xref: /original-bsd/sys/ufs/ffs/ffs_alloc.c (revision 3b235ced)
1 /*
2  * Copyright (c) 1982, 1986, 1989, 1993
3  *	The Regents of the University of California.  All rights reserved.
4  *
5  * %sccs.include.redist.c%
6  *
7  *	@(#)ffs_alloc.c	8.7 (Berkeley) 02/14/94
8  */
9 
10 #include <sys/param.h>
11 #include <sys/systm.h>
12 #include <sys/buf.h>
13 #include <sys/proc.h>
14 #include <sys/vnode.h>
15 #include <sys/mount.h>
16 #include <sys/kernel.h>
17 #include <sys/syslog.h>
18 
19 #include <vm/vm.h>
20 
21 #include <ufs/ufs/quota.h>
22 #include <ufs/ufs/inode.h>
23 
24 #include <ufs/ffs/fs.h>
25 #include <ufs/ffs/ffs_extern.h>
26 
27 extern u_long nextgennumber;
28 
29 static daddr_t	ffs_alloccg __P((struct inode *, int, daddr_t, int));
30 static daddr_t	ffs_alloccgblk __P((struct fs *, struct cg *, daddr_t));
31 static daddr_t	ffs_clusteralloc __P((struct inode *, int, daddr_t, int));
32 static ino_t	ffs_dirpref __P((struct fs *));
33 static daddr_t	ffs_fragextend __P((struct inode *, int, long, int, int));
34 static void	ffs_fserr __P((struct fs *, u_int, char *));
35 static u_long	ffs_hashalloc
36 		    __P((struct inode *, int, long, int, u_long (*)()));
37 static ino_t	ffs_nodealloccg __P((struct inode *, int, daddr_t, int));
38 static daddr_t	ffs_mapsearch __P((struct fs *, struct cg *, daddr_t, int));
39 
40 /*
41  * Allocate a block in the file system.
42  *
43  * The size of the requested block is given, which must be some
44  * multiple of fs_fsize and <= fs_bsize.
45  * A preference may be optionally specified. If a preference is given
46  * the following hierarchy is used to allocate a block:
47  *   1) allocate the requested block.
48  *   2) allocate a rotationally optimal block in the same cylinder.
49  *   3) allocate a block in the same cylinder group.
50  *   4) quadradically rehash into other cylinder groups, until an
51  *      available block is located.
52  * If no block preference is given the following heirarchy is used
53  * to allocate a block:
54  *   1) allocate a block in the cylinder group that contains the
55  *      inode for the file.
56  *   2) quadradically rehash into other cylinder groups, until an
57  *      available block is located.
58  */
59 ffs_alloc(ip, lbn, bpref, size, cred, bnp)
60 	register struct inode *ip;
61 	daddr_t lbn, bpref;
62 	int size;
63 	struct ucred *cred;
64 	daddr_t *bnp;
65 {
66 	register struct fs *fs;
67 	daddr_t bno;
68 	int cg, error;
69 
70 	*bnp = 0;
71 	fs = ip->i_fs;
72 #ifdef DIAGNOSTIC
73 	if ((u_int)size > fs->fs_bsize || fragoff(fs, size) != 0) {
74 		printf("dev = 0x%x, bsize = %d, size = %d, fs = %s\n",
75 		    ip->i_dev, fs->fs_bsize, size, fs->fs_fsmnt);
76 		panic("ffs_alloc: bad size");
77 	}
78 	if (cred == NOCRED)
79 		panic("ffs_alloc: missing credential\n");
80 #endif /* DIAGNOSTIC */
81 	if (size == fs->fs_bsize && fs->fs_cstotal.cs_nbfree == 0)
82 		goto nospace;
83 	if (cred->cr_uid != 0 && freespace(fs, fs->fs_minfree) <= 0)
84 		goto nospace;
85 #ifdef QUOTA
86 	if (error = chkdq(ip, (long)btodb(size), cred, 0))
87 		return (error);
88 #endif
89 	if (bpref >= fs->fs_size)
90 		bpref = 0;
91 	if (bpref == 0)
92 		cg = ino_to_cg(fs, ip->i_number);
93 	else
94 		cg = dtog(fs, bpref);
95 	bno = (daddr_t)ffs_hashalloc(ip, cg, (long)bpref, size,
96 	    (u_long (*)())ffs_alloccg);
97 	if (bno > 0) {
98 		ip->i_blocks += btodb(size);
99 		ip->i_flag |= IN_CHANGE | IN_UPDATE;
100 		*bnp = bno;
101 		return (0);
102 	}
103 #ifdef QUOTA
104 	/*
105 	 * Restore user's disk quota because allocation failed.
106 	 */
107 	(void) chkdq(ip, (long)-btodb(size), cred, FORCE);
108 #endif
109 nospace:
110 	ffs_fserr(fs, cred->cr_uid, "file system full");
111 	uprintf("\n%s: write failed, file system is full\n", fs->fs_fsmnt);
112 	return (ENOSPC);
113 }
114 
115 /*
116  * Reallocate a fragment to a bigger size
117  *
118  * The number and size of the old block is given, and a preference
119  * and new size is also specified. The allocator attempts to extend
120  * the original block. Failing that, the regular block allocator is
121  * invoked to get an appropriate block.
122  */
123 ffs_realloccg(ip, lbprev, bpref, osize, nsize, cred, bpp)
124 	register struct inode *ip;
125 	daddr_t lbprev;
126 	daddr_t bpref;
127 	int osize, nsize;
128 	struct ucred *cred;
129 	struct buf **bpp;
130 {
131 	register struct fs *fs;
132 	struct buf *bp;
133 	int cg, request, error;
134 	daddr_t bprev, bno;
135 
136 	*bpp = 0;
137 	fs = ip->i_fs;
138 #ifdef DIAGNOSTIC
139 	if ((u_int)osize > fs->fs_bsize || fragoff(fs, osize) != 0 ||
140 	    (u_int)nsize > fs->fs_bsize || fragoff(fs, nsize) != 0) {
141 		printf(
142 		    "dev = 0x%x, bsize = %d, osize = %d, nsize = %d, fs = %s\n",
143 		    ip->i_dev, fs->fs_bsize, osize, nsize, fs->fs_fsmnt);
144 		panic("ffs_realloccg: bad size");
145 	}
146 	if (cred == NOCRED)
147 		panic("ffs_realloccg: missing credential\n");
148 #endif /* DIAGNOSTIC */
149 	if (cred->cr_uid != 0 && freespace(fs, fs->fs_minfree) <= 0)
150 		goto nospace;
151 	if ((bprev = ip->i_db[lbprev]) == 0) {
152 		printf("dev = 0x%x, bsize = %d, bprev = %d, fs = %s\n",
153 		    ip->i_dev, fs->fs_bsize, bprev, fs->fs_fsmnt);
154 		panic("ffs_realloccg: bad bprev");
155 	}
156 	/*
157 	 * Allocate the extra space in the buffer.
158 	 */
159 	if (error = bread(ITOV(ip), lbprev, osize, NOCRED, &bp)) {
160 		brelse(bp);
161 		return (error);
162 	}
163 #ifdef QUOTA
164 	if (error = chkdq(ip, (long)btodb(nsize - osize), cred, 0)) {
165 		brelse(bp);
166 		return (error);
167 	}
168 #endif
169 	/*
170 	 * Check for extension in the existing location.
171 	 */
172 	cg = dtog(fs, bprev);
173 	if (bno = ffs_fragextend(ip, cg, (long)bprev, osize, nsize)) {
174 		if (bp->b_blkno != fsbtodb(fs, bno))
175 			panic("bad blockno");
176 		ip->i_blocks += btodb(nsize - osize);
177 		ip->i_flag |= IN_CHANGE | IN_UPDATE;
178 		allocbuf(bp, nsize);
179 		bp->b_flags |= B_DONE;
180 		bzero((char *)bp->b_data + osize, (u_int)nsize - osize);
181 		*bpp = bp;
182 		return (0);
183 	}
184 	/*
185 	 * Allocate a new disk location.
186 	 */
187 	if (bpref >= fs->fs_size)
188 		bpref = 0;
189 	switch ((int)fs->fs_optim) {
190 	case FS_OPTSPACE:
191 		/*
192 		 * Allocate an exact sized fragment. Although this makes
193 		 * best use of space, we will waste time relocating it if
194 		 * the file continues to grow. If the fragmentation is
195 		 * less than half of the minimum free reserve, we choose
196 		 * to begin optimizing for time.
197 		 */
198 		request = nsize;
199 		if (fs->fs_minfree < 5 ||
200 		    fs->fs_cstotal.cs_nffree >
201 		    fs->fs_dsize * fs->fs_minfree / (2 * 100))
202 			break;
203 		log(LOG_NOTICE, "%s: optimization changed from SPACE to TIME\n",
204 			fs->fs_fsmnt);
205 		fs->fs_optim = FS_OPTTIME;
206 		break;
207 	case FS_OPTTIME:
208 		/*
209 		 * At this point we have discovered a file that is trying to
210 		 * grow a small fragment to a larger fragment. To save time,
211 		 * we allocate a full sized block, then free the unused portion.
212 		 * If the file continues to grow, the `ffs_fragextend' call
213 		 * above will be able to grow it in place without further
214 		 * copying. If aberrant programs cause disk fragmentation to
215 		 * grow within 2% of the free reserve, we choose to begin
216 		 * optimizing for space.
217 		 */
218 		request = fs->fs_bsize;
219 		if (fs->fs_cstotal.cs_nffree <
220 		    fs->fs_dsize * (fs->fs_minfree - 2) / 100)
221 			break;
222 		log(LOG_NOTICE, "%s: optimization changed from TIME to SPACE\n",
223 			fs->fs_fsmnt);
224 		fs->fs_optim = FS_OPTSPACE;
225 		break;
226 	default:
227 		printf("dev = 0x%x, optim = %d, fs = %s\n",
228 		    ip->i_dev, fs->fs_optim, fs->fs_fsmnt);
229 		panic("ffs_realloccg: bad optim");
230 		/* NOTREACHED */
231 	}
232 	bno = (daddr_t)ffs_hashalloc(ip, cg, (long)bpref, request,
233 	    (u_long (*)())ffs_alloccg);
234 	if (bno > 0) {
235 		bp->b_blkno = fsbtodb(fs, bno);
236 		(void) vnode_pager_uncache(ITOV(ip));
237 		ffs_blkfree(ip, bprev, (long)osize);
238 		if (nsize < request)
239 			ffs_blkfree(ip, bno + numfrags(fs, nsize),
240 			    (long)(request - nsize));
241 		ip->i_blocks += btodb(nsize - osize);
242 		ip->i_flag |= IN_CHANGE | IN_UPDATE;
243 		allocbuf(bp, nsize);
244 		bp->b_flags |= B_DONE;
245 		bzero((char *)bp->b_data + osize, (u_int)nsize - osize);
246 		*bpp = bp;
247 		return (0);
248 	}
249 #ifdef QUOTA
250 	/*
251 	 * Restore user's disk quota because allocation failed.
252 	 */
253 	(void) chkdq(ip, (long)-btodb(nsize - osize), cred, FORCE);
254 #endif
255 	brelse(bp);
256 nospace:
257 	/*
258 	 * no space available
259 	 */
260 	ffs_fserr(fs, cred->cr_uid, "file system full");
261 	uprintf("\n%s: write failed, file system is full\n", fs->fs_fsmnt);
262 	return (ENOSPC);
263 }
264 
265 /*
266  * Reallocate a sequence of blocks into a contiguous sequence of blocks.
267  *
268  * The vnode and an array of buffer pointers for a range of sequential
269  * logical blocks to be made contiguous is given. The allocator attempts
270  * to find a range of sequential blocks starting as close as possible to
271  * an fs_rotdelay offset from the end of the allocation for the logical
272  * block immediately preceeding the current range. If successful, the
273  * physical block numbers in the buffer pointers and in the inode are
274  * changed to reflect the new allocation. If unsuccessful, the allocation
275  * is left unchanged. The success in doing the reallocation is returned.
276  * Note that the error return is not reflected back to the user. Rather
277  * the previous block allocation will be used.
278  */
279 int
280 ffs_reallocblks(ap)
281 	struct vop_reallocblks_args /* {
282 		struct vnode *a_vp;
283 		struct cluster_save *a_buflist;
284 	} */ *ap;
285 {
286 	struct fs *fs;
287 	struct inode *ip;
288 	struct vnode *vp;
289 	struct buf *sbp, *ebp;
290 	daddr_t *bap, *sbap, *ebap;
291 	struct cluster_save *buflist;
292 	daddr_t start_lbn, end_lbn, soff, eoff, newblk, blkno;
293 	struct indir start_ap[NIADDR + 1], end_ap[NIADDR + 1], *idp;
294 	int i, len, start_lvl, end_lvl, pref, ssize;
295 
296 	vp = ap->a_vp;
297 	ip = VTOI(vp);
298 	fs = ip->i_fs;
299 	if (fs->fs_contigsumsize <= 0)
300 		return (ENOSPC);
301 	buflist = ap->a_buflist;
302 	len = buflist->bs_nchildren;
303 	start_lbn = buflist->bs_children[0]->b_lblkno;
304 	end_lbn = start_lbn + len - 1;
305 #ifdef DIAGNOSTIC
306 	for (i = 1; i < len; i++)
307 		if (buflist->bs_children[i]->b_lblkno != start_lbn + i)
308 			panic("ffs_reallocblks: non-cluster");
309 #endif
310 	/*
311 	 * If the latest allocation is in a new cylinder group, assume that
312 	 * the filesystem has decided to move and do not force it back to
313 	 * the previous cylinder group.
314 	 */
315 	if (dtog(fs, dbtofsb(fs, buflist->bs_children[0]->b_blkno)) !=
316 	    dtog(fs, dbtofsb(fs, buflist->bs_children[len - 1]->b_blkno)))
317 		return (ENOSPC);
318 	if (ufs_getlbns(vp, start_lbn, start_ap, &start_lvl) ||
319 	    ufs_getlbns(vp, end_lbn, end_ap, &end_lvl))
320 		return (ENOSPC);
321 	/*
322 	 * Get the starting offset and block map for the first block.
323 	 */
324 	if (start_lvl == 0) {
325 		sbap = &ip->i_db[0];
326 		soff = start_lbn;
327 	} else {
328 		idp = &start_ap[start_lvl - 1];
329 		if (bread(vp, idp->in_lbn, (int)fs->fs_bsize, NOCRED, &sbp)) {
330 			brelse(sbp);
331 			return (ENOSPC);
332 		}
333 		sbap = (daddr_t *)sbp->b_data;
334 		soff = idp->in_off;
335 	}
336 	/*
337 	 * Find the preferred location for the cluster.
338 	 */
339 	pref = ffs_blkpref(ip, start_lbn, soff, sbap);
340 	/*
341 	 * If the block range spans two block maps, get the second map.
342 	 */
343 	if (end_lvl == 0 || (idp = &end_ap[end_lvl - 1])->in_off + 1 >= len) {
344 		ssize = len;
345 	} else {
346 #ifdef DIAGNOSTIC
347 		if (start_ap[start_lvl-1].in_lbn == idp->in_lbn)
348 			panic("ffs_reallocblk: start == end");
349 #endif
350 		ssize = len - (idp->in_off + 1);
351 		if (bread(vp, idp->in_lbn, (int)fs->fs_bsize, NOCRED, &ebp))
352 			goto fail;
353 		ebap = (daddr_t *)ebp->b_data;
354 	}
355 	/*
356 	 * Search the block map looking for an allocation of the desired size.
357 	 */
358 	if ((newblk = (daddr_t)ffs_hashalloc(ip, dtog(fs, pref), (long)pref,
359 	    len, (u_long (*)())ffs_clusteralloc)) == 0)
360 		goto fail;
361 	/*
362 	 * We have found a new contiguous block.
363 	 *
364 	 * First we have to replace the old block pointers with the new
365 	 * block pointers in the inode and indirect blocks associated
366 	 * with the file.
367 	 */
368 	blkno = newblk;
369 	for (bap = &sbap[soff], i = 0; i < len; i++, blkno += fs->fs_frag) {
370 		if (i == ssize)
371 			bap = ebap;
372 #ifdef DIAGNOSTIC
373 		if (buflist->bs_children[i]->b_blkno != fsbtodb(fs, *bap))
374 			panic("ffs_reallocblks: alloc mismatch");
375 #endif
376 		*bap++ = blkno;
377 	}
378 	/*
379 	 * Next we must write out the modified inode and indirect blocks.
380 	 * The writes are synchronous since the old block values may have
381 	 * been written to disk.
382 	 *
383 	 * These writes should be changed to be bdwrites (and the VOP_UPDATE
384 	 * dropped) when a flag has been added to the buffers and inodes
385 	 * to detect when they have been written. It should be set when the
386 	 * cluster is started and cleared whenever the buffer or inode is
387 	 * flushed. We can then check below to see if it is set, and do
388 	 * the synchronous write only when it has been cleared.
389 	 */
390 	if (sbap != &ip->i_db[0]) {
391 		bwrite(sbp);
392 	} else {
393 		ip->i_flag |= IN_CHANGE | IN_UPDATE;
394 		VOP_UPDATE(vp, &time, &time, MNT_WAIT);
395 	}
396 	if (ssize < len)
397 		bwrite(ebp);
398 	/*
399 	 * Last, free the old blocks and assign the new blocks to the buffers.
400 	 */
401 	for (blkno = newblk, i = 0; i < len; i++, blkno += fs->fs_frag) {
402 		ffs_blkfree(ip, dbtofsb(fs, buflist->bs_children[i]->b_blkno),
403 		    fs->fs_bsize);
404 		buflist->bs_children[i]->b_blkno = fsbtodb(fs, blkno);
405 	}
406 	return (0);
407 
408 fail:
409 	if (ssize < len)
410 		brelse(ebp);
411 	if (sbap != &ip->i_db[0])
412 		brelse(sbp);
413 	return (ENOSPC);
414 }
415 
416 /*
417  * Allocate an inode in the file system.
418  *
419  * If allocating a directory, use ffs_dirpref to select the inode.
420  * If allocating in a directory, the following hierarchy is followed:
421  *   1) allocate the preferred inode.
422  *   2) allocate an inode in the same cylinder group.
423  *   3) quadradically rehash into other cylinder groups, until an
424  *      available inode is located.
425  * If no inode preference is given the following heirarchy is used
426  * to allocate an inode:
427  *   1) allocate an inode in cylinder group 0.
428  *   2) quadradically rehash into other cylinder groups, until an
429  *      available inode is located.
430  */
431 ffs_valloc(ap)
432 	struct vop_valloc_args /* {
433 		struct vnode *a_pvp;
434 		int a_mode;
435 		struct ucred *a_cred;
436 		struct vnode **a_vpp;
437 	} */ *ap;
438 {
439 	register struct vnode *pvp = ap->a_pvp;
440 	register struct inode *pip;
441 	register struct fs *fs;
442 	register struct inode *ip;
443 	mode_t mode = ap->a_mode;
444 	ino_t ino, ipref;
445 	int cg, error;
446 
447 	*ap->a_vpp = NULL;
448 	pip = VTOI(pvp);
449 	fs = pip->i_fs;
450 	if (fs->fs_cstotal.cs_nifree == 0)
451 		goto noinodes;
452 
453 	if ((mode & IFMT) == IFDIR)
454 		ipref = ffs_dirpref(fs);
455 	else
456 		ipref = pip->i_number;
457 	if (ipref >= fs->fs_ncg * fs->fs_ipg)
458 		ipref = 0;
459 	cg = ino_to_cg(fs, ipref);
460 	ino = (ino_t)ffs_hashalloc(pip, cg, (long)ipref, mode, ffs_nodealloccg);
461 	if (ino == 0)
462 		goto noinodes;
463 	error = VFS_VGET(pvp->v_mount, ino, ap->a_vpp);
464 	if (error) {
465 		VOP_VFREE(pvp, ino, mode);
466 		return (error);
467 	}
468 	ip = VTOI(*ap->a_vpp);
469 	if (ip->i_mode) {
470 		printf("mode = 0%o, inum = %d, fs = %s\n",
471 		    ip->i_mode, ip->i_number, fs->fs_fsmnt);
472 		panic("ffs_valloc: dup alloc");
473 	}
474 	if (ip->i_blocks) {				/* XXX */
475 		printf("free inode %s/%d had %d blocks\n",
476 		    fs->fs_fsmnt, ino, ip->i_blocks);
477 		ip->i_blocks = 0;
478 	}
479 	ip->i_flags = 0;
480 	/*
481 	 * Set up a new generation number for this inode.
482 	 */
483 	if (++nextgennumber < (u_long)time.tv_sec)
484 		nextgennumber = time.tv_sec;
485 	ip->i_gen = nextgennumber;
486 	return (0);
487 noinodes:
488 	ffs_fserr(fs, ap->a_cred->cr_uid, "out of inodes");
489 	uprintf("\n%s: create/symlink failed, no inodes free\n", fs->fs_fsmnt);
490 	return (ENOSPC);
491 }
492 
493 /*
494  * Find a cylinder to place a directory.
495  *
496  * The policy implemented by this algorithm is to select from
497  * among those cylinder groups with above the average number of
498  * free inodes, the one with the smallest number of directories.
499  */
500 static ino_t
501 ffs_dirpref(fs)
502 	register struct fs *fs;
503 {
504 	int cg, minndir, mincg, avgifree;
505 
506 	avgifree = fs->fs_cstotal.cs_nifree / fs->fs_ncg;
507 	minndir = fs->fs_ipg;
508 	mincg = 0;
509 	for (cg = 0; cg < fs->fs_ncg; cg++)
510 		if (fs->fs_cs(fs, cg).cs_ndir < minndir &&
511 		    fs->fs_cs(fs, cg).cs_nifree >= avgifree) {
512 			mincg = cg;
513 			minndir = fs->fs_cs(fs, cg).cs_ndir;
514 		}
515 	return ((ino_t)(fs->fs_ipg * mincg));
516 }
517 
518 /*
519  * Select the desired position for the next block in a file.  The file is
520  * logically divided into sections. The first section is composed of the
521  * direct blocks. Each additional section contains fs_maxbpg blocks.
522  *
523  * If no blocks have been allocated in the first section, the policy is to
524  * request a block in the same cylinder group as the inode that describes
525  * the file. If no blocks have been allocated in any other section, the
526  * policy is to place the section in a cylinder group with a greater than
527  * average number of free blocks.  An appropriate cylinder group is found
528  * by using a rotor that sweeps the cylinder groups. When a new group of
529  * blocks is needed, the sweep begins in the cylinder group following the
530  * cylinder group from which the previous allocation was made. The sweep
531  * continues until a cylinder group with greater than the average number
532  * of free blocks is found. If the allocation is for the first block in an
533  * indirect block, the information on the previous allocation is unavailable;
534  * here a best guess is made based upon the logical block number being
535  * allocated.
536  *
537  * If a section is already partially allocated, the policy is to
538  * contiguously allocate fs_maxcontig blocks.  The end of one of these
539  * contiguous blocks and the beginning of the next is physically separated
540  * so that the disk head will be in transit between them for at least
541  * fs_rotdelay milliseconds.  This is to allow time for the processor to
542  * schedule another I/O transfer.
543  */
544 daddr_t
545 ffs_blkpref(ip, lbn, indx, bap)
546 	struct inode *ip;
547 	daddr_t lbn;
548 	int indx;
549 	daddr_t *bap;
550 {
551 	register struct fs *fs;
552 	register int cg;
553 	int avgbfree, startcg;
554 	daddr_t nextblk;
555 
556 	fs = ip->i_fs;
557 	if (indx % fs->fs_maxbpg == 0 || bap[indx - 1] == 0) {
558 		if (lbn < NDADDR) {
559 			cg = ino_to_cg(fs, ip->i_number);
560 			return (fs->fs_fpg * cg + fs->fs_frag);
561 		}
562 		/*
563 		 * Find a cylinder with greater than average number of
564 		 * unused data blocks.
565 		 */
566 		if (indx == 0 || bap[indx - 1] == 0)
567 			startcg =
568 			    ino_to_cg(fs, ip->i_number) + lbn / fs->fs_maxbpg;
569 		else
570 			startcg = dtog(fs, bap[indx - 1]) + 1;
571 		startcg %= fs->fs_ncg;
572 		avgbfree = fs->fs_cstotal.cs_nbfree / fs->fs_ncg;
573 		for (cg = startcg; cg < fs->fs_ncg; cg++)
574 			if (fs->fs_cs(fs, cg).cs_nbfree >= avgbfree) {
575 				fs->fs_cgrotor = cg;
576 				return (fs->fs_fpg * cg + fs->fs_frag);
577 			}
578 		for (cg = 0; cg <= startcg; cg++)
579 			if (fs->fs_cs(fs, cg).cs_nbfree >= avgbfree) {
580 				fs->fs_cgrotor = cg;
581 				return (fs->fs_fpg * cg + fs->fs_frag);
582 			}
583 		return (NULL);
584 	}
585 	/*
586 	 * One or more previous blocks have been laid out. If less
587 	 * than fs_maxcontig previous blocks are contiguous, the
588 	 * next block is requested contiguously, otherwise it is
589 	 * requested rotationally delayed by fs_rotdelay milliseconds.
590 	 */
591 	nextblk = bap[indx - 1] + fs->fs_frag;
592 	if (indx < fs->fs_maxcontig || bap[indx - fs->fs_maxcontig] +
593 	    blkstofrags(fs, fs->fs_maxcontig) != nextblk)
594 		return (nextblk);
595 	if (fs->fs_rotdelay != 0)
596 		/*
597 		 * Here we convert ms of delay to frags as:
598 		 * (frags) = (ms) * (rev/sec) * (sect/rev) /
599 		 *	((sect/frag) * (ms/sec))
600 		 * then round up to the next block.
601 		 */
602 		nextblk += roundup(fs->fs_rotdelay * fs->fs_rps * fs->fs_nsect /
603 		    (NSPF(fs) * 1000), fs->fs_frag);
604 	return (nextblk);
605 }
606 
607 /*
608  * Implement the cylinder overflow algorithm.
609  *
610  * The policy implemented by this algorithm is:
611  *   1) allocate the block in its requested cylinder group.
612  *   2) quadradically rehash on the cylinder group number.
613  *   3) brute force search for a free block.
614  */
615 /*VARARGS5*/
616 static u_long
617 ffs_hashalloc(ip, cg, pref, size, allocator)
618 	struct inode *ip;
619 	int cg;
620 	long pref;
621 	int size;	/* size for data blocks, mode for inodes */
622 	u_long (*allocator)();
623 {
624 	register struct fs *fs;
625 	long result;
626 	int i, icg = cg;
627 
628 	fs = ip->i_fs;
629 	/*
630 	 * 1: preferred cylinder group
631 	 */
632 	result = (*allocator)(ip, cg, pref, size);
633 	if (result)
634 		return (result);
635 	/*
636 	 * 2: quadratic rehash
637 	 */
638 	for (i = 1; i < fs->fs_ncg; i *= 2) {
639 		cg += i;
640 		if (cg >= fs->fs_ncg)
641 			cg -= fs->fs_ncg;
642 		result = (*allocator)(ip, cg, 0, size);
643 		if (result)
644 			return (result);
645 	}
646 	/*
647 	 * 3: brute force search
648 	 * Note that we start at i == 2, since 0 was checked initially,
649 	 * and 1 is always checked in the quadratic rehash.
650 	 */
651 	cg = (icg + 2) % fs->fs_ncg;
652 	for (i = 2; i < fs->fs_ncg; i++) {
653 		result = (*allocator)(ip, cg, 0, size);
654 		if (result)
655 			return (result);
656 		cg++;
657 		if (cg == fs->fs_ncg)
658 			cg = 0;
659 	}
660 	return (NULL);
661 }
662 
663 /*
664  * Determine whether a fragment can be extended.
665  *
666  * Check to see if the necessary fragments are available, and
667  * if they are, allocate them.
668  */
669 static daddr_t
670 ffs_fragextend(ip, cg, bprev, osize, nsize)
671 	struct inode *ip;
672 	int cg;
673 	long bprev;
674 	int osize, nsize;
675 {
676 	register struct fs *fs;
677 	register struct cg *cgp;
678 	struct buf *bp;
679 	long bno;
680 	int frags, bbase;
681 	int i, error;
682 
683 	fs = ip->i_fs;
684 	if (fs->fs_cs(fs, cg).cs_nffree < numfrags(fs, nsize - osize))
685 		return (NULL);
686 	frags = numfrags(fs, nsize);
687 	bbase = fragnum(fs, bprev);
688 	if (bbase > fragnum(fs, (bprev + frags - 1))) {
689 		/* cannot extend across a block boundary */
690 		return (NULL);
691 	}
692 	error = bread(ip->i_devvp, fsbtodb(fs, cgtod(fs, cg)),
693 		(int)fs->fs_cgsize, NOCRED, &bp);
694 	if (error) {
695 		brelse(bp);
696 		return (NULL);
697 	}
698 	cgp = (struct cg *)bp->b_data;
699 	if (!cg_chkmagic(cgp)) {
700 		brelse(bp);
701 		return (NULL);
702 	}
703 	cgp->cg_time = time.tv_sec;
704 	bno = dtogd(fs, bprev);
705 	for (i = numfrags(fs, osize); i < frags; i++)
706 		if (isclr(cg_blksfree(cgp), bno + i)) {
707 			brelse(bp);
708 			return (NULL);
709 		}
710 	/*
711 	 * the current fragment can be extended
712 	 * deduct the count on fragment being extended into
713 	 * increase the count on the remaining fragment (if any)
714 	 * allocate the extended piece
715 	 */
716 	for (i = frags; i < fs->fs_frag - bbase; i++)
717 		if (isclr(cg_blksfree(cgp), bno + i))
718 			break;
719 	cgp->cg_frsum[i - numfrags(fs, osize)]--;
720 	if (i != frags)
721 		cgp->cg_frsum[i - frags]++;
722 	for (i = numfrags(fs, osize); i < frags; i++) {
723 		clrbit(cg_blksfree(cgp), bno + i);
724 		cgp->cg_cs.cs_nffree--;
725 		fs->fs_cstotal.cs_nffree--;
726 		fs->fs_cs(fs, cg).cs_nffree--;
727 	}
728 	fs->fs_fmod = 1;
729 	bdwrite(bp);
730 	return (bprev);
731 }
732 
733 /*
734  * Determine whether a block can be allocated.
735  *
736  * Check to see if a block of the appropriate size is available,
737  * and if it is, allocate it.
738  */
739 static daddr_t
740 ffs_alloccg(ip, cg, bpref, size)
741 	struct inode *ip;
742 	int cg;
743 	daddr_t bpref;
744 	int size;
745 {
746 	register struct fs *fs;
747 	register struct cg *cgp;
748 	struct buf *bp;
749 	register int i;
750 	int error, bno, frags, allocsiz;
751 
752 	fs = ip->i_fs;
753 	if (fs->fs_cs(fs, cg).cs_nbfree == 0 && size == fs->fs_bsize)
754 		return (NULL);
755 	error = bread(ip->i_devvp, fsbtodb(fs, cgtod(fs, cg)),
756 		(int)fs->fs_cgsize, NOCRED, &bp);
757 	if (error) {
758 		brelse(bp);
759 		return (NULL);
760 	}
761 	cgp = (struct cg *)bp->b_data;
762 	if (!cg_chkmagic(cgp) ||
763 	    (cgp->cg_cs.cs_nbfree == 0 && size == fs->fs_bsize)) {
764 		brelse(bp);
765 		return (NULL);
766 	}
767 	cgp->cg_time = time.tv_sec;
768 	if (size == fs->fs_bsize) {
769 		bno = ffs_alloccgblk(fs, cgp, bpref);
770 		bdwrite(bp);
771 		return (bno);
772 	}
773 	/*
774 	 * check to see if any fragments are already available
775 	 * allocsiz is the size which will be allocated, hacking
776 	 * it down to a smaller size if necessary
777 	 */
778 	frags = numfrags(fs, size);
779 	for (allocsiz = frags; allocsiz < fs->fs_frag; allocsiz++)
780 		if (cgp->cg_frsum[allocsiz] != 0)
781 			break;
782 	if (allocsiz == fs->fs_frag) {
783 		/*
784 		 * no fragments were available, so a block will be
785 		 * allocated, and hacked up
786 		 */
787 		if (cgp->cg_cs.cs_nbfree == 0) {
788 			brelse(bp);
789 			return (NULL);
790 		}
791 		bno = ffs_alloccgblk(fs, cgp, bpref);
792 		bpref = dtogd(fs, bno);
793 		for (i = frags; i < fs->fs_frag; i++)
794 			setbit(cg_blksfree(cgp), bpref + i);
795 		i = fs->fs_frag - frags;
796 		cgp->cg_cs.cs_nffree += i;
797 		fs->fs_cstotal.cs_nffree += i;
798 		fs->fs_cs(fs, cg).cs_nffree += i;
799 		fs->fs_fmod = 1;
800 		cgp->cg_frsum[i]++;
801 		bdwrite(bp);
802 		return (bno);
803 	}
804 	bno = ffs_mapsearch(fs, cgp, bpref, allocsiz);
805 	if (bno < 0) {
806 		brelse(bp);
807 		return (NULL);
808 	}
809 	for (i = 0; i < frags; i++)
810 		clrbit(cg_blksfree(cgp), bno + i);
811 	cgp->cg_cs.cs_nffree -= frags;
812 	fs->fs_cstotal.cs_nffree -= frags;
813 	fs->fs_cs(fs, cg).cs_nffree -= frags;
814 	fs->fs_fmod = 1;
815 	cgp->cg_frsum[allocsiz]--;
816 	if (frags != allocsiz)
817 		cgp->cg_frsum[allocsiz - frags]++;
818 	bdwrite(bp);
819 	return (cg * fs->fs_fpg + bno);
820 }
821 
822 /*
823  * Allocate a block in a cylinder group.
824  *
825  * This algorithm implements the following policy:
826  *   1) allocate the requested block.
827  *   2) allocate a rotationally optimal block in the same cylinder.
828  *   3) allocate the next available block on the block rotor for the
829  *      specified cylinder group.
830  * Note that this routine only allocates fs_bsize blocks; these
831  * blocks may be fragmented by the routine that allocates them.
832  */
833 static daddr_t
834 ffs_alloccgblk(fs, cgp, bpref)
835 	register struct fs *fs;
836 	register struct cg *cgp;
837 	daddr_t bpref;
838 {
839 	daddr_t bno, blkno;
840 	int cylno, pos, delta;
841 	short *cylbp;
842 	register int i;
843 
844 	if (bpref == 0 || dtog(fs, bpref) != cgp->cg_cgx) {
845 		bpref = cgp->cg_rotor;
846 		goto norot;
847 	}
848 	bpref = blknum(fs, bpref);
849 	bpref = dtogd(fs, bpref);
850 	/*
851 	 * if the requested block is available, use it
852 	 */
853 	if (ffs_isblock(fs, cg_blksfree(cgp), fragstoblks(fs, bpref))) {
854 		bno = bpref;
855 		goto gotit;
856 	}
857 	/*
858 	 * check for a block available on the same cylinder
859 	 */
860 	cylno = cbtocylno(fs, bpref);
861 	if (cg_blktot(cgp)[cylno] == 0)
862 		goto norot;
863 	if (fs->fs_cpc == 0) {
864 		/*
865 		 * Block layout information is not available.
866 		 * Leaving bpref unchanged means we take the
867 		 * next available free block following the one
868 		 * we just allocated. Hopefully this will at
869 		 * least hit a track cache on drives of unknown
870 		 * geometry (e.g. SCSI).
871 		 */
872 		goto norot;
873 	}
874 	/*
875 	 * check the summary information to see if a block is
876 	 * available in the requested cylinder starting at the
877 	 * requested rotational position and proceeding around.
878 	 */
879 	cylbp = cg_blks(fs, cgp, cylno);
880 	pos = cbtorpos(fs, bpref);
881 	for (i = pos; i < fs->fs_nrpos; i++)
882 		if (cylbp[i] > 0)
883 			break;
884 	if (i == fs->fs_nrpos)
885 		for (i = 0; i < pos; i++)
886 			if (cylbp[i] > 0)
887 				break;
888 	if (cylbp[i] > 0) {
889 		/*
890 		 * found a rotational position, now find the actual
891 		 * block. A panic if none is actually there.
892 		 */
893 		pos = cylno % fs->fs_cpc;
894 		bno = (cylno - pos) * fs->fs_spc / NSPB(fs);
895 		if (fs_postbl(fs, pos)[i] == -1) {
896 			printf("pos = %d, i = %d, fs = %s\n",
897 			    pos, i, fs->fs_fsmnt);
898 			panic("ffs_alloccgblk: cyl groups corrupted");
899 		}
900 		for (i = fs_postbl(fs, pos)[i];; ) {
901 			if (ffs_isblock(fs, cg_blksfree(cgp), bno + i)) {
902 				bno = blkstofrags(fs, (bno + i));
903 				goto gotit;
904 			}
905 			delta = fs_rotbl(fs)[i];
906 			if (delta <= 0 ||
907 			    delta + i > fragstoblks(fs, fs->fs_fpg))
908 				break;
909 			i += delta;
910 		}
911 		printf("pos = %d, i = %d, fs = %s\n", pos, i, fs->fs_fsmnt);
912 		panic("ffs_alloccgblk: can't find blk in cyl");
913 	}
914 norot:
915 	/*
916 	 * no blocks in the requested cylinder, so take next
917 	 * available one in this cylinder group.
918 	 */
919 	bno = ffs_mapsearch(fs, cgp, bpref, (int)fs->fs_frag);
920 	if (bno < 0)
921 		return (NULL);
922 	cgp->cg_rotor = bno;
923 gotit:
924 	blkno = fragstoblks(fs, bno);
925 	ffs_clrblock(fs, cg_blksfree(cgp), (long)blkno);
926 	ffs_clusteracct(fs, cgp, blkno, -1);
927 	cgp->cg_cs.cs_nbfree--;
928 	fs->fs_cstotal.cs_nbfree--;
929 	fs->fs_cs(fs, cgp->cg_cgx).cs_nbfree--;
930 	cylno = cbtocylno(fs, bno);
931 	cg_blks(fs, cgp, cylno)[cbtorpos(fs, bno)]--;
932 	cg_blktot(cgp)[cylno]--;
933 	fs->fs_fmod = 1;
934 	return (cgp->cg_cgx * fs->fs_fpg + bno);
935 }
936 
937 /*
938  * Determine whether a cluster can be allocated.
939  *
940  * We do not currently check for optimal rotational layout if there
941  * are multiple choices in the same cylinder group. Instead we just
942  * take the first one that we find following bpref.
943  */
944 static daddr_t
945 ffs_clusteralloc(ip, cg, bpref, len)
946 	struct inode *ip;
947 	int cg;
948 	daddr_t bpref;
949 	int len;
950 {
951 	register struct fs *fs;
952 	register struct cg *cgp;
953 	struct buf *bp;
954 	int i, run, bno, bit, map;
955 	u_char *mapp;
956 
957 	fs = ip->i_fs;
958 	if (fs->fs_cs(fs, cg).cs_nbfree < len)
959 		return (NULL);
960 	if (bread(ip->i_devvp, fsbtodb(fs, cgtod(fs, cg)), (int)fs->fs_cgsize,
961 	    NOCRED, &bp))
962 		goto fail;
963 	cgp = (struct cg *)bp->b_data;
964 	if (!cg_chkmagic(cgp))
965 		goto fail;
966 	/*
967 	 * Check to see if a cluster of the needed size (or bigger) is
968 	 * available in this cylinder group.
969 	 */
970 	for (i = len; i <= fs->fs_contigsumsize; i++)
971 		if (cg_clustersum(cgp)[i] > 0)
972 			break;
973 	if (i > fs->fs_contigsumsize)
974 		goto fail;
975 	/*
976 	 * Search the cluster map to find a big enough cluster.
977 	 * We take the first one that we find, even if it is larger
978 	 * than we need as we prefer to get one close to the previous
979 	 * block allocation. We do not search before the current
980 	 * preference point as we do not want to allocate a block
981 	 * that is allocated before the previous one (as we will
982 	 * then have to wait for another pass of the elevator
983 	 * algorithm before it will be read). We prefer to fail and
984 	 * be recalled to try an allocation in the next cylinder group.
985 	 */
986 	if (dtog(fs, bpref) != cg)
987 		bpref = 0;
988 	else
989 		bpref = fragstoblks(fs, dtogd(fs, blknum(fs, bpref)));
990 	mapp = &cg_clustersfree(cgp)[bpref / NBBY];
991 	map = *mapp++;
992 	bit = 1 << (bpref % NBBY);
993 	for (run = 0, i = bpref; i < cgp->cg_nclusterblks; i++) {
994 		if ((map & bit) == 0) {
995 			run = 0;
996 		} else {
997 			run++;
998 			if (run == len)
999 				break;
1000 		}
1001 		if ((i & (NBBY - 1)) != (NBBY - 1)) {
1002 			bit <<= 1;
1003 		} else {
1004 			map = *mapp++;
1005 			bit = 1;
1006 		}
1007 	}
1008 	if (i == cgp->cg_nclusterblks)
1009 		goto fail;
1010 	/*
1011 	 * Allocate the cluster that we have found.
1012 	 */
1013 	bno = cg * fs->fs_fpg + blkstofrags(fs, i - run + 1);
1014 	len = blkstofrags(fs, len);
1015 	for (i = 0; i < len; i += fs->fs_frag)
1016 		if (ffs_alloccgblk(fs, cgp, bno + i) != bno + i)
1017 			panic("ffs_clusteralloc: lost block");
1018 	brelse(bp);
1019 	return (bno);
1020 
1021 fail:
1022 	brelse(bp);
1023 	return (0);
1024 }
1025 
1026 /*
1027  * Determine whether an inode can be allocated.
1028  *
1029  * Check to see if an inode is available, and if it is,
1030  * allocate it using the following policy:
1031  *   1) allocate the requested inode.
1032  *   2) allocate the next available inode after the requested
1033  *      inode in the specified cylinder group.
1034  */
1035 static ino_t
1036 ffs_nodealloccg(ip, cg, ipref, mode)
1037 	struct inode *ip;
1038 	int cg;
1039 	daddr_t ipref;
1040 	int mode;
1041 {
1042 	register struct fs *fs;
1043 	register struct cg *cgp;
1044 	struct buf *bp;
1045 	int error, start, len, loc, map, i;
1046 
1047 	fs = ip->i_fs;
1048 	if (fs->fs_cs(fs, cg).cs_nifree == 0)
1049 		return (NULL);
1050 	error = bread(ip->i_devvp, fsbtodb(fs, cgtod(fs, cg)),
1051 		(int)fs->fs_cgsize, NOCRED, &bp);
1052 	if (error) {
1053 		brelse(bp);
1054 		return (NULL);
1055 	}
1056 	cgp = (struct cg *)bp->b_data;
1057 	if (!cg_chkmagic(cgp) || cgp->cg_cs.cs_nifree == 0) {
1058 		brelse(bp);
1059 		return (NULL);
1060 	}
1061 	cgp->cg_time = time.tv_sec;
1062 	if (ipref) {
1063 		ipref %= fs->fs_ipg;
1064 		if (isclr(cg_inosused(cgp), ipref))
1065 			goto gotit;
1066 	}
1067 	start = cgp->cg_irotor / NBBY;
1068 	len = howmany(fs->fs_ipg - cgp->cg_irotor, NBBY);
1069 	loc = skpc(0xff, len, &cg_inosused(cgp)[start]);
1070 	if (loc == 0) {
1071 		len = start + 1;
1072 		start = 0;
1073 		loc = skpc(0xff, len, &cg_inosused(cgp)[0]);
1074 		if (loc == 0) {
1075 			printf("cg = %d, irotor = %d, fs = %s\n",
1076 			    cg, cgp->cg_irotor, fs->fs_fsmnt);
1077 			panic("ffs_nodealloccg: map corrupted");
1078 			/* NOTREACHED */
1079 		}
1080 	}
1081 	i = start + len - loc;
1082 	map = cg_inosused(cgp)[i];
1083 	ipref = i * NBBY;
1084 	for (i = 1; i < (1 << NBBY); i <<= 1, ipref++) {
1085 		if ((map & i) == 0) {
1086 			cgp->cg_irotor = ipref;
1087 			goto gotit;
1088 		}
1089 	}
1090 	printf("fs = %s\n", fs->fs_fsmnt);
1091 	panic("ffs_nodealloccg: block not in map");
1092 	/* NOTREACHED */
1093 gotit:
1094 	setbit(cg_inosused(cgp), ipref);
1095 	cgp->cg_cs.cs_nifree--;
1096 	fs->fs_cstotal.cs_nifree--;
1097 	fs->fs_cs(fs, cg).cs_nifree--;
1098 	fs->fs_fmod = 1;
1099 	if ((mode & IFMT) == IFDIR) {
1100 		cgp->cg_cs.cs_ndir++;
1101 		fs->fs_cstotal.cs_ndir++;
1102 		fs->fs_cs(fs, cg).cs_ndir++;
1103 	}
1104 	bdwrite(bp);
1105 	return (cg * fs->fs_ipg + ipref);
1106 }
1107 
1108 /*
1109  * Free a block or fragment.
1110  *
1111  * The specified block or fragment is placed back in the
1112  * free map. If a fragment is deallocated, a possible
1113  * block reassembly is checked.
1114  */
1115 ffs_blkfree(ip, bno, size)
1116 	register struct inode *ip;
1117 	daddr_t bno;
1118 	long size;
1119 {
1120 	register struct fs *fs;
1121 	register struct cg *cgp;
1122 	struct buf *bp;
1123 	daddr_t blkno;
1124 	int i, error, cg, blk, frags, bbase;
1125 
1126 	fs = ip->i_fs;
1127 	if ((u_int)size > fs->fs_bsize || fragoff(fs, size) != 0) {
1128 		printf("dev = 0x%x, bsize = %d, size = %d, fs = %s\n",
1129 		    ip->i_dev, fs->fs_bsize, size, fs->fs_fsmnt);
1130 		panic("blkfree: bad size");
1131 	}
1132 	cg = dtog(fs, bno);
1133 	if ((u_int)bno >= fs->fs_size) {
1134 		printf("bad block %d, ino %d\n", bno, ip->i_number);
1135 		ffs_fserr(fs, ip->i_uid, "bad block");
1136 		return;
1137 	}
1138 	error = bread(ip->i_devvp, fsbtodb(fs, cgtod(fs, cg)),
1139 		(int)fs->fs_cgsize, NOCRED, &bp);
1140 	if (error) {
1141 		brelse(bp);
1142 		return;
1143 	}
1144 	cgp = (struct cg *)bp->b_data;
1145 	if (!cg_chkmagic(cgp)) {
1146 		brelse(bp);
1147 		return;
1148 	}
1149 	cgp->cg_time = time.tv_sec;
1150 	bno = dtogd(fs, bno);
1151 	if (size == fs->fs_bsize) {
1152 		blkno = fragstoblks(fs, bno);
1153 		if (ffs_isblock(fs, cg_blksfree(cgp), blkno)) {
1154 			printf("dev = 0x%x, block = %d, fs = %s\n",
1155 			    ip->i_dev, bno, fs->fs_fsmnt);
1156 			panic("blkfree: freeing free block");
1157 		}
1158 		ffs_setblock(fs, cg_blksfree(cgp), blkno);
1159 		ffs_clusteracct(fs, cgp, blkno, 1);
1160 		cgp->cg_cs.cs_nbfree++;
1161 		fs->fs_cstotal.cs_nbfree++;
1162 		fs->fs_cs(fs, cg).cs_nbfree++;
1163 		i = cbtocylno(fs, bno);
1164 		cg_blks(fs, cgp, i)[cbtorpos(fs, bno)]++;
1165 		cg_blktot(cgp)[i]++;
1166 	} else {
1167 		bbase = bno - fragnum(fs, bno);
1168 		/*
1169 		 * decrement the counts associated with the old frags
1170 		 */
1171 		blk = blkmap(fs, cg_blksfree(cgp), bbase);
1172 		ffs_fragacct(fs, blk, cgp->cg_frsum, -1);
1173 		/*
1174 		 * deallocate the fragment
1175 		 */
1176 		frags = numfrags(fs, size);
1177 		for (i = 0; i < frags; i++) {
1178 			if (isset(cg_blksfree(cgp), bno + i)) {
1179 				printf("dev = 0x%x, block = %d, fs = %s\n",
1180 				    ip->i_dev, bno + i, fs->fs_fsmnt);
1181 				panic("blkfree: freeing free frag");
1182 			}
1183 			setbit(cg_blksfree(cgp), bno + i);
1184 		}
1185 		cgp->cg_cs.cs_nffree += i;
1186 		fs->fs_cstotal.cs_nffree += i;
1187 		fs->fs_cs(fs, cg).cs_nffree += i;
1188 		/*
1189 		 * add back in counts associated with the new frags
1190 		 */
1191 		blk = blkmap(fs, cg_blksfree(cgp), bbase);
1192 		ffs_fragacct(fs, blk, cgp->cg_frsum, 1);
1193 		/*
1194 		 * if a complete block has been reassembled, account for it
1195 		 */
1196 		blkno = fragstoblks(fs, bbase);
1197 		if (ffs_isblock(fs, cg_blksfree(cgp), blkno)) {
1198 			cgp->cg_cs.cs_nffree -= fs->fs_frag;
1199 			fs->fs_cstotal.cs_nffree -= fs->fs_frag;
1200 			fs->fs_cs(fs, cg).cs_nffree -= fs->fs_frag;
1201 			ffs_clusteracct(fs, cgp, blkno, 1);
1202 			cgp->cg_cs.cs_nbfree++;
1203 			fs->fs_cstotal.cs_nbfree++;
1204 			fs->fs_cs(fs, cg).cs_nbfree++;
1205 			i = cbtocylno(fs, bbase);
1206 			cg_blks(fs, cgp, i)[cbtorpos(fs, bbase)]++;
1207 			cg_blktot(cgp)[i]++;
1208 		}
1209 	}
1210 	fs->fs_fmod = 1;
1211 	bdwrite(bp);
1212 }
1213 
1214 /*
1215  * Free an inode.
1216  *
1217  * The specified inode is placed back in the free map.
1218  */
1219 int
1220 ffs_vfree(ap)
1221 	struct vop_vfree_args /* {
1222 		struct vnode *a_pvp;
1223 		ino_t a_ino;
1224 		int a_mode;
1225 	} */ *ap;
1226 {
1227 	register struct fs *fs;
1228 	register struct cg *cgp;
1229 	register struct inode *pip;
1230 	ino_t ino = ap->a_ino;
1231 	struct buf *bp;
1232 	int error, cg;
1233 
1234 	pip = VTOI(ap->a_pvp);
1235 	fs = pip->i_fs;
1236 	if ((u_int)ino >= fs->fs_ipg * fs->fs_ncg)
1237 		panic("ifree: range: dev = 0x%x, ino = %d, fs = %s\n",
1238 		    pip->i_dev, ino, fs->fs_fsmnt);
1239 	cg = ino_to_cg(fs, ino);
1240 	error = bread(pip->i_devvp, fsbtodb(fs, cgtod(fs, cg)),
1241 		(int)fs->fs_cgsize, NOCRED, &bp);
1242 	if (error) {
1243 		brelse(bp);
1244 		return (0);
1245 	}
1246 	cgp = (struct cg *)bp->b_data;
1247 	if (!cg_chkmagic(cgp)) {
1248 		brelse(bp);
1249 		return (0);
1250 	}
1251 	cgp->cg_time = time.tv_sec;
1252 	ino %= fs->fs_ipg;
1253 	if (isclr(cg_inosused(cgp), ino)) {
1254 		printf("dev = 0x%x, ino = %d, fs = %s\n",
1255 		    pip->i_dev, ino, fs->fs_fsmnt);
1256 		if (fs->fs_ronly == 0)
1257 			panic("ifree: freeing free inode");
1258 	}
1259 	clrbit(cg_inosused(cgp), ino);
1260 	if (ino < cgp->cg_irotor)
1261 		cgp->cg_irotor = ino;
1262 	cgp->cg_cs.cs_nifree++;
1263 	fs->fs_cstotal.cs_nifree++;
1264 	fs->fs_cs(fs, cg).cs_nifree++;
1265 	if ((ap->a_mode & IFMT) == IFDIR) {
1266 		cgp->cg_cs.cs_ndir--;
1267 		fs->fs_cstotal.cs_ndir--;
1268 		fs->fs_cs(fs, cg).cs_ndir--;
1269 	}
1270 	fs->fs_fmod = 1;
1271 	bdwrite(bp);
1272 	return (0);
1273 }
1274 
1275 /*
1276  * Find a block of the specified size in the specified cylinder group.
1277  *
1278  * It is a panic if a request is made to find a block if none are
1279  * available.
1280  */
1281 static daddr_t
1282 ffs_mapsearch(fs, cgp, bpref, allocsiz)
1283 	register struct fs *fs;
1284 	register struct cg *cgp;
1285 	daddr_t bpref;
1286 	int allocsiz;
1287 {
1288 	daddr_t bno;
1289 	int start, len, loc, i;
1290 	int blk, field, subfield, pos;
1291 
1292 	/*
1293 	 * find the fragment by searching through the free block
1294 	 * map for an appropriate bit pattern
1295 	 */
1296 	if (bpref)
1297 		start = dtogd(fs, bpref) / NBBY;
1298 	else
1299 		start = cgp->cg_frotor / NBBY;
1300 	len = howmany(fs->fs_fpg, NBBY) - start;
1301 	loc = scanc((u_int)len, (u_char *)&cg_blksfree(cgp)[start],
1302 		(u_char *)fragtbl[fs->fs_frag],
1303 		(u_char)(1 << (allocsiz - 1 + (fs->fs_frag % NBBY))));
1304 	if (loc == 0) {
1305 		len = start + 1;
1306 		start = 0;
1307 		loc = scanc((u_int)len, (u_char *)&cg_blksfree(cgp)[0],
1308 			(u_char *)fragtbl[fs->fs_frag],
1309 			(u_char)(1 << (allocsiz - 1 + (fs->fs_frag % NBBY))));
1310 		if (loc == 0) {
1311 			printf("start = %d, len = %d, fs = %s\n",
1312 			    start, len, fs->fs_fsmnt);
1313 			panic("ffs_alloccg: map corrupted");
1314 			/* NOTREACHED */
1315 		}
1316 	}
1317 	bno = (start + len - loc) * NBBY;
1318 	cgp->cg_frotor = bno;
1319 	/*
1320 	 * found the byte in the map
1321 	 * sift through the bits to find the selected frag
1322 	 */
1323 	for (i = bno + NBBY; bno < i; bno += fs->fs_frag) {
1324 		blk = blkmap(fs, cg_blksfree(cgp), bno);
1325 		blk <<= 1;
1326 		field = around[allocsiz];
1327 		subfield = inside[allocsiz];
1328 		for (pos = 0; pos <= fs->fs_frag - allocsiz; pos++) {
1329 			if ((blk & field) == subfield)
1330 				return (bno + pos);
1331 			field <<= 1;
1332 			subfield <<= 1;
1333 		}
1334 	}
1335 	printf("bno = %d, fs = %s\n", bno, fs->fs_fsmnt);
1336 	panic("ffs_alloccg: block not in map");
1337 	return (-1);
1338 }
1339 
1340 /*
1341  * Update the cluster map because of an allocation or free.
1342  *
1343  * Cnt == 1 means free; cnt == -1 means allocating.
1344  */
1345 ffs_clusteracct(fs, cgp, blkno, cnt)
1346 	struct fs *fs;
1347 	struct cg *cgp;
1348 	daddr_t blkno;
1349 	int cnt;
1350 {
1351 	long *sump;
1352 	u_char *freemapp, *mapp;
1353 	int i, start, end, forw, back, map, bit;
1354 
1355 	if (fs->fs_contigsumsize <= 0)
1356 		return;
1357 	freemapp = cg_clustersfree(cgp);
1358 	sump = cg_clustersum(cgp);
1359 	/*
1360 	 * Allocate or clear the actual block.
1361 	 */
1362 	if (cnt > 0)
1363 		setbit(freemapp, blkno);
1364 	else
1365 		clrbit(freemapp, blkno);
1366 	/*
1367 	 * Find the size of the cluster going forward.
1368 	 */
1369 	start = blkno + 1;
1370 	end = start + fs->fs_contigsumsize;
1371 	if (end >= cgp->cg_nclusterblks)
1372 		end = cgp->cg_nclusterblks;
1373 	mapp = &freemapp[start / NBBY];
1374 	map = *mapp++;
1375 	bit = 1 << (start % NBBY);
1376 	for (i = start; i < end; i++) {
1377 		if ((map & bit) == 0)
1378 			break;
1379 		if ((i & (NBBY - 1)) != (NBBY - 1)) {
1380 			bit <<= 1;
1381 		} else {
1382 			map = *mapp++;
1383 			bit = 1;
1384 		}
1385 	}
1386 	forw = i - start;
1387 	/*
1388 	 * Find the size of the cluster going backward.
1389 	 */
1390 	start = blkno - 1;
1391 	end = start - fs->fs_contigsumsize;
1392 	if (end < 0)
1393 		end = -1;
1394 	mapp = &freemapp[start / NBBY];
1395 	map = *mapp--;
1396 	bit = 1 << (start % NBBY);
1397 	for (i = start; i > end; i--) {
1398 		if ((map & bit) == 0)
1399 			break;
1400 		if ((i & (NBBY - 1)) != 0) {
1401 			bit >>= 1;
1402 		} else {
1403 			map = *mapp--;
1404 			bit = 1 << (NBBY - 1);
1405 		}
1406 	}
1407 	back = start - i;
1408 	/*
1409 	 * Account for old cluster and the possibly new forward and
1410 	 * back clusters.
1411 	 */
1412 	i = back + forw + 1;
1413 	if (i > fs->fs_contigsumsize)
1414 		i = fs->fs_contigsumsize;
1415 	sump[i] += cnt;
1416 	if (back > 0)
1417 		sump[back] -= cnt;
1418 	if (forw > 0)
1419 		sump[forw] -= cnt;
1420 }
1421 
1422 /*
1423  * Fserr prints the name of a file system with an error diagnostic.
1424  *
1425  * The form of the error message is:
1426  *	fs: error message
1427  */
1428 static void
1429 ffs_fserr(fs, uid, cp)
1430 	struct fs *fs;
1431 	u_int uid;
1432 	char *cp;
1433 {
1434 
1435 	log(LOG_ERR, "uid %d on %s: %s\n", uid, fs->fs_fsmnt, cp);
1436 }
1437