1 /* 2 * Copyright (c) 1982, 1986, 1989, 1993 3 * The Regents of the University of California. All rights reserved. 4 * 5 * Redistribution and use in source and binary forms, with or without 6 * modification, are permitted provided that the following conditions 7 * are met: 8 * 1. Redistributions of source code must retain the above copyright 9 * notice, this list of conditions and the following disclaimer. 10 * 2. Redistributions in binary form must reproduce the above copyright 11 * notice, this list of conditions and the following disclaimer in the 12 * documentation and/or other materials provided with the distribution. 13 * 3. All advertising materials mentioning features or use of this software 14 * must display the following acknowledgement: 15 * This product includes software developed by the University of 16 * California, Berkeley and its contributors. 17 * 4. Neither the name of the University nor the names of its contributors 18 * may be used to endorse or promote products derived from this software 19 * without specific prior written permission. 20 * 21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 24 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 31 * SUCH DAMAGE. 32 * 33 * @(#)ffs_alloc.c 8.18 (Berkeley) 5/26/95 34 * $FreeBSD: src/sys/ufs/ffs/ffs_alloc.c,v 1.64.2.2 2001/09/21 19:15:21 dillon Exp $ 35 */ 36 37 #include "opt_quota.h" 38 39 #include <sys/param.h> 40 #include <sys/systm.h> 41 #include <sys/buf.h> 42 #include <sys/conf.h> 43 #include <sys/proc.h> 44 #include <sys/vnode.h> 45 #include <sys/mount.h> 46 #include <sys/kernel.h> 47 #include <sys/sysctl.h> 48 #include <sys/syslog.h> 49 50 #include <sys/taskqueue.h> 51 #include <machine/inttypes.h> 52 53 #include <sys/buf2.h> 54 55 #include "quota.h" 56 #include "inode.h" 57 #include "ufs_extern.h" 58 #include "ufsmount.h" 59 60 #include "fs.h" 61 #include "ffs_extern.h" 62 63 typedef ufs_daddr_t allocfcn_t (struct inode *ip, int cg, ufs_daddr_t bpref, 64 int size); 65 66 static ufs_daddr_t ffs_alloccg (struct inode *, int, ufs_daddr_t, int); 67 static ufs_daddr_t 68 ffs_alloccgblk (struct inode *, struct buf *, ufs_daddr_t); 69 static void ffs_blkfree_cg(struct fs *, struct vnode *, cdev_t , ino_t, 70 uint32_t , ufs_daddr_t, long ); 71 #ifdef DIAGNOSTIC 72 static int ffs_checkblk (struct inode *, ufs_daddr_t, long); 73 #endif 74 static void ffs_clusteracct (struct fs *, struct cg *, ufs_daddr_t, 75 int); 76 static ufs_daddr_t ffs_clusteralloc (struct inode *, int, ufs_daddr_t, 77 int); 78 static ino_t ffs_dirpref (struct inode *); 79 static ufs_daddr_t ffs_fragextend (struct inode *, int, long, int, int); 80 static void ffs_fserr (struct fs *, uint, char *); 81 static u_long ffs_hashalloc 82 (struct inode *, int, long, int, allocfcn_t *); 83 static ino_t ffs_nodealloccg (struct inode *, int, ufs_daddr_t, int); 84 static ufs_daddr_t ffs_mapsearch (struct fs *, struct cg *, ufs_daddr_t, 85 int); 86 87 /* 88 * Allocate a block in the filesystem. 89 * 90 * The size of the requested block is given, which must be some 91 * multiple of fs_fsize and <= fs_bsize. 92 * A preference may be optionally specified. If a preference is given 93 * the following hierarchy is used to allocate a block: 94 * 1) allocate the requested block. 95 * 2) allocate a rotationally optimal block in the same cylinder. 96 * 3) allocate a block in the same cylinder group. 97 * 4) quadradically rehash into other cylinder groups, until an 98 * available block is located. 99 * If no block preference is given the following heirarchy is used 100 * to allocate a block: 101 * 1) allocate a block in the cylinder group that contains the 102 * inode for the file. 103 * 2) quadradically rehash into other cylinder groups, until an 104 * available block is located. 105 */ 106 int 107 ffs_alloc(struct inode *ip, ufs_daddr_t lbn, ufs_daddr_t bpref, int size, 108 struct ucred *cred, ufs_daddr_t *bnp) 109 { 110 struct fs *fs; 111 ufs_daddr_t bno; 112 int cg; 113 #ifdef QUOTA 114 int error; 115 #endif 116 117 *bnp = 0; 118 fs = ip->i_fs; 119 #ifdef DIAGNOSTIC 120 if ((uint)size > fs->fs_bsize || fragoff(fs, size) != 0) { 121 kprintf("dev = %s, bsize = %ld, size = %d, fs = %s\n", 122 devtoname(ip->i_dev), (long)fs->fs_bsize, size, 123 fs->fs_fsmnt); 124 panic("ffs_alloc: bad size"); 125 } 126 if (cred == NOCRED) 127 panic("ffs_alloc: missing credential"); 128 #endif /* DIAGNOSTIC */ 129 if (size == fs->fs_bsize && fs->fs_cstotal.cs_nbfree == 0) 130 goto nospace; 131 if (cred->cr_uid != 0 && 132 freespace(fs, fs->fs_minfree) - numfrags(fs, size) < 0) 133 goto nospace; 134 #ifdef QUOTA 135 error = ufs_chkdq(ip, (long)btodb(size), cred, 0); 136 if (error) 137 return (error); 138 #endif 139 if (bpref >= fs->fs_size) 140 bpref = 0; 141 if (bpref == 0) 142 cg = ino_to_cg(fs, ip->i_number); 143 else 144 cg = dtog(fs, bpref); 145 bno = (ufs_daddr_t)ffs_hashalloc(ip, cg, (long)bpref, size, 146 ffs_alloccg); 147 if (bno > 0) { 148 ip->i_blocks += btodb(size); 149 ip->i_flag |= IN_CHANGE | IN_UPDATE; 150 *bnp = bno; 151 return (0); 152 } 153 #ifdef QUOTA 154 /* 155 * Restore user's disk quota because allocation failed. 156 */ 157 (void) ufs_chkdq(ip, (long)-btodb(size), cred, FORCE); 158 #endif 159 nospace: 160 ffs_fserr(fs, cred->cr_uid, "filesystem full"); 161 uprintf("\n%s: write failed, filesystem is full\n", fs->fs_fsmnt); 162 return (ENOSPC); 163 } 164 165 /* 166 * Reallocate a fragment to a bigger size 167 * 168 * The number and size of the old block is given, and a preference 169 * and new size is also specified. The allocator attempts to extend 170 * the original block. Failing that, the regular block allocator is 171 * invoked to get an appropriate block. 172 */ 173 int 174 ffs_realloccg(struct inode *ip, ufs_daddr_t lbprev, ufs_daddr_t bpref, 175 int osize, int nsize, struct ucred *cred, struct buf **bpp) 176 { 177 struct fs *fs; 178 struct buf *bp; 179 int cg, request, error; 180 ufs_daddr_t bprev, bno; 181 182 *bpp = NULL; 183 fs = ip->i_fs; 184 #ifdef DIAGNOSTIC 185 if ((uint)osize > fs->fs_bsize || fragoff(fs, osize) != 0 || 186 (uint)nsize > fs->fs_bsize || fragoff(fs, nsize) != 0) { 187 kprintf( 188 "dev = %s, bsize = %ld, osize = %d, nsize = %d, fs = %s\n", 189 devtoname(ip->i_dev), (long)fs->fs_bsize, osize, 190 nsize, fs->fs_fsmnt); 191 panic("ffs_realloccg: bad size"); 192 } 193 if (cred == NOCRED) 194 panic("ffs_realloccg: missing credential"); 195 #endif /* DIAGNOSTIC */ 196 if (cred->cr_uid != 0 && 197 freespace(fs, fs->fs_minfree) - numfrags(fs, nsize - osize) < 0) 198 goto nospace; 199 if ((bprev = ip->i_db[lbprev]) == 0) { 200 kprintf("dev = %s, bsize = %ld, bprev = %ld, fs = %s\n", 201 devtoname(ip->i_dev), (long)fs->fs_bsize, (long)bprev, 202 fs->fs_fsmnt); 203 panic("ffs_realloccg: bad bprev"); 204 } 205 /* 206 * Allocate the extra space in the buffer. 207 */ 208 error = bread(ITOV(ip), lblktodoff(fs, lbprev), osize, &bp); 209 if (error) { 210 brelse(bp); 211 return (error); 212 } 213 214 if(bp->b_bio2.bio_offset == NOOFFSET) { 215 if( lbprev >= NDADDR) 216 panic("ffs_realloccg: lbprev out of range"); 217 bp->b_bio2.bio_offset = fsbtodoff(fs, bprev); 218 } 219 220 #ifdef QUOTA 221 error = ufs_chkdq(ip, (long)btodb(nsize - osize), cred, 0); 222 if (error) { 223 brelse(bp); 224 return (error); 225 } 226 #endif 227 /* 228 * Check for extension in the existing location. 229 */ 230 cg = dtog(fs, bprev); 231 bno = ffs_fragextend(ip, cg, (long)bprev, osize, nsize); 232 if (bno) { 233 if (bp->b_bio2.bio_offset != fsbtodoff(fs, bno)) 234 panic("ffs_realloccg: bad blockno"); 235 ip->i_blocks += btodb(nsize - osize); 236 ip->i_flag |= IN_CHANGE | IN_UPDATE; 237 allocbuf(bp, nsize); 238 bzero((char *)bp->b_data + osize, (uint)nsize - osize); 239 *bpp = bp; 240 return (0); 241 } 242 /* 243 * Allocate a new disk location. 244 */ 245 if (bpref >= fs->fs_size) 246 bpref = 0; 247 switch ((int)fs->fs_optim) { 248 case FS_OPTSPACE: 249 /* 250 * Allocate an exact sized fragment. Although this makes 251 * best use of space, we will waste time relocating it if 252 * the file continues to grow. If the fragmentation is 253 * less than half of the minimum free reserve, we choose 254 * to begin optimizing for time. 255 */ 256 request = nsize; 257 if (fs->fs_minfree <= 5 || 258 fs->fs_cstotal.cs_nffree > 259 (off_t)fs->fs_dsize * fs->fs_minfree / (2 * 100)) 260 break; 261 log(LOG_NOTICE, "%s: optimization changed from SPACE to TIME\n", 262 fs->fs_fsmnt); 263 fs->fs_optim = FS_OPTTIME; 264 break; 265 case FS_OPTTIME: 266 /* 267 * At this point we have discovered a file that is trying to 268 * grow a small fragment to a larger fragment. To save time, 269 * we allocate a full sized block, then free the unused portion. 270 * If the file continues to grow, the `ffs_fragextend' call 271 * above will be able to grow it in place without further 272 * copying. If aberrant programs cause disk fragmentation to 273 * grow within 2% of the free reserve, we choose to begin 274 * optimizing for space. 275 */ 276 request = fs->fs_bsize; 277 if (fs->fs_cstotal.cs_nffree < 278 (off_t)fs->fs_dsize * (fs->fs_minfree - 2) / 100) 279 break; 280 log(LOG_NOTICE, "%s: optimization changed from TIME to SPACE\n", 281 fs->fs_fsmnt); 282 fs->fs_optim = FS_OPTSPACE; 283 break; 284 default: 285 kprintf("dev = %s, optim = %ld, fs = %s\n", 286 devtoname(ip->i_dev), (long)fs->fs_optim, fs->fs_fsmnt); 287 panic("ffs_realloccg: bad optim"); 288 /* NOTREACHED */ 289 } 290 bno = (ufs_daddr_t)ffs_hashalloc(ip, cg, (long)bpref, request, 291 ffs_alloccg); 292 if (bno > 0) { 293 bp->b_bio2.bio_offset = fsbtodoff(fs, bno); 294 if (!DOINGSOFTDEP(ITOV(ip))) 295 ffs_blkfree(ip, bprev, (long)osize); 296 if (nsize < request) 297 ffs_blkfree(ip, bno + numfrags(fs, nsize), 298 (long)(request - nsize)); 299 ip->i_blocks += btodb(nsize - osize); 300 ip->i_flag |= IN_CHANGE | IN_UPDATE; 301 allocbuf(bp, nsize); 302 bzero((char *)bp->b_data + osize, (uint)nsize - osize); 303 *bpp = bp; 304 return (0); 305 } 306 #ifdef QUOTA 307 /* 308 * Restore user's disk quota because allocation failed. 309 */ 310 (void) ufs_chkdq(ip, (long)-btodb(nsize - osize), cred, FORCE); 311 #endif 312 brelse(bp); 313 nospace: 314 /* 315 * no space available 316 */ 317 ffs_fserr(fs, cred->cr_uid, "filesystem full"); 318 uprintf("\n%s: write failed, filesystem is full\n", fs->fs_fsmnt); 319 return (ENOSPC); 320 } 321 322 SYSCTL_NODE(_vfs, OID_AUTO, ffs, CTLFLAG_RW, 0, "FFS filesystem"); 323 324 /* 325 * Reallocate a sequence of blocks into a contiguous sequence of blocks. 326 * 327 * The vnode and an array of buffer pointers for a range of sequential 328 * logical blocks to be made contiguous is given. The allocator attempts 329 * to find a range of sequential blocks starting as close as possible to 330 * an fs_rotdelay offset from the end of the allocation for the logical 331 * block immediately preceeding the current range. If successful, the 332 * physical block numbers in the buffer pointers and in the inode are 333 * changed to reflect the new allocation. If unsuccessful, the allocation 334 * is left unchanged. The success in doing the reallocation is returned. 335 * Note that the error return is not reflected back to the user. Rather 336 * the previous block allocation will be used. 337 */ 338 static int doasyncfree = 1; 339 SYSCTL_INT(_vfs_ffs, FFS_ASYNCFREE, doasyncfree, CTLFLAG_RW, &doasyncfree, 0, ""); 340 341 static int doreallocblks = 1; 342 SYSCTL_INT(_vfs_ffs, FFS_REALLOCBLKS, doreallocblks, CTLFLAG_RW, &doreallocblks, 0, ""); 343 344 #ifdef DEBUG 345 static volatile int prtrealloc = 0; 346 #endif 347 348 /* 349 * ffs_reallocblks(struct vnode *a_vp, struct cluster_save *a_buflist) 350 */ 351 int 352 ffs_reallocblks(struct vop_reallocblks_args *ap) 353 { 354 struct fs *fs; 355 struct inode *ip; 356 struct vnode *vp; 357 struct buf *sbp, *ebp; 358 ufs_daddr_t *bap, *sbap, *ebap = NULL; 359 struct cluster_save *buflist; 360 ufs_daddr_t start_lbn, end_lbn, soff, newblk, blkno; 361 #ifdef DIAGNOSTIC 362 off_t boffset; 363 #endif 364 struct indir start_ap[NIADDR + 1], end_ap[NIADDR + 1], *idp; 365 int i, len, slen, start_lvl, end_lvl, pref, ssize; 366 367 if (doreallocblks == 0) 368 return (ENOSPC); 369 vp = ap->a_vp; 370 ip = VTOI(vp); 371 fs = ip->i_fs; 372 if (fs->fs_contigsumsize <= 0) 373 return (ENOSPC); 374 buflist = ap->a_buflist; 375 len = buflist->bs_nchildren; 376 start_lbn = lblkno(fs, buflist->bs_children[0]->b_loffset); 377 end_lbn = start_lbn + len - 1; 378 #ifdef DIAGNOSTIC 379 for (i = 0; i < len; i++) 380 if (!ffs_checkblk(ip, 381 dofftofsb(fs, buflist->bs_children[i]->b_bio2.bio_offset), fs->fs_bsize)) 382 panic("ffs_reallocblks: unallocated block 1"); 383 for (i = 1; i < len; i++) { 384 if (buflist->bs_children[i]->b_loffset != lblktodoff(fs, start_lbn) + lblktodoff(fs, i)) 385 panic("ffs_reallocblks: non-logical cluster"); 386 } 387 boffset = buflist->bs_children[0]->b_bio2.bio_offset; 388 ssize = (int)fsbtodoff(fs, fs->fs_frag); 389 for (i = 1; i < len - 1; i++) 390 if (buflist->bs_children[i]->b_bio2.bio_offset != boffset + (i * ssize)) 391 panic("ffs_reallocblks: non-physical cluster %d", i); 392 #endif 393 /* 394 * If the latest allocation is in a new cylinder group, assume that 395 * the filesystem has decided to move and do not force it back to 396 * the previous cylinder group. 397 */ 398 if (dtog(fs, dofftofsb(fs, buflist->bs_children[0]->b_bio2.bio_offset)) != 399 dtog(fs, dofftofsb(fs, buflist->bs_children[len - 1]->b_bio2.bio_offset))) 400 return (ENOSPC); 401 if (ufs_getlbns(vp, start_lbn, start_ap, &start_lvl) || 402 ufs_getlbns(vp, end_lbn, end_ap, &end_lvl)) 403 return (ENOSPC); 404 /* 405 * Get the starting offset and block map for the first block and 406 * the number of blocks that will fit into sbap starting at soff. 407 */ 408 if (start_lvl == 0) { 409 sbap = &ip->i_db[0]; 410 soff = start_lbn; 411 slen = NDADDR - soff; 412 } else { 413 idp = &start_ap[start_lvl - 1]; 414 if (bread(vp, lblktodoff(fs, idp->in_lbn), (int)fs->fs_bsize, &sbp)) { 415 brelse(sbp); 416 return (ENOSPC); 417 } 418 sbap = (ufs_daddr_t *)sbp->b_data; 419 soff = idp->in_off; 420 slen = fs->fs_nindir - soff; 421 } 422 /* 423 * Find the preferred location for the cluster. 424 */ 425 pref = ffs_blkpref(ip, start_lbn, soff, sbap); 426 427 /* 428 * If the block range spans two block maps, get the second map. 429 */ 430 if (end_lvl == 0 || (idp = &end_ap[end_lvl - 1])->in_off + 1 >= len) { 431 ssize = len; 432 } else { 433 #ifdef DIAGNOSTIC 434 if (start_ap[start_lvl-1].in_lbn == idp->in_lbn) 435 panic("ffs_reallocblk: start == end"); 436 #endif 437 ssize = len - (idp->in_off + 1); 438 if (bread(vp, lblktodoff(fs, idp->in_lbn), (int)fs->fs_bsize, &ebp)) 439 goto fail; 440 ebap = (ufs_daddr_t *)ebp->b_data; 441 } 442 443 /* 444 * Make sure we aren't spanning more then two blockmaps. ssize is 445 * our calculation of the span we have to scan in the first blockmap, 446 * while slen is our calculation of the number of entries available 447 * in the first blockmap (from soff). 448 */ 449 if (ssize > slen) { 450 panic("ffs_reallocblks: range spans more then two blockmaps!" 451 " start_lbn %ld len %d (%d/%d)", 452 (long)start_lbn, len, slen, ssize); 453 } 454 /* 455 * Search the block map looking for an allocation of the desired size. 456 */ 457 if ((newblk = (ufs_daddr_t)ffs_hashalloc(ip, dtog(fs, pref), (long)pref, 458 len, ffs_clusteralloc)) == 0) 459 goto fail; 460 /* 461 * We have found a new contiguous block. 462 * 463 * First we have to replace the old block pointers with the new 464 * block pointers in the inode and indirect blocks associated 465 * with the file. 466 */ 467 #ifdef DEBUG 468 if (prtrealloc) 469 kprintf("realloc: ino %ju, lbns %d-%d\n\told:", 470 (uintmax_t)ip->i_number, start_lbn, end_lbn); 471 #endif 472 blkno = newblk; 473 for (bap = &sbap[soff], i = 0; i < len; i++, blkno += fs->fs_frag) { 474 if (i == ssize) { 475 bap = ebap; 476 soff = -i; 477 } 478 #ifdef DIAGNOSTIC 479 if (!ffs_checkblk(ip, 480 dofftofsb(fs, buflist->bs_children[i]->b_bio2.bio_offset), fs->fs_bsize)) 481 panic("ffs_reallocblks: unallocated block 2"); 482 if (dofftofsb(fs, buflist->bs_children[i]->b_bio2.bio_offset) != *bap) 483 panic("ffs_reallocblks: alloc mismatch"); 484 #endif 485 #ifdef DEBUG 486 if (prtrealloc) 487 kprintf(" %d,", *bap); 488 #endif 489 if (DOINGSOFTDEP(vp)) { 490 if (sbap == &ip->i_db[0] && i < ssize) 491 softdep_setup_allocdirect(ip, start_lbn + i, 492 blkno, *bap, fs->fs_bsize, fs->fs_bsize, 493 buflist->bs_children[i]); 494 else 495 softdep_setup_allocindir_page(ip, start_lbn + i, 496 i < ssize ? sbp : ebp, soff + i, blkno, 497 *bap, buflist->bs_children[i]); 498 } 499 *bap++ = blkno; 500 } 501 /* 502 * Next we must write out the modified inode and indirect blocks. 503 * For strict correctness, the writes should be synchronous since 504 * the old block values may have been written to disk. In practise 505 * they are almost never written, but if we are concerned about 506 * strict correctness, the `doasyncfree' flag should be set to zero. 507 * 508 * The test on `doasyncfree' should be changed to test a flag 509 * that shows whether the associated buffers and inodes have 510 * been written. The flag should be set when the cluster is 511 * started and cleared whenever the buffer or inode is flushed. 512 * We can then check below to see if it is set, and do the 513 * synchronous write only when it has been cleared. 514 */ 515 if (sbap != &ip->i_db[0]) { 516 if (doasyncfree) 517 bdwrite(sbp); 518 else 519 bwrite(sbp); 520 } else { 521 ip->i_flag |= IN_CHANGE | IN_UPDATE; 522 if (!doasyncfree) 523 ffs_update(vp, 1); 524 } 525 if (ssize < len) { 526 if (doasyncfree) 527 bdwrite(ebp); 528 else 529 bwrite(ebp); 530 } 531 /* 532 * Last, free the old blocks and assign the new blocks to the buffers. 533 */ 534 #ifdef DEBUG 535 if (prtrealloc) 536 kprintf("\n\tnew:"); 537 #endif 538 for (blkno = newblk, i = 0; i < len; i++, blkno += fs->fs_frag) { 539 if (!DOINGSOFTDEP(vp)) 540 ffs_blkfree(ip, 541 dofftofsb(fs, buflist->bs_children[i]->b_bio2.bio_offset), 542 fs->fs_bsize); 543 buflist->bs_children[i]->b_bio2.bio_offset = fsbtodoff(fs, blkno); 544 #ifdef DIAGNOSTIC 545 if (!ffs_checkblk(ip, 546 dofftofsb(fs, buflist->bs_children[i]->b_bio2.bio_offset), fs->fs_bsize)) 547 panic("ffs_reallocblks: unallocated block 3"); 548 #endif 549 #ifdef DEBUG 550 if (prtrealloc) 551 kprintf(" %d,", blkno); 552 #endif 553 } 554 #ifdef DEBUG 555 if (prtrealloc) { 556 prtrealloc--; 557 kprintf("\n"); 558 } 559 #endif 560 return (0); 561 562 fail: 563 if (ssize < len) 564 brelse(ebp); 565 if (sbap != &ip->i_db[0]) 566 brelse(sbp); 567 return (ENOSPC); 568 } 569 570 /* 571 * Allocate an inode in the filesystem. 572 * 573 * If allocating a directory, use ffs_dirpref to select the inode. 574 * If allocating in a directory, the following hierarchy is followed: 575 * 1) allocate the preferred inode. 576 * 2) allocate an inode in the same cylinder group. 577 * 3) quadradically rehash into other cylinder groups, until an 578 * available inode is located. 579 * If no inode preference is given the following heirarchy is used 580 * to allocate an inode: 581 * 1) allocate an inode in cylinder group 0. 582 * 2) quadradically rehash into other cylinder groups, until an 583 * available inode is located. 584 */ 585 int 586 ffs_valloc(struct vnode *pvp, int mode, struct ucred *cred, struct vnode **vpp) 587 { 588 struct inode *pip; 589 struct fs *fs; 590 struct inode *ip; 591 ino_t ino, ipref; 592 int cg, error; 593 594 *vpp = NULL; 595 pip = VTOI(pvp); 596 fs = pip->i_fs; 597 if (fs->fs_cstotal.cs_nifree == 0) 598 goto noinodes; 599 600 if ((mode & IFMT) == IFDIR) 601 ipref = ffs_dirpref(pip); 602 else 603 ipref = pip->i_number; 604 if (ipref >= fs->fs_ncg * fs->fs_ipg) 605 ipref = 0; 606 cg = ino_to_cg(fs, ipref); 607 /* 608 * Track number of dirs created one after another 609 * in a same cg without intervening by files. 610 */ 611 if ((mode & IFMT) == IFDIR) { 612 if (fs->fs_contigdirs[cg] < 255) 613 fs->fs_contigdirs[cg]++; 614 } else { 615 if (fs->fs_contigdirs[cg] > 0) 616 fs->fs_contigdirs[cg]--; 617 } 618 ino = (ino_t)ffs_hashalloc(pip, cg, (long)ipref, mode, 619 (allocfcn_t *)ffs_nodealloccg); 620 if (ino == 0) 621 goto noinodes; 622 error = VFS_VGET(pvp->v_mount, NULL, ino, vpp); 623 if (error) { 624 ffs_vfree(pvp, ino, mode); 625 return (error); 626 } 627 ip = VTOI(*vpp); 628 if (ip->i_mode) { 629 kprintf("mode = 0%o, inum = %lu, fs = %s\n", 630 ip->i_mode, (u_long)ip->i_number, fs->fs_fsmnt); 631 panic("ffs_valloc: dup alloc"); 632 } 633 if (ip->i_blocks) { /* XXX */ 634 kprintf("free inode %s/%lu had %ld blocks\n", 635 fs->fs_fsmnt, (u_long)ino, (long)ip->i_blocks); 636 ip->i_blocks = 0; 637 } 638 ip->i_flags = 0; 639 /* 640 * Set up a new generation number for this inode. 641 */ 642 if (ip->i_gen == 0 || ++ip->i_gen == 0) 643 ip->i_gen = krandom() / 2 + 1; 644 return (0); 645 noinodes: 646 ffs_fserr(fs, cred->cr_uid, "out of inodes"); 647 uprintf("\n%s: create/symlink failed, no inodes free\n", fs->fs_fsmnt); 648 return (ENOSPC); 649 } 650 651 /* 652 * Find a cylinder group to place a directory. 653 * 654 * The policy implemented by this algorithm is to allocate a 655 * directory inode in the same cylinder group as its parent 656 * directory, but also to reserve space for its files inodes 657 * and data. Restrict the number of directories which may be 658 * allocated one after another in the same cylinder group 659 * without intervening allocation of files. 660 * 661 * If we allocate a first level directory then force allocation 662 * in another cylinder group. 663 */ 664 static ino_t 665 ffs_dirpref(struct inode *pip) 666 { 667 struct fs *fs; 668 int cg, prefcg, dirsize, cgsize; 669 int64_t dirsize64; 670 int avgifree, avgbfree, avgndir, curdirsize; 671 int minifree, minbfree, maxndir; 672 int mincg, minndir; 673 int maxcontigdirs; 674 675 fs = pip->i_fs; 676 677 avgifree = fs->fs_cstotal.cs_nifree / fs->fs_ncg; 678 avgbfree = fs->fs_cstotal.cs_nbfree / fs->fs_ncg; 679 avgndir = fs->fs_cstotal.cs_ndir / fs->fs_ncg; 680 681 /* 682 * Force allocation in another cg if creating a first level dir. 683 */ 684 if (ITOV(pip)->v_flag & VROOT) { 685 prefcg = karc4random() % fs->fs_ncg; 686 mincg = prefcg; 687 minndir = fs->fs_ipg; 688 for (cg = prefcg; cg < fs->fs_ncg; cg++) 689 if (fs->fs_cs(fs, cg).cs_ndir < minndir && 690 fs->fs_cs(fs, cg).cs_nifree >= avgifree && 691 fs->fs_cs(fs, cg).cs_nbfree >= avgbfree) { 692 mincg = cg; 693 minndir = fs->fs_cs(fs, cg).cs_ndir; 694 } 695 for (cg = 0; cg < prefcg; cg++) 696 if (fs->fs_cs(fs, cg).cs_ndir < minndir && 697 fs->fs_cs(fs, cg).cs_nifree >= avgifree && 698 fs->fs_cs(fs, cg).cs_nbfree >= avgbfree) { 699 mincg = cg; 700 minndir = fs->fs_cs(fs, cg).cs_ndir; 701 } 702 return ((ino_t)(fs->fs_ipg * mincg)); 703 } 704 705 /* 706 * Count various limits which used for 707 * optimal allocation of a directory inode. 708 */ 709 maxndir = min(avgndir + fs->fs_ipg / 16, fs->fs_ipg); 710 minifree = avgifree - avgifree / 4; 711 if (minifree < 1) 712 minifree = 1; 713 minbfree = avgbfree - avgbfree / 4; 714 if (minbfree < 1) 715 minbfree = 1; 716 cgsize = fs->fs_fsize * fs->fs_fpg; 717 718 /* 719 * fs_avgfilesize and fs_avgfpdir are user-settable entities and 720 * multiplying them may overflow a 32 bit integer. 721 */ 722 dirsize64 = fs->fs_avgfilesize * (int64_t)fs->fs_avgfpdir; 723 if (dirsize64 > 0x7fffffff) { 724 maxcontigdirs = 1; 725 } else { 726 dirsize = (int)dirsize64; 727 curdirsize = avgndir ? 728 (cgsize - avgbfree * fs->fs_bsize) / avgndir : 0; 729 if (dirsize < curdirsize) 730 dirsize = curdirsize; 731 maxcontigdirs = min((avgbfree * fs->fs_bsize) / dirsize, 255); 732 if (fs->fs_avgfpdir > 0) 733 maxcontigdirs = min(maxcontigdirs, 734 fs->fs_ipg / fs->fs_avgfpdir); 735 if (maxcontigdirs == 0) 736 maxcontigdirs = 1; 737 } 738 739 /* 740 * Limit number of dirs in one cg and reserve space for 741 * regular files, but only if we have no deficit in 742 * inodes or space. 743 */ 744 prefcg = ino_to_cg(fs, pip->i_number); 745 for (cg = prefcg; cg < fs->fs_ncg; cg++) 746 if (fs->fs_cs(fs, cg).cs_ndir < maxndir && 747 fs->fs_cs(fs, cg).cs_nifree >= minifree && 748 fs->fs_cs(fs, cg).cs_nbfree >= minbfree) { 749 if (fs->fs_contigdirs[cg] < maxcontigdirs) 750 return ((ino_t)(fs->fs_ipg * cg)); 751 } 752 for (cg = 0; cg < prefcg; cg++) 753 if (fs->fs_cs(fs, cg).cs_ndir < maxndir && 754 fs->fs_cs(fs, cg).cs_nifree >= minifree && 755 fs->fs_cs(fs, cg).cs_nbfree >= minbfree) { 756 if (fs->fs_contigdirs[cg] < maxcontigdirs) 757 return ((ino_t)(fs->fs_ipg * cg)); 758 } 759 /* 760 * This is a backstop when we have deficit in space. 761 */ 762 for (cg = prefcg; cg < fs->fs_ncg; cg++) 763 if (fs->fs_cs(fs, cg).cs_nifree >= avgifree) 764 return ((ino_t)(fs->fs_ipg * cg)); 765 for (cg = 0; cg < prefcg; cg++) 766 if (fs->fs_cs(fs, cg).cs_nifree >= avgifree) 767 break; 768 return ((ino_t)(fs->fs_ipg * cg)); 769 } 770 771 /* 772 * Select the desired position for the next block in a file. The file is 773 * logically divided into sections. The first section is composed of the 774 * direct blocks. Each additional section contains fs_maxbpg blocks. 775 * 776 * If no blocks have been allocated in the first section, the policy is to 777 * request a block in the same cylinder group as the inode that describes 778 * the file. If no blocks have been allocated in any other section, the 779 * policy is to place the section in a cylinder group with a greater than 780 * average number of free blocks. An appropriate cylinder group is found 781 * by using a rotor that sweeps the cylinder groups. When a new group of 782 * blocks is needed, the sweep begins in the cylinder group following the 783 * cylinder group from which the previous allocation was made. The sweep 784 * continues until a cylinder group with greater than the average number 785 * of free blocks is found. If the allocation is for the first block in an 786 * indirect block, the information on the previous allocation is unavailable; 787 * here a best guess is made based upon the logical block number being 788 * allocated. 789 * 790 * If a section is already partially allocated, the policy is to 791 * contiguously allocate fs_maxcontig blocks. The end of one of these 792 * contiguous blocks and the beginning of the next is physically separated 793 * so that the disk head will be in transit between them for at least 794 * fs_rotdelay milliseconds. This is to allow time for the processor to 795 * schedule another I/O transfer. 796 */ 797 ufs_daddr_t 798 ffs_blkpref(struct inode *ip, ufs_daddr_t lbn, int indx, ufs_daddr_t *bap) 799 { 800 struct fs *fs; 801 int cg; 802 int avgbfree, startcg; 803 ufs_daddr_t nextblk; 804 805 fs = ip->i_fs; 806 if (indx % fs->fs_maxbpg == 0 || bap[indx - 1] == 0) { 807 if (lbn < NDADDR + NINDIR(fs)) { 808 cg = ino_to_cg(fs, ip->i_number); 809 return (fs->fs_fpg * cg + fs->fs_frag); 810 } 811 /* 812 * Find a cylinder with greater than average number of 813 * unused data blocks. 814 */ 815 if (indx == 0 || bap[indx - 1] == 0) 816 startcg = 817 ino_to_cg(fs, ip->i_number) + lbn / fs->fs_maxbpg; 818 else 819 startcg = dtog(fs, bap[indx - 1]) + 1; 820 startcg %= fs->fs_ncg; 821 avgbfree = fs->fs_cstotal.cs_nbfree / fs->fs_ncg; 822 for (cg = startcg; cg < fs->fs_ncg; cg++) 823 if (fs->fs_cs(fs, cg).cs_nbfree >= avgbfree) { 824 fs->fs_cgrotor = cg; 825 return (fs->fs_fpg * cg + fs->fs_frag); 826 } 827 for (cg = 0; cg <= startcg; cg++) 828 if (fs->fs_cs(fs, cg).cs_nbfree >= avgbfree) { 829 fs->fs_cgrotor = cg; 830 return (fs->fs_fpg * cg + fs->fs_frag); 831 } 832 return (0); 833 } 834 /* 835 * One or more previous blocks have been laid out. If less 836 * than fs_maxcontig previous blocks are contiguous, the 837 * next block is requested contiguously, otherwise it is 838 * requested rotationally delayed by fs_rotdelay milliseconds. 839 */ 840 nextblk = bap[indx - 1] + fs->fs_frag; 841 if (fs->fs_rotdelay == 0 || indx < fs->fs_maxcontig || 842 bap[indx - fs->fs_maxcontig] + 843 blkstofrags(fs, fs->fs_maxcontig) != nextblk) 844 return (nextblk); 845 /* 846 * Here we convert ms of delay to frags as: 847 * (frags) = (ms) * (rev/sec) * (sect/rev) / 848 * ((sect/frag) * (ms/sec)) 849 * then round up to the next block. 850 */ 851 nextblk += roundup(fs->fs_rotdelay * fs->fs_rps * fs->fs_nsect / 852 (NSPF(fs) * 1000), fs->fs_frag); 853 return (nextblk); 854 } 855 856 /* 857 * Implement the cylinder overflow algorithm. 858 * 859 * The policy implemented by this algorithm is: 860 * 1) allocate the block in its requested cylinder group. 861 * 2) quadradically rehash on the cylinder group number. 862 * 3) brute force search for a free block. 863 */ 864 /*VARARGS5*/ 865 static u_long 866 ffs_hashalloc(struct inode *ip, int cg, long pref, 867 int size, /* size for data blocks, mode for inodes */ 868 allocfcn_t *allocator) 869 { 870 struct fs *fs; 871 long result; /* XXX why not same type as we return? */ 872 int i, icg = cg; 873 874 fs = ip->i_fs; 875 /* 876 * 1: preferred cylinder group 877 */ 878 result = (*allocator)(ip, cg, pref, size); 879 if (result) 880 return (result); 881 /* 882 * 2: quadratic rehash 883 */ 884 for (i = 1; i < fs->fs_ncg; i *= 2) { 885 cg += i; 886 if (cg >= fs->fs_ncg) 887 cg -= fs->fs_ncg; 888 result = (*allocator)(ip, cg, 0, size); 889 if (result) 890 return (result); 891 } 892 /* 893 * 3: brute force search 894 * Note that we start at i == 2, since 0 was checked initially, 895 * and 1 is always checked in the quadratic rehash. 896 */ 897 cg = (icg + 2) % fs->fs_ncg; 898 for (i = 2; i < fs->fs_ncg; i++) { 899 result = (*allocator)(ip, cg, 0, size); 900 if (result) 901 return (result); 902 cg++; 903 if (cg == fs->fs_ncg) 904 cg = 0; 905 } 906 return (0); 907 } 908 909 /* 910 * Determine whether a fragment can be extended. 911 * 912 * Check to see if the necessary fragments are available, and 913 * if they are, allocate them. 914 */ 915 static ufs_daddr_t 916 ffs_fragextend(struct inode *ip, int cg, long bprev, int osize, int nsize) 917 { 918 struct fs *fs; 919 struct cg *cgp; 920 struct buf *bp; 921 long bno; 922 int frags, bbase; 923 int i, error; 924 uint8_t *blksfree; 925 926 fs = ip->i_fs; 927 if (fs->fs_cs(fs, cg).cs_nffree < numfrags(fs, nsize - osize)) 928 return (0); 929 frags = numfrags(fs, nsize); 930 bbase = fragnum(fs, bprev); 931 if (bbase > fragnum(fs, (bprev + frags - 1))) { 932 /* cannot extend across a block boundary */ 933 return (0); 934 } 935 KKASSERT(blknum(fs, bprev) == blknum(fs, bprev + frags - 1)); 936 error = bread(ip->i_devvp, fsbtodoff(fs, cgtod(fs, cg)), 937 (int)fs->fs_cgsize, &bp); 938 if (error) { 939 brelse(bp); 940 return (0); 941 } 942 cgp = (struct cg *)bp->b_data; 943 if (!cg_chkmagic(cgp)) { 944 brelse(bp); 945 return (0); 946 } 947 cgp->cg_time = time_second; 948 bno = dtogd(fs, bprev); 949 blksfree = cg_blksfree(cgp); 950 for (i = numfrags(fs, osize); i < frags; i++) { 951 if (isclr(blksfree, bno + i)) { 952 brelse(bp); 953 return (0); 954 } 955 } 956 957 /* 958 * the current fragment can be extended 959 * deduct the count on fragment being extended into 960 * increase the count on the remaining fragment (if any) 961 * allocate the extended piece 962 * 963 * ---oooooooooonnnnnnn111---- 964 * [-----frags-----] 965 * ^ ^ 966 * bbase fs_frag 967 */ 968 for (i = frags; i < fs->fs_frag - bbase; i++) { 969 if (isclr(blksfree, bno + i)) 970 break; 971 } 972 973 /* 974 * Size of original free frag is [i - numfrags(fs, osize)] 975 * Size of remaining free frag is [i - frags] 976 */ 977 cgp->cg_frsum[i - numfrags(fs, osize)]--; 978 if (i != frags) 979 cgp->cg_frsum[i - frags]++; 980 for (i = numfrags(fs, osize); i < frags; i++) { 981 clrbit(blksfree, bno + i); 982 cgp->cg_cs.cs_nffree--; 983 fs->fs_cstotal.cs_nffree--; 984 fs->fs_cs(fs, cg).cs_nffree--; 985 } 986 fs->fs_fmod = 1; 987 if (DOINGSOFTDEP(ITOV(ip))) 988 softdep_setup_blkmapdep(bp, fs, bprev); 989 bdwrite(bp); 990 return (bprev); 991 } 992 993 /* 994 * Determine whether a block can be allocated. 995 * 996 * Check to see if a block of the appropriate size is available, 997 * and if it is, allocate it. 998 */ 999 static ufs_daddr_t 1000 ffs_alloccg(struct inode *ip, int cg, ufs_daddr_t bpref, int size) 1001 { 1002 struct fs *fs; 1003 struct cg *cgp; 1004 struct buf *bp; 1005 int i; 1006 ufs_daddr_t bno, blkno; 1007 int allocsiz, error, frags; 1008 uint8_t *blksfree; 1009 1010 fs = ip->i_fs; 1011 if (fs->fs_cs(fs, cg).cs_nbfree == 0 && size == fs->fs_bsize) 1012 return (0); 1013 error = bread(ip->i_devvp, fsbtodoff(fs, cgtod(fs, cg)), 1014 (int)fs->fs_cgsize, &bp); 1015 if (error) { 1016 brelse(bp); 1017 return (0); 1018 } 1019 cgp = (struct cg *)bp->b_data; 1020 if (!cg_chkmagic(cgp) || 1021 (cgp->cg_cs.cs_nbfree == 0 && size == fs->fs_bsize)) { 1022 brelse(bp); 1023 return (0); 1024 } 1025 cgp->cg_time = time_second; 1026 if (size == fs->fs_bsize) { 1027 bno = ffs_alloccgblk(ip, bp, bpref); 1028 bdwrite(bp); 1029 return (bno); 1030 } 1031 /* 1032 * Check to see if any fragments of sufficient size are already 1033 * available. Fit the data into a larger fragment if necessary, 1034 * before allocating a whole new block. 1035 */ 1036 blksfree = cg_blksfree(cgp); 1037 frags = numfrags(fs, size); 1038 for (allocsiz = frags; allocsiz < fs->fs_frag; allocsiz++) { 1039 if (cgp->cg_frsum[allocsiz] != 0) 1040 break; 1041 } 1042 if (allocsiz == fs->fs_frag) { 1043 /* 1044 * No fragments were available, allocate a whole block and 1045 * cut the requested fragment (of size frags) out of it. 1046 */ 1047 if (cgp->cg_cs.cs_nbfree == 0) { 1048 brelse(bp); 1049 return (0); 1050 } 1051 bno = ffs_alloccgblk(ip, bp, bpref); 1052 bpref = dtogd(fs, bno); 1053 for (i = frags; i < fs->fs_frag; i++) 1054 setbit(blksfree, bpref + i); 1055 1056 /* 1057 * Calculate the number of free frags still remaining after 1058 * we have cut out the requested allocation. Indicate that 1059 * a fragment of that size is now available for future 1060 * allocation. 1061 */ 1062 i = fs->fs_frag - frags; 1063 cgp->cg_cs.cs_nffree += i; 1064 fs->fs_cstotal.cs_nffree += i; 1065 fs->fs_cs(fs, cg).cs_nffree += i; 1066 fs->fs_fmod = 1; 1067 cgp->cg_frsum[i]++; 1068 bdwrite(bp); 1069 return (bno); 1070 } 1071 1072 /* 1073 * cg_frsum[] has told us that a free fragment of allocsiz size is 1074 * available. Find it, then clear the bitmap bits associated with 1075 * the size we want. 1076 */ 1077 bno = ffs_mapsearch(fs, cgp, bpref, allocsiz); 1078 if (bno < 0) { 1079 brelse(bp); 1080 return (0); 1081 } 1082 for (i = 0; i < frags; i++) 1083 clrbit(blksfree, bno + i); 1084 cgp->cg_cs.cs_nffree -= frags; 1085 fs->fs_cstotal.cs_nffree -= frags; 1086 fs->fs_cs(fs, cg).cs_nffree -= frags; 1087 fs->fs_fmod = 1; 1088 1089 /* 1090 * Account for the allocation. The original searched size that we 1091 * found is no longer available. If we cut out a smaller piece then 1092 * a smaller fragment is now available. 1093 */ 1094 cgp->cg_frsum[allocsiz]--; 1095 if (frags != allocsiz) 1096 cgp->cg_frsum[allocsiz - frags]++; 1097 blkno = cg * fs->fs_fpg + bno; 1098 if (DOINGSOFTDEP(ITOV(ip))) 1099 softdep_setup_blkmapdep(bp, fs, blkno); 1100 bdwrite(bp); 1101 return ((u_long)blkno); 1102 } 1103 1104 /* 1105 * Allocate a block in a cylinder group. 1106 * 1107 * This algorithm implements the following policy: 1108 * 1) allocate the requested block. 1109 * 2) allocate a rotationally optimal block in the same cylinder. 1110 * 3) allocate the next available block on the block rotor for the 1111 * specified cylinder group. 1112 * Note that this routine only allocates fs_bsize blocks; these 1113 * blocks may be fragmented by the routine that allocates them. 1114 */ 1115 static ufs_daddr_t 1116 ffs_alloccgblk(struct inode *ip, struct buf *bp, ufs_daddr_t bpref) 1117 { 1118 struct fs *fs; 1119 struct cg *cgp; 1120 ufs_daddr_t bno, blkno; 1121 int cylno, pos, delta; 1122 short *cylbp; 1123 int i; 1124 uint8_t *blksfree; 1125 1126 fs = ip->i_fs; 1127 cgp = (struct cg *)bp->b_data; 1128 blksfree = cg_blksfree(cgp); 1129 if (bpref == 0 || dtog(fs, bpref) != cgp->cg_cgx) { 1130 bpref = cgp->cg_rotor; 1131 goto norot; 1132 } 1133 bpref = blknum(fs, bpref); 1134 bpref = dtogd(fs, bpref); 1135 /* 1136 * if the requested block is available, use it 1137 */ 1138 if (ffs_isblock(fs, blksfree, fragstoblks(fs, bpref))) { 1139 bno = bpref; 1140 goto gotit; 1141 } 1142 if (fs->fs_nrpos <= 1 || fs->fs_cpc == 0) { 1143 /* 1144 * Block layout information is not available. 1145 * Leaving bpref unchanged means we take the 1146 * next available free block following the one 1147 * we just allocated. Hopefully this will at 1148 * least hit a track cache on drives of unknown 1149 * geometry (e.g. SCSI). 1150 */ 1151 goto norot; 1152 } 1153 /* 1154 * check for a block available on the same cylinder 1155 */ 1156 cylno = cbtocylno(fs, bpref); 1157 if (cg_blktot(cgp)[cylno] == 0) 1158 goto norot; 1159 /* 1160 * check the summary information to see if a block is 1161 * available in the requested cylinder starting at the 1162 * requested rotational position and proceeding around. 1163 */ 1164 cylbp = cg_blks(fs, cgp, cylno); 1165 pos = cbtorpos(fs, bpref); 1166 for (i = pos; i < fs->fs_nrpos; i++) 1167 if (cylbp[i] > 0) 1168 break; 1169 if (i == fs->fs_nrpos) 1170 for (i = 0; i < pos; i++) 1171 if (cylbp[i] > 0) 1172 break; 1173 if (cylbp[i] > 0) { 1174 /* 1175 * found a rotational position, now find the actual 1176 * block. A panic if none is actually there. 1177 */ 1178 pos = cylno % fs->fs_cpc; 1179 bno = (cylno - pos) * fs->fs_spc / NSPB(fs); 1180 if (fs_postbl(fs, pos)[i] == -1) { 1181 kprintf("pos = %d, i = %d, fs = %s\n", 1182 pos, i, fs->fs_fsmnt); 1183 panic("ffs_alloccgblk: cyl groups corrupted"); 1184 } 1185 for (i = fs_postbl(fs, pos)[i];; ) { 1186 if (ffs_isblock(fs, blksfree, bno + i)) { 1187 bno = blkstofrags(fs, (bno + i)); 1188 goto gotit; 1189 } 1190 delta = fs_rotbl(fs)[i]; 1191 if (delta <= 0 || 1192 delta + i > fragstoblks(fs, fs->fs_fpg)) 1193 break; 1194 i += delta; 1195 } 1196 kprintf("pos = %d, i = %d, fs = %s\n", pos, i, fs->fs_fsmnt); 1197 panic("ffs_alloccgblk: can't find blk in cyl"); 1198 } 1199 norot: 1200 /* 1201 * no blocks in the requested cylinder, so take next 1202 * available one in this cylinder group. 1203 */ 1204 bno = ffs_mapsearch(fs, cgp, bpref, (int)fs->fs_frag); 1205 if (bno < 0) 1206 return (0); 1207 cgp->cg_rotor = bno; 1208 gotit: 1209 blkno = fragstoblks(fs, bno); 1210 ffs_clrblock(fs, blksfree, (long)blkno); 1211 ffs_clusteracct(fs, cgp, blkno, -1); 1212 cgp->cg_cs.cs_nbfree--; 1213 fs->fs_cstotal.cs_nbfree--; 1214 fs->fs_cs(fs, cgp->cg_cgx).cs_nbfree--; 1215 cylno = cbtocylno(fs, bno); 1216 cg_blks(fs, cgp, cylno)[cbtorpos(fs, bno)]--; 1217 cg_blktot(cgp)[cylno]--; 1218 fs->fs_fmod = 1; 1219 blkno = cgp->cg_cgx * fs->fs_fpg + bno; 1220 if (DOINGSOFTDEP(ITOV(ip))) 1221 softdep_setup_blkmapdep(bp, fs, blkno); 1222 return (blkno); 1223 } 1224 1225 /* 1226 * Determine whether a cluster can be allocated. 1227 * 1228 * We do not currently check for optimal rotational layout if there 1229 * are multiple choices in the same cylinder group. Instead we just 1230 * take the first one that we find following bpref. 1231 */ 1232 static ufs_daddr_t 1233 ffs_clusteralloc(struct inode *ip, int cg, ufs_daddr_t bpref, int len) 1234 { 1235 struct fs *fs; 1236 struct cg *cgp; 1237 struct buf *bp; 1238 int i, got, run, bno, bit, map; 1239 u_char *mapp; 1240 int32_t *lp; 1241 uint8_t *blksfree; 1242 1243 fs = ip->i_fs; 1244 if (fs->fs_maxcluster[cg] < len) 1245 return (0); 1246 if (bread(ip->i_devvp, fsbtodoff(fs, cgtod(fs, cg)), 1247 (int)fs->fs_cgsize, &bp)) { 1248 goto fail; 1249 } 1250 cgp = (struct cg *)bp->b_data; 1251 if (!cg_chkmagic(cgp)) 1252 goto fail; 1253 1254 /* 1255 * Check to see if a cluster of the needed size (or bigger) is 1256 * available in this cylinder group. 1257 */ 1258 lp = &cg_clustersum(cgp)[len]; 1259 for (i = len; i <= fs->fs_contigsumsize; i++) 1260 if (*lp++ > 0) 1261 break; 1262 if (i > fs->fs_contigsumsize) { 1263 /* 1264 * This is the first time looking for a cluster in this 1265 * cylinder group. Update the cluster summary information 1266 * to reflect the true maximum sized cluster so that 1267 * future cluster allocation requests can avoid reading 1268 * the cylinder group map only to find no clusters. 1269 */ 1270 lp = &cg_clustersum(cgp)[len - 1]; 1271 for (i = len - 1; i > 0; i--) 1272 if (*lp-- > 0) 1273 break; 1274 fs->fs_maxcluster[cg] = i; 1275 goto fail; 1276 } 1277 /* 1278 * Search the cluster map to find a big enough cluster. 1279 * We take the first one that we find, even if it is larger 1280 * than we need as we prefer to get one close to the previous 1281 * block allocation. We do not search before the current 1282 * preference point as we do not want to allocate a block 1283 * that is allocated before the previous one (as we will 1284 * then have to wait for another pass of the elevator 1285 * algorithm before it will be read). We prefer to fail and 1286 * be recalled to try an allocation in the next cylinder group. 1287 */ 1288 if (dtog(fs, bpref) != cg) 1289 bpref = 0; 1290 else 1291 bpref = fragstoblks(fs, dtogd(fs, blknum(fs, bpref))); 1292 mapp = &cg_clustersfree(cgp)[bpref / NBBY]; 1293 map = *mapp++; 1294 bit = 1 << (bpref % NBBY); 1295 for (run = 0, got = bpref; got < cgp->cg_nclusterblks; got++) { 1296 if ((map & bit) == 0) { 1297 run = 0; 1298 } else { 1299 run++; 1300 if (run == len) 1301 break; 1302 } 1303 if ((got & (NBBY - 1)) != (NBBY - 1)) { 1304 bit <<= 1; 1305 } else { 1306 map = *mapp++; 1307 bit = 1; 1308 } 1309 } 1310 if (got >= cgp->cg_nclusterblks) 1311 goto fail; 1312 /* 1313 * Allocate the cluster that we have found. 1314 */ 1315 blksfree = cg_blksfree(cgp); 1316 for (i = 1; i <= len; i++) { 1317 if (!ffs_isblock(fs, blksfree, got - run + i)) 1318 panic("ffs_clusteralloc: map mismatch"); 1319 } 1320 bno = cg * fs->fs_fpg + blkstofrags(fs, got - run + 1); 1321 if (dtog(fs, bno) != cg) 1322 panic("ffs_clusteralloc: allocated out of group"); 1323 len = blkstofrags(fs, len); 1324 for (i = 0; i < len; i += fs->fs_frag) { 1325 if ((got = ffs_alloccgblk(ip, bp, bno + i)) != bno + i) 1326 panic("ffs_clusteralloc: lost block"); 1327 } 1328 bdwrite(bp); 1329 return (bno); 1330 1331 fail: 1332 brelse(bp); 1333 return (0); 1334 } 1335 1336 /* 1337 * Determine whether an inode can be allocated. 1338 * 1339 * Check to see if an inode is available, and if it is, 1340 * allocate it using the following policy: 1341 * 1) allocate the requested inode. 1342 * 2) allocate the next available inode after the requested 1343 * inode in the specified cylinder group. 1344 * 3) the inode must not already be in the inode hash table. We 1345 * can encounter such a case because the vnode reclamation sequence 1346 * frees the bit 1347 * 3) the inode must not already be in the inode hash, otherwise it 1348 * may be in the process of being deallocated. This can occur 1349 * because the bitmap is updated before the inode is removed from 1350 * hash. If we were to reallocate the inode the caller could wind 1351 * up returning a vnode/inode combination which is in an indeterminate 1352 * state. 1353 */ 1354 static ino_t 1355 ffs_nodealloccg(struct inode *ip, int cg, ufs_daddr_t ipref, int mode) 1356 { 1357 struct ufsmount *ump; 1358 struct fs *fs; 1359 struct cg *cgp; 1360 struct buf *bp; 1361 uint8_t *inosused; 1362 uint8_t map; 1363 int error, len, arraysize, i; 1364 int icheckmiss; 1365 ufs_daddr_t ibase; 1366 struct vnode *vp; 1367 1368 vp = ITOV(ip); 1369 ump = VFSTOUFS(vp->v_mount); 1370 fs = ip->i_fs; 1371 if (fs->fs_cs(fs, cg).cs_nifree == 0) 1372 return (0); 1373 error = bread(ip->i_devvp, fsbtodoff(fs, cgtod(fs, cg)), 1374 (int)fs->fs_cgsize, &bp); 1375 if (error) { 1376 brelse(bp); 1377 return (0); 1378 } 1379 cgp = (struct cg *)bp->b_data; 1380 if (!cg_chkmagic(cgp) || cgp->cg_cs.cs_nifree == 0) { 1381 brelse(bp); 1382 return (0); 1383 } 1384 inosused = cg_inosused(cgp); 1385 icheckmiss = 0; 1386 1387 /* 1388 * Quick check, reuse the most recently free inode or continue 1389 * a scan from where we left off the last time. 1390 */ 1391 ibase = cg * fs->fs_ipg; 1392 if (ipref) { 1393 ipref %= fs->fs_ipg; 1394 if (isclr(inosused, ipref)) { 1395 if (ufs_ihashcheck(ump, ip->i_dev, ibase + ipref) == 0) 1396 goto gotit; 1397 } 1398 } 1399 1400 /* 1401 * Scan the inode bitmap starting at irotor, be sure to handle 1402 * the edge case by going back to the beginning of the array. 1403 * 1404 * If the number of inodes is not byte-aligned, the unused bits 1405 * should be set to 1. This will be sanity checked in gotit. Note 1406 * that we have to be sure not to overlap the beginning and end 1407 * when irotor is in the middle of a byte as this will cause the 1408 * same bitmap byte to be checked twice. To solve this problem we 1409 * just convert everything to a byte index for the loop. 1410 */ 1411 ipref = (cgp->cg_irotor % fs->fs_ipg) >> 3; /* byte index */ 1412 len = (fs->fs_ipg + 7) >> 3; /* byte size */ 1413 arraysize = len; 1414 1415 while (len > 0) { 1416 map = inosused[ipref]; 1417 if (map != 255) { 1418 for (i = 0; i < NBBY; ++i) { 1419 /* 1420 * If we find a free bit we have to make sure 1421 * that the inode is not in the middle of 1422 * being destroyed. The inode should not exist 1423 * in the inode hash. 1424 * 1425 * Adjust the rotor to try to hit the 1426 * quick-check up above. 1427 */ 1428 if ((map & (1 << i)) == 0) { 1429 if (ufs_ihashcheck(ump, ip->i_dev, ibase + (ipref << 3) + i) == 0) { 1430 ipref = (ipref << 3) + i; 1431 cgp->cg_irotor = (ipref + 1) % fs->fs_ipg; 1432 goto gotit; 1433 } 1434 ++icheckmiss; 1435 } 1436 } 1437 } 1438 1439 /* 1440 * Setup for the next byte, start at the beginning again if 1441 * we hit the end of the array. 1442 */ 1443 if (++ipref == arraysize) 1444 ipref = 0; 1445 --len; 1446 } 1447 if (icheckmiss == cgp->cg_cs.cs_nifree) { 1448 brelse(bp); 1449 return(0); 1450 } 1451 kprintf("fs = %s\n", fs->fs_fsmnt); 1452 panic("ffs_nodealloccg: block not in map, icheckmiss/nfree %d/%d", 1453 icheckmiss, cgp->cg_cs.cs_nifree); 1454 /* NOTREACHED */ 1455 1456 /* 1457 * ipref is a bit index as of the gotit label. 1458 */ 1459 gotit: 1460 KKASSERT(ipref >= 0 && ipref < fs->fs_ipg); 1461 cgp->cg_time = time_second; 1462 if (DOINGSOFTDEP(ITOV(ip))) 1463 softdep_setup_inomapdep(bp, ip, ibase + ipref); 1464 setbit(inosused, ipref); 1465 cgp->cg_cs.cs_nifree--; 1466 fs->fs_cstotal.cs_nifree--; 1467 fs->fs_cs(fs, cg).cs_nifree--; 1468 fs->fs_fmod = 1; 1469 if ((mode & IFMT) == IFDIR) { 1470 cgp->cg_cs.cs_ndir++; 1471 fs->fs_cstotal.cs_ndir++; 1472 fs->fs_cs(fs, cg).cs_ndir++; 1473 } 1474 bdwrite(bp); 1475 return (ibase + ipref); 1476 } 1477 1478 /* 1479 * Free a block or fragment. 1480 * 1481 * The specified block or fragment is placed back in the 1482 * free map. If a fragment is deallocated, a possible 1483 * block reassembly is checked. 1484 */ 1485 void 1486 ffs_blkfree_cg(struct fs * fs, struct vnode * i_devvp, cdev_t i_dev, ino_t i_number, 1487 uint32_t i_din_uid, ufs_daddr_t bno, long size) 1488 { 1489 struct cg *cgp; 1490 struct buf *bp; 1491 ufs_daddr_t blkno; 1492 int i, error, cg, blk, frags, bbase; 1493 uint8_t *blksfree; 1494 1495 VOP_FREEBLKS(i_devvp, fsbtodoff(fs, bno), size); 1496 if ((uint)size > fs->fs_bsize || fragoff(fs, size) != 0 || 1497 fragnum(fs, bno) + numfrags(fs, size) > fs->fs_frag) { 1498 kprintf("dev=%s, bno = %ld, bsize = %ld, size = %ld, fs = %s\n", 1499 devtoname(i_dev), (long)bno, (long)fs->fs_bsize, size, 1500 fs->fs_fsmnt); 1501 panic("ffs_blkfree: bad size"); 1502 } 1503 cg = dtog(fs, bno); 1504 if ((uint)bno >= fs->fs_size) { 1505 kprintf("bad block %ld, ino %lu\n", 1506 (long)bno, (u_long)i_number); 1507 ffs_fserr(fs, i_din_uid, "bad block"); 1508 return; 1509 } 1510 1511 /* 1512 * Load the cylinder group 1513 */ 1514 error = bread(i_devvp, fsbtodoff(fs, cgtod(fs, cg)), 1515 (int)fs->fs_cgsize, &bp); 1516 if (error) { 1517 brelse(bp); 1518 return; 1519 } 1520 cgp = (struct cg *)bp->b_data; 1521 if (!cg_chkmagic(cgp)) { 1522 brelse(bp); 1523 return; 1524 } 1525 cgp->cg_time = time_second; 1526 bno = dtogd(fs, bno); 1527 blksfree = cg_blksfree(cgp); 1528 1529 if (size == fs->fs_bsize) { 1530 /* 1531 * Free a whole block 1532 */ 1533 blkno = fragstoblks(fs, bno); 1534 if (!ffs_isfreeblock(fs, blksfree, blkno)) { 1535 kprintf("dev = %s, block = %ld, fs = %s\n", 1536 devtoname(i_dev), (long)bno, fs->fs_fsmnt); 1537 panic("ffs_blkfree: freeing free block"); 1538 } 1539 ffs_setblock(fs, blksfree, blkno); 1540 ffs_clusteracct(fs, cgp, blkno, 1); 1541 cgp->cg_cs.cs_nbfree++; 1542 fs->fs_cstotal.cs_nbfree++; 1543 fs->fs_cs(fs, cg).cs_nbfree++; 1544 i = cbtocylno(fs, bno); 1545 cg_blks(fs, cgp, i)[cbtorpos(fs, bno)]++; 1546 cg_blktot(cgp)[i]++; 1547 } else { 1548 /* 1549 * Free a fragment within a block. 1550 * 1551 * bno is the starting block number of the fragment being 1552 * freed. 1553 * 1554 * bbase is the starting block number for the filesystem 1555 * block containing the fragment. 1556 * 1557 * blk is the current bitmap for the fragments within the 1558 * filesystem block containing the fragment. 1559 * 1560 * frags is the number of fragments being freed 1561 * 1562 * Call ffs_fragacct() to account for the removal of all 1563 * current fragments, then adjust the bitmap to free the 1564 * requested fragment, and finally call ffs_fragacct() again 1565 * to regenerate the accounting. 1566 */ 1567 bbase = bno - fragnum(fs, bno); 1568 blk = blkmap(fs, blksfree, bbase); 1569 ffs_fragacct(fs, blk, cgp->cg_frsum, -1); 1570 frags = numfrags(fs, size); 1571 for (i = 0; i < frags; i++) { 1572 if (isset(blksfree, bno + i)) { 1573 kprintf("dev = %s, block = %ld, fs = %s\n", 1574 devtoname(i_dev), (long)(bno + i), 1575 fs->fs_fsmnt); 1576 panic("ffs_blkfree: freeing free frag"); 1577 } 1578 setbit(blksfree, bno + i); 1579 } 1580 cgp->cg_cs.cs_nffree += i; 1581 fs->fs_cstotal.cs_nffree += i; 1582 fs->fs_cs(fs, cg).cs_nffree += i; 1583 1584 /* 1585 * Add back in counts associated with the new frags 1586 */ 1587 blk = blkmap(fs, blksfree, bbase); 1588 ffs_fragacct(fs, blk, cgp->cg_frsum, 1); 1589 1590 /* 1591 * If a complete block has been reassembled, account for it 1592 */ 1593 blkno = fragstoblks(fs, bbase); 1594 if (ffs_isblock(fs, blksfree, blkno)) { 1595 cgp->cg_cs.cs_nffree -= fs->fs_frag; 1596 fs->fs_cstotal.cs_nffree -= fs->fs_frag; 1597 fs->fs_cs(fs, cg).cs_nffree -= fs->fs_frag; 1598 ffs_clusteracct(fs, cgp, blkno, 1); 1599 cgp->cg_cs.cs_nbfree++; 1600 fs->fs_cstotal.cs_nbfree++; 1601 fs->fs_cs(fs, cg).cs_nbfree++; 1602 i = cbtocylno(fs, bbase); 1603 cg_blks(fs, cgp, i)[cbtorpos(fs, bbase)]++; 1604 cg_blktot(cgp)[i]++; 1605 } 1606 } 1607 fs->fs_fmod = 1; 1608 bdwrite(bp); 1609 } 1610 1611 struct ffs_blkfree_trim_params { 1612 struct task task; 1613 ufs_daddr_t bno; 1614 long size; 1615 1616 /* 1617 * With TRIM, inode pointer is gone in the callback but we still need 1618 * the following fields for ffs_blkfree_cg() 1619 */ 1620 struct vnode *i_devvp; 1621 struct fs *i_fs; 1622 cdev_t i_dev; 1623 ino_t i_number; 1624 uint32_t i_din_uid; 1625 }; 1626 1627 1628 static void 1629 ffs_blkfree_trim_task(void *ctx, int pending) 1630 { 1631 struct ffs_blkfree_trim_params *tp; 1632 1633 tp = ctx; 1634 ffs_blkfree_cg(tp->i_fs, tp->i_devvp, tp->i_dev, tp->i_number, 1635 tp->i_din_uid, tp->bno, tp->size); 1636 kfree(tp, M_TEMP); 1637 } 1638 1639 1640 1641 static void 1642 ffs_blkfree_trim_completed(struct bio *biop) 1643 { 1644 struct buf *bp = biop->bio_buf; 1645 struct ffs_blkfree_trim_params *tp; 1646 1647 tp = bp->b_bio1.bio_caller_info1.ptr; 1648 TASK_INIT(&tp->task, 0, ffs_blkfree_trim_task, tp); 1649 tp = biop->bio_caller_info1.ptr; 1650 taskqueue_enqueue(taskqueue_swi, &tp->task); 1651 biodone(biop); 1652 } 1653 1654 1655 /* 1656 * If TRIM is enabled, we TRIM the blocks first then free them. We do this 1657 * after TRIM is finished and the callback handler is called. The logic here 1658 * is that we free the blocks before updating the bitmap so that we don't 1659 * reuse a block before we actually trim it, which would result in trimming 1660 * a valid block. 1661 */ 1662 void 1663 ffs_blkfree(struct inode *ip, ufs_daddr_t bno, long size) 1664 { 1665 struct mount *mp = ip->i_devvp->v_mount; 1666 struct ffs_blkfree_trim_params *tp; 1667 1668 if (!(mp->mnt_flag & MNT_TRIM)) { 1669 ffs_blkfree_cg(ip->i_fs, ip->i_devvp,ip->i_dev,ip->i_number, 1670 ip->i_uid, bno, size); 1671 return; 1672 } 1673 1674 struct buf *bp; 1675 1676 tp = kmalloc(sizeof(struct ffs_blkfree_trim_params), M_TEMP, M_WAITOK); 1677 tp->bno = bno; 1678 tp->i_fs= ip->i_fs; 1679 tp->i_devvp = ip->i_devvp; 1680 tp->i_dev = ip->i_dev; 1681 tp->i_din_uid = ip->i_uid; 1682 tp->i_number = ip->i_number; 1683 tp->size = size; 1684 1685 bp = getnewbuf(0,0,0,1); 1686 BUF_KERNPROC(bp); 1687 bp->b_cmd = BUF_CMD_FREEBLKS; 1688 bp->b_bio1.bio_offset = fsbtodoff(ip->i_fs, bno); 1689 bp->b_bcount = size; 1690 bp->b_bio1.bio_caller_info1.ptr = tp; 1691 bp->b_bio1.bio_done = ffs_blkfree_trim_completed; 1692 vn_strategy(ip->i_devvp, &bp->b_bio1); 1693 } 1694 1695 #ifdef DIAGNOSTIC 1696 /* 1697 * Verify allocation of a block or fragment. Returns true if block or 1698 * fragment is allocated, false if it is free. 1699 */ 1700 static int 1701 ffs_checkblk(struct inode *ip, ufs_daddr_t bno, long size) 1702 { 1703 struct fs *fs; 1704 struct cg *cgp; 1705 struct buf *bp; 1706 int i, error, frags, free; 1707 uint8_t *blksfree; 1708 1709 fs = ip->i_fs; 1710 if ((uint)size > fs->fs_bsize || fragoff(fs, size) != 0) { 1711 kprintf("bsize = %ld, size = %ld, fs = %s\n", 1712 (long)fs->fs_bsize, size, fs->fs_fsmnt); 1713 panic("ffs_checkblk: bad size"); 1714 } 1715 if ((uint)bno >= fs->fs_size) 1716 panic("ffs_checkblk: bad block %d", bno); 1717 error = bread(ip->i_devvp, fsbtodoff(fs, cgtod(fs, dtog(fs, bno))), 1718 (int)fs->fs_cgsize, &bp); 1719 if (error) 1720 panic("ffs_checkblk: cg bread failed"); 1721 cgp = (struct cg *)bp->b_data; 1722 if (!cg_chkmagic(cgp)) 1723 panic("ffs_checkblk: cg magic mismatch"); 1724 blksfree = cg_blksfree(cgp); 1725 bno = dtogd(fs, bno); 1726 if (size == fs->fs_bsize) { 1727 free = ffs_isblock(fs, blksfree, fragstoblks(fs, bno)); 1728 } else { 1729 frags = numfrags(fs, size); 1730 for (free = 0, i = 0; i < frags; i++) 1731 if (isset(blksfree, bno + i)) 1732 free++; 1733 if (free != 0 && free != frags) 1734 panic("ffs_checkblk: partially free fragment"); 1735 } 1736 brelse(bp); 1737 return (!free); 1738 } 1739 #endif /* DIAGNOSTIC */ 1740 1741 /* 1742 * Free an inode. 1743 */ 1744 int 1745 ffs_vfree(struct vnode *pvp, ino_t ino, int mode) 1746 { 1747 if (DOINGSOFTDEP(pvp)) { 1748 softdep_freefile(pvp, ino, mode); 1749 return (0); 1750 } 1751 return (ffs_freefile(pvp, ino, mode)); 1752 } 1753 1754 /* 1755 * Do the actual free operation. 1756 * The specified inode is placed back in the free map. 1757 */ 1758 int 1759 ffs_freefile(struct vnode *pvp, ino_t ino, int mode) 1760 { 1761 struct fs *fs; 1762 struct cg *cgp; 1763 struct inode *pip; 1764 struct buf *bp; 1765 int error, cg; 1766 uint8_t *inosused; 1767 1768 pip = VTOI(pvp); 1769 fs = pip->i_fs; 1770 if ((uint)ino >= fs->fs_ipg * fs->fs_ncg) 1771 panic("ffs_vfree: range: dev = (%d,%d), ino = %"PRId64", fs = %s", 1772 major(pip->i_dev), minor(pip->i_dev), ino, fs->fs_fsmnt); 1773 cg = ino_to_cg(fs, ino); 1774 error = bread(pip->i_devvp, fsbtodoff(fs, cgtod(fs, cg)), 1775 (int)fs->fs_cgsize, &bp); 1776 if (error) { 1777 brelse(bp); 1778 return (error); 1779 } 1780 cgp = (struct cg *)bp->b_data; 1781 if (!cg_chkmagic(cgp)) { 1782 brelse(bp); 1783 return (0); 1784 } 1785 cgp->cg_time = time_second; 1786 inosused = cg_inosused(cgp); 1787 ino %= fs->fs_ipg; 1788 if (isclr(inosused, ino)) { 1789 kprintf("dev = %s, ino = %lu, fs = %s\n", 1790 devtoname(pip->i_dev), (u_long)ino, fs->fs_fsmnt); 1791 if (fs->fs_ronly == 0) 1792 panic("ffs_vfree: freeing free inode"); 1793 } 1794 clrbit(inosused, ino); 1795 if (ino < cgp->cg_irotor) 1796 cgp->cg_irotor = ino; 1797 cgp->cg_cs.cs_nifree++; 1798 fs->fs_cstotal.cs_nifree++; 1799 fs->fs_cs(fs, cg).cs_nifree++; 1800 if ((mode & IFMT) == IFDIR) { 1801 cgp->cg_cs.cs_ndir--; 1802 fs->fs_cstotal.cs_ndir--; 1803 fs->fs_cs(fs, cg).cs_ndir--; 1804 } 1805 fs->fs_fmod = 1; 1806 bdwrite(bp); 1807 return (0); 1808 } 1809 1810 /* 1811 * Find a block of the specified size in the specified cylinder group. 1812 * 1813 * It is a panic if a request is made to find a block if none are 1814 * available. 1815 */ 1816 static ufs_daddr_t 1817 ffs_mapsearch(struct fs *fs, struct cg *cgp, ufs_daddr_t bpref, int allocsiz) 1818 { 1819 ufs_daddr_t bno; 1820 int start, len, loc, i; 1821 int blk, field, subfield, pos; 1822 uint8_t *blksfree; 1823 1824 /* 1825 * find the fragment by searching through the free block 1826 * map for an appropriate bit pattern. 1827 */ 1828 if (bpref) 1829 start = dtogd(fs, bpref) / NBBY; 1830 else 1831 start = cgp->cg_frotor / NBBY; 1832 blksfree = cg_blksfree(cgp); 1833 len = howmany(fs->fs_fpg, NBBY) - start; 1834 loc = scanc((uint)len, (u_char *)&blksfree[start], 1835 (u_char *)fragtbl[fs->fs_frag], 1836 (u_char)(1 << (allocsiz - 1 + (fs->fs_frag % NBBY)))); 1837 if (loc == 0) { 1838 len = start + 1; /* XXX why overlap here? */ 1839 start = 0; 1840 loc = scanc((uint)len, (u_char *)&blksfree[0], 1841 (u_char *)fragtbl[fs->fs_frag], 1842 (u_char)(1 << (allocsiz - 1 + (fs->fs_frag % NBBY)))); 1843 if (loc == 0) { 1844 kprintf("start = %d, len = %d, fs = %s\n", 1845 start, len, fs->fs_fsmnt); 1846 panic("ffs_alloccg: map corrupted"); 1847 /* NOTREACHED */ 1848 } 1849 } 1850 bno = (start + len - loc) * NBBY; 1851 cgp->cg_frotor = bno; 1852 /* 1853 * found the byte in the map 1854 * sift through the bits to find the selected frag 1855 */ 1856 for (i = bno + NBBY; bno < i; bno += fs->fs_frag) { 1857 blk = blkmap(fs, blksfree, bno); 1858 blk <<= 1; 1859 field = around[allocsiz]; 1860 subfield = inside[allocsiz]; 1861 for (pos = 0; pos <= fs->fs_frag - allocsiz; pos++) { 1862 if ((blk & field) == subfield) 1863 return (bno + pos); 1864 field <<= 1; 1865 subfield <<= 1; 1866 } 1867 } 1868 kprintf("bno = %lu, fs = %s\n", (u_long)bno, fs->fs_fsmnt); 1869 panic("ffs_alloccg: block not in map"); 1870 return (-1); 1871 } 1872 1873 /* 1874 * Update the cluster map because of an allocation or free. 1875 * 1876 * Cnt == 1 means free; cnt == -1 means allocating. 1877 */ 1878 static void 1879 ffs_clusteracct(struct fs *fs, struct cg *cgp, ufs_daddr_t blkno, int cnt) 1880 { 1881 int32_t *sump; 1882 int32_t *lp; 1883 u_char *freemapp, *mapp; 1884 int i, start, end, forw, back, map, bit; 1885 1886 if (fs->fs_contigsumsize <= 0) 1887 return; 1888 freemapp = cg_clustersfree(cgp); 1889 sump = cg_clustersum(cgp); 1890 /* 1891 * Allocate or clear the actual block. 1892 */ 1893 if (cnt > 0) 1894 setbit(freemapp, blkno); 1895 else 1896 clrbit(freemapp, blkno); 1897 /* 1898 * Find the size of the cluster going forward. 1899 */ 1900 start = blkno + 1; 1901 end = start + fs->fs_contigsumsize; 1902 if (end >= cgp->cg_nclusterblks) 1903 end = cgp->cg_nclusterblks; 1904 mapp = &freemapp[start / NBBY]; 1905 map = *mapp++; 1906 bit = 1 << (start % NBBY); 1907 for (i = start; i < end; i++) { 1908 if ((map & bit) == 0) 1909 break; 1910 if ((i & (NBBY - 1)) != (NBBY - 1)) { 1911 bit <<= 1; 1912 } else { 1913 map = *mapp++; 1914 bit = 1; 1915 } 1916 } 1917 forw = i - start; 1918 /* 1919 * Find the size of the cluster going backward. 1920 */ 1921 start = blkno - 1; 1922 end = start - fs->fs_contigsumsize; 1923 if (end < 0) 1924 end = -1; 1925 mapp = &freemapp[start / NBBY]; 1926 map = *mapp--; 1927 bit = 1 << (start % NBBY); 1928 for (i = start; i > end; i--) { 1929 if ((map & bit) == 0) 1930 break; 1931 if ((i & (NBBY - 1)) != 0) { 1932 bit >>= 1; 1933 } else { 1934 map = *mapp--; 1935 bit = 1 << (NBBY - 1); 1936 } 1937 } 1938 back = start - i; 1939 /* 1940 * Account for old cluster and the possibly new forward and 1941 * back clusters. 1942 */ 1943 i = back + forw + 1; 1944 if (i > fs->fs_contigsumsize) 1945 i = fs->fs_contigsumsize; 1946 sump[i] += cnt; 1947 if (back > 0) 1948 sump[back] -= cnt; 1949 if (forw > 0) 1950 sump[forw] -= cnt; 1951 /* 1952 * Update cluster summary information. 1953 */ 1954 lp = &sump[fs->fs_contigsumsize]; 1955 for (i = fs->fs_contigsumsize; i > 0; i--) 1956 if (*lp-- > 0) 1957 break; 1958 fs->fs_maxcluster[cgp->cg_cgx] = i; 1959 } 1960 1961 /* 1962 * Fserr prints the name of a filesystem with an error diagnostic. 1963 * 1964 * The form of the error message is: 1965 * fs: error message 1966 */ 1967 static void 1968 ffs_fserr(struct fs *fs, uint uid, char *cp) 1969 { 1970 struct thread *td = curthread; 1971 struct proc *p; 1972 1973 if ((p = td->td_proc) != NULL) { 1974 log(LOG_ERR, "pid %d (%s), uid %d on %s: %s\n", p ? p->p_pid : -1, 1975 p ? p->p_comm : "-", uid, fs->fs_fsmnt, cp); 1976 } else { 1977 log(LOG_ERR, "system thread %p, uid %d on %s: %s\n", 1978 td, uid, fs->fs_fsmnt, cp); 1979 } 1980 } 1981