EXT2_HAS_RO_COMPAT_FEATURE(fs, EXT2F_ROCOMPAT_LARGEFILE))
		fs->e2fs_maxfilesize = 0x7fffffff;
	else {
		fs->e2fs_maxfilesize = 0xffffffffffff;
		if (EXT2_HAS_RO_COMPAT_FEATURE(fs, EXT2F_ROCOMPAT_HUGE_FILE))
			fs->e2fs_maxfilesize = 0x7fffffffffffffff;
	}
	if (es->e4fs_flags & E2FS_UNSIGNED_HASH) {
		fs->e2fs_uhash = 3;
	} else if ((es->e4fs_flags & E2FS_SIGNED_HASH) == 0) {
#ifdef __CHAR_UNSIGNED__
		es->e4fs_flags |= E2FS_UNSIGNED_HASH;
		fs->e2fs_uhash = 3;
#else
		es->e4fs_flags |= E2FS_SIGNED_HASH;
#endif
	}

	return (0);
}

/*
 * Reload all incore data for a filesystem (used after running fsck on
 * the root filesystem and finding things to fix). The filesystem must
 * be mounted read-only.
 *
 * Things to do to update the mount:
 *	1) invalidate all cached meta-data.
 *	2) re-read superblock from disk.
 *	3) invalidate all cluster summary information.
 *	4) invalidate all inactive vnodes.
 *	5) invalidate all cached file data.
 *	6) re-read inode data for all active vnodes.
 * XXX we are missing some steps, in particular # 3, this has to be reviewed.
 */
static int
ext2_reload(struct mount *mp, struct thread *td)
{
	struct vnode *vp, *mvp, *devvp;
	struct inode *ip;
	struct buf *bp;
	struct ext2fs *es;
	struct m_ext2fs *fs;
	struct csum *sump;
	int error, i;
	int32_t *lp;

	if ((mp->mnt_flag & MNT_RDONLY) == 0)
		return (EINVAL);
	/*
	 * Step 1: invalidate all cached meta-data.
	 */
	devvp = VFSTOEXT2(mp)->um_devvp;
	vn_lock(devvp, LK_EXCLUSIVE | LK_RETRY);
	if (vinvalbuf(devvp, 0, 0, 0) != 0)
		panic("ext2_reload: dirty1");
	VOP_UNLOCK(devvp, 0);

	/*
	 * Step 2: re-read superblock from disk.
	 * constants have been adjusted for ext2
	 */
	if ((error = bread(devvp, SBLOCK, SBSIZE, NOCRED, &bp)) != 0)
		return (error);
	es = (struct ext2fs *)bp->b_data;
	if (ext2_check_sb_compat(es, devvp->v_rdev, 0) != 0) {
		brelse(bp);
		return (EIO);		/* XXX needs translation */
	}
	fs = VFSTOEXT2(mp)->um_e2fs;
	bcopy(bp->b_data, fs->e2fs, sizeof(struct ext2fs));

	if ((error = compute_sb_data(devvp, es, fs)) != 0) {
		brelse(bp);
		return (error);
	}
#ifdef UNKLAR
	if (fs->fs_sbsize < SBSIZE)
		bp->b_flags |= B_INVAL;
#endif
	brelse(bp);

	/*
	 * Step 3: invalidate all cluster summary information.
	 */
	if (fs->e2fs_contigsumsize > 0) {
		lp = fs->e2fs_maxcluster;
		sump = fs->e2fs_clustersum;
		for (i = 0; i < fs->e2fs_gcount; i++, sump++) {
			*lp++ = fs->e2fs_contigsumsize;
			sump->cs_init = 0;
			bzero(sump->cs_sum, fs->e2fs_contigsumsize + 1);
		}
	}

loop:
	MNT_VNODE_FOREACH_ALL(vp, mp, mvp) {
		/*
		 * Step 4: invalidate all cached file data.
		 */
		if (vget(vp, LK_EXCLUSIVE | LK_INTERLOCK, td)) {
			MNT_VNODE_FOREACH_ALL_ABORT(mp, mvp);
			goto loop;
		}
		if (vinvalbuf(vp, 0, 0, 0))
			panic("ext2_reload: dirty2");

		/*
		 * Step 5: re-read inode data for all active vnodes.
		 */
		ip = VTOI(vp);
		error = bread(devvp, fsbtodb(fs, ino_to_fsba(fs, ip->i_number)),
		    (int)fs->e2fs_bsize, NOCRED, &bp);
		if (error) {
			VOP_UNLOCK(vp, 0);
			vrele(vp);
			MNT_VNODE_FOREACH_ALL_ABORT(mp, mvp);
			return (error);
		}
		ext2_ei2i((struct ext2fs_dinode *)((char *)bp->b_data +
		    EXT2_INODE_SIZE(fs) * ino_to_fsbo(fs, ip->i_number)), ip);
		brelse(bp);
		VOP_UNLOCK(vp, 0);
		vrele(vp);
	}
	return (0);
}

/*
 * Common code for mount and mountroot.
 */
static int
ext2_mountfs(struct vnode *devvp, struct mount *mp)
{
	struct ext2mount *ump;
	struct buf *bp;
	struct m_ext2fs *fs;
	struct ext2fs *es;
	struct cdev *dev = devvp->v_rdev;
	struct g_consumer *cp;
	struct bufobj *bo;
	struct csum *sump;
	int error;
	int ronly;
	int i;
	u_long size;
	int32_t *lp;
	int32_t e2fs_maxcontig;

	ronly = vfs_flagopt(mp->mnt_optnew, "ro", NULL, 0);
	/* XXX: use VOP_ACESS to check FS perms */
	g_topology_lock();
	error = g_vfs_open(devvp, &cp, "ext2fs", ronly ? 0 : 1);
	g_topology_unlock();
	VOP_UNLOCK(devvp, 0);
	if (error)
		return (error);

	/* XXX: should we check for some sectorsize or 512 instead? */
	if (((SBSIZE % cp->provider->sectorsize) != 0) ||
	    (SBSIZE < cp->provider->sectorsize)) {
		g_topology_lock();
		g_vfs_close(cp);
		g_topology_unlock();
		return (EINVAL);
	}

	bo = &devvp->v_bufobj;
	bo->bo_private = cp;
	bo->bo_ops = g_vfs_bufops;
	if (devvp->v_rdev->si_iosize_max != 0)
		mp->mnt_iosize_max = devvp->v_rdev->si_iosize_max;
	if (mp->mnt_iosize_max > MAXPHYS)
		mp->mnt_iosize_max = MAXPHYS;

	bp = NULL;
	ump = NULL;
	if ((error = bread(devvp, SBLOCK, SBSIZE, NOCRED, &bp)) != 0)
		goto out;
	es = (struct ext2fs *)bp->b_data;
	if (ext2_check_sb_compat(es, dev, ronly) != 0) {
		error = EINVAL;		/* XXX needs translation */
		goto out;
	}
	if ((es->e2fs_state & E2FS_ISCLEAN) == 0 ||
	    (es->e2fs_state & E2FS_ERRORS)) {
		if (ronly || (mp->mnt_flag & MNT_FORCE)) {
			printf(
"WARNING: Filesystem was not properly dismounted\n");
		} else {
			printf(
"WARNING: R/W mount denied.  Filesystem is not clean - run fsck\n");
			error = EPERM;
			goto out;
		}
	}
	ump = malloc(sizeof(*ump), M_EXT2MNT, M_WAITOK | M_ZERO);

	/*
	 * I don't know whether this is the right strategy. Note that
	 * we dynamically allocate both an m_ext2fs and an ext2fs
	 * while Linux keeps the super block in a locked buffer.
	 */
	ump->um_e2fs = malloc(sizeof(struct m_ext2fs),
	    M_EXT2MNT, M_WAITOK | M_ZERO);
	ump->um_e2fs->e2fs = malloc(sizeof(struct ext2fs),
	    M_EXT2MNT, M_WAITOK);
	mtx_init(EXT2_MTX(ump), "EXT2FS", "EXT2FS Lock", MTX_DEF);
	bcopy(es, ump->um_e2fs->e2fs, (u_int)sizeof(struct ext2fs));
	if ((error = compute_sb_data(devvp, ump->um_e2fs->e2fs, ump->um_e2fs)))
		goto out;

	/*
	 * Calculate the maximum contiguous blocks and size of cluster summary
	 * array.  In FFS this is done by newfs; however, the superblock
	 * in ext2fs doesn't have these variables, so we can calculate
	 * them here.
	 */
	e2fs_maxcontig = MAX(1, MAXPHYS / ump->um_e2fs->e2fs_bsize);
	ump->um_e2fs->e2fs_contigsumsize = MIN(e2fs_maxcontig, EXT2_MAXCONTIG);
	if (ump->um_e2fs->e2fs_contigsumsize > 0) {
		size = ump->um_e2fs->e2fs_gcount * sizeof(int32_t);
		ump->um_e2fs->e2fs_maxcluster = malloc(size, M_EXT2MNT, M_WAITOK);
		size = ump->um_e2fs->e2fs_gcount * sizeof(struct csum);
		ump->um_e2fs->e2fs_clustersum = malloc(size, M_EXT2MNT, M_WAITOK);
		lp = ump->um_e2fs->e2fs_maxcluster;
		sump = ump->um_e2fs->e2fs_clustersum;
		for (i = 0; i < ump->um_e2fs->e2fs_gcount; i++, sump++) {
			*lp++ = ump->um_e2fs->e2fs_contigsumsize;
			sump->cs_init = 0;
			sump->cs_sum = malloc((ump->um_e2fs->e2fs_contigsumsize + 1) *
			    sizeof(int32_t), M_EXT2MNT, M_WAITOK | M_ZERO);
		}
	}

	brelse(bp);
	bp = NULL;
	fs = ump->um_e2fs;
	fs->e2fs_ronly = ronly;	/* ronly is set according to mnt_flags */

	/*
	 * If the fs is not mounted read-only, make sure the super block is
	 * always written back on a sync().
	 */
	fs->e2fs_wasvalid = fs->e2fs->e2fs_state & E2FS_ISCLEAN ? 1 : 0;
	if (ronly == 0) {
		fs->e2fs_fmod = 1;	/* mark it modified */
		fs->e2fs->e2fs_state &= ~E2FS_ISCLEAN;	/* set fs invalid */
	}
	mp->mnt_data = ump;
	mp->mnt_stat.f_fsid.val[0] = dev2udev(dev);
	mp->mnt_stat.f_fsid.val[1] = mp->mnt_vfc->vfc_typenum;
	mp->mnt_maxsymlinklen = EXT2_MAXSYMLINKLEN;
	MNT_ILOCK(mp);
	mp->mnt_flag |= MNT_LOCAL;
	MNT_IUNLOCK(mp);
	ump->um_mountp = mp;
	ump->um_dev = dev;
	ump->um_devvp = devvp;
	ump->um_bo = &devvp->v_bufobj;
	ump->um_cp = cp;

	/*
	 * Setting those two parameters allowed us to use
	 * ufs_bmap w/o changse!
	 */
	ump->um_nindir = EXT2_ADDR_PER_BLOCK(fs);
	ump->um_bptrtodb = fs->e2fs->e2fs_log_bsize + 1;
	ump->um_seqinc = EXT2_FRAGS_PER_BLOCK(fs);
	if (ronly == 0)
		ext2_sbupdate(ump, MNT_WAIT);
	/*
	 * Initialize filesystem stat information in mount struct.
	 */
	MNT_ILOCK(mp);
	mp->mnt_kern_flag |= MNTK_LOOKUP_SHARED | MNTK_EXTENDED_SHARED |
	    MNTK_USES_BCACHE;
	MNT_IUNLOCK(mp);
	return (0);
out:
	if (bp)
		brelse(bp);
	if (cp != NULL) {
		g_topology_lock();
		g_vfs_close(cp);
		g_topology_unlock();
	}
	if (ump) {
		mtx_destroy(EXT2_MTX(ump));
		free(ump->um_e2fs->e2fs_gd, M_EXT2MNT);
		free(ump->um_e2fs->e2fs_contigdirs, M_EXT2MNT);
		free(ump->um_e2fs->e2fs, M_EXT2MNT);
		free(ump->um_e2fs, M_EXT2MNT);
		free(ump, M_EXT2MNT);
		mp->mnt_data = NULL;
	}
	return (error);
}

/*
 * Unmount system call.
 */
static int
ext2_unmount(struct mount *mp, int mntflags)
{
	struct ext2mount *ump;
	struct m_ext2fs *fs;
	struct csum *sump;
	int error, flags, i, ronly;

	flags = 0;
	if (mntflags & MNT_FORCE) {
		if (mp->mnt_flag & MNT_ROOTFS)
			return (EINVAL);
		flags |= FORCECLOSE;
	}
	if ((error = ext2_flushfiles(mp, flags, curthread)) != 0)
		return (error);
	ump = VFSTOEXT2(mp);
	fs = ump->um_e2fs;
	ronly = fs->e2fs_ronly;
	if (ronly == 0 && ext2_cgupdate(ump, MNT_WAIT) == 0) {
		if (fs->e2fs_wasvalid)
			fs->e2fs->e2fs_state |= E2FS_ISCLEAN;
		ext2_sbupdate(ump, MNT_WAIT);
	}

	g_topology_lock();
	g_vfs_close(ump->um_cp);
	g_topology_unlock();
	vrele(ump->um_devvp);
	sump = fs->e2fs_clustersum;
	for (i = 0; i < fs->e2fs_gcount; i++, sump++)
		free(sump->cs_sum, M_EXT2MNT);
	free(fs->e2fs_clustersum, M_EXT2MNT);
	free(fs->e2fs_maxcluster, M_EXT2MNT);
	free(fs->e2fs_gd, M_EXT2MNT);
	free(fs->e2fs_contigdirs, M_EXT2MNT);
	free(fs->e2fs, M_EXT2MNT);
	free(fs, M_EXT2MNT);
	free(ump, M_EXT2MNT);
	mp->mnt_data = NULL;
	MNT_ILOCK(mp);
	mp->mnt_flag &= ~MNT_LOCAL;
	MNT_IUNLOCK(mp);
	return (error);
}

/*
 * Flush out all the files in a filesystem.
 */
static int
ext2_flushfiles(struct mount *mp, int flags, struct thread *td)
{
	int error;

	error = vflush(mp, 0, flags, td);
	return (error);
}

/*
 * Get filesystem statistics.
 */
int
ext2_statfs(struct mount *mp, struct statfs *sbp)
{
	struct ext2mount *ump;
	struct m_ext2fs *fs;
	uint32_t overhead, overhead_per_group, ngdb;
	int i, ngroups;

	ump = VFSTOEXT2(mp);
	fs = ump->um_e2fs;
	if (fs->e2fs->e2fs_magic != E2FS_MAGIC)
		panic("ext2_statfs");

	/*
	 * Compute the overhead (FS structures)
	 */
	overhead_per_group =
	    1 /* block bitmap */ +
	    1 /* inode bitmap */ +
	    fs->e2fs_itpg;
	overhead = fs->e2fs->e2fs_first_dblock +
	    fs->e2fs_gcount * overhead_per_group;
	if (fs->e2fs->e2fs_rev > E2FS_REV0 &&
	    fs->e2fs->e2fs_features_rocompat & EXT2F_ROCOMPAT_SPARSESUPER) {
		for (i = 0, ngroups = 0; i < fs->e2fs_gcount; i++) {
			if (ext2_cg_has_sb(fs, i))
				ngroups++;
		}
	} else {
		ngroups = fs->e2fs_gcount;
	}
	ngdb = fs->e2fs_gdbcount;
	if (fs->e2fs->e2fs_rev > E2FS_REV0 &&
	    fs->e2fs->e2fs_features_compat & EXT2F_COMPAT_RESIZE)
		ngdb += fs->e2fs->e2fs_reserved_ngdb;
	overhead += ngroups * (1 /* superblock */ + ngdb);

	sbp->f_bsize = EXT2_FRAG_SIZE(fs);
	sbp->f_iosize = EXT2_BLOCK_SIZE(fs);
	sbp->f_blocks = fs->e2fs->e2fs_bcount - overhead;
	sbp->f_bfree = fs->e2fs->e2fs_fbcount;
	sbp->f_bavail = sbp->f_bfree - fs->e2fs->e2fs_rbcount;
	sbp->f_files = fs->e2fs->e2fs_icount;
	sbp->f_ffree = fs->e2fs->e2fs_ficount;
	return (0);
}

/*
 * Go through the disk queues to initiate sandbagged IO;
 * go through the inodes to write those that have been modified;
 * initiate the writing of the super block if it has been modified.
 *
 * Note: we are always called with the filesystem marked `MPBUSY'.
 */
static int
ext2_sync(struct mount *mp, int waitfor)
{
	struct vnode *mvp, *vp;
	struct thread *td;
	struct inode *ip;
	struct ext2mount *ump = VFSTOEXT2(mp);
	struct m_ext2fs *fs;
	int error, allerror = 0;

	td = curthread;
	fs = ump->um_e2fs;
	if (fs->e2fs_fmod != 0 && fs->e2fs_ronly != 0) {		/* XXX */
		printf("fs = %s\n", fs->e2fs_fsmnt);
		panic("ext2_sync: rofs mod");
	}

	/*
	 * Write back each (modified) inode.
	 */
loop:
	MNT_VNODE_FOREACH_ALL(vp, mp, mvp) {
		if (vp->v_type == VNON) {
			VI_UNLOCK(vp);
			continue;
		}
		ip = VTOI(vp);
		if ((ip->i_flag &
		    (IN_ACCESS | IN_CHANGE | IN_MODIFIED | IN_UPDATE)) == 0 &&
		    (vp->v_bufobj.bo_dirty.bv_cnt == 0 ||
		    waitfor == MNT_LAZY)) {
			VI_UNLOCK(vp);
			continue;
		}
		error = vget(vp, LK_EXCLUSIVE | LK_NOWAIT | LK_INTERLOCK, td);
		if (error) {
			if (error == ENOENT) {
				MNT_VNODE_FOREACH_ALL_ABORT(mp, mvp);
				goto loop;
			}
			continue;
		}
		if ((error = VOP_FSYNC(vp, waitfor, td)) != 0)
			allerror = error;
		VOP_UNLOCK(vp, 0);
		vrele(vp);
	}

	/*
	 * Force stale filesystem control information to be flushed.
	 */
	if (waitfor != MNT_LAZY) {
		vn_lock(ump->um_devvp, LK_EXCLUSIVE | LK_RETRY);
		if ((error = VOP_FSYNC(ump->um_devvp, waitfor, td)) != 0)
			allerror = error;
		VOP_UNLOCK(ump->um_devvp, 0);
	}

	/*
	 * Write back modified superblock.
	 */
	if (fs->e2fs_fmod != 0) {
		fs->e2fs_fmod = 0;
		fs->e2fs->e2fs_wtime = time_second;
		if ((error = ext2_cgupdate(ump, waitfor)) != 0)
			allerror = error;
	}
	return (allerror);
}

/*
 * Look up an EXT2FS dinode number to find its incore vnode, otherwise read it
 * in from disk.  If it is in core, wait for the lock bit to clear, then
 * return the inode locked.  Detection and handling of mount points must be
 * done by the calling routine.
 */
static int
ext2_vget(struct mount *mp, ino_t ino, int flags, struct vnode **vpp)
{
	struct m_ext2fs *fs;
	struct inode *ip;
	struct ext2mount *ump;
	struct buf *bp;
	struct vnode *vp;
	struct thread *td;
	int i, error;
	int used_blocks;

	td = curthread;
	error = vfs_hash_get(mp, ino, flags, td, vpp, NULL, NULL);
	if (error || *vpp != NULL)
		return (error);

	ump = VFSTOEXT2(mp);
	ip = malloc(sizeof(struct inode), M_EXT2NODE, M_WAITOK | M_ZERO);

	/* Allocate a new vnode/inode. */
	if ((error = getnewvnode("ext2fs", mp, &ext2_vnodeops, &vp)) != 0) {
		*vpp = NULL;
		free(ip, M_EXT2NODE);
		return (error);
	}
	vp->v_data = ip;
	ip->i_vnode = vp;
	ip->i_e2fs = fs = ump->um_e2fs;
	ip->i_ump = ump;
	ip->i_number = ino;

	lockmgr(vp->v_vnlock, LK_EXCLUSIVE, NULL);
	error = insmntque(vp, mp);
	if (error != 0) {
		free(ip, M_EXT2NODE);
		*vpp = NULL;
		return (error);
	}
	error = vfs_hash_insert(vp, ino, flags, td, vpp, NULL, NULL);
	if (error || *vpp != NULL)
		return (error);

	/* Read in the disk contents for the inode, copy into the inode. */
	if ((error = bread(ump->um_devvp, fsbtodb(fs, ino_to_fsba(fs, ino)),
	    (int)fs->e2fs_bsize, NOCRED, &bp)) != 0) {
		/*
		 * The inode does not contain anything useful, so it would
		 * be misleading to leave it on its hash chain. With mode
		 * still zero, it will be unlinked and returned to the free
		 * list by vput().
		 */
		brelse(bp);
		vput(vp);
		*vpp = NULL;
		return (error);
	}
	/* convert ext2 inode to dinode */
	ext2_ei2i((struct ext2fs_dinode *)((char *)bp->b_data + EXT2_INODE_SIZE(fs) *
	    ino_to_fsbo(fs, ino)), ip);
	ip->i_block_group = ino_to_cg(fs, ino);
	ip->i_next_alloc_block = 0;
	ip->i_next_alloc_goal = 0;

	/*
	 * Now we want to make sure that block pointers for unused
	 * blocks are zeroed out - ext2_balloc depends on this
	 * although for regular files and directories only
	 *
	 * If IN_E4EXTENTS is enabled, unused blocks are not zeroed
	 * out because we could corrupt the extent tree.
	 */
	if (!(ip->i_flag & IN_E4EXTENTS) &&
	    (S_ISDIR(ip->i_mode) || S_ISREG(ip->i_mode))) {
		used_blocks = howmany(ip->i_size, fs->e2fs_bsize);
		for (i = used_blocks; i < EXT2_NDIR_BLOCKS; i++)
			ip->i_db[i] = 0;
	}
#ifdef EXT2FS_DEBUG
	ext2_print_inode(ip);
#endif
	bqrelse(bp);

	/*
	 * Initialize the vnode from the inode, check for aliases.
	 * Note that the underlying vnode may have changed.
	 */
	if ((error = ext2_vinit(mp, &ext2_fifoops, &vp)) != 0) {
		vput(vp);
		*vpp = NULL;
		return (error);
	}

	/*
	 * Finish inode initialization.
	 */

	*vpp = vp;
	return (0);
}

/*
 * File handle to vnode
 *
 * Have to be really careful about stale file handles:
 * - check that the inode number is valid
 * - call ext2_vget() to get the locked inode
 * - check for an unallocated inode (i_mode == 0)
 * - check that the given client host has export rights and return
 *   those rights via. exflagsp and credanonp
 */
static int
ext2_fhtovp(struct mount *mp, struct fid *fhp, int flags, struct vnode **vpp)
{
	struct inode *ip;
	struct ufid *ufhp;
	struct vnode *nvp;
	struct m_ext2fs *fs;
	int error;

	ufhp = (struct ufid *)fhp;
	fs = VFSTOEXT2(mp)->um_e2fs;
	if (ufhp->ufid_ino < EXT2_ROOTINO ||
	    ufhp->ufid_ino > fs->e2fs_gcount * fs->e2fs->e2fs_ipg)
		return (ESTALE);

	error = VFS_VGET(mp, ufhp->ufid_ino, LK_EXCLUSIVE, &nvp);
	if (error) {
		*vpp = NULLVP;
		return (error);
	}
	ip = VTOI(nvp);
	if (ip->i_mode == 0 ||
	    ip->i_gen != ufhp->ufid_gen || ip->i_nlink <= 0) {
		vput(nvp);
		*vpp = NULLVP;
		return (ESTALE);
	}
	*vpp = nvp;
	vnode_create_vobject(*vpp, 0, curthread);
	return (0);
}

/*
 * Write a superblock and associated information back to disk.
 */
static int
ext2_sbupdate(struct ext2mount *mp, int waitfor)
{
	struct m_ext2fs *fs = mp->um_e2fs;
	struct ext2fs *es = fs->e2fs;
	struct buf *bp;
	int error = 0;

	bp = getblk(mp->um_devvp, SBLOCK, SBSIZE, 0, 0, 0);
	bcopy((caddr_t)es, bp->b_data, (u_int)sizeof(struct ext2fs));
	if (waitfor == MNT_WAIT)
		error = bwrite(bp);
	else
		bawrite(bp);

	/*
	 * The buffers for group descriptors, inode bitmaps and block bitmaps
	 * are not busy at this point and are (hopefully) written by the
	 * usual sync mechanism. No need to write them here.
	 */
	return (error);
}
int
ext2_cgupdate(struct ext2mount *mp, int waitfor)
{
	struct m_ext2fs *fs = mp->um_e2fs;
	struct buf *bp;
	int i, error = 0, allerror = 0;

	allerror = ext2_sbupdate(mp, waitfor);

	/* Update gd csums */
	if (EXT2_HAS_RO_COMPAT_FEATURE(fs, EXT2F_ROCOMPAT_GDT_CSUM))
		ext2_gd_csum_set(fs);

	for (i = 0; i < fs->e2fs_gdbcount; i++) {
		bp = getblk(mp->um_devvp, fsbtodb(fs,
		    fs->e2fs->e2fs_first_dblock +
		    1 /* superblock */ + i), fs->e2fs_bsize, 0, 0, 0);
		e2fs_cgsave(&fs->e2fs_gd[
		    i * fs->e2fs_bsize / sizeof(struct ext2_gd)],
		    (struct ext2_gd *)bp->b_data, fs->e2fs_bsize);
		if (waitfor == MNT_WAIT)
			error = bwrite(bp);
		else
			bawrite(bp);
	}

	if (!allerror && error)
		allerror = error;
	return (allerror);
}

/*
 * Return the root of a filesystem.
 */
static int
ext2_root(struct mount *mp, int flags, struct vnode **vpp)
{
	struct vnode *nvp;
	int error;

	error = VFS_VGET(mp, EXT2_ROOTINO, LK_EXCLUSIVE, &nvp);
	if (error)
		return (error);
	*vpp = nvp;
	return (0);
}