1 /*- 2 * modified for Lites 1.1 3 * 4 * Aug 1995, Godmar Back (gback@cs.utah.edu) 5 * University of Utah, Department of Computer Science 6 */ 7 /*- 8 * SPDX-License-Identifier: BSD-3-Clause 9 * 10 * Copyright (c) 1982, 1986, 1989, 1993 11 * The Regents of the University of California. All rights reserved. 12 * 13 * Redistribution and use in source and binary forms, with or without 14 * modification, are permitted provided that the following conditions 15 * are met: 16 * 1. Redistributions of source code must retain the above copyright 17 * notice, this list of conditions and the following disclaimer. 18 * 2. Redistributions in binary form must reproduce the above copyright 19 * notice, this list of conditions and the following disclaimer in the 20 * documentation and/or other materials provided with the distribution. 21 * 3. Neither the name of the University nor the names of its contributors 22 * may be used to endorse or promote products derived from this software 23 * without specific prior written permission. 24 * 25 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 27 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 28 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 29 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 30 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 31 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 32 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 33 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 34 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 35 * SUCH DAMAGE. 36 * 37 * @(#)ffs_balloc.c 8.4 (Berkeley) 9/23/93 38 * $FreeBSD$ 39 */ 40 41 #include <sys/param.h> 42 #include <sys/systm.h> 43 #include <sys/bio.h> 44 #include <sys/buf.h> 45 #include <sys/limits.h> 46 #include <sys/lock.h> 47 #include <sys/mount.h> 48 #include <sys/vnode.h> 49 50 #include <fs/ext2fs/fs.h> 51 #include <fs/ext2fs/inode.h> 52 #include <fs/ext2fs/ext2fs.h> 53 #include <fs/ext2fs/ext2_dinode.h> 54 #include <fs/ext2fs/ext2_extern.h> 55 #include <fs/ext2fs/ext2_mount.h> 56 57 static int 58 ext2_ext_balloc(struct inode *ip, uint32_t lbn, int size, 59 struct ucred *cred, struct buf **bpp, int flags) 60 { 61 struct m_ext2fs *fs; 62 struct buf *bp = NULL; 63 struct vnode *vp = ITOV(ip); 64 daddr_t newblk; 65 int blks, error, allocated; 66 67 fs = ip->i_e2fs; 68 blks = howmany(size, fs->e2fs_bsize); 69 70 error = ext4_ext_get_blocks(ip, lbn, blks, cred, NULL, &allocated, &newblk); 71 if (error) 72 return (error); 73 74 if (allocated) { 75 bp = getblk(vp, lbn, fs->e2fs_bsize, 0, 0, 0); 76 if(!bp) 77 return (EIO); 78 } else { 79 error = bread(vp, lbn, fs->e2fs_bsize, NOCRED, &bp); 80 if (error) { 81 return (error); 82 } 83 } 84 85 86 bp->b_blkno = fsbtodb(fs, newblk); 87 if (flags & BA_CLRBUF) 88 vfs_bio_clrbuf(bp); 89 90 *bpp = bp; 91 92 return (error); 93 } 94 95 /* 96 * Balloc defines the structure of filesystem storage 97 * by allocating the physical blocks on a device given 98 * the inode and the logical block number in a file. 99 */ 100 int 101 ext2_balloc(struct inode *ip, e2fs_lbn_t lbn, int size, struct ucred *cred, 102 struct buf **bpp, int flags) 103 { 104 struct m_ext2fs *fs; 105 struct ext2mount *ump; 106 struct buf *bp, *nbp; 107 struct vnode *vp = ITOV(ip); 108 struct indir indirs[EXT2_NIADDR + 2]; 109 e4fs_daddr_t nb, newb; 110 e2fs_daddr_t *bap, pref; 111 int num, i, error; 112 113 *bpp = NULL; 114 if (lbn < 0) 115 return (EFBIG); 116 fs = ip->i_e2fs; 117 ump = ip->i_ump; 118 119 /* 120 * check if this is a sequential block allocation. 121 * If so, increment next_alloc fields to allow ext2_blkpref 122 * to make a good guess 123 */ 124 if (lbn == ip->i_next_alloc_block + 1) { 125 ip->i_next_alloc_block++; 126 ip->i_next_alloc_goal++; 127 } 128 129 if (ip->i_flag & IN_E4EXTENTS) 130 return (ext2_ext_balloc(ip, lbn, size, cred, bpp, flags)); 131 132 /* 133 * The first EXT2_NDADDR blocks are direct blocks 134 */ 135 if (lbn < EXT2_NDADDR) { 136 nb = ip->i_db[lbn]; 137 /* 138 * no new block is to be allocated, and no need to expand 139 * the file 140 */ 141 if (nb != 0) { 142 error = bread(vp, lbn, fs->e2fs_bsize, NOCRED, &bp); 143 if (error) { 144 return (error); 145 } 146 bp->b_blkno = fsbtodb(fs, nb); 147 if (ip->i_size >= (lbn + 1) * fs->e2fs_bsize) { 148 *bpp = bp; 149 return (0); 150 } 151 } else { 152 EXT2_LOCK(ump); 153 error = ext2_alloc(ip, lbn, 154 ext2_blkpref(ip, lbn, (int)lbn, &ip->i_db[0], 0), 155 fs->e2fs_bsize, cred, &newb); 156 if (error) 157 return (error); 158 /* 159 * If the newly allocated block exceeds 32-bit limit, 160 * we can not use it in file block maps. 161 */ 162 if (newb > UINT_MAX) 163 return (EFBIG); 164 bp = getblk(vp, lbn, fs->e2fs_bsize, 0, 0, 0); 165 bp->b_blkno = fsbtodb(fs, newb); 166 if (flags & BA_CLRBUF) 167 vfs_bio_clrbuf(bp); 168 } 169 ip->i_db[lbn] = dbtofsb(fs, bp->b_blkno); 170 ip->i_flag |= IN_CHANGE | IN_UPDATE; 171 *bpp = bp; 172 return (0); 173 } 174 /* 175 * Determine the number of levels of indirection. 176 */ 177 pref = 0; 178 if ((error = ext2_getlbns(vp, lbn, indirs, &num)) != 0) 179 return (error); 180 #ifdef INVARIANTS 181 if (num < 1) 182 panic("ext2_balloc: ext2_getlbns returned indirect block"); 183 #endif 184 /* 185 * Fetch the first indirect block allocating if necessary. 186 */ 187 --num; 188 nb = ip->i_ib[indirs[0].in_off]; 189 if (nb == 0) { 190 EXT2_LOCK(ump); 191 pref = ext2_blkpref(ip, lbn, indirs[0].in_off + 192 EXT2_NDIR_BLOCKS, &ip->i_db[0], 0); 193 if ((error = ext2_alloc(ip, lbn, pref, fs->e2fs_bsize, cred, 194 &newb))) 195 return (error); 196 if (newb > UINT_MAX) 197 return (EFBIG); 198 nb = newb; 199 bp = getblk(vp, indirs[1].in_lbn, fs->e2fs_bsize, 0, 0, 0); 200 bp->b_blkno = fsbtodb(fs, newb); 201 vfs_bio_clrbuf(bp); 202 /* 203 * Write synchronously so that indirect blocks 204 * never point at garbage. 205 */ 206 if ((error = bwrite(bp)) != 0) { 207 ext2_blkfree(ip, nb, fs->e2fs_bsize); 208 return (error); 209 } 210 ip->i_ib[indirs[0].in_off] = newb; 211 ip->i_flag |= IN_CHANGE | IN_UPDATE; 212 } 213 /* 214 * Fetch through the indirect blocks, allocating as necessary. 215 */ 216 for (i = 1;;) { 217 error = bread(vp, 218 indirs[i].in_lbn, (int)fs->e2fs_bsize, NOCRED, &bp); 219 if (error) { 220 return (error); 221 } 222 bap = (e2fs_daddr_t *)bp->b_data; 223 nb = bap[indirs[i].in_off]; 224 if (i == num) 225 break; 226 i += 1; 227 if (nb != 0) { 228 bqrelse(bp); 229 continue; 230 } 231 EXT2_LOCK(ump); 232 if (pref == 0) 233 pref = ext2_blkpref(ip, lbn, indirs[i].in_off, bap, 234 bp->b_lblkno); 235 error = ext2_alloc(ip, lbn, pref, (int)fs->e2fs_bsize, cred, &newb); 236 if (error) { 237 brelse(bp); 238 return (error); 239 } 240 if (newb > UINT_MAX) 241 return (EFBIG); 242 nb = newb; 243 nbp = getblk(vp, indirs[i].in_lbn, fs->e2fs_bsize, 0, 0, 0); 244 nbp->b_blkno = fsbtodb(fs, nb); 245 vfs_bio_clrbuf(nbp); 246 /* 247 * Write synchronously so that indirect blocks 248 * never point at garbage. 249 */ 250 if ((error = bwrite(nbp)) != 0) { 251 ext2_blkfree(ip, nb, fs->e2fs_bsize); 252 brelse(bp); 253 return (error); 254 } 255 bap[indirs[i - 1].in_off] = nb; 256 /* 257 * If required, write synchronously, otherwise use 258 * delayed write. 259 */ 260 if (flags & IO_SYNC) { 261 bwrite(bp); 262 } else { 263 if (bp->b_bufsize == fs->e2fs_bsize) 264 bp->b_flags |= B_CLUSTEROK; 265 bdwrite(bp); 266 } 267 } 268 /* 269 * Get the data block, allocating if necessary. 270 */ 271 if (nb == 0) { 272 EXT2_LOCK(ump); 273 pref = ext2_blkpref(ip, lbn, indirs[i].in_off, &bap[0], 274 bp->b_lblkno); 275 if ((error = ext2_alloc(ip, 276 lbn, pref, (int)fs->e2fs_bsize, cred, &newb)) != 0) { 277 brelse(bp); 278 return (error); 279 } 280 if (newb > UINT_MAX) 281 return (EFBIG); 282 nb = newb; 283 nbp = getblk(vp, lbn, fs->e2fs_bsize, 0, 0, 0); 284 nbp->b_blkno = fsbtodb(fs, nb); 285 if (flags & BA_CLRBUF) 286 vfs_bio_clrbuf(nbp); 287 bap[indirs[i].in_off] = nb; 288 /* 289 * If required, write synchronously, otherwise use 290 * delayed write. 291 */ 292 if (flags & IO_SYNC) { 293 bwrite(bp); 294 } else { 295 if (bp->b_bufsize == fs->e2fs_bsize) 296 bp->b_flags |= B_CLUSTEROK; 297 bdwrite(bp); 298 } 299 *bpp = nbp; 300 return (0); 301 } 302 brelse(bp); 303 if (flags & BA_CLRBUF) { 304 int seqcount = (flags & BA_SEQMASK) >> BA_SEQSHIFT; 305 306 if (seqcount && (vp->v_mount->mnt_flag & MNT_NOCLUSTERR) == 0) { 307 error = cluster_read(vp, ip->i_size, lbn, 308 (int)fs->e2fs_bsize, NOCRED, 309 MAXBSIZE, seqcount, 0, &nbp); 310 } else { 311 error = bread(vp, lbn, (int)fs->e2fs_bsize, NOCRED, &nbp); 312 } 313 if (error) { 314 brelse(nbp); 315 return (error); 316 } 317 } else { 318 nbp = getblk(vp, lbn, fs->e2fs_bsize, 0, 0, 0); 319 nbp->b_blkno = fsbtodb(fs, nb); 320 } 321 *bpp = nbp; 322 return (0); 323 } 324