xref: /netbsd-src/sys/ufs/lfs/lfs_syscalls.c (revision 413d532bcc3f62d122e56d92e13ac64825a40baf)
1 /*	$NetBSD: lfs_syscalls.c,v 1.155 2014/04/17 18:15:49 pgoyette Exp $	*/
2 
3 /*-
4  * Copyright (c) 1999, 2000, 2001, 2002, 2003, 2007, 2007, 2008
5  *    The NetBSD Foundation, Inc.
6  * All rights reserved.
7  *
8  * This code is derived from software contributed to The NetBSD Foundation
9  * by Konrad E. Schroder <perseant@hhhh.org>.
10  *
11  * Redistribution and use in source and binary forms, with or without
12  * modification, are permitted provided that the following conditions
13  * are met:
14  * 1. Redistributions of source code must retain the above copyright
15  *    notice, this list of conditions and the following disclaimer.
16  * 2. Redistributions in binary form must reproduce the above copyright
17  *    notice, this list of conditions and the following disclaimer in the
18  *    documentation and/or other materials provided with the distribution.
19  *
20  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
21  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
22  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
23  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
24  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
25  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
26  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
27  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
28  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
29  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
30  * POSSIBILITY OF SUCH DAMAGE.
31  */
32 /*-
33  * Copyright (c) 1991, 1993, 1994
34  *	The Regents of the University of California.  All rights reserved.
35  *
36  * Redistribution and use in source and binary forms, with or without
37  * modification, are permitted provided that the following conditions
38  * are met:
39  * 1. Redistributions of source code must retain the above copyright
40  *    notice, this list of conditions and the following disclaimer.
41  * 2. Redistributions in binary form must reproduce the above copyright
42  *    notice, this list of conditions and the following disclaimer in the
43  *    documentation and/or other materials provided with the distribution.
44  * 3. Neither the name of the University nor the names of its contributors
45  *    may be used to endorse or promote products derived from this software
46  *    without specific prior written permission.
47  *
48  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
49  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
50  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
51  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
52  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
53  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
54  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
55  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
56  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
57  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
58  * SUCH DAMAGE.
59  *
60  *	@(#)lfs_syscalls.c	8.10 (Berkeley) 5/14/95
61  */
62 
63 #include <sys/cdefs.h>
64 __KERNEL_RCSID(0, "$NetBSD: lfs_syscalls.c,v 1.155 2014/04/17 18:15:49 pgoyette Exp $");
65 
66 #ifndef LFS
67 # define LFS		/* for prototypes in syscallargs.h */
68 #endif
69 
70 #include <sys/param.h>
71 #include <sys/systm.h>
72 #include <sys/proc.h>
73 #include <sys/buf.h>
74 #include <sys/mount.h>
75 #include <sys/vnode.h>
76 #include <sys/kernel.h>
77 #include <sys/kauth.h>
78 #include <sys/syscallargs.h>
79 
80 #include <ufs/lfs/ulfs_inode.h>
81 #include <ufs/lfs/ulfsmount.h>
82 #include <ufs/lfs/ulfs_extern.h>
83 
84 #include <ufs/lfs/lfs.h>
85 #include <ufs/lfs/lfs_kernel.h>
86 #include <ufs/lfs/lfs_extern.h>
87 
88 struct buf *lfs_fakebuf(struct lfs *, struct vnode *, int, size_t, void *);
89 int lfs_fasthashget(dev_t, ino_t, struct vnode **);
90 
91 pid_t lfs_cleaner_pid = 0;
92 
93 /*
94  * sys_lfs_markv:
95  *
96  * This will mark inodes and blocks dirty, so they are written into the log.
97  * It will block until all the blocks have been written.  The segment create
98  * time passed in the block_info and inode_info structures is used to decide
99  * if the data is valid for each block (in case some process dirtied a block
100  * or inode that is being cleaned between the determination that a block is
101  * live and the lfs_markv call).
102  *
103  *  0 on success
104  * -1/errno is return on error.
105  */
106 #ifdef USE_64BIT_SYSCALLS
107 int
108 sys_lfs_markv(struct lwp *l, const struct sys_lfs_markv_args *uap, register_t *retval)
109 {
110 	/* {
111 		syscallarg(fsid_t *) fsidp;
112 		syscallarg(struct block_info *) blkiov;
113 		syscallarg(int) blkcnt;
114 	} */
115 	BLOCK_INFO *blkiov;
116 	int blkcnt, error;
117 	fsid_t fsid;
118 	struct lfs *fs;
119 	struct mount *mntp;
120 
121 	error = kauth_authorize_system(l->l_cred, KAUTH_SYSTEM_LFS,
122 	    KAUTH_REQ_SYSTEM_LFS_MARKV, NULL, NULL, NULL);
123 	if (error)
124 		return (error);
125 
126 	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
127 		return (error);
128 
129 	if ((mntp = vfs_getvfs(fsidp)) == NULL)
130 		return (ENOENT);
131 	fs = VFSTOULFS(mntp)->um_lfs;
132 
133 	blkcnt = SCARG(uap, blkcnt);
134 	if ((u_int) blkcnt > LFS_MARKV_MAXBLKCNT)
135 		return (EINVAL);
136 
137 	KERNEL_LOCK(1, NULL);
138 	blkiov = lfs_malloc(fs, blkcnt * sizeof(BLOCK_INFO), LFS_NB_BLKIOV);
139 	if ((error = copyin(SCARG(uap, blkiov), blkiov,
140 			    blkcnt * sizeof(BLOCK_INFO))) != 0)
141 		goto out;
142 
143 	if ((error = lfs_markv(p, &fsid, blkiov, blkcnt)) == 0)
144 		copyout(blkiov, SCARG(uap, blkiov),
145 			blkcnt * sizeof(BLOCK_INFO));
146     out:
147 	lfs_free(fs, blkiov, LFS_NB_BLKIOV);
148 	KERNEL_UNLOCK_ONE(NULL);
149 	return error;
150 }
151 #else
152 int
153 sys_lfs_markv(struct lwp *l, const struct sys_lfs_markv_args *uap, register_t *retval)
154 {
155 	/* {
156 		syscallarg(fsid_t *) fsidp;
157 		syscallarg(struct block_info *) blkiov;
158 		syscallarg(int) blkcnt;
159 	} */
160 	BLOCK_INFO *blkiov;
161 	BLOCK_INFO_15 *blkiov15;
162 	int i, blkcnt, error;
163 	fsid_t fsid;
164 	struct lfs *fs;
165 	struct mount *mntp;
166 
167 	error = kauth_authorize_system(l->l_cred, KAUTH_SYSTEM_LFS,
168 	    KAUTH_REQ_SYSTEM_LFS_MARKV, NULL, NULL, NULL);
169 	if (error)
170 		return (error);
171 
172 	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
173 		return (error);
174 
175 	if ((mntp = vfs_getvfs(&fsid)) == NULL)
176 		return (ENOENT);
177 	fs = VFSTOULFS(mntp)->um_lfs;
178 
179 	blkcnt = SCARG(uap, blkcnt);
180 	if ((u_int) blkcnt > LFS_MARKV_MAXBLKCNT)
181 		return (EINVAL);
182 
183 	KERNEL_LOCK(1, NULL);
184 	blkiov = lfs_malloc(fs, blkcnt * sizeof(BLOCK_INFO), LFS_NB_BLKIOV);
185 	blkiov15 = lfs_malloc(fs, blkcnt * sizeof(BLOCK_INFO_15), LFS_NB_BLKIOV);
186 	if ((error = copyin(SCARG(uap, blkiov), blkiov15,
187 			    blkcnt * sizeof(BLOCK_INFO_15))) != 0)
188 		goto out;
189 
190 	for (i = 0; i < blkcnt; i++) {
191 		blkiov[i].bi_inode     = blkiov15[i].bi_inode;
192 		blkiov[i].bi_lbn       = blkiov15[i].bi_lbn;
193 		blkiov[i].bi_daddr     = blkiov15[i].bi_daddr;
194 		blkiov[i].bi_segcreate = blkiov15[i].bi_segcreate;
195 		blkiov[i].bi_version   = blkiov15[i].bi_version;
196 		blkiov[i].bi_bp	       = blkiov15[i].bi_bp;
197 		blkiov[i].bi_size      = blkiov15[i].bi_size;
198 	}
199 
200 	if ((error = lfs_markv(l->l_proc, &fsid, blkiov, blkcnt)) == 0) {
201 		for (i = 0; i < blkcnt; i++) {
202 			blkiov15[i].bi_inode	 = blkiov[i].bi_inode;
203 			blkiov15[i].bi_lbn	 = blkiov[i].bi_lbn;
204 			blkiov15[i].bi_daddr	 = blkiov[i].bi_daddr;
205 			blkiov15[i].bi_segcreate = blkiov[i].bi_segcreate;
206 			blkiov15[i].bi_version	 = blkiov[i].bi_version;
207 			blkiov15[i].bi_bp	 = blkiov[i].bi_bp;
208 			blkiov15[i].bi_size	 = blkiov[i].bi_size;
209 		}
210 		copyout(blkiov15, SCARG(uap, blkiov),
211 			blkcnt * sizeof(BLOCK_INFO_15));
212 	}
213     out:
214 	lfs_free(fs, blkiov, LFS_NB_BLKIOV);
215 	lfs_free(fs, blkiov15, LFS_NB_BLKIOV);
216 	KERNEL_UNLOCK_ONE(NULL);
217 	return error;
218 }
219 #endif
220 
221 #define	LFS_MARKV_MAX_BLOCKS	(LFS_MAX_BUFS)
222 
223 int
224 lfs_markv(struct proc *p, fsid_t *fsidp, BLOCK_INFO *blkiov,
225     int blkcnt)
226 {
227 	BLOCK_INFO *blkp;
228 	IFILE *ifp;
229 	struct buf *bp;
230 	struct inode *ip = NULL;
231 	struct lfs *fs;
232 	struct mount *mntp;
233 	struct vnode *vp = NULL;
234 	ino_t lastino;
235 	daddr_t b_daddr, v_daddr;
236 	int cnt, error;
237 	int do_again = 0;
238 	int numrefed = 0;
239 	ino_t maxino;
240 	size_t obsize;
241 
242 	/* number of blocks/inodes that we have already bwrite'ed */
243 	int nblkwritten, ninowritten;
244 
245 	if ((mntp = vfs_getvfs(fsidp)) == NULL)
246 		return (ENOENT);
247 
248 	fs = VFSTOULFS(mntp)->um_lfs;
249 
250 	if (fs->lfs_ronly)
251 		return EROFS;
252 
253 	maxino = (lfs_fragstoblks(fs, VTOI(fs->lfs_ivnode)->i_ffs1_blocks) -
254 		      fs->lfs_cleansz - fs->lfs_segtabsz) * fs->lfs_ifpb;
255 
256 	cnt = blkcnt;
257 
258 	if ((error = vfs_busy(mntp, NULL)) != 0)
259 		return (error);
260 
261 	/*
262 	 * This seglock is just to prevent the fact that we might have to sleep
263 	 * from allowing the possibility that our blocks might become
264 	 * invalid.
265 	 *
266 	 * It is also important to note here that unless we specify SEGM_CKP,
267 	 * any Ifile blocks that we might be asked to clean will never get
268 	 * to the disk.
269 	 */
270 	lfs_seglock(fs, SEGM_CLEAN | SEGM_CKP | SEGM_SYNC);
271 
272 	/* Mark blocks/inodes dirty.  */
273 	error = 0;
274 
275 	/* these were inside the initialization for the for loop */
276 	v_daddr = LFS_UNUSED_DADDR;
277 	lastino = LFS_UNUSED_INUM;
278 	nblkwritten = ninowritten = 0;
279 	for (blkp = blkiov; cnt--; ++blkp)
280 	{
281 		/* Bounds-check incoming data, avoid panic for failed VGET */
282 		if (blkp->bi_inode <= 0 || blkp->bi_inode >= maxino) {
283 			error = EINVAL;
284 			goto err3;
285 		}
286 		/*
287 		 * Get the IFILE entry (only once) and see if the file still
288 		 * exists.
289 		 */
290 		if (lastino != blkp->bi_inode) {
291 			/*
292 			 * Finish the old file, if there was one.  The presence
293 			 * of a usable vnode in vp is signaled by a valid v_daddr.
294 			 */
295 			if (v_daddr != LFS_UNUSED_DADDR) {
296 				lfs_vunref(vp);
297 				numrefed--;
298 			}
299 
300 			/*
301 			 * Start a new file
302 			 */
303 			lastino = blkp->bi_inode;
304 			if (blkp->bi_inode == LFS_IFILE_INUM)
305 				v_daddr = fs->lfs_idaddr;
306 			else {
307 				LFS_IENTRY(ifp, fs, blkp->bi_inode, bp);
308 				/* XXX fix for force write */
309 				v_daddr = ifp->if_daddr;
310 				brelse(bp, 0);
311 			}
312 			if (v_daddr == LFS_UNUSED_DADDR)
313 				continue;
314 
315 			/* Get the vnode/inode. */
316 			error = lfs_fastvget(mntp, blkp->bi_inode, v_daddr,
317 					   &vp,
318 					   (blkp->bi_lbn == LFS_UNUSED_LBN
319 					    ? blkp->bi_bp
320 					    : NULL));
321 
322 			if (!error) {
323 				numrefed++;
324 			}
325 			if (error) {
326 				DLOG((DLOG_CLEAN, "lfs_markv: lfs_fastvget"
327 				      " failed with %d (ino %d, segment %d)\n",
328 				      error, blkp->bi_inode,
329 				      lfs_dtosn(fs, blkp->bi_daddr)));
330 				/*
331 				 * If we got EAGAIN, that means that the
332 				 * Inode was locked.  This is
333 				 * recoverable: just clean the rest of
334 				 * this segment, and let the cleaner try
335 				 * again with another.	(When the
336 				 * cleaner runs again, this segment will
337 				 * sort high on the list, since it is
338 				 * now almost entirely empty.) But, we
339 				 * still set v_daddr = LFS_UNUSED_ADDR
340 				 * so as not to test this over and over
341 				 * again.
342 				 */
343 				if (error == EAGAIN) {
344 					error = 0;
345 					do_again++;
346 				}
347 #ifdef DIAGNOSTIC
348 				else if (error != ENOENT)
349 					panic("lfs_markv VFS_VGET FAILED");
350 #endif
351 				/* lastino = LFS_UNUSED_INUM; */
352 				v_daddr = LFS_UNUSED_DADDR;
353 				vp = NULL;
354 				ip = NULL;
355 				continue;
356 			}
357 			ip = VTOI(vp);
358 			ninowritten++;
359 		} else if (v_daddr == LFS_UNUSED_DADDR) {
360 			/*
361 			 * This can only happen if the vnode is dead (or
362 			 * in any case we can't get it...e.g., it is
363 			 * inlocked).  Keep going.
364 			 */
365 			continue;
366 		}
367 
368 		/* Past this point we are guaranteed that vp, ip are valid. */
369 
370 		/* Can't clean VU_DIROP directories in case of truncation */
371 		/* XXX - maybe we should mark removed dirs specially? */
372 		if (vp->v_type == VDIR && (vp->v_uflag & VU_DIROP)) {
373 			do_again++;
374 			continue;
375 		}
376 
377 		/* If this BLOCK_INFO didn't contain a block, keep going. */
378 		if (blkp->bi_lbn == LFS_UNUSED_LBN) {
379 			/* XXX need to make sure that the inode gets written in this case */
380 			/* XXX but only write the inode if it's the right one */
381 			if (blkp->bi_inode != LFS_IFILE_INUM) {
382 				LFS_IENTRY(ifp, fs, blkp->bi_inode, bp);
383 				if (ifp->if_daddr == blkp->bi_daddr) {
384 					mutex_enter(&lfs_lock);
385 					LFS_SET_UINO(ip, IN_CLEANING);
386 					mutex_exit(&lfs_lock);
387 				}
388 				brelse(bp, 0);
389 			}
390 			continue;
391 		}
392 
393 		b_daddr = 0;
394 		if (VOP_BMAP(vp, blkp->bi_lbn, NULL, &b_daddr, NULL) ||
395 		    LFS_DBTOFSB(fs, b_daddr) != blkp->bi_daddr)
396 		{
397 			if (lfs_dtosn(fs, LFS_DBTOFSB(fs, b_daddr)) ==
398 			    lfs_dtosn(fs, blkp->bi_daddr))
399 			{
400 				DLOG((DLOG_CLEAN, "lfs_markv: wrong da same seg: %llx vs %llx\n",
401 				      (long long)blkp->bi_daddr, (long long)LFS_DBTOFSB(fs, b_daddr)));
402 			}
403 			do_again++;
404 			continue;
405 		}
406 
407 		/*
408 		 * Check block sizes.  The blocks being cleaned come from
409 		 * disk, so they should have the same size as their on-disk
410 		 * counterparts.
411 		 */
412 		if (blkp->bi_lbn >= 0)
413 			obsize = lfs_blksize(fs, ip, blkp->bi_lbn);
414 		else
415 			obsize = fs->lfs_bsize;
416 		/* Check for fragment size change */
417 		if (blkp->bi_lbn >= 0 && blkp->bi_lbn < ULFS_NDADDR) {
418 			obsize = ip->i_lfs_fragsize[blkp->bi_lbn];
419 		}
420 		if (obsize != blkp->bi_size) {
421 			DLOG((DLOG_CLEAN, "lfs_markv: ino %d lbn %lld wrong"
422 			      " size (%ld != %d), try again\n",
423 			      blkp->bi_inode, (long long)blkp->bi_lbn,
424 			      (long) obsize, blkp->bi_size));
425 			do_again++;
426 			continue;
427 		}
428 
429 		/*
430 		 * If we get to here, then we are keeping the block.  If
431 		 * it is an indirect block, we want to actually put it
432 		 * in the buffer cache so that it can be updated in the
433 		 * finish_meta section.	 If it's not, we need to
434 		 * allocate a fake buffer so that writeseg can perform
435 		 * the copyin and write the buffer.
436 		 */
437 		if (ip->i_number != LFS_IFILE_INUM && blkp->bi_lbn >= 0) {
438 			/* Data Block */
439 			bp = lfs_fakebuf(fs, vp, blkp->bi_lbn,
440 					 blkp->bi_size, blkp->bi_bp);
441 			/* Pretend we used bread() to get it */
442 			bp->b_blkno = LFS_FSBTODB(fs, blkp->bi_daddr);
443 		} else {
444 			/* Indirect block or ifile */
445 			if (blkp->bi_size != fs->lfs_bsize &&
446 			    ip->i_number != LFS_IFILE_INUM)
447 				panic("lfs_markv: partial indirect block?"
448 				    " size=%d\n", blkp->bi_size);
449 			bp = getblk(vp, blkp->bi_lbn, blkp->bi_size, 0, 0);
450 			if (!(bp->b_oflags & (BO_DONE|BO_DELWRI))) {
451 				/*
452 				 * The block in question was not found
453 				 * in the cache; i.e., the block that
454 				 * getblk() returned is empty.	So, we
455 				 * can (and should) copy in the
456 				 * contents, because we've already
457 				 * determined that this was the right
458 				 * version of this block on disk.
459 				 *
460 				 * And, it can't have changed underneath
461 				 * us, because we have the segment lock.
462 				 */
463 				error = copyin(blkp->bi_bp, bp->b_data, blkp->bi_size);
464 				if (error)
465 					goto err2;
466 			}
467 		}
468 		if ((error = lfs_bwrite_ext(bp, BW_CLEAN)) != 0)
469 			goto err2;
470 
471 		nblkwritten++;
472 		/*
473 		 * XXX should account indirect blocks and ifile pages as well
474 		 */
475 		if (nblkwritten + lfs_lblkno(fs, ninowritten * sizeof (struct ulfs1_dinode))
476 		    > LFS_MARKV_MAX_BLOCKS) {
477 			DLOG((DLOG_CLEAN, "lfs_markv: writing %d blks %d inos\n",
478 			      nblkwritten, ninowritten));
479 			lfs_segwrite(mntp, SEGM_CLEAN);
480 			nblkwritten = ninowritten = 0;
481 		}
482 	}
483 
484 	/*
485 	 * Finish the old file, if there was one
486 	 */
487 	if (v_daddr != LFS_UNUSED_DADDR) {
488 		lfs_vunref(vp);
489 		numrefed--;
490 	}
491 
492 #ifdef DIAGNOSTIC
493 	if (numrefed != 0)
494 		panic("lfs_markv: numrefed=%d", numrefed);
495 #endif
496 	DLOG((DLOG_CLEAN, "lfs_markv: writing %d blks %d inos (check point)\n",
497 	      nblkwritten, ninowritten));
498 
499 	/*
500 	 * The last write has to be SEGM_SYNC, because of calling semantics.
501 	 * It also has to be SEGM_CKP, because otherwise we could write
502 	 * over the newly cleaned data contained in a checkpoint, and then
503 	 * we'd be unhappy at recovery time.
504 	 */
505 	lfs_segwrite(mntp, SEGM_CLEAN | SEGM_CKP | SEGM_SYNC);
506 
507 	lfs_segunlock(fs);
508 
509 	vfs_unbusy(mntp, false, NULL);
510 	if (error)
511 		return (error);
512 	else if (do_again)
513 		return EAGAIN;
514 
515 	return 0;
516 
517 err2:
518 	DLOG((DLOG_CLEAN, "lfs_markv err2\n"));
519 
520 	/*
521 	 * XXX we're here because copyin() failed.
522 	 * XXX it means that we can't trust the cleanerd.  too bad.
523 	 * XXX how can we recover from this?
524 	 */
525 
526 err3:
527 	/*
528 	 * XXX should do segwrite here anyway?
529 	 */
530 
531 	if (v_daddr != LFS_UNUSED_DADDR) {
532 		lfs_vunref(vp);
533 		--numrefed;
534 	}
535 
536 	lfs_segunlock(fs);
537 	vfs_unbusy(mntp, false, NULL);
538 #ifdef DIAGNOSTIC
539 	if (numrefed != 0)
540 		panic("lfs_markv: numrefed=%d", numrefed);
541 #endif
542 
543 	return (error);
544 }
545 
546 /*
547  * sys_lfs_bmapv:
548  *
549  * This will fill in the current disk address for arrays of blocks.
550  *
551  *  0 on success
552  * -1/errno is return on error.
553  */
554 #ifdef USE_64BIT_SYSCALLS
555 int
556 sys_lfs_bmapv(struct lwp *l, const struct sys_lfs_bmapv_args *uap, register_t *retval)
557 {
558 	/* {
559 		syscallarg(fsid_t *) fsidp;
560 		syscallarg(struct block_info *) blkiov;
561 		syscallarg(int) blkcnt;
562 	} */
563 	BLOCK_INFO *blkiov;
564 	int blkcnt, error;
565 	fsid_t fsid;
566 	struct lfs *fs;
567 	struct mount *mntp;
568 
569 	error = kauth_authorize_system(l->l_cred, KAUTH_SYSTEM_LFS,
570 	    KAUTH_REQ_SYSTEM_LFS_BMAPV, NULL, NULL, NULL);
571 	if (error)
572 		return (error);
573 
574 	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
575 		return (error);
576 
577 	if ((mntp = vfs_getvfs(&fsid)) == NULL)
578 		return (ENOENT);
579 	fs = VFSTOULFS(mntp)->um_lfs;
580 
581 	blkcnt = SCARG(uap, blkcnt);
582 	if ((u_int) blkcnt > SIZE_T_MAX / sizeof(BLOCK_INFO))
583 		return (EINVAL);
584 	KERNEL_LOCK(1, NULL);
585 	blkiov = lfs_malloc(fs, blkcnt * sizeof(BLOCK_INFO), LFS_NB_BLKIOV);
586 	if ((error = copyin(SCARG(uap, blkiov), blkiov,
587 			    blkcnt * sizeof(BLOCK_INFO))) != 0)
588 		goto out;
589 
590 	if ((error = lfs_bmapv(p, &fsid, blkiov, blkcnt)) == 0)
591 		copyout(blkiov, SCARG(uap, blkiov),
592 			blkcnt * sizeof(BLOCK_INFO));
593     out:
594 	lfs_free(fs, blkiov, LFS_NB_BLKIOV);
595 	KERNEL_UNLOCK_ONE(NULL);
596 	return error;
597 }
598 #else
599 int
600 sys_lfs_bmapv(struct lwp *l, const struct sys_lfs_bmapv_args *uap, register_t *retval)
601 {
602 	/* {
603 		syscallarg(fsid_t *) fsidp;
604 		syscallarg(struct block_info *) blkiov;
605 		syscallarg(int) blkcnt;
606 	} */
607 	BLOCK_INFO *blkiov;
608 	BLOCK_INFO_15 *blkiov15;
609 	int i, blkcnt, error;
610 	fsid_t fsid;
611 	struct lfs *fs;
612 	struct mount *mntp;
613 
614 	error = kauth_authorize_system(l->l_cred, KAUTH_SYSTEM_LFS,
615 	    KAUTH_REQ_SYSTEM_LFS_BMAPV, NULL, NULL, NULL);
616 	if (error)
617 		return (error);
618 
619 	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
620 		return (error);
621 
622 	if ((mntp = vfs_getvfs(&fsid)) == NULL)
623 		return (ENOENT);
624 	fs = VFSTOULFS(mntp)->um_lfs;
625 
626 	blkcnt = SCARG(uap, blkcnt);
627 	if ((size_t) blkcnt > SIZE_T_MAX / sizeof(BLOCK_INFO))
628 		return (EINVAL);
629 	KERNEL_LOCK(1, NULL);
630 	blkiov = lfs_malloc(fs, blkcnt * sizeof(BLOCK_INFO), LFS_NB_BLKIOV);
631 	blkiov15 = lfs_malloc(fs, blkcnt * sizeof(BLOCK_INFO_15), LFS_NB_BLKIOV);
632 	if ((error = copyin(SCARG(uap, blkiov), blkiov15,
633 			    blkcnt * sizeof(BLOCK_INFO_15))) != 0)
634 		goto out;
635 
636 	for (i = 0; i < blkcnt; i++) {
637 		blkiov[i].bi_inode     = blkiov15[i].bi_inode;
638 		blkiov[i].bi_lbn       = blkiov15[i].bi_lbn;
639 		blkiov[i].bi_daddr     = blkiov15[i].bi_daddr;
640 		blkiov[i].bi_segcreate = blkiov15[i].bi_segcreate;
641 		blkiov[i].bi_version   = blkiov15[i].bi_version;
642 		blkiov[i].bi_bp	       = blkiov15[i].bi_bp;
643 		blkiov[i].bi_size      = blkiov15[i].bi_size;
644 	}
645 
646 	if ((error = lfs_bmapv(l->l_proc, &fsid, blkiov, blkcnt)) == 0) {
647 		for (i = 0; i < blkcnt; i++) {
648 			blkiov15[i].bi_inode	 = blkiov[i].bi_inode;
649 			blkiov15[i].bi_lbn	 = blkiov[i].bi_lbn;
650 			blkiov15[i].bi_daddr	 = blkiov[i].bi_daddr;
651 			blkiov15[i].bi_segcreate = blkiov[i].bi_segcreate;
652 			blkiov15[i].bi_version	 = blkiov[i].bi_version;
653 			blkiov15[i].bi_bp	 = blkiov[i].bi_bp;
654 			blkiov15[i].bi_size	 = blkiov[i].bi_size;
655 		}
656 		copyout(blkiov15, SCARG(uap, blkiov),
657 			blkcnt * sizeof(BLOCK_INFO_15));
658 	}
659     out:
660 	lfs_free(fs, blkiov, LFS_NB_BLKIOV);
661 	lfs_free(fs, blkiov15, LFS_NB_BLKIOV);
662 	KERNEL_UNLOCK_ONE(NULL);
663 	return error;
664 }
665 #endif
666 
667 int
668 lfs_bmapv(struct proc *p, fsid_t *fsidp, BLOCK_INFO *blkiov, int blkcnt)
669 {
670 	BLOCK_INFO *blkp;
671 	IFILE *ifp;
672 	struct buf *bp;
673 	struct inode *ip = NULL;
674 	struct lfs *fs;
675 	struct mount *mntp;
676 	struct ulfsmount *ump;
677 	struct vnode *vp;
678 	ino_t lastino;
679 	daddr_t v_daddr;
680 	int cnt, error;
681 	int numrefed = 0;
682 
683 	lfs_cleaner_pid = p->p_pid;
684 
685 	if ((mntp = vfs_getvfs(fsidp)) == NULL)
686 		return (ENOENT);
687 
688 	ump = VFSTOULFS(mntp);
689 	if ((error = vfs_busy(mntp, NULL)) != 0)
690 		return (error);
691 
692 	cnt = blkcnt;
693 
694 	fs = VFSTOULFS(mntp)->um_lfs;
695 
696 	error = 0;
697 
698 	/* these were inside the initialization for the for loop */
699 	v_daddr = LFS_UNUSED_DADDR;
700 	lastino = LFS_UNUSED_INUM;
701 	for (blkp = blkiov; cnt--; ++blkp)
702 	{
703 		/*
704 		 * Get the IFILE entry (only once) and see if the file still
705 		 * exists.
706 		 */
707 		if (lastino != blkp->bi_inode) {
708 			/*
709 			 * Finish the old file, if there was one.  The presence
710 			 * of a usable vnode in vp is signaled by a valid
711 			 * v_daddr.
712 			 */
713 			if (v_daddr != LFS_UNUSED_DADDR) {
714 				lfs_vunref(vp);
715 				if (VTOI(vp)->i_lfs_iflags & LFSI_BMAP) {
716 					mutex_enter(vp->v_interlock);
717 					if (vget(vp, LK_NOWAIT) == 0) {
718 						if (! vrecycle(vp))
719 							vrele(vp);
720 					}
721 				}
722 				numrefed--;
723 			}
724 
725 			/*
726 			 * Start a new file
727 			 */
728 			lastino = blkp->bi_inode;
729 			if (blkp->bi_inode == LFS_IFILE_INUM)
730 				v_daddr = fs->lfs_idaddr;
731 			else {
732 				LFS_IENTRY(ifp, fs, blkp->bi_inode, bp);
733 				v_daddr = ifp->if_daddr;
734 				brelse(bp, 0);
735 			}
736 			if (v_daddr == LFS_UNUSED_DADDR) {
737 				blkp->bi_daddr = LFS_UNUSED_DADDR;
738 				continue;
739 			}
740 			/*
741 			 * A regular call to VFS_VGET could deadlock
742 			 * here.  Instead, we try an unlocked access.
743 			 */
744 			mutex_enter(&ulfs_ihash_lock);
745 			vp = ulfs_ihashlookup(ump->um_dev, blkp->bi_inode);
746 			if (vp != NULL)
747 				mutex_enter(vp->v_interlock);
748 			if (vp != NULL && vdead_check(vp, VDEAD_NOWAIT) == 0) {
749 				ip = VTOI(vp);
750 				mutex_exit(&ulfs_ihash_lock);
751 				if (lfs_vref(vp)) {
752 					v_daddr = LFS_UNUSED_DADDR;
753 					continue;
754 				}
755 				numrefed++;
756 			} else {
757 				if (vp != NULL)
758 					mutex_exit(vp->v_interlock);
759 				mutex_exit(&ulfs_ihash_lock);
760 				/*
761 				 * Don't VFS_VGET if we're being unmounted,
762 				 * since we hold vfs_busy().
763 				 */
764 				if (mntp->mnt_iflag & IMNT_UNMOUNT) {
765 					v_daddr = LFS_UNUSED_DADDR;
766 					continue;
767 				}
768 				error = VFS_VGET(mntp, blkp->bi_inode, &vp);
769 				if (error) {
770 					DLOG((DLOG_CLEAN, "lfs_bmapv: vget ino"
771 					      "%d failed with %d",
772 					      blkp->bi_inode,error));
773 					v_daddr = LFS_UNUSED_DADDR;
774 					continue;
775 				} else {
776 					KASSERT(VOP_ISLOCKED(vp));
777 					VTOI(vp)->i_lfs_iflags |= LFSI_BMAP;
778 					VOP_UNLOCK(vp);
779 					numrefed++;
780 				}
781 			}
782 			ip = VTOI(vp);
783 		} else if (v_daddr == LFS_UNUSED_DADDR) {
784 			/*
785 			 * This can only happen if the vnode is dead.
786 			 * Keep going.	Note that we DO NOT set the
787 			 * bi_addr to anything -- if we failed to get
788 			 * the vnode, for example, we want to assume
789 			 * conservatively that all of its blocks *are*
790 			 * located in the segment in question.
791 			 * lfs_markv will throw them out if we are
792 			 * wrong.
793 			 */
794 			/* blkp->bi_daddr = LFS_UNUSED_DADDR; */
795 			continue;
796 		}
797 
798 		/* Past this point we are guaranteed that vp, ip are valid. */
799 
800 		if (blkp->bi_lbn == LFS_UNUSED_LBN) {
801 			/*
802 			 * We just want the inode address, which is
803 			 * conveniently in v_daddr.
804 			 */
805 			blkp->bi_daddr = v_daddr;
806 		} else {
807 			daddr_t bi_daddr;
808 
809 			/* XXX ondisk32 */
810 			error = VOP_BMAP(vp, blkp->bi_lbn, NULL,
811 					 &bi_daddr, NULL);
812 			if (error)
813 			{
814 				blkp->bi_daddr = LFS_UNUSED_DADDR;
815 				continue;
816 			}
817 			blkp->bi_daddr = LFS_DBTOFSB(fs, bi_daddr);
818 			/* Fill in the block size, too */
819 			if (blkp->bi_lbn >= 0)
820 				blkp->bi_size = lfs_blksize(fs, ip, blkp->bi_lbn);
821 			else
822 				blkp->bi_size = fs->lfs_bsize;
823 		}
824 	}
825 
826 	/*
827 	 * Finish the old file, if there was one.  The presence
828 	 * of a usable vnode in vp is signaled by a valid v_daddr.
829 	 */
830 	if (v_daddr != LFS_UNUSED_DADDR) {
831 		lfs_vunref(vp);
832 		/* Recycle as above. */
833 		if (ip->i_lfs_iflags & LFSI_BMAP) {
834 			mutex_enter(vp->v_interlock);
835 			if (vget(vp, LK_NOWAIT) == 0) {
836 				if (! vrecycle(vp))
837 					vrele(vp);
838 			}
839 		}
840 		numrefed--;
841 	}
842 
843 #ifdef DIAGNOSTIC
844 	if (numrefed != 0)
845 		panic("lfs_bmapv: numrefed=%d", numrefed);
846 #endif
847 
848 	vfs_unbusy(mntp, false, NULL);
849 
850 	return 0;
851 }
852 
853 /*
854  * sys_lfs_segclean:
855  *
856  * Mark the segment clean.
857  *
858  *  0 on success
859  * -1/errno is return on error.
860  */
861 int
862 sys_lfs_segclean(struct lwp *l, const struct sys_lfs_segclean_args *uap, register_t *retval)
863 {
864 	/* {
865 		syscallarg(fsid_t *) fsidp;
866 		syscallarg(u_long) segment;
867 	} */
868 	struct lfs *fs;
869 	struct mount *mntp;
870 	fsid_t fsid;
871 	int error;
872 	unsigned long segnum;
873 
874 	error = kauth_authorize_system(l->l_cred, KAUTH_SYSTEM_LFS,
875 	    KAUTH_REQ_SYSTEM_LFS_SEGCLEAN, NULL, NULL, NULL);
876 	if (error)
877 		return (error);
878 
879 	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
880 		return (error);
881 	if ((mntp = vfs_getvfs(&fsid)) == NULL)
882 		return (ENOENT);
883 
884 	fs = VFSTOULFS(mntp)->um_lfs;
885 	segnum = SCARG(uap, segment);
886 
887 	if ((error = vfs_busy(mntp, NULL)) != 0)
888 		return (error);
889 
890 	KERNEL_LOCK(1, NULL);
891 	lfs_seglock(fs, SEGM_PROT);
892 	error = lfs_do_segclean(fs, segnum);
893 	lfs_segunlock(fs);
894 	KERNEL_UNLOCK_ONE(NULL);
895 	vfs_unbusy(mntp, false, NULL);
896 	return error;
897 }
898 
899 /*
900  * Actually mark the segment clean.
901  * Must be called with the segment lock held.
902  */
903 int
904 lfs_do_segclean(struct lfs *fs, unsigned long segnum)
905 {
906 	extern int lfs_dostats;
907 	struct buf *bp;
908 	CLEANERINFO *cip;
909 	SEGUSE *sup;
910 
911 	if (lfs_dtosn(fs, fs->lfs_curseg) == segnum) {
912 		return (EBUSY);
913 	}
914 
915 	LFS_SEGENTRY(sup, fs, segnum, bp);
916 	if (sup->su_nbytes) {
917 		DLOG((DLOG_CLEAN, "lfs_segclean: not cleaning segment %lu:"
918 		      " %d live bytes\n", segnum, sup->su_nbytes));
919 		brelse(bp, 0);
920 		return (EBUSY);
921 	}
922 	if (sup->su_flags & SEGUSE_ACTIVE) {
923 		DLOG((DLOG_CLEAN, "lfs_segclean: not cleaning segment %lu:"
924 		      " segment is active\n", segnum));
925 		brelse(bp, 0);
926 		return (EBUSY);
927 	}
928 	if (!(sup->su_flags & SEGUSE_DIRTY)) {
929 		DLOG((DLOG_CLEAN, "lfs_segclean: not cleaning segment %lu:"
930 		      " segment is already clean\n", segnum));
931 		brelse(bp, 0);
932 		return (EALREADY);
933 	}
934 
935 	fs->lfs_avail += lfs_segtod(fs, 1);
936 	if (sup->su_flags & SEGUSE_SUPERBLOCK)
937 		fs->lfs_avail -= lfs_btofsb(fs, LFS_SBPAD);
938 	if (fs->lfs_version > 1 && segnum == 0 &&
939 	    fs->lfs_start < lfs_btofsb(fs, LFS_LABELPAD))
940 		fs->lfs_avail -= lfs_btofsb(fs, LFS_LABELPAD) - fs->lfs_start;
941 	mutex_enter(&lfs_lock);
942 	fs->lfs_bfree += sup->su_nsums * lfs_btofsb(fs, fs->lfs_sumsize) +
943 		lfs_btofsb(fs, sup->su_ninos * fs->lfs_ibsize);
944 	fs->lfs_dmeta -= sup->su_nsums * lfs_btofsb(fs, fs->lfs_sumsize) +
945 		lfs_btofsb(fs, sup->su_ninos * fs->lfs_ibsize);
946 	if (fs->lfs_dmeta < 0)
947 		fs->lfs_dmeta = 0;
948 	mutex_exit(&lfs_lock);
949 	sup->su_flags &= ~SEGUSE_DIRTY;
950 	LFS_WRITESEGENTRY(sup, fs, segnum, bp);
951 
952 	LFS_CLEANERINFO(cip, fs, bp);
953 	++cip->clean;
954 	--cip->dirty;
955 	fs->lfs_nclean = cip->clean;
956 	cip->bfree = fs->lfs_bfree;
957 	mutex_enter(&lfs_lock);
958 	cip->avail = fs->lfs_avail - fs->lfs_ravail - fs->lfs_favail;
959 	wakeup(&fs->lfs_avail);
960 	mutex_exit(&lfs_lock);
961 	(void) LFS_BWRITE_LOG(bp);
962 
963 	if (lfs_dostats)
964 		++lfs_stats.segs_reclaimed;
965 
966 	return (0);
967 }
968 
969 /*
970  * This will block until a segment in file system fsid is written.  A timeout
971  * in milliseconds may be specified which will awake the cleaner automatically.
972  * An fsid of -1 means any file system, and a timeout of 0 means forever.
973  */
974 int
975 lfs_segwait(fsid_t *fsidp, struct timeval *tv)
976 {
977 	struct mount *mntp;
978 	void *addr;
979 	u_long timeout;
980 	int error;
981 
982 	KERNEL_LOCK(1, NULL);
983 	if (fsidp == NULL || (mntp = vfs_getvfs(fsidp)) == NULL)
984 		addr = &lfs_allclean_wakeup;
985 	else
986 		addr = &VFSTOULFS(mntp)->um_lfs->lfs_nextseg;
987 	/*
988 	 * XXX THIS COULD SLEEP FOREVER IF TIMEOUT IS {0,0}!
989 	 * XXX IS THAT WHAT IS INTENDED?
990 	 */
991 	timeout = tvtohz(tv);
992 	error = tsleep(addr, PCATCH | PVFS, "segment", timeout);
993 	KERNEL_UNLOCK_ONE(NULL);
994 	return (error == ERESTART ? EINTR : 0);
995 }
996 
997 /*
998  * sys_lfs_segwait:
999  *
1000  * System call wrapper around lfs_segwait().
1001  *
1002  *  0 on success
1003  *  1 on timeout
1004  * -1/errno is return on error.
1005  */
1006 int
1007 sys___lfs_segwait50(struct lwp *l, const struct sys___lfs_segwait50_args *uap,
1008     register_t *retval)
1009 {
1010 	/* {
1011 		syscallarg(fsid_t *) fsidp;
1012 		syscallarg(struct timeval *) tv;
1013 	} */
1014 	struct timeval atv;
1015 	fsid_t fsid;
1016 	int error;
1017 
1018 	/* XXX need we be su to segwait? */
1019 	error = kauth_authorize_system(l->l_cred, KAUTH_SYSTEM_LFS,
1020 	    KAUTH_REQ_SYSTEM_LFS_SEGWAIT, NULL, NULL, NULL);
1021 	if (error)
1022 		return (error);
1023 	if ((error = copyin(SCARG(uap, fsidp), &fsid, sizeof(fsid_t))) != 0)
1024 		return (error);
1025 
1026 	if (SCARG(uap, tv)) {
1027 		error = copyin(SCARG(uap, tv), &atv, sizeof(struct timeval));
1028 		if (error)
1029 			return (error);
1030 		if (itimerfix(&atv))
1031 			return (EINVAL);
1032 	} else /* NULL or invalid */
1033 		atv.tv_sec = atv.tv_usec = 0;
1034 	return lfs_segwait(&fsid, &atv);
1035 }
1036 
1037 /*
1038  * VFS_VGET call specialized for the cleaner.  The cleaner already knows the
1039  * daddr from the ifile, so don't look it up again.  If the cleaner is
1040  * processing IINFO structures, it may have the ondisk inode already, so
1041  * don't go retrieving it again.
1042  *
1043  * we lfs_vref, and it is the caller's responsibility to lfs_vunref
1044  * when finished.
1045  */
1046 
1047 int
1048 lfs_fasthashget(dev_t dev, ino_t ino, struct vnode **vpp)
1049 {
1050 	struct vnode *vp;
1051 
1052 	mutex_enter(&ulfs_ihash_lock);
1053 	if ((vp = ulfs_ihashlookup(dev, ino)) != NULL) {
1054 		mutex_enter(vp->v_interlock);
1055 		mutex_exit(&ulfs_ihash_lock);
1056 		if (vdead_check(vp, VDEAD_NOWAIT) != 0) {
1057 			DLOG((DLOG_CLEAN, "lfs_fastvget: ino %d dead\n",
1058 			      ino));
1059 			lfs_stats.clean_vnlocked++;
1060 			mutex_exit(vp->v_interlock);
1061 			return EAGAIN;
1062 		}
1063 		if (lfs_vref(vp)) {
1064 			DLOG((DLOG_CLEAN, "lfs_fastvget: lfs_vref failed"
1065 			      " for ino %d\n", ino));
1066 			lfs_stats.clean_inlocked++;
1067 			return EAGAIN;
1068 		}
1069 	} else {
1070 		mutex_exit(&ulfs_ihash_lock);
1071 	}
1072 	*vpp = vp;
1073 
1074 	return (0);
1075 }
1076 
1077 int
1078 lfs_fastvget(struct mount *mp, ino_t ino, daddr_t daddr, struct vnode **vpp,
1079 	     struct ulfs1_dinode *dinp)
1080 {
1081 	struct inode *ip;
1082 	struct ulfs1_dinode *dip;
1083 	struct vnode *vp;
1084 	struct ulfsmount *ump;
1085 	dev_t dev;
1086 	int error, retries;
1087 	struct buf *bp;
1088 	struct lfs *fs;
1089 
1090 	ump = VFSTOULFS(mp);
1091 	dev = ump->um_dev;
1092 	fs = ump->um_lfs;
1093 
1094 	/*
1095 	 * Wait until the filesystem is fully mounted before allowing vget
1096 	 * to complete.	 This prevents possible problems with roll-forward.
1097 	 */
1098 	mutex_enter(&lfs_lock);
1099 	while (fs->lfs_flags & LFS_NOTYET) {
1100 		mtsleep(&fs->lfs_flags, PRIBIO+1, "lfs_fnotyet", 0,
1101 			&lfs_lock);
1102 	}
1103 	mutex_exit(&lfs_lock);
1104 
1105 	/*
1106 	 * This is playing fast and loose.  Someone may have the inode
1107 	 * locked, in which case they are going to be distinctly unhappy
1108 	 * if we trash something.
1109 	 */
1110 
1111 	error = lfs_fasthashget(dev, ino, vpp);
1112 	if (error != 0 || *vpp != NULL)
1113 		return (error);
1114 
1115 	/*
1116 	 * getnewvnode(9) will call vfs_busy, which will block if the
1117 	 * filesystem is being unmounted; but umount(9) is waiting for
1118 	 * us because we're already holding the fs busy.
1119 	 * XXXMP
1120 	 */
1121 	if (mp->mnt_iflag & IMNT_UNMOUNT) {
1122 		*vpp = NULL;
1123 		return EDEADLK;
1124 	}
1125 	error = getnewvnode(VT_LFS, mp, lfs_vnodeop_p, NULL, &vp);
1126 	if (error) {
1127 		*vpp = NULL;
1128 		return (error);
1129 	}
1130 
1131 	mutex_enter(&ulfs_hashlock);
1132 	error = lfs_fasthashget(dev, ino, vpp);
1133 	if (error != 0 || *vpp != NULL) {
1134 		mutex_exit(&ulfs_hashlock);
1135 		ungetnewvnode(vp);
1136 		return (error);
1137 	}
1138 
1139 	/* Allocate new vnode/inode. */
1140 	lfs_vcreate(mp, ino, vp);
1141 
1142 	/*
1143 	 * Put it onto its hash chain and lock it so that other requests for
1144 	 * this inode will block if they arrive while we are sleeping waiting
1145 	 * for old data structures to be purged or for the contents of the
1146 	 * disk portion of this inode to be read.
1147 	 */
1148 	ip = VTOI(vp);
1149 	ulfs_ihashins(ip);
1150 	mutex_exit(&ulfs_hashlock);
1151 
1152 #ifdef notyet
1153 	/* Not found in the cache => this vnode was loaded only for cleaning. */
1154 	ip->i_lfs_iflags |= LFSI_BMAP;
1155 #endif
1156 
1157 	/*
1158 	 * XXX
1159 	 * This may not need to be here, logically it should go down with
1160 	 * the i_devvp initialization.
1161 	 * Ask Kirk.
1162 	 */
1163 	ip->i_lfs = fs;
1164 
1165 	/* Read in the disk contents for the inode, copy into the inode. */
1166 	if (dinp) {
1167 		error = copyin(dinp, ip->i_din.ffs1_din, sizeof (struct ulfs1_dinode));
1168 		if (error) {
1169 			DLOG((DLOG_CLEAN, "lfs_fastvget: dinode copyin failed"
1170 			      " for ino %d\n", ino));
1171 			ulfs_ihashrem(ip);
1172 
1173 			/* Unlock and discard unneeded inode. */
1174 			VOP_UNLOCK(vp);
1175 			lfs_vunref(vp);
1176 			*vpp = NULL;
1177 			return (error);
1178 		}
1179 		if (ip->i_number != ino)
1180 			panic("lfs_fastvget: I was fed the wrong inode!");
1181 	} else {
1182 		retries = 0;
1183 	    again:
1184 		error = bread(ump->um_devvp, LFS_FSBTODB(fs, daddr), fs->lfs_ibsize,
1185 			      NOCRED, 0, &bp);
1186 		if (error) {
1187 			DLOG((DLOG_CLEAN, "lfs_fastvget: bread failed (%d)\n",
1188 			      error));
1189 			/*
1190 			 * The inode does not contain anything useful, so it
1191 			 * would be misleading to leave it on its hash chain.
1192 			 * Iput() will return it to the free list.
1193 			 */
1194 			ulfs_ihashrem(ip);
1195 
1196 			/* Unlock and discard unneeded inode. */
1197 			VOP_UNLOCK(vp);
1198 			lfs_vunref(vp);
1199 			*vpp = NULL;
1200 			return (error);
1201 		}
1202 		dip = lfs_ifind(ump->um_lfs, ino, bp);
1203 		if (dip == NULL) {
1204 			/* Assume write has not completed yet; try again */
1205 			brelse(bp, BC_INVAL);
1206 			++retries;
1207 			if (retries > LFS_IFIND_RETRIES)
1208 				panic("lfs_fastvget: dinode not found");
1209 			DLOG((DLOG_CLEAN, "lfs_fastvget: dinode not found,"
1210 			      " retrying...\n"));
1211 			goto again;
1212 		}
1213 		*ip->i_din.ffs1_din = *dip;
1214 		brelse(bp, 0);
1215 	}
1216 	lfs_vinit(mp, &vp);
1217 
1218 	*vpp = vp;
1219 
1220 	KASSERT(VOP_ISLOCKED(vp));
1221 	VOP_UNLOCK(vp);
1222 
1223 	return (0);
1224 }
1225 
1226 /*
1227  * Make up a "fake" cleaner buffer, copy the data from userland into it.
1228  */
1229 struct buf *
1230 lfs_fakebuf(struct lfs *fs, struct vnode *vp, int lbn, size_t size, void *uaddr)
1231 {
1232 	struct buf *bp;
1233 	int error;
1234 
1235 	KASSERT(VTOI(vp)->i_number != LFS_IFILE_INUM);
1236 
1237 	bp = lfs_newbuf(VTOI(vp)->i_lfs, vp, lbn, size, LFS_NB_CLEAN);
1238 	error = copyin(uaddr, bp->b_data, size);
1239 	if (error) {
1240 		lfs_freebuf(fs, bp);
1241 		return NULL;
1242 	}
1243 	KDASSERT(bp->b_iodone == lfs_callback);
1244 
1245 #if 0
1246 	mutex_enter(&lfs_lock);
1247 	++fs->lfs_iocount;
1248 	mutex_exit(&lfs_lock);
1249 #endif
1250 	bp->b_bufsize = size;
1251 	bp->b_bcount = size;
1252 	return (bp);
1253 }
1254