37 */
38
39#include <sys/param.h>
40#include <sys/systm.h>
41#include <sys/conf.h>
42#include <sys/vnode.h>
43#include <sys/stat.h>
44#include <sys/mount.h>
45#include <sys/sysctl.h>
46#include <sys/syslog.h>
47#include <sys/buf.h>
48
49#include <fs/ext2fs/inode.h>
50#include <fs/ext2fs/ext2_mount.h>
51#include <fs/ext2fs/ext2fs.h>
52#include <fs/ext2fs/fs.h>
53#include <fs/ext2fs/ext2_extern.h>
54
55static daddr_t ext2_alloccg(struct inode *, int, daddr_t, int);
56static daddr_t ext2_clusteralloc(struct inode *, int, daddr_t, int);
57static u_long ext2_dirpref(struct inode *);
58static void ext2_fserr(struct m_ext2fs *, uid_t, char *);
59static u_long ext2_hashalloc(struct inode *, int, long, int,
60 daddr_t (*)(struct inode *, int, daddr_t,
61 int));
62static daddr_t ext2_nodealloccg(struct inode *, int, daddr_t, int);
63static daddr_t ext2_mapsearch(struct m_ext2fs *, char *, daddr_t);
64
65/*
66 * Allocate a block in the file system.
67 *
68 * A preference may be optionally specified. If a preference is given
69 * the following hierarchy is used to allocate a block:
70 * 1) allocate the requested block.
71 * 2) allocate a rotationally optimal block in the same cylinder.
72 * 3) allocate a block in the same cylinder group.
73 * 4) quadradically rehash into other cylinder groups, until an
74 * available block is located.
75 * If no block preference is given the following hierarchy is used
76 * to allocate a block:
77 * 1) allocate a block in the cylinder group that contains the
78 * inode for the file.
79 * 2) quadradically rehash into other cylinder groups, until an
80 * available block is located.
81 */
82int
83ext2_alloc(ip, lbn, bpref, size, cred, bnp)
84 struct inode *ip;
85 int32_t lbn, bpref;
86 int size;
87 struct ucred *cred;
88 int32_t *bnp;
89{
90 struct m_ext2fs *fs;
91 struct ext2mount *ump;
92 int32_t bno;
93 int cg;
94 *bnp = 0;
95 fs = ip->i_e2fs;
96 ump = ip->i_ump;
97 mtx_assert(EXT2_MTX(ump), MA_OWNED);
98#ifdef DIAGNOSTIC
99 if ((u_int)size > fs->e2fs_bsize || blkoff(fs, size) != 0) {
100 vn_printf(ip->i_devvp, "bsize = %lu, size = %d, fs = %s\n",
101 (long unsigned int)fs->e2fs_bsize, size, fs->e2fs_fsmnt);
102 panic("ext2_alloc: bad size");
103 }
104 if (cred == NOCRED)
105 panic("ext2_alloc: missing credential");
106#endif /* DIAGNOSTIC */
107 if (size == fs->e2fs_bsize && fs->e2fs->e2fs_fbcount == 0)
108 goto nospace;
109 if (cred->cr_uid != 0 &&
110 fs->e2fs->e2fs_fbcount < fs->e2fs->e2fs_rbcount)
111 goto nospace;
112 if (bpref >= fs->e2fs->e2fs_bcount)
113 bpref = 0;
114 if (bpref == 0)
115 cg = ino_to_cg(fs, ip->i_number);
116 else
117 cg = dtog(fs, bpref);
118 bno = (daddr_t)ext2_hashalloc(ip, cg, bpref, fs->e2fs_bsize,
119 ext2_alloccg);
120 if (bno > 0) {
121 /* set next_alloc fields as done in block_getblk */
122 ip->i_next_alloc_block = lbn;
123 ip->i_next_alloc_goal = bno;
124
125 ip->i_blocks += btodb(fs->e2fs_bsize);
126 ip->i_flag |= IN_CHANGE | IN_UPDATE;
127 *bnp = bno;
128 return (0);
129 }
130nospace:
131 EXT2_UNLOCK(ump);
132 ext2_fserr(fs, cred->cr_uid, "file system full");
133 uprintf("\n%s: write failed, file system is full\n", fs->e2fs_fsmnt);
134 return (ENOSPC);
135}
136
137/*
138 * Reallocate a sequence of blocks into a contiguous sequence of blocks.
139 *
140 * The vnode and an array of buffer pointers for a range of sequential
141 * logical blocks to be made contiguous is given. The allocator attempts
142 * to find a range of sequential blocks starting as close as possible to
143 * an fs_rotdelay offset from the end of the allocation for the logical
144 * block immediately preceding the current range. If successful, the
145 * physical block numbers in the buffer pointers and in the inode are
146 * changed to reflect the new allocation. If unsuccessful, the allocation
147 * is left unchanged. The success in doing the reallocation is returned.
148 * Note that the error return is not reflected back to the user. Rather
149 * the previous block allocation will be used.
150 */
151
152static SYSCTL_NODE(_vfs, OID_AUTO, ext2fs, CTLFLAG_RW, 0, "EXT2FS filesystem");
153
154static int doasyncfree = 0;
155SYSCTL_INT(_vfs_ext2fs, OID_AUTO, doasyncfree, CTLFLAG_RW, &doasyncfree, 0,
156 "Use asychronous writes to update block pointers when freeing blocks");
157
158static int doreallocblks = 0;
159SYSCTL_INT(_vfs_ext2fs, OID_AUTO, doreallocblks, CTLFLAG_RW, &doreallocblks, 0, "");
160
161int
162ext2_reallocblks(ap)
163 struct vop_reallocblks_args /* {
164 struct vnode *a_vp;
165 struct cluster_save *a_buflist;
166 } */ *ap;
167{
168 struct m_ext2fs *fs;
169 struct inode *ip;
170 struct vnode *vp;
171 struct buf *sbp, *ebp;
| 37 */
38
39#include <sys/param.h>
40#include <sys/systm.h>
41#include <sys/conf.h>
42#include <sys/vnode.h>
43#include <sys/stat.h>
44#include <sys/mount.h>
45#include <sys/sysctl.h>
46#include <sys/syslog.h>
47#include <sys/buf.h>
48
49#include <fs/ext2fs/inode.h>
50#include <fs/ext2fs/ext2_mount.h>
51#include <fs/ext2fs/ext2fs.h>
52#include <fs/ext2fs/fs.h>
53#include <fs/ext2fs/ext2_extern.h>
54
55static daddr_t ext2_alloccg(struct inode *, int, daddr_t, int);
56static daddr_t ext2_clusteralloc(struct inode *, int, daddr_t, int);
57static u_long ext2_dirpref(struct inode *);
58static void ext2_fserr(struct m_ext2fs *, uid_t, char *);
59static u_long ext2_hashalloc(struct inode *, int, long, int,
60 daddr_t (*)(struct inode *, int, daddr_t,
61 int));
62static daddr_t ext2_nodealloccg(struct inode *, int, daddr_t, int);
63static daddr_t ext2_mapsearch(struct m_ext2fs *, char *, daddr_t);
64
65/*
66 * Allocate a block in the file system.
67 *
68 * A preference may be optionally specified. If a preference is given
69 * the following hierarchy is used to allocate a block:
70 * 1) allocate the requested block.
71 * 2) allocate a rotationally optimal block in the same cylinder.
72 * 3) allocate a block in the same cylinder group.
73 * 4) quadradically rehash into other cylinder groups, until an
74 * available block is located.
75 * If no block preference is given the following hierarchy is used
76 * to allocate a block:
77 * 1) allocate a block in the cylinder group that contains the
78 * inode for the file.
79 * 2) quadradically rehash into other cylinder groups, until an
80 * available block is located.
81 */
82int
83ext2_alloc(ip, lbn, bpref, size, cred, bnp)
84 struct inode *ip;
85 int32_t lbn, bpref;
86 int size;
87 struct ucred *cred;
88 int32_t *bnp;
89{
90 struct m_ext2fs *fs;
91 struct ext2mount *ump;
92 int32_t bno;
93 int cg;
94 *bnp = 0;
95 fs = ip->i_e2fs;
96 ump = ip->i_ump;
97 mtx_assert(EXT2_MTX(ump), MA_OWNED);
98#ifdef DIAGNOSTIC
99 if ((u_int)size > fs->e2fs_bsize || blkoff(fs, size) != 0) {
100 vn_printf(ip->i_devvp, "bsize = %lu, size = %d, fs = %s\n",
101 (long unsigned int)fs->e2fs_bsize, size, fs->e2fs_fsmnt);
102 panic("ext2_alloc: bad size");
103 }
104 if (cred == NOCRED)
105 panic("ext2_alloc: missing credential");
106#endif /* DIAGNOSTIC */
107 if (size == fs->e2fs_bsize && fs->e2fs->e2fs_fbcount == 0)
108 goto nospace;
109 if (cred->cr_uid != 0 &&
110 fs->e2fs->e2fs_fbcount < fs->e2fs->e2fs_rbcount)
111 goto nospace;
112 if (bpref >= fs->e2fs->e2fs_bcount)
113 bpref = 0;
114 if (bpref == 0)
115 cg = ino_to_cg(fs, ip->i_number);
116 else
117 cg = dtog(fs, bpref);
118 bno = (daddr_t)ext2_hashalloc(ip, cg, bpref, fs->e2fs_bsize,
119 ext2_alloccg);
120 if (bno > 0) {
121 /* set next_alloc fields as done in block_getblk */
122 ip->i_next_alloc_block = lbn;
123 ip->i_next_alloc_goal = bno;
124
125 ip->i_blocks += btodb(fs->e2fs_bsize);
126 ip->i_flag |= IN_CHANGE | IN_UPDATE;
127 *bnp = bno;
128 return (0);
129 }
130nospace:
131 EXT2_UNLOCK(ump);
132 ext2_fserr(fs, cred->cr_uid, "file system full");
133 uprintf("\n%s: write failed, file system is full\n", fs->e2fs_fsmnt);
134 return (ENOSPC);
135}
136
137/*
138 * Reallocate a sequence of blocks into a contiguous sequence of blocks.
139 *
140 * The vnode and an array of buffer pointers for a range of sequential
141 * logical blocks to be made contiguous is given. The allocator attempts
142 * to find a range of sequential blocks starting as close as possible to
143 * an fs_rotdelay offset from the end of the allocation for the logical
144 * block immediately preceding the current range. If successful, the
145 * physical block numbers in the buffer pointers and in the inode are
146 * changed to reflect the new allocation. If unsuccessful, the allocation
147 * is left unchanged. The success in doing the reallocation is returned.
148 * Note that the error return is not reflected back to the user. Rather
149 * the previous block allocation will be used.
150 */
151
152static SYSCTL_NODE(_vfs, OID_AUTO, ext2fs, CTLFLAG_RW, 0, "EXT2FS filesystem");
153
154static int doasyncfree = 0;
155SYSCTL_INT(_vfs_ext2fs, OID_AUTO, doasyncfree, CTLFLAG_RW, &doasyncfree, 0,
156 "Use asychronous writes to update block pointers when freeing blocks");
157
158static int doreallocblks = 0;
159SYSCTL_INT(_vfs_ext2fs, OID_AUTO, doreallocblks, CTLFLAG_RW, &doreallocblks, 0, "");
160
161int
162ext2_reallocblks(ap)
163 struct vop_reallocblks_args /* {
164 struct vnode *a_vp;
165 struct cluster_save *a_buflist;
166 } */ *ap;
167{
168 struct m_ext2fs *fs;
169 struct inode *ip;
170 struct vnode *vp;
171 struct buf *sbp, *ebp;
|
173 struct ext2mount *ump;
174 struct cluster_save *buflist;
175 struct indir start_ap[NIADDR + 1], end_ap[NIADDR + 1], *idp;
176 int32_t start_lbn, end_lbn, soff, newblk, blkno;
177 int i, len, start_lvl, end_lvl, pref, ssize;
178
179 if (doreallocblks == 0)
180 return (ENOSPC);
181
182 vp = ap->a_vp;
183 ip = VTOI(vp);
184 fs = ip->i_e2fs;
185 ump = ip->i_ump;
186
187 if (fs->e2fs_contigsumsize <= 0)
188 return (ENOSPC);
189
190 buflist = ap->a_buflist;
191 len = buflist->bs_nchildren;
192 start_lbn = buflist->bs_children[0]->b_lblkno;
193 end_lbn = start_lbn + len - 1;
194#ifdef DIAGNOSTIC
195 for (i = 1; i < len; i++)
196 if (buflist->bs_children[i]->b_lblkno != start_lbn + i)
197 panic("ext2_reallocblks: non-cluster");
198#endif
199 /*
200 * If the cluster crosses the boundary for the first indirect
201 * block, leave space for the indirect block. Indirect blocks
202 * are initially laid out in a position after the last direct
203 * block. Block reallocation would usually destroy locality by
204 * moving the indirect block out of the way to make room for
205 * data blocks if we didn't compensate here. We should also do
206 * this for other indirect block boundaries, but it is only
207 * important for the first one.
208 */
209 if (start_lbn < NDADDR && end_lbn >= NDADDR)
210 return (ENOSPC);
211 /*
212 * If the latest allocation is in a new cylinder group, assume that
213 * the filesystem has decided to move and do not force it back to
214 * the previous cylinder group.
215 */
216 if (dtog(fs, dbtofsb(fs, buflist->bs_children[0]->b_blkno)) !=
217 dtog(fs, dbtofsb(fs, buflist->bs_children[len - 1]->b_blkno)))
218 return (ENOSPC);
219 if (ext2_getlbns(vp, start_lbn, start_ap, &start_lvl) ||
220 ext2_getlbns(vp, end_lbn, end_ap, &end_lvl))
221 return (ENOSPC);
222 /*
223 * Get the starting offset and block map for the first block.
224 */
225 if (start_lvl == 0) {
226 sbap = &ip->i_db[0];
227 soff = start_lbn;
228 } else {
229 idp = &start_ap[start_lvl - 1];
230 if (bread(vp, idp->in_lbn, (int)fs->e2fs_bsize, NOCRED, &sbp)) {
231 brelse(sbp);
232 return (ENOSPC);
233 }
234 sbap = (int32_t *)sbp->b_data;
235 soff = idp->in_off;
236 }
237 /*
238 * If the block range spans two block maps, get the second map.
239 */
240 if (end_lvl == 0 || (idp = &end_ap[end_lvl - 1])->in_off + 1 >= len) {
241 ssize = len;
242 } else {
243#ifdef DIAGNOSTIC
244 if (start_ap[start_lvl-1].in_lbn == idp->in_lbn)
245 panic("ext2_reallocblk: start == end");
246#endif
247 ssize = len - (idp->in_off + 1);
248 if (bread(vp, idp->in_lbn, (int)fs->e2fs_bsize, NOCRED, &ebp))
249 goto fail;
250 ebap = (int32_t *)ebp->b_data;
251 }
252 /*
253 * Find the preferred location for the cluster.
254 */
255 EXT2_LOCK(ump);
256 pref = ext2_blkpref(ip, start_lbn, soff, sbap, 0);
257 /*
258 * Search the block map looking for an allocation of the desired size.
259 */
260 if ((newblk = (int32_t)ext2_hashalloc(ip, dtog(fs, pref), pref,
261 len, ext2_clusteralloc)) == 0){
262 EXT2_UNLOCK(ump);
263 goto fail;
264 }
265 /*
266 * We have found a new contiguous block.
267 *
268 * First we have to replace the old block pointers with the new
269 * block pointers in the inode and indirect blocks associated
270 * with the file.
271 */
272#ifdef DEBUG
273 printf("realloc: ino %d, lbns %jd-%jd\n\told:", ip->i_number,
274 (intmax_t)start_lbn, (intmax_t)end_lbn);
275#endif /* DEBUG */
276 blkno = newblk;
277 for (bap = &sbap[soff], i = 0; i < len; i++, blkno += fs->e2fs_fpb) {
278 if (i == ssize) {
279 bap = ebap;
280 soff = -i;
281 }
282#ifdef DIAGNOSTIC
283 if (buflist->bs_children[i]->b_blkno != fsbtodb(fs, *bap))
284 panic("ext2_reallocblks: alloc mismatch");
285#endif
286#ifdef DEBUG
287 printf(" %d,", *bap);
288#endif /* DEBUG */
289 *bap++ = blkno;
290 }
291 /*
292 * Next we must write out the modified inode and indirect blocks.
293 * For strict correctness, the writes should be synchronous since
294 * the old block values may have been written to disk. In practise
295 * they are almost never written, but if we are concerned about
296 * strict correctness, the `doasyncfree' flag should be set to zero.
297 *
298 * The test on `doasyncfree' should be changed to test a flag
299 * that shows whether the associated buffers and inodes have
300 * been written. The flag should be set when the cluster is
301 * started and cleared whenever the buffer or inode is flushed.
302 * We can then check below to see if it is set, and do the
303 * synchronous write only when it has been cleared.
304 */
305 if (sbap != &ip->i_db[0]) {
306 if (doasyncfree)
307 bdwrite(sbp);
308 else
309 bwrite(sbp);
310 } else {
311 ip->i_flag |= IN_CHANGE | IN_UPDATE;
312 if (!doasyncfree)
313 ext2_update(vp, 1);
314 }
315 if (ssize < len) {
316 if (doasyncfree)
317 bdwrite(ebp);
318 else
319 bwrite(ebp);
320 }
321 /*
322 * Last, free the old blocks and assign the new blocks to the buffers.
323 */
324#ifdef DEBUG
325 printf("\n\tnew:");
326#endif /* DEBUG */
327 for (blkno = newblk, i = 0; i < len; i++, blkno += fs->e2fs_fpb) {
328 ext2_blkfree(ip, dbtofsb(fs, buflist->bs_children[i]->b_blkno),
329 fs->e2fs_bsize);
330 buflist->bs_children[i]->b_blkno = fsbtodb(fs, blkno);
331#ifdef DEBUG
332 printf(" %d,", blkno);
333#endif /* DEBUG */
334 }
335#ifdef DEBUG
336 printf("\n");
337#endif /* DEBUG */
338 return (0);
339
340fail:
341 if (ssize < len)
342 brelse(ebp);
343 if (sbap != &ip->i_db[0])
344 brelse(sbp);
345 return (ENOSPC);
346}
347
348/*
349 * Allocate an inode in the file system.
350 *
351 */
352int
353ext2_valloc(pvp, mode, cred, vpp)
354 struct vnode *pvp;
355 int mode;
356 struct ucred *cred;
357 struct vnode **vpp;
358{
359 struct timespec ts;
360 struct inode *pip;
361 struct m_ext2fs *fs;
362 struct inode *ip;
363 struct ext2mount *ump;
364 ino_t ino, ipref;
365 int i, error, cg;
366
367 *vpp = NULL;
368 pip = VTOI(pvp);
369 fs = pip->i_e2fs;
370 ump = pip->i_ump;
371
372 EXT2_LOCK(ump);
373 if (fs->e2fs->e2fs_ficount == 0)
374 goto noinodes;
375 /*
376 * If it is a directory then obtain a cylinder group based on
377 * ext2_dirpref else obtain it using ino_to_cg. The preferred inode is
378 * always the next inode.
379 */
380 if ((mode & IFMT) == IFDIR) {
381 cg = ext2_dirpref(pip);
382 if (fs->e2fs_contigdirs[cg] < 255)
383 fs->e2fs_contigdirs[cg]++;
384 } else {
385 cg = ino_to_cg(fs, pip->i_number);
386 if (fs->e2fs_contigdirs[cg] > 0)
387 fs->e2fs_contigdirs[cg]--;
388 }
389 ipref = cg * fs->e2fs->e2fs_ipg + 1;
390 ino = (ino_t)ext2_hashalloc(pip, cg, (long)ipref, mode, ext2_nodealloccg);
391
392 if (ino == 0)
393 goto noinodes;
394 error = VFS_VGET(pvp->v_mount, ino, LK_EXCLUSIVE, vpp);
395 if (error) {
396 ext2_vfree(pvp, ino, mode);
397 return (error);
398 }
399 ip = VTOI(*vpp);
400
401 /*
402 * The question is whether using VGET was such good idea at all:
403 * Linux doesn't read the old inode in when it is allocating a
404 * new one. I will set at least i_size and i_blocks to zero.
405 */
406 ip->i_size = 0;
407 ip->i_blocks = 0;
408 ip->i_mode = 0;
409 ip->i_flags = 0;
410 /* now we want to make sure that the block pointers are zeroed out */
411 for (i = 0; i < NDADDR; i++)
412 ip->i_db[i] = 0;
413 for (i = 0; i < NIADDR; i++)
414 ip->i_ib[i] = 0;
415
416 /*
417 * Set up a new generation number for this inode.
418 * XXX check if this makes sense in ext2
419 */
420 if (ip->i_gen == 0 || ++ip->i_gen == 0)
421 ip->i_gen = random() / 2 + 1;
422
423 vfs_timestamp(&ts);
424 ip->i_birthtime = ts.tv_sec;
425 ip->i_birthnsec = ts.tv_nsec;
426
427/*
428printf("ext2_valloc: allocated inode %d\n", ino);
429*/
430 return (0);
431noinodes:
432 EXT2_UNLOCK(ump);
433 ext2_fserr(fs, cred->cr_uid, "out of inodes");
434 uprintf("\n%s: create/symlink failed, no inodes free\n", fs->e2fs_fsmnt);
435 return (ENOSPC);
436}
437
438/*
439 * Find a cylinder to place a directory.
440 *
441 * The policy implemented by this algorithm is to allocate a
442 * directory inode in the same cylinder group as its parent
443 * directory, but also to reserve space for its files inodes
444 * and data. Restrict the number of directories which may be
445 * allocated one after another in the same cylinder group
446 * without intervening allocation of files.
447 *
448 * If we allocate a first level directory then force allocation
449 * in another cylinder group.
450 *
451 */
452static u_long
453ext2_dirpref(struct inode *pip)
454{
455 struct m_ext2fs *fs;
456 int cg, prefcg, dirsize, cgsize;
457 int avgifree, avgbfree, avgndir, curdirsize;
458 int minifree, minbfree, maxndir;
459 int mincg, minndir;
460 int maxcontigdirs;
461
462 mtx_assert(EXT2_MTX(pip->i_ump), MA_OWNED);
463 fs = pip->i_e2fs;
464
465 avgifree = fs->e2fs->e2fs_ficount / fs->e2fs_gcount;
466 avgbfree = fs->e2fs->e2fs_fbcount / fs->e2fs_gcount;
467 avgndir = fs->e2fs_total_dir / fs->e2fs_gcount;
468
469 /*
470 * Force allocation in another cg if creating a first level dir.
471 */
472 ASSERT_VOP_LOCKED(ITOV(pip), "ext2fs_dirpref");
473 if (ITOV(pip)->v_vflag & VV_ROOT) {
474 prefcg = arc4random() % fs->e2fs_gcount;
475 mincg = prefcg;
476 minndir = fs->e2fs_ipg;
477 for (cg = prefcg; cg < fs->e2fs_gcount; cg++)
478 if (fs->e2fs_gd[cg].ext2bgd_ndirs < minndir &&
479 fs->e2fs_gd[cg].ext2bgd_nifree >= avgifree &&
480 fs->e2fs_gd[cg].ext2bgd_nbfree >= avgbfree) {
481 mincg = cg;
482 minndir = fs->e2fs_gd[cg].ext2bgd_ndirs;
483 }
484 for (cg = 0; cg < prefcg; cg++)
485 if (fs->e2fs_gd[cg].ext2bgd_ndirs < minndir &&
486 fs->e2fs_gd[cg].ext2bgd_nifree >= avgifree &&
487 fs->e2fs_gd[cg].ext2bgd_nbfree >= avgbfree) {
488 mincg = cg;
489 minndir = fs->e2fs_gd[cg].ext2bgd_ndirs;
490 }
491
492 return (mincg);
493 }
494
495 /*
496 * Count various limits which used for
497 * optimal allocation of a directory inode.
498 */
499 maxndir = min(avgndir + fs->e2fs_ipg / 16, fs->e2fs_ipg);
500 minifree = avgifree - avgifree / 4;
501 if (minifree < 1)
502 minifree = 1;
503 minbfree = avgbfree - avgbfree / 4;
504 if (minbfree < 1)
505 minbfree = 1;
506 cgsize = fs->e2fs_fsize * fs->e2fs_fpg;
507 dirsize = AVGDIRSIZE;
508 curdirsize = avgndir ? (cgsize - avgbfree * fs->e2fs_bsize) / avgndir : 0;
509 if (dirsize < curdirsize)
510 dirsize = curdirsize;
511 if (dirsize <= 0)
512 maxcontigdirs = 0; /* dirsize overflowed */
513 else
514 maxcontigdirs = min((avgbfree * fs->e2fs_bsize) / dirsize, 255);
515 maxcontigdirs = min(maxcontigdirs, fs->e2fs_ipg / AFPDIR);
516 if (maxcontigdirs == 0)
517 maxcontigdirs = 1;
518
519 /*
520 * Limit number of dirs in one cg and reserve space for
521 * regular files, but only if we have no deficit in
522 * inodes or space.
523 */
524 prefcg = ino_to_cg(fs, pip->i_number);
525 for (cg = prefcg; cg < fs->e2fs_gcount; cg++)
526 if (fs->e2fs_gd[cg].ext2bgd_ndirs < maxndir &&
527 fs->e2fs_gd[cg].ext2bgd_nifree >= minifree &&
528 fs->e2fs_gd[cg].ext2bgd_nbfree >= minbfree) {
529 if (fs->e2fs_contigdirs[cg] < maxcontigdirs)
530 return (cg);
531 }
532 for (cg = 0; cg < prefcg; cg++)
533 if (fs->e2fs_gd[cg].ext2bgd_ndirs < maxndir &&
534 fs->e2fs_gd[cg].ext2bgd_nifree >= minifree &&
535 fs->e2fs_gd[cg].ext2bgd_nbfree >= minbfree) {
536 if (fs->e2fs_contigdirs[cg] < maxcontigdirs)
537 return (cg);
538 }
539 /*
540 * This is a backstop when we have deficit in space.
541 */
542 for (cg = prefcg; cg < fs->e2fs_gcount; cg++)
543 if (fs->e2fs_gd[cg].ext2bgd_nifree >= avgifree)
544 return (cg);
545 for (cg = 0; cg < prefcg; cg++)
546 if (fs->e2fs_gd[cg].ext2bgd_nifree >= avgifree)
547 break;
548 return (cg);
549}
550
551/*
552 * Select the desired position for the next block in a file.
553 *
554 * we try to mimic what Remy does in inode_getblk/block_getblk
555 *
556 * we note: blocknr == 0 means that we're about to allocate either
557 * a direct block or a pointer block at the first level of indirection
558 * (In other words, stuff that will go in i_db[] or i_ib[])
559 *
560 * blocknr != 0 means that we're allocating a block that is none
561 * of the above. Then, blocknr tells us the number of the block
562 * that will hold the pointer
563 */
564int32_t
565ext2_blkpref(ip, lbn, indx, bap, blocknr)
566 struct inode *ip;
567 int32_t lbn;
568 int indx;
569 int32_t *bap;
570 int32_t blocknr;
571{
572 int tmp;
573 mtx_assert(EXT2_MTX(ip->i_ump), MA_OWNED);
574
575 /* if the next block is actually what we thought it is,
576 then set the goal to what we thought it should be
577 */
578 if (ip->i_next_alloc_block == lbn && ip->i_next_alloc_goal != 0)
579 return ip->i_next_alloc_goal;
580
581 /* now check whether we were provided with an array that basically
582 tells us previous blocks to which we want to stay closeby
583 */
584 if (bap)
585 for (tmp = indx - 1; tmp >= 0; tmp--)
586 if (bap[tmp])
587 return bap[tmp];
588
589 /* else let's fall back to the blocknr, or, if there is none,
590 follow the rule that a block should be allocated near its inode
591 */
592 return blocknr ? blocknr :
593 (int32_t)(ip->i_block_group *
594 EXT2_BLOCKS_PER_GROUP(ip->i_e2fs)) +
595 ip->i_e2fs->e2fs->e2fs_first_dblock;
596}
597
598/*
599 * Implement the cylinder overflow algorithm.
600 *
601 * The policy implemented by this algorithm is:
602 * 1) allocate the block in its requested cylinder group.
603 * 2) quadradically rehash on the cylinder group number.
604 * 3) brute force search for a free block.
605 */
606static u_long
607ext2_hashalloc(struct inode *ip, int cg, long pref, int size,
608 daddr_t (*allocator)(struct inode *, int, daddr_t, int))
609{
610 struct m_ext2fs *fs;
611 ino_t result;
612 int i, icg = cg;
613
614 mtx_assert(EXT2_MTX(ip->i_ump), MA_OWNED);
615 fs = ip->i_e2fs;
616 /*
617 * 1: preferred cylinder group
618 */
619 result = (*allocator)(ip, cg, pref, size);
620 if (result)
621 return (result);
622 /*
623 * 2: quadratic rehash
624 */
625 for (i = 1; i < fs->e2fs_gcount; i *= 2) {
626 cg += i;
627 if (cg >= fs->e2fs_gcount)
628 cg -= fs->e2fs_gcount;
629 result = (*allocator)(ip, cg, 0, size);
630 if (result)
631 return (result);
632 }
633 /*
634 * 3: brute force search
635 * Note that we start at i == 2, since 0 was checked initially,
636 * and 1 is always checked in the quadratic rehash.
637 */
638 cg = (icg + 2) % fs->e2fs_gcount;
639 for (i = 2; i < fs->e2fs_gcount; i++) {
640 result = (*allocator)(ip, cg, 0, size);
641 if (result)
642 return (result);
643 cg++;
644 if (cg == fs->e2fs_gcount)
645 cg = 0;
646 }
647 return (0);
648}
649
650/*
651 * Determine whether a block can be allocated.
652 *
653 * Check to see if a block of the appropriate size is available,
654 * and if it is, allocate it.
655 */
656static daddr_t
657ext2_alloccg(struct inode *ip, int cg, daddr_t bpref, int size)
658{
659 struct m_ext2fs *fs;
660 struct buf *bp;
661 struct ext2mount *ump;
662 daddr_t bno, runstart, runlen;
663 int bit, loc, end, error, start;
664 char *bbp;
665 /* XXX ondisk32 */
666 fs = ip->i_e2fs;
667 ump = ip->i_ump;
668 if (fs->e2fs_gd[cg].ext2bgd_nbfree == 0)
669 return (0);
670 EXT2_UNLOCK(ump);
671 error = bread(ip->i_devvp, fsbtodb(fs,
672 fs->e2fs_gd[cg].ext2bgd_b_bitmap),
673 (int)fs->e2fs_bsize, NOCRED, &bp);
674 if (error) {
675 brelse(bp);
676 EXT2_LOCK(ump);
677 return (0);
678 }
679 if (fs->e2fs_gd[cg].ext2bgd_nbfree == 0) {
680 /*
681 * Another thread allocated the last block in this
682 * group while we were waiting for the buffer.
683 */
684 brelse(bp);
685 EXT2_LOCK(ump);
686 return (0);
687 }
688 bbp = (char *)bp->b_data;
689
690 if (dtog(fs, bpref) != cg)
691 bpref = 0;
692 if (bpref != 0) {
693 bpref = dtogd(fs, bpref);
694 /*
695 * if the requested block is available, use it
696 */
697 if (isclr(bbp, bpref)) {
698 bno = bpref;
699 goto gotit;
700 }
701 }
702 /*
703 * no blocks in the requested cylinder, so take next
704 * available one in this cylinder group.
705 * first try to get 8 contigous blocks, then fall back to a single
706 * block.
707 */
708 if (bpref)
709 start = dtogd(fs, bpref) / NBBY;
710 else
711 start = 0;
712 end = howmany(fs->e2fs->e2fs_fpg, NBBY) - start;
713retry:
714 runlen = 0;
715 runstart = 0;
716 for (loc = start; loc < end; loc++) {
717 if (bbp[loc] == (char)0xff) {
718 runlen = 0;
719 continue;
720 }
721
722 /* Start of a run, find the number of high clear bits. */
723 if (runlen == 0) {
724 bit = fls(bbp[loc]);
725 runlen = NBBY - bit;
726 runstart = loc * NBBY + bit;
727 } else if (bbp[loc] == 0) {
728 /* Continue a run. */
729 runlen += NBBY;
730 } else {
731 /*
732 * Finish the current run. If it isn't long
733 * enough, start a new one.
734 */
735 bit = ffs(bbp[loc]) - 1;
736 runlen += bit;
737 if (runlen >= 8) {
738 bno = runstart;
739 goto gotit;
740 }
741
742 /* Run was too short, start a new one. */
743 bit = fls(bbp[loc]);
744 runlen = NBBY - bit;
745 runstart = loc * NBBY + bit;
746 }
747
748 /* If the current run is long enough, use it. */
749 if (runlen >= 8) {
750 bno = runstart;
751 goto gotit;
752 }
753 }
754 if (start != 0) {
755 end = start;
756 start = 0;
757 goto retry;
758 }
759
760 bno = ext2_mapsearch(fs, bbp, bpref);
761 if (bno < 0){
762 brelse(bp);
763 EXT2_LOCK(ump);
764 return (0);
765 }
766gotit:
767#ifdef DIAGNOSTIC
768 if (isset(bbp, bno)) {
769 printf("ext2fs_alloccgblk: cg=%d bno=%jd fs=%s\n",
770 cg, (intmax_t)bno, fs->e2fs_fsmnt);
771 panic("ext2fs_alloccg: dup alloc");
772 }
773#endif
774 setbit(bbp, bno);
775 EXT2_LOCK(ump);
776 ext2_clusteracct(fs, bbp, cg, bno, -1);
777 fs->e2fs->e2fs_fbcount--;
778 fs->e2fs_gd[cg].ext2bgd_nbfree--;
779 fs->e2fs_fmod = 1;
780 EXT2_UNLOCK(ump);
781 bdwrite(bp);
782 return (cg * fs->e2fs->e2fs_fpg + fs->e2fs->e2fs_first_dblock + bno);
783}
784
785/*
786 * Determine whether a cluster can be allocated.
787 */
788static daddr_t
789ext2_clusteralloc(struct inode *ip, int cg, daddr_t bpref, int len)
790{
791 struct m_ext2fs *fs;
792 struct ext2mount *ump;
793 struct buf *bp;
794 char *bbp;
795 int bit, error, got, i, loc, run;
796 int32_t *lp;
797 daddr_t bno;
798
799 fs = ip->i_e2fs;
800 ump = ip->i_ump;
801
802 if (fs->e2fs_maxcluster[cg] < len)
803 return (0);
804
805 EXT2_UNLOCK(ump);
806 error = bread(ip->i_devvp,
807 fsbtodb(fs, fs->e2fs_gd[cg].ext2bgd_b_bitmap),
808 (int)fs->e2fs_bsize, NOCRED, &bp);
809 if (error)
810 goto fail_lock;
811
812 bbp = (char *)bp->b_data;
813 bp->b_xflags |= BX_BKGRDWRITE;
814
815 EXT2_LOCK(ump);
816 /*
817 * Check to see if a cluster of the needed size (or bigger) is
818 * available in this cylinder group.
819 */
820 lp = &fs->e2fs_clustersum[cg].cs_sum[len];
821 for (i = len; i <= fs->e2fs_contigsumsize; i++)
822 if (*lp++ > 0)
823 break;
824 if (i > fs->e2fs_contigsumsize) {
825 /*
826 * Update the cluster summary information to reflect
827 * the true maximum-sized cluster so that future cluster
828 * allocation requests can avoid reading the bitmap only
829 * to find no cluster.
830 */
831 lp = &fs->e2fs_clustersum[cg].cs_sum[len - 1];
832 for (i = len - 1; i > 0; i--)
833 if (*lp-- > 0)
834 break;
835 fs->e2fs_maxcluster[cg] = i;
836 goto fail;
837 }
838 EXT2_UNLOCK(ump);
839
840 /* Search the bitmap to find a big enough cluster like in FFS. */
841 if (dtog(fs, bpref) != cg)
842 bpref = 0;
843 if (bpref != 0)
844 bpref = dtogd(fs, bpref);
845 loc = bpref / NBBY;
846 bit = 1 << (bpref % NBBY);
847 for (run = 0, got = bpref; got < fs->e2fs->e2fs_fpg; got++) {
848 if ((bbp[loc] & bit) != 0)
849 run = 0;
850 else {
851 run++;
852 if (run == len)
853 break;
854 }
855 if ((got & (NBBY - 1)) != (NBBY - 1))
856 bit <<= 1;
857 else {
858 loc++;
859 bit = 1;
860 }
861 }
862
863 if (got >= fs->e2fs->e2fs_fpg)
864 goto fail_lock;
865
866 /* Allocate the cluster that we found. */
867 for (i = 1; i < len; i++)
868 if (!isclr(bbp, got - run + i))
869 panic("ext2_clusteralloc: map mismatch");
870
871 bno = got - run + 1;
872 if (bno >= fs->e2fs->e2fs_fpg)
873 panic("ext2_clusteralloc: allocated out of group");
874
875 EXT2_LOCK(ump);
876 for (i = 0; i < len; i += fs->e2fs_fpb) {
877 setbit(bbp, bno + i);
878 ext2_clusteracct(fs, bbp, cg, bno + i, -1);
879 fs->e2fs->e2fs_fbcount--;
880 fs->e2fs_gd[cg].ext2bgd_nbfree--;
881 }
882 fs->e2fs_fmod = 1;
883 EXT2_UNLOCK(ump);
884
885 bdwrite(bp);
886 return (cg * fs->e2fs->e2fs_fpg + fs->e2fs->e2fs_first_dblock + bno);
887
888fail_lock:
889 EXT2_LOCK(ump);
890fail:
891 brelse(bp);
892 return (0);
893}
894
895/*
896 * Determine whether an inode can be allocated.
897 *
898 * Check to see if an inode is available, and if it is,
899 * allocate it using tode in the specified cylinder group.
900 */
901static daddr_t
902ext2_nodealloccg(struct inode *ip, int cg, daddr_t ipref, int mode)
903{
904 struct m_ext2fs *fs;
905 struct buf *bp;
906 struct ext2mount *ump;
907 int error, start, len;
908 char *ibp, *loc;
909 ipref--; /* to avoid a lot of (ipref -1) */
910 if (ipref == -1)
911 ipref = 0;
912 fs = ip->i_e2fs;
913 ump = ip->i_ump;
914 if (fs->e2fs_gd[cg].ext2bgd_nifree == 0)
915 return (0);
916 EXT2_UNLOCK(ump);
917 error = bread(ip->i_devvp, fsbtodb(fs,
918 fs->e2fs_gd[cg].ext2bgd_i_bitmap),
919 (int)fs->e2fs_bsize, NOCRED, &bp);
920 if (error) {
921 brelse(bp);
922 EXT2_LOCK(ump);
923 return (0);
924 }
925 if (fs->e2fs_gd[cg].ext2bgd_nifree == 0) {
926 /*
927 * Another thread allocated the last i-node in this
928 * group while we were waiting for the buffer.
929 */
930 brelse(bp);
931 EXT2_LOCK(ump);
932 return (0);
933 }
934 ibp = (char *)bp->b_data;
935 if (ipref) {
936 ipref %= fs->e2fs->e2fs_ipg;
937 if (isclr(ibp, ipref))
938 goto gotit;
939 }
940 start = ipref / NBBY;
941 len = howmany(fs->e2fs->e2fs_ipg - ipref, NBBY);
942 loc = memcchr(&ibp[start], 0xff, len);
943 if (loc == NULL) {
944 len = start + 1;
945 start = 0;
946 loc = memcchr(&ibp[start], 0xff, len);
947 if (loc == NULL) {
948 printf("cg = %d, ipref = %lld, fs = %s\n",
949 cg, (long long)ipref, fs->e2fs_fsmnt);
950 panic("ext2fs_nodealloccg: map corrupted");
951 /* NOTREACHED */
952 }
953 }
954 ipref = (loc - ibp) * NBBY + ffs(~*loc) - 1;
955gotit:
956 setbit(ibp, ipref);
957 EXT2_LOCK(ump);
958 fs->e2fs_gd[cg].ext2bgd_nifree--;
959 fs->e2fs->e2fs_ficount--;
960 fs->e2fs_fmod = 1;
961 if ((mode & IFMT) == IFDIR) {
962 fs->e2fs_gd[cg].ext2bgd_ndirs++;
963 fs->e2fs_total_dir++;
964 }
965 EXT2_UNLOCK(ump);
966 bdwrite(bp);
967 return (cg * fs->e2fs->e2fs_ipg + ipref +1);
968}
969
970/*
971 * Free a block or fragment.
972 *
973 */
974void
975ext2_blkfree(ip, bno, size)
976 struct inode *ip;
977 int32_t bno;
978 long size;
979{
980 struct m_ext2fs *fs;
981 struct buf *bp;
982 struct ext2mount *ump;
983 int cg, error;
984 char *bbp;
985
986 fs = ip->i_e2fs;
987 ump = ip->i_ump;
988 cg = dtog(fs, bno);
989 if ((u_int)bno >= fs->e2fs->e2fs_bcount) {
990 printf("bad block %lld, ino %llu\n", (long long)bno,
991 (unsigned long long)ip->i_number);
992 ext2_fserr(fs, ip->i_uid, "bad block");
993 return;
994 }
995 error = bread(ip->i_devvp,
996 fsbtodb(fs, fs->e2fs_gd[cg].ext2bgd_b_bitmap),
997 (int)fs->e2fs_bsize, NOCRED, &bp);
998 if (error) {
999 brelse(bp);
1000 return;
1001 }
1002 bbp = (char *)bp->b_data;
1003 bno = dtogd(fs, bno);
1004 if (isclr(bbp, bno)) {
1005 printf("block = %lld, fs = %s\n",
1006 (long long)bno, fs->e2fs_fsmnt);
1007 panic("blkfree: freeing free block");
1008 }
1009 clrbit(bbp, bno);
1010 EXT2_LOCK(ump);
1011 ext2_clusteracct(fs, bbp, cg, bno, 1);
1012 fs->e2fs->e2fs_fbcount++;
1013 fs->e2fs_gd[cg].ext2bgd_nbfree++;
1014 fs->e2fs_fmod = 1;
1015 EXT2_UNLOCK(ump);
1016 bdwrite(bp);
1017}
1018
1019/*
1020 * Free an inode.
1021 *
1022 */
1023int
1024ext2_vfree(pvp, ino, mode)
1025 struct vnode *pvp;
1026 ino_t ino;
1027 int mode;
1028{
1029 struct m_ext2fs *fs;
1030 struct inode *pip;
1031 struct buf *bp;
1032 struct ext2mount *ump;
1033 int error, cg;
1034 char * ibp;
1035/* mode_t save_i_mode; */
1036
1037 pip = VTOI(pvp);
1038 fs = pip->i_e2fs;
1039 ump = pip->i_ump;
1040 if ((u_int)ino > fs->e2fs_ipg * fs->e2fs_gcount)
1041 panic("ext2_vfree: range: devvp = %p, ino = %ju, fs = %s",
1042 pip->i_devvp, (uintmax_t)ino, fs->e2fs_fsmnt);
1043
1044 cg = ino_to_cg(fs, ino);
1045 error = bread(pip->i_devvp,
1046 fsbtodb(fs, fs->e2fs_gd[cg].ext2bgd_i_bitmap),
1047 (int)fs->e2fs_bsize, NOCRED, &bp);
1048 if (error) {
1049 brelse(bp);
1050 return (0);
1051 }
1052 ibp = (char *)bp->b_data;
1053 ino = (ino - 1) % fs->e2fs->e2fs_ipg;
1054 if (isclr(ibp, ino)) {
1055 printf("ino = %llu, fs = %s\n",
1056 (unsigned long long)ino, fs->e2fs_fsmnt);
1057 if (fs->e2fs_ronly == 0)
1058 panic("ifree: freeing free inode");
1059 }
1060 clrbit(ibp, ino);
1061 EXT2_LOCK(ump);
1062 fs->e2fs->e2fs_ficount++;
1063 fs->e2fs_gd[cg].ext2bgd_nifree++;
1064 if ((mode & IFMT) == IFDIR) {
1065 fs->e2fs_gd[cg].ext2bgd_ndirs--;
1066 fs->e2fs_total_dir--;
1067 }
1068 fs->e2fs_fmod = 1;
1069 EXT2_UNLOCK(ump);
1070 bdwrite(bp);
1071 return (0);
1072}
1073
1074/*
1075 * Find a block in the specified cylinder group.
1076 *
1077 * It is a panic if a request is made to find a block if none are
1078 * available.
1079 */
1080static daddr_t
1081ext2_mapsearch(struct m_ext2fs *fs, char *bbp, daddr_t bpref)
1082{
1083 char *loc;
1084 int start, len;
1085
1086 /*
1087 * find the fragment by searching through the free block
1088 * map for an appropriate bit pattern
1089 */
1090 if (bpref)
1091 start = dtogd(fs, bpref) / NBBY;
1092 else
1093 start = 0;
1094 len = howmany(fs->e2fs->e2fs_fpg, NBBY) - start;
1095 loc = memcchr(&bbp[start], 0xff, len);
1096 if (loc == NULL) {
1097 len = start + 1;
1098 start = 0;
1099 loc = memcchr(&bbp[start], 0xff, len);
1100 if (loc == NULL) {
1101 printf("start = %d, len = %d, fs = %s\n",
1102 start, len, fs->e2fs_fsmnt);
1103 panic("ext2fs_alloccg: map corrupted");
1104 /* NOTREACHED */
1105 }
1106 }
1107 return ((loc - bbp) * NBBY + ffs(~*loc) - 1);
1108}
1109
1110/*
1111 * Fserr prints the name of a file system with an error diagnostic.
1112 *
1113 * The form of the error message is:
1114 * fs: error message
1115 */
1116static void
1117ext2_fserr(fs, uid, cp)
1118 struct m_ext2fs *fs;
1119 uid_t uid;
1120 char *cp;
1121{
1122
1123 log(LOG_ERR, "uid %u on %s: %s\n", uid, fs->e2fs_fsmnt, cp);
1124}
1125
1126int
1127cg_has_sb(int i)
1128{
1129 int a3, a5, a7;
1130
1131 if (i == 0 || i == 1)
1132 return 1;
1133 for (a3 = 3, a5 = 5, a7 = 7;
1134 a3 <= i || a5 <= i || a7 <= i;
1135 a3 *= 3, a5 *= 5, a7 *= 7)
1136 if (i == a3 || i == a5 || i == a7)
1137 return 1;
1138 return 0;
1139}
| 173 struct ext2mount *ump;
174 struct cluster_save *buflist;
175 struct indir start_ap[NIADDR + 1], end_ap[NIADDR + 1], *idp;
176 int32_t start_lbn, end_lbn, soff, newblk, blkno;
177 int i, len, start_lvl, end_lvl, pref, ssize;
178
179 if (doreallocblks == 0)
180 return (ENOSPC);
181
182 vp = ap->a_vp;
183 ip = VTOI(vp);
184 fs = ip->i_e2fs;
185 ump = ip->i_ump;
186
187 if (fs->e2fs_contigsumsize <= 0)
188 return (ENOSPC);
189
190 buflist = ap->a_buflist;
191 len = buflist->bs_nchildren;
192 start_lbn = buflist->bs_children[0]->b_lblkno;
193 end_lbn = start_lbn + len - 1;
194#ifdef DIAGNOSTIC
195 for (i = 1; i < len; i++)
196 if (buflist->bs_children[i]->b_lblkno != start_lbn + i)
197 panic("ext2_reallocblks: non-cluster");
198#endif
199 /*
200 * If the cluster crosses the boundary for the first indirect
201 * block, leave space for the indirect block. Indirect blocks
202 * are initially laid out in a position after the last direct
203 * block. Block reallocation would usually destroy locality by
204 * moving the indirect block out of the way to make room for
205 * data blocks if we didn't compensate here. We should also do
206 * this for other indirect block boundaries, but it is only
207 * important for the first one.
208 */
209 if (start_lbn < NDADDR && end_lbn >= NDADDR)
210 return (ENOSPC);
211 /*
212 * If the latest allocation is in a new cylinder group, assume that
213 * the filesystem has decided to move and do not force it back to
214 * the previous cylinder group.
215 */
216 if (dtog(fs, dbtofsb(fs, buflist->bs_children[0]->b_blkno)) !=
217 dtog(fs, dbtofsb(fs, buflist->bs_children[len - 1]->b_blkno)))
218 return (ENOSPC);
219 if (ext2_getlbns(vp, start_lbn, start_ap, &start_lvl) ||
220 ext2_getlbns(vp, end_lbn, end_ap, &end_lvl))
221 return (ENOSPC);
222 /*
223 * Get the starting offset and block map for the first block.
224 */
225 if (start_lvl == 0) {
226 sbap = &ip->i_db[0];
227 soff = start_lbn;
228 } else {
229 idp = &start_ap[start_lvl - 1];
230 if (bread(vp, idp->in_lbn, (int)fs->e2fs_bsize, NOCRED, &sbp)) {
231 brelse(sbp);
232 return (ENOSPC);
233 }
234 sbap = (int32_t *)sbp->b_data;
235 soff = idp->in_off;
236 }
237 /*
238 * If the block range spans two block maps, get the second map.
239 */
240 if (end_lvl == 0 || (idp = &end_ap[end_lvl - 1])->in_off + 1 >= len) {
241 ssize = len;
242 } else {
243#ifdef DIAGNOSTIC
244 if (start_ap[start_lvl-1].in_lbn == idp->in_lbn)
245 panic("ext2_reallocblk: start == end");
246#endif
247 ssize = len - (idp->in_off + 1);
248 if (bread(vp, idp->in_lbn, (int)fs->e2fs_bsize, NOCRED, &ebp))
249 goto fail;
250 ebap = (int32_t *)ebp->b_data;
251 }
252 /*
253 * Find the preferred location for the cluster.
254 */
255 EXT2_LOCK(ump);
256 pref = ext2_blkpref(ip, start_lbn, soff, sbap, 0);
257 /*
258 * Search the block map looking for an allocation of the desired size.
259 */
260 if ((newblk = (int32_t)ext2_hashalloc(ip, dtog(fs, pref), pref,
261 len, ext2_clusteralloc)) == 0){
262 EXT2_UNLOCK(ump);
263 goto fail;
264 }
265 /*
266 * We have found a new contiguous block.
267 *
268 * First we have to replace the old block pointers with the new
269 * block pointers in the inode and indirect blocks associated
270 * with the file.
271 */
272#ifdef DEBUG
273 printf("realloc: ino %d, lbns %jd-%jd\n\told:", ip->i_number,
274 (intmax_t)start_lbn, (intmax_t)end_lbn);
275#endif /* DEBUG */
276 blkno = newblk;
277 for (bap = &sbap[soff], i = 0; i < len; i++, blkno += fs->e2fs_fpb) {
278 if (i == ssize) {
279 bap = ebap;
280 soff = -i;
281 }
282#ifdef DIAGNOSTIC
283 if (buflist->bs_children[i]->b_blkno != fsbtodb(fs, *bap))
284 panic("ext2_reallocblks: alloc mismatch");
285#endif
286#ifdef DEBUG
287 printf(" %d,", *bap);
288#endif /* DEBUG */
289 *bap++ = blkno;
290 }
291 /*
292 * Next we must write out the modified inode and indirect blocks.
293 * For strict correctness, the writes should be synchronous since
294 * the old block values may have been written to disk. In practise
295 * they are almost never written, but if we are concerned about
296 * strict correctness, the `doasyncfree' flag should be set to zero.
297 *
298 * The test on `doasyncfree' should be changed to test a flag
299 * that shows whether the associated buffers and inodes have
300 * been written. The flag should be set when the cluster is
301 * started and cleared whenever the buffer or inode is flushed.
302 * We can then check below to see if it is set, and do the
303 * synchronous write only when it has been cleared.
304 */
305 if (sbap != &ip->i_db[0]) {
306 if (doasyncfree)
307 bdwrite(sbp);
308 else
309 bwrite(sbp);
310 } else {
311 ip->i_flag |= IN_CHANGE | IN_UPDATE;
312 if (!doasyncfree)
313 ext2_update(vp, 1);
314 }
315 if (ssize < len) {
316 if (doasyncfree)
317 bdwrite(ebp);
318 else
319 bwrite(ebp);
320 }
321 /*
322 * Last, free the old blocks and assign the new blocks to the buffers.
323 */
324#ifdef DEBUG
325 printf("\n\tnew:");
326#endif /* DEBUG */
327 for (blkno = newblk, i = 0; i < len; i++, blkno += fs->e2fs_fpb) {
328 ext2_blkfree(ip, dbtofsb(fs, buflist->bs_children[i]->b_blkno),
329 fs->e2fs_bsize);
330 buflist->bs_children[i]->b_blkno = fsbtodb(fs, blkno);
331#ifdef DEBUG
332 printf(" %d,", blkno);
333#endif /* DEBUG */
334 }
335#ifdef DEBUG
336 printf("\n");
337#endif /* DEBUG */
338 return (0);
339
340fail:
341 if (ssize < len)
342 brelse(ebp);
343 if (sbap != &ip->i_db[0])
344 brelse(sbp);
345 return (ENOSPC);
346}
347
348/*
349 * Allocate an inode in the file system.
350 *
351 */
352int
353ext2_valloc(pvp, mode, cred, vpp)
354 struct vnode *pvp;
355 int mode;
356 struct ucred *cred;
357 struct vnode **vpp;
358{
359 struct timespec ts;
360 struct inode *pip;
361 struct m_ext2fs *fs;
362 struct inode *ip;
363 struct ext2mount *ump;
364 ino_t ino, ipref;
365 int i, error, cg;
366
367 *vpp = NULL;
368 pip = VTOI(pvp);
369 fs = pip->i_e2fs;
370 ump = pip->i_ump;
371
372 EXT2_LOCK(ump);
373 if (fs->e2fs->e2fs_ficount == 0)
374 goto noinodes;
375 /*
376 * If it is a directory then obtain a cylinder group based on
377 * ext2_dirpref else obtain it using ino_to_cg. The preferred inode is
378 * always the next inode.
379 */
380 if ((mode & IFMT) == IFDIR) {
381 cg = ext2_dirpref(pip);
382 if (fs->e2fs_contigdirs[cg] < 255)
383 fs->e2fs_contigdirs[cg]++;
384 } else {
385 cg = ino_to_cg(fs, pip->i_number);
386 if (fs->e2fs_contigdirs[cg] > 0)
387 fs->e2fs_contigdirs[cg]--;
388 }
389 ipref = cg * fs->e2fs->e2fs_ipg + 1;
390 ino = (ino_t)ext2_hashalloc(pip, cg, (long)ipref, mode, ext2_nodealloccg);
391
392 if (ino == 0)
393 goto noinodes;
394 error = VFS_VGET(pvp->v_mount, ino, LK_EXCLUSIVE, vpp);
395 if (error) {
396 ext2_vfree(pvp, ino, mode);
397 return (error);
398 }
399 ip = VTOI(*vpp);
400
401 /*
402 * The question is whether using VGET was such good idea at all:
403 * Linux doesn't read the old inode in when it is allocating a
404 * new one. I will set at least i_size and i_blocks to zero.
405 */
406 ip->i_size = 0;
407 ip->i_blocks = 0;
408 ip->i_mode = 0;
409 ip->i_flags = 0;
410 /* now we want to make sure that the block pointers are zeroed out */
411 for (i = 0; i < NDADDR; i++)
412 ip->i_db[i] = 0;
413 for (i = 0; i < NIADDR; i++)
414 ip->i_ib[i] = 0;
415
416 /*
417 * Set up a new generation number for this inode.
418 * XXX check if this makes sense in ext2
419 */
420 if (ip->i_gen == 0 || ++ip->i_gen == 0)
421 ip->i_gen = random() / 2 + 1;
422
423 vfs_timestamp(&ts);
424 ip->i_birthtime = ts.tv_sec;
425 ip->i_birthnsec = ts.tv_nsec;
426
427/*
428printf("ext2_valloc: allocated inode %d\n", ino);
429*/
430 return (0);
431noinodes:
432 EXT2_UNLOCK(ump);
433 ext2_fserr(fs, cred->cr_uid, "out of inodes");
434 uprintf("\n%s: create/symlink failed, no inodes free\n", fs->e2fs_fsmnt);
435 return (ENOSPC);
436}
437
438/*
439 * Find a cylinder to place a directory.
440 *
441 * The policy implemented by this algorithm is to allocate a
442 * directory inode in the same cylinder group as its parent
443 * directory, but also to reserve space for its files inodes
444 * and data. Restrict the number of directories which may be
445 * allocated one after another in the same cylinder group
446 * without intervening allocation of files.
447 *
448 * If we allocate a first level directory then force allocation
449 * in another cylinder group.
450 *
451 */
452static u_long
453ext2_dirpref(struct inode *pip)
454{
455 struct m_ext2fs *fs;
456 int cg, prefcg, dirsize, cgsize;
457 int avgifree, avgbfree, avgndir, curdirsize;
458 int minifree, minbfree, maxndir;
459 int mincg, minndir;
460 int maxcontigdirs;
461
462 mtx_assert(EXT2_MTX(pip->i_ump), MA_OWNED);
463 fs = pip->i_e2fs;
464
465 avgifree = fs->e2fs->e2fs_ficount / fs->e2fs_gcount;
466 avgbfree = fs->e2fs->e2fs_fbcount / fs->e2fs_gcount;
467 avgndir = fs->e2fs_total_dir / fs->e2fs_gcount;
468
469 /*
470 * Force allocation in another cg if creating a first level dir.
471 */
472 ASSERT_VOP_LOCKED(ITOV(pip), "ext2fs_dirpref");
473 if (ITOV(pip)->v_vflag & VV_ROOT) {
474 prefcg = arc4random() % fs->e2fs_gcount;
475 mincg = prefcg;
476 minndir = fs->e2fs_ipg;
477 for (cg = prefcg; cg < fs->e2fs_gcount; cg++)
478 if (fs->e2fs_gd[cg].ext2bgd_ndirs < minndir &&
479 fs->e2fs_gd[cg].ext2bgd_nifree >= avgifree &&
480 fs->e2fs_gd[cg].ext2bgd_nbfree >= avgbfree) {
481 mincg = cg;
482 minndir = fs->e2fs_gd[cg].ext2bgd_ndirs;
483 }
484 for (cg = 0; cg < prefcg; cg++)
485 if (fs->e2fs_gd[cg].ext2bgd_ndirs < minndir &&
486 fs->e2fs_gd[cg].ext2bgd_nifree >= avgifree &&
487 fs->e2fs_gd[cg].ext2bgd_nbfree >= avgbfree) {
488 mincg = cg;
489 minndir = fs->e2fs_gd[cg].ext2bgd_ndirs;
490 }
491
492 return (mincg);
493 }
494
495 /*
496 * Count various limits which used for
497 * optimal allocation of a directory inode.
498 */
499 maxndir = min(avgndir + fs->e2fs_ipg / 16, fs->e2fs_ipg);
500 minifree = avgifree - avgifree / 4;
501 if (minifree < 1)
502 minifree = 1;
503 minbfree = avgbfree - avgbfree / 4;
504 if (minbfree < 1)
505 minbfree = 1;
506 cgsize = fs->e2fs_fsize * fs->e2fs_fpg;
507 dirsize = AVGDIRSIZE;
508 curdirsize = avgndir ? (cgsize - avgbfree * fs->e2fs_bsize) / avgndir : 0;
509 if (dirsize < curdirsize)
510 dirsize = curdirsize;
511 if (dirsize <= 0)
512 maxcontigdirs = 0; /* dirsize overflowed */
513 else
514 maxcontigdirs = min((avgbfree * fs->e2fs_bsize) / dirsize, 255);
515 maxcontigdirs = min(maxcontigdirs, fs->e2fs_ipg / AFPDIR);
516 if (maxcontigdirs == 0)
517 maxcontigdirs = 1;
518
519 /*
520 * Limit number of dirs in one cg and reserve space for
521 * regular files, but only if we have no deficit in
522 * inodes or space.
523 */
524 prefcg = ino_to_cg(fs, pip->i_number);
525 for (cg = prefcg; cg < fs->e2fs_gcount; cg++)
526 if (fs->e2fs_gd[cg].ext2bgd_ndirs < maxndir &&
527 fs->e2fs_gd[cg].ext2bgd_nifree >= minifree &&
528 fs->e2fs_gd[cg].ext2bgd_nbfree >= minbfree) {
529 if (fs->e2fs_contigdirs[cg] < maxcontigdirs)
530 return (cg);
531 }
532 for (cg = 0; cg < prefcg; cg++)
533 if (fs->e2fs_gd[cg].ext2bgd_ndirs < maxndir &&
534 fs->e2fs_gd[cg].ext2bgd_nifree >= minifree &&
535 fs->e2fs_gd[cg].ext2bgd_nbfree >= minbfree) {
536 if (fs->e2fs_contigdirs[cg] < maxcontigdirs)
537 return (cg);
538 }
539 /*
540 * This is a backstop when we have deficit in space.
541 */
542 for (cg = prefcg; cg < fs->e2fs_gcount; cg++)
543 if (fs->e2fs_gd[cg].ext2bgd_nifree >= avgifree)
544 return (cg);
545 for (cg = 0; cg < prefcg; cg++)
546 if (fs->e2fs_gd[cg].ext2bgd_nifree >= avgifree)
547 break;
548 return (cg);
549}
550
551/*
552 * Select the desired position for the next block in a file.
553 *
554 * we try to mimic what Remy does in inode_getblk/block_getblk
555 *
556 * we note: blocknr == 0 means that we're about to allocate either
557 * a direct block or a pointer block at the first level of indirection
558 * (In other words, stuff that will go in i_db[] or i_ib[])
559 *
560 * blocknr != 0 means that we're allocating a block that is none
561 * of the above. Then, blocknr tells us the number of the block
562 * that will hold the pointer
563 */
564int32_t
565ext2_blkpref(ip, lbn, indx, bap, blocknr)
566 struct inode *ip;
567 int32_t lbn;
568 int indx;
569 int32_t *bap;
570 int32_t blocknr;
571{
572 int tmp;
573 mtx_assert(EXT2_MTX(ip->i_ump), MA_OWNED);
574
575 /* if the next block is actually what we thought it is,
576 then set the goal to what we thought it should be
577 */
578 if (ip->i_next_alloc_block == lbn && ip->i_next_alloc_goal != 0)
579 return ip->i_next_alloc_goal;
580
581 /* now check whether we were provided with an array that basically
582 tells us previous blocks to which we want to stay closeby
583 */
584 if (bap)
585 for (tmp = indx - 1; tmp >= 0; tmp--)
586 if (bap[tmp])
587 return bap[tmp];
588
589 /* else let's fall back to the blocknr, or, if there is none,
590 follow the rule that a block should be allocated near its inode
591 */
592 return blocknr ? blocknr :
593 (int32_t)(ip->i_block_group *
594 EXT2_BLOCKS_PER_GROUP(ip->i_e2fs)) +
595 ip->i_e2fs->e2fs->e2fs_first_dblock;
596}
597
598/*
599 * Implement the cylinder overflow algorithm.
600 *
601 * The policy implemented by this algorithm is:
602 * 1) allocate the block in its requested cylinder group.
603 * 2) quadradically rehash on the cylinder group number.
604 * 3) brute force search for a free block.
605 */
606static u_long
607ext2_hashalloc(struct inode *ip, int cg, long pref, int size,
608 daddr_t (*allocator)(struct inode *, int, daddr_t, int))
609{
610 struct m_ext2fs *fs;
611 ino_t result;
612 int i, icg = cg;
613
614 mtx_assert(EXT2_MTX(ip->i_ump), MA_OWNED);
615 fs = ip->i_e2fs;
616 /*
617 * 1: preferred cylinder group
618 */
619 result = (*allocator)(ip, cg, pref, size);
620 if (result)
621 return (result);
622 /*
623 * 2: quadratic rehash
624 */
625 for (i = 1; i < fs->e2fs_gcount; i *= 2) {
626 cg += i;
627 if (cg >= fs->e2fs_gcount)
628 cg -= fs->e2fs_gcount;
629 result = (*allocator)(ip, cg, 0, size);
630 if (result)
631 return (result);
632 }
633 /*
634 * 3: brute force search
635 * Note that we start at i == 2, since 0 was checked initially,
636 * and 1 is always checked in the quadratic rehash.
637 */
638 cg = (icg + 2) % fs->e2fs_gcount;
639 for (i = 2; i < fs->e2fs_gcount; i++) {
640 result = (*allocator)(ip, cg, 0, size);
641 if (result)
642 return (result);
643 cg++;
644 if (cg == fs->e2fs_gcount)
645 cg = 0;
646 }
647 return (0);
648}
649
650/*
651 * Determine whether a block can be allocated.
652 *
653 * Check to see if a block of the appropriate size is available,
654 * and if it is, allocate it.
655 */
656static daddr_t
657ext2_alloccg(struct inode *ip, int cg, daddr_t bpref, int size)
658{
659 struct m_ext2fs *fs;
660 struct buf *bp;
661 struct ext2mount *ump;
662 daddr_t bno, runstart, runlen;
663 int bit, loc, end, error, start;
664 char *bbp;
665 /* XXX ondisk32 */
666 fs = ip->i_e2fs;
667 ump = ip->i_ump;
668 if (fs->e2fs_gd[cg].ext2bgd_nbfree == 0)
669 return (0);
670 EXT2_UNLOCK(ump);
671 error = bread(ip->i_devvp, fsbtodb(fs,
672 fs->e2fs_gd[cg].ext2bgd_b_bitmap),
673 (int)fs->e2fs_bsize, NOCRED, &bp);
674 if (error) {
675 brelse(bp);
676 EXT2_LOCK(ump);
677 return (0);
678 }
679 if (fs->e2fs_gd[cg].ext2bgd_nbfree == 0) {
680 /*
681 * Another thread allocated the last block in this
682 * group while we were waiting for the buffer.
683 */
684 brelse(bp);
685 EXT2_LOCK(ump);
686 return (0);
687 }
688 bbp = (char *)bp->b_data;
689
690 if (dtog(fs, bpref) != cg)
691 bpref = 0;
692 if (bpref != 0) {
693 bpref = dtogd(fs, bpref);
694 /*
695 * if the requested block is available, use it
696 */
697 if (isclr(bbp, bpref)) {
698 bno = bpref;
699 goto gotit;
700 }
701 }
702 /*
703 * no blocks in the requested cylinder, so take next
704 * available one in this cylinder group.
705 * first try to get 8 contigous blocks, then fall back to a single
706 * block.
707 */
708 if (bpref)
709 start = dtogd(fs, bpref) / NBBY;
710 else
711 start = 0;
712 end = howmany(fs->e2fs->e2fs_fpg, NBBY) - start;
713retry:
714 runlen = 0;
715 runstart = 0;
716 for (loc = start; loc < end; loc++) {
717 if (bbp[loc] == (char)0xff) {
718 runlen = 0;
719 continue;
720 }
721
722 /* Start of a run, find the number of high clear bits. */
723 if (runlen == 0) {
724 bit = fls(bbp[loc]);
725 runlen = NBBY - bit;
726 runstart = loc * NBBY + bit;
727 } else if (bbp[loc] == 0) {
728 /* Continue a run. */
729 runlen += NBBY;
730 } else {
731 /*
732 * Finish the current run. If it isn't long
733 * enough, start a new one.
734 */
735 bit = ffs(bbp[loc]) - 1;
736 runlen += bit;
737 if (runlen >= 8) {
738 bno = runstart;
739 goto gotit;
740 }
741
742 /* Run was too short, start a new one. */
743 bit = fls(bbp[loc]);
744 runlen = NBBY - bit;
745 runstart = loc * NBBY + bit;
746 }
747
748 /* If the current run is long enough, use it. */
749 if (runlen >= 8) {
750 bno = runstart;
751 goto gotit;
752 }
753 }
754 if (start != 0) {
755 end = start;
756 start = 0;
757 goto retry;
758 }
759
760 bno = ext2_mapsearch(fs, bbp, bpref);
761 if (bno < 0){
762 brelse(bp);
763 EXT2_LOCK(ump);
764 return (0);
765 }
766gotit:
767#ifdef DIAGNOSTIC
768 if (isset(bbp, bno)) {
769 printf("ext2fs_alloccgblk: cg=%d bno=%jd fs=%s\n",
770 cg, (intmax_t)bno, fs->e2fs_fsmnt);
771 panic("ext2fs_alloccg: dup alloc");
772 }
773#endif
774 setbit(bbp, bno);
775 EXT2_LOCK(ump);
776 ext2_clusteracct(fs, bbp, cg, bno, -1);
777 fs->e2fs->e2fs_fbcount--;
778 fs->e2fs_gd[cg].ext2bgd_nbfree--;
779 fs->e2fs_fmod = 1;
780 EXT2_UNLOCK(ump);
781 bdwrite(bp);
782 return (cg * fs->e2fs->e2fs_fpg + fs->e2fs->e2fs_first_dblock + bno);
783}
784
785/*
786 * Determine whether a cluster can be allocated.
787 */
788static daddr_t
789ext2_clusteralloc(struct inode *ip, int cg, daddr_t bpref, int len)
790{
791 struct m_ext2fs *fs;
792 struct ext2mount *ump;
793 struct buf *bp;
794 char *bbp;
795 int bit, error, got, i, loc, run;
796 int32_t *lp;
797 daddr_t bno;
798
799 fs = ip->i_e2fs;
800 ump = ip->i_ump;
801
802 if (fs->e2fs_maxcluster[cg] < len)
803 return (0);
804
805 EXT2_UNLOCK(ump);
806 error = bread(ip->i_devvp,
807 fsbtodb(fs, fs->e2fs_gd[cg].ext2bgd_b_bitmap),
808 (int)fs->e2fs_bsize, NOCRED, &bp);
809 if (error)
810 goto fail_lock;
811
812 bbp = (char *)bp->b_data;
813 bp->b_xflags |= BX_BKGRDWRITE;
814
815 EXT2_LOCK(ump);
816 /*
817 * Check to see if a cluster of the needed size (or bigger) is
818 * available in this cylinder group.
819 */
820 lp = &fs->e2fs_clustersum[cg].cs_sum[len];
821 for (i = len; i <= fs->e2fs_contigsumsize; i++)
822 if (*lp++ > 0)
823 break;
824 if (i > fs->e2fs_contigsumsize) {
825 /*
826 * Update the cluster summary information to reflect
827 * the true maximum-sized cluster so that future cluster
828 * allocation requests can avoid reading the bitmap only
829 * to find no cluster.
830 */
831 lp = &fs->e2fs_clustersum[cg].cs_sum[len - 1];
832 for (i = len - 1; i > 0; i--)
833 if (*lp-- > 0)
834 break;
835 fs->e2fs_maxcluster[cg] = i;
836 goto fail;
837 }
838 EXT2_UNLOCK(ump);
839
840 /* Search the bitmap to find a big enough cluster like in FFS. */
841 if (dtog(fs, bpref) != cg)
842 bpref = 0;
843 if (bpref != 0)
844 bpref = dtogd(fs, bpref);
845 loc = bpref / NBBY;
846 bit = 1 << (bpref % NBBY);
847 for (run = 0, got = bpref; got < fs->e2fs->e2fs_fpg; got++) {
848 if ((bbp[loc] & bit) != 0)
849 run = 0;
850 else {
851 run++;
852 if (run == len)
853 break;
854 }
855 if ((got & (NBBY - 1)) != (NBBY - 1))
856 bit <<= 1;
857 else {
858 loc++;
859 bit = 1;
860 }
861 }
862
863 if (got >= fs->e2fs->e2fs_fpg)
864 goto fail_lock;
865
866 /* Allocate the cluster that we found. */
867 for (i = 1; i < len; i++)
868 if (!isclr(bbp, got - run + i))
869 panic("ext2_clusteralloc: map mismatch");
870
871 bno = got - run + 1;
872 if (bno >= fs->e2fs->e2fs_fpg)
873 panic("ext2_clusteralloc: allocated out of group");
874
875 EXT2_LOCK(ump);
876 for (i = 0; i < len; i += fs->e2fs_fpb) {
877 setbit(bbp, bno + i);
878 ext2_clusteracct(fs, bbp, cg, bno + i, -1);
879 fs->e2fs->e2fs_fbcount--;
880 fs->e2fs_gd[cg].ext2bgd_nbfree--;
881 }
882 fs->e2fs_fmod = 1;
883 EXT2_UNLOCK(ump);
884
885 bdwrite(bp);
886 return (cg * fs->e2fs->e2fs_fpg + fs->e2fs->e2fs_first_dblock + bno);
887
888fail_lock:
889 EXT2_LOCK(ump);
890fail:
891 brelse(bp);
892 return (0);
893}
894
895/*
896 * Determine whether an inode can be allocated.
897 *
898 * Check to see if an inode is available, and if it is,
899 * allocate it using tode in the specified cylinder group.
900 */
901static daddr_t
902ext2_nodealloccg(struct inode *ip, int cg, daddr_t ipref, int mode)
903{
904 struct m_ext2fs *fs;
905 struct buf *bp;
906 struct ext2mount *ump;
907 int error, start, len;
908 char *ibp, *loc;
909 ipref--; /* to avoid a lot of (ipref -1) */
910 if (ipref == -1)
911 ipref = 0;
912 fs = ip->i_e2fs;
913 ump = ip->i_ump;
914 if (fs->e2fs_gd[cg].ext2bgd_nifree == 0)
915 return (0);
916 EXT2_UNLOCK(ump);
917 error = bread(ip->i_devvp, fsbtodb(fs,
918 fs->e2fs_gd[cg].ext2bgd_i_bitmap),
919 (int)fs->e2fs_bsize, NOCRED, &bp);
920 if (error) {
921 brelse(bp);
922 EXT2_LOCK(ump);
923 return (0);
924 }
925 if (fs->e2fs_gd[cg].ext2bgd_nifree == 0) {
926 /*
927 * Another thread allocated the last i-node in this
928 * group while we were waiting for the buffer.
929 */
930 brelse(bp);
931 EXT2_LOCK(ump);
932 return (0);
933 }
934 ibp = (char *)bp->b_data;
935 if (ipref) {
936 ipref %= fs->e2fs->e2fs_ipg;
937 if (isclr(ibp, ipref))
938 goto gotit;
939 }
940 start = ipref / NBBY;
941 len = howmany(fs->e2fs->e2fs_ipg - ipref, NBBY);
942 loc = memcchr(&ibp[start], 0xff, len);
943 if (loc == NULL) {
944 len = start + 1;
945 start = 0;
946 loc = memcchr(&ibp[start], 0xff, len);
947 if (loc == NULL) {
948 printf("cg = %d, ipref = %lld, fs = %s\n",
949 cg, (long long)ipref, fs->e2fs_fsmnt);
950 panic("ext2fs_nodealloccg: map corrupted");
951 /* NOTREACHED */
952 }
953 }
954 ipref = (loc - ibp) * NBBY + ffs(~*loc) - 1;
955gotit:
956 setbit(ibp, ipref);
957 EXT2_LOCK(ump);
958 fs->e2fs_gd[cg].ext2bgd_nifree--;
959 fs->e2fs->e2fs_ficount--;
960 fs->e2fs_fmod = 1;
961 if ((mode & IFMT) == IFDIR) {
962 fs->e2fs_gd[cg].ext2bgd_ndirs++;
963 fs->e2fs_total_dir++;
964 }
965 EXT2_UNLOCK(ump);
966 bdwrite(bp);
967 return (cg * fs->e2fs->e2fs_ipg + ipref +1);
968}
969
970/*
971 * Free a block or fragment.
972 *
973 */
974void
975ext2_blkfree(ip, bno, size)
976 struct inode *ip;
977 int32_t bno;
978 long size;
979{
980 struct m_ext2fs *fs;
981 struct buf *bp;
982 struct ext2mount *ump;
983 int cg, error;
984 char *bbp;
985
986 fs = ip->i_e2fs;
987 ump = ip->i_ump;
988 cg = dtog(fs, bno);
989 if ((u_int)bno >= fs->e2fs->e2fs_bcount) {
990 printf("bad block %lld, ino %llu\n", (long long)bno,
991 (unsigned long long)ip->i_number);
992 ext2_fserr(fs, ip->i_uid, "bad block");
993 return;
994 }
995 error = bread(ip->i_devvp,
996 fsbtodb(fs, fs->e2fs_gd[cg].ext2bgd_b_bitmap),
997 (int)fs->e2fs_bsize, NOCRED, &bp);
998 if (error) {
999 brelse(bp);
1000 return;
1001 }
1002 bbp = (char *)bp->b_data;
1003 bno = dtogd(fs, bno);
1004 if (isclr(bbp, bno)) {
1005 printf("block = %lld, fs = %s\n",
1006 (long long)bno, fs->e2fs_fsmnt);
1007 panic("blkfree: freeing free block");
1008 }
1009 clrbit(bbp, bno);
1010 EXT2_LOCK(ump);
1011 ext2_clusteracct(fs, bbp, cg, bno, 1);
1012 fs->e2fs->e2fs_fbcount++;
1013 fs->e2fs_gd[cg].ext2bgd_nbfree++;
1014 fs->e2fs_fmod = 1;
1015 EXT2_UNLOCK(ump);
1016 bdwrite(bp);
1017}
1018
1019/*
1020 * Free an inode.
1021 *
1022 */
1023int
1024ext2_vfree(pvp, ino, mode)
1025 struct vnode *pvp;
1026 ino_t ino;
1027 int mode;
1028{
1029 struct m_ext2fs *fs;
1030 struct inode *pip;
1031 struct buf *bp;
1032 struct ext2mount *ump;
1033 int error, cg;
1034 char * ibp;
1035/* mode_t save_i_mode; */
1036
1037 pip = VTOI(pvp);
1038 fs = pip->i_e2fs;
1039 ump = pip->i_ump;
1040 if ((u_int)ino > fs->e2fs_ipg * fs->e2fs_gcount)
1041 panic("ext2_vfree: range: devvp = %p, ino = %ju, fs = %s",
1042 pip->i_devvp, (uintmax_t)ino, fs->e2fs_fsmnt);
1043
1044 cg = ino_to_cg(fs, ino);
1045 error = bread(pip->i_devvp,
1046 fsbtodb(fs, fs->e2fs_gd[cg].ext2bgd_i_bitmap),
1047 (int)fs->e2fs_bsize, NOCRED, &bp);
1048 if (error) {
1049 brelse(bp);
1050 return (0);
1051 }
1052 ibp = (char *)bp->b_data;
1053 ino = (ino - 1) % fs->e2fs->e2fs_ipg;
1054 if (isclr(ibp, ino)) {
1055 printf("ino = %llu, fs = %s\n",
1056 (unsigned long long)ino, fs->e2fs_fsmnt);
1057 if (fs->e2fs_ronly == 0)
1058 panic("ifree: freeing free inode");
1059 }
1060 clrbit(ibp, ino);
1061 EXT2_LOCK(ump);
1062 fs->e2fs->e2fs_ficount++;
1063 fs->e2fs_gd[cg].ext2bgd_nifree++;
1064 if ((mode & IFMT) == IFDIR) {
1065 fs->e2fs_gd[cg].ext2bgd_ndirs--;
1066 fs->e2fs_total_dir--;
1067 }
1068 fs->e2fs_fmod = 1;
1069 EXT2_UNLOCK(ump);
1070 bdwrite(bp);
1071 return (0);
1072}
1073
1074/*
1075 * Find a block in the specified cylinder group.
1076 *
1077 * It is a panic if a request is made to find a block if none are
1078 * available.
1079 */
1080static daddr_t
1081ext2_mapsearch(struct m_ext2fs *fs, char *bbp, daddr_t bpref)
1082{
1083 char *loc;
1084 int start, len;
1085
1086 /*
1087 * find the fragment by searching through the free block
1088 * map for an appropriate bit pattern
1089 */
1090 if (bpref)
1091 start = dtogd(fs, bpref) / NBBY;
1092 else
1093 start = 0;
1094 len = howmany(fs->e2fs->e2fs_fpg, NBBY) - start;
1095 loc = memcchr(&bbp[start], 0xff, len);
1096 if (loc == NULL) {
1097 len = start + 1;
1098 start = 0;
1099 loc = memcchr(&bbp[start], 0xff, len);
1100 if (loc == NULL) {
1101 printf("start = %d, len = %d, fs = %s\n",
1102 start, len, fs->e2fs_fsmnt);
1103 panic("ext2fs_alloccg: map corrupted");
1104 /* NOTREACHED */
1105 }
1106 }
1107 return ((loc - bbp) * NBBY + ffs(~*loc) - 1);
1108}
1109
1110/*
1111 * Fserr prints the name of a file system with an error diagnostic.
1112 *
1113 * The form of the error message is:
1114 * fs: error message
1115 */
1116static void
1117ext2_fserr(fs, uid, cp)
1118 struct m_ext2fs *fs;
1119 uid_t uid;
1120 char *cp;
1121{
1122
1123 log(LOG_ERR, "uid %u on %s: %s\n", uid, fs->e2fs_fsmnt, cp);
1124}
1125
1126int
1127cg_has_sb(int i)
1128{
1129 int a3, a5, a7;
1130
1131 if (i == 0 || i == 1)
1132 return 1;
1133 for (a3 = 3, a5 = 5, a7 = 7;
1134 a3 <= i || a5 <= i || a7 <= i;
1135 a3 *= 3, a5 *= 5, a7 *= 7)
1136 if (i == a3 || i == a5 || i == a7)
1137 return 1;
1138 return 0;
1139}
|