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
|
/* $OpenBSD: vnode_pager.c,v 1.6 1997/11/06 05:59:39 csapuntz Exp $ */
/* $NetBSD: vnode_pager.c,v 1.19 1996/03/16 23:15:27 christos Exp $ */
/*
* Copyright (c) 1990 University of Utah.
* Copyright (c) 1991, 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. 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.
*
* @(#)vnode_pager.c 8.10 (Berkeley) 5/14/95
*/
/*
* Page to/from files (vnodes).
*
* TODO:
* pageouts
* fix credential use (uses current process credentials now)
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/proc.h>
#include <sys/malloc.h>
#include <sys/vnode.h>
#include <sys/uio.h>
#include <sys/mount.h>
#include <vm/vm.h>
#include <vm/vm_page.h>
#include <vm/vnode_pager.h>
struct pagerlst vnode_pager_list; /* list of managed vnodes */
#ifdef DEBUG
int vpagerdebug = 0x00;
#define VDB_FOLLOW 0x01
#define VDB_INIT 0x02
#define VDB_IO 0x04
#define VDB_FAIL 0x08
#define VDB_ALLOC 0x10
#define VDB_SIZE 0x20
#endif
static vm_pager_t vnode_pager_alloc
__P((caddr_t, vm_size_t, vm_prot_t, vm_offset_t));
static void vnode_pager_cluster
__P((vm_pager_t, vm_offset_t,
vm_offset_t *, vm_offset_t *));
static void vnode_pager_dealloc __P((vm_pager_t));
static int vnode_pager_getpage
__P((vm_pager_t, vm_page_t *, int, boolean_t));
static boolean_t vnode_pager_haspage __P((vm_pager_t, vm_offset_t));
static void vnode_pager_init __P((void));
static int vnode_pager_io
__P((vn_pager_t, vm_page_t *, int,
boolean_t, enum uio_rw));
static boolean_t vnode_pager_putpage
__P((vm_pager_t, vm_page_t *, int, boolean_t));
struct pagerops vnodepagerops = {
vnode_pager_init,
vnode_pager_alloc,
vnode_pager_dealloc,
vnode_pager_getpage,
vnode_pager_putpage,
vnode_pager_haspage,
vnode_pager_cluster
};
static void
vnode_pager_init()
{
#ifdef DEBUG
if (vpagerdebug & VDB_FOLLOW)
printf("vnode_pager_init()\n");
#endif
TAILQ_INIT(&vnode_pager_list);
}
/*
* Allocate (or lookup) pager for a vnode.
* Handle is a vnode pointer.
*/
static vm_pager_t
vnode_pager_alloc(handle, size, prot, foff)
caddr_t handle;
vm_size_t size;
vm_prot_t prot;
vm_offset_t foff;
{
register vm_pager_t pager;
register vn_pager_t vnp;
vm_object_t object;
struct vattr vattr;
struct vnode *vp;
struct proc *p = curproc; /* XXX */
#ifdef DEBUG
if (vpagerdebug & (VDB_FOLLOW|VDB_ALLOC))
printf("vnode_pager_alloc(%p, %lx, %x)\n", handle, size, prot);
#endif
/*
* Pageout to vnode, no can do yet.
*/
if (handle == NULL)
return(NULL);
/*
* Vnodes keep a pointer to any associated pager so no need to
* lookup with vm_pager_lookup.
*/
vp = (struct vnode *)handle;
pager = (vm_pager_t)vp->v_vmdata;
if (pager == NULL) {
/*
* Allocate pager structures
*/
pager = (vm_pager_t)malloc(sizeof *pager, M_VMPAGER, M_WAITOK);
if (pager == NULL)
return(NULL);
vnp = (vn_pager_t)malloc(sizeof *vnp, M_VMPGDATA, M_WAITOK);
if (vnp == NULL) {
free((caddr_t)pager, M_VMPAGER);
return(NULL);
}
/*
* And an object of the appropriate size
*/
if (VOP_GETATTR(vp, &vattr, p->p_ucred, p) == 0) {
object = vm_object_allocate(round_page(vattr.va_size));
vm_object_enter(object, pager);
vm_object_setpager(object, pager, 0, TRUE);
} else {
free((caddr_t)vnp, M_VMPGDATA);
free((caddr_t)pager, M_VMPAGER);
return(NULL);
}
/*
* Hold a reference to the vnode and initialize pager data.
*/
VREF(vp);
vnp->vnp_flags = 0;
vnp->vnp_vp = vp;
vnp->vnp_size = vattr.va_size;
TAILQ_INSERT_TAIL(&vnode_pager_list, pager, pg_list);
pager->pg_handle = handle;
pager->pg_type = PG_VNODE;
pager->pg_flags = 0;
pager->pg_ops = &vnodepagerops;
pager->pg_data = vnp;
vp->v_vmdata = (caddr_t)pager;
} else {
/*
* vm_object_lookup() will remove the object from the
* cache if found and also gain a reference to the object.
*/
object = vm_object_lookup(pager);
#ifdef DEBUG
vnp = (vn_pager_t)pager->pg_data;
#endif
}
#ifdef DEBUG
if (vpagerdebug & VDB_ALLOC)
printf("vnode_pager_setup: vp %p sz %lx pager %p object %p\n",
vp, vnp->vnp_size, pager, object);
#endif
return(pager);
}
static void
vnode_pager_dealloc(pager)
vm_pager_t pager;
{
register vn_pager_t vnp = (vn_pager_t)pager->pg_data;
register struct vnode *vp;
#ifdef NOTDEF
struct proc *p = curproc; /* XXX */
#endif
#ifdef DEBUG
if (vpagerdebug & VDB_FOLLOW)
printf("vnode_pager_dealloc(%p)\n", pager);
#endif
if ((vp = vnp->vnp_vp) != NULL) {
vp->v_vmdata = NULL;
vp->v_flag &= ~VTEXT;
#if NOTDEF
/* can hang if done at reboot on NFS FS */
(void) VOP_FSYNC(vp, p->p_ucred, p);
#endif
vrele(vp);
}
TAILQ_REMOVE(&vnode_pager_list, pager, pg_list);
free((caddr_t)vnp, M_VMPGDATA);
free((caddr_t)pager, M_VMPAGER);
}
static int
vnode_pager_getpage(pager, mlist, npages, sync)
vm_pager_t pager;
vm_page_t *mlist;
int npages;
boolean_t sync;
{
#ifdef DEBUG
if (vpagerdebug & VDB_FOLLOW)
printf("vnode_pager_getpage(%p, %p, %x, %x)\n",
pager, mlist, npages, sync);
#endif
return(vnode_pager_io((vn_pager_t)pager->pg_data,
mlist, npages, sync, UIO_READ));
}
static boolean_t
vnode_pager_putpage(pager, mlist, npages, sync)
vm_pager_t pager;
vm_page_t *mlist;
int npages;
boolean_t sync;
{
int err;
#ifdef DEBUG
if (vpagerdebug & VDB_FOLLOW)
printf("vnode_pager_putpage(%p, %p, %x, %x)\n",
pager, mlist, npages, sync);
#endif
if (pager == NULL)
return (FALSE); /* ??? */
err = vnode_pager_io((vn_pager_t)pager->pg_data,
mlist, npages, sync, UIO_WRITE);
/*
* If the operation was successful, mark the pages clean.
*/
if (err == VM_PAGER_OK) {
while (npages--) {
(*mlist)->flags |= PG_CLEAN;
pmap_clear_modify(VM_PAGE_TO_PHYS(*mlist));
mlist++;
}
}
return(err);
}
static boolean_t
vnode_pager_haspage(pager, offset)
vm_pager_t pager;
vm_offset_t offset;
{
struct proc *p = curproc; /* XXX */
vn_pager_t vnp = (vn_pager_t)pager->pg_data;
daddr_t bn;
int err;
#ifdef DEBUG
if (vpagerdebug & VDB_FOLLOW)
printf("vnode_pager_haspage(%p, %lx)\n", pager, offset);
#endif
/*
* Offset beyond end of file, do not have the page
* Lock the vnode first to make sure we have the most recent
* version of the size.
*/
vn_lock(vnp->vnp_vp, LK_EXCLUSIVE | LK_RETRY, p);
if (offset >= vnp->vnp_size) {
VOP_UNLOCK(vnp->vnp_vp, 0, p);
#ifdef DEBUG
if (vpagerdebug & (VDB_FAIL|VDB_SIZE))
printf("vnode_pager_haspage: pg %p, off %lx, size %lx\n",
pager, offset, vnp->vnp_size);
#endif
return(FALSE);
}
/*
* Read the index to find the disk block to read
* from. If there is no block, report that we don't
* have this data.
*
* Assumes that the vnode has whole page or nothing.
*/
err = VOP_BMAP(vnp->vnp_vp,
offset / vnp->vnp_vp->v_mount->mnt_stat.f_iosize,
(struct vnode **)0, &bn, NULL);
VOP_UNLOCK(vnp->vnp_vp, 0, p);
if (err) {
#ifdef DEBUG
if (vpagerdebug & VDB_FAIL)
printf("vnode_pager_haspage: BMAP err %d, pg %p, off %lx\n",
err, pager, offset);
#endif
return(TRUE);
}
return((long)bn < 0 ? FALSE : TRUE);
}
static void
vnode_pager_cluster(pager, offset, loffset, hoffset)
vm_pager_t pager;
vm_offset_t offset;
vm_offset_t *loffset;
vm_offset_t *hoffset;
{
vn_pager_t vnp = (vn_pager_t)pager->pg_data;
vm_offset_t loff, hoff;
#ifdef DEBUG
if (vpagerdebug & VDB_FOLLOW)
printf("vnode_pager_cluster(%p, %lx) ", pager, offset);
#endif
loff = offset;
if (loff >= vnp->vnp_size)
panic("vnode_pager_cluster: bad offset");
/*
* XXX could use VOP_BMAP to get maxcontig value
*/
hoff = loff + MAXBSIZE;
if (hoff > round_page(vnp->vnp_size))
hoff = round_page(vnp->vnp_size);
*loffset = loff;
*hoffset = hoff;
#ifdef DEBUG
if (vpagerdebug & VDB_FOLLOW)
printf("returns [%lx-%lx]\n", loff, hoff);
#endif
}
/*
* (XXX)
* Lets the VM system know about a change in size for a file.
* If this vnode is mapped into some address space (i.e. we have a pager
* for it) we adjust our own internal size and flush any cached pages in
* the associated object that are affected by the size change.
*
* Note: this routine may be invoked as a result of a pager put
* operation (possibly at object termination time), so we must be careful.
*/
void
vnode_pager_setsize(vp, nsize)
struct vnode *vp;
u_long nsize;
{
register vn_pager_t vnp;
register vm_object_t object;
vm_pager_t pager;
/*
* Not a mapped vnode
*/
if (vp == NULL || vp->v_type != VREG || vp->v_vmdata == NULL)
return;
/*
* Hasn't changed size
*/
pager = (vm_pager_t)vp->v_vmdata;
vnp = (vn_pager_t)pager->pg_data;
if (nsize == vnp->vnp_size)
return;
/*
* No object.
* This can happen during object termination since
* vm_object_page_clean is called after the object
* has been removed from the hash table, and clean
* may cause vnode write operations which can wind
* up back here.
*/
object = vm_object_lookup(pager);
if (object == NULL)
return;
#ifdef DEBUG
if (vpagerdebug & (VDB_FOLLOW|VDB_SIZE))
printf("vnode_pager_setsize: vp %p obj %p osz %ld nsz %ld\n",
vp, object, vnp->vnp_size, nsize);
#endif
/*
* File has shrunk.
* Toss any cached pages beyond the new EOF.
*/
if (nsize < vnp->vnp_size) {
vm_object_lock(object);
vm_object_page_remove(object,
(vm_offset_t)nsize, vnp->vnp_size);
vm_object_unlock(object);
}
vnp->vnp_size = (vm_offset_t)nsize;
vm_object_deallocate(object);
}
void
vnode_pager_umount(mp)
register struct mount *mp;
{
struct proc *p = curproc; /* XXX */
vm_pager_t pager, npager;
struct vnode *vp;
for (pager = vnode_pager_list.tqh_first; pager != NULL; pager = npager){
/*
* Save the next pointer now since uncaching may
* terminate the object and render pager invalid
*/
npager = pager->pg_list.tqe_next;
vp = ((vn_pager_t)pager->pg_data)->vnp_vp;
if (mp == (struct mount *)0 || vp->v_mount == mp) {
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p);
(void) vnode_pager_uncache(vp);
VOP_UNLOCK(vp, 0, p);
}
}
}
/*
* Remove vnode associated object from the object cache.
*
* XXX unlock the vnode if it is currently locked.
* We must do this since uncaching the object may result in its
* destruction which may initiate paging activity which may necessitate
* re-locking the vnode.
*/
boolean_t
vnode_pager_uncache(vp)
register struct vnode *vp;
{
struct proc *p = curproc; /* XXX */
vm_object_t object;
boolean_t uncached;
vm_pager_t pager;
/*
* Not a mapped vnode
*/
if (vp->v_type != VREG || (pager = (vm_pager_t)vp->v_vmdata) == NULL)
return (TRUE);
#ifdef DEBUG
if (!VOP_ISLOCKED(vp)) {
#ifdef NFSCLIENT
extern int (**nfsv2_vnodeop_p) __P((void *));
extern int (**spec_nfsv2nodeop_p) __P((void *));
#ifdef FIFO
extern int (**fifo_nfsv2nodeop_p) __P((void *));
#endif
if (vp->v_op != nfsv2_vnodeop_p
&& vp->v_op != spec_nfsv2nodeop_p
#ifdef FIFO
&& vp->v_op != fifo_nfsv2nodeop_p
#endif
)
#endif
panic("vnode_pager_uncache: vnode not locked!");
}
#endif
/*
* Must use vm_object_lookup() as it actually removes
* the object from the cache list.
*/
object = vm_object_lookup(pager);
if (object) {
uncached = (object->ref_count <= 1);
VOP_UNLOCK(vp, 0, p);
pager_cache(object, FALSE);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p);
} else
uncached = TRUE;
return(uncached);
}
static int
vnode_pager_io(vnp, mlist, npages, sync, rw)
register vn_pager_t vnp;
vm_page_t *mlist;
int npages;
boolean_t sync;
enum uio_rw rw;
{
struct uio auio;
struct iovec aiov;
vm_offset_t kva, foff;
int error, size;
struct proc *p = curproc; /* XXX */
/* XXX */
vm_page_t m;
if (npages != 1)
panic("vnode_pager_io: cannot handle multiple pages");
m = *mlist;
/* XXX */
#ifdef DEBUG
if (vpagerdebug & VDB_FOLLOW)
printf("vnode_pager_io(%p, %p, %c): vnode %p\n",
vnp, m, rw == UIO_READ ? 'R' : 'W', vnp->vnp_vp);
#endif
foff = m->offset + m->object->paging_offset;
/*
* Allocate a kernel virtual address and initialize so that
* we can use VOP_READ/WRITE routines.
*/
kva = vm_pager_map_pages(mlist, npages, sync);
if (kva == NULL)
return(VM_PAGER_AGAIN);
/*
* After all of the potentially blocking operations have been
* performed, we can do the size checks:
* read beyond EOF (returns error)
* short read
*/
vn_lock(vnp->vnp_vp, LK_EXCLUSIVE | LK_RETRY, p);
if (foff >= vnp->vnp_size) {
VOP_UNLOCK(vnp->vnp_vp, 0, p);
vm_pager_unmap_pages(kva, npages);
#ifdef DEBUG
if (vpagerdebug & VDB_SIZE)
printf("vnode_pager_io: vp %p, off %ld size %ld\n",
vnp->vnp_vp, foff, vnp->vnp_size);
#endif
return(VM_PAGER_BAD);
}
if (foff + PAGE_SIZE > vnp->vnp_size)
size = vnp->vnp_size - foff;
else
size = PAGE_SIZE;
aiov.iov_base = (caddr_t)kva;
aiov.iov_len = size;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_offset = foff;
auio.uio_segflg = UIO_SYSSPACE;
auio.uio_rw = rw;
auio.uio_resid = size;
auio.uio_procp = (struct proc *)0;
#ifdef DEBUG
if (vpagerdebug & VDB_IO)
printf("vnode_pager_io: vp %p kva %lx foff %lx size %x",
vnp->vnp_vp, kva, foff, size);
#endif
if (rw == UIO_READ)
error = VOP_READ(vnp->vnp_vp, &auio, 0, p->p_ucred);
else
error = VOP_WRITE(vnp->vnp_vp, &auio, 0, p->p_ucred);
VOP_UNLOCK(vnp->vnp_vp, 0, p);
#ifdef DEBUG
if (vpagerdebug & VDB_IO) {
if (error || auio.uio_resid)
printf(" returns error %x, resid %x",
error, auio.uio_resid);
printf("\n");
}
#endif
if (!error) {
register int count = size - auio.uio_resid;
if (count == 0)
error = EINVAL;
else if (count != PAGE_SIZE && rw == UIO_READ)
bzero((void *)(kva + count), PAGE_SIZE - count);
}
vm_pager_unmap_pages(kva, npages);
return (error ? VM_PAGER_ERROR : VM_PAGER_OK);
}
|