summaryrefslogtreecommitdiff
path: root/sys/ufs/lfs/lfs_syscalls.c
blob: 45363d78ce87ee3432200c2977c5c5520f4042f5 (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
/*	$OpenBSD: lfs_syscalls.c,v 1.6 2002/03/14 01:27:15 millert Exp $	*/
/*	$NetBSD: lfs_syscalls.c,v 1.10 1996/02/09 22:28:56 christos Exp $	*/

/*-
 * Copyright (c) 1991, 1993, 1994
 *	The Regents of the University of California.  All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. 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.
 *
 *	@(#)lfs_syscalls.c	8.10 (Berkeley) 5/14/95
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/proc.h>
#include <sys/buf.h>
#include <sys/mount.h>
#include <sys/vnode.h>
#include <sys/malloc.h>
#include <sys/kernel.h>

#include <sys/syscallargs.h>

#include <ufs/ufs/quota.h>
#include <ufs/ufs/inode.h>
#include <ufs/ufs/ufsmount.h>
#include <ufs/ufs/ufs_extern.h>

#include <ufs/lfs/lfs.h>
#include <ufs/lfs/lfs_extern.h>

#define BUMP_FIP(SP) \
	(SP)->fip = (FINFO *) (&(SP)->fip->fi_blocks[(SP)->fip->fi_nblocks])

#define INC_FINFO(SP) ++((SEGSUM *)((SP)->segsum))->ss_nfinfo
#define DEC_FINFO(SP) --((SEGSUM *)((SP)->segsum))->ss_nfinfo

/*
 * Before committing to add something to a segment summary, make sure there
 * is enough room.  S is the bytes added to the summary.
 */
#define	CHECK_SEG(s)			\
if (sp->sum_bytes_left < (s)) {		\
	(void) lfs_writeseg(fs, sp);	\
}
struct buf *lfs_fakebuf(struct vnode *, int, size_t, caddr_t);

int debug_cleaner = 0;
int clean_vnlocked = 0;
int clean_inlocked = 0;

/*
 * lfs_markv:
 *
 * This will mark inodes and blocks dirty, so they are written into the log.
 * It will block until all the blocks have been written.  The segment create
 * time passed in the block_info and inode_info structures is used to decide
 * if the data is valid for each block (in case some process dirtied a block
 * or inode that is being cleaned between the determination that a block is
 * live and the lfs_markv call).
 *
 *  0 on success
 * -1/errno is return on error.
 */
int
lfs_markv(p, v, retval)
	struct proc *p;
	void *v;
	register_t *retval;
{
	struct lfs_markv_args /* {
		syscallarg(fsid_t *) fsidp;
		syscallarg(struct block_info *) blkiov;
		syscallarg(int) blkcnt;
	} */ *uap = v;
	struct segment *sp;
	BLOCK_INFO *blkp;
	IFILE *ifp;
	struct buf *bp, **bpp;
	struct inode *ip;
	struct lfs *fs;
	struct mount *mntp;
	struct vnode *vp;
	fsid_t fsid;
	void *start;
	ino_t lastino;
	ufs_daddr_t b_daddr, v_daddr;
	u_long bsize;
	int cnt, error;

	if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
		return (error);

	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
		return (error);
	if ((mntp = getvfs(&fsid)) == NULL)
		return (EINVAL);

	cnt = SCARG(uap, blkcnt);
	start = malloc(cnt * sizeof(BLOCK_INFO), M_SEGMENT, M_WAITOK);
	error = copyin(SCARG(uap, blkiov), start, cnt * sizeof(BLOCK_INFO));
	if (error)
		goto err1;

	/* Mark blocks/inodes dirty.  */
	fs = VFSTOUFS(mntp)->um_lfs;
	bsize = fs->lfs_bsize;
	error = 0;

	lfs_seglock(fs, SEGM_SYNC | SEGM_CLEAN);
	sp = fs->lfs_sp;
	for (v_daddr = LFS_UNUSED_DADDR, lastino = LFS_UNUSED_INUM,
	    blkp = start; cnt--; ++blkp) {
		/*
		 * Get the IFILE entry (only once) and see if the file still
		 * exists.
		 */
		if (lastino != blkp->bi_inode) {
			if (lastino != LFS_UNUSED_INUM) {
				/* Finish up last file */
				if (sp->fip->fi_nblocks == 0) {
					DEC_FINFO(sp);
					sp->sum_bytes_left +=
					    sizeof(FINFO) - sizeof(ufs_daddr_t);
				} else {
					lfs_updatemeta(sp);
					BUMP_FIP(sp);
				}

				lfs_writeinode(fs, sp, ip);
				lfs_vunref(vp);
			}

			/* Start a new file */
			CHECK_SEG(sizeof(FINFO));
			sp->sum_bytes_left -= sizeof(FINFO) - sizeof(ufs_daddr_t);
			INC_FINFO(sp);
			sp->start_lbp = &sp->fip->fi_blocks[0];
			sp->vp = NULL;
			sp->fip->fi_version = blkp->bi_version;
			sp->fip->fi_nblocks = 0;
			sp->fip->fi_ino = blkp->bi_inode;
			lastino = blkp->bi_inode;
			if (blkp->bi_inode == LFS_IFILE_INUM)
				v_daddr = fs->lfs_idaddr;
			else {
				LFS_IENTRY(ifp, fs, blkp->bi_inode, bp);
				v_daddr = ifp->if_daddr;
				brelse(bp);
			}
			if (v_daddr == LFS_UNUSED_DADDR)
				continue;

			/* Get the vnode/inode. */
			if (lfs_fastvget(mntp, blkp->bi_inode, v_daddr, &vp,
			    blkp->bi_lbn == LFS_UNUSED_LBN ? 
			    blkp->bi_bp : NULL)) {
#ifdef DIAGNOSTIC
				printf("lfs_markv: VFS_VGET failed (%d)\n",
				    blkp->bi_inode);
				panic("lfs_markv VFS_VGET FAILED");
#endif
				lastino = LFS_UNUSED_INUM;
				v_daddr = LFS_UNUSED_DADDR;
				continue;
			}
			sp->vp = vp;
			ip = VTOI(vp);
		} else if (v_daddr == LFS_UNUSED_DADDR)
			continue;

		/* If this BLOCK_INFO didn't contain a block, keep going. */
		if (blkp->bi_lbn == LFS_UNUSED_LBN)
			continue;
		if (VOP_BMAP(vp, blkp->bi_lbn, NULL, &b_daddr, NULL) ||
		    b_daddr != blkp->bi_daddr)
			continue;
		/*
		 * If we got to here, then we are keeping the block.  If it
		 * is an indirect block, we want to actually put it in the
		 * buffer cache so that it can be updated in the finish_meta
		 * section.  If it's not, we need to allocate a fake buffer
		 * so that writeseg can perform the copyin and write the buffer.
		 */
		if (blkp->bi_lbn >= 0)	/* Data Block */
			bp = lfs_fakebuf(vp, blkp->bi_lbn, bsize,
			    blkp->bi_bp);
		else {
			bp = getblk(vp, blkp->bi_lbn, bsize, 0, 0);
			if (!(bp->b_flags & (B_DELWRI | B_DONE | B_CACHE)) &&
			    (error = copyin(blkp->bi_bp, bp->b_data,
			    bsize)))
				goto err2;
			if ((error = VOP_BWRITE(bp)) != 0);
				goto err2;
		}
		while (lfs_gatherblock(sp, bp, NULL));
	}
	if (sp->vp) {
		if (sp->fip->fi_nblocks == 0) {
			DEC_FINFO(sp);
			sp->sum_bytes_left +=
			    sizeof(FINFO) - sizeof(ufs_daddr_t);
		} else
			lfs_updatemeta(sp);

		lfs_writeinode(fs, sp, ip);
		lfs_vunref(vp);
	}
	(void) lfs_writeseg(fs, sp);
	lfs_segunlock(fs);
	free(start, M_SEGMENT);
	return (error);

/*
 * XXX
 * If we come in to error 2, we might have indirect blocks that were
 * updated and now have bad block pointers.  I don't know what to do
 * about this.
 */

err2:	lfs_vunref(vp);
	/* Free up fakebuffers */
	for (bpp = --sp->cbpp; bpp >= sp->bpp; --bpp)
		if ((*bpp)->b_flags & B_CALL) {
			brelvp(*bpp);
			free(*bpp, M_SEGMENT);
		} else
			brelse(*bpp);
	lfs_segunlock(fs);
err1:	
	free(start, M_SEGMENT);
	return (error);
}

/*
 * lfs_bmapv:
 *
 * This will fill in the current disk address for arrays of blocks.
 *
 *  0 on success
 * -1/errno is return on error.
 */
int
lfs_bmapv(p, v, retval)
	struct proc *p;
	void *v;
	register_t *retval;
{
	struct lfs_bmapv_args /* {
		syscallarg(fsid_t *) fsidp;
		syscallarg(struct block_info *) blkiov;
		syscallarg(int) blkcnt;
	} */ *uap = v;
	BLOCK_INFO *blkp;
	struct mount *mntp;
	struct ufsmount *ump;
	struct vnode *vp;
	fsid_t fsid;
	void *start;
	ufs_daddr_t daddr;
	int cnt, error, step;

	if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
		return (error);

	error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t));
	if (error)
		return (error);
	if ((mntp = getvfs(&fsid)) == NULL)
		return (EINVAL);

	cnt = SCARG(uap, blkcnt);
	start = blkp = malloc(cnt * sizeof(BLOCK_INFO), M_SEGMENT, M_WAITOK);
	error = copyin(SCARG(uap, blkiov), blkp, cnt * sizeof(BLOCK_INFO));
	if (error) {
		free(blkp, M_SEGMENT);
		return (error);
	}

	for (step = cnt; step--; ++blkp) {
		if (blkp->bi_lbn == LFS_UNUSED_LBN)
			continue;
		/*
		 * A regular call to VFS_VGET could deadlock
		 * here.  Instead, we try an unlocked access.
		 */
		ump = VFSTOUFS(mntp);
		if ((vp =
		    ufs_ihashlookup(ump->um_dev, blkp->bi_inode)) != NULL) {
			if (VOP_BMAP(vp, blkp->bi_lbn, NULL, &daddr, NULL))
				daddr = LFS_UNUSED_DADDR;
		} else if (VFS_VGET(mntp, blkp->bi_inode, &vp))
			daddr = LFS_UNUSED_DADDR;
		else  {
			if (VOP_BMAP(vp, blkp->bi_lbn, NULL, &daddr, NULL))
				daddr = LFS_UNUSED_DADDR;
			vput(vp);
		}
		blkp->bi_daddr = daddr;
        }
	copyout(start, SCARG(uap, blkiov), cnt * sizeof(BLOCK_INFO));
	free(start, M_SEGMENT);
	return (0);
}

/*
 * lfs_segclean:
 *
 * Mark the segment clean.
 *
 *  0 on success
 * -1/errno is return on error.
 */
int
lfs_segclean(p, v, retval)
	struct proc *p;
	void *v;
	register_t *retval;
{
	struct lfs_segclean_args /* {
		syscallarg(fsid_t *) fsidp;
		syscallarg(u_long) segment;
	} */ *uap = v;
	CLEANERINFO *cip;
	SEGUSE *sup;
	struct buf *bp;
	struct mount *mntp;
	struct lfs *fs;
	fsid_t fsid;
	int error;

	if ((error = suser(p->p_ucred, &p->p_acflag)) != 0)
		return (error);

	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
		return (error);
	if ((mntp = getvfs(&fsid)) == NULL)
		return (EINVAL);

	fs = VFSTOUFS(mntp)->um_lfs;

	if (datosn(fs, fs->lfs_curseg) == SCARG(uap, segment))
		return (EBUSY);

	LFS_SEGENTRY(sup, fs, SCARG(uap, segment), bp);
	if (sup->su_flags & SEGUSE_ACTIVE) {
		brelse(bp);
		return (EBUSY);
	}
	fs->lfs_avail += fsbtodb(fs, fs->lfs_ssize) - 1;
	fs->lfs_bfree += (sup->su_nsums * LFS_SUMMARY_SIZE / DEV_BSIZE) +
	    sup->su_ninos * btodb(fs->lfs_bsize);
	sup->su_flags &= ~SEGUSE_DIRTY;
	(void) VOP_BWRITE(bp);

	LFS_CLEANERINFO(cip, fs, bp);
	++cip->clean;
	--cip->dirty;
	(void) VOP_BWRITE(bp);
	wakeup(&fs->lfs_avail);
	return (0);
}

/*
 * lfs_segwait:
 *
 * This will block until a segment in file system fsid is written.  A timeout
 * in milliseconds may be specified which will awake the cleaner automatically.
 * An fsid of -1 means any file system, and a timeout of 0 means forever.
 *
 *  0 on success
 *  1 on timeout
 * -1/errno is return on error.
 */
int
lfs_segwait(p, v, retval)
	struct proc *p;
	void *v;
	register_t *retval;
{
	struct lfs_segwait_args /* {
		syscallarg(fsid_t *) fsidp;
		syscallarg(struct timeval *) tv;
	} */ *uap = v;
	extern int lfs_allclean_wakeup;
	struct mount *mntp;
	struct timeval atv;
	fsid_t fsid;
	void *addr;
	u_long timeout;
	int error, s;

	if ((error = suser(p->p_ucred, &p->p_acflag)) != 0) {
		return (error);
	}
#ifdef WHEN_QUADS_WORK
	if (error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t)))
		return (error);
	if (fsid == (fsid_t)-1)
		addr = &lfs_allclean_wakeup;
	else {
		if ((mntp = getvfs(&fsid)) == NULL)
			return (EINVAL);
		addr = &VFSTOUFS(mntp)->um_lfs->lfs_nextseg;
	}
#else
	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
		return (error);
	if ((mntp = getvfs(&fsid)) == NULL)
		addr = &lfs_allclean_wakeup;
	else
		addr = &VFSTOUFS(mntp)->um_lfs->lfs_nextseg;
#endif

	if (SCARG(uap, tv)) {
		error = copyin(SCARG(uap, tv), &atv, sizeof(struct timeval));
		if (error)
			return (error);
		if (itimerfix(&atv))
			return (EINVAL);
		s = splclock();
		timeradd(&atv, &time, &atv);
		timeout = hzto(&atv);
		splx(s);
	} else
		timeout = 0;

	error = tsleep(addr, PCATCH | PUSER, "segment", timeout);
	return (error == ERESTART ? EINTR : 0);
}

/*
 * VFS_VGET call specialized for the cleaner.  The cleaner already knows the
 * daddr from the ifile, so don't look it up again.  If the cleaner is
 * processing IINFO structures, it may have the ondisk inode already, so
 * don't go retrieving it again.
 */
int
lfs_fastvget(mp, ino, daddr, vpp, dinp)
	struct mount *mp;
	ino_t ino;
	ufs_daddr_t daddr;
	struct vnode **vpp;
	struct dinode *dinp;
{
	register struct inode *ip;
	struct vnode *vp;
	struct ufsmount *ump;
	struct buf *bp;
	dev_t dev;
	int error;

	ump = VFSTOUFS(mp);
	dev = ump->um_dev;
	/*
	 * This is playing fast and loose.  Someone may have the inode
	 * locked, in which case they are going to be distinctly unhappy
	 * if we trash something.
	 */
	if ((*vpp = ufs_ihashlookup(dev, ino)) != NULL) {
		lfs_vref(*vpp);
		if ((*vpp)->v_flag & VXLOCK)
			clean_vnlocked++;
		ip = VTOI(*vpp);
		if (ip->i_flag & IN_LOCKED)
			clean_inlocked++;
		if (!(ip->i_flag & IN_MODIFIED))
			++ump->um_lfs->lfs_uinodes;
		ip->i_flag |= IN_MODIFIED;
		return (0);
	}

	/* Allocate new vnode/inode. */
	if ((error = lfs_vcreate(mp, ino, &vp)) != 0) {
		*vpp = NULL;
		return (error);
	}

	/*
	 * Put it onto its hash chain and lock it so that other requests for
	 * this inode will block if they arrive while we are sleeping waiting
	 * for old data structures to be purged or for the contents of the
	 * disk portion of this inode to be read.
	 */
	ip = VTOI(vp);
	ufs_ihashins(ip);

	/*
	 * XXX
	 * This may not need to be here, logically it should go down with
	 * the i_devvp initialization.
	 * Ask Kirk.
	 */
	ip->i_lfs = ump->um_lfs;

	/* Read in the disk contents for the inode, copy into the inode. */
	if (dinp) {
		error = copyin(dinp, &ip->i_din.ffs_din, sizeof(struct dinode));
		if (error)
			return (error);
	}
	else {
		error = bread(ump->um_devvp, daddr,
			      (int)ump->um_lfs->lfs_bsize, NOCRED, &bp);
		if (error) {
			/*
			 * The inode does not contain anything useful, so it
			 * would be misleading to leave it on its hash chain.
			 * Iput() will return it to the free list.
			 */
			ufs_ihashrem(ip);

			/* Unlock and discard unneeded inode. */
			lfs_vunref(vp);
			brelse(bp);
			*vpp = NULL;
			return (error);
		}
		ip->i_din.ffs_din =
		    *lfs_ifind(ump->um_lfs, ino, (struct dinode *)bp->b_data);
		brelse(bp);
	}

	/*
	 * Initialize the vnode from the inode, check for aliases.  In all
	 * cases re-init ip, the underlying vnode/inode may have changed.
	 */
	error = ufs_vinit(mp, lfs_specop_p, LFS_FIFOOPS, &vp);
	if (error) {
		lfs_vunref(vp);
		*vpp = NULL;
		return (error);
	}
	/*
	 * Finish inode initialization now that aliasing has been resolved.
	 */
	ip->i_devvp = ump->um_devvp;
	ip->i_flag |= IN_MODIFIED;
	++ump->um_lfs->lfs_uinodes;
	VREF(ip->i_devvp);
	*vpp = vp;
	return (0);
}
struct buf *
lfs_fakebuf(vp, lbn, size, uaddr)
	struct vnode *vp;
	int lbn;
	size_t size;
	caddr_t uaddr;
{
	struct buf *bp;

	bp = lfs_newbuf(vp, lbn, 0);
	bp->b_saveaddr = uaddr;
	bp->b_bufsize = size;
	bp->b_bcount = size;
	bp->b_flags |= B_INVAL;
	return (bp);
}