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
|
/* $OpenBSD: vnd.c,v 1.68 2007/02/01 01:02:05 pedro Exp $ */
/* $NetBSD: vnd.c,v 1.26 1996/03/30 23:06:11 christos Exp $ */
/*
* Copyright (c) 1988 University of Utah.
* Copyright (c) 1990, 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* the Systems Programming Group of the University of Utah Computer
* Science Department.
*
* 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.
*
* from: Utah $Hdr: vn.c 1.13 94/04/02$
*
* @(#)vn.c 8.6 (Berkeley) 4/1/94
*/
/*
* Vnode disk driver.
*
* Block/character interface to a vnode. Allows one to treat a file
* as a disk (e.g. build a filesystem in it, mount it, etc.).
*
* NOTE 1: This uses either the VOP_BMAP/VOP_STRATEGY interface to the
* vnode or simple VOP_READ/VOP_WRITE. The former is suitable for swapping
* as it doesn't distort the local buffer cache. The latter is good for
* building disk images as it keeps the cache consistent after the block
* device is closed.
*
* NOTE 2: There is a security issue involved with this driver.
* Once mounted all access to the contents of the "mapped" file via
* the special file is controlled by the permissions on the special
* file, the protection of the mapped file is ignored (effectively,
* by using root credentials in all transactions).
*
* NOTE 3: Doesn't interact with leases, should it?
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/namei.h>
#include <sys/proc.h>
#include <sys/errno.h>
#include <sys/buf.h>
#include <sys/malloc.h>
#include <sys/pool.h>
#include <sys/ioctl.h>
#include <sys/disklabel.h>
#include <sys/device.h>
#include <sys/disk.h>
#include <sys/stat.h>
#include <sys/mount.h>
#include <sys/vnode.h>
#include <sys/file.h>
#include <sys/uio.h>
#include <sys/conf.h>
#include <crypto/blf.h>
#include <miscfs/specfs/specdev.h>
#include <dev/vndioctl.h>
#ifdef VNDDEBUG
int dovndcluster = 1;
int vnddebug = 0x00;
#define VDB_FOLLOW 0x01
#define VDB_INIT 0x02
#define VDB_IO 0x04
#define DNPRINTF(f, p...) do { if ((f) & vnddebug) printf(p); } while (0)
#else
#define DNPRINTF(f, p...) /* nothing */
#endif /* VNDDEBUG */
/*
* vndunit is a bit weird. have to reconstitute the dev_t for
* DISKUNIT(), but with the minor masked off.
*/
#define vndunit(x) DISKUNIT(makedev(major(x), minor(x) & 0x7ff))
#define vndsimple(x) (minor(x) & 0x800)
/* same as MAKEDISKDEV, preserving the vndsimple() property */
#define VNDLABELDEV(dev) \
makedev(major(dev), DISKMINOR(vndunit(dev), RAW_PART) | \
(vndsimple(dev) ? 0x800 : 0))
struct vndbuf {
struct buf vb_buf;
struct buf *vb_obp;
};
/*
* struct vndbuf allocator
*/
struct pool vndbufpl;
#define getvndbuf() pool_get(&vndbufpl, PR_WAITOK)
#define putvndbuf(vbp) pool_put(&vndbufpl, vbp);
struct vnd_softc {
struct device sc_dev;
struct disk sc_dk;
char sc_file[VNDNLEN]; /* file we're covering */
int sc_flags; /* flags */
size_t sc_size; /* size of vnd in blocks */
struct vnode *sc_vp; /* vnode */
struct ucred *sc_cred; /* credentials */
struct buf sc_tab; /* transfer queue */
blf_ctx *sc_keyctx; /* key context */
};
/* sc_flags */
#define VNF_ALIVE 0x0001
#define VNF_INITED 0x0002
#define VNF_WANTED 0x0040
#define VNF_LOCKED 0x0080
#define VNF_LABELLING 0x0100
#define VNF_WLABEL 0x0200
#define VNF_HAVELABEL 0x0400
#define VNF_SIMPLE 0x1000
#define VNF_READONLY 0x2000
#define VNDRW(v) ((v)->sc_flags & VNF_READONLY ? FREAD : FREAD|FWRITE)
struct vnd_softc *vnd_softc;
int numvnd = 0;
struct dkdriver vnddkdriver = { vndstrategy };
/* called by main() at boot time */
void vndattach(int);
void vndclear(struct vnd_softc *);
void vndstart(struct vnd_softc *);
int vndsetcred(struct vnd_softc *, struct ucred *);
void vndiodone(struct buf *);
void vndshutdown(void);
void vndgetdisklabel(dev_t, struct vnd_softc *);
void vndencrypt(struct vnd_softc *, caddr_t, size_t, daddr_t, int);
int vndlock(struct vnd_softc *);
void vndunlock(struct vnd_softc *);
void
vndencrypt(struct vnd_softc *vnd, caddr_t addr, size_t size, daddr_t off,
int encrypt)
{
int i, bsize;
u_char iv[8];
bsize = dbtob(1);
for (i = 0; i < size/bsize; i++) {
bzero(iv, sizeof(iv));
bcopy((u_char *)&off, iv, sizeof(off));
blf_ecb_encrypt(vnd->sc_keyctx, iv, sizeof(iv));
if (encrypt)
blf_cbc_encrypt(vnd->sc_keyctx, iv, addr, bsize);
else
blf_cbc_decrypt(vnd->sc_keyctx, iv, addr, bsize);
addr += bsize;
off++;
}
}
void
vndattach(int num)
{
char *mem;
u_long size;
if (num <= 0)
return;
size = num * sizeof(struct vnd_softc);
mem = malloc(size, M_DEVBUF, M_NOWAIT);
if (mem == NULL) {
printf("WARNING: no memory for vnode disks\n");
return;
}
bzero(mem, size);
vnd_softc = (struct vnd_softc *)mem;
numvnd = num;
pool_init(&vndbufpl, sizeof(struct vndbuf), 0, 0, 0, "vndbufpl", NULL);
pool_setlowat(&vndbufpl, 16);
pool_sethiwat(&vndbufpl, 1024);
}
int
vndopen(dev_t dev, int flags, int mode, struct proc *p)
{
int unit = vndunit(dev);
struct vnd_softc *sc;
int error = 0, part, pmask;
DNPRINTF(VDB_FOLLOW, "vndopen(%x, %x, %x, %p)\n", dev, flags, mode, p);
if (unit >= numvnd)
return (ENXIO);
sc = &vnd_softc[unit];
if ((error = vndlock(sc)) != 0)
return (error);
if ((flags & FWRITE) && (sc->sc_flags & VNF_READONLY)) {
error = EROFS;
goto bad;
}
if ((sc->sc_flags & VNF_INITED) &&
(sc->sc_flags & VNF_HAVELABEL) == 0) {
sc->sc_flags |= VNF_HAVELABEL;
vndgetdisklabel(dev, sc);
}
part = DISKPART(dev);
pmask = 1 << part;
/*
* If any partition is open, all succeeding openings must be of the
* same type or read-only.
*/
if (sc->sc_dk.dk_openmask) {
if (((sc->sc_flags & VNF_SIMPLE) != 0) !=
(vndsimple(dev) != 0) && (flags & FWRITE)) {
error = EBUSY;
goto bad;
}
} else if (vndsimple(dev))
sc->sc_flags |= VNF_SIMPLE;
else
sc->sc_flags &= ~VNF_SIMPLE;
/* Check that the partition exists. */
if (part != RAW_PART &&
((sc->sc_flags & VNF_HAVELABEL) == 0 ||
part >= sc->sc_dk.dk_label->d_npartitions ||
sc->sc_dk.dk_label->d_partitions[part].p_fstype == FS_UNUSED)) {
error = ENXIO;
goto bad;
}
/* Prevent our unit from being unconfigured while open. */
switch (mode) {
case S_IFCHR:
sc->sc_dk.dk_copenmask |= pmask;
break;
case S_IFBLK:
sc->sc_dk.dk_bopenmask |= pmask;
break;
}
sc->sc_dk.dk_openmask =
sc->sc_dk.dk_copenmask | sc->sc_dk.dk_bopenmask;
error = 0;
bad:
vndunlock(sc);
return (error);
}
/*
* Load the label information on the named device
*/
void
vndgetdisklabel(dev_t dev, struct vnd_softc *sc)
{
struct disklabel *lp = sc->sc_dk.dk_label;
char *errstring = NULL;
bzero(lp, sizeof(struct disklabel));
bzero(sc->sc_dk.dk_cpulabel, sizeof(struct cpu_disklabel));
lp->d_secsize = 512;
lp->d_ntracks = 1;
lp->d_nsectors = 100;
lp->d_ncylinders = sc->sc_size / 100;
lp->d_secpercyl = 100; /* lp->d_ntracks * lp->d_nsectors */
strncpy(lp->d_typename, "vnd device", sizeof(lp->d_typename));
lp->d_type = DTYPE_VND;
strncpy(lp->d_packname, "fictitious", sizeof(lp->d_packname));
lp->d_secperunit = sc->sc_size;
lp->d_rpm = 3600;
lp->d_interleave = 1;
lp->d_flags = 0;
lp->d_partitions[RAW_PART].p_offset = 0;
lp->d_partitions[RAW_PART].p_size = lp->d_secperunit;
lp->d_partitions[RAW_PART].p_fstype = FS_UNUSED;
lp->d_npartitions = RAW_PART + 1;
lp->d_magic = DISKMAGIC;
lp->d_magic2 = DISKMAGIC;
lp->d_checksum = dkcksum(lp);
/* Call the generic disklabel extraction routine */
errstring = readdisklabel(VNDLABELDEV(dev), vndstrategy, lp,
sc->sc_dk.dk_cpulabel, 0);
if (errstring) {
DNPRINTF(VDB_IO, "%s: %s\n", sc->sc_dev.dv_xname,
errstring);
return;
}
}
int
vndclose(dev_t dev, int flags, int mode, struct proc *p)
{
int unit = vndunit(dev);
struct vnd_softc *sc;
int error = 0, part;
DNPRINTF(VDB_FOLLOW, "vndclose(%x, %x, %x, %p)\n", dev, flags, mode, p);
if (unit >= numvnd)
return (ENXIO);
sc = &vnd_softc[unit];
if ((error = vndlock(sc)) != 0)
return (error);
part = DISKPART(dev);
/* ...that much closer to allowing unconfiguration... */
switch (mode) {
case S_IFCHR:
sc->sc_dk.dk_copenmask &= ~(1 << part);
break;
case S_IFBLK:
sc->sc_dk.dk_bopenmask &= ~(1 << part);
break;
}
sc->sc_dk.dk_openmask =
sc->sc_dk.dk_copenmask | sc->sc_dk.dk_bopenmask;
vndunlock(sc);
return (0);
}
/*
* Two methods are used, the traditional buffercache bypassing and the
* newer, cache-coherent on unmount, one.
*
* Former method:
* Break the request into bsize pieces and submit using VOP_BMAP/VOP_STRATEGY.
* Note that this driver can only be used for swapping over NFS on the hp
* since nfs_strategy on the vax cannot handle u-areas and page tables.
*
* Latter method:
* Repack the buffer into an uio structure and use VOP_READ/VOP_WRITE to
* access the underlying file.
*/
void
vndstrategy(struct buf *bp)
{
int unit = vndunit(bp->b_dev);
struct vnd_softc *vnd = &vnd_softc[unit];
struct vndbuf *nbp;
int bsize;
off_t bn;
caddr_t addr;
size_t resid;
int sz, flags, error, s;
struct iovec aiov;
struct uio auio;
struct proc *p = curproc;
DNPRINTF(VDB_FOLLOW, "vndstrategy(%p): unit %d\n", bp, unit);
if ((vnd->sc_flags & VNF_INITED) == 0) {
bp->b_error = ENXIO;
bp->b_flags |= B_ERROR;
s = splbio();
biodone(bp);
splx(s);
return;
}
bn = bp->b_blkno;
bp->b_resid = bp->b_bcount;
if (bn < 0) {
bp->b_error = EINVAL;
bp->b_flags |= B_ERROR;
s = splbio();
biodone(bp);
splx(s);
return;
}
/* If we have a label, do a boundary check. */
if (vnd->sc_flags & VNF_HAVELABEL) {
if (bounds_check_with_label(bp, vnd->sc_dk.dk_label,
vnd->sc_dk.dk_cpulabel, 1) <= 0) {
s = splbio();
biodone(bp);
splx(s);
return;
}
/*
* bounds_check_with_label() changes bp->b_resid, reset it
*/
bp->b_resid = bp->b_bcount;
}
sz = howmany(bp->b_bcount, DEV_BSIZE);
/* No bypassing of buffer cache? */
if (vndsimple(bp->b_dev)) {
/* Loop until all queued requests are handled. */
for (;;) {
int part = DISKPART(bp->b_dev);
int off = vnd->sc_dk.dk_label->d_partitions[part].p_offset;
aiov.iov_base = bp->b_data;
auio.uio_resid = aiov.iov_len = bp->b_bcount;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_offset = dbtob((off_t)(bp->b_blkno + off));
auio.uio_segflg = UIO_SYSSPACE;
auio.uio_procp = p;
vn_lock(vnd->sc_vp, LK_EXCLUSIVE | LK_RETRY, p);
if (bp->b_flags & B_READ) {
auio.uio_rw = UIO_READ;
bp->b_error = VOP_READ(vnd->sc_vp, &auio, 0,
vnd->sc_cred);
if (vnd->sc_keyctx)
vndencrypt(vnd, bp->b_data,
bp->b_bcount, bp->b_blkno, 0);
} else {
if (vnd->sc_keyctx)
vndencrypt(vnd, bp->b_data,
bp->b_bcount, bp->b_blkno, 1);
auio.uio_rw = UIO_WRITE;
bp->b_error = VOP_WRITE(vnd->sc_vp, &auio, 0,
vnd->sc_cred);
/* Data in buffer cache needs to be in clear */
if (vnd->sc_keyctx)
vndencrypt(vnd, bp->b_data,
bp->b_bcount, bp->b_blkno, 0);
}
VOP_UNLOCK(vnd->sc_vp, 0, p);
if (bp->b_error)
bp->b_flags |= B_ERROR;
bp->b_resid = auio.uio_resid;
s = splbio();
biodone(bp);
splx(s);
/* If nothing more is queued, we are done. */
if (!vnd->sc_tab.b_active)
return;
/*
* Dequeue now since lower level strategy
* routine might queue using same links.
*/
s = splbio();
bp = vnd->sc_tab.b_actf;
vnd->sc_tab.b_actf = bp->b_actf;
vnd->sc_tab.b_active--;
splx(s);
}
}
/* The old-style buffercache bypassing method. */
bn += vnd->sc_dk.dk_label->d_partitions[DISKPART(bp->b_dev)].p_offset;
bn = dbtob(bn);
bsize = vnd->sc_vp->v_mount->mnt_stat.f_iosize;
addr = bp->b_data;
flags = bp->b_flags | B_CALL;
for (resid = bp->b_resid; resid; resid -= sz) {
struct vnode *vp;
daddr64_t nbn;
int off, s, nra;
nra = 0;
vn_lock(vnd->sc_vp, LK_RETRY | LK_EXCLUSIVE, p);
error = VOP_BMAP(vnd->sc_vp, bn / bsize, &vp, &nbn, &nra);
VOP_UNLOCK(vnd->sc_vp, 0, p);
if (error == 0 && (long)nbn == -1)
error = EIO;
#ifdef VNDDEBUG
if (!dovndcluster)
nra = 0;
#endif
if ((off = bn % bsize) != 0)
sz = bsize - off;
else
sz = (1 + nra) * bsize;
if (resid < sz)
sz = resid;
DNPRINTF(VDB_IO, "vndstrategy: vp %p/%p bn %x/%x sz %x\n",
vnd->sc_vp, vp, bn, nbn, sz);
s = splbio();
nbp = getvndbuf();
splx(s);
nbp->vb_buf.b_flags = flags;
nbp->vb_buf.b_bcount = sz;
nbp->vb_buf.b_bufsize = bp->b_bufsize;
nbp->vb_buf.b_error = 0;
if (vp->v_type == VBLK || vp->v_type == VCHR)
nbp->vb_buf.b_dev = vp->v_rdev;
else
nbp->vb_buf.b_dev = NODEV;
nbp->vb_buf.b_data = addr;
nbp->vb_buf.b_blkno = nbn + btodb(off);
nbp->vb_buf.b_proc = bp->b_proc;
nbp->vb_buf.b_iodone = vndiodone;
nbp->vb_buf.b_vp = vp;
nbp->vb_buf.b_dirtyoff = bp->b_dirtyoff;
nbp->vb_buf.b_dirtyend = bp->b_dirtyend;
nbp->vb_buf.b_validoff = bp->b_validoff;
nbp->vb_buf.b_validend = bp->b_validend;
LIST_INIT(&nbp->vb_buf.b_dep);
/* save a reference to the old buffer */
nbp->vb_obp = bp;
/*
* If there was an error or a hole in the file...punt.
* Note that we deal with this after the nbp allocation.
* This ensures that we properly clean up any operations
* that we have already fired off.
*
* XXX we could deal with holes here but it would be
* a hassle (in the write case).
* We must still however charge for the write even if there
* was an error.
*/
if (error) {
nbp->vb_buf.b_error = error;
nbp->vb_buf.b_flags |= B_ERROR;
bp->b_resid -= (resid - sz);
s = splbio();
/* charge for the write */
if ((nbp->vb_buf.b_flags & B_READ) == 0)
nbp->vb_buf.b_vp->v_numoutput++;
biodone(&nbp->vb_buf);
splx(s);
return;
}
/*
* Just sort by block number
*/
nbp->vb_buf.b_cylinder = nbp->vb_buf.b_blkno;
s = splbio();
disksort(&vnd->sc_tab, &nbp->vb_buf);
vnd->sc_tab.b_active++;
vndstart(vnd);
splx(s);
bn += sz;
addr += sz;
}
}
/*
* Feed requests sequentially.
* We do it this way to keep from flooding NFS servers if we are connected
* to an NFS file. This places the burden on the client rather than the
* server.
*/
void
vndstart(struct vnd_softc *vnd)
{
struct buf *bp;
/*
* Dequeue now since lower level strategy routine might
* queue using same links
*/
bp = vnd->sc_tab.b_actf;
vnd->sc_tab.b_actf = bp->b_actf;
DNPRINTF(VDB_IO,
"vndstart(%d): bp %p vp %p blkno %x addr %p cnt %lx\n",
vnd-vnd_softc, bp, bp->b_vp, bp->b_blkno, bp->b_data,
bp->b_bcount);
/* Instrumentation. */
disk_busy(&vnd->sc_dk);
if ((bp->b_flags & B_READ) == 0)
bp->b_vp->v_numoutput++;
VOP_STRATEGY(bp);
}
void
vndiodone(struct buf *bp)
{
struct vndbuf *vbp = (struct vndbuf *) bp;
struct buf *pbp = vbp->vb_obp;
struct vnd_softc *vnd = &vnd_softc[vndunit(pbp->b_dev)];
splassert(IPL_BIO);
DNPRINTF(VDB_IO,
"vndiodone(%d): vbp %p vp %p blkno %x addr %p cnt %lx\n",
vnd-vnd_softc, vbp, vbp->vb_buf.b_vp, vbp->vb_buf.b_blkno,
vbp->vb_buf.b_data, vbp->vb_buf.b_bcount);
if (vbp->vb_buf.b_error) {
DNPRINTF(VDB_IO, "vndiodone: vbp %p error %d\n", vbp,
vbp->vb_buf.b_error);
pbp->b_flags |= B_ERROR;
pbp->b_error = biowait(&vbp->vb_buf);
}
pbp->b_resid -= vbp->vb_buf.b_bcount;
putvndbuf(vbp);
if (vnd->sc_tab.b_active) {
disk_unbusy(&vnd->sc_dk, (pbp->b_bcount - pbp->b_resid),
(pbp->b_flags & B_READ));
if (!vnd->sc_tab.b_actf)
vnd->sc_tab.b_active--;
}
if (pbp->b_resid == 0) {
DNPRINTF(VDB_IO, "vndiodone: pbp %p iodone\n", pbp);
biodone(pbp);
}
}
/* ARGSUSED */
int
vndread(dev_t dev, struct uio *uio, int flags)
{
int unit = vndunit(dev);
struct vnd_softc *sc;
DNPRINTF(VDB_FOLLOW, "vndread(%x, %p)\n", dev, uio);
if (unit >= numvnd)
return (ENXIO);
sc = &vnd_softc[unit];
if ((sc->sc_flags & VNF_INITED) == 0)
return (ENXIO);
return (physio(vndstrategy, NULL, dev, B_READ, minphys, uio));
}
/* ARGSUSED */
int
vndwrite(dev_t dev, struct uio *uio, int flags)
{
int unit = vndunit(dev);
struct vnd_softc *sc;
DNPRINTF(VDB_FOLLOW, "vndwrite(%x, %p)\n", dev, uio);
if (unit >= numvnd)
return (ENXIO);
sc = &vnd_softc[unit];
if ((sc->sc_flags & VNF_INITED) == 0)
return (ENXIO);
return (physio(vndstrategy, NULL, dev, B_WRITE, minphys, uio));
}
/* ARGSUSED */
int
vndioctl(dev_t dev, u_long cmd, caddr_t addr, int flag, struct proc *p)
{
int unit = vndunit(dev);
struct vnd_softc *vnd;
struct vnd_ioctl *vio;
struct vnd_user *vnu;
struct vattr vattr;
struct nameidata nd;
int error, part, pmask, s;
DNPRINTF(VDB_FOLLOW, "vndioctl(%x, %lx, %p, %x, %p): unit %d\n",
dev, cmd, addr, flag, p, unit);
error = suser(p, 0);
if (error)
return (error);
if (unit >= numvnd)
return (ENXIO);
vnd = &vnd_softc[unit];
vio = (struct vnd_ioctl *)addr;
switch (cmd) {
case VNDIOCSET:
if (vnd->sc_flags & VNF_INITED)
return (EBUSY);
if (!(vnd->sc_flags & VNF_SIMPLE) && vio->vnd_keylen)
return (EINVAL);
if ((error = vndlock(vnd)) != 0)
return (error);
if ((error = copyinstr(vio->vnd_file, vnd->sc_file,
sizeof(vnd->sc_file), NULL))) {
vndunlock(vnd);
return (error);
}
bzero(vnd->sc_dev.dv_xname, sizeof(vnd->sc_dev.dv_xname));
if (snprintf(vnd->sc_dev.dv_xname, sizeof(vnd->sc_dev.dv_xname),
"vnd%d", unit) >= sizeof(vnd->sc_dev.dv_xname)) {
printf("VNDIOCSET: device name too long\n");
vndunlock(vnd);
return(ENXIO);
}
/*
* Open for read and write first. This lets vn_open() weed out
* directories, sockets, etc. so we don't have to worry about
* them.
*/
NDINIT(&nd, LOOKUP, FOLLOW, UIO_USERSPACE, vio->vnd_file, p);
vnd->sc_flags &= ~VNF_READONLY;
error = vn_open(&nd, FREAD|FWRITE, 0);
if (error == EROFS) {
vnd->sc_flags |= VNF_READONLY;
error = vn_open(&nd, FREAD, 0);
}
if (error) {
vndunlock(vnd);
return (error);
}
error = VOP_GETATTR(nd.ni_vp, &vattr, p->p_ucred, p);
if (error) {
VOP_UNLOCK(nd.ni_vp, 0, p);
(void) vn_close(nd.ni_vp, VNDRW(vnd), p->p_ucred, p);
vndunlock(vnd);
return (error);
}
VOP_UNLOCK(nd.ni_vp, 0, p);
vnd->sc_vp = nd.ni_vp;
vnd->sc_size = btodb(vattr.va_size); /* note truncation */
if ((error = vndsetcred(vnd, p->p_ucred)) != 0) {
(void) vn_close(nd.ni_vp, VNDRW(vnd), p->p_ucred, p);
vndunlock(vnd);
return (error);
}
if (vio->vnd_keylen > 0) {
char key[128];
if (vio->vnd_keylen > sizeof(key))
vio->vnd_keylen = sizeof(key);
if ((error = copyin(vio->vnd_key, key,
vio->vnd_keylen)) != 0) {
(void) vn_close(nd.ni_vp, VNDRW(vnd),
p->p_ucred, p);
vndunlock(vnd);
return (error);
}
vnd->sc_keyctx = malloc(sizeof(*vnd->sc_keyctx), M_DEVBUF,
M_WAITOK);
blf_key(vnd->sc_keyctx, key, vio->vnd_keylen);
bzero(key, vio->vnd_keylen);
} else
vnd->sc_keyctx = NULL;
vio->vnd_size = dbtob((off_t)vnd->sc_size);
vnd->sc_flags |= VNF_INITED;
DNPRINTF(VDB_INIT, "vndioctl: SET vp %p size %llx\n",
vnd->sc_vp, (unsigned long long)vnd->sc_size);
/* Attach the disk. */
vnd->sc_dk.dk_driver = &vnddkdriver;
vnd->sc_dk.dk_name = vnd->sc_dev.dv_xname;
disk_attach(&vnd->sc_dk);
vndunlock(vnd);
break;
case VNDIOCCLR:
if ((vnd->sc_flags & VNF_INITED) == 0)
return (ENXIO);
if ((error = vndlock(vnd)) != 0)
return (error);
/*
* Don't unconfigure if any other partitions are open
* or if both the character and block flavors of this
* partition are open.
*/
part = DISKPART(dev);
pmask = (1 << part);
if ((vnd->sc_dk.dk_openmask & ~pmask) ||
((vnd->sc_dk.dk_bopenmask & pmask) &&
(vnd->sc_dk.dk_copenmask & pmask))) {
vndunlock(vnd);
return (EBUSY);
}
vndclear(vnd);
DNPRINTF(VDB_INIT, "vndioctl: CLRed\n");
/* Free crypto key */
if (vnd->sc_keyctx) {
bzero(vnd->sc_keyctx, sizeof(*vnd->sc_keyctx));
free(vnd->sc_keyctx, M_DEVBUF);
}
/* Detatch the disk. */
disk_detach(&vnd->sc_dk);
/* This must be atomic. */
s = splhigh();
vndunlock(vnd);
bzero(vnd, sizeof(struct vnd_softc));
splx(s);
break;
case VNDIOCGET:
vnu = (struct vnd_user *)addr;
if (vnu->vnu_unit == -1)
vnu->vnu_unit = unit;
if (vnu->vnu_unit >= numvnd)
return (ENXIO);
if (vnu->vnu_unit < 0)
return (EINVAL);
vnd = &vnd_softc[vnu->vnu_unit];
if (vnd->sc_flags & VNF_INITED) {
error = VOP_GETATTR(vnd->sc_vp, &vattr, p->p_ucred, p);
if (error)
return (error);
strlcpy(vnu->vnu_file, vnd->sc_file,
sizeof(vnu->vnu_file));
vnu->vnu_dev = vattr.va_fsid;
vnu->vnu_ino = vattr.va_fileid;
} else {
vnu->vnu_dev = 0;
vnu->vnu_ino = 0;
}
break;
case DIOCGDINFO:
if ((vnd->sc_flags & VNF_HAVELABEL) == 0)
return (ENOTTY);
*(struct disklabel *)addr = *(vnd->sc_dk.dk_label);
return (0);
case DIOCGPART:
if ((vnd->sc_flags & VNF_HAVELABEL) == 0)
return (ENOTTY);
((struct partinfo *)addr)->disklab = vnd->sc_dk.dk_label;
((struct partinfo *)addr)->part =
&vnd->sc_dk.dk_label->d_partitions[DISKPART(dev)];
return (0);
case DIOCWDINFO:
case DIOCSDINFO:
if ((vnd->sc_flags & VNF_HAVELABEL) == 0)
return (ENOTTY);
if ((flag & FWRITE) == 0)
return (EBADF);
if ((error = vndlock(vnd)) != 0)
return (error);
vnd->sc_flags |= VNF_LABELLING;
error = setdisklabel(vnd->sc_dk.dk_label,
(struct disklabel *)addr, /*vnd->sc_dk.dk_openmask : */0,
vnd->sc_dk.dk_cpulabel);
if (error == 0) {
if (cmd == DIOCWDINFO)
error = writedisklabel(MAKEDISKDEV(major(dev),
DISKUNIT(dev), RAW_PART),
vndstrategy, vnd->sc_dk.dk_label,
vnd->sc_dk.dk_cpulabel);
}
vnd->sc_flags &= ~VNF_LABELLING;
vndunlock(vnd);
return (error);
case DIOCWLABEL:
if ((flag & FWRITE) == 0)
return (EBADF);
if (*(int *)addr)
vnd->sc_flags |= VNF_WLABEL;
else
vnd->sc_flags &= ~VNF_WLABEL;
return (0);
default:
return (ENOTTY);
}
return (0);
}
/*
* Duplicate the current processes' credentials. Since we are called only
* as the result of a SET ioctl and only root can do that, any future access
* to this "disk" is essentially as root. Note that credentials may change
* if some other uid can write directly to the mapped file (NFS).
*/
int
vndsetcred(struct vnd_softc *vnd, struct ucred *cred)
{
struct uio auio;
struct iovec aiov;
char *tmpbuf;
int error;
struct proc *p = curproc;
vnd->sc_cred = crdup(cred);
tmpbuf = malloc(DEV_BSIZE, M_TEMP, M_WAITOK);
/* XXX: Horrible kludge to establish credentials for NFS */
aiov.iov_base = tmpbuf;
aiov.iov_len = MIN(DEV_BSIZE, dbtob((off_t)vnd->sc_size));
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_offset = 0;
auio.uio_rw = UIO_READ;
auio.uio_segflg = UIO_SYSSPACE;
auio.uio_resid = aiov.iov_len;
vn_lock(vnd->sc_vp, LK_RETRY | LK_EXCLUSIVE, p);
error = VOP_READ(vnd->sc_vp, &auio, 0, vnd->sc_cred);
VOP_UNLOCK(vnd->sc_vp, 0, p);
free(tmpbuf, M_TEMP);
return (error);
}
void
vndshutdown(void)
{
struct vnd_softc *vnd;
for (vnd = &vnd_softc[0]; vnd < &vnd_softc[numvnd]; vnd++)
if (vnd->sc_flags & VNF_INITED)
vndclear(vnd);
}
void
vndclear(struct vnd_softc *vnd)
{
struct vnode *vp = vnd->sc_vp;
struct proc *p = curproc; /* XXX */
DNPRINTF(VDB_FOLLOW, "vndclear(%p): vp %p\n", vnd, vp);
vnd->sc_flags &= ~VNF_INITED;
if (vp == NULL)
panic("vndioctl: null vp");
(void) vn_close(vp, VNDRW(vnd), vnd->sc_cred, p);
crfree(vnd->sc_cred);
vnd->sc_vp = NULL;
vnd->sc_cred = NULL;
vnd->sc_size = 0;
}
int
vndsize(dev_t dev)
{
int unit = vndunit(dev);
struct vnd_softc *vnd = &vnd_softc[unit];
if (unit >= numvnd || (vnd->sc_flags & VNF_INITED) == 0)
return (-1);
return (vnd->sc_size);
}
int
vnddump(dev_t dev, daddr_t blkno, caddr_t va, size_t size)
{
/* Not implemented. */
return (ENXIO);
}
/*
* Wait interruptibly for an exclusive lock.
*
* XXX
* Several drivers do this; it should be abstracted and made MP-safe.
*/
int
vndlock(struct vnd_softc *sc)
{
int error;
while ((sc->sc_flags & VNF_LOCKED) != 0) {
sc->sc_flags |= VNF_WANTED;
if ((error = tsleep(sc, PRIBIO | PCATCH, "vndlck", 0)) != 0)
return (error);
}
sc->sc_flags |= VNF_LOCKED;
return (0);
}
/*
* Unlock and wake up any waiters.
*/
void
vndunlock(struct vnd_softc *sc)
{
sc->sc_flags &= ~VNF_LOCKED;
if ((sc->sc_flags & VNF_WANTED) != 0) {
sc->sc_flags &= ~VNF_WANTED;
wakeup(sc);
}
}
|