lfs_inode.c revision 1.24
1 /* $NetBSD: lfs_inode.c,v 1.24 1999/04/12 00:30:08 perseant Exp $ */ 2 3 /*- 4 * Copyright (c) 1999 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. All advertising materials mentioning features or use of this software 51 * must display the following acknowledgement: 52 * This product includes software developed by the University of 53 * California, Berkeley and its contributors. 54 * 4. Neither the name of the University nor the names of its contributors 55 * may be used to endorse or promote products derived from this software 56 * without specific prior written permission. 57 * 58 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 59 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 60 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 61 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 62 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 63 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 64 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 65 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 66 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 67 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 68 * SUCH DAMAGE. 69 * 70 * @(#)lfs_inode.c 8.9 (Berkeley) 5/8/95 71 */ 72 73 #if defined(_KERNEL) && !defined(_LKM) 74 #include "opt_quota.h" 75 #endif 76 77 #include <sys/param.h> 78 #include <sys/systm.h> 79 #include <sys/mount.h> 80 #include <sys/proc.h> 81 #include <sys/file.h> 82 #include <sys/buf.h> 83 #include <sys/vnode.h> 84 #include <sys/kernel.h> 85 #include <sys/malloc.h> 86 87 #include <vm/vm.h> 88 89 #include <ufs/ufs/quota.h> 90 #include <ufs/ufs/inode.h> 91 #include <ufs/ufs/ufsmount.h> 92 #include <ufs/ufs/ufs_extern.h> 93 94 #include <ufs/lfs/lfs.h> 95 #include <ufs/lfs/lfs_extern.h> 96 97 static int lfs_vinvalbuf __P((struct vnode *, struct ucred *, struct proc *, ufs_daddr_t)); 98 99 /* Search a block for a specific dinode. */ 100 struct dinode * 101 lfs_ifind(fs, ino, dip) 102 struct lfs *fs; 103 ino_t ino; 104 register struct dinode *dip; 105 { 106 register int cnt; 107 register struct dinode *ldip; 108 109 for (cnt = INOPB(fs), ldip = dip + (cnt - 1); cnt--; --ldip) 110 if (ldip->di_inumber == ino) 111 return (ldip); 112 113 panic("lfs_ifind: dinode %u not found", ino); 114 /* NOTREACHED */ 115 } 116 117 int 118 lfs_update(v) 119 void *v; 120 { 121 struct vop_update_args /* { 122 struct vnode *a_vp; 123 struct timespec *a_access; 124 struct timespec *a_modify; 125 int a_waitfor; 126 } */ *ap = v; 127 struct inode *ip; 128 struct vnode *vp = ap->a_vp; 129 int mod, oflag; 130 struct timespec ts; 131 132 if (vp->v_mount->mnt_flag & MNT_RDONLY) 133 return (0); 134 ip = VTOI(vp); 135 136 /* 137 * If we are called from vinvalbuf, and the file's blocks have 138 * already been scheduled for writing, but the writes have not 139 * yet completed, lfs_vflush will not be called, and vinvalbuf 140 * will cause a panic. So, we must wait until any pending write 141 * for our inode completes, if we are called with LFS_SYNC set. 142 */ 143 while((ap->a_waitfor & LFS_SYNC) && WRITEINPROG(vp)) { 144 #ifdef DEBUG_LFS 145 printf("lfs_update: sleeping on inode %d\n",ip->i_number); 146 #endif 147 tsleep(vp, (PRIBIO+1), "lfs_update", 0); 148 } 149 mod = ip->i_flag & IN_MODIFIED; 150 oflag = ip->i_flag; 151 TIMEVAL_TO_TIMESPEC(&time, &ts); 152 LFS_ITIMES(ip, 153 ap->a_access ? ap->a_access : &ts, 154 ap->a_modify ? ap->a_modify : &ts, &ts); 155 if (!mod && (ip->i_flag & IN_MODIFIED)) 156 ip->i_lfs->lfs_uinodes++; 157 if ((ip->i_flag & (IN_MODIFIED|IN_CLEANING)) == 0) { 158 return (0); 159 } 160 161 /* If sync, push back the vnode and any dirty blocks it may have. */ 162 return (ap->a_waitfor & LFS_SYNC ? lfs_vflush(vp) : 0); 163 } 164 165 /* Update segment usage information when removing a block. */ 166 #define UPDATE_SEGUSE \ 167 if (lastseg != -1) { \ 168 LFS_SEGENTRY(sup, fs, lastseg, sup_bp); \ 169 if (num > sup->su_nbytes) { \ 170 panic("lfs_truncate: negative bytes in segment %d\n", \ 171 lastseg); \ 172 sup->su_nbytes = 0; \ 173 } else \ 174 sup->su_nbytes -= num; \ 175 e1 = VOP_BWRITE(sup_bp); \ 176 fragsreleased += numfrags(fs, num); \ 177 } 178 179 #define SEGDEC(S) { \ 180 if (daddr != 0) { \ 181 if (lastseg != (seg = datosn(fs, daddr))) { \ 182 UPDATE_SEGUSE; \ 183 num = (S); \ 184 lastseg = seg; \ 185 } else \ 186 num += (S); \ 187 } \ 188 } 189 190 /* 191 * Truncate the inode ip to at most length size. Update segment usage 192 * table information. 193 */ 194 /* ARGSUSED */ 195 int 196 lfs_truncate(v) 197 void *v; 198 { 199 struct vop_truncate_args /* { 200 struct vnode *a_vp; 201 off_t a_length; 202 int a_flags; 203 struct ucred *a_cred; 204 struct proc *a_p; 205 } */ *ap = v; 206 register struct indir *inp; 207 register int i; 208 register ufs_daddr_t *daddrp; 209 register struct vnode *vp = ap->a_vp; 210 off_t length = ap->a_length; 211 struct buf *bp, *sup_bp; 212 struct ifile *ifp; 213 struct inode *ip; 214 struct lfs *fs; 215 struct indir a[NIADDR + 2], a_end[NIADDR + 2]; 216 SEGUSE *sup; 217 ufs_daddr_t daddr, lastblock, lbn, olastblock; 218 ufs_daddr_t oldsize_lastblock, oldsize_newlast, newsize; 219 long off, a_released, fragsreleased, i_released; 220 int e1, e2, depth, lastseg, num, offset, seg, freesize; 221 222 ip = VTOI(vp); 223 224 if (vp->v_type == VLNK && vp->v_mount->mnt_maxsymlinklen > 0) { 225 #ifdef DIAGNOSTIC 226 if (length != 0) 227 panic("lfs_truncate: partial truncate of symlink"); 228 #endif 229 bzero((char *)&ip->i_ffs_shortlink, (u_int)ip->i_ffs_size); 230 ip->i_ffs_size = 0; 231 ip->i_flag |= IN_CHANGE | IN_UPDATE; 232 return (VOP_UPDATE(vp, NULL, NULL, 0)); 233 } 234 uvm_vnp_setsize(vp, length); 235 236 fs = ip->i_lfs; 237 lfs_imtime(fs); 238 239 /* If length is larger than the file, just update the times. */ 240 if (ip->i_ffs_size <= length) { 241 ip->i_flag |= IN_CHANGE | IN_UPDATE; 242 return (VOP_UPDATE(vp, NULL, NULL, 0)); 243 } 244 245 /* 246 * Make sure no writes happen while we're truncating. 247 * Otherwise, blocks which are accounted for on the inode 248 * *and* which have been created for cleaning can coexist, 249 * and cause us to overcount, and panic below. 250 * 251 * XXX KS - too restrictive? Maybe only when cleaning? 252 */ 253 while(fs->lfs_seglock) { 254 tsleep(&fs->lfs_seglock, (PRIBIO+1), "lfs_truncate", 0); 255 } 256 257 /* 258 * Calculate index into inode's block list of last direct and indirect 259 * blocks (if any) which we want to keep. Lastblock is 0 when the 260 * file is truncated to 0. 261 */ 262 lastblock = lblkno(fs, length + fs->lfs_bsize - 1); 263 olastblock = lblkno(fs, ip->i_ffs_size + fs->lfs_bsize - 1) - 1; 264 265 /* 266 * Update the size of the file. If the file is not being truncated to 267 * a block boundry, the contents of the partial block following the end 268 * must be zero'd in case it ever becomes accessible again 269 * because of subsequent file growth. For this part of the code, 270 * oldsize_newlast refers to the old size of the new last block in the 271 * file. 272 */ 273 offset = blkoff(fs, length); 274 lbn = lblkno(fs, length); 275 oldsize_newlast = blksize(fs, ip, lbn); 276 277 /* Now set oldsize to the current size of the current last block */ 278 oldsize_lastblock = blksize(fs, ip, olastblock); 279 if (offset == 0) 280 ip->i_ffs_size = length; 281 else { 282 #ifdef QUOTA 283 if ((e1 = getinoquota(ip)) != 0) 284 return (e1); 285 #endif 286 if ((e1 = bread(vp, lbn, oldsize_newlast, NOCRED, &bp)) != 0) { 287 printf("lfs_truncate: bread: %d\n",e1); 288 brelse(bp); 289 return (e1); 290 } 291 ip->i_ffs_size = length; 292 (void)uvm_vnp_uncache(vp); 293 newsize = blksize(fs, ip, lbn); 294 bzero((char *)bp->b_data + offset, (u_int)(newsize - offset)); 295 #ifdef DEBUG 296 if(bp->b_flags & B_CALL) 297 panic("Can't allocbuf malloced buffer!"); 298 else 299 #endif 300 allocbuf(bp, newsize); 301 if(oldsize_newlast > newsize) 302 ip->i_ffs_blocks -= btodb(oldsize_newlast - newsize); 303 if ((e1 = VOP_BWRITE(bp)) != 0) { 304 printf("lfs_truncate: bwrite: %d\n",e1); 305 return (e1); 306 } 307 } 308 /* 309 * Modify sup->su_nbyte counters for each deleted block; keep track 310 * of number of blocks removed for ip->i_ffs_blocks. 311 */ 312 fragsreleased = 0; 313 num = 0; 314 lastseg = -1; 315 316 for (lbn = olastblock; lbn >= lastblock;) { 317 /* XXX use run length from bmap array to make this faster */ 318 ufs_bmaparray(vp, lbn, &daddr, a, &depth, NULL); 319 if (lbn == olastblock) { 320 for (i = NIADDR + 2; i--;) 321 a_end[i] = a[i]; 322 freesize = oldsize_lastblock; 323 } else 324 freesize = fs->lfs_bsize; 325 switch (depth) { 326 case 0: /* Direct block. */ 327 daddr = ip->i_ffs_db[lbn]; 328 SEGDEC(freesize); 329 ip->i_ffs_db[lbn] = 0; 330 --lbn; 331 break; 332 #ifdef DIAGNOSTIC 333 case 1: /* An indirect block. */ 334 panic("lfs_truncate: ufs_bmaparray returned depth 1"); 335 /* NOTREACHED */ 336 #endif 337 default: /* Chain of indirect blocks. */ 338 inp = a + --depth; 339 if (inp->in_off > 0 && lbn != lastblock) { 340 lbn -= inp->in_off < lbn - lastblock ? 341 inp->in_off : lbn - lastblock; 342 break; 343 } 344 for (; depth && (inp->in_off == 0 || lbn == lastblock); 345 --inp, --depth) { 346 if (bread(vp, 347 inp->in_lbn, fs->lfs_bsize, NOCRED, &bp)) 348 panic("lfs_truncate: bread bno %d", 349 inp->in_lbn); 350 daddrp = (ufs_daddr_t *)bp->b_data + inp->in_off; 351 for (i = inp->in_off; 352 i++ <= a_end[depth].in_off;) { 353 daddr = *daddrp++; 354 SEGDEC(freesize); 355 } 356 a_end[depth].in_off = NINDIR(fs) - 1; 357 if (inp->in_off == 0) 358 brelse (bp); 359 else { 360 bzero((ufs_daddr_t *)bp->b_data + 361 inp->in_off, fs->lfs_bsize - 362 inp->in_off * sizeof(ufs_daddr_t)); 363 if ((e1 = VOP_BWRITE(bp)) != 0) { 364 printf("lfs_truncate: indir bwrite: %d\n",e1); 365 return (e1); 366 } 367 } 368 } 369 if (depth == 0 && a[1].in_off == 0) { 370 off = a[0].in_off; 371 daddr = ip->i_ffs_ib[off]; 372 SEGDEC(freesize); 373 ip->i_ffs_ib[off] = 0; 374 } 375 if (lbn == lastblock || lbn <= NDADDR) 376 --lbn; 377 else { 378 lbn -= NINDIR(fs); 379 if (lbn < lastblock) 380 lbn = lastblock; 381 } 382 } 383 } 384 UPDATE_SEGUSE; 385 386 /* If truncating the file to 0, update the version number. */ 387 if (length == 0) { 388 LFS_IENTRY(ifp, fs, ip->i_number, bp); 389 ++ifp->if_version; 390 (void) VOP_BWRITE(bp); 391 } 392 #ifdef DIAGNOSTIC 393 if (ip->i_ffs_blocks < fragstodb(fs, fragsreleased)) { 394 panic("lfs_truncate: frag count < 0 (%d<%ld), ino %d\n", 395 ip->i_ffs_blocks, fragstodb(fs, fragsreleased), 396 ip->i_number); 397 fragsreleased = dbtofrags(fs, ip->i_ffs_blocks); 398 } 399 #endif 400 ip->i_ffs_blocks -= fragstodb(fs, fragsreleased); 401 fs->lfs_bfree += fragstodb(fs, fragsreleased); 402 ip->i_flag |= IN_CHANGE | IN_UPDATE; 403 /* 404 * Traverse dirty block list counting number of dirty buffers 405 * that are being deleted out of the cache, so that the lfs_avail 406 * field can be updated. 407 */ 408 a_released = 0; 409 i_released = 0; 410 for (bp = vp->v_dirtyblkhd.lh_first; bp; bp = bp->b_vnbufs.le_next) { 411 412 /* XXX KS - Don't miscount if we're not truncating to zero. */ 413 if(length>0 && !(bp->b_lblkno >= 0 && bp->b_lblkno > lastblock) 414 && !(bp->b_lblkno < 0 && bp->b_lblkno < -lastblock-NIADDR)) 415 continue; 416 417 if (bp->b_flags & B_LOCKED) { 418 a_released += numfrags(fs, bp->b_bcount); 419 /* 420 * XXX 421 * When buffers are created in the cache, their block 422 * number is set equal to their logical block number. 423 * If that is still true, we are assuming that the 424 * blocks are new (not yet on disk) and weren't 425 * counted above. However, there is a slight chance 426 * that a block's disk address is equal to its logical 427 * block number in which case, we'll get an overcounting 428 * here. 429 */ 430 if (bp->b_blkno == bp->b_lblkno) { 431 i_released += numfrags(fs, bp->b_bcount); 432 } 433 } 434 } 435 fragsreleased = i_released; 436 #ifdef DIAGNOSTIC 437 if (fragsreleased > dbtofrags(fs, ip->i_ffs_blocks)) { 438 printf("lfs_inode: %ld frags released > %d in inode %d\n", 439 fragsreleased, dbtofrags(fs, ip->i_ffs_blocks), 440 ip->i_number); 441 fragsreleased = dbtofrags(fs, ip->i_ffs_blocks); 442 } 443 #endif 444 fs->lfs_bfree += fragstodb(fs, fragsreleased); 445 ip->i_ffs_blocks -= fragstodb(fs, fragsreleased); 446 #ifdef DIAGNOSTIC 447 if (length == 0 && ip->i_ffs_blocks != 0) { 448 printf("lfs_inode: trunc to zero, but %d blocks left on inode %d\n", 449 ip->i_ffs_blocks, ip->i_number); 450 panic("lfs_inode\n"); 451 } 452 #endif 453 fs->lfs_avail += fragstodb(fs, a_released); 454 if(length>0) 455 e1 = lfs_vinvalbuf(vp, ap->a_cred, ap->a_p, lastblock-1); 456 else 457 e1 = vinvalbuf(vp, 0, ap->a_cred, ap->a_p, 0, 0); 458 e2 = VOP_UPDATE(vp, NULL, NULL, 0); 459 if(e1) 460 printf("lfs_truncate: vinvalbuf: %d\n",e1); 461 if(e2) 462 printf("lfs_truncate: update: %d\n",e2); 463 464 return (e1 ? e1 : e2 ? e2 : 0); 465 } 466 467 /* 468 * Get rid of blocks a la vinvalbuf; but only blocks that are of a higher 469 * lblkno than the file size allows. 470 */ 471 int 472 lfs_vinvalbuf(vp, cred, p, maxblk) 473 register struct vnode *vp; 474 struct ucred *cred; 475 struct proc *p; 476 ufs_daddr_t maxblk; 477 { 478 register struct buf *bp; 479 struct buf *nbp, *blist; 480 int i, s, error, dirty; 481 482 top: 483 dirty=0; 484 for (i=0;i<2;i++) { 485 if(i==0) 486 blist = vp->v_cleanblkhd.lh_first; 487 else /* i == 1 */ 488 blist = vp->v_dirtyblkhd.lh_first; 489 490 for (bp = blist; bp; bp = nbp) { 491 nbp = bp->b_vnbufs.le_next; 492 493 s = splbio(); 494 if (bp->b_flags & B_GATHERED) { 495 error = tsleep(vp, PRIBIO+1, "lfs_vin2", 0); 496 splx(s); 497 if(error) 498 return error; 499 goto top; 500 } 501 if (bp->b_flags & B_BUSY) { 502 bp->b_flags |= B_WANTED; 503 error = tsleep((caddr_t)bp, 504 (PRIBIO + 1), "lfs_vinval", 0); 505 splx(s); 506 if (error) 507 return (error); 508 goto top; 509 } 510 511 bp->b_flags |= B_BUSY; 512 splx(s); 513 if((bp->b_lblkno >= 0 && bp->b_lblkno > maxblk) 514 || (bp->b_lblkno < 0 && bp->b_lblkno < -maxblk-(NIADDR-1))) 515 { 516 bp->b_flags |= B_INVAL | B_VFLUSH; 517 if(bp->b_flags & B_CALL) { 518 lfs_freebuf(bp); 519 } else { 520 brelse(bp); 521 } 522 ++dirty; 523 } else { 524 /* 525 * This buffer is still on its free list. 526 * So don't brelse, but wake up any sleepers. 527 */ 528 bp->b_flags &= ~B_BUSY; 529 if(bp->b_flags & B_WANTED) { 530 bp->b_flags &= ~(B_WANTED|B_AGE); 531 wakeup(bp); 532 } 533 } 534 } 535 } 536 if(dirty) 537 goto top; 538 return (0); 539 } 540
Indexes created Fri Sep 26 08:10:20 GMT 2025