lfs_inode.c revision 1.95
1 /* $NetBSD: lfs_inode.c,v 1.95 2005/04/23 19:47:51 perseant Exp $ */ 2 3 /*- 4 * Copyright (c) 1999, 2000, 2001, 2002, 2003 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Konrad E. Schroder <perseant (at) hhhh.org>. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 3. All advertising materials mentioning features or use of this software 19 * must display the following acknowledgement: 20 * This product includes software developed by the NetBSD 21 * Foundation, Inc. and its contributors. 22 * 4. Neither the name of The NetBSD Foundation nor the names of its 23 * contributors may be used to endorse or promote products derived 24 * from this software without specific prior written permission. 25 * 26 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 27 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 28 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 29 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 30 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 31 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 32 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 33 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 34 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 35 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 36 * POSSIBILITY OF SUCH DAMAGE. 37 */ 38 /* 39 * Copyright (c) 1986, 1989, 1991, 1993 40 * The Regents of the University of California. All rights reserved. 41 * 42 * Redistribution and use in source and binary forms, with or without 43 * modification, are permitted provided that the following conditions 44 * are met: 45 * 1. Redistributions of source code must retain the above copyright 46 * notice, this list of conditions and the following disclaimer. 47 * 2. Redistributions in binary form must reproduce the above copyright 48 * notice, this list of conditions and the following disclaimer in the 49 * documentation and/or other materials provided with the distribution. 50 * 3. Neither the name of the University nor the names of its contributors 51 * may be used to endorse or promote products derived from this software 52 * without specific prior written permission. 53 * 54 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 55 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 56 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 57 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 58 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 59 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 60 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 61 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 62 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 63 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 64 * SUCH DAMAGE. 65 * 66 * @(#)lfs_inode.c 8.9 (Berkeley) 5/8/95 67 */ 68 69 #include <sys/cdefs.h> 70 __KERNEL_RCSID(0, "$NetBSD: lfs_inode.c,v 1.95 2005/04/23 19:47:51 perseant Exp $"); 71 72 #if defined(_KERNEL_OPT) 73 #include "opt_quota.h" 74 #endif 75 76 #include <sys/param.h> 77 #include <sys/systm.h> 78 #include <sys/mount.h> 79 #include <sys/proc.h> 80 #include <sys/file.h> 81 #include <sys/buf.h> 82 #include <sys/vnode.h> 83 #include <sys/kernel.h> 84 #include <sys/trace.h> 85 #include <sys/resourcevar.h> 86 87 #include <ufs/ufs/quota.h> 88 #include <ufs/ufs/inode.h> 89 #include <ufs/ufs/ufsmount.h> 90 #include <ufs/ufs/ufs_extern.h> 91 92 #include <ufs/lfs/lfs.h> 93 #include <ufs/lfs/lfs_extern.h> 94 95 static int lfs_update_seguse(struct lfs *, long, size_t); 96 static int lfs_indirtrunc (struct inode *, daddr_t, daddr_t, 97 daddr_t, int, long *, long *, long *, size_t *, 98 struct proc *); 99 static int lfs_blkfree (struct lfs *, daddr_t, size_t, long *, size_t *); 100 static int lfs_vtruncbuf(struct vnode *, daddr_t, int, int); 101 102 /* Search a block for a specific dinode. */ 103 struct ufs1_dinode * 104 lfs_ifind(struct lfs *fs, ino_t ino, struct buf *bp) 105 { 106 struct ufs1_dinode *dip = (struct ufs1_dinode *)bp->b_data; 107 struct ufs1_dinode *ldip, *fin; 108 109 ASSERT_NO_SEGLOCK(fs); 110 /* 111 * Read the inode block backwards, since later versions of the 112 * inode will supercede earlier ones. Though it is unlikely, it is 113 * possible that the same inode will appear in the same inode block. 114 */ 115 fin = dip + INOPB(fs); 116 for (ldip = fin - 1; ldip >= dip; --ldip) 117 if (ldip->di_inumber == ino) 118 return (ldip); 119 120 printf("searched %d entries\n", (int)(fin - dip)); 121 printf("offset is 0x%x (seg %d)\n", fs->lfs_offset, 122 dtosn(fs, fs->lfs_offset)); 123 printf("block is 0x%llx (seg %lld)\n", 124 (unsigned long long)dbtofsb(fs, bp->b_blkno), 125 (long long)dtosn(fs, dbtofsb(fs, bp->b_blkno))); 126 127 return NULL; 128 } 129 130 int 131 lfs_update(void *v) 132 { 133 struct vop_update_args /* { 134 struct vnode *a_vp; 135 struct timespec *a_access; 136 struct timespec *a_modify; 137 int a_flags; 138 } */ *ap = v; 139 struct inode *ip; 140 struct vnode *vp = ap->a_vp; 141 struct timespec ts; 142 struct lfs *fs = VFSTOUFS(vp->v_mount)->um_lfs; 143 int s; 144 int flags; 145 146 ASSERT_NO_SEGLOCK(fs); 147 if (vp->v_mount->mnt_flag & MNT_RDONLY) 148 return (0); 149 ip = VTOI(vp); 150 151 /* 152 * If we are called from vinvalbuf, and the file's blocks have 153 * already been scheduled for writing, but the writes have not 154 * yet completed, lfs_vflush will not be called, and vinvalbuf 155 * will cause a panic. So, we must wait until any pending write 156 * for our inode completes, if we are called with UPDATE_WAIT set. 157 */ 158 s = splbio(); 159 simple_lock(&vp->v_interlock); 160 while ((ap->a_flags & (UPDATE_WAIT|UPDATE_DIROP)) == UPDATE_WAIT && 161 WRITEINPROG(vp)) { 162 DLOG((DLOG_SEG, "lfs_update: sleeping on ino %d" 163 " (in progress)\n", ip->i_number)); 164 ltsleep(vp, (PRIBIO+1), "lfs_update", 0, &vp->v_interlock); 165 } 166 simple_unlock(&vp->v_interlock); 167 splx(s); 168 TIMEVAL_TO_TIMESPEC(&time, &ts); 169 LFS_ITIMES(ip, 170 ap->a_access ? ap->a_access : &ts, 171 ap->a_modify ? ap->a_modify : &ts, &ts); 172 if (ap->a_flags & UPDATE_CLOSE) 173 flags = ip->i_flag & (IN_MODIFIED | IN_ACCESSED | IN_CLEANING); 174 else 175 flags = ip->i_flag & (IN_MODIFIED | IN_CLEANING); 176 if (flags == 0) 177 return (0); 178 179 /* If sync, push back the vnode and any dirty blocks it may have. */ 180 if ((ap->a_flags & (UPDATE_WAIT|UPDATE_DIROP)) == UPDATE_WAIT) { 181 /* Avoid flushing VDIROP. */ 182 simple_lock(&fs->lfs_interlock); 183 ++fs->lfs_diropwait; 184 while (vp->v_flag & VDIROP) { 185 DLOG((DLOG_DIROP, "lfs_update: sleeping on inode %d" 186 " (dirops)\n", ip->i_number)); 187 DLOG((DLOG_DIROP, "lfs_update: vflags 0x%x, iflags" 188 " 0x%x\n", vp->v_flag, ip->i_flag)); 189 if (fs->lfs_dirops == 0) 190 lfs_flush_fs(fs, SEGM_SYNC); 191 else 192 ltsleep(&fs->lfs_writer, PRIBIO+1, "lfs_fsync", 193 0, &fs->lfs_interlock); 194 /* XXX KS - by falling out here, are we writing the vn 195 twice? */ 196 } 197 --fs->lfs_diropwait; 198 simple_unlock(&fs->lfs_interlock); 199 return lfs_vflush(vp); 200 } 201 return 0; 202 } 203 204 #define SINGLE 0 /* index of single indirect block */ 205 #define DOUBLE 1 /* index of double indirect block */ 206 #define TRIPLE 2 /* index of triple indirect block */ 207 /* 208 * Truncate the inode oip to at most length size, freeing the 209 * disk blocks. 210 */ 211 /* VOP_BWRITE 1 + NIADDR + VOP_BALLOC == 2 + 2*NIADDR times */ 212 213 int 214 lfs_truncate(void *v) 215 { 216 struct vop_truncate_args /* { 217 struct vnode *a_vp; 218 off_t a_length; 219 int a_flags; 220 struct ucred *a_cred; 221 struct proc *a_p; 222 } */ *ap = v; 223 struct vnode *ovp = ap->a_vp; 224 struct genfs_node *gp = VTOG(ovp); 225 daddr_t lastblock; 226 struct inode *oip = VTOI(ovp); 227 daddr_t bn, lbn, lastiblock[NIADDR], indir_lbn[NIADDR]; 228 /* XXX ondisk32 */ 229 int32_t newblks[NDADDR + NIADDR]; 230 off_t length = ap->a_length; 231 struct lfs *fs; 232 struct buf *bp; 233 int offset, size, level; 234 long count, rcount, blocksreleased = 0, real_released = 0; 235 int i, ioflag, nblocks; 236 int aflags, error, allerror = 0; 237 off_t osize; 238 long lastseg; 239 size_t bc; 240 int obufsize, odb; 241 int usepc; 242 struct ufsmount *ump = oip->i_ump; 243 244 if (length < 0) 245 return (EINVAL); 246 247 /* 248 * Just return and not update modification times. 249 */ 250 if (oip->i_size == length) 251 return (0); 252 253 if (ovp->v_type == VLNK && 254 (oip->i_size < ump->um_maxsymlinklen || 255 (ump->um_maxsymlinklen == 0 && 256 oip->i_ffs1_blocks == 0))) { 257 #ifdef DIAGNOSTIC 258 if (length != 0) 259 panic("lfs_truncate: partial truncate of symlink"); 260 #endif 261 memset((char *)SHORTLINK(oip), 0, (u_int)oip->i_size); 262 oip->i_size = oip->i_ffs1_size = 0; 263 oip->i_flag |= IN_CHANGE | IN_UPDATE; 264 return (VOP_UPDATE(ovp, NULL, NULL, 0)); 265 } 266 if (oip->i_size == length) { 267 oip->i_flag |= IN_CHANGE | IN_UPDATE; 268 return (VOP_UPDATE(ovp, NULL, NULL, 0)); 269 } 270 #ifdef QUOTA 271 if ((error = getinoquota(oip)) != 0) 272 return (error); 273 #endif 274 fs = oip->i_lfs; 275 lfs_imtime(fs); 276 osize = oip->i_size; 277 ioflag = ap->a_flags; 278 usepc = (ovp->v_type == VREG && ovp != fs->lfs_ivnode); 279 280 ASSERT_NO_SEGLOCK(fs); 281 /* 282 * Lengthen the size of the file. We must ensure that the 283 * last byte of the file is allocated. Since the smallest 284 * value of osize is 0, length will be at least 1. 285 */ 286 if (osize < length) { 287 if (length > ump->um_maxfilesize) 288 return (EFBIG); 289 aflags = B_CLRBUF; 290 if (ioflag & IO_SYNC) 291 aflags |= B_SYNC; 292 if (usepc) { 293 if (lblkno(fs, osize) < NDADDR && 294 lblkno(fs, osize) != lblkno(fs, length) && 295 blkroundup(fs, osize) != osize) { 296 off_t eob; 297 298 eob = blkroundup(fs, osize); 299 error = ufs_balloc_range(ovp, osize, 300 eob - osize, ap->a_cred, aflags); 301 if (error) 302 return error; 303 if (ioflag & IO_SYNC) { 304 ovp->v_size = eob; 305 simple_lock(&ovp->v_interlock); 306 VOP_PUTPAGES(ovp, 307 trunc_page(osize & fs->lfs_bmask), 308 round_page(eob), 309 PGO_CLEANIT | PGO_SYNCIO); 310 } 311 } 312 error = ufs_balloc_range(ovp, length - 1, 1, ap->a_cred, 313 aflags); 314 if (error) { 315 (void) VOP_TRUNCATE(ovp, osize, 316 ioflag & IO_SYNC, 317 ap->a_cred, ap->a_p); 318 return error; 319 } 320 uvm_vnp_setsize(ovp, length); 321 oip->i_flag |= IN_CHANGE | IN_UPDATE; 322 KASSERT(ovp->v_size == oip->i_size); 323 oip->i_lfs_hiblk = lblkno(fs, oip->i_size + fs->lfs_bsize - 1) - 1; 324 return (VOP_UPDATE(ovp, NULL, NULL, 0)); 325 } else { 326 error = lfs_reserve(fs, ovp, NULL, 327 btofsb(fs, (NIADDR + 2) << fs->lfs_bshift)); 328 if (error) 329 return (error); 330 error = VOP_BALLOC(ovp, length - 1, 1, ap->a_cred, 331 aflags, &bp); 332 lfs_reserve(fs, ovp, NULL, 333 -btofsb(fs, (NIADDR + 2) << fs->lfs_bshift)); 334 if (error) 335 return (error); 336 oip->i_ffs1_size = oip->i_size = length; 337 uvm_vnp_setsize(ovp, length); 338 (void) VOP_BWRITE(bp); 339 oip->i_flag |= IN_CHANGE | IN_UPDATE; 340 oip->i_lfs_hiblk = lblkno(fs, oip->i_size + fs->lfs_bsize - 1) - 1; 341 return (VOP_UPDATE(ovp, NULL, NULL, 0)); 342 } 343 } 344 345 if ((error = lfs_reserve(fs, ovp, NULL, 346 btofsb(fs, (2 * NIADDR + 3) << fs->lfs_bshift))) != 0) 347 return (error); 348 349 /* 350 * Shorten the size of the file. If the file is not being 351 * truncated to a block boundary, the contents of the 352 * partial block following the end of the file must be 353 * zero'ed in case it ever becomes accessible again because 354 * of subsequent file growth. Directories however are not 355 * zero'ed as they should grow back initialized to empty. 356 */ 357 offset = blkoff(fs, length); 358 lastseg = -1; 359 bc = 0; 360 361 if (ovp != fs->lfs_ivnode) 362 lfs_seglock(fs, SEGM_PROT); 363 if (offset == 0) { 364 oip->i_size = oip->i_ffs1_size = length; 365 } else if (!usepc) { 366 lbn = lblkno(fs, length); 367 aflags = B_CLRBUF; 368 if (ioflag & IO_SYNC) 369 aflags |= B_SYNC; 370 error = VOP_BALLOC(ovp, length - 1, 1, ap->a_cred, aflags, &bp); 371 if (error) { 372 lfs_reserve(fs, ovp, NULL, 373 -btofsb(fs, (2 * NIADDR + 3) << fs->lfs_bshift)); 374 goto errout; 375 } 376 obufsize = bp->b_bufsize; 377 odb = btofsb(fs, bp->b_bcount); 378 oip->i_size = oip->i_ffs1_size = length; 379 size = blksize(fs, oip, lbn); 380 if (ovp->v_type != VDIR) 381 memset((char *)bp->b_data + offset, 0, 382 (u_int)(size - offset)); 383 allocbuf(bp, size, 1); 384 if ((bp->b_flags & (B_LOCKED | B_CALL)) == B_LOCKED) { 385 simple_lock(&lfs_subsys_lock); 386 locked_queue_bytes -= obufsize - bp->b_bufsize; 387 simple_unlock(&lfs_subsys_lock); 388 } 389 if (bp->b_flags & B_DELWRI) 390 fs->lfs_avail += odb - btofsb(fs, size); 391 (void) VOP_BWRITE(bp); 392 } else { /* vp->v_type == VREG && length < osize && offset != 0 */ 393 /* 394 * When truncating a regular file down to a non-block-aligned 395 * size, we must zero the part of last block which is past 396 * the new EOF. We must synchronously flush the zeroed pages 397 * to disk since the new pages will be invalidated as soon 398 * as we inform the VM system of the new, smaller size. 399 * We must do this before acquiring the GLOCK, since fetching 400 * the pages will acquire the GLOCK internally. 401 * So there is a window where another thread could see a whole 402 * zeroed page past EOF, but that's life. 403 */ 404 daddr_t lbn; 405 voff_t eoz; 406 407 aflags = ioflag & IO_SYNC ? B_SYNC : 0; 408 error = ufs_balloc_range(ovp, length - 1, 1, ap->a_cred, 409 aflags); 410 if (error) { 411 lfs_reserve(fs, ovp, NULL, 412 -btofsb(fs, (2 * NIADDR + 3) << fs->lfs_bshift)); 413 goto errout; 414 } 415 lbn = lblkno(fs, length); 416 size = blksize(fs, oip, lbn); 417 eoz = MIN(lblktosize(fs, lbn) + size, osize); 418 uvm_vnp_zerorange(ovp, length, eoz - length); 419 if (round_page(eoz) > round_page(length)) { 420 simple_lock(&ovp->v_interlock); 421 error = VOP_PUTPAGES(ovp, round_page(length), 422 round_page(eoz), 423 PGO_CLEANIT | PGO_DEACTIVATE | 424 ((ioflag & IO_SYNC) ? PGO_SYNCIO : 0)); 425 if (error) { 426 lfs_reserve(fs, ovp, NULL, 427 -btofsb(fs, (2 * NIADDR + 3) << fs->lfs_bshift)); 428 goto errout; 429 } 430 } 431 } 432 433 lockmgr(&gp->g_glock, LK_EXCLUSIVE, NULL); 434 435 oip->i_size = oip->i_ffs1_size = length; 436 uvm_vnp_setsize(ovp, length); 437 /* 438 * Calculate index into inode's block list of 439 * last direct and indirect blocks (if any) 440 * which we want to keep. Lastblock is -1 when 441 * the file is truncated to 0. 442 */ 443 lastblock = lblkno(fs, length + fs->lfs_bsize - 1) - 1; 444 lastiblock[SINGLE] = lastblock - NDADDR; 445 lastiblock[DOUBLE] = lastiblock[SINGLE] - NINDIR(fs); 446 lastiblock[TRIPLE] = lastiblock[DOUBLE] - NINDIR(fs) * NINDIR(fs); 447 nblocks = btofsb(fs, fs->lfs_bsize); 448 /* 449 * Record changed file and block pointers before we start 450 * freeing blocks. lastiblock values are also normalized to -1 451 * for calls to lfs_indirtrunc below. 452 */ 453 memcpy((caddr_t)newblks, (caddr_t)&oip->i_ffs1_db[0], sizeof newblks); 454 for (level = TRIPLE; level >= SINGLE; level--) 455 if (lastiblock[level] < 0) { 456 newblks[NDADDR+level] = 0; 457 lastiblock[level] = -1; 458 } 459 for (i = NDADDR - 1; i > lastblock; i--) 460 newblks[i] = 0; 461 462 oip->i_size = oip->i_ffs1_size = osize; 463 error = lfs_vtruncbuf(ovp, lastblock + 1, 0, 0); 464 if (error && !allerror) 465 allerror = error; 466 467 /* 468 * Indirect blocks first. 469 */ 470 indir_lbn[SINGLE] = -NDADDR; 471 indir_lbn[DOUBLE] = indir_lbn[SINGLE] - NINDIR(fs) - 1; 472 indir_lbn[TRIPLE] = indir_lbn[DOUBLE] - NINDIR(fs) * NINDIR(fs) - 1; 473 for (level = TRIPLE; level >= SINGLE; level--) { 474 bn = oip->i_ffs1_ib[level]; 475 if (bn != 0) { 476 error = lfs_indirtrunc(oip, indir_lbn[level], 477 bn, lastiblock[level], 478 level, &count, &rcount, 479 &lastseg, &bc, ap->a_p); 480 if (error) 481 allerror = error; 482 real_released += rcount; 483 blocksreleased += count; 484 if (lastiblock[level] < 0) { 485 if (oip->i_ffs1_ib[level] > 0) 486 real_released += nblocks; 487 blocksreleased += nblocks; 488 oip->i_ffs1_ib[level] = 0; 489 lfs_blkfree(fs, bn, fs->lfs_bsize, &lastseg, &bc); 490 lfs_deregister_block(ovp, bn); 491 } 492 } 493 if (lastiblock[level] >= 0) 494 goto done; 495 } 496 497 /* 498 * All whole direct blocks or frags. 499 */ 500 for (i = NDADDR - 1; i > lastblock; i--) { 501 long bsize, obsize; 502 503 bn = oip->i_ffs1_db[i]; 504 if (bn == 0) 505 continue; 506 bsize = blksize(fs, oip, i); 507 if (oip->i_ffs1_db[i] > 0) { 508 /* Check for fragment size changes */ 509 obsize = oip->i_lfs_fragsize[i]; 510 real_released += btofsb(fs, obsize); 511 oip->i_lfs_fragsize[i] = 0; 512 } else 513 obsize = 0; 514 blocksreleased += btofsb(fs, bsize); 515 oip->i_ffs1_db[i] = 0; 516 lfs_blkfree(fs, bn, obsize, &lastseg, &bc); 517 lfs_deregister_block(ovp, bn); 518 } 519 if (lastblock < 0) 520 goto done; 521 522 /* 523 * Finally, look for a change in size of the 524 * last direct block; release any frags. 525 */ 526 bn = oip->i_ffs1_db[lastblock]; 527 if (bn != 0) { 528 long oldspace, newspace; 529 #if 0 530 long olddspace; 531 #endif 532 533 /* 534 * Calculate amount of space we're giving 535 * back as old block size minus new block size. 536 */ 537 oldspace = blksize(fs, oip, lastblock); 538 #if 0 539 olddspace = oip->i_lfs_fragsize[lastblock]; 540 #endif 541 542 oip->i_size = oip->i_ffs1_size = length; 543 newspace = blksize(fs, oip, lastblock); 544 if (newspace == 0) 545 panic("itrunc: newspace"); 546 if (oldspace - newspace > 0) { 547 blocksreleased += btofsb(fs, oldspace - newspace); 548 } 549 #if 0 550 if (bn > 0 && olddspace - newspace > 0) { 551 /* No segment accounting here, just vnode */ 552 real_released += btofsb(fs, olddspace - newspace); 553 } 554 #endif 555 } 556 557 done: 558 /* Finish segment accounting corrections */ 559 lfs_update_seguse(fs, lastseg, bc); 560 #ifdef DIAGNOSTIC 561 for (level = SINGLE; level <= TRIPLE; level++) 562 if ((newblks[NDADDR + level] == 0) != 563 (oip->i_ffs1_ib[level]) == 0) { 564 panic("lfs itrunc1"); 565 } 566 for (i = 0; i < NDADDR; i++) 567 if ((newblks[i] == 0) != (oip->i_ffs1_db[i] == 0)) { 568 panic("lfs itrunc2"); 569 } 570 if (length == 0 && 571 (!LIST_EMPTY(&ovp->v_cleanblkhd) || !LIST_EMPTY(&ovp->v_dirtyblkhd))) 572 panic("lfs itrunc3"); 573 #endif /* DIAGNOSTIC */ 574 /* 575 * Put back the real size. 576 */ 577 oip->i_size = oip->i_ffs1_size = length; 578 oip->i_lfs_effnblks -= blocksreleased; 579 oip->i_ffs1_blocks -= real_released; 580 simple_lock(&fs->lfs_interlock); 581 fs->lfs_bfree += blocksreleased; 582 simple_unlock(&fs->lfs_interlock); 583 #ifdef DIAGNOSTIC 584 if (oip->i_size == 0 && 585 (oip->i_ffs1_blocks != 0 || oip->i_lfs_effnblks != 0)) { 586 printf("lfs_truncate: truncate to 0 but %d blks/%d effblks\n", 587 oip->i_ffs1_blocks, oip->i_lfs_effnblks); 588 panic("lfs_truncate: persistent blocks"); 589 } 590 #endif 591 oip->i_flag |= IN_CHANGE; 592 #ifdef QUOTA 593 (void) chkdq(oip, -blocksreleased, NOCRED, 0); 594 #endif 595 lfs_reserve(fs, ovp, NULL, 596 -btofsb(fs, (2 * NIADDR + 3) << fs->lfs_bshift)); 597 lockmgr(&gp->g_glock, LK_RELEASE, NULL); 598 errout: 599 oip->i_lfs_hiblk = lblkno(fs, oip->i_size + fs->lfs_bsize - 1) - 1; 600 if (ovp != fs->lfs_ivnode) 601 lfs_segunlock(fs); 602 return (allerror ? allerror : error); 603 } 604 605 /* Update segment and avail usage information when removing a block. */ 606 static int 607 lfs_blkfree(struct lfs *fs, daddr_t daddr, size_t bsize, long *lastseg, 608 size_t *num) 609 { 610 long seg; 611 int error = 0; 612 613 ASSERT_SEGLOCK(fs); 614 bsize = fragroundup(fs, bsize); 615 if (daddr > 0) { 616 if (*lastseg != (seg = dtosn(fs, daddr))) { 617 error = lfs_update_seguse(fs, *lastseg, *num); 618 *num = bsize; 619 *lastseg = seg; 620 } else 621 *num += bsize; 622 } 623 624 return error; 625 } 626 627 /* Finish the accounting updates for a segment. */ 628 static int 629 lfs_update_seguse(struct lfs *fs, long lastseg, size_t num) 630 { 631 SEGUSE *sup; 632 struct buf *bp; 633 634 ASSERT_SEGLOCK(fs); 635 if (lastseg < 0 || num == 0) 636 return 0; 637 638 LFS_SEGENTRY(sup, fs, lastseg, bp); 639 if (num > sup->su_nbytes) { 640 printf("lfs_truncate: segment %ld short by %ld\n", 641 lastseg, (long)num - sup->su_nbytes); 642 panic("lfs_truncate: negative bytes"); 643 sup->su_nbytes = num; 644 } 645 sup->su_nbytes -= num; 646 LFS_WRITESEGENTRY(sup, fs, lastseg, bp); 647 648 return 0; 649 } 650 651 /* 652 * Release blocks associated with the inode ip and stored in the indirect 653 * block bn. Blocks are free'd in LIFO order up to (but not including) 654 * lastbn. If level is greater than SINGLE, the block is an indirect block 655 * and recursive calls to indirtrunc must be used to cleanse other indirect 656 * blocks. 657 * 658 * NB: triple indirect blocks are untested. 659 */ 660 static int 661 lfs_indirtrunc(struct inode *ip, daddr_t lbn, daddr_t dbn, 662 daddr_t lastbn, int level, long *countp, 663 long *rcountp, long *lastsegp, size_t *bcp, struct proc *p) 664 { 665 int i; 666 struct buf *bp; 667 struct lfs *fs = ip->i_lfs; 668 int32_t *bap; /* XXX ondisk32 */ 669 struct vnode *vp; 670 daddr_t nb, nlbn, last; 671 int32_t *copy = NULL; /* XXX ondisk32 */ 672 long blkcount, rblkcount, factor; 673 int nblocks, blocksreleased = 0, real_released = 0; 674 int error = 0, allerror = 0; 675 676 ASSERT_SEGLOCK(fs); 677 /* 678 * Calculate index in current block of last 679 * block to be kept. -1 indicates the entire 680 * block so we need not calculate the index. 681 */ 682 factor = 1; 683 for (i = SINGLE; i < level; i++) 684 factor *= NINDIR(fs); 685 last = lastbn; 686 if (lastbn > 0) 687 last /= factor; 688 nblocks = btofsb(fs, fs->lfs_bsize); 689 /* 690 * Get buffer of block pointers, zero those entries corresponding 691 * to blocks to be free'd, and update on disk copy first. Since 692 * double(triple) indirect before single(double) indirect, calls 693 * to bmap on these blocks will fail. However, we already have 694 * the on disk address, so we have to set the b_blkno field 695 * explicitly instead of letting bread do everything for us. 696 */ 697 vp = ITOV(ip); 698 bp = getblk(vp, lbn, (int)fs->lfs_bsize, 0, 0); 699 if (bp->b_flags & (B_DONE | B_DELWRI)) { 700 /* Braces must be here in case trace evaluates to nothing. */ 701 trace(TR_BREADHIT, pack(vp, fs->lfs_bsize), lbn); 702 } else { 703 trace(TR_BREADMISS, pack(vp, fs->lfs_bsize), lbn); 704 p->p_stats->p_ru.ru_inblock++; /* pay for read */ 705 bp->b_flags |= B_READ; 706 if (bp->b_bcount > bp->b_bufsize) 707 panic("lfs_indirtrunc: bad buffer size"); 708 bp->b_blkno = fsbtodb(fs, dbn); 709 VOP_STRATEGY(vp, bp); 710 error = biowait(bp); 711 } 712 if (error) { 713 brelse(bp); 714 *countp = *rcountp = 0; 715 return (error); 716 } 717 718 bap = (int32_t *)bp->b_data; /* XXX ondisk32 */ 719 if (lastbn >= 0) { 720 copy = (int32_t *)lfs_malloc(fs, fs->lfs_bsize, LFS_NB_IBLOCK); 721 memcpy((caddr_t)copy, (caddr_t)bap, (u_int)fs->lfs_bsize); 722 memset((caddr_t)&bap[last + 1], 0, 723 /* XXX ondisk32 */ 724 (u_int)(NINDIR(fs) - (last + 1)) * sizeof (int32_t)); 725 error = VOP_BWRITE(bp); 726 if (error) 727 allerror = error; 728 bap = copy; 729 } 730 731 /* 732 * Recursively free totally unused blocks. 733 */ 734 for (i = NINDIR(fs) - 1, nlbn = lbn + 1 - i * factor; i > last; 735 i--, nlbn += factor) { 736 nb = bap[i]; 737 if (nb == 0) 738 continue; 739 if (level > SINGLE) { 740 error = lfs_indirtrunc(ip, nlbn, nb, 741 (daddr_t)-1, level - 1, 742 &blkcount, &rblkcount, 743 lastsegp, bcp, p); 744 if (error) 745 allerror = error; 746 blocksreleased += blkcount; 747 real_released += rblkcount; 748 } 749 lfs_blkfree(fs, nb, fs->lfs_bsize, lastsegp, bcp); 750 if (bap[i] > 0) 751 real_released += nblocks; 752 blocksreleased += nblocks; 753 } 754 755 /* 756 * Recursively free last partial block. 757 */ 758 if (level > SINGLE && lastbn >= 0) { 759 last = lastbn % factor; 760 nb = bap[i]; 761 if (nb != 0) { 762 error = lfs_indirtrunc(ip, nlbn, nb, 763 last, level - 1, &blkcount, 764 &rblkcount, lastsegp, bcp, p); 765 if (error) 766 allerror = error; 767 real_released += rblkcount; 768 blocksreleased += blkcount; 769 } 770 } 771 772 if (copy != NULL) { 773 lfs_free(fs, copy, LFS_NB_IBLOCK); 774 } else { 775 if (bp->b_flags & B_DELWRI) { 776 LFS_UNLOCK_BUF(bp); 777 fs->lfs_avail += btofsb(fs, bp->b_bcount); 778 wakeup(&fs->lfs_avail); 779 } 780 bp->b_flags |= B_INVAL; 781 brelse(bp); 782 } 783 784 *countp = blocksreleased; 785 *rcountp = real_released; 786 return (allerror); 787 } 788 789 /* 790 * Destroy any in core blocks past the truncation length. 791 * Inlined from vtruncbuf, so that lfs_avail could be updated. 792 * We take the seglock to prevent cleaning from occurring while we are 793 * invalidating blocks. 794 */ 795 static int 796 lfs_vtruncbuf(struct vnode *vp, daddr_t lbn, int slpflag, int slptimeo) 797 { 798 struct buf *bp, *nbp; 799 int s, error; 800 struct lfs *fs; 801 voff_t off; 802 803 off = round_page((voff_t)lbn << vp->v_mount->mnt_fs_bshift); 804 simple_lock(&vp->v_interlock); 805 error = VOP_PUTPAGES(vp, off, 0, PGO_FREE | PGO_SYNCIO); 806 if (error) 807 return error; 808 809 fs = VTOI(vp)->i_lfs; 810 s = splbio(); 811 812 ASSERT_SEGLOCK(fs); 813 restart: 814 for (bp = LIST_FIRST(&vp->v_cleanblkhd); bp; bp = nbp) { 815 nbp = LIST_NEXT(bp, b_vnbufs); 816 if (bp->b_lblkno < lbn) 817 continue; 818 simple_lock(&bp->b_interlock); 819 if (bp->b_flags & B_BUSY) { 820 bp->b_flags |= B_WANTED; 821 error = ltsleep(bp, slpflag | (PRIBIO + 1) | PNORELOCK, 822 "lfs_vtruncbuf", slptimeo, &bp->b_interlock); 823 if (error) { 824 splx(s); 825 return (error); 826 } 827 goto restart; 828 } 829 bp->b_flags |= B_BUSY | B_INVAL | B_VFLUSH; 830 if (bp->b_flags & B_DELWRI) { 831 bp->b_flags &= ~B_DELWRI; 832 fs->lfs_avail += btofsb(fs, bp->b_bcount); 833 wakeup(&fs->lfs_avail); 834 } 835 LFS_UNLOCK_BUF(bp); 836 simple_unlock(&bp->b_interlock); 837 brelse(bp); 838 } 839 840 for (bp = LIST_FIRST(&vp->v_dirtyblkhd); bp; bp = nbp) { 841 nbp = LIST_NEXT(bp, b_vnbufs); 842 if (bp->b_lblkno < lbn) 843 continue; 844 simple_lock(&bp->b_interlock); 845 if (bp->b_flags & B_BUSY) { 846 bp->b_flags |= B_WANTED; 847 error = ltsleep(bp, slpflag | (PRIBIO + 1) | PNORELOCK, 848 "lfs_vtruncbuf", slptimeo, &bp->b_interlock); 849 if (error) { 850 splx(s); 851 return (error); 852 } 853 goto restart; 854 } 855 bp->b_flags |= B_BUSY | B_INVAL | B_VFLUSH; 856 if (bp->b_flags & B_DELWRI) { 857 bp->b_flags &= ~B_DELWRI; 858 fs->lfs_avail += btofsb(fs, bp->b_bcount); 859 wakeup(&fs->lfs_avail); 860 } 861 LFS_UNLOCK_BUF(bp); 862 simple_unlock(&bp->b_interlock); 863 brelse(bp); 864 } 865 866 splx(s); 867 868 return (0); 869 } 870 871
Indexes created Sat Sep 27 14:09:57 GMT 2025