bsd/ufs/ufs/ufs_bmap.c

   1 /*
   2  * Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
   3  *
   4  * @APPLE_LICENSE_HEADER_START@
   5  *
   6  * The contents of this file constitute Original Code as defined in and
   7  * are subject to the Apple Public Source License Version 1.1 (the
   8  * "License").  You may not use this file except in compliance with the
   9  * License.  Please obtain a copy of the License at
  10  * http://www.apple.com/publicsource and read it before using this file.
  11  *
  12  * This Original Code and all software distributed under the License are
  13  * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  14  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  15  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  16  * FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT.  Please see the
  17  * License for the specific language governing rights and limitations
  18  * under the License.
  19  *
  20  * @APPLE_LICENSE_HEADER_END@
  21  */
  22 /* Copyright (c) 1995 NeXT Computer, Inc. All Rights Reserved */
  23 /*
  24  * Copyright (c) 1989, 1991, 1993
  25  *      The Regents of the University of California.  All rights reserved.
  26  * (c) UNIX System Laboratories, Inc.
  27  * All or some portions of this file are derived from material licensed
  28  * to the University of California by American Telephone and Telegraph
  29  * Co. or Unix System Laboratories, Inc. and are reproduced herein with
  30  * the permission of UNIX System Laboratories, Inc.
  31  *
  32  * Redistribution and use in source and binary forms, with or without
  33  * modification, are permitted provided that the following conditions
  34  * are met:
  35  * 1. Redistributions of source code must retain the above copyright
  36  *    notice, this list of conditions and the following disclaimer.
  37  * 2. Redistributions in binary form must reproduce the above copyright
  38  *    notice, this list of conditions and the following disclaimer in the
  39  *    documentation and/or other materials provided with the distribution.
  40  * 3. All advertising materials mentioning features or use of this software
  41  *    must display the following acknowledgement:
  42  *      This product includes software developed by the University of
  43  *      California, Berkeley and its contributors.
  44  * 4. Neither the name of the University nor the names of its contributors
  45  *    may be used to endorse or promote products derived from this software
  46  *    without specific prior written permission.
  47  *
  48  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  49  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  50  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  51  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  52  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  53  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  54  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  55  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  56  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  57  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  58  * SUCH DAMAGE.
  59  *
  60  *      @(#)ufs_bmap.c  8.7 (Berkeley) 3/21/95
  61  */
  62 /*
  63  * HISTORY
  64  * 11-July-97  Umesh Vaishampayan (umeshv@apple.com)
  65  *      Cleanup. Fixed compilation error when tracing is turned on.
  66  */
  67 #include <rev_endian_fs.h>
  68 #include <sys/param.h>
  69 #include <sys/buf.h>
  70 #include <sys/proc_internal.h>  /* for p_stats */
  71 #include <sys/vnode_internal.h>
  72 #include <sys/mount_internal.h>
  73 #include <sys/resourcevar.h>
  74 #include <sys/trace.h>
  75 #include <sys/quota.h>
  76
  77 #include <miscfs/specfs/specdev.h>
  78
  79 #include <ufs/ufs/quota.h>
  80 #include <ufs/ufs/inode.h>
  81 #include <ufs/ufs/ufsmount.h>
  82 #include <ufs/ufs/ufs_extern.h>
  83 #if REV_ENDIAN_FS
  84 #include <ufs/ufs/ufs_byte_order.h>
  85 #include <architecture/byte_order.h>
  86 #endif /* REV_ENDIAN_FS */
  87
  88
  89 /*
  90  * Indirect blocks are now on the vnode for the file.  They are given negative
  91  * logical block numbers.  Indirect blocks are addressed by the negative
  92  * address of the first data block to which they point.  Double indirect blocks
  93  * are addressed by one less than the address of the first indirect block to
  94  * which they point.  Triple indirect blocks are addressed by one less than
  95  * the address of the first double indirect block to which they point.
  96  *
  97  * ufs_bmaparray does the bmap conversion, and if requested returns the
  98  * array of logical blocks which must be traversed to get to a block.
  99  * Each entry contains the offset into that block that gets you to the
 100  * next block and the disk address of the block (if it is assigned).
 101  */
 102
 103 int
 104 ufs_bmaparray(vp, bn, bnp, ap, nump, runp)
 105         vnode_t     vp;
 106         ufs_daddr_t bn;
 107         ufs_daddr_t *bnp;
 108         struct indir *ap;
 109         int *nump;
 110         int *runp;
 111 {
 112         register struct inode *ip;
 113         struct buf *bp;
 114         struct ufsmount *ump;
 115         struct mount *mp;
 116         struct vnode *devvp;
 117         struct indir a[NIADDR], *xap;
 118         ufs_daddr_t daddr;
 119         long metalbn;
 120         int error, maxrun, num;
 121 #if REV_ENDIAN_FS
 122         int rev_endian=0;
 123 #endif /* REV_ENDIAN_FS */
 124
 125         ip = VTOI(vp);
 126         mp = vp->v_mount;
 127         ump = VFSTOUFS(mp);
 128
 129 #if REV_ENDIAN_FS
 130         rev_endian=(mp->mnt_flag & MNT_REVEND);
 131 #endif /* REV_ENDIAN_FS */
 132
 133 #if DIAGNOSTIC
 134         if (ap != NULL && nump == NULL || ap == NULL && nump != NULL)
 135                 panic("ufs_bmaparray: invalid arguments");
 136 #endif
 137
 138         if (runp) {
 139                 /*
 140                  * XXX
 141                  * If MAXPHYSIO is the largest transfer the disks can handle,
 142                  * we probably want maxrun to be 1 block less so that we
 143                  * don't create a block larger than the device can handle.
 144                  */
 145                 *runp = 0;
 146                 maxrun = MAXPHYSIO / mp->mnt_vfsstat.f_iosize - 1;
 147         }
 148
 149         xap = ap == NULL ? a : ap;
 150         if (!nump)
 151                 nump = &num;
 152         if (error = ufs_getlbns(vp, bn, xap, nump))
 153                 return (error);
 154
 155         num = *nump;
 156         if (num == 0) {
 157                 *bnp = blkptrtodb(ump, ip->i_db[bn]);
 158                 if (*bnp == 0)
 159                         *bnp = -1;
 160                 else if (runp)
 161                         for (++bn; bn < NDADDR && *runp < maxrun &&
 162                             is_sequential(ump, ip->i_db[bn - 1], ip->i_db[bn]);
 163                             ++bn, ++*runp);
 164                 return (0);
 165         }
 166
 167
 168         /* Get disk address out of indirect block array */
 169         daddr = ip->i_ib[xap->in_off];
 170
 171         devvp = VFSTOUFS(vp->v_mount)->um_devvp;
 172         for (bp = NULL, ++xap; --num; ++xap) {
 173                 ufs_daddr_t *dataptr;
 174                 int bop;
 175
 176                 if ((metalbn = xap->in_lbn) == bn)
 177                        /*
 178                         * found the indirect block we were
 179                         * looking for... exit the loop
 180                         */
 181                        break;
 182
 183                 if (daddr == 0)
 184                         bop = BLK_ONLYVALID | BLK_META;
 185                 else
 186                         bop = BLK_META;
 187
 188                 if (bp)
 189                         buf_brelse(bp);
 190                 bp = buf_getblk(vp, (daddr64_t)((unsigned)metalbn), mp->mnt_vfsstat.f_iosize, 0, 0, bop);
 191
 192                 if (bp == 0) {
 193                         /*
 194                          * Exit the loop if there is no disk address assigned yet and
 195                          * the indirect block isn't in the cache
 196                          */
 197                         break;
 198                 }
 199                 /*
 200                  * If we get here, we've either got the block in the cache
 201                  * or we have a disk address for it, go fetch it.
 202                  */
 203                 xap->in_exists = 1;
 204
 205                 if (buf_valid(bp)) {
 206                         trace(TR_BREADHIT, pack(vp, mp->mnt_vfsstat.f_iosize), metalbn);
 207                 }
 208                 else {
 209                         trace(TR_BREADMISS, pack(vp, mp->mnt_vfsstat.f_iosize), metalbn);
 210                         buf_setblkno(bp, blkptrtodb(ump, (daddr64_t)((unsigned)daddr)));
 211                         buf_setflags(bp, B_READ);
 212                         VNOP_STRATEGY(bp);
 213                         current_proc()->p_stats->p_ru.ru_inblock++;     /* XXX */
 214                         if (error = (int)buf_biowait(bp)) {
 215                                 buf_brelse(bp);
 216                                 return (error);
 217                         }
 218                 }
 219                 dataptr = (ufs_daddr_t *)buf_dataptr(bp);
 220                 daddr = dataptr[xap->in_off];
 221 #if REV_ENDIAN_FS
 222                 if (rev_endian)
 223                         daddr = NXSwapLong(daddr);
 224 #endif /* REV_ENDIAN_FS */
 225                 if (num == 1 && daddr && runp) {
 226 #if REV_ENDIAN_FS
 227                 if (rev_endian) {
 228                         for (bn = xap->in_off + 1;
 229                             bn < MNINDIR(ump) && *runp < maxrun &&
 230                             is_sequential(ump,
 231                             NXSwapLong(dataptr[bn - 1]),
 232                             NXSwapLong(dataptr[bn]));
 233                             ++bn, ++*runp);
 234                  } else {
 235 #endif /* REV_ENDIAN_FS */
 236                         for (bn = xap->in_off + 1;
 237                             bn < MNINDIR(ump) && *runp < maxrun &&
 238                             is_sequential(ump,
 239                             dataptr[bn - 1],
 240                             dataptr[bn]);
 241                             ++bn, ++*runp);
 242 #if REV_ENDIAN_FS
 243                 }
 244 #endif /* REV_ENDIAN_FS */
 245                 }
 246         }
 247         if (bp)
 248                 buf_brelse(bp);
 249
 250         daddr = blkptrtodb(ump, daddr);
 251         *bnp = daddr == 0 ? -1 : daddr;
 252         return (0);
 253 }
 254
 255 /*
 256  * Create an array of logical block number/offset pairs which represent the
 257  * path of indirect blocks required to access a data block.  The first "pair"
 258  * contains the logical block number of the appropriate single, double or
 259  * triple indirect block and the offset into the inode indirect block array.
 260  * Note, the logical block number of the inode single/double/triple indirect
 261  * block appears twice in the array, once with the offset into the i_ib and
 262  * once with the offset into the page itself.
 263  */
 264 int
 265 ufs_getlbns(vp, bn, ap, nump)
 266         struct vnode *vp;
 267         ufs_daddr_t bn;
 268         struct indir *ap;
 269         int *nump;
 270 {
 271         long metalbn, realbn;
 272         struct ufsmount *ump;
 273         int blockcnt, i, numlevels, off;
 274
 275         ump = VFSTOUFS(vp->v_mount);
 276         if (nump)
 277                 *nump = 0;
 278         numlevels = 0;
 279         realbn = bn;
 280         if ((long)bn < 0)
 281                 bn = -(long)bn;
 282
 283         /* The first NDADDR blocks are direct blocks. */
 284         if (bn < NDADDR)
 285                 return (0);
 286
 287         /*
 288          * Determine the number of levels of indirection.  After this loop
 289          * is done, blockcnt indicates the number of data blocks possible
 290          * at the given level of indirection, and NIADDR - i is the number
 291          * of levels of indirection needed to locate the requested block.
 292          */
 293         for (blockcnt = 1, i = NIADDR, bn -= NDADDR;; i--, bn -= blockcnt) {
 294                 if (i == 0)
 295                         return (EFBIG);
 296                 blockcnt *= MNINDIR(ump);
 297                 if (bn < blockcnt)
 298                         break;
 299         }
 300
 301         /* Calculate the address of the first meta-block. */
 302         if (realbn >= 0)
 303                 metalbn = -(realbn - bn + NIADDR - i);
 304         else
 305                 metalbn = -(-realbn - bn + NIADDR - i);
 306
 307         /*
 308          * At each iteration, off is the offset into the bap array which is
 309          * an array of disk addresses at the current level of indirection.
 310          * The logical block number and the offset in that block are stored
 311          * into the argument array.
 312          */
 313         ap->in_lbn = metalbn;
 314         ap->in_off = off = NIADDR - i;
 315         ap->in_exists = 0;
 316         ap++;
 317         for (++numlevels; i <= NIADDR; i++) {
 318                 /* If searching for a meta-data block, quit when found. */
 319                 if (metalbn == realbn)
 320                         break;
 321
 322                 blockcnt /= MNINDIR(ump);
 323                 off = (bn / blockcnt) % MNINDIR(ump);
 324
 325                 ++numlevels;
 326                 ap->in_lbn = metalbn;
 327                 ap->in_off = off;
 328                 ap->in_exists = 0;
 329                 ++ap;
 330
 331                 metalbn -= -1 + off * blockcnt;
 332         }
 333         if (nump)
 334                 *nump = numlevels;
 335         return (0);
 336 }
 337 /*
 338  * blockmap converts a file offsetto its physical block
 339  * number on the disk... it optionally returns the physically
 340  * contiguous size.
 341  */
 342 int
 343 ufs_blockmap(ap)
 344         struct vnop_blockmap_args /* {
 345                 struct vnode *a_vp;
 346                 off_t a_foffset;
 347                 size_t a_size;
 348                 daddr64_t *a_bpn;
 349                 size_t *a_run;
 350                 void *a_poff;
 351                 int a_flags;
 352         } */ *ap;
 353 {
 354         vnode_t     vp   = ap->a_vp;
 355         daddr64_t * bnp  = ap->a_bpn;
 356         size_t    * runp = ap->a_run;
 357         int         size = ap->a_size;
 358         struct fs * fs;
 359         struct inode *ip;
 360         ufs_daddr_t lbn;
 361         ufs_daddr_t daddr = 0;
 362         int     devBlockSize = 0;
 363         int     retsize = 0;
 364         int     error = 0;
 365         int     nblks;
 366
 367         ip = VTOI(vp);
 368         fs = ip->i_fs;
 369
 370         lbn = (ufs_daddr_t)lblkno(fs, ap->a_foffset);
 371         devBlockSize = vfs_devblocksize(vnode_mount(vp));
 372
 373         if (blkoff(fs, ap->a_foffset))
 374                 panic("ufs_blockmap; allocation requested inside a block");
 375
 376         if (size % devBlockSize)
 377                 panic("ufs_blockmap: size is not multiple of device block size\n");
 378
 379         if ((error = ufs_bmaparray(vp, lbn, &daddr, NULL, NULL, &nblks)))
 380                 return (error);
 381
 382         if (bnp)
 383                 *bnp = (daddr64_t)daddr;
 384
 385         if (ap->a_poff)
 386                 *(int *)ap->a_poff = 0;
 387
 388         if (runp) {
 389                 if (lbn < 0) {
 390                         /*
 391                          * we're dealing with the indirect blocks
 392                          * which are always fs_bsize in size
 393                          */
 394                         retsize = (nblks + 1) * fs->fs_bsize;
 395                 } else if (daddr == -1 || nblks == 0) {
 396                         /*
 397                          * we're dealing with a 'hole'... UFS doesn't
 398                          * have a clean way to determine it's size
 399                          * or
 400                          * there's are no physically contiguous blocks
 401                          * so
 402                          * just return the size of the lbn we started with
 403                          */
 404                         retsize = blksize(fs, ip, lbn);
 405                 } else {
 406                         /*
 407                          * we have 1 or more blocks that are physically contiguous
 408                          * to our starting block number... the orignal block + (nblks - 1)
 409                          * blocks must be full sized since only the last block can be
 410                          * composed of fragments...
 411                          */
 412                          retsize = nblks * fs->fs_bsize;
 413
 414                          /*
 415                           * now compute the size of the last block and add it in
 416                           */
 417                          retsize += blksize(fs, ip, (lbn + nblks));
 418                 }
 419                 if (retsize < size)
 420                         *runp = retsize;
 421                 else
 422                         *runp = size;
 423         }
 424         return (0);
 425 }