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
|
/*
* Copyright (c) 1994, 1995 The Regents of the University of California.
* Copyright (c) 1994, 1995 Jan-Simon Pendry.
* All rights reserved.
*
* This code is derived from software donated to Berkeley by
* Jan-Simon Pendry.
*
* 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.
*
* @(#)union_vfsops.c 8.20 (Berkeley) 5/20/95
* $Id: union_vfsops.c,v 1.29 1998/05/06 05:29:37 msmith Exp $
*/
/*
* Union Layer
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/proc.h>
#include <sys/vnode.h>
#include <sys/mount.h>
#include <sys/namei.h>
#include <sys/malloc.h>
#include <sys/filedesc.h>
#include <miscfs/union/union.h>
static MALLOC_DEFINE(M_UNIONFSMNT, "UNION mount", "UNION mount structure");
extern int union_init __P((struct vfsconf *));
extern int union_fhtovp __P((struct mount *mp, struct fid *fidp,
struct mbuf *nam, struct vnode **vpp,
int *exflagsp, struct ucred **credanonp));
static int union_mount __P((struct mount *mp, char *path, caddr_t data,
struct nameidata *ndp, struct proc *p));
extern int union_quotactl __P((struct mount *mp, int cmd, uid_t uid,
caddr_t arg, struct proc *p));
static int union_root __P((struct mount *mp, struct vnode **vpp));
static int union_start __P((struct mount *mp, int flags, struct proc *p));
static int union_statfs __P((struct mount *mp, struct statfs *sbp,
struct proc *p));
extern int union_sync __P((struct mount *mp, int waitfor,
struct ucred *cred, struct proc *p));
static int union_unmount __P((struct mount *mp, int mntflags,
struct proc *p));
extern int union_vget __P((struct mount *mp, ino_t ino,
struct vnode **vpp));
extern int union_vptofh __P((struct vnode *vp, struct fid *fhp));
/*
* Mount union filesystem
*/
static int
union_mount(mp, path, data, ndp, p)
struct mount *mp;
char *path;
caddr_t data;
struct nameidata *ndp;
struct proc *p;
{
int error = 0;
struct union_args args;
struct vnode *lowerrootvp = NULLVP;
struct vnode *upperrootvp = NULLVP;
struct union_mount *um = 0;
struct ucred *cred = 0;
char *cp = 0;
int len;
u_int size;
#ifdef UNION_DIAGNOSTIC
printf("union_mount(mp = %x)\n", mp);
#endif
/*
* Disable clustered write, otherwise system becomes unstable.
*/
mp->mnt_flag |= MNT_NOCLUSTERW;
/*
* Update is a no-op
*/
if (mp->mnt_flag & MNT_UPDATE) {
/*
* Need to provide.
* 1. a way to convert between rdonly and rdwr mounts.
* 2. support for nfs exports.
*/
error = EOPNOTSUPP;
goto bad;
}
/*
* Get argument
*/
error = copyin(data, (caddr_t)&args, sizeof(struct union_args));
if (error)
goto bad;
lowerrootvp = mp->mnt_vnodecovered;
VREF(lowerrootvp);
/*
* Unlock lower node to avoid deadlock.
*/
if (lowerrootvp->v_op == union_vnodeop_p)
VOP_UNLOCK(lowerrootvp, 0, p);
/*
* Find upper node.
*/
NDINIT(ndp, LOOKUP, FOLLOW|WANTPARENT,
UIO_USERSPACE, args.target, p);
error = namei(ndp);
if (lowerrootvp->v_op == union_vnodeop_p)
vn_lock(lowerrootvp, LK_EXCLUSIVE | LK_RETRY, p);
if (error)
goto bad;
upperrootvp = ndp->ni_vp;
vrele(ndp->ni_dvp);
ndp->ni_dvp = NULL;
/*
* Check multi union mount to avoid `lock myself again' panic.
*/
if (upperrootvp == VTOUNION(lowerrootvp)->un_uppervp) {
#ifdef DIAGNOSTIC
printf("union_mount: multi union mount?\n");
#endif
error = EDEADLK;
goto bad;
}
if (upperrootvp->v_type != VDIR) {
error = EINVAL;
goto bad;
}
um = (struct union_mount *) malloc(sizeof(struct union_mount),
M_UNIONFSMNT, M_WAITOK); /* XXX */
/*
* Keep a held reference to the target vnodes.
* They are vrele'd in union_unmount.
*
* Depending on the _BELOW flag, the filesystems are
* viewed in a different order. In effect, this is the
* same as providing a mount under option to the mount syscall.
*/
um->um_op = args.mntflags & UNMNT_OPMASK;
switch (um->um_op) {
case UNMNT_ABOVE:
um->um_lowervp = lowerrootvp;
um->um_uppervp = upperrootvp;
break;
case UNMNT_BELOW:
um->um_lowervp = upperrootvp;
um->um_uppervp = lowerrootvp;
break;
case UNMNT_REPLACE:
vrele(lowerrootvp);
lowerrootvp = NULLVP;
um->um_uppervp = upperrootvp;
um->um_lowervp = lowerrootvp;
break;
default:
error = EINVAL;
goto bad;
}
/*
* Unless the mount is readonly, ensure that the top layer
* supports whiteout operations
*/
if ((mp->mnt_flag & MNT_RDONLY) == 0) {
error = VOP_WHITEOUT(um->um_uppervp, (struct componentname *) 0, LOOKUP);
if (error)
goto bad;
}
um->um_cred = p->p_ucred;
crhold(um->um_cred);
um->um_cmode = UN_DIRMODE &~ p->p_fd->fd_cmask;
/*
* Depending on what you think the MNT_LOCAL flag might mean,
* you may want the && to be || on the conditional below.
* At the moment it has been defined that the filesystem is
* only local if it is all local, ie the MNT_LOCAL flag implies
* that the entire namespace is local. If you think the MNT_LOCAL
* flag implies that some of the files might be stored locally
* then you will want to change the conditional.
*/
if (um->um_op == UNMNT_ABOVE) {
if (((um->um_lowervp == NULLVP) ||
(um->um_lowervp->v_mount->mnt_flag & MNT_LOCAL)) &&
(um->um_uppervp->v_mount->mnt_flag & MNT_LOCAL))
mp->mnt_flag |= MNT_LOCAL;
}
/*
* Copy in the upper layer's RDONLY flag. This is for the benefit
* of lookup() which explicitly checks the flag, rather than asking
* the filesystem for its own opinion. This means, that an update
* mount of the underlying filesystem to go from rdonly to rdwr
* will leave the unioned view as read-only.
*/
mp->mnt_flag |= (um->um_uppervp->v_mount->mnt_flag & MNT_RDONLY);
mp->mnt_data = (qaddr_t) um;
vfs_getnewfsid(mp);
(void) copyinstr(path, mp->mnt_stat.f_mntonname, MNAMELEN - 1, &size);
bzero(mp->mnt_stat.f_mntonname + size, MNAMELEN - size);
switch (um->um_op) {
case UNMNT_ABOVE:
cp = "<above>:";
break;
case UNMNT_BELOW:
cp = "<below>:";
break;
case UNMNT_REPLACE:
cp = "";
break;
}
len = strlen(cp);
bcopy(cp, mp->mnt_stat.f_mntfromname, len);
cp = mp->mnt_stat.f_mntfromname + len;
len = MNAMELEN - len;
(void) copyinstr(args.target, cp, len - 1, &size);
bzero(cp + size, len - size);
(void)union_statfs(mp, &mp->mnt_stat, p);
#ifdef UNION_DIAGNOSTIC
printf("union_mount: from %s, on %s\n",
mp->mnt_stat.f_mntfromname, mp->mnt_stat.f_mntonname);
#endif
return (0);
bad:
if (um)
free(um, M_UNIONFSMNT);
if (cred)
crfree(cred);
if (upperrootvp)
vrele(upperrootvp);
if (lowerrootvp)
vrele(lowerrootvp);
return (error);
}
/*
* VFS start. Nothing needed here - the start routine
* on the underlying filesystem(s) will have been called
* when that filesystem was mounted.
*/
static int
union_start(mp, flags, p)
struct mount *mp;
int flags;
struct proc *p;
{
return (0);
}
/*
* Free reference to union layer
*/
static int
union_unmount(mp, mntflags, p)
struct mount *mp;
int mntflags;
struct proc *p;
{
struct union_mount *um = MOUNTTOUNIONMOUNT(mp);
struct vnode *um_rootvp;
int error;
int freeing;
int flags = 0;
#ifdef UNION_DIAGNOSTIC
printf("union_unmount(mp = %x)\n", mp);
#endif
if (mntflags & MNT_FORCE)
flags |= FORCECLOSE;
if (error = union_root(mp, &um_rootvp))
return (error);
/*
* Keep flushing vnodes from the mount list.
* This is needed because of the un_pvp held
* reference to the parent vnode.
* If more vnodes have been freed on a given pass,
* the try again. The loop will iterate at most
* (d) times, where (d) is the maximum tree depth
* in the filesystem.
*/
for (freeing = 0; vflush(mp, um_rootvp, flags) != 0;) {
struct vnode *vp;
int n;
/* count #vnodes held on mount list */
for (n = 0, vp = mp->mnt_vnodelist.lh_first;
vp != NULLVP;
vp = vp->v_mntvnodes.le_next)
n++;
/* if this is unchanged then stop */
if (n == freeing)
break;
/* otherwise try once more time */
freeing = n;
}
/* At this point the root vnode should have a single reference */
if (um_rootvp->v_usecount > 1) {
vput(um_rootvp);
return (EBUSY);
}
#ifdef UNION_DIAGNOSTIC
vprint("union root", um_rootvp);
#endif
/*
* Discard references to upper and lower target vnodes.
*/
if (um->um_lowervp)
vrele(um->um_lowervp);
vrele(um->um_uppervp);
crfree(um->um_cred);
/*
* Release reference on underlying root vnode
*/
vput(um_rootvp);
/*
* And blow it away for future re-use
*/
vgone(um_rootvp);
/*
* Finally, throw away the union_mount structure
*/
free(mp->mnt_data, M_UNIONFSMNT); /* XXX */
mp->mnt_data = 0;
return (0);
}
static int
union_root(mp, vpp)
struct mount *mp;
struct vnode **vpp;
{
struct proc *p = curproc; /* XXX */
struct union_mount *um = MOUNTTOUNIONMOUNT(mp);
int error;
int loselock;
int lockadj = 0;
if (um->um_lowervp && um->um_op != UNMNT_BELOW &&
VOP_ISLOCKED(um->um_lowervp)) {
VREF(um->um_lowervp);
VOP_UNLOCK(um->um_lowervp, 0, p);
lockadj = 1;
}
/*
* Return locked reference to root.
*/
VREF(um->um_uppervp);
if ((um->um_op == UNMNT_BELOW) &&
VOP_ISLOCKED(um->um_uppervp)) {
loselock = 1;
} else {
vn_lock(um->um_uppervp, LK_EXCLUSIVE | LK_RETRY, p);
loselock = 0;
}
if (um->um_lowervp)
VREF(um->um_lowervp);
error = union_allocvp(vpp, mp,
(struct vnode *) 0,
(struct vnode *) 0,
(struct componentname *) 0,
um->um_uppervp,
um->um_lowervp,
1);
if (error) {
if (loselock)
vrele(um->um_uppervp);
else
vput(um->um_uppervp);
if (um->um_lowervp)
vrele(um->um_lowervp);
} else {
if (loselock)
VTOUNION(*vpp)->un_flags &= ~UN_ULOCK;
}
if (lockadj) {
vn_lock(um->um_lowervp, LK_EXCLUSIVE | LK_RETRY, p);
vrele(um->um_lowervp);
}
return (error);
}
static int
union_statfs(mp, sbp, p)
struct mount *mp;
struct statfs *sbp;
struct proc *p;
{
int error;
struct union_mount *um = MOUNTTOUNIONMOUNT(mp);
struct statfs mstat;
int lbsize;
#ifdef UNION_DIAGNOSTIC
printf("union_statfs(mp = %x, lvp = %x, uvp = %x)\n", mp,
um->um_lowervp,
um->um_uppervp);
#endif
bzero(&mstat, sizeof(mstat));
if (um->um_lowervp) {
error = VFS_STATFS(um->um_lowervp->v_mount, &mstat, p);
if (error)
return (error);
}
/* now copy across the "interesting" information and fake the rest */
#if 0
sbp->f_type = mstat.f_type;
sbp->f_flags = mstat.f_flags;
sbp->f_bsize = mstat.f_bsize;
sbp->f_iosize = mstat.f_iosize;
#endif
lbsize = mstat.f_bsize;
sbp->f_blocks = mstat.f_blocks;
sbp->f_bfree = mstat.f_bfree;
sbp->f_bavail = mstat.f_bavail;
sbp->f_files = mstat.f_files;
sbp->f_ffree = mstat.f_ffree;
error = VFS_STATFS(um->um_uppervp->v_mount, &mstat, p);
if (error)
return (error);
sbp->f_flags = mstat.f_flags;
sbp->f_bsize = mstat.f_bsize;
sbp->f_iosize = mstat.f_iosize;
/*
* if the lower and upper blocksizes differ, then frig the
* block counts so that the sizes reported by df make some
* kind of sense. none of this makes sense though.
*/
if (mstat.f_bsize != lbsize)
sbp->f_blocks = ((off_t) sbp->f_blocks * lbsize) / mstat.f_bsize;
/*
* The "total" fields count total resources in all layers,
* the "free" fields count only those resources which are
* free in the upper layer (since only the upper layer
* is writeable).
*/
sbp->f_blocks += mstat.f_blocks;
sbp->f_bfree = mstat.f_bfree;
sbp->f_bavail = mstat.f_bavail;
sbp->f_files += mstat.f_files;
sbp->f_ffree = mstat.f_ffree;
if (sbp != &mp->mnt_stat) {
sbp->f_type = mp->mnt_vfc->vfc_typenum;
bcopy(&mp->mnt_stat.f_fsid, &sbp->f_fsid, sizeof(sbp->f_fsid));
bcopy(mp->mnt_stat.f_mntonname, sbp->f_mntonname, MNAMELEN);
bcopy(mp->mnt_stat.f_mntfromname, sbp->f_mntfromname, MNAMELEN);
}
return (0);
}
/*
* XXX - Assumes no data cached at union layer.
*/
#define union_sync ((int (*) __P((struct mount *, int, struct ucred *, \
struct proc *)))nullop)
#define union_fhtovp ((int (*) __P((struct mount *, struct fid *, \
struct sockaddr *, struct vnode **, int *, struct ucred **)))eopnotsupp)
#define union_quotactl ((int (*) __P((struct mount *, int, uid_t, caddr_t, \
struct proc *)))eopnotsupp)
#define union_sysctl ((int (*) __P((int *, u_int, void *, size_t *, void *, \
size_t, struct proc *)))eopnotsupp)
#define union_vget ((int (*) __P((struct mount *, ino_t, struct vnode **))) \
eopnotsupp)
#define union_vptofh ((int (*) __P((struct vnode *, struct fid *)))eopnotsupp)
static struct vfsops union_vfsops = {
union_mount,
union_start,
union_unmount,
union_root,
union_quotactl,
union_statfs,
union_sync,
union_vget,
union_fhtovp,
union_vptofh,
union_init,
};
VFS_SET(union_vfsops, union, VFCF_LOOPBACK);
|