summaryrefslogtreecommitdiffstats
path: root/sys/gnu/fs/ext2fs/ext2_lookup.c
blob: c11aa5af4d233fc11cad7669ddc4a3a6d03e848b (plain)
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
/*
 *  modified for Lites 1.1
 *
 *  Aug 1995, Godmar Back (gback@cs.utah.edu)
 *  University of Utah, Department of Computer Science
 */
/*
 * Copyright (c) 1989, 1993
 *	The Regents of the University of California.  All rights reserved.
 * (c) UNIX System Laboratories, Inc.
 * All or some portions of this file are derived from material licensed
 * to the University of California by American Telephone and Telegraph
 * Co. or Unix System Laboratories, Inc. and are reproduced herein with
 * the permission of UNIX System Laboratories, Inc.
 *
 * 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. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *	This product includes software developed by the University of
 *	California, Berkeley and its contributors.
 * 4. 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.
 *
 *	@(#)ufs_lookup.c	8.6 (Berkeley) 4/1/94
 * $FreeBSD$
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/namei.h>
#include <sys/bio.h>
#include <sys/buf.h>
#include <sys/mount.h>
#include <sys/vnode.h>
#include <sys/malloc.h>
#include <sys/dirent.h>
#include <sys/sysctl.h>

#include <ufs/ufs/dir.h>

#include <gnu/ext2fs/inode.h>
#include <gnu/ext2fs/ext2_mount.h>
#include <gnu/ext2fs/ext2_extern.h>
#include <gnu/ext2fs/ext2_fs.h>
#include <gnu/ext2fs/ext2_fs_sb.h>

#ifdef DIAGNOSTIC
static int dirchk = 1;
#else
static int dirchk = 0;
#endif

SYSCTL_NODE(_vfs, OID_AUTO, e2fs, CTLFLAG_RD, 0, "EXT2FS filesystem");
SYSCTL_INT(_vfs_e2fs, OID_AUTO, dircheck, CTLFLAG_RW, &dirchk, 0, "");

/* 
   DIRBLKSIZE in ffs is DEV_BSIZE (in most cases 512)
   while it is the native blocksize in ext2fs - thus, a #define
   is no longer appropriate
*/
#undef  DIRBLKSIZ

static u_char ext2_ft_to_dt[] = {
	DT_UNKNOWN,		/* EXT2_FT_UNKNOWN */
	DT_REG,			/* EXT2_FT_REG_FILE */
	DT_DIR,			/* EXT2_FT_DIR */
	DT_CHR,			/* EXT2_FT_CHRDEV */
	DT_BLK,			/* EXT2_FT_BLKDEV */
	DT_FIFO,		/* EXT2_FT_FIFO */
	DT_SOCK,		/* EXT2_FT_SOCK */
	DT_LNK,			/* EXT2_FT_SYMLINK */
};
#define	FTTODT(ft)						\
    ((ft) > sizeof(ext2_ft_to_dt) / sizeof(ext2_ft_to_dt[0]) ?	\
    DT_UNKNOWN : ext2_ft_to_dt[(ft)])

static u_char dt_to_ext2_ft[] = {
	EXT2_FT_UNKNOWN,	/* DT_UNKNOWN */
	EXT2_FT_FIFO,		/* DT_FIFO */
	EXT2_FT_CHRDEV,		/* DT_CHR */
	EXT2_FT_UNKNOWN,	/* unused */
	EXT2_FT_DIR,		/* DT_DIR */
	EXT2_FT_UNKNOWN,	/* unused */
	EXT2_FT_BLKDEV,		/* DT_BLK */
	EXT2_FT_UNKNOWN,	/* unused */
	EXT2_FT_REG_FILE,	/* DT_REG */
	EXT2_FT_UNKNOWN,	/* unused */
	EXT2_FT_SYMLINK,	/* DT_LNK */
	EXT2_FT_UNKNOWN,	/* unused */
	EXT2_FT_SOCK,		/* DT_SOCK */
	EXT2_FT_UNKNOWN,	/* unused */
	EXT2_FT_UNKNOWN,	/* DT_WHT */
};
#define	DTTOFT(dt)						\
    ((dt) > sizeof(dt_to_ext2_ft) / sizeof(dt_to_ext2_ft[0]) ?	\
    EXT2_FT_UNKNOWN : dt_to_ext2_ft[(dt)])

static int	ext2_dirbadentry(struct vnode *dp, struct ext2_dir_entry_2 *de,
		    int entryoffsetinblock);

/*
 * Vnode op for reading directories.
 *
 * The routine below assumes that the on-disk format of a directory
 * is the same as that defined by <sys/dirent.h>. If the on-disk
 * format changes, then it will be necessary to do a conversion
 * from the on-disk format that read returns to the format defined
 * by <sys/dirent.h>.
 */
/*
 * this is exactly what we do here - the problem is that the conversion
 * will blow up some entries by four bytes, so it can't be done in place.
 * This is too bad. Right now the conversion is done entry by entry, the
 * converted entry is sent via uiomove. 
 *
 * XXX allocate a buffer, convert as many entries as possible, then send
 * the whole buffer to uiomove
 */
int
ext2_readdir(ap)
        struct vop_readdir_args /* {
                struct vnode *a_vp;
                struct uio *a_uio;
                struct ucred *a_cred;
        } */ *ap;
{
        struct uio *uio = ap->a_uio;
        int count, error;

	struct ext2_dir_entry_2 *edp, *dp;
	int ncookies;
	struct dirent dstdp;
	struct uio auio;
	struct iovec aiov;
	caddr_t dirbuf;
	int DIRBLKSIZ = VTOI(ap->a_vp)->i_e2fs->s_blocksize;
	int readcnt;
	off_t startoffset = uio->uio_offset;

	count = uio->uio_resid;
	/*
	 * Avoid complications for partial directory entries by adjusting
	 * the i/o to end at a block boundary.  Don't give up (like ufs
	 * does) if the initial adjustment gives a negative count, since
	 * many callers don't supply a large enough buffer.  The correct
	 * size is a little larger than DIRBLKSIZ to allow for expansion
	 * of directory entries, but some callers just use 512.
	 */
	count -= (uio->uio_offset + count) & (DIRBLKSIZ -1);
	if (count <= 0)
		count += DIRBLKSIZ;

#ifdef EXT2FS_DEBUG
	printf("ext2_readdir: uio_offset = %lld, uio_resid = %d, count = %d\n", 
	    uio->uio_offset, uio->uio_resid, count);
#endif

	auio = *uio;
	auio.uio_iov = &aiov;
	auio.uio_iovcnt = 1;
	auio.uio_resid = count;
	auio.uio_segflg = UIO_SYSSPACE;
	aiov.iov_len = count;
	MALLOC(dirbuf, caddr_t, count, M_TEMP, 0);
	aiov.iov_base = dirbuf;
	error = VOP_READ(ap->a_vp, &auio, 0, ap->a_cred);
	if (error == 0) {
		readcnt = count - auio.uio_resid;
		edp = (struct ext2_dir_entry_2 *)&dirbuf[readcnt];
		ncookies = 0;
		bzero(&dstdp, offsetof(struct dirent, d_name));
		for (dp = (struct ext2_dir_entry_2 *)dirbuf; 
		    !error && uio->uio_resid > 0 && dp < edp; ) {
			/*-
			 * "New" ext2fs directory entries differ in 3 ways
			 * from ufs on-disk ones:
			 * - the name is not necessarily NUL-terminated.
			 * - the file type field always exists and always
			 * follows the name length field.
			 * - the file type is encoded in a different way.
			 *
			 * "Old" ext2fs directory entries need no special
			 * conversions, since they binary compatible with
			 * "new" entries having a file type of 0 (i.e.,
			 * EXT2_FT_UNKNOWN).  Splitting the old name length
			 * field didn't make a mess like it did in ufs,
			 * because ext2fs uses a machine-dependent disk
			 * layout.
			 */
			dstdp.d_fileno = dp->inode;
			dstdp.d_type = FTTODT(dp->file_type);
			dstdp.d_namlen = dp->name_len;
			dstdp.d_reclen = GENERIC_DIRSIZ(&dstdp);
			bcopy(dp->name, dstdp.d_name, dstdp.d_namlen);
			bzero(dstdp.d_name + dstdp.d_namlen,
			    dstdp.d_reclen - offsetof(struct dirent, d_name) -
			    dstdp.d_namlen);

			if (dp->rec_len > 0) {
				if(dstdp.d_reclen <= uio->uio_resid) {
					/* advance dp */
					dp = (struct ext2_dir_entry_2 *)
					    ((char *)dp + dp->rec_len); 
					error = 
					  uiomove((caddr_t)&dstdp,
						  dstdp.d_reclen, uio);
					if (!error)
						ncookies++;
				} else
					break;
			} else {
				error = EIO;
				break;
			}
		}
		/* we need to correct uio_offset */
		uio->uio_offset = startoffset + (caddr_t)dp - dirbuf;

		if (!error && ap->a_ncookies != NULL) {
			u_long *cookiep, *cookies, *ecookies;
			off_t off;

			if (uio->uio_segflg != UIO_SYSSPACE || uio->uio_iovcnt != 1)
				panic("ext2fs_readdir: unexpected uio from NFS server");
			MALLOC(cookies, u_long *, ncookies * sizeof(u_long), M_TEMP,
			       0);
			off = startoffset;
			for (dp = (struct ext2_dir_entry_2 *)dirbuf,
			     cookiep = cookies, ecookies = cookies + ncookies;
			     cookiep < ecookies;
			     dp = (struct ext2_dir_entry_2 *)((caddr_t) dp + dp->rec_len)) {
				off += dp->rec_len;
				*cookiep++ = (u_long) off;
			}
			*ap->a_ncookies = ncookies;
			*ap->a_cookies = cookies;
		}
	}
	FREE(dirbuf, M_TEMP);
	if (ap->a_eofflag)
		*ap->a_eofflag = VTOI(ap->a_vp)->i_size <= uio->uio_offset;
        return (error);
}

/*
 * Convert a component of a pathname into a pointer to a locked inode.
 * This is a very central and rather complicated routine.
 * If the file system is not maintained in a strict tree hierarchy,
 * this can result in a deadlock situation (see comments in code below).
 *
 * The cnp->cn_nameiop argument is LOOKUP, CREATE, RENAME, or DELETE depending
 * on whether the name is to be looked up, created, renamed, or deleted.
 * When CREATE, RENAME, or DELETE is specified, information usable in
 * creating, renaming, or deleting a directory entry may be calculated.
 * If flag has LOCKPARENT or'ed into it and the target of the pathname
 * exists, lookup returns both the target and its parent directory locked.
 * When creating or renaming and LOCKPARENT is specified, the target may
 * not be ".".  When deleting and LOCKPARENT is specified, the target may
 * be "."., but the caller must check to ensure it does an vrele and vput
 * instead of two vputs.
 *
 * Overall outline of ufs_lookup:
 *
 *	search for name in directory, to found or notfound
 * notfound:
 *	if creating, return locked directory, leaving info on available slots
 *	else return error
 * found:
 *	if at end of path and deleting, return information to allow delete
 *	if at end of path and rewriting (RENAME and LOCKPARENT), lock target
 *	  inode and return info to allow rewrite
 *	if not at end, add name to cache; if at end and neither creating
 *	  nor deleting, add name to cache
 */
int
ext2_lookup(ap)
	struct vop_cachedlookup_args /* {
		struct vnode *a_dvp;
		struct vnode **a_vpp;
		struct componentname *a_cnp;
	} */ *ap;
{
	struct vnode *vdp;		/* vnode for directory being searched */
	struct inode *dp;		/* inode for directory being searched */
	struct buf *bp;			/* a buffer of directory entries */
	struct ext2_dir_entry_2 *ep;	/* the current directory entry */
	int entryoffsetinblock;		/* offset of ep in bp's buffer */
	enum {NONE, COMPACT, FOUND} slotstatus;
	doff_t slotoffset;		/* offset of area with free space */
	int slotsize;			/* size of area at slotoffset */
	int slotfreespace;		/* amount of space free in slot */
	int slotneeded;			/* size of the entry we're seeking */
	int numdirpasses;		/* strategy for directory search */
	doff_t endsearch;		/* offset to end directory search */
	doff_t prevoff;			/* prev entry dp->i_offset */
	struct vnode *pdp;		/* saved dp during symlink work */
	struct vnode *tdp;		/* returned by VFS_VGET */
	doff_t enduseful;		/* pointer past last used dir slot */
	u_long bmask;			/* block offset mask */
	int lockparent;			/* 1 => lockparent flag is set */
	int wantparent;			/* 1 => wantparent or lockparent flag */
	int namlen, error;
	struct vnode **vpp = ap->a_vpp;
	struct componentname *cnp = ap->a_cnp;
	struct ucred *cred = cnp->cn_cred;
	int flags = cnp->cn_flags;
	int nameiop = cnp->cn_nameiop;
	struct thread *td = cnp->cn_thread;

	int	DIRBLKSIZ = VTOI(ap->a_dvp)->i_e2fs->s_blocksize;

	bp = NULL;
	slotoffset = -1;
	*vpp = NULL;
	vdp = ap->a_dvp;
	dp = VTOI(vdp);
	lockparent = flags & LOCKPARENT;
	wantparent = flags & (LOCKPARENT|WANTPARENT);

	/*
	 * We now have a segment name to search for, and a directory to search.
	 */

	/*
	 * Suppress search for slots unless creating
	 * file and at end of pathname, in which case
	 * we watch for a place to put the new file in
	 * case it doesn't already exist.
	 */
	slotstatus = FOUND;
	slotfreespace = slotsize = slotneeded = 0;
	if ((nameiop == CREATE || nameiop == RENAME) &&
	    (flags & ISLASTCN)) {
		slotstatus = NONE;
		slotneeded = EXT2_DIR_REC_LEN(cnp->cn_namelen); 
		/* was
		slotneeded = (sizeof(struct direct) - MAXNAMLEN +
			cnp->cn_namelen + 3) &~ 3; */
	}

	/*
	 * If there is cached information on a previous search of
	 * this directory, pick up where we last left off.
	 * We cache only lookups as these are the most common
	 * and have the greatest payoff. Caching CREATE has little
	 * benefit as it usually must search the entire directory
	 * to determine that the entry does not exist. Caching the
	 * location of the last DELETE or RENAME has not reduced
	 * profiling time and hence has been removed in the interest
	 * of simplicity.
	 */
	bmask = VFSTOEXT2(vdp->v_mount)->um_mountp->mnt_stat.f_iosize - 1;
	if (nameiop != LOOKUP || dp->i_diroff == 0 ||
	    dp->i_diroff > dp->i_size) {
		entryoffsetinblock = 0;
		dp->i_offset = 0;
		numdirpasses = 1;
	} else {
		dp->i_offset = dp->i_diroff;
		if ((entryoffsetinblock = dp->i_offset & bmask) &&
		    (error = ext2_blkatoff(vdp, (off_t)dp->i_offset, NULL,
		    &bp)))
			return (error);
		numdirpasses = 2;
		nchstats.ncs_2passes++;
	}
	prevoff = dp->i_offset;
	endsearch = roundup(dp->i_size, DIRBLKSIZ);
	enduseful = 0;

searchloop:
	while (dp->i_offset < endsearch) {
		/*
		 * If necessary, get the next directory block.
		 */
		if ((dp->i_offset & bmask) == 0) {
			if (bp != NULL)
				brelse(bp);
			if ((error =
			    ext2_blkatoff(vdp, (off_t)dp->i_offset, NULL,
			    &bp)) != 0)
				return (error);
			entryoffsetinblock = 0;
		}
		/*
		 * If still looking for a slot, and at a DIRBLKSIZE
		 * boundary, have to start looking for free space again.
		 */
		if (slotstatus == NONE &&
		    (entryoffsetinblock & (DIRBLKSIZ - 1)) == 0) {
			slotoffset = -1;
			slotfreespace = 0;
		}
		/*
		 * Get pointer to next entry.
		 * Full validation checks are slow, so we only check
		 * enough to insure forward progress through the
		 * directory. Complete checks can be run by setting
		 * "vfs.e2fs.dirchk" to be true.
		 */
		ep = (struct ext2_dir_entry_2 *)
			((char *)bp->b_data + entryoffsetinblock);
		if (ep->rec_len == 0 ||
		    (dirchk && ext2_dirbadentry(vdp, ep, entryoffsetinblock))) {
			int i;
			ext2_dirbad(dp, dp->i_offset, "mangled entry");
			i = DIRBLKSIZ - (entryoffsetinblock & (DIRBLKSIZ - 1));
			dp->i_offset += i;
			entryoffsetinblock += i;
			continue;
		}

		/*
		 * If an appropriate sized slot has not yet been found,
		 * check to see if one is available. Also accumulate space
		 * in the current block so that we can determine if
		 * compaction is viable.
		 */
		if (slotstatus != FOUND) {
			int size = ep->rec_len;

			if (ep->inode != 0)
				size -= EXT2_DIR_REC_LEN(ep->name_len);
			if (size > 0) {
				if (size >= slotneeded) {
					slotstatus = FOUND;
					slotoffset = dp->i_offset;
					slotsize = ep->rec_len;
				} else if (slotstatus == NONE) {
					slotfreespace += size;
					if (slotoffset == -1)
						slotoffset = dp->i_offset;
					if (slotfreespace >= slotneeded) {
						slotstatus = COMPACT;
						slotsize = dp->i_offset +
						      ep->rec_len - slotoffset;
					}
				}
			}
		}

		/*
		 * Check for a name match.
		 */
		if (ep->inode) {
			namlen = ep->name_len;
			if (namlen == cnp->cn_namelen &&
			    !bcmp(cnp->cn_nameptr, ep->name,
				(unsigned)namlen)) {
				/*
				 * Save directory entry's inode number and
				 * reclen in ndp->ni_ufs area, and release
				 * directory buffer.
				 */
				dp->i_ino = ep->inode;
				dp->i_reclen = ep->rec_len;
				goto found;
			}
		}
		prevoff = dp->i_offset;
		dp->i_offset += ep->rec_len;
		entryoffsetinblock += ep->rec_len;
		if (ep->inode)
			enduseful = dp->i_offset;
	}
/* notfound: */
	/*
	 * If we started in the middle of the directory and failed
	 * to find our target, we must check the beginning as well.
	 */
	if (numdirpasses == 2) {
		numdirpasses--;
		dp->i_offset = 0;
		endsearch = dp->i_diroff;
		goto searchloop;
	}
	if (bp != NULL)
		brelse(bp);
	/*
	 * If creating, and at end of pathname and current
	 * directory has not been removed, then can consider
	 * allowing file to be created.
	 */
	if ((nameiop == CREATE || nameiop == RENAME) &&
	    (flags & ISLASTCN) && dp->i_nlink != 0) {
		/*
		 * Access for write is interpreted as allowing
		 * creation of files in the directory.
		 */
		if ((error = VOP_ACCESS(vdp, VWRITE, cred, cnp->cn_thread)) != 0)
			return (error);
		/*
		 * Return an indication of where the new directory
		 * entry should be put.  If we didn't find a slot,
		 * then set dp->i_count to 0 indicating
		 * that the new slot belongs at the end of the
		 * directory. If we found a slot, then the new entry
		 * can be put in the range from dp->i_offset to
		 * dp->i_offset + dp->i_count.
		 */
		if (slotstatus == NONE) {
			dp->i_offset = roundup(dp->i_size, DIRBLKSIZ);
			dp->i_count = 0;
			enduseful = dp->i_offset;
		} else {
			dp->i_offset = slotoffset;
			dp->i_count = slotsize;
			if (enduseful < slotoffset + slotsize)
				enduseful = slotoffset + slotsize;
		}
		dp->i_endoff = roundup(enduseful, DIRBLKSIZ);
		dp->i_flag |= IN_CHANGE | IN_UPDATE;
		/*
		 * We return with the directory locked, so that
		 * the parameters we set up above will still be
		 * valid if we actually decide to do a direnter().
		 * We return ni_vp == NULL to indicate that the entry
		 * does not currently exist; we leave a pointer to
		 * the (locked) directory inode in ndp->ni_dvp.
		 * The pathname buffer is saved so that the name
		 * can be obtained later.
		 *
		 * NB - if the directory is unlocked, then this
		 * information cannot be used.
		 */
		cnp->cn_flags |= SAVENAME;
		if (!lockparent)
			VOP_UNLOCK(vdp, 0, td);
		return (EJUSTRETURN);
	}
	/*
	 * Insert name into cache (as non-existent) if appropriate.
	 */
	if ((cnp->cn_flags & MAKEENTRY) && nameiop != CREATE)
		cache_enter(vdp, *vpp, cnp);
	return (ENOENT);

found:
	if (numdirpasses == 2)
		nchstats.ncs_pass2++;
	/*
	 * Check that directory length properly reflects presence
	 * of this entry.
	 */
	if (entryoffsetinblock + EXT2_DIR_REC_LEN(ep->name_len)
		> dp->i_size) {
		ext2_dirbad(dp, dp->i_offset, "i_size too small");
		dp->i_size = entryoffsetinblock+EXT2_DIR_REC_LEN(ep->name_len);
		dp->i_flag |= IN_CHANGE | IN_UPDATE;
	}
	brelse(bp);

	/*
	 * Found component in pathname.
	 * If the final component of path name, save information
	 * in the cache as to where the entry was found.
	 */
	if ((flags & ISLASTCN) && nameiop == LOOKUP)
		dp->i_diroff = dp->i_offset &~ (DIRBLKSIZ - 1);

	/*
	 * If deleting, and at end of pathname, return
	 * parameters which can be used to remove file.
	 * If the wantparent flag isn't set, we return only
	 * the directory (in ndp->ni_dvp), otherwise we go
	 * on and lock the inode, being careful with ".".
	 */
	if (nameiop == DELETE && (flags & ISLASTCN)) {
		/*
		 * Write access to directory required to delete files.
		 */
		if ((error = VOP_ACCESS(vdp, VWRITE, cred, cnp->cn_thread)) != 0)
			return (error);
		/*
		 * Return pointer to current entry in dp->i_offset,
		 * and distance past previous entry (if there
		 * is a previous entry in this block) in dp->i_count.
		 * Save directory inode pointer in ndp->ni_dvp for dirremove().
		 */
		if ((dp->i_offset & (DIRBLKSIZ - 1)) == 0)
			dp->i_count = 0;
		else
			dp->i_count = dp->i_offset - prevoff;
		if (dp->i_number == dp->i_ino) {
			VREF(vdp);
			*vpp = vdp;
			return (0);
		}
		if ((error = VFS_VGET(vdp->v_mount, dp->i_ino, LK_EXCLUSIVE,
		    &tdp)) != 0)
			return (error);
		/*
		 * If directory is "sticky", then user must own
		 * the directory, or the file in it, else she
		 * may not delete it (unless she's root). This
		 * implements append-only directories.
		 */
		if ((dp->i_mode & ISVTX) &&
		    cred->cr_uid != 0 &&
		    cred->cr_uid != dp->i_uid &&
		    VTOI(tdp)->i_uid != cred->cr_uid) {
			vput(tdp);
			return (EPERM);
		}
		*vpp = tdp;
		if (!lockparent)
			VOP_UNLOCK(vdp, 0, td);
		return (0);
	}

	/*
	 * If rewriting (RENAME), return the inode and the
	 * information required to rewrite the present directory
	 * Must get inode of directory entry to verify it's a
	 * regular file, or empty directory.
	 */
	if (nameiop == RENAME && wantparent &&
	    (flags & ISLASTCN)) {
		if ((error = VOP_ACCESS(vdp, VWRITE, cred, cnp->cn_thread)) != 0)
			return (error);
		/*
		 * Careful about locking second inode.
		 * This can only occur if the target is ".".
		 */
		if (dp->i_number == dp->i_ino)
			return (EISDIR);
		if ((error = VFS_VGET(vdp->v_mount, dp->i_ino, LK_EXCLUSIVE,
		    &tdp)) != 0)
			return (error);
		*vpp = tdp;
		cnp->cn_flags |= SAVENAME;
		if (!lockparent)
			VOP_UNLOCK(vdp, 0, td);
		return (0);
	}

	/*
	 * Step through the translation in the name.  We do not `vput' the
	 * directory because we may need it again if a symbolic link
	 * is relative to the current directory.  Instead we save it
	 * unlocked as "pdp".  We must get the target inode before unlocking
	 * the directory to insure that the inode will not be removed
	 * before we get it.  We prevent deadlock by always fetching
	 * inodes from the root, moving down the directory tree. Thus
	 * when following backward pointers ".." we must unlock the
	 * parent directory before getting the requested directory.
	 * There is a potential race condition here if both the current
	 * and parent directories are removed before the VFS_VGET for the
	 * inode associated with ".." returns.  We hope that this occurs
	 * infrequently since we cannot avoid this race condition without
	 * implementing a sophisticated deadlock detection algorithm.
	 * Note also that this simple deadlock detection scheme will not
	 * work if the file system has any hard links other than ".."
	 * that point backwards in the directory structure.
	 */
	pdp = vdp;
	if (flags & ISDOTDOT) {
		VOP_UNLOCK(pdp, 0, td);	/* race to get the inode */
		if ((error = VFS_VGET(vdp->v_mount, dp->i_ino, LK_EXCLUSIVE,
		    &tdp)) != 0) {
			vn_lock(pdp, LK_EXCLUSIVE | LK_RETRY, td);
			return (error);
		}
		if (lockparent && (flags & ISLASTCN) &&
		    (error = vn_lock(pdp, LK_EXCLUSIVE, td))) {
			vput(tdp);
			return (error);
		}
		*vpp = tdp;
	} else if (dp->i_number == dp->i_ino) {
		VREF(vdp);	/* we want ourself, ie "." */
		*vpp = vdp;
	} else {
		if ((error = VFS_VGET(vdp->v_mount, dp->i_ino, LK_EXCLUSIVE,
		    &tdp)) != 0)
			return (error);
		if (!lockparent || !(flags & ISLASTCN))
			VOP_UNLOCK(pdp, 0, td);
		*vpp = tdp;
	}

	/*
	 * Insert name into cache if appropriate.
	 */
	if (cnp->cn_flags & MAKEENTRY)
		cache_enter(vdp, *vpp, cnp);
	return (0);
}

void
ext2_dirbad(ip, offset, how)
	struct inode *ip;
	doff_t offset;
	char *how;
{
	struct mount *mp;

	mp = ITOV(ip)->v_mount;
	(void)printf("%s: bad dir ino %lu at offset %ld: %s\n",
	    mp->mnt_stat.f_mntonname, (u_long)ip->i_number, (long)offset, how);
	if ((mp->mnt_flag & MNT_RDONLY) == 0)
		panic("ext2_dirbad: bad dir");
}

/*
 * Do consistency checking on a directory entry:
 *	record length must be multiple of 4
 *	entry must fit in rest of its DIRBLKSIZ block
 *	record must be large enough to contain entry
 *	name is not longer than MAXNAMLEN
 *	name must be as long as advertised, and null terminated
 */
/*
 *	changed so that it confirms to ext2_check_dir_entry
 */
static int
ext2_dirbadentry(dp, de, entryoffsetinblock)
	struct vnode *dp;
	struct ext2_dir_entry_2 *de;
	int entryoffsetinblock;
{
	int	DIRBLKSIZ = VTOI(dp)->i_e2fs->s_blocksize;

        char * error_msg = NULL;

        if (de->rec_len < EXT2_DIR_REC_LEN(1))
                error_msg = "rec_len is smaller than minimal";
        else if (de->rec_len % 4 != 0)
                error_msg = "rec_len % 4 != 0";
        else if (de->rec_len < EXT2_DIR_REC_LEN(de->name_len))
                error_msg = "reclen is too small for name_len";
        else if (entryoffsetinblock + de->rec_len > DIRBLKSIZ)
                error_msg = "directory entry across blocks";
        /* else LATER
	     if (de->inode > dir->i_sb->u.ext2_sb.s_es->s_inodes_count)
                error_msg = "inode out of bounds";
	*/

        if (error_msg != NULL) {
                printf("bad directory entry: %s\n", error_msg);
                printf("offset=%d, inode=%lu, rec_len=%u, name_len=%u\n",
			entryoffsetinblock, (unsigned long)de->inode,
			de->rec_len, de->name_len);
        }
        return error_msg == NULL ? 0 : 1;
}

/*
 * Write a directory entry after a call to namei, using the parameters
 * that it left in nameidata.  The argument ip is the inode which the new
 * directory entry will refer to.  Dvp is a pointer to the directory to
 * be written, which was left locked by namei. Remaining parameters
 * (dp->i_offset, dp->i_count) indicate how the space for the new
 * entry is to be obtained.
 */
int
ext2_direnter(ip, dvp, cnp)
	struct inode *ip;
	struct vnode *dvp;
	struct componentname *cnp;
{
	struct ext2_dir_entry_2 *ep, *nep;
	struct inode *dp;
	struct buf *bp;
	struct ext2_dir_entry_2 newdir;
	struct iovec aiov;
	struct uio auio;
	u_int dsize;
	int error, loc, newentrysize, spacefree;
	char *dirbuf;
	int     DIRBLKSIZ = ip->i_e2fs->s_blocksize;


#if DIAGNOSTIC
	if ((cnp->cn_flags & SAVENAME) == 0)
		panic("direnter: missing name");
#endif
	dp = VTOI(dvp);
	newdir.inode = ip->i_number;
	newdir.name_len = cnp->cn_namelen;
	if (EXT2_HAS_INCOMPAT_FEATURE(ip->i_e2fs->s_es,
	    EXT2_FEATURE_INCOMPAT_FILETYPE))
		newdir.file_type = DTTOFT(IFTODT(ip->i_mode));
	else
		newdir.file_type = EXT2_FT_UNKNOWN;
	bcopy(cnp->cn_nameptr, newdir.name, (unsigned)cnp->cn_namelen + 1);
	newentrysize = EXT2_DIR_REC_LEN(newdir.name_len);
	if (dp->i_count == 0) {
		/*
		 * If dp->i_count is 0, then namei could find no
		 * space in the directory. Here, dp->i_offset will
		 * be on a directory block boundary and we will write the
		 * new entry into a fresh block.
		 */
		if (dp->i_offset & (DIRBLKSIZ - 1))
			panic("ext2_direnter: newblk");
		auio.uio_offset = dp->i_offset;
		newdir.rec_len = DIRBLKSIZ;
		auio.uio_resid = newentrysize;
		aiov.iov_len = newentrysize;
		aiov.iov_base = (caddr_t)&newdir;
		auio.uio_iov = &aiov;
		auio.uio_iovcnt = 1;
		auio.uio_rw = UIO_WRITE;
		auio.uio_segflg = UIO_SYSSPACE;
		auio.uio_td = (struct thread *)0;
		error = VOP_WRITE(dvp, &auio, IO_SYNC, cnp->cn_cred);
		if (DIRBLKSIZ >
		    VFSTOEXT2(dvp->v_mount)->um_mountp->mnt_stat.f_bsize)
			/* XXX should grow with balloc() */
			panic("ext2_direnter: frag size");
		else if (!error) {
			dp->i_size = roundup(dp->i_size, DIRBLKSIZ);
			dp->i_flag |= IN_CHANGE;
		}
		return (error);
	}

	/*
	 * If dp->i_count is non-zero, then namei found space
	 * for the new entry in the range dp->i_offset to
	 * dp->i_offset + dp->i_count in the directory.
	 * To use this space, we may have to compact the entries located
	 * there, by copying them together towards the beginning of the
	 * block, leaving the free space in one usable chunk at the end.
	 */

	/*
	 * Increase size of directory if entry eats into new space.
	 * This should never push the size past a new multiple of
	 * DIRBLKSIZE.
	 *
	 * N.B. - THIS IS AN ARTIFACT OF 4.2 AND SHOULD NEVER HAPPEN.
	 */
	if (dp->i_offset + dp->i_count > dp->i_size)
		dp->i_size = dp->i_offset + dp->i_count;
	/*
	 * Get the block containing the space for the new directory entry.
	 */
	if ((error = ext2_blkatoff(dvp, (off_t)dp->i_offset, &dirbuf,
	    &bp)) != 0)
		return (error);
	/*
	 * Find space for the new entry. In the simple case, the entry at
	 * offset base will have the space. If it does not, then namei
	 * arranged that compacting the region dp->i_offset to
	 * dp->i_offset + dp->i_count would yield the
	 * space.
	 */
	ep = (struct ext2_dir_entry_2 *)dirbuf;
	dsize = EXT2_DIR_REC_LEN(ep->name_len);
	spacefree = ep->rec_len - dsize;
	for (loc = ep->rec_len; loc < dp->i_count; ) {
		nep = (struct ext2_dir_entry_2 *)(dirbuf + loc);
		if (ep->inode) {
			/* trim the existing slot */
			ep->rec_len = dsize;
			ep = (struct ext2_dir_entry_2 *)((char *)ep + dsize);
		} else {
			/* overwrite; nothing there; header is ours */
			spacefree += dsize;
		}
		dsize = EXT2_DIR_REC_LEN(nep->name_len);
		spacefree += nep->rec_len - dsize;
		loc += nep->rec_len;
		bcopy((caddr_t)nep, (caddr_t)ep, dsize);
	}
	/*
	 * Update the pointer fields in the previous entry (if any),
	 * copy in the new entry, and write out the block.
	 */
	if (ep->inode == 0) {
		if (spacefree + dsize < newentrysize)
			panic("ext2_direnter: compact1");
		newdir.rec_len = spacefree + dsize;
	} else {
		if (spacefree < newentrysize)
			panic("ext2_direnter: compact2");
		newdir.rec_len = spacefree;
		ep->rec_len = dsize;
		ep = (struct ext2_dir_entry_2 *)((char *)ep + dsize);
	}
	bcopy((caddr_t)&newdir, (caddr_t)ep, (u_int)newentrysize);
	error = BUF_WRITE(bp);
	dp->i_flag |= IN_CHANGE | IN_UPDATE;
	if (!error && dp->i_endoff && dp->i_endoff < dp->i_size)
		error = ext2_truncate(dvp, (off_t)dp->i_endoff, IO_SYNC,
		    cnp->cn_cred, cnp->cn_thread);
	return (error);
}

/*
 * Remove a directory entry after a call to namei, using
 * the parameters which it left in nameidata. The entry
 * dp->i_offset contains the offset into the directory of the
 * entry to be eliminated.  The dp->i_count field contains the
 * size of the previous record in the directory.  If this
 * is 0, the first entry is being deleted, so we need only
 * zero the inode number to mark the entry as free.  If the
 * entry is not the first in the directory, we must reclaim
 * the space of the now empty record by adding the record size
 * to the size of the previous entry.
 */
int
ext2_dirremove(dvp, cnp)
	struct vnode *dvp;
	struct componentname *cnp;
{
	struct inode *dp;
	struct ext2_dir_entry_2 *ep;
	struct buf *bp;
	int error;
	 
	dp = VTOI(dvp);
	if (dp->i_count == 0) {
		/*
		 * First entry in block: set d_ino to zero.
		 */
		if ((error =
		    ext2_blkatoff(dvp, (off_t)dp->i_offset, (char **)&ep,
		    &bp)) != 0)
			return (error);
		ep->inode = 0;
		error = BUF_WRITE(bp);
		dp->i_flag |= IN_CHANGE | IN_UPDATE;
		return (error);
	}
	/*
	 * Collapse new free space into previous entry.
	 */
	if ((error = ext2_blkatoff(dvp, (off_t)(dp->i_offset - dp->i_count),
	    (char **)&ep, &bp)) != 0)
		return (error);
	ep->rec_len += dp->i_reclen;
	error = BUF_WRITE(bp);
	dp->i_flag |= IN_CHANGE | IN_UPDATE;
	return (error);
}

/*
 * Rewrite an existing directory entry to point at the inode
 * supplied.  The parameters describing the directory entry are
 * set up by a call to namei.
 */
int
ext2_dirrewrite(dp, ip, cnp)
	struct inode *dp, *ip;
	struct componentname *cnp;
{
	struct buf *bp;
	struct ext2_dir_entry_2 *ep;
	struct vnode *vdp = ITOV(dp);
	int error;

	if ((error = ext2_blkatoff(vdp, (off_t)dp->i_offset, (char **)&ep,
	    &bp)) != 0)
		return (error);
	ep->inode = ip->i_number;
	if (EXT2_HAS_INCOMPAT_FEATURE(ip->i_e2fs->s_es,
	    EXT2_FEATURE_INCOMPAT_FILETYPE))
		ep->file_type = DTTOFT(IFTODT(ip->i_mode));
	else
		ep->file_type = EXT2_FT_UNKNOWN;
	error = BUF_WRITE(bp);
	dp->i_flag |= IN_CHANGE | IN_UPDATE;
	return (error);
}

/*
 * Check if a directory is empty or not.
 * Inode supplied must be locked.
 *
 * Using a struct dirtemplate here is not precisely
 * what we want, but better than using a struct direct.
 *
 * NB: does not handle corrupted directories.
 */
int
ext2_dirempty(ip, parentino, cred)
	struct inode *ip;
	ino_t parentino;
	struct ucred *cred;
{
	off_t off;
	struct dirtemplate dbuf;
	struct ext2_dir_entry_2 *dp = (struct ext2_dir_entry_2 *)&dbuf;
	int error, count, namlen;
		 
#define	MINDIRSIZ (sizeof (struct dirtemplate) / 2)

	for (off = 0; off < ip->i_size; off += dp->rec_len) {
		error = vn_rdwr(UIO_READ, ITOV(ip), (caddr_t)dp, MINDIRSIZ,
		    off, UIO_SYSSPACE, IO_NODELOCKED | IO_NOMACCHECK, cred,
		    NOCRED, &count, (struct thread *)0);
		/*
		 * Since we read MINDIRSIZ, residual must
		 * be 0 unless we're at end of file.
		 */
		if (error || count != 0)
			return (0);
		/* avoid infinite loops */
		if (dp->rec_len == 0)
			return (0);
		/* skip empty entries */
		if (dp->inode == 0)
			continue;
		/* accept only "." and ".." */
		namlen = dp->name_len;
		if (namlen > 2)
			return (0);
		if (dp->name[0] != '.')
			return (0);
		/*
		 * At this point namlen must be 1 or 2.
		 * 1 implies ".", 2 implies ".." if second
		 * char is also "."
		 */
		if (namlen == 1)
			continue;
		if (dp->name[1] == '.' && dp->inode == parentino)
			continue;
		return (0);
	}
	return (1);
}

/*
 * Check if source directory is in the path of the target directory.
 * Target is supplied locked, source is unlocked.
 * The target is always vput before returning.
 */
int
ext2_checkpath(source, target, cred)
	struct inode *source, *target;
	struct ucred *cred;
{
	struct vnode *vp;
	int error, rootino, namlen;
	struct dirtemplate dirbuf;

	vp = ITOV(target);
	if (target->i_number == source->i_number) {
		error = EEXIST;
		goto out;
	}
	rootino = ROOTINO;
	error = 0;
	if (target->i_number == rootino)
		goto out;

	for (;;) {
		if (vp->v_type != VDIR) {
			error = ENOTDIR;
			break;
		}
		error = vn_rdwr(UIO_READ, vp, (caddr_t)&dirbuf,
			sizeof (struct dirtemplate), (off_t)0, UIO_SYSSPACE,
			IO_NODELOCKED | IO_NOMACCHECK, cred, NOCRED, (int *)0,
			(struct thread *)0);
		if (error != 0)
			break;
		namlen = dirbuf.dotdot_type;	/* like ufs little-endian */
		if (namlen != 2 ||
		    dirbuf.dotdot_name[0] != '.' ||
		    dirbuf.dotdot_name[1] != '.') {
			error = ENOTDIR;
			break;
		}
		if (dirbuf.dotdot_ino == source->i_number) {
			error = EINVAL;
			break;
		}
		if (dirbuf.dotdot_ino == rootino)
			break;
		vput(vp);
		if ((error = VFS_VGET(vp->v_mount, dirbuf.dotdot_ino,
		    LK_EXCLUSIVE, &vp)) != 0) {
			vp = NULL;
			break;
		}
	}

out:
	if (error == ENOTDIR)
		printf("checkpath: .. not a directory\n");
	if (vp != NULL)
		vput(vp);
	return (error);
}

OpenPOWER on IntegriCloud