lfs_vfsops.c revision 1.30 1 /* $NetBSD: lfs_vfsops.c,v 1.30 1999/04/11 23:31:09 perseant Exp $ */
2
3 /*-
4 * Copyright (c) 1999 The NetBSD Foundation, Inc.
5 * All rights reserved.
6 *
7 * This code is derived from software contributed to The NetBSD Foundation
8 * by Konrad E. Schroder <perseant (at) hhhh.org>.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 * 3. All advertising materials mentioning features or use of this software
19 * must display the following acknowledgement:
20 * This product includes software developed by the NetBSD
21 * Foundation, Inc. and its contributors.
22 * 4. Neither the name of The NetBSD Foundation nor the names of its
23 * contributors may be used to endorse or promote products derived
24 * from this software without specific prior written permission.
25 *
26 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
27 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
28 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
29 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
30 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
31 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
32 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
33 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
34 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
36 * POSSIBILITY OF SUCH DAMAGE.
37 */
38 /*-
39 * Copyright (c) 1989, 1991, 1993, 1994
40 * The Regents of the University of California. All rights reserved.
41 *
42 * Redistribution and use in source and binary forms, with or without
43 * modification, are permitted provided that the following conditions
44 * are met:
45 * 1. Redistributions of source code must retain the above copyright
46 * notice, this list of conditions and the following disclaimer.
47 * 2. Redistributions in binary form must reproduce the above copyright
48 * notice, this list of conditions and the following disclaimer in the
49 * documentation and/or other materials provided with the distribution.
50 * 3. All advertising materials mentioning features or use of this software
51 * must display the following acknowledgement:
52 * This product includes software developed by the University of
53 * California, Berkeley and its contributors.
54 * 4. Neither the name of the University nor the names of its contributors
55 * may be used to endorse or promote products derived from this software
56 * without specific prior written permission.
57 *
58 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
59 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
60 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
61 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
62 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
63 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
64 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
65 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
66 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
67 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
68 * SUCH DAMAGE.
69 *
70 * @(#)lfs_vfsops.c 8.20 (Berkeley) 6/10/95
71 */
72
73 #if defined(_KERNEL) && !defined(_LKM)
74 #include "opt_quota.h"
75 #endif
76
77 #include <sys/param.h>
78 #include <sys/systm.h>
79 #include <sys/namei.h>
80 #include <sys/proc.h>
81 #include <sys/kernel.h>
82 #include <sys/vnode.h>
83 #include <sys/mount.h>
84 #include <sys/buf.h>
85 #include <sys/mbuf.h>
86 #include <sys/file.h>
87 #include <sys/disklabel.h>
88 #include <sys/ioctl.h>
89 #include <sys/errno.h>
90 #include <sys/malloc.h>
91 #include <sys/pool.h>
92 #include <sys/socket.h>
93 #include <vm/vm.h>
94 #include <sys/sysctl.h>
95
96 #include <miscfs/specfs/specdev.h>
97
98 #include <ufs/ufs/quota.h>
99 #include <ufs/ufs/inode.h>
100 #include <ufs/ufs/ufsmount.h>
101 #include <ufs/ufs/ufs_extern.h>
102
103 #include <ufs/lfs/lfs.h>
104 #include <ufs/lfs/lfs_extern.h>
105
106 int lfs_mountfs __P((struct vnode *, struct mount *, struct proc *));
107
108 extern struct vnodeopv_desc lfs_vnodeop_opv_desc;
109 extern struct vnodeopv_desc lfs_specop_opv_desc;
110 extern struct vnodeopv_desc lfs_fifoop_opv_desc;
111
112 struct vnodeopv_desc *lfs_vnodeopv_descs[] = {
113 &lfs_vnodeop_opv_desc,
114 &lfs_specop_opv_desc,
115 &lfs_fifoop_opv_desc,
116 NULL,
117 };
118
119 struct vfsops lfs_vfsops = {
120 MOUNT_LFS,
121 lfs_mount,
122 ufs_start,
123 lfs_unmount,
124 ufs_root,
125 ufs_quotactl,
126 lfs_statfs,
127 lfs_sync,
128 lfs_vget,
129 lfs_fhtovp,
130 lfs_vptofh,
131 lfs_init,
132 lfs_sysctl,
133 NULL,
134 ufs_check_export,
135 lfs_vnodeopv_descs,
136 };
137
138 struct pool lfs_inode_pool;
139
140 /*
141 * Initialize the filesystem, most work done by ufs_init.
142 */
143 void
144 lfs_init()
145 {
146 ufs_init();
147
148 /*
149 * XXX Same structure as FFS inodes? Should we share a common pool?
150 */
151 pool_init(&lfs_inode_pool, sizeof(struct inode), 0, 0, 0,
152 "lfsinopl", 0, pool_page_alloc_nointr, pool_page_free_nointr,
153 M_LFSNODE);
154 }
155
156 /*
157 * Called by main() when ufs is going to be mounted as root.
158 */
159 int
160 lfs_mountroot()
161 {
162 extern struct vnode *rootvp;
163 struct mount *mp;
164 struct proc *p = curproc; /* XXX */
165 int error;
166
167 /*
168 * Get vnodes for swapdev and rootdev.
169 */
170 if ((error = bdevvp(rootdev, &rootvp))) {
171 printf("lfs_mountroot: can't setup bdevvp's");
172 return (error);
173 }
174 if ((error = vfs_rootmountalloc(MOUNT_LFS, "root_device", &mp)))
175 return (error);
176 if ((error = lfs_mountfs(rootvp, mp, p))) {
177 mp->mnt_op->vfs_refcount--;
178 vfs_unbusy(mp);
179 free(mp, M_MOUNT);
180 return (error);
181 }
182 simple_lock(&mountlist_slock);
183 CIRCLEQ_INSERT_TAIL(&mountlist, mp, mnt_list);
184 simple_unlock(&mountlist_slock);
185 (void)lfs_statfs(mp, &mp->mnt_stat, p);
186 vfs_unbusy(mp);
187 return (0);
188 }
189
190 /*
191 * VFS Operations.
192 *
193 * mount system call
194 */
195 int
196 lfs_mount(mp, path, data, ndp, p)
197 register struct mount *mp;
198 const char *path;
199 void *data;
200 struct nameidata *ndp;
201 struct proc *p;
202 {
203 struct vnode *devvp;
204 struct ufs_args args;
205 struct ufsmount *ump = NULL;
206 register struct lfs *fs = NULL; /* LFS */
207 size_t size;
208 int error;
209 mode_t accessmode;
210
211 error = copyin(data, (caddr_t)&args, sizeof (struct ufs_args));
212 if (error)
213 return (error);
214
215 #if 0
216 /* Until LFS can do NFS right. XXX */
217 if (args.export.ex_flags & MNT_EXPORTED)
218 return (EINVAL);
219 #endif
220
221 /*
222 * If updating, check whether changing from read-only to
223 * read/write; if there is no device name, that's all we do.
224 */
225 if (mp->mnt_flag & MNT_UPDATE) {
226 ump = VFSTOUFS(mp);
227 fs = ump->um_lfs;
228 if (fs->lfs_ronly && (mp->mnt_flag & MNT_WANTRDWR)) {
229 /*
230 * If upgrade to read-write by non-root, then verify
231 * that user has necessary permissions on the device.
232 */
233 if (p->p_ucred->cr_uid != 0) {
234 vn_lock(ump->um_devvp, LK_EXCLUSIVE | LK_RETRY);
235 error = VOP_ACCESS(ump->um_devvp, VREAD|VWRITE,
236 p->p_ucred, p);
237 VOP_UNLOCK(ump->um_devvp, 0);
238 if (error)
239 return (error);
240 }
241 fs->lfs_ronly = 0;
242 }
243 if (args.fspec == 0) {
244 /*
245 * Process export requests.
246 */
247 return (vfs_export(mp, &ump->um_export, &args.export));
248 }
249 }
250 /*
251 * Not an update, or updating the name: look up the name
252 * and verify that it refers to a sensible block device.
253 */
254 NDINIT(ndp, LOOKUP, FOLLOW, UIO_USERSPACE, args.fspec, p);
255 if ((error = namei(ndp)) != 0)
256 return (error);
257 devvp = ndp->ni_vp;
258 if (devvp->v_type != VBLK) {
259 vrele(devvp);
260 return (ENOTBLK);
261 }
262 if (major(devvp->v_rdev) >= nblkdev) {
263 vrele(devvp);
264 return (ENXIO);
265 }
266 /*
267 * If mount by non-root, then verify that user has necessary
268 * permissions on the device.
269 */
270 if (p->p_ucred->cr_uid != 0) {
271 accessmode = VREAD;
272 if ((mp->mnt_flag & MNT_RDONLY) == 0)
273 accessmode |= VWRITE;
274 vn_lock(devvp, LK_EXCLUSIVE | LK_RETRY);
275 error = VOP_ACCESS(devvp, accessmode, p->p_ucred, p);
276 if (error) {
277 vput(devvp);
278 return (error);
279 }
280 VOP_UNLOCK(devvp, 0);
281 }
282 if ((mp->mnt_flag & MNT_UPDATE) == 0)
283 error = lfs_mountfs(devvp, mp, p); /* LFS */
284 else {
285 if (devvp != ump->um_devvp)
286 error = EINVAL; /* needs translation */
287 else
288 vrele(devvp);
289 }
290 if (error) {
291 vrele(devvp);
292 return (error);
293 }
294 ump = VFSTOUFS(mp);
295 fs = ump->um_lfs; /* LFS */
296 (void)copyinstr(path, fs->lfs_fsmnt, sizeof(fs->lfs_fsmnt) - 1, &size);
297 bzero(fs->lfs_fsmnt + size, sizeof(fs->lfs_fsmnt) - size);
298 bcopy(fs->lfs_fsmnt, mp->mnt_stat.f_mntonname, MNAMELEN);
299 (void) copyinstr(args.fspec, mp->mnt_stat.f_mntfromname, MNAMELEN - 1,
300 &size);
301 bzero(mp->mnt_stat.f_mntfromname + size, MNAMELEN - size);
302 return (0);
303 }
304
305 /*
306 * Common code for mount and mountroot
307 * LFS specific
308 */
309 int
310 lfs_mountfs(devvp, mp, p)
311 register struct vnode *devvp;
312 struct mount *mp;
313 struct proc *p;
314 {
315 extern struct vnode *rootvp;
316 struct dlfs *dfs, *adfs;
317 register struct lfs *fs;
318 register struct ufsmount *ump;
319 struct vnode *vp;
320 struct buf *bp, *abp;
321 struct partinfo dpart;
322 dev_t dev;
323 int error, i, ronly, size;
324 struct ucred *cred;
325 SEGUSE *sup;
326
327 cred = p ? p->p_ucred : NOCRED;
328 /*
329 * Disallow multiple mounts of the same device.
330 * Disallow mounting of a device that is currently in use
331 * (except for root, which might share swap device for miniroot).
332 * Flush out any old buffers remaining from a previous use.
333 */
334 if ((error = vfs_mountedon(devvp)) != 0)
335 return (error);
336 if (vcount(devvp) > 1 && devvp != rootvp)
337 return (EBUSY);
338 if ((error = vinvalbuf(devvp, V_SAVE, cred, p, 0, 0)) != 0)
339 return (error);
340
341 ronly = (mp->mnt_flag & MNT_RDONLY) != 0;
342 error = VOP_OPEN(devvp, ronly ? FREAD : FREAD|FWRITE, FSCRED, p);
343 if (error)
344 return (error);
345 if (VOP_IOCTL(devvp, DIOCGPART, (caddr_t)&dpart, FREAD, cred, p) != 0)
346 size = DEV_BSIZE;
347 else
348 size = dpart.disklab->d_secsize;
349
350 /* Don't free random space on error. */
351 bp = NULL;
352 ump = NULL;
353
354 /* Read in the superblock. */
355 error = bread(devvp, LFS_LABELPAD / size, LFS_SBPAD, cred, &bp);
356 if (error)
357 goto out;
358 dfs = (struct dlfs *)bp->b_data;
359
360 /*
361 * Check the second superblock to see which is newer; then mount
362 * using the older of the two. This is necessary to ensure that
363 * the filesystem is valid if it was not unmounted cleanly.
364 */
365 error = bread(devvp, dfs->dlfs_sboffs[1], LFS_SBPAD, cred, &abp);
366 if (error)
367 goto out;
368 adfs = (struct dlfs *)abp->b_data;
369
370 if (adfs->dlfs_tstamp < dfs->dlfs_tstamp) /* XXX KS - 1s resolution? */
371 dfs = adfs;
372
373 /* Check the basics. */
374 if (dfs->dlfs_magic != LFS_MAGIC || dfs->dlfs_bsize > MAXBSIZE ||
375 dfs->dlfs_bsize < sizeof(struct dlfs)) {
376 error = EINVAL; /* XXX needs translation */
377 goto out;
378 }
379
380 /* Allocate the mount structure, copy the superblock into it. */
381 fs = malloc(sizeof(struct lfs), M_UFSMNT, M_WAITOK);
382 memcpy(&fs->lfs_dlfs, dfs, sizeof(struct dlfs));
383 ump = malloc(sizeof *ump, M_UFSMNT, M_WAITOK);
384 memset((caddr_t)ump, 0, sizeof *ump);
385 ump->um_lfs = fs;
386 if (sizeof(struct lfs) < LFS_SBPAD) /* XXX why? */
387 bp->b_flags |= B_INVAL;
388 brelse(bp);
389 bp = NULL;
390 brelse(abp);
391 abp = NULL;
392
393 /* Set up the I/O information */
394 fs->lfs_iocount = 0;
395 fs->lfs_dirvcount = 0;
396 fs->lfs_activesb = 0;
397 #ifdef LFS_CANNOT_ROLLFW
398 fs->lfs_sbactive = NULL;
399 #endif
400 #ifdef LFS_TRACK_IOS
401 for (i=0;i<LFS_THROTTLE;i++)
402 fs->lfs_pending[i] = LFS_UNUSED_DADDR;
403 #endif
404
405 /* Set up the ifile and lock aflags */
406 fs->lfs_doifile = 0;
407 fs->lfs_writer = 0;
408 fs->lfs_dirops = 0;
409 fs->lfs_seglock = 0;
410
411 /* Set the file system readonly/modify bits. */
412 fs->lfs_ronly = ronly;
413 if (ronly == 0)
414 fs->lfs_fmod = 1;
415
416 /* Initialize the mount structure. */
417 dev = devvp->v_rdev;
418 mp->mnt_data = (qaddr_t)ump;
419 mp->mnt_stat.f_fsid.val[0] = (long)dev;
420 mp->mnt_stat.f_fsid.val[1] = makefstype(MOUNT_LFS);
421 mp->mnt_maxsymlinklen = fs->lfs_maxsymlinklen;
422 mp->mnt_flag |= MNT_LOCAL;
423 ump->um_flags = 0;
424 ump->um_mountp = mp;
425 ump->um_dev = dev;
426 ump->um_devvp = devvp;
427 ump->um_bptrtodb = 0;
428 ump->um_seqinc = 1 << fs->lfs_fsbtodb;
429 ump->um_nindir = fs->lfs_nindir;
430 for (i = 0; i < MAXQUOTAS; i++)
431 ump->um_quotas[i] = NULLVP;
432 devvp->v_specflags |= SI_MOUNTEDON;
433
434 /*
435 * We use the ifile vnode for almost every operation. Instead of
436 * retrieving it from the hash table each time we retrieve it here,
437 * artificially increment the reference count and keep a pointer
438 * to it in the incore copy of the superblock.
439 */
440 if ((error = VFS_VGET(mp, LFS_IFILE_INUM, &vp)) != 0)
441 goto out;
442 fs->lfs_ivnode = vp;
443 VREF(vp);
444 vput(vp);
445
446 /*
447 * Mark the current segment as ACTIVE, since we're going to
448 * be writing to it.
449 */
450 LFS_SEGENTRY(sup, fs, datosn(fs, fs->lfs_offset), bp);
451 sup->su_flags |= SEGUSE_DIRTY | SEGUSE_ACTIVE;
452 (void) VOP_BWRITE(bp);
453
454 return (0);
455 out:
456 if (bp)
457 brelse(bp);
458 if (abp)
459 brelse(abp);
460 (void)VOP_CLOSE(devvp, ronly ? FREAD : FREAD|FWRITE, cred, p);
461 if (ump) {
462 free(ump->um_lfs, M_UFSMNT);
463 free(ump, M_UFSMNT);
464 mp->mnt_data = (qaddr_t)0;
465 }
466 return (error);
467 }
468
469 /*
470 * unmount system call
471 */
472 int
473 lfs_unmount(mp, mntflags, p)
474 struct mount *mp;
475 int mntflags;
476 struct proc *p;
477 {
478 register struct ufsmount *ump;
479 register struct lfs *fs;
480 int error, flags, ronly;
481 extern int lfs_allclean_wakeup;
482
483 flags = 0;
484 if (mntflags & MNT_FORCE)
485 flags |= FORCECLOSE;
486
487 ump = VFSTOUFS(mp);
488 fs = ump->um_lfs;
489 #ifdef QUOTA
490 if (mp->mnt_flag & MNT_QUOTA) {
491 int i;
492 error = vflush(mp, fs->lfs_ivnode, SKIPSYSTEM|flags);
493 if (error)
494 return (error);
495 for (i = 0; i < MAXQUOTAS; i++) {
496 if (ump->um_quotas[i] == NULLVP)
497 continue;
498 quotaoff(p, mp, i);
499 }
500 /*
501 * Here we fall through to vflush again to ensure
502 * that we have gotten rid of all the system vnodes.
503 */
504 }
505 #endif
506 if ((error = vflush(mp, fs->lfs_ivnode, flags)) != 0)
507 return (error);
508 fs->lfs_clean = 1;
509 if ((error = VFS_SYNC(mp, 1, p->p_ucred, p)) != 0)
510 return (error);
511 if (fs->lfs_ivnode->v_dirtyblkhd.lh_first)
512 panic("lfs_unmount: still dirty blocks on ifile vnode\n");
513 vrele(fs->lfs_ivnode);
514 vgone(fs->lfs_ivnode);
515
516 ronly = !fs->lfs_ronly;
517 ump->um_devvp->v_specflags &= ~SI_MOUNTEDON;
518 error = VOP_CLOSE(ump->um_devvp,
519 ronly ? FREAD : FREAD|FWRITE, NOCRED, p);
520 vrele(ump->um_devvp);
521
522 /* XXX KS - wake up the cleaner so it can die */
523 wakeup(&fs->lfs_nextseg);
524 wakeup(&lfs_allclean_wakeup);
525
526 free(fs, M_UFSMNT);
527 free(ump, M_UFSMNT);
528 mp->mnt_data = (qaddr_t)0;
529 mp->mnt_flag &= ~MNT_LOCAL;
530 return (error);
531 }
532
533 /*
534 * Get file system statistics.
535 */
536 int
537 lfs_statfs(mp, sbp, p)
538 struct mount *mp;
539 register struct statfs *sbp;
540 struct proc *p;
541 {
542 register struct lfs *fs;
543 register struct ufsmount *ump;
544
545 ump = VFSTOUFS(mp);
546 fs = ump->um_lfs;
547 if (fs->lfs_magic != LFS_MAGIC)
548 panic("lfs_statfs: magic");
549 sbp->f_type = 0;
550 sbp->f_bsize = fs->lfs_fsize;
551 sbp->f_iosize = fs->lfs_bsize;
552 sbp->f_blocks = dbtofrags(fs, fs->lfs_dsize);
553 sbp->f_bfree = dbtofrags(fs, fs->lfs_bfree);
554 /*
555 * To compute the available space. Subtract the minimum free
556 * from the total number of blocks in the file system. Set avail
557 * to the smaller of this number and fs->lfs_bfree.
558 *
559 * XXX KS - is my modification below what is desired? (This
560 * will, e.g., change the report when the cleaner runs.)
561 */
562 #if 0
563 sbp->f_bavail = (long) ((u_int64_t) fs->lfs_dsize * (u_int64_t)
564 (100 - fs->lfs_minfree) / (u_int64_t) 100);
565 sbp->f_bavail =
566 sbp->f_bavail > fs->lfs_bfree ? fs->lfs_bfree : sbp->f_bavail;
567 #else
568 sbp->f_bavail = (long) ((u_int64_t) fs->lfs_dsize * (u_int64_t)
569 (100 - fs->lfs_minfree) / (u_int64_t) 100)
570 - (u_int64_t)(fs->lfs_dsize - fs->lfs_bfree);
571 #endif
572 sbp->f_bavail = dbtofrags(fs, sbp->f_bavail);
573 sbp->f_files = fs->lfs_nfiles;
574 sbp->f_ffree = sbp->f_bfree * INOPB(fs);
575 if (sbp != &mp->mnt_stat) {
576 bcopy(mp->mnt_stat.f_mntonname, sbp->f_mntonname, MNAMELEN);
577 bcopy(mp->mnt_stat.f_mntfromname, sbp->f_mntfromname, MNAMELEN);
578 }
579 strncpy(sbp->f_fstypename, mp->mnt_op->vfs_name, MFSNAMELEN);
580 return (0);
581 }
582
583 /*
584 * Go through the disk queues to initiate sandbagged IO;
585 * go through the inodes to write those that have been modified;
586 * initiate the writing of the super block if it has been modified.
587 *
588 * Note: we are always called with the filesystem marked `MPBUSY'.
589 */
590 int
591 lfs_sync(mp, waitfor, cred, p)
592 struct mount *mp;
593 int waitfor;
594 struct ucred *cred;
595 struct proc *p;
596 {
597 int error;
598 struct lfs *fs;
599
600 fs = ((struct ufsmount *)mp->mnt_data)->ufsmount_u.lfs;
601 while(fs->lfs_dirops)
602 error = tsleep(&fs->lfs_dirops, PRIBIO + 1, "lfs_dirops", 0);
603 fs->lfs_writer++;
604
605 /* All syncs must be checkpoints until roll-forward is implemented. */
606 error = lfs_segwrite(mp, SEGM_CKP | (waitfor ? SEGM_SYNC : 0));
607 if(--fs->lfs_writer==0)
608 wakeup(&fs->lfs_dirops);
609 #ifdef QUOTA
610 qsync(mp);
611 #endif
612 return (error);
613 }
614
615 #ifdef USE_UFS_HASHLOCK
616 extern struct lock ufs_hashlock;
617 #endif
618
619 /*
620 * Look up an LFS dinode number to find its incore vnode. If not already
621 * in core, read it in from the specified device. Return the inode locked.
622 * Detection and handling of mount points must be done by the calling routine.
623 */
624 int
625 lfs_vget(mp, ino, vpp)
626 struct mount *mp;
627 ino_t ino;
628 struct vnode **vpp;
629 {
630 register struct lfs *fs;
631 register struct inode *ip;
632 struct buf *bp;
633 struct ifile *ifp;
634 struct vnode *vp;
635 struct ufsmount *ump;
636 ufs_daddr_t daddr;
637 dev_t dev;
638 int error;
639 #ifdef LFS_ATIME_IFILE
640 struct timespec ts;
641 #endif
642
643 ump = VFSTOUFS(mp);
644 dev = ump->um_dev;
645
646 #ifdef USE_UFS_HASHLOCK
647 do {
648 #endif
649 if ((*vpp = ufs_ihashget(dev, ino)) != NULL)
650 return (0);
651 #ifdef USE_UFS_HASHLOCK
652 } while (lockmgr(&ufs_hashlock, LK_EXCLUSIVE|LK_SLEEPFAIL, 0));
653 #endif
654
655 /* Translate the inode number to a disk address. */
656 fs = ump->um_lfs;
657 if (ino == LFS_IFILE_INUM)
658 daddr = fs->lfs_idaddr;
659 else {
660 LFS_IENTRY(ifp, fs, ino, bp);
661 daddr = ifp->if_daddr;
662 #ifdef LFS_ATIME_IFILE
663 ts = ifp->if_atime; /* structure copy */
664 #endif
665 brelse(bp);
666 if (daddr == LFS_UNUSED_DADDR) {
667 #ifdef USE_UFS_HASHLOCK
668 lockmgr(&ufs_hashlock, LK_RELEASE, 0);
669 #endif
670 return (ENOENT);
671 }
672 }
673
674 /* Allocate new vnode/inode. */
675 if ((error = lfs_vcreate(mp, ino, &vp)) != 0) {
676 *vpp = NULL;
677 #ifdef USE_UFS_HASHLOCK
678 lockmgr(&ufs_hashlock, LK_RELEASE, 0);
679 #endif
680 return (error);
681 }
682
683 /*
684 * Put it onto its hash chain and lock it so that other requests for
685 * this inode will block if they arrive while we are sleeping waiting
686 * for old data structures to be purged or for the contents of the
687 * disk portion of this inode to be read.
688 */
689 ip = VTOI(vp);
690 ufs_ihashins(ip);
691 #ifdef USE_UFS_HASHLOCK
692 lockmgr(&ufs_hashlock, LK_RELEASE, 0);
693 #endif
694
695 /*
696 * XXX
697 * This may not need to be here, logically it should go down with
698 * the i_devvp initialization.
699 * Ask Kirk.
700 */
701 ip->i_lfs = ump->um_lfs;
702
703 /* Read in the disk contents for the inode, copy into the inode. */
704 error = bread(ump->um_devvp, daddr, (int)fs->lfs_bsize, NOCRED, &bp);
705 if (error) {
706 /*
707 * The inode does not contain anything useful, so it would
708 * be misleading to leave it on its hash chain. With mode
709 * still zero, it will be unlinked and returned to the free
710 * list by vput().
711 */
712 vput(vp);
713 brelse(bp);
714 *vpp = NULL;
715 return (error);
716 }
717 ip->i_din.ffs_din = *lfs_ifind(fs, ino, (struct dinode *)bp->b_data);
718 #ifdef LFS_ATIME_IFILE
719 ip->i_ffs_atime = ts.tv_sec;
720 ip->i_ffs_atimensec = ts.tv_nsec;
721 #endif
722 brelse(bp);
723
724 /*
725 * Initialize the vnode from the inode, check for aliases. In all
726 * cases re-init ip, the underlying vnode/inode may have changed.
727 */
728 error = ufs_vinit(mp, lfs_specop_p, lfs_fifoop_p, &vp);
729 if (error) {
730 vput(vp);
731 *vpp = NULL;
732 return (error);
733 }
734 if(vp->v_type == VNON) {
735 printf("lfs_vget: ino %d is type VNON! (ifmt %o)\n", ip->i_number, (ip->i_ffs_mode&IFMT)>>12);
736 #ifdef DDB
737 Debugger();
738 #endif
739 }
740 /*
741 * Finish inode initialization now that aliasing has been resolved.
742 */
743 ip->i_devvp = ump->um_devvp;
744 VREF(ip->i_devvp);
745 *vpp = vp;
746
747 return (0);
748 }
749
750 /*
751 * File handle to vnode
752 *
753 * Have to be really careful about stale file handles:
754 * - check that the inode number is valid
755 * - call lfs_vget() to get the locked inode
756 * - check for an unallocated inode (i_mode == 0)
757 *
758 * XXX
759 * use ifile to see if inode is allocated instead of reading off disk
760 * what is the relationship between my generational number and the NFS
761 * generational number.
762 */
763 int
764 lfs_fhtovp(mp, fhp, vpp)
765 register struct mount *mp;
766 struct fid *fhp;
767 struct vnode **vpp;
768 {
769 register struct ufid *ufhp;
770
771 ufhp = (struct ufid *)fhp;
772 if (ufhp->ufid_ino < ROOTINO)
773 return (ESTALE);
774 return (ufs_fhtovp(mp, ufhp, vpp));
775 }
776
777 /*
778 * Vnode pointer to File handle
779 */
780 /* ARGSUSED */
781 int
782 lfs_vptofh(vp, fhp)
783 struct vnode *vp;
784 struct fid *fhp;
785 {
786 register struct inode *ip;
787 register struct ufid *ufhp;
788
789 ip = VTOI(vp);
790 ufhp = (struct ufid *)fhp;
791 ufhp->ufid_len = sizeof(struct ufid);
792 ufhp->ufid_ino = ip->i_number;
793 ufhp->ufid_gen = ip->i_ffs_gen;
794 return (0);
795 }
796
797 int
798 lfs_sysctl(name, namelen, oldp, oldlenp, newp, newlen, p)
799 int *name;
800 u_int namelen;
801 void *oldp;
802 size_t *oldlenp;
803 void *newp;
804 size_t newlen;
805 struct proc *p;
806 {
807 extern int lfs_writeindir, lfs_dostats, lfs_clean_vnhead;
808 extern struct lfs_stats lfs_stats;
809 int error;
810
811 /* all sysctl names at this level are terminal */
812 if (namelen != 1)
813 return (ENOTDIR);
814
815 switch (name[0]) {
816 case LFS_WRITEINDIR:
817 return (sysctl_int(oldp, oldlenp, newp, newlen,
818 &lfs_writeindir));
819 case LFS_CLEAN_VNHEAD:
820 return (sysctl_int(oldp, oldlenp, newp, newlen,
821 &lfs_clean_vnhead));
822 case LFS_DOSTATS:
823 if((error = sysctl_int(oldp, oldlenp, newp, newlen,
824 &lfs_dostats)))
825 return error;
826 if(lfs_dostats == 0)
827 memset(&lfs_stats,0,sizeof(lfs_stats));
828 return 0;
829 case LFS_STATS:
830 return (sysctl_rdstruct(oldp, oldlenp, newp,
831 &lfs_stats, sizeof(lfs_stats)));
832 default:
833 return (EOPNOTSUPP);
834 }
835 /* NOTREACHED */
836 }
837